Contents

Foreword xii

Chapter authors xiv

Editorial panel XV

Protection symbols used in circuit diagrams xvi

12 Protection of generators, transformers, generator-transformer units and

transformer feeders J.Rushton and K. G.M.Mewes I
12.1 Introduction 1
12.2 Performance requirements 2
12.2.1 Generator faults 2
12.2.2 Transformer faults 4
12.3 Generator protection systems 7
12.3.1 Unbiased differential protection 7
12.3.2 Biased differential protection 9
12.3.3 Back-up overcurrent and earth-fault protection 12
12.3.4 Negative phase-sequence protection 17
12.3.5 Interturn fault protection 21
12.3.6 Loss of excitation (field failure) protection 22
12.3.'/ Protection against pole-slipping 24
12.3.8 Rotor earth-fault protection 2"/
12.3.9 Sensitive power protection 29
12.3.10 Low forward power interlock 29
12.3.11 Overspeed protection 30
12.3.12 Underexcitation limiting 31
12.3.13 Mechanical and hydraulic trips 31
12.4 Gas-turbine driven generators 34
12.4.1 Direct connected, gas-turbine sets 34
12.4.2 Transformer connected, gas-turbine sets 36
12.5 Transformer protection 39
12.5.1 Unbiased differential protection 39
12.5.2 Biased differential protection 42
12.5.3 Restricted earth-fault protection 45
12.5.4 Overcurrent protection 47
Contents

12.5.5 Directional overcurrent protection 48

12.5.6 Interlocked overcurrent protection 50
12.5.7 Standby earth-fault protection 51
12.5.8 Tank earth-fault protection 52
12.5.9 Winding temperature protection 52
12.5.10 Gas generation and oil-surge protection 53
12.6 Protection schemes for typical transformers 54
12.6.1 Distribution transformers 54
12.6.2 Two-winding transmission transformers 55
12.6.3 Station transformers 57
12.6.4 Autotransformers for transmission 57
12.7 Protection system for generator transformer units 60
12.7.1 Biased differential protection 63
12.7.2 Stator earth-fault protection 64
12.7.3 Tripping arrangements 65
12.7.4 Generator transformer overfluxing protection 65
12.8 Transformer feeder protection 66
12.8.1 Overall protection for feeder and transformer 67
12.8.2 Separate protection for feeder and transformer 69
12.8.3 Intertripping 74
12.8.4 Neutral displacement protection 77
12.8.5 Directional overcurrent protection 79
12.8.6 Typical protection arrangements for transformer
feeders 79
12.9 Bibliography 80

13 Busbar protection L. C. W.Frerk 81

13.1 History of the development of busbar protection 81
13.2 General considerations 81
13.2.1 The basic philosophy of busbar protection 81
13.2.2 Earth-fault protection versus phase and earth-fault
protection 83
13.3 The clearance of busbar faults by non-unit circuit protection 83
13.3.1 Back-up overcurrent and earth-fault relays 83
13.3.2 Distance protection 84
13.4 Unit systems of busbar protection for metalclad distribution
switchgear 86
13.4.1 General considerations 86
13.4.2 Frame earth systems 87
13.5 Unit systems of busbar protection for transmission substations 93
13.5.1 General considerations 93
13.5.2 Current balance using circulating current principle 94
13.5.3 Connections for circulating current busbar
protection 95
13.5.4 The influence of c.t. performance on through-fault
stability 103
13.5.5 Basic principles of high-impedance circulating current
busbar protection: stability 103
13.5.6 Basic principles of high-impedance circulating current
busbar protection: operation 108
13.5.7 Extension of the basic principles to busbar protection 111
13.5.8 Typos of high-impedance relays 112
13.5.9 Practical high-impedance installations 113
 Contents

13.6 Practical considerations 118

13.6.1 Factors affecting the position of c.t.s in busbars 118
13.6.2 Effect of c.t. location in outgoing circuits 120
13.6.3 Multiple check zones 123
13.6.4 Busbar selector auxiliary switches 127
13.6.5 C.T. test links 130
13.6.6 Precautions against maloperation of busbar
protection 131
13.6.7 Tripping and alarm circuit arrangements 132
13.6.8 Back-tripping 134
13.6.9 Test facilities 139
13.6.10 Fault settings 140
13.6.11 Stability limits 142
13.7 Circuit breaker fail protection 142
13.7.1 Principle of operation 142
13.7.2 Precautions against maloperation 144
13.7.3 Current check relay settings 145
13.7.4 Circuit breaker fail timer settings 145
13.8 Terminology 147
13.9 Bibliography 148

14 Protection of motors, reactors, boosters and capacitors P.M.Dolby 149

14.1 Introduction 149
14.2 Motors 149
14.2.1 Characteristics of d.c. and a.c. motors 149
14.2.2 Application of d.c. and a.c. motors 162
14.2.3 Motor control 163
14.2.4 Types of fault 164
14.2.5 A.C. and d.c. motor protection 164
14.3 Reactors 181
14.3.1 The place of reactors in a power system 181
14.3.2 Types of reactor 182
14.3.3 Reactor rating 185
14.3.4 Reactor application 186
14.3.5 Reactor protection 190
14.4 Boosters 199
14.4.1 The place of boosters in a power system 199
14.4.2 Transformer tap-changing 200
14.4.3 Booster transformers 200
14.4.4 The moving-coil regulator 205
14.4.5 The induction regulator 210
14.4.6 Protection of boosters 213
14.5 Capacitors 215
14.5.1 Capacitors in an interconnected power system 215
14.5.2 Series-connected capacitors 216
14.5.3 Shunt-connected capacitors 217
14.5.4 Series or shunt connection 219
14.5.5 The capacitor unit 221
14.5.6 Protection of capacitors 222
14.5.6.1 Series capacitor internal protection 223
14.5.6.2 Series capacitor external protection 224
14.5.6.3 Shunt capacitor internal protection 230
14.5.6.4 Shunt capacitor external protection 235
Contents

14.5.6.5 Protection of synchronous shunt

compensators 237
14.6 Bibliography 245

15 The application of protection to rural distribution systems J.Harris 247

15.1 Introduction 247
15.2 Fuses 250
15.2.1 Types employed 250
15.2.2 Application 250
15.3 Automatic circuit reclosing 255
15.3.1 Principle 255
15.3.2 Repeater fuses 257
15.3.3 Pole-mounted automatic circuit reclosers 257
15.3.4 Substation circuit breakers 269
15.4 Sensitive earth-fault protection 275
15.5 Arc-suppression coils 278
15.6 Performance/cost comparison of protective equipment for
rural systems 280
15.7 Primary networks in rural areas 282
15.8 Bibliography 289

16 The application of protection to urban and metropolitan systems

K.A.J. Coates 291
16.1 Introduction 291
16.2 Characteristics of urban and metropolitan areas 292
16.3 Distribution system protection - radial l.v. systems 294
16.3.1 Services 294
16.3.2 L.V. cables 294
16.3.3 Substation transformers 295
16.3.4 H.V. cables 297
16.3.5 Primary substations 302
16.4 Distribution system protection - interconnected 1.v. systems 303
16.4.1 The l.v. network 304
16.4.2 Substations 304
16.4.3 The h.v. network 305
16.4.4 Protection on a distributed block l.v.
interconnected system 306
16.4.5 Supply to large point loads 313
16.4.6 Supply to h.v. consumers 316
16.5 Private generation 316
16.6 Future trends 319
16.7 Bibliography 319

17 The application of protection to transmission systems

J. C. Whittaker 321
17.1 General principles of application of protection to transmission
systems 321
17.1.1 Introduction 321
17.1.2 System design considerations 321
17.1.3 Factors which influence the choice of protection 322
17.1.3.1 Plant to be protected 322
17.1.3.2 Probability of various types of fault 323
17.1.3.3 Load and fault currents 323
 Contents

17.1.3.4 Voltage and current ratings of

protected plant 323
17.1.3.5 Necessity or otherwise for high-speed
operation 324
17.1.3.6 Importance of security of supply 325
17.1.3.7 Compatibility with existing protection 327
17.1.3.8 Availability of signalling channels 327
17.1.3.9 Cost 328
17.2 Main and back-up protection and location of current
transformers 328
17.2.1 Main and back-up protection 328
17.2.1.1 Main protection 329
17.2.1.2 Back-up protection 330
17.2.2 Effect of location of current transformers in
determining protection to be provided 334
17.2.3 Two-stage overcurrent protection 337
17.3 Intertripping and protection signalling 338
17.3.1 General 338
17.3.2 D.C. signalling 340
17.3.3 Post Office signalling 342
17.3.4 Carrier signalling 343
17.3.5 Fault throwing 343
17.4 Automatic switching 344
17.4.1 Design and application of automatic switching
equipment 344
17.4.2 High-speed automatic reclosing 345
17.4.3 Delayed automatic reclosing 345
17.4.4 Equipment design and programming 348
17.4.5 Commissioning 356
17.5 Economic considerations 357
17.6 Typical protection applications in a major transmission
system 358
17.6.1 Feeder protection 358
17.6.1.1 Protection for a long overload feeder 358
17.6.1.2 Protection for a short overhead feeder 362
17.6.1.3 Protection for an underground feeder 363
17.6.2 Protection for a transformer 365
17.6.3 Protection for banked transformers 373
17.6.4 Protection for a teed feeder 376
17.6.5 Protection for a transformer-feeder 383
17.6.6 Protection for a double-busbar station 385
17.6.7 Protection for mesh stations 387
17.6.8 Protection for complex primary circuit configurations 387
17.7 Typical protection applications in a minor transmission
system or major distribution system 392
17.7.1 Protection for a feeder 392
17.7.2 Protection for a transformer 393
17.7.3 Protection for banked transformers and dual
secondary transformers 393
17.7.4 Protection for a teed feeder 393
17.7.5 Protection for a transformer feeder 393
17.7.6 Protection for a double busbar and mesh station 397
Contents

17.7.7 Protection for complex primary circuit configurations 397

17.8 Bibliography 398

18 Testing, commissioning and management of protection E.C.Smith and

D.Hay 399
18.1 Introduction 399
18.2 Contractual obligations 399
18.3 The mental approach to commissioning tests 400
18.4 Commissioning tests 400
18.4.1 Reasons for commissioning tests 400
18.4.2 Planning of commissioning tests 401
18.4.3 Inspection prior to testing 405
18.4.4 The tests 407
18.4.5 Phasing tests 433
18.4.6 Closing up 435
18.4.7 On-load tests 435
18.4.8 Modification to existing substations 437
18.5 Routine maintenance tests 437
18.5.1 Causes and effects of deterioration 438
18.5.2 Frequency of routine maintenance 438
18.5.3 Inspection and tests 439
18.5.4 Maintenance of busbar protection, back-tripping
and circuit-breaker fail protection at double-busbar
type substations 441
18.5.5 Maintenance and testing of intertripping and
protection signalling equipment 443
18.6 Fault investigation 446
18.6.1 Primary faults 446
18.6.2 Faults on the protective equipment 446
18.6.3 Faults on solid-state equipment 447
18.7 The avoidance of errors when testing 448
18.7.1 General 448
18.8 The test equipment 450
18.9 Records 463
8.9.1 Relay settings 463
8.9.2 Test results 464
Index 467
 Chapter 12'
Protection of generators, transformers,
generator-transformer units and
transformer feeders
by J.Rushton, revised by K.G.M.Mewes
, , II

12.1 Introduction

In the early days of electricity, supply generators were operated by the Supply
Company to supply a local load. Economically it was desirable to generate at the
distribution system voltage. As the size of local networks grew, higher distribution
voltages became necessary, and by the time the 132 kV National Grid was
established many of the larger supply companies were generating and distributing at
a voltage of 33 kV.
With the development of the 132/275/400kV systems, the overall pattern of
generation changed resulting in larger stations sited to take advantage of fuel and
cooling water supplies, and operated over an interconnected transmission network
to give lowest cost and maximum efficiency. Generators were connected to the
main transmission busbar via an associated transformer, and this arrangement
permitted the additional facility of voltage and power factor control by transformer
tappings.
Optimisation of machine design has resulted in standard machine ratings, of
which the following is a typical selection:

Generator output Voltage of Generated voltage

MW h.v. connection
• , , , •

60 132 kV 11.8 kV
120 132 kV 13.8 kV
500 275/400 kV 22 kV
660 400 kV 22 kV
i, i 1 ,i

Increase in outputs has been achieved without proportional increase in frame size
by employing more efficient cooling methods. The protective systems applied to
small directly connected generators are equally applicable to large transformer
connected units, but the smaller operating margins available with these units (due
2 Protection of generators and transformers

to their smaller frame size) require more comprehensive schemes of high perform-
ance protection.

12.2 Performance requirements

The high costs associated with large generating and transforming plants accentuate
the need for reliable, high speed schemes of protection to:

(a) minimise fault damage and so reduce the possible need to replace the plant
(capital outlay)
(b) reduce repair outage time and so minimise the need to run lower merit (less
cost-efficient) plant in order to meet the demand (revenue expenditure)
(c) assist in maintaining system stability.

The degree of protection to be provided for the plant is determined by protection

engineers in consultation with plant designers and system operation engineers,
the objective being to provide a minimum of protection consistent with adequate
coverage of all conditions liable to cause damage or affect the continuity of supply.
Before considering in detail the many forms of protection fitted to generators and
transformers, it is desirable to consider the origin and effects of faults and other
system disturbances so that the significance of the protection arrangements may be
appreciated.

12.2.1 Generator faults

(a) Stator faults: Stator faults involve the main current carrying conductors and
must therefore be cleared quickly from the power system by a complete shutdown
of the generator. They may be faults to earth, between phases or between turns of a
phase, singly or in combination. The great danger from all faults is the possibility of
damage to the laminations of the stator core and stator windings due to the heat
generated at the point of fault. If the damage so caused is other than superficial, the
stator would have to be dismantled, the damaged laminations and windings replaced
and the stator rebuilt, all of which is a lengthy and costly process.
Limitation of generator stator earth-fault current by means of resistance earthing
is normal practice (see Chapter 1) and serves, among other things, to minimise core
burning.
Phase-to-phase faults and interturn faults are both less common than earth faults.
It is relatively easy to provide protection for phase-to-phase faults, but interturn
faults are, on the other hand more difficult to detect and protection is not usually
provided. Generally speaking, interturn faults quickly involve contact with earth via
the stator core and are then tripped by stator earth-fault protection.

(b) Rotor faults: Rotor faults may be either to earth or between turns and may
be caused by the severe mechanical and thermal stresses acting upon the winding
insulation; these are aggravated by a variable load cycle.
 Protection of generators and transformers 3

The field system is not normally connected to earth so that a single earth fault
does not give rise to any fault current. However, a second fault to earth would short
circuit part of the field winding and thereby produce an asymmetrical field system,
and unbalanced forces on the rotor. Such forces will cause excess pressure on
bearings and shaft distortion, if not quickly removed.
Under the general heading of rotor faults can be included loss of excitation. This
may be caused by an open circuit in the main field winding or a failure elsewhere in
the excitation system.
Loss of excitation in a generator connected to a large interconnected power
system results in a loss of synchronism and slightly increased generator speed, since
the power input to the machine is unchanged. The machine behaves as an induction
generator drawing its exciting current from the remainder of the system in the form
of wattless current whose magnitude approximates to that of the full load rating of
the machine. This may cause overheating of the stator winding and increased rotor
losses due to the currents induced in the rotor body and damper winding. This
condition should not be allowed to persist indefinitely and corrective action either
to restore the field, or to off-load and shut down the machine should be taken.
With generator outputs above half rated load, pole-slipping caused by weak field
condition, would cause severe voltage variations which may, in turn, cause opera-
tion of the undervoltage protection on the boiler auxiliaries. The resultant
operation of 'loss of boiler f'tring' protection would then shut down the generator
unit. Other generators connected to the same busbar may also be caused to 'swing'
and system instability would result. Pole slipping may also result from insufficiently
fast clearance of a system fault and require the tripping of the unit.

(c) Mechanical conditions: The mechanical conditions requiring consideration are

overspeed due to sudden loss of load, loss of drive due to prime mover failure and
loss of condenser vacuum.
The problem of overspeed limitation, particularly in relation to sudden load
changes, is considered later. With modem large units it is essential to anticipate
overspeed and take corrective action. Mechanical overspeed devices which operate
on the steam stop valves are invariably fitted.
In the event of failure of the prime mover, a generator will continue to run
synchronously drawing power from the system. This can sometimes lead to a
dangerous mechanical condition if allowed to persist, although the condition is
immediately obvious to the attendant.
Sets having an internal combustion prime mover must be protected against
engine failure, where, if the alternator continues to motor serious engine damage
may result.
Vacuum failure (or low vacuum) detection is necessary to prevent a rise of
condenser pressure which might lead to shattering of 1.p. casing and condensers.

(d) External faults: Turboalternators must be protected against the effects of

sustained external faults, for example faults on lines or busbars which are not
4 Protection of generators and transformers

cleared by the appropriate protection. The main condition of interest is that of

an unsymmetrical fault producing negative phase sequence currents in the stator
winding. The effect of these currents is to produce a field rotating in opposite sense
to the d.c. field system producing a flux which cuts the rotor at twice the rotational
frequency thereby inducing double frequency currents in the field system and the
rotor body. These currents produce severe rotor heating and modern machines have
a limited negative phase sequence current capability.
Automatic tripping is therefore required for the higher negative phase sequence
current conditions.
This capability limit applies to all modem hydrogen-cooled machines and many
air-cooled machines, but some of the older air-cooled machines are designed to
withstand full negative sequence currents continuously. In large modern alternators,
particularly those employing direct cooling of the stator and rotor conductors, the
temperature rise caused by abnormally high stator currents is more rapid than in
the less highly rated machines and the capability limit is therefore lower.

12.2.2. Transformer faults

(a) Faults within the transformer tank: These may comprise phase-to-earth,
phase-to-phase, or intertum faults on the windings, interwinding faults, tap changer
faults, insulator bushing failure and core overheating due to failure of core insula-
tion. The possibility of damage is high for these faults as is the risk of fire, and
short fault clearance times are advantageous.
The connections of the power transformer and the method of earthing play an
important part in determining current magnitude available for relay operation, and
each case requires separate consideration. Figs. 12.2.2A and B give the current
distribution under fault conditions for various transformer arrangements based on
the performance of typical transformers.
For a resistance earthed, star-connected winding, a winding-to-earth fault will
give rise to a current dependent on the value of the earthing resistor and the
distance of the fault from the neutral end of the winding. The effective ratio of
transformation between the primary winding and the short circuited portion of the
secondary winding varies with the fault position. The current flowing through the
transformer terminals is therefore, for all practical purposes, proportional to the
square of the percentage of the winding short circuited. This is illustrated in
Fig. 12.2.2A(a~.
For a solidly earthed star winding, the fault current bears no simple relationship
to the distance of the fault from the neutral end since the effective reactance of the
fault path changes with fault position. Fig. 12.2.2A(b)shows that the minimum
value of fault current occurs for a fault 30 to 40% from the neutral end.
For a delta connected winding the minimum voltage on the delta winding is at
the centre of one phase and is 50% normal phase-to-earth voltage,and an illustration
of the approximate method of calculation is given in Fig. 12.2.2B. The range of
values of fault current varies less than with the star connected winding. The value of
fault current depends upon system earthing arrangements, and the curves of
6 Protection of generators and transformers

. . . . . . . ,~ a~¢~'~f"~__
||,"
2~ ., .-, t ' ~ ¢ " ~ p

., ~Y"~¢"¢"~ p

t
" v

m 11
o

T
I
I
Ii." V I
2,~3Zl.: I
I z 1-
I
I
o
L .... _1

(a) A p p r o x i m a t e method for calculating fault current

ZE - O
L.-

•. 3
M

P.u. I

0 50 I00

Distance ~¢ o f f a u l t f r o m e n d o f w i n d i n g ( % )

(h) Typical fault current values

Fig, 12.2.2B Transformer fault current for delta winding

Fig. 12.2.2B(b) show that the minimum value of fault current occurs for a fault at
the centre of one phase winding. The impedance of the winding under such
conditions will be between two and four times the normal value.
Phase-to-phase faults rarely occur on a power transformer. Clearly such faults
will give rise to large currents.
Interturn faults are more likely to occur than phase-to-phase faults. The interturn
insulation on a power transformer is not so great as the interwinding insulation, and
the possibility exists of breakdown between turns.
 Protection of generators and transformers 7

A short circuit of a few turns of the winding will produce a heavy current in the
faulted loop and a very small terminal current. In this respect it has some similarity
to a neutral end fault on a solidly earthed star winding.
Core faults can occur due to lamination insulation becoming short circuited.
This can cause serious overheating due to eddy current losses. Core clamping bolts
must always be insulated to prevent this trouble. If core insulation becomes
defective (due, possibly, to the failure of core bolt insulation or debris in the tank),
it must be detected quickly.

(b) Faults on transformer connections: These may comprise any type of normal
system fault on open copperwork connections or flashover of co-ordinating gaps.
Damage due to such faults is not usually great though they may constitute a serious
hazard to power system stability if not cleared quickly. Faults between the current
transformers and the associated circuit breaker have to be included in this category.

(c) Overheating: Failure of the cooling system will cause overheating and conse-
quent danger of damage to the windings.

(d) Faults external to the transformer zone: These will be of the usual range of
system earth and phase faults to be cleared by appropriate external protection
systems. They will affect, therefore, only the requirement of transformer back-up
protection.

12.3 Generator protection systems

A generator is invariably protected against phase-to-phase faults by overall

differential protection using either a biased or unbiased relay. The unbiased relay
is usually of the attracted armature pattern and the biased relay uses either an
induction-disc or rectifier moving coil element. In all cases a high speed relay is
used.

12.3.1 Unbiased differential protection

The basic circuit for generator differential protection shown in Fig. 12.3.1A is
the simplest arrangement of a balanced current system in which only two current
transformers per phase are required to balance together. The theory of balanced
systems is discussed in Chapter 4 which gives the design parameters for the correct
operation of an unbiased relay system under power system transient conditions.
For most applications to the protection of a direct-connected generator, an
instantaneous unbiased relay is used and this gives an adequate and predictable
performance.
The fault setting required from the differential protection is determined by the
value of the neutral earthing resistor and also by the amount of winding to be
protected. For the winding fault shown in Fig. 12.3.1A the earth-fault current is
 8 Protection of generators and transformers

R Y B

I r Fault

I '

, - k,.° !
I' Iv i '-- --1

I Generator
' I w,od,.
Bushars
I " " ' - T ~ ' - "
| !-
, ~ ~ , I I''
Stabilising
L___j resistors

Relay

Fig. 12.3.1A Overall differential (circulating current) protection of direct-connected generator

1.0
C
0
a_
0
Q,

/:b'//:
0.9 Protection setting
as per cent o f
generator rated current:
e~
= IS%
~ 0.8
(a) = s%

//'// Y
o,u

C
o.,,
(b)= ~0%
(c) = 2 0 %
* 0.7
I1

<~ o.6 A
0 0.5 1.0 1.5 2.0

Setting current l R o f neutral

earthing resistor as a multiple
o f generator rated current
Fig. 12.3.1B Relation between neutral resistor setting, protection sensitivity and amount of
generator winding protected

IF - (1 - x)E/R which must equal the primary fault setting of the differential
protection, Is, for minimum operating current.
Thus,
E E
Is=(1-x) andlR R
R =--
 Protection of generators and transformers 9

whence x = (1 - IS/IR), IR bein~ the earthing resistor current setting.

The relationship is shown graphically in Fig. 12.3.1B.
With a resistor designed to pass rated current of the machine on a fault at the
machine terminals, it is clear that a differential protection setting of 10% will protect
90% of the machine winding, the unprotected portion being the 10% at the neutral
end where fault risk is reduced.
The differential protection will give greater than 90% cover for phase-to-phase
faults since the fault current is limited only by the machine impedance which is
small when the faulted portion of the winding is small.

12.3.2 Biased differential protection

Occasionally, there is some advantage in using a biased differential relay. The effect
of the bias feature is to enable the impedance of the relay operating coil circuit to
be reduced for a given value of through fault stability. The voltage drop across the
current transformers at setting is correspondingly reduced and the magnetising
currents have a negligible effect on the primary fault setting. The bias feature is
obtained by circulating the through fault current through an additional winding
which exerts a restraining force on the relay. The basic circuit connections are
shown in Fig. 12.3.2A. Nominally no current flows in the operating coil under
through fault conditions but, due to imperfect matching of the current transformers,
some spill current may be present. This spill current will flow in the relay operating
circuit but will not cause operation unless the relay operating/bias setting ratio is
exceeded. The magnitude of relay operating coil current to cause operation thus
increases as the circulating current increases in a f'Lxed relationship determined by
the fundamental constants of tile relay circuit. This is illustrated in Fig. 12.3.2B.
Referring to Fig. 12.3.2A

relay operating force = K ( I I - I2 )No

(h +h)
relay restraining force = K ' Nr + S
2

where No and Nr are the operating and restraint coil turns, respectively, and S is the
spring restraint force.
The electrical operating and restraint forces are equal at the balance point of the
relay when the spring restraint is zero, whence
I1 - I2 Nr
n

~(I~ +I2) No

This equation shows that the characteristic has a slope determined by the ratio
Nr/No, and the relay bias is defined as the difference current (11
- I 2 ) divided by
10 Protection of generators and transformers

the mean circulating current, which is fundamentally a constant ratio for all current
magnitudes. This is plotted in Fig. 12.3.2C.
The percentage bias feature reduces c.t. requirements under transient through
fault conditions. With the correct combination of bias and operating circuit
resistance, through fault stability can be obtained for any value of through fault
current. The required value of stabilising resistance is invariably quite low, thus
giving a lower relay operating voltage than that required for an unbiased relay. The
M M
iiii

~
Nr !
I2 )

N r = T o t a l restraint coil turns

N O = O p e r a t i n g coil turns

Fig. 1 2 . 3 . 2 A Basic circuit connection for biased differential protection

4 - I
I
I T r i p area

! 3-

,-.:,.
= -
e~

•.: P
i.
s_ I~ _,~" /
¢,J
I ~ % ~ Stable area I
O
¢j
¢d0
¢1

t_
QJ

. . . . R e l a y setting

2 4 6 8 10 12 14 16 18 20

Mean bias coil current ( p . u . ) . I 1 + 12

2
Fig. 12.3.2B Biased relay characteristic illustrating performance under internal and external
faults
 Protection of generators and transformers 11

450
eel

400- ~ + ,
¢~1 cq
Trip area
350-

300-
._= ~,
~ E E
250- ~ ° ~ ~

Z/,l'/,'/,
200-

150-
I/J//,//,//i
I00-
//1//,// I/
I/1//I///
.E 40%
50-

0 100 200 300

7/// / / , I / / / / /
400 500 600 700 800 900 1000
~ 30% 0 200 400 600 800 1000 1200
20% 0 200 400 600 800 1000 1200 1400 1600 1800 2000
~ 10% 0 400 800 1200 1600 2000
Mean b i a s c u r r e n t i n percent o f r a t e d c u r r e n t .
I 1 + 12 f o r 2 w i n d i n g t r a n s f o r m e r s , I 1 + I 2 + I 3 f o r w i n d i n g t r a n s f o r m e r s
2 3

I1 R
C

12 R O
t

Three winding transformer

13 R R A u x . c.t. (Restraining coils)

O A u x . c.t. (Operating coil)

Fig. 12.3.2C Percentage differential relay restraint characteristic

biased relay is thus most suitable for cases where c.t. performance is limited or
where leads are long.
Care must be taken that the c.t.s do not saturate too severely under high current
internal fault conditions since any magnetising current will flow through the relay
bias coils and may inhibit operation.
High speed biased differential protection may be particularly suitable when it is
necessary to use existing current transformers and their design is not suited to an
unbiased system.
12 Protection of generators and transformers

12.3.3 Back-up overcurrent and earth-fault protection

Inverse, definite minimum time relays are generally fitted to provide back-up over-
current and earth-fault protection of a generator, and its h.v. connections, primarily
to provide 'last resort' tripping in the event of failure of the main protection.
Ideally, it should be set t o trip in the shortest possible time so as to minimise the
risk of loss of system stability.
The minimum permissible relay setting is determined by the requirement that
tripping must not occur for external h.v. system faults which may be more
discriminatively cleared by other forms of protection. The relay setting should be

Fig. 12.3.3A Direct connection of overcurrent protection

Fig. 12.3.3B Interposed connection of overcurrent protection

 Protection of generators and transformers 13

Generator Gen. transformer

500 MW 600 MVA
22kV ( ~ H.V.
•It-"
- ~i - -T , ~ ! system
400 kV
..ff
F
_ I °.,,

SEQ/
30 MV transformer
U
['!

Fig. 12.3.3C Diagram of transformer connected generator showing overcurrent and n.p.s.
relays

chosen to provide adequate grading margins with negative phase sequence back-up
protection.
Overcurrent protection normally uses c.t.s in the h.v. circuit breaker which, in
the case of a transformer connected generator, is on the star-connected side of the
generator transformer. Two types of relay connection may be used, either:

(a) direct connection to star-connected current transformers (Fig. 12.3.3A)

where the relay is supplied with the phase currents, or
(b) connection to star-delta interposing current transformers (ratio 1/0.578)
(Fig. 12.3.3B) where the relay is supplied with phase-difference currents so
avoiding any zero-sequence currents present in the primary fault current.

Generally, the star-connected c.t. arrangement is used, but it may be seen from
Fig. 12.3.3D that the star-delta interposed connection appears to give a more
acceptable range of operating times for the more severe types of external h.v.
system fault. It also appears to provide better back-up to the negative phase
sequence protection described in Section 12.3.4.
The following analysis refers particularly to the generator-transformer connected
unit illustrated in Fig. 12.3.3C
External-system faults: The overcurrent relay must not operate for external-system
faults for which the generator will remain in synchronism, and settings chosen on
the basis of 2.0 s minimum operating time will embody high safety margins. The
relay-fault-current curves for both cases are shown in Fig. 12.3.3D. The highest
fault current for star-connected current transformers is obtained for the single-
phase-to-earth fault, while, for delta-connected current transformers, the phase-
to-phase fault gives the worst case. Fault settings for the two types of relay
connection are determined in Table 12.3.3A for a minimum operating time of 2 s
under external-system.fault conditions.
14 Protection of generators and transformers
T F XS

0.04

3 .~-'~" ~"" ~- . _ _ _ .

I ,, li --|
1 2 3

Time, S

a and g P h a s e - t o - e a r t h fault
b and e P h a s e - t o - p h a s e - t o - e a r t h fault
c and d P h a s e - t o - p h a s e fault
f 3-phase fault
delta-connected auxiliary current transformers
star-connected current transformers

Fig. 12.3.3D Variation o f relay current for star- and delta-connected current transformers
under ex ternal-faulr conditions
16 Protection of generators and transformers

It is seen from Table 12.3.3 A that a lower time-multiplier setting is possible with
the delta-connected current transformers because the effect of the delta connection
is to reduce the variations in relay-current level for different types of fault. Thus,
with delta-connected current transformers, the maximum operating time for an
external fault, using conventional inverse-definite-minimum-time-lag (i.d.m.t.l.)
relay characteristics is 4.0 s (for a single phase-to-earth fault), while the maximum
operating time with star-connected current transformers is 5.7 (for the 3-phase-fault
case). Since longer fault-clearance times can be accepted for single-phase-to-earth
faults, the use of delta-connected current transformers seems preferable.

Faults internal to generator: The back-up protection should operate quickly for
internal generator faults, since it then provides discriminative tripping in the event
of failure of the main protection. Fault currents for an internal fault will usually be
higher than those for external-system faults. Fig. 12.3.3E shows the relay-operating-
time curves for different types of fault. The curves apply to a simple system which

2 -
r~

°,,,,

al
t~

O 1 -

(i) (ii) 5 10 15 20
Fault power x 103 MVA

a Star-connected current transformers: all faults except phase-to-phase

b S t a r - c o n n e c t e d current transformers: phase-to-phase fault
c Delta-connected current transformers: phase-to-earth fault
d Delta-connected current transformers: all faults except
phase-to-earth
e E x t r e m e l y inverse relay characteristic for c o n d i t i o n a
f E x t r e m e l y inverse relay characteristic for c o n d i t i o n d
Note: Curves a - d apply for a conventional i.d.m.t.l, relay to
BS 1 4 2 : 1 9 6 6

(i) M a x i m u m external-fault power for delta-connected current

transformers
(ii) M a x i m u m external-fault power for star-connected current
transformers

Fig. 12.3.3E Relay operating times for generator internal faults fed from h.v, system
 Protection of generators and transformers 17

neglects the infeed of the faulted generator and assumes an equivalent, system
having equal positive-, negative, and zero-sequence infeed at the busbar. For the
star-connected current transformers, the infeed from the phase-to-phase fault will
be lower than for other faults, while, for delta-connected current transformers, the
phase-to-earth fault will give the lower infeed.
It is seen from the curves that the tripping times obtained from the use of delta-
connected current transformers are lower because of the lower t.m.s, setting needed
for external fault grading.
Fig. 12.3.3E also shows the characteristics of typical extremely inverse
overcurrent relays. This type of characteristic gives greatly reduced fault-clearance
times at the higher current levels, and, where delta-connected current transformers
are used, the back-up-clearance times for faults within the generator could greatly
assist in meeting the system-stability requirements.
12.3.4 Negative phase-sequence protection

As previously mentioned, there is a Limit to the degree of stator-current unbalance

a turboalternator can withstand before it suffers serious rotor overheating. Faults
, , , ,i i

Permissible 122t

Cooling Medium Type a Type b

12C = o.15 12C = 0 . I 0
i

Air 2O o

,, ,,, i

Hydrogen 0.5 p.s.i.g. 2O 13

O
__
., 3
Hydrogen 15 p.s.i.g. 15 10
o_
Hydrogen 30 p.s.i.g. 12 7
i
_ _ ,- _ _

oL .

= 2 Hydrogen 4 5 - 6 0 p.s.i.g. ~.s/2.s

¢J

z
P4
u

2 3
t (Permissible time in seconds)
Multiply by 122t factor from table

Type (a). Machines without directly cooled rotor copper.

Type (b). Machines with directly cooled rotor copper.
Fig. 12.3.4A Negative phase-sequence current capability for modern alternators
18 Protection of generators and transformers

external to the generator are usually cleared quickly by circuit protection, but
failure of remote protection to operate, or its associated circuit breaker to trip,
would leave the faulted circuit connected to the generator. All faults other than
one involving all three phases of the primary system give rise to a system of un-
balanced currents which may be resolved into its positive-, negative-and zero-
sequence components. By definition, the negative sequence components rotate in a
direction contrary to the d.c. field system of the generator. The stator flux thus
produced therefore cuts the rotor at twice its rotational velocity, thereby inducing
double-frequency currents in the field system and rotor body. The resultant eddy
R ! S
I i i _
'
"-"- ~.
or" ' ~IrZ

Y
- > . . . . . .

z R

_ ~
B lb
l " , . m

Sequence n e t w o r k , relay and c.t. c o n n e c t i o n s

Ir

IrZ
Positive sequence value of I x = irR + ibZ + iy - 0 IrR

Ibz

,r ,bJ , / ly
l lb R

IrR .lrz Positive sequence currents

lbR
Negative sequence value
or I X : irR + ibz + iy = I
I X = ly 1~ Z

Negative sequence currents. (Relay impedance X = 0)

Fig. 12.3.4B Generatornegative phase-sequenceprotection using simple bridge network

 Protection of generators and transformers 19

currents may be very large, so causing severe overheating of the rotor at those
points where the circulating eddy current is concentrated by winding slots. To
disconnect the machine before damaging temperatures are attained, a negative
IR
R .---...=
Iy
y .-...--- °

IB
B m

:N 2N

• i

~ fZ ~"
Zr
f
-, J

Zb .--~ /
i : .. I
I - ~ 11

I R, I r and VZr
I R and I r

Positive seq uence /f, V Zr

Iy
I B and Ib I Iy I B, I b and VZb

Vzb Vz +V
. A
r Zb
I R,I rand VZr / 4
/ I R and Ir

t Vz r ,
o ~ . ~ Negative sequence o ~"~' ,~ Vzr + VZb
Iy J ~ ' ~ l Band I b Iy ' ~ ~ ~ I B, I b and VZb
Phasor diagram for Phasor diagram for
trip unit (T) alarm unit ( A )

Fig. 1 2 . 3 . 4 C Generator negative phase-sequence relay (General Electric Company Ltd.J

 20 Protection of generators and transformers

phase sequence (n.p.s) current measuring relay is applied whose (n.p.s. current) 2 x
time characteristic matches that of the machine. The relay is so designed as to give
warning when the maximum continuous n.p.s, current rating of the generator is

T F x S

\. .-
0

2.0 i , x' \ a
\ •
, \
b\ '\
1.5 \ \
'\ \
0

1,1

u c ..\,, ' >:"

'b L ~..,~
'_.~
d
e~

d ~ ~ ~
_.~ ..~..-.'~ ~ ' ~
~ 1.0 k qt

..~ •
~
--" -'- *""
ii a~" I I * Ib.~
,B
II

:g
• , ~ ,

0.5 h
--. u -- . . . . - - --- --- J - " "-" " " -'- " "

i
! I

1 2
Time, S
a
C u r r e n t against time curve for 12t = 3.O
b C u r r e n t against time curve for 12t = 2.0
cand d Phase-to-phase fault
e and g Phase-to-earth fault
f and h Phase-to-phase-to-eart h fault
unloaded machine
mmmm,,,mm~m
loaded m a c h i n e

Fig. 12.3.4D Generator n.p.s-current/tirne curves for 500 MW generator loaded and unloaded
(h.v.. circuit breaker open), including effect of a.v.r.
 Protection of generators and transformers 21

exceeded (12 c) and to trip when the value of (12)2 t, where t < 100 s exceeds the
capability of the generator. Fig. 12.3.4A shows the capabilities of modern British
turboalternator sets.
Typical n.p.s, networks are shown in Figs. 12.3.4 B, C and D.
The output of the network is proportional to the generator n.p.s, current in each
case and is fed into a relay with an inverse square law characteristic. The overall
arrangement is matched to the generator n.p.s, current capability curve such that
tripping occurs when the curve value is reached. An alarm unit may also be pro-
vided with a range of settings to cover the required range of I2c. The required
inverse square law characteristic is obtained by use of an appropriately designed

(.'.T °
Negative
sequence Comparator
input network
Integrator

Trip

Definite [
Starter time . ~ _ Alarm

Fig. 12.3.4E Staticn.p.s, relay (Type CTN, GEC Measurements)

induction disc relay, a thermal replica relay (Fig. 12.3.4C) or a 'static' relay in
which the characteristic is shaped by a resistor/Zener diode circuit. (Fig. 12.3.4E)
The settings of the n.p.s, relay are determined by the stator n.p.s, current
capability of the generator. Fig. 12.3.4D shows the n.p.s, currents for a typical 500
MW machine for different types of fault condition. The unloaded case (shown
dotted) gives lower n.p.s, currents than the loaded case, except for a phase-to-earth
fault, and the maximum current, and therefore the minimum operating time, are
given for a phase-to-phase fault on a loaded machine. This gives a required tripping
time of 1.3 s for ~ t = 2-0 and 1.9 s for I~ t = 3.0.
12.3.5 Interturn fault protection

The differential current protection described in Section 12.3.1 cannot detect inter-
turn faults which remain clear of earth, since there is a balance of the currents
entering and leaving the winding despite the presence of a large current circulating
the shorted turns. Interturn faults are not normally protected against because of the
technical difficulty of so doing. If intertum faults occur in the stator slots, they
quickly develop into faults to earth and are cleared by the stator earth fault protec-
tion. There is, however, the possibility that they may occur at the winding ends and
so cause extensive damage to the generator before the fault evolves to one detectable
by other protection.
 22 Protection of generators and transformers

12.3.6 Loss of excitation (field failure) protection

Loss of excitation results in a generator losing synchronism and running above

synchronous speed. Operating as an induction generator, it would produce its main
flux from wattless stator current drawn from the power system to which it was still
connected. Excitation under these conditions requires components of reactive
~urrent which may well exceed the rating of the generator and so overload the stator

Rotor earth
fault relay

Generator
field

Shunt

Main
exciter
T| --I - -

I
Alarm ~ =~ - - - - ~

Fig. 12.3.6A Generator field failure protection using sensitive relay

(Diagram also shows simple rotor E/F scheme)

Generator
field

Shunt

Main [~!1 ~ " ~ 1 1

exciter 1 I
' P-'U.,U-- I
T I I
e- ~- "-=~:::¢~" i
I
Alarm 5 sees "-=-----,,--0 I C . . . . .
t. . . . . J
1
Trip 60 sees ~ - -~ ~ - - 1
Fig. 12.3.6B Generator field failure protection using underpower relay
 Protection of generators and transformers 23

rL
Alternative |&~/
position IX,~j
for v.t.
I
I
I
L. . . . . . . . . ~ I - --~--~1
(a) Relay c o n n e c t i o n s __
Mho relay

Typicalpowerswinglocus System fault locus

1A 1B
-R

/ I
Mean locus of
i m p e d a n c e vector
on loss of field

Relay characteristic

-X

(b) Relay c h a r a c t e r i s t i c s h o w i n g mean locus o f i n p u t

i m p e d a n c e at g e n e r a t o r t e r m i n a l s u n d e r loss o f e x c i t a t i o n
c o n d i t i o n s . N.B. Effects of r o t o r saliency not i n c l u d e d .

I A - Lagging c u r r e n t (heavy load)

I B - Lagging c u r r e n t ( l i g h t . l o a d )

Fig. 12.3.6C Loss of excitation protection using offset impedance measuring relay at
generator terminals

winding. Additionally the slip frequency currents induced in the damper windings
of the rotor would cause abnormal heating of the rotor.
Operation as an induction generator does not, therefore, damage the set immedi-
ately, but, as the higher ratings of modern machines are obtained by advanced
cooling techniques rather than increased frame size, the short thermal time constants
require the machine to be deloaded and tripped in a matter of seconds.
Fig. 12.3.6A illustrates a method commonly used to provide field failure and
rotor earth-fault protection on the smaller generators of a decade ago. The relay
 24 Protection of generators and transformers

combination operates when the field current is zero (or cyclically passing through a
zero during asynchronous running) and trips or alarms after a predetermined time
delay set on relay ?'2.
An alternative method employed a d.c. underpower relay in the field circuit as
shown in Fig. 12.3.6B. This relay is responsive to power flow and is held inoperative
by the normal flow of current from the main exciter to the field windings. Should
the power fall below a certain level, or reverse its direction (due to induced a.c. at
slip frequency) the relay contacts would close after a time lag to give an alarm.
As modern large generators may be required to operate with very low values of
excitation, both of the above schemes would be unsuitable. When a generator loses
synchronism, the quantity which changes most is its impedance as measured at the
stator terminals. Loss of field will cause the terminal voltage of the generator to
begin to fall, while the current begins to increase. The apparent impedance of the
machine will therefore be seen to decrease and its power factor to change. A relay
designed to detect the change of impedance from the normal load value may
therefore be used to provide protection against asynchronous operation resulting
from the loss of excitation (Fig. 12.3.6C).

12.3.7 Protection against pole-slipping

Generally speaking, pole-slipping protection was not provided on earlier, small

machines but on larger, modern sets, the severe mechanical torque osciUatiom
accompanying pole slipping conditions require prompt action to be taken if damage
to the generating plant is to beprevented and power-system disturbance minimised.
Pole-slipping conditions do not normally occur for system faults correctly cleared
by high speed main protection
Pole slipping, either between generators and the system, or between two sections
of the system, results in a flow of synchronising power, which reverses in direction
twice every slip cycle. When displayed in the complex impedance plane, this flow of
synchronising power is characterised by a cyclic change in the effective load impe-
dance VII as measured at the terminals of the two asynchronous systems and the
load impedance locus passes between the +R and - R quadrants as the real power
flow reverses in direction. The transition point on the/X axis corresponds approxi-
mately to the instant when the two systems are 180 ° out of phase (assuming the
load impedances are largely reactive). At this instant only reactive power is flowing
and the system voltage collapses to zero at the electrical midpoint of the two
systems. This midpoint is the centre about which pole-slipping can be comidered to
be taking place and its location, with respect to the relaying point, can be deter-
mined from the apparent load impedance to the point where the locus crosses the
IX axis. The load impedances seen from the terminals of individual generators either
side of the pole-slipping centre will depend on their respective contributions to the
total synchronising power flowing at any instant. Measurement of the magnitude,
direction and rate of change of load impedance relative to a generator's terminals
provides a convenient and generally reliable means of detecting whether pole
 Protection o f generators and transformers 25

I_=, Stage 1 , v I Stage 2 __1

I I -I

System

I i
I I
I
I I
I -- --
I
I
I
~
detection unit
F-C______] Counters
'
I
I
'I [~ S t a g e 1 and 2 trip
'I
L_ . . . . . _ _ _ _ _ _ _ . . . . . _J
Simplified block diagram

¢ - System impedance
angle
Ohm 1 ZI,, - Generator reach
jX
Offset m h o stage 2 ZR 1 "Offset mho reach
limit of reach \ ZR2 - System reach

%
Ohm 2
Restrain
71111
H.V. directional Operate
control relay

Locus of pole
slipping centre
(Stage 2)

'~K Locus of pole

Relay characteristics slipping centre
(Stage ! )
ZI.,
For operation: Offset mho stage 1
t 2 -t I <[" approx 40 ms. limit of reach

Fig. 12.3.7A Offset mho type pole slipping protection (GEC Measurements Ltd.)
 26 Protection o f generators and transformers

I ...... . .
Stage
. .
1 . . . .
_ .I _ . .
Stage
.
2 __1
I~ I -I

System

r- - T_.ZI_.. -]
] FPole.slipping [
II ~ t i o n unit I

I I
II Stage 1 and 2 trip
I[

L w . . . . . . . ,..,,. . . . . . . . . A

Simplified block diagram

- System impedance
angle
6 - Lens angle
ZR2 Z F - Generator reach
ZR2 - System reach
ZRI " Reactance
\
element reach
\

Reactance element,

Locus of pole
- ' - " - - slipping centre
H.V. directional (Stage 2)
re lay ----

--- ..~ nl/~

Ohm
line
? ~ ~ Locus of pole
slipping centre
(Stage 1)

F'or operation locus must enter lens

from left or right, traverse whole
Relay lens and remain within the lens for
characteristics a minimum of 50 ms.

Fig. 1 2 . 3 . 7 B Lenticulartype pole slipping protection (Brown BoveriJ

 Protection of generators and transformers 27

slipping is taking place. Individual generators, generally those on the side of the
larger system, although contributing synchronising power will not see a pole
slipping centre (i.e. their impedance loci will not cross t h e / X axis) if their real
power output remains positive, but care needs to be exercised in designing and
setting pole-slipping protection based on load impedance measurement if satisfactory
discrimination is to be obtained.
The pole-slipping detection relays are connected to c.t.s and v.t.s on the l.v.
side of the generator transformer and the directional control relay to c.t.s and v.t.s
fitted on the h.v. side of the transformer. Schematic diagrams of two manufacturer's
schemes are illustrated in Fig. 12.3.7A and B.
In both cases, the protection is of the impedance-measuring type, having two
stages:

Stage 1 reaching through the generator to the neutral point and into the trans-
former to cover the entire winding as far as, but not beyond, the h.v. terminals
(the directional control relay inhibiting operation if the system centre occurs
beyond the transformer h.v. terminals).; and
Stage 2 operating after an adjustable number of pole slips, whether the system
centre is within the generator transformer unit or out on the system.
The pole-slipping relays are capable of detecting the first pole-slip condition
when a slip, corresponding to the speed of pole-slipping is in the range +0.1% to
+10% on a 50 Hz basis. The protection must, of course, remain unoperated for
steady state loading, power swinging and correctly cleared system fault conditions,
although some of its component relays may operate.

12.3.8 Rotor earth-fault protection

Two methods are available for the detection of earth faults in the rotor circuit. One
method utilises a high resistance connected across the rotor circuit, the centre point
of which is connected to earth through the coil of a sensitive relay (Fig. 12.3.6A).
This relay will detect earth faults over most of the rotor circuit. There is, however,
a blind spot at the centre point of the field winding which is at equipotential with
the midpoint of the resistor under earth-fault conditions. This blind spot can be
examined by arranging a tapping switch which, when operated, shifts the relay
connection from the centre of the resistor to a point a little to one side. Alterna-
tively, one half of the resistor can be replaced by a nonlinear resistor which, since it
will change its value for different values of rotor voltage, will continuously vary
the effective resistor tapping voltage as the field conditions change. This method
has been used in the USA.
A second method utilises a small power pack connected to the positive pole of
the field circuit, in series with which is connected the fault detecting relay and a
high resistance (Fig. 12.3.8A). A fault at any point in the field system will pass a
cdrrent of sufficient magnitude through the relay to cause operation. The field
winding earth-fault relays are usually instantaneous in operation and are connected
28 Protection of generators and transformers

Field
circuit
breaker
I
I

Rotor

Dis-
Exciter
charge Field
windin
Resistor

Sensitive
moving-coil
relay
~ - . . . . . -I~ Alarm

Current limiting resistor

30 V d.c.

Auxiliary Q ~
supply
a.c.

_L
.¢.

Fig. 12.3.8A Rotor earth-fault detection using negative~potential biasing device

for alarm only as there is no immediate danger to the set. This device is similar in
principle to the battery negative biasing device referred to in Section 7.2.
It will be seen that with both of these schemes the earth return path is through
the body of the rotor. Since, however, the contact between rotor and stator is
through the bearing oil film which is non- (or intermittently) conducting, it is
essential to earth the rotor shaft by an additional earthed brush. This also serves
the essential purpose of discharging static electricity induced in the turbine rotor
by steam friction, thereby preventing the beating surfaces from pitting.
In some modern machines the traditional direct current exciter is replaced by an
alternator with a rotating armature, the field winding of the main generator being
supplied through rectifiers carded on the rotor, and a 'brushless' design is thereby
achieved. The inaccessibility of the main d.c. field circuit makes direct detection
of rotor earth-faults impossible.
The single earth-fault (or diode open circmt) would produce little change in the
level of excitation. A more severe interturn fault (or diode short circuit)would, in
all probability, tend to cause abnormal vibration of th_~set and a vibration detector
 Protection of generators and transformers 29

is sometimes employed to initiate deloading and tripping. The diodes are themselves
protected by fuses and a fuse failure protection relay is normally included.

12.3.9 Sensitive power protection

Reverse power protection has in the past been fitted to detect failure of the prime
mover (that is the engine or turbine) which it does by measuring the power drawn
from the system by the generator when it is motoring.
The power taken by the generator under such conditions is very low, being from
about 10% for engine driven sets to 2% for turboaltemators. The power factor of
the current depends, of course, upon the excitation level and may thus be either
leading or lagging. The wattmetric relay must, therefore, be highly sensitive and
have an accurate quadrature adjustment. It must also be connected as a 'true'
wattmeter and not be fitted with phase angle or low voltage compensation, as are
directional relays for fault power applications.
The wattmetric relay must be associated with a time lag relay to prevent
operation due to power swings.

12.3.10 Low forward power interlock

A more recent application of a sensitive power relay is on large modem alternators

where it is desirable to delay tripping of the electrical load on the generator until
after the steam has been tripped, so that the steam entrained in the turbine is used
in supplying electrical power rather than causing the set to overspeed, that is the
generator acts as a 'brake'. This feature is only provided for 'non-urgent' tripping
conditions. The contact of the wattmetric relay is associated with a time lag relay
whose contact is connected in series with the 'non-urgent' tripping contacts, and
with this arrangement the generator is not allowed to trip until the power output of
the generator falls below a small preset value. (Fig. 12.3.13 A).
Operation of the relay at low levels of forward power permits full protection
during run up and prior to synchronising, since the relay contacts will be closed
during this period.
To provide the highest degree of security under all conditions the relay contacts
are arranged to close when the generated power falls to 0.5% of full load output.
Such an accurate setting requires the current and voltage transformers to be of high
accuracy. Two relays are normally used and have their contacts connected in
parallel.
The generator internal fault tripping devices(for example differential protection,
etc.) are arranged to trip independently of this sensitive power interlock.
An alternative method of preventing the electrical tripping of a generator until
after its steam supply had been cut off was to use a pressure operated switch
connected into a suitable point on the turbine. The setting of such a switch had to
be not appreciably higher than the no-load steam pressure and the contacts were
set to be on the point of opening under no-load conditions.
30 Protection of generators and transformers

Two pressure switches with parallel connected contacts were sometimes used to
give additional safeguard against mechanical failure. A time delay relay was
associated with the pressure switch so that transient overshoot of the switch would
not permit tripping of the generator circuit breaker under unsafe conditions.
The method was recommended only for non-reheat machines in view of the
additional complications where reheat is used. In the case of reheat machines, three
possible contingencies must be guarded against:

(a) failure to close of the main steam supply valves

(b) failureto close of the valves after the reheater, and
(c) failure of both sets of valves.

Two series-connected pressure switches were required to cover these contin-

gencies; one operated by the pressure difference across the h.p. cylinder, and the
other by the pressure at some point in the turbine down-stream of the reheat valves.
In view of these additional complications the use of pressure switches is not now
recommended for reheat machines.

12.3.11 Overspeed protection

It is essential for the governing system of a turbine generator to incorporate safety

devices to prevent dangerous overspeeding. The importance of these devices is such
that routine testing facilities should be included to check the correct functioning of
the valve gear on normal load, and normal operating procedure should verify the
operation of all safety devices prior to going on load and after removal of load
before shut down.
Most large turbines are now fitted with overspeed limiting gear designed to
detect sudden loss of load and to close emergency valves immediately, to limit the
magnitude of the temporary speed rise.
In a typical scheme this is achieved by monitoring the electrical output of the
generator using a wattmetric relay. This relay will detect a sudden loss of output
and operate instantaneously to close its contacts. A second relay monitors the
steam input to the turbines at a chosen stage and the contacts are held dosed when
the steam pressure is in the full load region.
A sudden loss of load will give instantaneous operation of the output relay but
the steam input relay does not operate immediately because steam is being
admitted to and expanding in the turbine. Under this condition the emergency
valve solenoids are energised giving instantaneous control of steam admission. The
emergency valves remain closed until falling pressure or restoration of load restores
the machines to normal control.
The action of this equipment is clearly much faster than that obtainable from
the governing system which requires an actual overspeed to produce a response and
take corrective action. It is for this reason that overspeed limiting equipment of the
type described is often installed where reheat turbines are used, because the long
 Protect~on of generators and transformers 31

steam pipes of relatively large bore interconnecting each reheater and the reheat
sections of the associated boiler plant present special problems due to the large
volume of steam entrained.
The overspeed limiting equipment then operates additionally into the interceptor
emergency stop valves associated with each interceptor steam chest to give instan-
taneous control of the steam entering the turbine at all stages.
In the ultimate, overspeeding of the machine beyond the safe limit (10%) will
cause operation of the overspeed bolts, and shut the stop valves.

12.3.12 Underexcitation limiting

The operating characteristics of a cylindrical rotor a.c. generator feeding an inter-

connected system are such that a def'mite positive minimum of rotor excitation
must be maintained to ensure synchronous stability.
From the design criteria this limit can be described and is usually termed the
'theoretical stability limit'. To guarantee stability in service a safety margin is
added to the theoretical limit to make a 'practical stability limit'.
If the control of excitation is by manual adjustment the operating point of the
generator is usually maintained by observation and adjustment within an area
bounded by the practical stability limit in the leading (underexcited) reactive
zone, the stator and rotor heating limits in the lagging (overexcited) zone, the kW
limit being set by the turbine capacity or steam conditions.
When an automatic voltage regulator is employed to control excitation, measures
must be taken to ensure that excitation is not reduced below a safe limit. This limit
will depend upon the form of voltage regulator and will correspond to the practical
stability limit for most electromechanical regulators, but may approach the
theoretical stability limit for regulators having no dead band and a high speed of
response. If the level of excitation approaches the limit a subsidiary control loop is
employed to feed a modifying signal into the automatic voltage regulator control
loop to prevent further lowering, and in certain cases, to raise excitation in addition
to giving an alarm.
The initiating signal for the under excitation limiting device may be taken from a
measure of rotor angle, or of rotor excitation current, or of a derivation of the
generator operating point from a measure of watts, revs and terminal voltage.

12.3.13 Mechanical and hydraulic trips

In addition to the electrical protection discussed earlier in the chapter, various

mechanical and hydraulic devices are employed to detect abnormalities in the
turbine and its boiler. Some of the devices intiate alarms to the operator, and these
are not dealt with here, but others trip the unit automatically and so, to illustrate
complete protection of a modern, large, turbo-generator, they have been included
in Fig. 12.3.13A.

(a) Local turbine tn'p level: a grab-handle on the turbine pillar enabling an
32 Protectionof generators and transformers

Electrical
solenoid on
I
I turbine to
release fluid ! -J
& trip turbine
LOCAL TURBINF steam valves TURBINE"

OVF.RSPEEI) TRIP
TRIP LEVER

"
~ !

t
"'
. = . ,. . /~ i[ VALVESI
STEAM I

:~ LOW S T E A M I N L E T P r E s s u r E
LOSS OV L U B R I C A T I N G O I L -
Itt~~ Relay
-Rl ..... fluid / I
._~ -"I LOSS0 ~ s~.e~:o (;OV~,NO,_T,~,~ I
switch [ BOILER
, ,
t" ' L O S S OI.' B O I L E R ,WATER , j _ " -- ~i I.'iRIN(.;
TRIP

•- - ' - Electrical signal

EMERGENCY
. . . . PUSH
. . . B. U .T T.I ) N REMOTE:
' & LOCAL --'-- Mechanlcal/hydraulic
signal

A - Contact which closes on

+ operation of sensitive power

relay for detecting cut-off
o f steam supply to turbine

•-'i'NEGATIVE PHASE S E Q U E N C E ' ,, ,,

H GEN. S T A T O R E/V INSTANTANE:OUS

a.

--- { TEM:'ERA'rURI,:
GEN. TRANS. WDG. .[

,. -" ~ . Ta^N~. O~'~^~C ~O~:':.

GI L
GEN. T"ANs..u(.'HHOLZ
GI : su r~;,.:
(

"-" GI
GEN. T R A N S . H.V,, O V E R C U R R E N T
--- GEN. T R A N S . H.V. REI".
GI
--- E'iRST
I.'i MAIN~EN.
L (
. E'EEI~L:R PROT.

~= "" SECOND M A I N GEN. FEEDER PRO3r.

~1 "" INTERLOCKED OVE:RCURRENT

I BREAKER E:AIL
UNIT
,0- TRANS.
L.V.C.B.

~ ~, UmT TR^N;'."OV~:R^LL'~OT. E'IELD

~ SWITCHES
1 " - - I UN,T T~^NS. H.V. O V ~ : ~ C ~ k H ~ T
.V.C.B.
J H'V" B U S B A R PROTECTION . . . . . . I
J

Fig. 12.3.13A Tripping logic for a modern large turbogenerator

 Protection of generators and transformers 33

attendant to shut-down the set locally should the need arise.

(b) Overspeed trtp: a centrifugal device initiating tripping if the turbine speed
rises 10% above the synchronous speed.
(c) Low steam inlet pressure: a device sensing loss of turbine steam pressure due
to loss of boiler firing.
(d) Loss of lubricating oil: an off-pressure device monitoring the supply of
lubricating oil to the turbine bearings.
(e) Loss of boiler water: a water-level detector initiating boiler firing tripping
to prevent extensive damage being done to the boiler tubing due to overheating.

(f) Vacuum trip: independent vacuum-operated unloading control and tripping

devices. The control progressively reduces the steam flow into the turbine as the
vacuum falls over a predetermined range, but if the vacuum falls below a pre-
determined minimum value a vacuum operated switch initiates tripping.
(g) Emergency push button: operation of an emergency push button initiates
tripping of the turbine steam valves. The remaining main plant will trip via the
low power interlock contact A.
To enchance the integrity of modem turbogenerator protection tripping, the
trip-initiating devices are divided between several, independent d.c. tripping supplies,
typically, as follows:

H.V. switchgear, first supply

Busbar protection
Generator feeder, first main protection relay 1
Interlocked overcurrent relay
Generator transformer h.v. overcurrent relay

H. V. switchgear, second supply

Breaker fail protection
Generator feeder, second main protection relay 1

Power station, prst supply

Generator feeder, first main protection relay 2
Unit transformer overall protection
Unit transformer Buchholz surge
Generator stator ElF relay (instantaneous, low resistance)
Generator transformer overall protection
34 Protection of generators and transformers

Negative phase-sequence protection

Generator transformer Buchholz surge
Generator transformer winding temperature trip
Emergency push button
Generator stator E/F relay (1 st i.d.m.t., high resistance)
Loss of excitation protection
Loss of vacuum trip
Low steam inlet pressure trip
Unit transformer 1.v. standby E/F
Pole-slipping protection
Loss of stator water flow
Loss of lubricating oil
Loss of speed governor trip
Loss of boiler water

Power station, second supply

Generator feeder, second main protection relay 2
Unit transformer 1.v. restricted E/F
Unit transformer h.v. overcurrent
Generator stator E/F relay (2nd i.d.m.t., high resistance)
Generator stator E/F relay (i.d.m.t., low resistance)
Generator transformer, h.v. restricted E/F
Negative phase-sequence protection

12.4 Gas-turbine driven generators

12.4.1 Direct connected, gas-turbine sets

At all modem power stations in the United Kingdom, gas-turbine powered

generators have been installed to provide:

(a) an independent supply to the auxiliaries of the main steam driven units in the
event of unacceptably low frequency on the grid system
(b) additional generation into the grid system if required by system conditions

(c) start-up supplies to a station detached from the grid system

(d) independent supply to ensure operation of essential drives, e.g. main bearing
lubricating oil in the event of loss of normal supplied. This duty is, in effect, a
back-up to the d.c. battery system.
 Protection of generators and transformers 35

! 1 kV unit board

[ :1
E~-E ,-~
E~-E 'I
[

' 1
i I
i i
I
I
I I
mmm .
I
I
I
i
I

1.6 MVA 35 MW I
I " I
" I
I
I _
i
E -E'i Et~, E J' L[2.L~
s t 1
-E t. . . .

415V
Stage 1 trips h.v.c.b, o n l y
Stage 2 trips gas t u r b i n e
9

Fig. 12.4.1A Protective relay a.c. circuits for power station emergency gas-turbine set

Fig. 12.4.1A shows the protection of a 3 5 MW, 11 kV generator and its connections
(but not the gas-turbine). It comprises:

differential current protection covering the generator and its outgoing connections
loss of excitation protection
standby earth-fault protection (2 stages)
- stage 1 trips the h.v. circuit breaker only
- stage 2 trips gas-turbine
voltage-controlled overcurrent protection
negative phase-sequence protection

The teed unit transformer and its h.v. connections are protected by overcurrent
and restricted earth-fault relays.
36 Protection of generators and transformers

12.4.2 Transformer connected, gas-turbine sets

The functions of main gas-turbine generating plant are to provide:

(a) Normal, economic generation close to load centres, giving added security to
load demand by reducing its dependence on remote generation and its
associated transmission system
(b) Emergency generation at times when severe system disturbances cause a local
generation deficit
(c) Peak-lopping capacity for system frequency regulation purposes, thereby
reducing the amount of national spinning reserve required
(d) Reactive compensation whilst running on load, or synchronous compensation
(with clutch disengaged).
Typically gas-turbine sets are each of 70 MN capacity powered by either two
35 MN or four 17.5 MW power turbines (Fig. 12.4.2.4,). The protection arrangement
and tripping logic are illustrated in Figs. 12.4.2B and 12.4.2C respectively. A
particular point of interest in Fig. 12.4.2B is the absence of biased differential c.t.s
in the unit transformer tee-off connections. In this case the low rating (high
impedance) of the unit transformer limits through fault currents on the 415 volt
system to a value well below the differential protection fault setting.

415Vsw.bd.

I,, ,

,722 . J
i °.,, !

,, ,

Gas Gas
gen. I A 2 gen. I B I
,, r , ,,,,,,

,1
Power
turbine l A
Electrical
generator --tlup:;:~rlB
,,! 71
Gas Gas
gen. I A ! gen. 1 B2
i
Fig.12.4.2A Key diagram of gas-turbine generator
 Protection o f generators and transformers 37

1 32 k V

F .,-,,-- ,,_._,..

E'
I
I
! I
! I

~-~~
I

I
U
l-'S
'75 M V A
1 1 / 1 3 2 kV
I
I
m !m
g , i 500 kVA

II '7o MW

415V . . . . . .

Fig. 12.4.2B Protective relay a.c. circuits for main gas-turbine generator unit
 Next Page
38 Protection of generators and transformers

GAS G E N E R A T O R I A I T R I P S FUEL V A L V E
GAS G E N E R A T O R IA2 T R I P S SYSTEM
I A I & IA2
POWER T U R B I N E ! A T R I P S

POWER T U R B I N E IA O V E R S P E E D

E M E R G E N C Y PUSH B U T T O N

GAS G E N E R A T O R IB! T R I P S
FUEL V A L V E
GAS G E N E R A T O R IB2 T R I P S SYSTEM
POWER T U R B I N E IB T R I P S IBI & IB2

POWER T U R B I N E IB O V E R S P E E I )

I A & IB
FUEL SYSTEMS
BOTH T R I P P E D

l
ELECTRICAL GENI'RATOR

LUB. OIL P R E S S U R E LOW

UNIT T R A N S F . .
T H R U S T B E A R I N G D R A I N TEMP. E X C E S S I V E
L.V.C.B.
AIR I N L E T D I F F E R E N T I A L PRESS. E X C E S S I V E
O U T L E T C O O L I N G AIR TEMP. E X C E S S I V E
B E A R I N G TEMP. E X C E S S I V E
B E A R I N G V I B R A T I O N EXCESSIV E H.V.C.B.
LOW F R E Q U E N C Y (40 H z ) T R I P i 32 kV

N E G A T I V E PHASE S E Q U E N C E T R I P
S T A T O R E A R T H F A U L T IDMT ( F I R S T )
S T A T O R E A R T H F A U L T IDMT ( S E C O N D ) SUPPRESS VOLTS
LOSS O1" E X C I T A T I O N BY A.V.R.

A.V.R. F A I L

GENERATOR TRANSFORMER
BUCHHOLZ SURGE
OV E R A L L P R O T E C T I O N
WINDING T E M P E R A T U R E
H.V. R E S T R I C T E D E A R T H F A U L T

OV E R F L U X I N G P R O T E C T I O N

UNIT T R A N S F O R M E R

B U C H H O L Z SURGE;
H.V. O V E R C U R R E N T
L.V. R E S T R I C T E D E A R T H F A U L T

GENERATOR FEEDER PROTECTION

BUSBAR P R O T E C T I O N
G E N E R A T O R H.V. O V E R C U R R E N T

Fig. 1 2 . 4 . 2 C Protection tripping logic for main gas-turbine generator

Previous Page Protection of generators and transformers 39

12.5 Transformer protection

The protection provided for a power transformer depends to some extent upon its
size and rating, and will comprise a number of systems each designed to provide the
requisite degree of protection for the different fault conditions outlined in Section
12.2.2. For large units high speed protection is essential.

12.5.1 Unbiased differential protection

The unbiased differential overall transformer protection system is similar in

principle to that for a generator. In this case, however, the differential protection
system compares h.v. and 1.v. currents, which are in a known relationship under
healthy conditions, rather than the same current entering and leaving the protected
apparatus, as for generator protection. It is for this reason that the transformer
differential protection system is capable of detecting interturn short circuits since
these change the effective overall transformation ratio of the power transformers.
The c.t. connections must be arranged to give a through fault balance taking into
account the transformer vector group reference with respect to windings, connec-
tions, and turns ratio. The guiding principles in establishing c.t. connections are
that (a) zero-sequence currents should be eliminated from or correctly compensated
in the relay circuits since, almost invariably, the transformer connections will not
permit transformation of zero-sequence currents, and (b) the phase shift due to the
through transformation of positive and negative-sequence currents must be
correctly compensated.

1R K Ydll ¢r~ I r

-- r

Ib
- b

. ~ . ~ N

(a) T r a n s f o r m e r connections C.T. ratio IS/I

~-x- pL (I Y - I R ) lr/l S
___jo -__Z_ > "
C.T. r a t i o ~'~V¢3 Ip (IB- IY) ly/l S
J
Ip/(..~) -~ Y
--L-(I R-IB) lh/l s
Ip . . . . -~ b
q
Aq

(b) R e l a y a n d c.t. c o n n e c t i o n s - -

Fig. 12.5.1A Overall differential protection of star/delta power transformer

40 Protection of generators and transformers

The elimination of zero-sequence currents from the relay circuits is usually

achieved by associating a delta c.t. connection with a star connected transformer
winding.
Considering the typical example of a star-delta transformer (Fig. 12.5.1A)
having a star/delta voltage transformation ratio of K: 1, then the line currents on the
delta side I~, ly and Ib are

K
I,= - ~ (& - &)

K
Ib = "-~ (IB- IR )

when IR, Iy and IB are the line currents on the star side. Balance is therefore
obtained using star connected c.t.s on the delta side of the transformer and delta
connected c.t.s on the star side of the transformer, as shown in Fig. 12.5.1A, since
with turns ratio K: ~/3 and Is = KIp

m
(ly - IR) etc.

If similar reasoning is applied to the transformer and earthing transformer

combination of Fig. 12.5.1B, the following relations are obtained
K I~
7

K
• ------- . . , ,

C ~ (~B - I R ) - / ~ 3

which shows that through fault balance is obtained by the addition of neutral c.t.s
to the delta side of the transformer, as shown in Fig. 12.5.1B. It will be noted that
the additional neutral c.t.s have a ratio which is three times that of the star
connected c.t.s in the delta side of the power transformer.
In Figs. 12.5.1A and B, Ip and Is are, respectively, the primary and secondary
winding rated currents.
 Protection of generators and transformers 41

Ydll
IR |r
R ~ r
ul ~--.--r
1

ib
~--,b

N
(a) Transformer connections
"~! n

C.T r a t i o
• Ip/( _~-
1 ) ('.T. ratio IS/I
R

2_1
Neutral
c.t. ratio
(b) Relay Ld c.t. connections
t- ls/(--~-3 )

III
,=,

Fig. 12.5.1B Overall differential protection of star/delta transformer with I.v. earthing
transformer

The effect of tap changing equipment upon the overall transformer ratio must
also be borne in mind in the application of a differential protection system. This, in
effect, changes the ratio of the transformer according to tap position so that a
differential protection system using a fixed c.t. ratio cannot give adequate balance.
The unbalance current resulting from inaccurate matching of transformer and c.t.
ratio will increase in magnitude as the through current increases, so that, for
example, a ratio difference producing 15% spill current at rated load will give rise
to a spill current of 150% for a through fault of 10 times rated load. An unbiased
high speed relay would require a fault setting in excess of this if stability is to be
achieved under through fault conditions. In practice, for optimum performance, the
c.t.s ratios are based on the transformer turns ratio at the mid-tap position.
An inverse time lag relay is used for unbiased differential protection systems,
often in conjunction with a high set instantaneous relay operating at a higher level
than the maximum through current spill. The relay current and time settings are
42 Protection of generators and transformers

adjusted so that it will grade correctly with back-up protection under maximum
unbalance conditions. Where both the inverse and high set relays are employed
typical current settings are 50% for the inverse time lag relay and 600% for the high
set relay.
Source Z = 5 % o n i 0 0 M V A
100/0.577 132/33 kV,30 MVA, 10% 600/1

570A 3950A

2
I 140A J,_ i "

9.87A 6.58A

9.87A 6.58A
~r s"

0 0

!A !B IC

I . I ) . M . T . relays
,l
I A = Current setting
A[ B C
0 . 5 = Time multiplier

O n internal fault (I,'1), ! A = ! B = 9 . 8 7 A ( N o I.v. infeed)

O n external fault (I.'2), I A = ! B = 3 . 2 9 A

Fig. 12.5.1C Rough-balance system of transformer differential protection showing relay

currents under internal and external fault conditions

An interesting extension of this system is the 'rough balance' protection, in

which the h.v. and 1.v.c.t. ratios are deliberately mismatched. The induction
pattern relay thus provides low speed differential protection and also a more
positive measure of back-up protection for external fault conditions.
A typical application of the scheme is shown in Fig. 12.5.1C.
Where high-speed protection is required with a low fault setting it is essential to
use a percentage bias relay to overcome the unbalance current effects due to tap
changing.

12.5.2 Biased differential protection

In protecting power transformers equipped with on load tap changing facilities the
overall differential protection must incorporate a bias feature if a low fault setting
and high operating speed are to be obtained. For smaller transformers the
percentage differential (biased) induction-disc relay depicted in Fig. 12.5.2A gives
 Protection of generators and transformers 43

.---% 0-'- -- --

(a) (h)

Fig. 12.5.2A Induction-type biased-differential .relay for overall transformer, generator or

genera tor/transformer pro tec tion

some improvement on the unbiased schemes described in the previous Section, but
is slow in operation. With the larger and more important transformers, the relay
current and time settings necessary to ensure stability on the magnetising inrush
currents produced by switching in the transformer are inadequate to provide high-
speed protection.
A high speed biased differential relay incorporating a harmonic restraint feature
will prevent relay operation under magnetising inrush current conditions. A typical
oscillogram of transformer magnetising inrush currents is shown in Fig. 12.5.2B, the
initial peak value of the inrush current in any phase depending on such factors as
the instant of switching, and the magnetic condition of the core. Maximum peak
values equal to 6-8 times the rated current of the transformer can occur. Insofar as
inrush current affects the operation of transformer differential protection relays
two aspects are of significance. First, that this current flows in one winding only of

Red phase

o
C'lcle
Instant of switching

Yellow phase

Blue phase _

Fig. 12.5.2B Oscfllogram o f magnetising-inrush currents on a three-phase transformer

 44 Protection of generators and transformers

l Bias
coil

t' 2

L2

~ Operating
coil
Instantaneous ~ H.S.
high-set "~
element
Stabilising resistor

LIC I = Fundamental frequency acceptor circuit

L2C 2 = Second harmonic acceptor circuit

L3C 3 = Third harmonic acceptor circuit

Fig. 12.5.2C GEC biased~ifferential transformer protection with harmonic restraint feature
(only one phase shown)

the power transformer (the winding being energised) and therefore appears to the
protection as an internal fault condition. Second, that this current differs from an
internal fault current insofar as its wave form comprises a high percentage of
harmonics. Of these, the second harmonic has particular prominence under all
'switching in' conditions, as Fig. 12.5.2B shows.
It thus follows that a relay designed to detect the second harmonic component
in the magnetising inrush current can be made to utilise this as a means of
discrimination between internal fault currents and magnetising inrush conditions.
This is achieved by the use of a second-harmonic filter which is arranged to inject
an additional bias current in the relay circuit proportional to the second-harmonic
component. Typical arrangements of practical schemes are shown in Figs. 12.5.2C
and 12.5.2D.
In Fig. 12.5.2C, the harmonic restraint feature is added to each phase separately,
 Protection of generators and transformers 45

mm I

L: I- LZ

R I ~ R I l.-..i.-J R

LC = S e c o n d harmonic restraint circuit

T = Transductor
R = Relay

Fig. 12.5.2D Reyrolle "Duobias" biased differentia/ transformer protection with harmonic
restraint feature

as shown, whereas in Fig. 12.5.2D a transductor relay is used with the harmonic
restraint derived from all three phases and used to bias all three transductors.
A sensitive relay element is essential to reduce c.t. burdens to a minimum, and
modern harmonic restrained relays utilise a sensitive moving coil or transductor
type relay as the basic comparator. Typical current settings for biased differential
transformer protection relays are 50-100% for the low-speed induction-disc relay
and 20-30% for the harmonic restraint relays. In both cases a range from 10-40% is
usually provided for the through current bias.

12.5.3 Restricted earth-fault protection

The difficulties inherent in the provision of an adequate earth fault sensitivity in

the overall differential protection system often require that restricted earth-fault
protection should be added to both windings of the transformer. Separate c.t.s can
46 Protection of generators and transformers

I I

I I I
, _ L _ L J_.,
I I
i F.T. I
L -- ..1----J

tL---
~---
.i.
] I I
I I N
i.J
I
t R.E.F. J_
,,,g,

Fig. 12.5.3A Overcurren t and restricted earth-fault pro tec tion

H.V. L.V.

.Fr'.
y
.rr-~.
~'L.

I'.T.

,I

m ,.,_.

I
I V I
Aux. c.t. Biased differential Aux. c.t.
relay

Fig. 12.5.3B C. T. circuit diagram for combined biased differential and restricted earth-fault
protection
 Protection of generators and transformers 47

be used for this purpose, as shown in Fig. 12.5.3A, or the restricted earth-fault
protection may be operated from the c.t.s associated with the overall differential
protection. A typical arrangement is shown in Fig. 12.5.3 B.
It will be seen that compensation for the power transformers h.v./l.v, star delta
connection is achieved by a similar star/delta connection of the auxiliary c.t.
associated with the h.v., c.t.s. The star/delta/star auxiliary c.t. associated with the
l.v., c.t.s provides the four-wire connection of the line c.t.s (necessary to connect
with the 1.v. neutral c.t. to give l.v.. restricted earth-fault protection)whilst
eliminating the zero sequence components of 1.v. earth-fault current from the
three-wire connection required by the phase fault differential relay. Ratio
correction for a mismatch of h.v. and 1.v. line c.t. ratios may be accommodated in
either, or both, sets of auxiliary c.t.s. For solidly earthed, star-connected
transformer windings, an effective setting of greater than 50% but less than 100% of
rated current is usually specified. For transformer delta windings connected to
resistance earthed systems, the setting specified should lie between 20 and 25% of
the neutral resistor current setting.
12.5.4 Overeurrent protection

The degree of protection afforded by an overcurrent relay of the i.d.m.t, type is

somewhat limited when applied to a transformer. Since the relay must not operate
under emergency loading conditions it requires a high current setting (often about
200% rating). Also, the time setting may have to be high in order to grade with
other overcurrent relays on the system. Clearly overcurrent relays provide negligible
protection for faults inside the transformer tank, and may be very slow even for
terminal faults where high fault currents are involved.
On large transformers, therefore, overcurrent relays are usually employed only as
back-up protection for terminal faults, or uncleared 1.v. system faults. In such cases
the overcurrent relays may be installed on one or both sides of the transformer,
according to requirements. Moreover the relays may trip only the side of the trans-
former with which they are associated, or they may trip both.
Two-stage overcurrent protection is sometimes employed instead of separate
i.d.m.t, overcurrent on both sides of a transformer. The two-stage scheme comprises
one i.d.m.t, overcurrent relay and one time-lag relay. The i.d.m.t, relay is usually
energised from current transformers on the h.v. side of the transformer, the source
of infeed. Operation of this relay trips the 1.v. breaker and starts the time-lag relay.
The latter requires a time setting, say 0-3 sec., so that it does not operate before the
1.v. breaker trips. If the fault persists the time-lag relay completes its operation and
trips the h.v. breaker. Since the contacts of most i.d.m.t, relays have a long reset
time it is usually necessary to connect in series with them the contact of an
instantaneous overcurrent element which has a high speed of reset, so that the time-
lag relay is de-energised as soon as the fault is cleared. The instantaneous elements
would have the same current setting range, and prescribed setting as the i.d.m.t.
relay. The advantages of two-stage overcurrent protection are mainly economic, the
chief saving being the cost of a set of three current transformers.
48 Protection of generators and transformers

A disadvantage of this arrangement is, however, that it does not provide dis-
criminative back up protection for faults on the transformer 1.v. connections.
Frequently, supergrid transformers have their l.v. connections separately protected
by differential protection (Fig. 12.6.4B) but if this protection should fail to operate
correctly, the two stage, h.v. overcurrent protection on the transformers feeding
the l.v. system would operate and randomly trip their l.v. breakers. If however each
transformer is equipped with l.v. overcurrent protection, breaker tripping would be
discriminative since individual l.v. system infeeds would aggregate in the
overcurrent protection of the faulted transformer to give it fastest operation.

12.5.5 Directional overcurrent protection

Directional overcurrent relays may be used to provide discriminative phase fault

protection for two parallel transformers where there is no 1.v. source infeed. Where
an l.v. source infeed exists care must be taken to ensure correct discrimination for
faults on the h.v. system external to the transformers.
An analysis of the various fault conditions on transformer circuits shows that the
(90 ° -45 °) arrangement obtained with a 90 ° connection of a 45 ° relay gives best
discrimination. Fi_g. 12.5.5A shows the basic connections for the directional over-
current relay, and Fig. 12.5.5B illustrates the performance of the relay under
phase-fault conditions. The circuit diagram illustrates the current flow in the
Bu.4hars

...... , ,, ,,, : - - -

('urrent
.~. circuits

V~ltage
circuits

I,¢ Y B

Fig. 12.5.5A Relay connections (90°) for directional overcurrent protection

 Protection of generators and transformers 49

Ydl

-q41 m
m ~ b T__

h.v°
i.v.
h usha r busbar

Iiealthy
c,,nditi,,ns J ~ h
B~" I~ y Y

I' a ult ~ x . YI J
c~nd it i<~ns h
VI
h
I
II~ Vry ~,,¢, V I

I~'\~" .,'¢ Y

vh ~ "~'- yn
v ~I'I/ B x

IyP k~Y
I -
iy
/
V!
Pl~as~Jrdiagram for relay at A Y I'has~Jr d i a g r a m for relay at B

V R V y V B = Star side voltages

V ~ V ~ V ~ = Star side relay coil voltages (not to scale)

R Y B
IR! Y = Star side fault (and relay) currents
V r Vy V h ~ Delta Side voltages

V t V t V u ~ l)eltaside relay cecil voltages (not to scale)

r y h
! r ly i h = Delta side fault (and relay) currents
Assumed fault impedance angle ~- 70 °

Fig. 12.5.5B Use of 9 0 ° connection of 45 ° relay for parallel transformers showing fault
performance
50 Protection of generators and transformers

parallel transformers and the phasor diagrams for the primary system voltages under
both the healthy and the faulty conditions are shown separately.
From a study of the phasor diagrams for the relays at A and B, it is seen that the
use of the (90 ° -45 ° ) arrangement gives positive operating torque for the relays on
all three phases of the faulty circuit; it being assumed that positive torque occurs
when the angle between the relay coil voltage and the relay current is within the
range of +90 ° t o - 9 0 ° . On the healthy circuit the currents are in the opposite
direction, and it is clear that all three relays will restrain in this case. If a similar
study is made for relays using the 30 ° connection (see Chapter 8) it is found that
one of the end B relays on the healthy circuit will operate incorrectly, thereby
causing tripping of the healthy circuit, and one of the relays on the faulty circuit
will restrain when it should in fact operate.
Directional overcurrent protection is also fitted to the 1.v. side of transformer
feeders to detect reverse flow of current through the transformer, such as will
occur for h.v. feeder faults. Typical applications are discussed under transformer
feeder protection (Section 12.9).
Directional overcurrent protection cannot normally be applied where generation
exists on the 1.v. side of the power transformer since reverse load flow is possible
under healthy system conditions. In such cases, however, the directional element
of the relay can be arranged to have an additional voltage restraint feature such that
relay operation only occurs for reverse current flow associated with a voltage
collapse denoting a fault condition.

12.5.6 Interlocked overcurrent protection

Where for economic reasons it is necessary to locate the protective current
transformers on one side of the circuit breaker only, a fault occurring between the
circuit breaker and its associated current transformer may not be adequately
guarded against.
In Fig. 12.5.6A, for example, a fault occurring at point F is detected by the
busbar protection which trips the circuit breaker. It is clear, therefore, that the
fault will not be disconnected, except possibly by back-up protection involving a
long clearance time. One solution would be to arrange the busbar protection to
trip both sides of the transformer. This would not be acceptable if the other side of
the transformer were connected, for example, to a single-switch or mesh substation,
or if the transformer was banked with another. In the case of the genuine busbar
fault it would cause unnecessary disconnection of healthy equipment.
To discriminate between a busbar fault and the fault at F, busbar protection trip
relay contacts are used to start an induction disc relay. If fault infeed persists for
more than its operating time (about 0-5s), the relay initiates tripping to the backing
breakers.
A later development of the interlocked overcurrent principle is now being
installed on the supergrid system. Known as 'circuit-breaker fail protection' its role
is to ensure selective automatic tripping of the breakers 'backing' a circuit breaker
that fails to clear a fault. If c.t.s associated with a circuit breaker carry current after
 Protection of generators and transformers 51

Busbar

Busbar
protection
Fault ~ I

,,

B.B. ~~ ~~ Unit
prot. protection
~ --

?:-'i
I I
I I
I.
. . . . . . ~ . . . . . . L~---
I

hJ
i

Fig. 12.5.6A A p p l i c a t i o n o f in terlocked o vercurren t p r o tec tion

an interval marginally longer than designed tripping time of the breaker, it must
be assumed it has either,
(a) failed to trip correctly, or that
(b) fault current continues to be fed into the protected zone from a remote end.
(In either case, a minimum number of additional breakers have to be tripped
to clear the fault from the system and this is achieved by the breaker fail
protection. The scheme is dealt with at greater length in Chapter 13).

12.5.7 Standby earth-fault protection

Back-up protection against l.v. earth faults can be provided by an induction pattern
relay energised from a current transformer connected in the power transformer 1.v.
neutral. For resistance earthed systems the relay should have a time setting high
enough to discriminate with the 1.v. network protection, and be arranged to trip the
transformer in the event of a sustained l.v. fault. The relay also serves to protect the
neutral earthing resistor against the effects of a sustained fault.
52 Protection of generators and transformers

Duplicate relays may be used, the coils being connected in series. The first (Stage
1) relay should have the shorter time setting, say 5 s, and should trip the
transformer 1.v. circuit breaker. The Stage 2 relay, say 7 s should trip the h.v. circuit
breakers either directly or, where appropriate, via intertripping equipment.
Where the transformer has an individual h.v. circuit breaker, as for example, at a
double busbar substation, there is no necessity to provide discriminative tripping
between h.v. and l.v. circuit breakers and only one stage is required. On mesh
connected substations, banked transformers, and transformer feeders, it is usual to
provide two stages and discriminative tripping.

12.5.8 Tank earth-fault protection

Where banked transformers are protected by a single overall differential protection

system there is no indication given by the protection regarding which transformer is
faulty. This indication can conveniently be provided by means of a tank earth-fault
relay.
The transformer tank is lightly insulated from earth and all earthed cable
sheaths, and then bonded to earth via a single copper strap, over which is mounted
a current transformer connected to a relay. Any earth fault within the transformer
tank will produce a current in the earthing strap which operates the relay.
Faults external to the transformer and its bushings are not covered by this
device, but these can usually be located by inspection.
This is the principle of frame-leakage protection which has been applied in the
past for the earth-fault protection of distribution transformers, as an alternative to
restricted earth-fault protection.

12.5.9 Winding temperature protection

Large transformers with forced cooling are usually fitted with winding temperature
devices to detect overloading of the transformer or failure of the cooling equipment.
The winding temperature device is of the type described in Chapter 6. The bulb
is situated in a special pocket located in the flow of hot oil and is, in addition,
heated by a small heater energised from a current transformer connected to
measure the transformer winding current. The device thus indicates the top oil
temperature of the transformer plus an increment proportional to the load on the
transformer, this increment being arranged to match the difference between top oil
and winding hot spot temperatures. FuU use is made of the transformer overload
capability by arranging the thermal time constant of the equipment to be similar to
that of the transformer.
Two winding temperature instruments are generally fitted to each transformer;
each instrument is fitted with two mercury switch contacts. Operation of one
instrument is arranged to start cooling fans and pumps, and to give an alarm. The
other instrument is arranged to give the same alarm and to trip the l.v. circuit
breaker.
 Protection of generators and transformers 53

Typical settings employed are:

Instrument 1 Instrument 2
Coolers In: 90°C Alarm ll0°C
Out: 65°C
Alarm 110 °C Trip 125°C

At one time, transformers were fitted with separate oil and winding temperature
instruments. The former measured the oil temperature only which did not give a
reliable indication of hot spot temperatures because of the long thermal time
constant of the oil (approximately 10 times that of the winding).
In the case of transformers without forced cooling, an oil temperature alarm
only is sometimes fitted.

12.5.10 Gas generation and oil-surge protection

All faults within the transformer tank give rise to generation of gas, which may be.
slow for minor or incipient faults or violent in the case of heavy faults.

\ /
Buchholz relay

Transformer
tank

View of relay
mounted in position

Secondary wiring
to trip and alarm circuits

Fig. 12,5.10A Installation of Buchholz gas- and oil-actuated relay

54 Protection of generators and transformers

The generation of gas is used as a means of fault detection in the gas and oil-
operated relay which comprises one or two hinged floats, buckets, or similar
buoyant masses which are inserted in the pipework (Fig. 12.5.10A)between
transformer tank and conservator and which are normally held in equilibrium by
the oil (see Chapter 6).
The rising bubbles produced by the slow generation of gas, due to a minor fault,
pass upwards towards the conservator but are trapped in the relay chamber causing
a fall in oil level inside it. This disturbs the equilibrium of the gas float, thereby
closing its contacts which would normally be connected to give an alarm.
A heavy fault will produce a rapid generation of gas causing violent displacement
of the oil which moves the surge float system of the relay in passing to the
conservator. This will result in closure of the surge float contacts which are
arranged to trip the transformer.
To relieve the violent surging of oil which may cause splitting of the transformer
tank wall and the ejection of its bushings, transformers of small rating are fitted
with explosion vents taking the form of a 'chimney' capped by a bursting-disc
diaphragm. For transformers of larger rating it becomes impossible to accommodate
the explosion vents and maintain their earth fault clearance from the live
conductors. The solution to the problem is to fit a spring-loaded pressure relief
diaphragm in the tank wall through which surging oil may escape into the bund
surrounding the transformer. 'Qualitrol devices' as they are called are equipped with
contacts which may be used for alarm or trip initiation.
The gas- and off-operated relay gives the best possible protection against such
conditions as defective coil bolt insulation and short-circuited laminations, and
incipient failure of the main insulation. The alarm element will also operate for low
oil conditions etc., as will the trip element if the condition deteriorates.
Analysis of the gas collected in the Buchholz relay chamber may frequently
assist in the diagnosis of the type of fault, and the rate of gas generation gives an
indication of the severity.
Where the tap changer selector switches are in a separate oil compartment from
the main transformer, protection can be provided by either a separate gas-and oil-
operated relay, or by arranging the oil pipework so that the tap change compartment
is connected to the transformer conservator via the main oil-and gas-operated
relays.

12.6 Protection schemes for typical transformers

The degree of protection provided depends to a great extent upon the size and
functional importance of the unit. A further important factor is the economic
aspect. The cost of protection for power transformers tends to be proportionally
higher than the cost of protection for other items of plant.

12.6.1 Distribution transformers

With distribution transformers the economic considerations are predominant, and

 Protection of generators and transformers 55

the minimum protection is usually provided consistent with acceptable overall

performance. High speed protection is not always necessary, particularly for phase-
faults as power system stability problems have rarely to be considered.
The larger sizes of distribution transformers, say about 5 or 10 MVA, are
equipped with on-load tap changing and may have forced cooling. The smaller sizes
generally have neither and are often equipped with fuses rather than circuit
breakers.
Larger distribution transformers are protected by overcurrent and earth-fault
protection. Where fault current can be fed from h.v. and l.v. sides the overcurrent
protection is usually fitted to both sides of the transformer. Fig. 12.6.1A shows a
typical arrangement.
Larger distribution transformers may also have overall differential protection in
which the restricted earth fault protection is incorporated. The protection of such
transformers in practice may differ little from transmission transformers described
in Section 12.6.2.
h.v. I.v.
_ _ M

(2
I
h.v.
I overcurrent
relay

fault overcurrent
relay
relay

Restricted earth
fault relay
Fig. 12.6.1A Overcurrent and earth-fault protection for a delta/star distribution transformer

12.6.2 Two-winding transmission transformers

Economic considerations in the protection of transformers connected to the

132 kV, 275 kV and 400 kV grid system tend to be outweighed by the need for
high speed fault clearance, necessary for system stability reasons, and the need to
minimise fault damage.
The two-winding transmission transformer normally has the h.v. winding solidly
earthed, and the 1.v. winding resistance earthed through an interconnected-star
earthing transformer. On-load tap changing and forced cooling are always provided.
To achieve high-speed discriminative fault clearance for both phase faults and
earth-faults differential protection with h.v. and 1.v. restricted earth fault is fitted.
The differential protection usually incorporates load bias and harmonic restraint
features.
56 Protection of generators and transformers

H.V.

To [
buswires
t FI.VI C O N N E C T I O N
,,, I
PROTE,CT,ION _
" " 1
! ' I I
--I---- - L B'USBAR
I ,PROTECTION __
I
I
I
I
I
I
I H.V. O V E R C U R R E N T
t
STAGE 2
I I
I ,I, H.V. O V E R C U R R E N T
I I STAGE l
I I j ~
. I I OVERALL
i--+--1, ~
, I I I BIAS DI I"L.
I----J I H.V.

. . . . . . J I
RE i.
I L.V.
R E !.'

, WINDING
TEMPERATURE
MA|N TRANS.
BUCHHOLZ
AUX. T R A N S .
BUCHHOLZ
I
I
-r i
STANI)BY
i EA RTH i-'A U LT 2
4
I
I
I
L_ STANDBY
EARTH FAULT I
E3 I
I
I
I
I
I
I
_J
__...

E]

(.'.T. circuits
D.C. circuits

Fig. 12.6.2A Typical grid transformer protection arrangement

 Protection of generators and transformers 57

Gas- and off-actuated relays are fitted to both the main transformer and its
associated earthing (auxiliary) transformer. Duplicate winding temperature
instruments are fitted for starting the cooling, and for alarm and trip.
Back-up protection is provided by overcurrent relays and 1.v. standby earth-fault
relays (one or two stages).
A typical 132 kV grid transformer protection arrangement is shown in
Fig. 12.6.2A.
On a single transformer installation the h.v. connections are normally included in
the zone of the overall differential and balanced earth-fault protections. Where
transformers are banked it is preferable to have separate overall differential and
balanced earth-fault protection for each transformer, and a separate circulating
current system for the protection of the h.v. interconnections. This requires current
transformers in the transformer h.v. bushings.
Banked transformers will invariably be fitted with two stage standby earth fault
protection.

12.6.3 Station transformers

At most major generating stations transformers are installed to provide a supply to

the station auxiliaries from the grid system. Though their MVA rating may be
comparatively low, these station transformers are important units, and their
protection is governed by technical rather than economic considerations.
Station transformers normally have on-load tap changing; but forced cooling is
not employed. The windings are usually star/delta/star, the h.v. winding being
solidly earthed and the 1.v. winding resistance earthed.
Since these transformers are connected to the grid system, internal fault currents
can reach very high multiples of transformer rating (for example 300 times c.t.
rating), and the protection arrangements must be capable of correct operation at
these high currents.
The protection fitted to a station transformer is similar to that applied to a grid
transformer except that winding temperature indicators may be omitted.

12.6.4 Autotransformers for transmission

The use of an autotransformer rather than a two winding transformer has

economical advantages where the h.v./1.v, voltage ratio is low, since in these
circumstances the autotransformer will be smaller for a given MVA rating. Auto-
transformers are used to interconnect the British 400 kV, 275 kV and 132 kV
systems, thus making possible the installation of units of high rating in situations
where the use of a two winding transformer would introduce transporta~tion
difficulties. Both h.v. and l.v. systems have common neutral earthing arrangements
(solid earthing in the case of the 400 kV, 275 kV and 132 kV systems).
Most autotransformers are fitted with delta-connected tertiary windings (voltage
typically 13 kV) and these tertiary windings may be used for the connection of a
58 Protection of generators and transformers

" -....~ " L

- "--L

[ [ [ Stabilising
resistors
Differential
relay

Fig. 12.6.4A Circulating current protection for an autotranstormer

(Note: A l l c.t.s, have identical ratios)

synchronous compensator. The protection of synchronous compensators is

described in Chapter 14.
Large autotransformers are fitted with 'on-load' tap changing and forced cooling
equipment. The tap changing is generally carried out at the lower voltage terminals,
and in consequence the tap changer has to be fully insulated and the diverter
switches are therefore mounted on top of the lower voltage terminal bushings.
The protection arrangements for large autotransformers are similar in most
respects to the protection of two winding transformers. A simpler differential
protection scheme, however, can usually be employed. Typical of the differential
protection for an autotransformer is the circulating current system shown in
Fig. 12.6.4A. Each phase winding forms a three-ended protection zone, and the
c.t.s in the l.v., h.v. and neutral ends are connected in parallel to form a circulating
current system.
All c.t.s have the same ratio and a simple instantaneous relay is used since the
protection is not affected by either magnetising inrush currents or tap changing.
The earth connection of the tertiary winding is normally taken from inside the zone
of the circulating current protection so that operation occurs for earth faults on the
tertiary winding.
With banked transformers it is usual to provide separate circulating current
protection for each transformer using a further circulating current scheme to cover
faults on the connections between the h.v. circuit breaker and transformer h.v.
bushings. (Fig. 12.6.4B). This arrangement has the advantage of distinguishing
co-ordinating gap flash-overs and h.v. connection faults and provides basic logic for
the automatic isolation of a faulty transformer and the restoration of healthy plant.
 Protection of generators and transformers 59

400 kV 275 kV

T -_~ . . . . . . w
• -,,-t

-~. . . . . ] B.-
'
"l .... I
I I
I I
I I

'
I
I
N ~---q I
~------~

I I
I
l_--J I
- I 1 fI- -
I
I I
I
To compensati~m I f
equipment ] I' I

r-~ f---
I
I
I

I_ I
I
I
I

-__4__~ t
[ I
I
1--
I

3 3 kV

F ig. 12.6.4 B Circulating c u r r e n t and back-up p r o t e c t i o n for auto a n d d o u b l e - w o u n d

transformers
60 Protection of generators and transformers

12.7 Protection system for generator transformer units

All large modern generators are directly connected to an individual power

transformer and the combination is switched on the h.v. side of the transformer at
the main transmission voltage. A unit transformer, which supplies the generator
auxiliary system, is connected at the generator terminals.

Station
transformer
f Generator
transformer

Unit
transformer

(a)Conventional

f Generator
transformer T I

l: ] (~
~) GeneratOrtransf~~:~
ormer Station
transformer

Unit/station
transformer /,~ ~.~
Unit
transformer
~~)L ( G'}
[:]

(b) Generator circuit breaker (c) Banked station transformer

Fig. 12.7A Alternative methods of connecting generators

 Protection of generators and transformers 61

Alternative methods of connecting generators are shown in Fig. 12.7A but for
each of the arrangements shown, the principles of protection are similar. In the
following discussion the conventional arrangement shown in Fig. 12.7A(a)is
assumed.
The generator winding and transformer l.v. windings form an isolated system
which therefore requires its own neutral earth. Common methods of earthing use a
low resistance, a high resistance (possibly using a distribution transformer and
secondary loading resistor) or a resistance-loaded voltage transformer. The earth-
fault current is usually limited to 200-300 A in the case of low resistance
earthing, to 1-10 A in the case of high resistance earthing, and to a negligible value
in the case of v.t.s earthing. Care must be taken with v.t. earthing to avoid
ferroresonance effects between the voltage transformer and the transformer
winding capacitances. The virtue claimed for this method is that the generator can
be left in service with an earth fault on the winding until it is convenient to transfer
the load, and arrange a planned shut down.
Modern practice in the United Kingdom tends toward the use of high resistance
earthing set to limit earth-fault current to 10A.
The protection systems for the generator transformer unit are generally similar
in principle to those applied individually to generators and transformers. Since the
generator stator and the transformer 1.v. (delta) windings form an isolated zone of
protection for earth faults, however, a simple stator earth-fault protection scheme
can be used. The differential protection system, moreover, can be arranged to cover
both the generator and main transformer.(Figs. 12.3.13A and 12.7B).

12.7.1 Biased differential protection

When generator and transformer are solidly connected there are no problems due to
magnetising current inrush, since both the generator and its transformer are excited
gradually up to their full operational values.
Through fault conditions can result in collapse of voltage, but the subsequent
inrush due to the recovery of the voltage is far less onerous than the switching-
inrush. Inrush current conditions being relatively light, no special harmonic restraint
feature need be applied to the overall differential protection as long as the fault
setting is not too low.
A biased differential relay with a setting of 20% and a bias of 20% is generally
satisfactory; the bias is required, of course, to cover transformer on load tap
changing.
The generator-transformer overall differential protection does not include the
unit transformer for which, because of its low rating, a separate overall differential
protection is provided (Fig. 12.7A). The fault settings of the unit transformer
differential protection can thus be related to its own rating rather than the rating of
the main generator.
64 Protection of generators and transformers

12.7.2 Stator earth-fault protection

A simple earth-fault protection can be applied using a relay operated from a current
transformer in the generator neutral connection, or in the secondary winding of the
distribution transformer, or from the secondary winding of the voltage transformer
in the case of v.t. earthing.
Where a low value earthing resistance is used to limit earth-fault current to
200 to 300 A, high-speed tripping from instantaneous relays is required to minimise
damage. A setting of 10% of maximum earth-fault current is the minimum
permissible if the possibility of relay operation, due to the zero sequence currents
transmitted by generator-transformer inter winding capacitance during external h.v.
earth-fault conditions is to be avoided. The protection arrangement, therefore,
usually comprises two relays, an instantaneous relay having a setting of 10% and an
i.d.m.t, induction disc relay, having a setting of 5%.
When the generator is earthed through a high resistance (limiting the maximum
stator earth-fault current to approximately 10 A)a longer time delay is permissible.
Duplicated i.d.m.t, relays, having 5% settings, are used in such cases.
Both of the above systems of protection leave the neutral ends of the stator
windings unprotected against earth faults (typically the bottom 5.10% of the
windings) and, although earth faults near the neutral ends are unlikely to be caused
by electrical stressing, faults due to mechanical stressing cannot be ruled out and
100% winding protection is required and two systems are available.
One method injects coded a.c. current into the generator neutral connection and
monitors its amplitude, as drawn by the total system to earth capacitance (i.e. that
of the generator stator, generator transformer 1.v. and unit transformer h.v.
windings and connections). The effect of an earth fault on the windings would be
to reduce the system impedance and so increase the level of injected current and
trip the set if the deviation from the datum level exceeded a predetermined value.
A second method makes use of the third harmonic voltages normally present on
the neutral connection of the generator. A fault near the neutral end of the machine
would reduce this voltage to near zero and so identify faults in those sections of the
windings not protected by the associated 50 Hz voltage sensing protection, set to
cover the upper 90-95% of the windings. A system of filters reject unwanted
frequencies from the complementary systems, so as to provide substantial overlap
in the zones covered by the two relays.
The first of the two described methods has the advantage of enabling a check for
low resistance to be made on the machine before it is run up - a significant facility,
particularly when returning a set to service after a long shutdown or maintenance
outage.
Where v.t. earthing is used, earth-fault current is negligible and an i.d.m.t, relay
may be used to give alarm only.
It should be noted that neither the direct nor the indirect connected generators
are provided with interturn fault protection. Such faults occur rarely in modem
generators but when they do, they rapidly involve earth and are cleared by the
earth-fault protection.
 Protection of generators and transformers. 65

12.7.3 Trippingarrangements

Tripping arrangements for large generator units require particularly careful

consideration. With modern high capacity generating units both turbines and
generators are designed to a comparatively smaller frame size per MVA than earlier
units, and in consequence, the relative inertia of the principal components is much
smaller. This may result in greater risk of damage and introduces additional control
problems under conditions of heavy load rejection.
For these reasons the complete tripping of a modern generator unit is carded out
only for internal fault conditions, and for all external faults the generator back-up
protection devices are arranged to trip the h.v. circuit breaker only, any speed rise
being dealt with by the governor equipment.
For very large units it is considered necessary to minimise the risk of a set
running away in the event of the turbine emergency stop and governing valves
failing to close when the protection operates. An interlock is therefore included to
prevent the generator circuit breaker from being opened until the steam supply to
the turbine has been cut off. This interlock is applied only to those protection
circuits for which a significant increase in the damage caused by the fault is unlikely
to result from delaying the opening of the generator circuit breaker until cut off of
the steam has been detected.

12.7.4 Generator transformer overfluxing protection

It is necessary to safeguard generator step-up transformers against the risk of

damage that may be caused if they are operated at flux density levels significantly
greater than their design maximum value (typically 1.9 T). Such conditions may
occur when a unit is on load, but are more likely to arise when it is on open circuit
under automatic voltage regulator (a.v.r.) or manual excitation control, particularly
when running up to, or down from, synchronous speed.
It follows from the fundamental transformer equation E = 4 4 4 (BA).tT that
B = k.E/ffl', that is flux density B is directly proportional to induced voltage E and
inversely proportional to frequency f and turns T, so that disproportional variations
in these quantities may give rise to core overfluxing. If the core flux density
increases to a point above saturation level, the flux will no longer be contained
within the core and so will extend into other (unlaminated)parts of the
transformer and give rise to eddy current circulation.
Depending on the quantity of flux, the dimensions of the metal path and its
physical properties, losses would be generated which may be manifested in four
ways:

(a) a large increase in magnetising current

(b) an increase in winding temperature
(c) an increase in transformer noise and vibration
(a) overheating of non-laminated metal parts affected by the stray fluxes.
66 Protection of generators and transformers

(;enerator transformer

i ii

I.'ield
I
I
_~
Exciter
'
I
I
t I
V.T.

[ .
. . . . L - - ~ l"/f c',mparato.r

1
TD 1 - Time delay relay
I I 0.5 to ! second
(i) Lower excitation I
(ii) Inhibit raise excitation
(iii) Alarm
TI)2 - Time delay relay
(i) Trip excitati()n system 5 to 30 seconds.
(ii) Alarm (Inhibited when h.v.
circuit breaker closed)

Fig. 12.7C Generatortransformer overfluxing protection

Since other protection devices do not afford complete protection against all of
these conditions, overfluxing protection is provided either as an integral feature of
the a.v.r, equipment itself, or as separate relay system (Fig. 12.7C). In both systems
the action of the overfluxing protection is to initiate a reduction of excitation by
normal excitation control and, if that should fail to produce the necessary
correction, to proceed after a short time delay (typically 5-7s) to trip the generator.

12.8 Transformer feeder protection

The transformer feeder results from the need for economy in switchgear, the
transformer being connected directly to the feeder without an intervening switch.
It is used mainly to provide a bulk supply from a major switching station. A second
case is that of interconnecting two low voltage systems by means of an h.v. line,
each end of which is directly connected to a transformer.
The problems of transformer feeder protection are usually similar to those
encountered in the protection of transformers and feeders as separate units. Some
simplification in the feeder protection is often possible by virtue of the impedance
of the transformer windings, and a high-speed protection is therefore obtainable
over the whole feeder length using distance or high set overcurrent relay principles.
The application of such schemes is dependent, however upon maximum and
 Protection of generators and transformers 67

minimum fault levels, feeder and transformer impedance, etc. and must be carefully
considered at the design stage.
The complete isolation of the transformer and feeder for a fault on either almost
invariably requires the use of intertripping or fault throwing.

12.8.1 Overall protection for feeder and transformer

It is possible to protect the feeder and transformer in one zone of protection using
a differential system. This necessitates pilot wires, but has the advantage that
current transformers are not required in the transformer bushings. The fault setting
of such schemes makes adequate protection of the transformer windings very
difficult, and this together with the long operating time (due to the need to stabilise
the relay against magnetising inrush conditions by means of time delay) limits the
application of this type of scheme to the smaller transformer circuits. In all cases it
is desirable to utilise separate high-speed earth fault protection.
The general principle of overall differential protection providing protection
against phase faults is shown in Fig. 12.8.1A. The summation transformers do not
incorporate an earth-fault winding since earth-fault balance is not required. An earth
fault on the delta side will not result in the appearance of zero-sequence current
on the star side and vice versa. In order to avoid blind spots for phase faults on the
star side (which gives a 2:1:1 current distribution on the delta side) the two
sections of the summation transformer have an unequal number of turns.

J'm ....

J ' L__~

J ' L.._

0 "~ 0
Earth
fault I

Fig. 12.8.1A Voltage balance system for delta/star transformer

A separate relay system is required to give satisfactory earth-fault protection,

and a typical arrangement of phase fault differential protection with instantaneous
restricted earth-fault protection is shown in Fig. 12.8.1B. On the earthed star side
the relay is energised from a core balance transformer which has primary windings
energised from line and neutral c.t.s. In the case of the delta winding the earth-fault
relay is connected in the residual c.t. circuit.
68 Protection of generators and transformers

b ~,' ~',' r~,-~" .....

!'i

_ic
e
i| i i

1,2

0 = Operating coil
R = Restraint c o i l

Fig. 1 2 . 8 . 1 B Overafl Solkor protection for transformer feeder using two pilot wires
Differential phase-fault protection with separate earth-fault protection
(A Reyrolle & Co. Ltd, J

BO-- j

~J'l._ ¸

Pab
U.C'. L.C. Pah YO'--

Y U.C.--.--..

Diagram of translay relay

U.C - Relay Upper Coil
!..('. : Relay Lo~er Coil

Fig. 1 2 , 8 . 1 C Translay phase and earth-fault differential protection for a delta/star transformer
using three pilot wires, (GEC Measurements Ltd.)
 Protection of generators and transformers 69

Fig. 12.8.1C shows a typical overall differential protection scheme giving both
phase and earth-fault protection to a transformer feeder. In this case two separate
relay elements are used at each end of the line. One relay element is energised from
R and B current and the second from the Y phase current and this arrangement
avoids a blind spot under 2:1:1 fault conditions (see Chapter 8). The interconnec-
tion between the relays at the two ends requires the use of three pilot wires as
shown.
Both types of scheme are suitable for use with 7/.029 pilot circuits up to about
15 miles in length and have fault settings in the range of 60-200%.
The separate earth-fault protection arrangement of Fig. 12.8.1B can have a low
fault setting with high speed of operation, and it is often possible to protect the
whole of the delta winding and a considerable portion, say 90%, of the star winding
under solid earthing conditions. Some difficulty may be experienced if the earth-
fault setting on the delta side of the power transformer has to be high to prevent
operation by residual capacitance current effects under external fault conditions;
this is described more fully in the general consideration of earth-fault protection.

12.8.2 Separate protection for feeder and transformer

Schemes employing separate protection for feeder and transformer are frequently
used.

Unit feeder protection systems

The choice is between pilot wire protection systems using privately owned or
rented GPO pilot circuits, and carrier current protection usually of the phase
comparison type. In either case the transformer protection is conventional and will
embody all the features discussed in Section 12.5.
Current transformers are required in the transformer bushings for the separate
zones of feeder and transformer protection. Intertripping is also required. The use
of separate unit schemes of feeder protection permits high speed tripping at both
feeder ends, and such methods will often be used over transformer interconnector
circuits, where fault power infeed may occur at both ends.

(a) P~'lotwire feeder protection with separate transformer protection: Pilot wires
may be either rented or privately owned and the protection scheme employed will
comprise any of the standard arrangements described in Chapter 8. Severe magnetis-
ing inrush suppressors, or equivalent stabilising means, may be necessary.
Post Office pilot wire systems are applicable only to two ended feeders and not
to teed (three-ended) feeders for which distance or high-set overcurrent protection
would normally be used.

(b) Carrier current protection with separate transformer protection: Phase

comparison carder protection is often applied to longer transformer interconnector
70 Protection of generators and transformers

circuits, this being identical in every way to the standard schemes described in
Chapter 8. Teed circuits cannot usually be protected by this method.

Non-unit feeder protection systems

The high impedance of the transformer will often permit the application of a high
speed non-unit scheme of feeder protection. Of these types two are in general use,
distance protection and high set overcurrent protection. The latter is the most
economical form of protection possible for feeder transformer circuits and has all
the attendant advantages of simplicity etc. It requires careful application and must
be thoroughly investigated at the design stage.
Distance protection is particularly suitable for the protection of teed
transformer feeders. In general, however, it tends to be more expensive than pilot
wire systems for the shorter transformer feeder circuits.

(c) Distance protection with separate transformer protection: A single zone

distance protection will give high speed phase fault protection to the whole feeder,
the setting reach normally being chosen to extend to the 'middle' of the trans-
former, that is feeder impedance +50% of transformer impedance; earth-fault
protection can invariably be provided by means of an instantaneous relay which
may in certain instances be required to embody a directional feature.
The distance relay must be carefully chosen to restrict the effect of the 'blind
spot' on close up faults inherent in all directional relays. This blind spot can be
avoided completely by the use of an impedance measuring relay with overcurrent
starting feature, Fig. 12.8.2A; this arrangement being applicable as long as the
remote l.v. fault infeed to a fault behind the relay does not exceed the overcurrent
starting relay setting.
Where a high remote infeed may occur, a directional distance scheme is required,
and the use of the polarised Mho relay offers the simplest solution for phase fault
protection since it combines the functions of direction and distance measurement
(see Fig. 12.8.2B). A memory action feature is invariably provided which reduces
the effects of 'dose up' faults but is not proof against 'switched on' faults.
A simple nondirectional-earth fault relay can be arranged to provide high speed
protection in many cases, particularly where a phase current check feature is
incorporated as described in the next section. The use of a directional earth-fault
relay with a definite current setting will provide high-speed earth-fault protection
covering all exigencies, and this is usually provided with the distance protection
scheme.

(d) High-set instantaneous overcurrent and instantaneous earth-fault protection:

Overcurrent protection of the i.d.m.t, type will normally be fitted to one or both
ends of the transformer feeders, either as a main or back-up protection feature. In
many instances an additional overcurrent relay of the high set instantaneous type
may be fitted, its current setting being chosen such that the relay operates for
faults on the line but is inoperative for faults through the transformer.
 Protection o f generators and transformers 71

!
| ! L - -

__L
,~ ll

!f!i
R 0

S = Overcurrent starting relay (controls ,~peration of Z).

Z = I m p e d a n c e measuring relay. (Operating coils normally
short-circuited by contacts on S)
E = Earth fault relay (contacts in series with s to trip).
R.O. = I m p e d a n c e relay operating and restraint coils

Fig. 12.8.2A Single zone impedance protection for transformer feeder with overcurrent
starting

The simple arrangement of Fig. 12.8.2C is suitable for application to relatively

low MVA rating (high impedance) transformers connected to a high MVA (low
impedance) source. The range of application of the simple scheme can be extended
by using the modified relay connections of Fig. 12.8.2D using delta connected
auxiliary current transformers so that the relay currents are similar under three-
phase and two-phase fault conditions.
Fundamentally the problem is that the relay must operate for a fault at the h.v.
terminals of the transformer under minimum plant conditions (fault at F l in
Figs. 12.8.2C and D), and yet remain inoperative for a three-phase fault on the 1.v.
side of the transformer under conditions of maximum offset (fault at F2). For the
relay connections shown in Fig. 12.8.2D three phase fault conditions need only be
considered, and the effect of the d.c. component in the fault current can be
neglected if a transient free relay is used; such a relay is unresponsive to current
levels than 90% of its setting (see Chapter 6).
The earth-fault relay must remain inoperative for all external faults, either on
the local busbar (at F3) or on the remote side of the transformer at F2. This latter
requirement is usually easy to fulfill, since most transformers are delta connected
72 Protection of generators and transformers

t
!

!) = Directional earth fault relay

E = Earth fault relay (controlled by contacts on D . )
M = I ' o l a r i s e d M h o relay

Fig. 12.8.2B Single zone Mho protection for transformer feeder with directional earth-fault
relay

I r- f n

i. 3 - , J . ~ .

P = Phase fault protection relay

E = Earth fault protection relay

Fig. 12.8.2C High-set overcurrent and earth-fault protection for transformer feeder

on one side and do not, therefore, transform zero sequence currents. The zero
sequence infeed from the transformer star winding for a busbar fault at F3 will tend
to operate the residually connected relay E~ (Fig. 12.8.2D), but willnot operate
relay E2 because of the effect of the delta winding in eliminating zero sequence
currents from the relay. The combination of relays E~ and E2 will therefore ensure
that the relay is inoperative for all external fault conditions (assuming no remote
infeed).
 Protection of generators and transformers 73

The residually connected earth-fault relays must be provided with the correct
value of stabilising resistor to ensure that relay operation does not occur due to c.t.
saturation under transient conditions.
This type of scheme is useful for the protection of important circuits feeding
industrial loads where simplicity and economy of equipment is all important.
As a general rule, this scheme can be applied successfully to transformer feeders
of up to 45 MVA rating and frequently to 60 and 90 MVA transformers when
conditions are favourable. The lower ohmic impedances of the larger units
sometimes precludes a successful design where large variations in source infeed are
possible.
In generating stations also where stations transformers are banked, it is usual to
apply separate differential protection to each transformer using c.t.s in the
transformer h.v. bushings. The instantaneous overcurrent feature is arranged to
protect the h.v. connections between circuit breaker and transformer.
Instantaneous earth-fault protection for the transformer l.v. winding is of course
conventional practice in power transformer protection.
In the case of instantaneous feeder earth-fault protection some care is necessary
in the choice of relay setting. For example, when the feeder is connected to the
unearthed winding of a power transformer a large residual capacitance current may

' , 0 - - -

Trip
P = Phase fault p r o t e c t i o n relay
E l - Residual earth fault relay
E 2 = Earth fault check relay

Fig, 1 2 . 8 . 2 D High-set overcurrent and earth-faultprotection for transformer feeder using delta
connected auxiliary c.t.s and earth-fault check feature
74 Protection of generators and transformers

R Y B

Iyc
,.'2q
IBC

IC

[1
I C = IBC + I y c = 3 x n o r m a l charging c u r r e n t / p h a s e

Iyc

~ IC

B Y

Fig. 12.8.2E Showing the effect of residual capacitance current on an earth-fault relay under
external earth-fault conditions

flow in the feeder under external earth-fault conditions. Fig. 12.8.2E shows the
way in which this current appears in the relay circuit, and it is usual for the relay
setting to be at least twice the residual capacitance current value.

12.8.3 Intertripping

lntertripping (the tripping of remote circuit breakers) is an essential requirement

for transformer feeders to ensure complete isolation under internal fault conditions.
The more common conditions requiring intertripping are:

(a) Buchholz relay operation for a transformer winding fault which may produce
insufficient current to operate the remote protection. It will trip the local breaker
directly but the remote breaker will require intertripping.

(b) Earth faults on a feeder connected to a delta or unearthed star winding which
will normally trip the feeder circuit breaker only. The fault will remain alive from
 Protection of generators and transformers 75

-o O ,,

. _ i

|."

(a) Unearthed system after clearance fr~m ~ne end

I
G 1

(h) Lt~x~fault current infeed I.v. earth fault x~ith

resistance earthing

Large ]~ /¢~
ge __~_ i ~ -'~ I"1 l" ~l,Sul"'

i i

(c) t.o~ fault current infeed (due tc~ parallel line being
c~ut o f service)

Fig. 12.8.3A Conditions requiring in tertripping

the l.v. side of the transformer and the condition will give rise to transient over
voltage due to intermittent arcing at the point of fault (Fig. 12.8.3A). Intertripping
is required here to disconnect the fault from the remaining end.

(c) Where the fault current infeed is limited due to resistance earthing or high
source impedance (Fig. 12.8.3A(b) and (c)). The fault current magnitude may not
be sufficient to trip both ends of the feeder, and intertripping is again required.
76 Protection of generators and transformers

The various methods commonly used for intertripping are described in Chapter 8.
These may be summarised as:
(a) The use of receive relays over private pilot circuits. Surge-proofing of the
intertripping may be required.
(b) The use of coded or frequency shift signalling over rented GPO or carrier
channels.
(c) Short-circuiting the stabilising signal of a unit protection scheme from
transformer protection; this is applicable for higher fault levels.
A fourth method of achieving remote tripping, particularly applicable to
transformer feeders, is the use of fault throwing switches. This is suitable for the
clearance of transformer faults and comprises an automatic switch of special design
connected between one phase and earth on the feeder side of the transformer.
Operation of the transformer protection will trip the local circuit breaker and
operate the fault throwing switch, which applies a single phase-to-earth fault at
the transformer terminals. This fault is detected by the line protection at the
remote end, which trips its circuit breaker. The arrangement is shown in
Fig. 12.8.3 B.

,j, O'A

Fig. 12.8.3B Application of fault throwing switches showing automatic disconnectors

Where intertripping is required only for transformer faults, either fault throwing,
or pilot, or carder systems are applicable. Fault throwing has advantages where
automatic isolation followed by autoreclosing is envisaged, because it permits
positive interlocking between the fault throwing switch and automatic isolator, and
reacts directly upon the power system without placing reliance on pilot circuits.
Where intertripping is required to cover low fault infeed conditions, a pilot
scheme is essential if high speed operation is required. The use of directional
overcurrent relays on the l.v. side of the power transformer will often provide
selective clearance but more slowly. As previously mentioned, however, a biased
directional relay is required where the transformer may be required to import or
export under difficult system conditions.
Pilot or carder intertripping is invariably used for important interconnectors
and fault throwing for bulk supply transformer feeders.
 Protection of generators and transformers 77

12.8.4 Neutral displacement protection

Where a delta or unearthed star winding is connected to a transformer feeder it is

essential to ensure that this unearthed winding cannot remain energised under
system earth-fault conditions. This condition will result in danger to life and
possible hazard to the sound phases due to intermittent arcing via system earth
capacitance.
Initiation of tripping may be by means of a neutral displacement relay arranged
to detect residual voltage to earth at the transformer using either voltage trans-
former or coupling capacitors. Typical arrangements are shown in Fig. 12.8.4A. The
relay will operate for external as well as internal feeder faults and must, therefore,
be provided with a time delay to ensure discrimination.

iii iii i

r i

T±T
.J- -L. . . L Coupling

T c'''~''°''
Relay
]

.I.- -
B_ _. i-
Voltage
transformer

: I

Voltage
transformer

Fig. 12.8.4A Neutral displacement schemes

78 Protection o f generators and transformers

I Iil30C LL~

(a) Separate unit systems

'~t ~'s+]
w ith intertripping

L ~

Ii~ Ii, I
ii,~, lq~l~----
30(.' 30(."
S DI

(h) N{}n-unit protection systems (~

I '~ 'i' a- ....... 'i'

L._ -" 3
(C) (.)verall unit pr~}tection system (~ W(W~

~.o i ~o<.
I i ~.,.
RES[ ~oc
Di

IN!
I T INr
S
i ~

(d) I)istance protection

NPI) - Neutral point displacement protection

Fig. 12.8.6A Typicalschemes for transformer feeder protection

 Protection of generators and transformers 79

12.8.5 Directional overcurrent protection

The application of directional overcurrent protection to transformer circuits was

described in Section 12.5.5 which illustrated the use of the (900-45 °) arrangements
for the relay voltage circuits. This, of course, applies equally to transformer feeders
as well as to transformers.
Directional overcurrent protection is applied extensively to parallel transformer
feeder circuits where there is no normal infeed from the l.v. busbars, but where an
infeed of fault current can occur through the l.v. sides of the transformers for a
fault on the feeder. The directional overcurrent protection is often used as back up
to the intertripping equipment, since it is the only means whereby the transformer
1.v. circuit breaker can be tripped in the event of a failure of intertripping equipment.

12.8.6 Typical protection arrangements for transformer feeders

Many variations are possible in the overall protection of transformer feeders,

depending upon the transformer and its functional importance. Some typical
arrangements are shown in Figs. 12.8.6A and 12.8.6B. As with all forms of
protection application, the best scheme for a particular system arrangement is that
giving the requisite performance for minimum cost. Transformer feeder applications
depend to a great extent upon the power system parameters, that is maximum and
minimum fault levels, line and transformer impedances, etc., and perhaps in more

Fig. 12.8.6B Tvpical teed feeder transformer protection scheme

80 Protectionof generators and transformers

than any other instance require a very careful investigation to ensure optimum
design. It should be noted that in Figs. 12.8.6A and 12.8.6B directional overcurrent
(OCD) and neutral displacement (ND) protection would only be employed where
there are two or more transformer feeders in parallel.

12.9 Bibliography

Books

Protection relays application guide (GEC Measurements Ltd., 2nd edn. 1975)

Articles

'Developments in electrical protection of large generators' by J H Naylor and H D

Nunney (Journ6es Internationales d'etude des centrales electriques modemes,
AIM, Lidge, October 1974)
'Some aspects of generator back-up protection in relation to synchronous-machine
performance during h.v. system faults' by V Cook and J Rushton (Proc. lEE,
July 1972)
'Protecting transformer circuits' by J Rushton and K T Yeates (Electr. Times,
January 1962)
'Pole-slipping protection' by A Stalewski; J L H Goody-and J A Dowries (Second
International conference on developments in power system protection, IEE,
London, June 1980)
 "' Cha'pter 13

Busbar protection
by L.C.W.Frerk

13.1 History of the development of busbar protection

Up to the mid 1930s, no widescale efforts had been made to protect busbars on a
unit basis. Also there was reluctance in arranging one protective equipment to cause
simultaneous tripping of a large number of circuits.
Before the British Grid System was built in the early 1930s, many undertakings
ran isolated from adjacent ones, and so the power available for busbar faults was
often relatively small, and damage due to these faults was generally not extensive.
By the late 1930s, the British Power Systems were extensively interconnected,
with a consequent increase in fault power.
A number of busbar faults occurred about this time, but due to their relatively
slow clearance from the system by overcurrent and earth-fault relays, considerable
damage resulted, especially in indoor stations.
These faults led to efforts being made to produce busbar protection in such a
form that it could be widely applied without itself being a further hazard to the
system.
Construction of the British 275 kV supergrid system began in about 1953, by
which time standard principles of busbar protection had been adopted for outdoor
switchgear at the higher voltages. At this time the emphasis was placed on the
avoidance of unwanted operations in order to give maximum security of supply.
With the introduction of 400 kV substations in the 1960s, the transient stability
of generators became the more important consideration and this led to a change of
emphasis so that fast operating times and reliable operation would be obtained for a
fault occurring within the protected zone, which in this case would be the busbars
and switchgear.

13.2 General considerations

13.2.1 The basic philosophy of busbar protection

Busbar protection is that combination of c.t.s and relays which are used to detect
82 Busbarprotection

faults occurring within the busbar zone of a busbar substation and which initiates
tripping of all those circuit breakers, the opening of which is necessary to isolate
these faults. In Fig. 13.2.1A which shows a double busbar substation with examples
of typical circuits, the dotted line encloses the busbar zone but for simplicity c.t.s
are not shown. Where unit protection is employed there may be one or more zones
of busbar protection within the overall zone shown in Fig. 13.2.1A, the boundary
of each being the circuit breakers (or the c.t.s thereof) connecting that zone to the
transmission system, generators, transformers and to other zones of busbar
protection. The circuit breakers themselves are included in the respective zones so
that a fault within a circuit breaker causes operation of the busbar protection. This
ensures interruption of all infeeds of fault current in the same way as for faults

Feeder Feeder
I /
I
/
I I
I I
I I
• A

I 1 i ...... i i
/ / / / I
I / I
I
I I
I
I
I
I
I
I
/
1 v
/
! /--El-/
Bus section
.....
/
!,
/
!
Id?
r
I
I
I

'
I
I I
L. J

/ I

t Transformer

Generator

Fig. 13.2.1A A double busbar substation showing {dotted) the zone covered by busbar
protection

occurring o n the busbars themselves or on bus selector disconnectors, busbar

section disconnectors and the connections between these items.
Although the protection of other zones of switchgear, such as the corners of
mesh substations, often uses similar principles they are not considered as busbar
protection within the scope of this chapter.
 Busbar protection 83

13.2.2 Earth-fault protection versus phase and earth-fault protection

In some substations dividing walls, which may be part of the building structure for
indoor switchgear, or steel plates separate the three phases one from another. In
other designs of switchgear, each phase conductor is supported within an earthed
tubular chamber. Such installations are commonly referred to as 'phase-segregated
switchgear' and provided that the segregation is maintained throughout the busbar
zone, it follows that all naturally occurring busbar faults must be between one or
more phases and earth. This being the case, then, for lower voltage systems the
whole installation need only be protected against earth faults. Three-phase faults on
phase-segregated switchgear can, however, still occur if the busbars are energised in
error with safety earths connected. In these circumstances the sum of the fault
current in the three phases (which is the net current flowing to earth) would be
small and unlikely to operate the relay(s) of an earth-fault protective system. It is
for this reason that phase and earth-fault protection is applied to phase-segregated
metalclad switchgear on the supergrid system. At lower voltages the choice will be
influenced by the additional cost of the phase and earth-fault protective systems.
With open terminal switchgear it was originally considered adequate to protect
against earth faults only, because experience has shown that most faults on busbars
and switchgear had involved earth. The consequences of relying on back-up pro-
tection to clear the few phase faults that did occur were, however, sufficiently
drastic to cause a change of policy and for many years now busbar protection
responding to phase-to.phase faults as well as earth-faults has been regarded as
essential. The only exceptions occur at those 33 kV, 11 kV and 6~5 kV substations
where the justification for providing busbar protection is marginal and the simplest
possible system is usually fitted.
Non-phase-segregated metalclad switchgear should, for obvious reasons, also be
provided with busbar protection which will operate for phase faults and earth
faults.
The method of achieving earth-fault only or phase- and earth-fault protection is
described later on in this Chapter.

13.3 The clearance of busbar faults by non-unit circuit protection

13.3.1 Back-up overcurrent and earth-fault relays

For the reasons given in Chapter 8, faults cleared by back-up overcurrent and earth-
fault protection will usually be nondiscriminative, causing unnecessary splitting of
the system which may leave some sections with insufficient generation to meet the
load, leading to widespread load shedding by voltage reduction and quite possibly
the disconnection of some consumers.
Reliance on back-up protection to clear busbar faults is nowadays confined to
radial systems at the lower distribution voltages of 11 kV and 6-6 kV where the
economics are such that even the simplest form of dedicated busbar protection can
84 8usbar protection

rarely be justified. Where permitted by the substation configuration, some improve-

ment in the security of supplies can be achieved by designing the protection so that,
as a first step, relays operate to trip a limited number of circuit breakers so as to
split the busbars into sections, each with its own infeed and as a second step to trip
those infeeds still carrying fault current after a time<lelay of, say, 04s. Where a
more discriminative scheme is necessary it will usually be of the frame-earth type
described in Section 13.4.2.

13.3.2 Distance protection

From Chapter 9 it was seen that feeders fitted with distance protection have a time
characteristic as shown in Fig. 13.3A. This means that the distance protection fitted
at end A of the feeder AB will, for a fault F on the busbars at B, open the circuit
breaker at A in the Zone 2 time, which will be set between 0-4 and 1.0s.
Where all feeders connected to a busbar station are protected by 3-zone distance
protection, this can, in some circumstances, provide a limited degree of busbar
protection. If, however, some of the circuits connected to the busbars concerned
are generators, one must consider how these would be tripped for a busbar fault.
The only generator protective systems which operate for faults on the high-voltage
circuit breaker or the remote side thereof are earth-fault and/or overcurrent and
negative phase-sequence protection. These relays would all be unacceptably slow in
clearing busbar faults and, due to the decrement of the generation contribution to
the fault current, overcurrent and earth-fault relays may even fail to operate at all.
Furthermore, the negative phase-sequence protection only operates for unbalanced
faults. Therefore distance protection cannot be relied on to clear busbar faults
where generation is connected to the same busbars.

Time _ 1.0
(secs)
Zone 2
. . . . . . .

Zone I

i A

protection
equipment

Fig. 13.3A Characteristicsof distance protection

 Busbar protection 85

z c

--o, I

j-~ . . . .
-o OE)- --El--
I
I
z r Bus section I
--0- I

--E]--
Z
.,,,,,m.-..-.
,-

"-" I-" --0 ,

(~= Circuit-breakers tripped Z - D i s t a n c e protectionequipment

Fig. 13.3B Limited busbar protection by distance protection

Fig. 13.3B shows how the protection of busbars can be afforded by distance
protection where the foregoing considerations do not apply. Here the infeed to
fault F on busbar B from the four feeders is interrupted by the operation of their
distance protection at the remote ends at substations A and C.
it is also to be noted that the whole of the busbars at station B are cleared by
remote distance protection, although it is only necessary for the bus section, two
feeder circuit breakers and one transformer circuit breaker to operate to clear
fault F.
On a closely interconnected system such as the 132 kV British Grid, where
there are a number of short feeders of insufficient length to permit protecting by
distance protection, unit protection has to be used. Such unit protection remains
stable for busbar faults, and so the back-up overcurrent relays, with their slow
operating time, have to be relied upon to clear the fault.
Better discrimination may be possible by also fitting distance protection to bus
coupler and section switches and possibly generators at largo double busbar sub-
stations, but perfect discrimination could still not be guaranteed under all
circumstances.
From the foregoing, it is apparent that non-unit types of protection such as
overcurrent, earth fault and distance cannot perform adequately for busbar faults
86 Busbarprotection

on a large interconnected system. Consequently, unit forms have been developed

to ensure short clearance times and correct discrimination for such faults. The only
present day application for distance protection as a limited form of busbar protec-
tion is on distribution systems at 33 kV.

13.4 Unit systems of busbar protection for metalclad distribution switchgear

13.4.1 General considerations

The main requirements of unit protection when fitted to protect busbars are that it
must:

(a) Have a short operating time, especially where fault levels are high, in order to
minimise damage to the switchgear and to assist system stability.

(b) Be certain to operate on internal faults. Busbar faultsare rare, especially so

for indoor metal clad equipments, as they are less affected by atmospheric pol-
lution etc. Only by very careful design and regular comprehensive routine testing of
the busbar protection can the desired reliability be achieved.

(c) Remain stable during all external faults. Since many more faults occur
externally to busbars than internally, busbar protection is called upon to stabilise
many more times than to operate. The maximum value of external fault current
occurs when a fault is adjacent to the busbar and may be equal to the rupturing
capacity of the switchgear. On the other hand, the current at which the busbar
protection is required to operate on an internal fault may only be one fiftieth of
this value. The protection in this instance is thus said to require a stability factor
of at least 50.

(d) Discriminate correctly, that is decide on which section of the busbars the
fault has occurred, and then trip rapidly only those circuit breakers connected to
that section. It is sometimes necessary to trip the remote ends of some circuits, but
this depends on the location of the current transformers in the switchgear and is
described in detail later.

(e) Be immune from maloperation. Since busbar protection has to trip a large
number of circuits, it is most important that it does not do so when there is not an
actual fault on the busbar. Thus, besides requiring a high stability factor, as discussed
in item (c) the equipment and circuitry should be as far as possible immune to the
effect of faults in wiring, auxiliary switches and human errors. These additional
precautions against maloperation are described in detail later.

This Section deals with frame-earth busbar protection which is commonly applied
to metalclad switchgear at voltages up to 11 kV. At higher distribution voltages, the
 8usbar protection 87

busbar protection uses the same principle as that for transmission substations dealt
with in Section 13_5.

13.4.2 Frame earth systems

With metalclad switchgear, a very simple form of busbar earth-fault protection can
be achieved by lightly insulating from earth all the metal framework. The frame-
work is then connected to earth at only one point and a current transformer fitted
over this connection. A typical physical arrangement is shown in Fig. 13.4.2A and
the main cable gland in Figs. 13.4.2C and 13.4.2D. It is seen that there are two
earth bars, which run the length of the switchboard. One is called the switchgear
bonding bar and interconnects each cubicle or framework. Where the circuit
breakers are withdrawable, e.g. truck type switchgear, the moving portion must
also have a heavy current earth connection to the bonding bar to provide a path for
any earth fault occurring within the circuit breaker itself. The second earth bar,
known as the cable-sheath earth bar, is lightly insulated from the switchgear cubicle
or framework and provides a direct earth connection for those cable sheaths which
require to be earthed; this will not be used for every circuit because, for example,
single-end or midpoint earthing may be specified. All cables with an earthed sheath,
irrespective of the point of earthing, require an insulated cable gland to ensure that
current from an earth fault within the switchboard can only flow to earth via the
Bus.bar ( o n l y one phase s h o w n )

Switchgear
h{}nding bar

ind ica tion I nsulat ed

~'-' ~ " '-J II g ~ . d for
r"7 ~ ' t ' ~ " 7",,-I II auxiliary
~ Z..~.Z_..~q [ [ ~ Frame-earth cable
curren t t r a n s f o r m e r

' " ~ • foundation

Circuit side I ~ /~
plug/socket Withdrawable
and main circuit breaker Cable sheath I'arth
cable gland (one only shown) earth bar electrode

Fig. 13.4.2A Cubicleswitchgear arranged for frame-leakage protection

88 8usbar protection

Switchgear framework Switchgear

3 4 IF
~ R I

All cable glands Cable sheath

Local infeed I insulated from earth bar
transformer switchgear

RI R2
- - - - - I Q
o- .l?c- ~1

OIC ~ 2
===r" To trip coils
010 --3
-r-'
., ~-- 4

Fig. 1 3 . 4 . 2 B Frame~arth busbar protection without check

switchgear bonding bar and frame earth-fault current transformer while, for cable
faults, any current returning down the sheath flows directly to earth and not via the
current transformer.
The bonding bar, earth bar, the interconnection between them and all the
bonding connections referred to above must be adequate to carry the rated short-
time current of the switchgear for the specified duration, which is usually three
seconds.
Relay R1 fed from this c.t. would detect any fault to the metalwork of the
switchgear as shown in Fig. 13.4.2 B.
Relay R1 could then operate a multicontact d.c. repeat relay R2 to trip all the
circuit breakers connected to the faulted busbar as shown.
Care has to be taken to ensure that the switchgear framework has no other
earth connection than that through the current transformer. The usual concrete
 Busbar protection 89

foundations used for switchgear are adequate for insulation from the general body
of the ground provided the holding down bolts are clear of any reinforcing rods.
Breaks must be made in any ventilating pipes or earthed conduits carrying_
secondary wiring, and both main and auxiliary cables must have insulated glands.
A typical main cable gland with an insulated barrier is shown in Fig. 13.4.2C.
It is possible to check this gland insulation only by disconnecting the earth
connections, since both sides of it are normally connected to earth. For reasons of
safety, this would require an outage of the complete switchboard. A more elaborate
gland which can readily be tested without breaking the main earth connections is
shown in Fig. 13.4.2D. Here, the insulation is in two parts with a metal island layer
between. Normally, this layer is joined to the main switchgear metalwork by a link
connection. The cable gland is tested by opening this link and the insulation
measured between the island layer and the cable sheath, and also between the island
layer and the switchgear metalwork. If both insulated sections are sound, then the
cable must be insulated from the switchgear metalwork.

Switchgear
bondin~ ba
Switchgear
I rf J()l I metalwork

earth
connection

~ RI
Gland insulation

! ,
' | -l.t | ~ , ~ l" 1 Wiped joint
--"- (.'able sheath
earth bar
Cable lead sheath

Fig. 1 3 . 4 . 2 C Simpleinsulated cable gland

The elementary scheme of busbar protection shown in Fig. 13.4.2B suffers from
the disadvantage that it may operate for spurious currents through the frame-earth
or c.t., which, for example, could be produced by insulation faults on secondary
circuit wiring on the switchgear. Also, inadvertent operation of relay R1 by, say, a
mechanical blow would cause all the switches to trip. This limitation can be
overcome by fitting an additional relay R3 which is called a check relay, as shown
in Fig. 13.4.2E.
90 Busbarprotection

Switchgear
bonding bar

l"rame~ork
earth
connection

~ R I
Link
Double insulation

island layer

Cable sheath
earth bar

Fig. 13.4.2D Double insulated cable gland

If the infeed to the switchgear is from a local transformer then this relay R3 can
be fed from a c.t. on the transformer neutral connection. Only an earth fault in the
switchgear will cause operation of both R3 and R1.
If there are no suitable transformer neutrals at the station, then a core-balance
c.t. fitted around the incoming cable could be used to supply R3. If more than one
cable (or local transformer) were feeding the substation, c.t.s must be fitted around
each cable (or neutral) and paralleled to ensure operation should one of the cables
(or the transformers) be out of service.

(a) Single busbar switchboard with one section circuit breaker: So far only a
single section switchboard has been considered. If a section circuit breaker is in-
eluded, then, when an earth occurs in one section, the busbar protection can be
arranged to open only the bus section and feeder and transformer circuit breakers
of that section to isolate completely the fault from the power system. To do this,
the framework of the bus section circuit breaker must be insulated from the
adjacent switchgear on both sides as shown in Fig. 13.4.2F. The three framework
sections are then separately earthed, with a relay fed from a c.t. on each earth
connection. The check relay R4 is shown in this case fed from a c.t. on the local
transformer neutrals. The d.c. tripping circuits are so arranged that the inadvertent
operation any of one a.c. or d.c. relay will not cause circuit breaker tripping.
 Busbar protection 91

i
l I "l |
] !
i~ I I ll..
°,
i i -
•

1 2 4 RI

' !1
1_
o

I
I
Local infeed I
transformer I
R3 I
I
I
J

R1 R2

I
R4 I
@
/ o..l_c~- - - - ? . ~ 1
"-"?d.~ 2
To trip coils
,'-'~ IO ' 3

~ 4

Fig. 13.4.2E Frame-earth busbar protection with neutral check

From these circuits it will be seen that the bus section circuit breaker always
trips on a busbar fault, and that for the fault Fa in the bus section itself, every
circuit breaker connected to the busbars must be tripped. This is necessary because
the relay R2 will operate for fault Fa on the primary connection or contacts on
the side connected to the L.H. busbars as well as for a fault in a similar position on
the right-hand side, that is it cannot discriminate between the two, and so the
whole switchboard is made dead.

(b) Double busbars: In the case of switchboards fitted with double busbars,
because of the great difficulty of insulating between main and reserve busbar
casings, it is not practical to discriminate between faults on one bar and the other.
Therefore it is necessary to employ the same principle as for single busbars. It
follows that the fault F~ shown in Fig. 13.4.2G would trip all feeder switches
 92 Busbarprotection

Insulation barriers

Bus section

R2 | R3

+,--~ . . . . . . . . .

I
-~+E +++~_!~+++.
To trip circuits To trip circuits To trip circui-ts
of 1, 2, 3 and of 1, 2, 3, 4, S, 6 of 4, 5, 6 and
bus section and bus section bus section

Fig. 13.4.2F Typical frame~arth protection for two sections o f busbars

 Busbar protection 93

Insulation barriers

Bus section

.-#.,,

)
)

Fig. 13.4.2G Frame-earth protection for double busbar switchboard

selected to the left-hand side of the main bar, the bus section switch, and all
feeders selected to the reserve bar. Auxiliary switches on the busbar selectors would
be connected to achieve this.

(c) Performanceof frame-earth protection: In Fig. 13.4.2E the frame earth.fault

relay R1 and the neutral (or core-balance) check relay R3 are usually both attracted
armature types and their operating times together with R2 do not usually exceed
80 ms. It is to be noted that the check relay will operate for earth faults both on
the busbars and also for all external faults near to the station. If the installation is
of sufficient importance, a fully discriminative check scheme can be achieved by
fitting extra c.t.s to every feeder. This arrangement is described in detail in Section
13.5 and would respond only to faults at the substation.

13.5 Unit systems of busbar protection for transmission substations

13.5.1 General considerations

The requirements of unit protection as set out in Section 13.4.1 apply equally to
94 Busbarprotection

busbar protection employed at the higher switchgear voltages including that on

the supergrid system, where the unit feature is achieved by comparing the currents
in each circuit connected to the protected busbar.
A number of systems were developed by protection relay manufacturers when
the installation of unit busbar protection was found to be necessary. Most of these
employed a bias feature to achieve the required stability and it was not uncommon
to use a current summation technique to enable earth faults and phase faults to
be detected by a single phase relay. However, as confidence in the high-impedance
circulating current differential protection principle increased, it was used in the UK
to the exclusion of other systems. Although some older substations are still
equipped with one or other of the earlier designs, they are no longer fitted on new
switchgear and are therefore not described here; readers requiring information
thereon are referred to the first edition of this book and to the bibliography.
High-impedance circulating current differential protection was introduced in
.Chapter 4 dealing with protective transformers. The principle is of sufficient
importance, however, to warrant a further explanation here, beginning from the
basic principles of the circulating current system.

13.5.2 Current balance using circulating current principle

A very simple form of unit protection of a circuit can be achieved by comparing

the currents entering it with those leaving, the circuit being healthy if they are
equal, and faulty if they differ by more than a certain amount.
In order to understand the principle more easily, the application to a short
single phase feeder carrying load current will first be considered, and Fig. 13.5.2A
shows the two equal ratio c.t.s so connected to each other that their secondary
currents flow in opposite directions in the relay R connected across them.
If the c.t.s were perfect, their secondary currents would be identical in both
magnitude and phase angle, but as no two c.t.s perform absolutely alike, then let
the two almost equal c.t. secondary currents caused by the load current be Ia and It,.

End A End B | To load

Feeder current ~ m | or__fault

F1

la _I b

R e l a y c u r r e n t is p h a s o r d i f f e r e n c e (1 a - l b ) A m p e r e s

Fig. 13.5.2A Circulating current protection (external fault)

 Busbar protection 95

The instanteneous values of c.t. currents are shown during the period of time
when the half-cycle of primary current is flowing from A to B. As these flow in
opposite directions through the relay, the resulting relay current would be their
difference (Ia - Ib ).
If a fault external to feeder AB were to occur at F~ beyond end B of the feeder,
the c.t. secondary currents la and Ib would flow exactly as for the load currents
considered above, but they would be increased in value many times. For this fault
at F l, relay R must not operate and so the c.t. d i f f e r e n c e current (Ia - Ib) flowing
through the relay must not exceed the value of the current needed to operate the
relay. This condition must remain satisfied up to the largest values of through
current that can flow into any fault just beyond end A or end B.

End A End B

t F2

la ~ la/~ R
t1' lb

Relay current is phasor sum (1a + Ib) Amperes

Fig. 13.5.2B Circulating current protection (internal fault)

Fig. 13.5.2B shows a distribution of instantaneous values of current when a

fault occurs inside the protected circuit at F2. Here it is seen that the relay now
receives the s u m of the c.t. secondary currents and the relay must be designed to
operate rapidly for this value of current. If no infeed to the fault existed at end B
then the relay current would be only I a and this must also positively operate the
relay down to the lowest value of fault current likely to exist on the primary
system, that is during minimum plant conditions.

13.5.3 Connections for circulating current busbar protection

For busbar protection the application of the circulating current principle is shown
in Fig. 13.5.3A for a busbar with four circuits connected to it. The c.t.s for this
protection are shown fitted on the feeder side of the circuit breakers so that the
protected zone includes all circuit breakers as well as the busbars.
The single phase system is still considered for simplicity, and the c.t.s are con-
nected in parallel with the relay R. For the fault at F1 on feeder Dan assumed primary
fault current distribution is shown, from which it will be seen that the summation
of the c.t. secondary currents results in zero current in the relay R and therefore
the busbar protection does not operate.
96 Busbarprotection

, ,,

,ooo/, ( " I !
~
T4,o
1000 A

T 250 A

.-" 1"1(1500 A)
Relay current = 1.0 - 0 . 2 5 - 0 . 4 5 - 0 . 3 - 0 A

Fig. 13.5.3A Circulating current busbar protection stabilising on an external fault

The fault in this case would be cleared by the feeder protection on circuit D.
If the fault were on the busbars as shown by F2 in Fig. 13.5.3B then for the
same primary fault current values as for the external fault, the relay current will
now be 1.5 A and so it will operate to trip the circuit breakers A, B, C and D to
isolate the fault.

F2(I S00A)

C.T. ratio /I~ ! 5

I OOO/I |

Relay current = 0.25 + 0.45 + 0.3 + 0.5 = 1.5A

Fig. 13.5.3B Circulating current busbar protection operating on an internal fault

(a) Single busbar arrangement: For simplicity, the application is first considered
on a single-phase basis. In order to reduce the risk of wrong operation on through
faults, the principle of having two independent lines of defence is used, as was done
in the frame earth systems. These two separate forms of protection have both to
operate before tripping can take place, and they are shown in Fig. 13.5.3C.
One group of c.t.s (a) (b) and (c) forms a circulating current system covering
zone l and a similar group (d) (e) (O covers zone 2" these are called discriminating
zones. A third group of c.t.s made up of (g) (h) (j) (k) covers both zones 1 and 2,
that is the whole of the busbars, and is used for a check system. Thus for a fault
F in zone 1, the discriminating zone 1 relay RI will operate together with the
 8usbar protection 97

Bus section
Section 1 Section 2

Zone 1 I / \ / ~ I Zone2

system

Check system I 1

Fig. 13.5.3C Circulating current duplicate line of defence busbar protection

check system relay R3 to trip all circuit breakers in zone 1, together with the bus
section circuit breaker.
To appreciate the purpose of employing a check system consider now the
condition of a healthy system carrying normal load currents. If one of the leads
to, say, c.t. (a) were to become open-circuited, then relay RI would have a current
flowing in it due to the lack of the c.t. current (a) balancing out those of c.t.s (b)
and (c). This current in R1 may not be large enough to operate it, but when an
external fault occurs the current could well be large enough to cause R1 to pick
up. Tripping of zone 1 would, however, still be prevented by the fact that the
check system relay R3 will not have operated.
The method of using separate check and discriminating systems also covers
against the inadvertent operation of any one relay due, for example, to vibration,
or operation by hand.

(b) Double busbar arrangement: Fig. 13_5.3D shows the arrangement required to
cover duplicate busbars, a single phase system still being considered at this stage to
help in more easily demonstrating the principles involved.
Here, four discriminating relays are required, one for each main and reserve
section of busbars. If feeder A on the left-hand side of the drawing were selected to
the reserve busbar, then the auxiliary switch operated by the selector would close,
so selecting its discriminating c.t. to the reserve zone 1 discriminating system, if
feeder B were selected to the main busbar, then the reserve zone 1 discriminating
system would only have two c.t.s connected to it, namely those of feeder A and
those of the bus coupler (b). Main zone 2 would only have two c.t.s selected to
it, namely feeder B and the c.t. (c) of the bus section.
In order to make one of the two fines of defence immune from defects which
might arise in auxiliary switches, none are fitted in the check system, which there-
98 Busbarprotection

Section I Section 2

Reserve ,--r
/Reserve bar d i s c o n n e c t o r s
.... /
Bus section
I

, A
A

Main '| i .... I LJ I I

Bus coupler [ Busbar
/3 /4
selectors
I - |

l)iscr, min- (
X [!1
ating c.t. C
.j-r
Chec : c.t f -
l-
(r) •
• k (e)
I.'q eder A L - ~ Feeder B
r
Auxiliary :%witches
Res~ ve z~ne I
---,[ (2 _ (41 Re_serve z<}l 2
1
Main zor:e | M a i n zone 2
[

t
(" - ( ' h e c k zone relay

M! -Main zone I "1

M2 - M a i n z~ne 2
Discriminating relays
R! - R e s e r v e z o n e I
R2 - Reserve z~me 2

Fig. 13.5.3D Circulating current protection for double busbars (a.c. circuits)

fore covers the whole substation and does not discriminate between busbar
sections. Thus the check system has only the feeder c.t.s permanently connected
to it; additional feeders would have their c.t.s added in parallel with these two.
The c.t.s in the bus section and bus coupler are permanently associated with the
discriminating zones on the opposite side of their breakers to include the breaker
in both zones on either side of it. This means that faults on bus couplers and
sections cause the loss of two sections of busbar.
In the double busbar arrangement shown there are disconnectors 5 and 6 fitted
in the reserve busbar, but no circuit breaker. Therefore, if the former are closed,
their associated auxiliary switches (5) and (6) join the buswires of zones 1 and 3
together, and under these circumstances a fault on the reserve busbar would require
all circuits selected on it to trip. This would result from the operation of R1, R2
and C.
The d.c. tripping circuits required for the a.c. scheme in Fig. 13.5.3D are shown
in Fig. 1 3 5 3E. Further busbar selector auxiliary switches are required in these
 Busbar protection 99

circuits to ensure that only the circuits selected to a particular zone are tripped.
Since both the bus coupler and bus section must be tripped from two zones they
need two tripping relays.
The check relay contact is connected in the common negative lead of the
individual tripping relays, whilst the discriminating relay contacts are in the positive
lead. This ensures that no tripping would occur if a positive supply were
inadvertently applied to the wiring of any trip relay, as, for example, at point P on
feeder A in Fig. 13.5 3E. This could happen because the leads to the disconnector
auxiliary switches may be of considerable length and pass through several junction
boxes.

Rl
(~i M
i(~I.... :
M Discriminating

switches

Reserve z o n e I . . . . . ) _~ ~6)O _- . . . . . . . . . . Reserve

. . . .z o n e 2

_
IF
..... l_
_.
_ | .,. ,
rI A4 IA3
:
I - I - i
I I I
I '
I

I Check I

L L I.J L IJ rela~/ L
C

I
T o trip To trip T o trip T o trip
feeder A bus c o u p l e r bus s e c t i o n feeder B

A I , A 2, A 3, A 4 - Z o n e d e f e c t i v e alarm relays
T - Individual circuit tripping relays

Fig. 13.5.3E Circulating current protection for double busbars (d.c.. circuits); individual
circuit- tripping rela Its
100 Busbarprotection

2!
R]

~, (z

To trip To trip To trip To trip

feeder bus bus feeder
A coupler section B

Fig. 13.5.3F Circulating current protection for double busbars (d.c. circuits)/multicontact
tripping relays

The individual circuit-tripping relays arrangement shown in Fig. 13.5.3E has the
advantage that regular proving tests of circuit tripping can readily include proving
the tripping from the busbar protection relay T. With the alternative arrangement
of multicontact tripping relays shown in Fig. 13.5.3.F, careful consideration has
to be given before operating such relays, in order to ensure that during 'trip testing'
only the desired circuit is tripped.

(c) Protection for earth faults: If only earth-fault protection of the busbars is
required, then the c.t.s can be connected as shown in Fig. 13.5.3G where it will be
seen that only one relay element is required since the c.t.s of all phases are
paralleled together.
If the outgoing circuits are cables (either three core or three single cores), an
alternative arrangement of c.t.s is to fit a single core-balance c.t. on each circuit,
as shown in Fig. 13.5.3G.
With this form of earth fault protection, as with frame earth-fault protection
described earlier, the check feature may be by a relay fed from the sum of the
currents in the neutrals of the local transformers connected to the busbars. The
disadvantage of this arrangement however, is that this relay will operate for external
earth faults. Also, the relay will not operate if there is no neutral current due to the
local transformers being out-of-service. The most reliable check arrangement is
current balance as shown in Fig. 13.5.3C.

(d) Protection for phase and earth faults: Fig. 13.5.3H shows the principle of
such a current balance system which is, in effect, three separate systems, each
covering one phase and joined together at the star point of each set of c.t.s to save
in multicores. For simplicity, part (a) shows only two primary circuits connected
to the three-phase busbars. For the blue to ycllc 0vbusbar fault F shown, both blue
 Busbar protection 101

_-r

"r
I
I
I
I Single
I pole relay
I
I
J.£ Infeed 1 lnfeed
I --
I

(a) Earth-fault protection using residually connected c.t.s

i ~ t I .111
R
i Y
B

Insulated
I
=s
cable glands

Core ~,Jt
balance
c.t.s

In fee d Infeed

(b) Earth-fault p r o t e c t i o n using core balance c.t.s

Fig. 13.5.3G Protection for earth faults

102 Busbarprotection

J ~ . y

J J - j

l i • I I
I m I
Three pole relay

Infeed I In feed

(a) Phase-to-phase fault. Current distributi~n

,_ i I ~ -

u
p m~m
T
I Ill Qml R
I
I
I
I
I
I
iZ
I
Infeed infeed

(b) Earth fault. Current distribution

Fig. 13.5.3H Combined phase and earth-fault protection

 Busbar protection 103

and yellow relays will operate. In the event of a blue phase-to-earth fault, as shown
at F in part (b) of Fig. 13.5.3H then only the blue element will operate.

13.5.4 The influence of c.t. performance on through.fault stability

The circulating current principle can be extended for application to various sub-
station configurations with multiple sections of single or double busbar switchgear,
but before doing this a practical limitation to the system described so far must be
eliminated. There would be no problem with such a system if the output of the
current transformers was a faithful reproduction of the primary current at all
times. However, one must allow for the fact that fault current may, as explained
in Chapter 4, contain a transient d.c. component current which can cause saturation
of the current transformer cores and distortion of the secondary current, effects
which can only be avoided by increasing the core size substantially. Since the c.t.s
which are connected together to form a circulating current zone will each generally
be carrying different magnitudes of current and hence different magnitudes of de
component, the degree of distortion will vary from one c.t. to another, with the
result that (for an external fault) the secondary current will not sum to zero and
there will be an unbalance (spill) current in the relay. If the magnitude of the spill
current exceeds the relay setting for sufficient time, operation of the relay will
occur and the protection will be unstable. Many early applications of circulating
current protection used relays with a relatively short operating time and these
occasionally operated incorrectly for an external fault. The relays were frequently
of the attracted armature type of fairly low impedance with a nominal operating
time of about 100 ms. At this time it was not clearly understood why the relay
misbehaved in this way. It was found, however, that if a relay with a much longer
operating time was used, for example an induction disc relay it did not-operate on
external faults. It is now known that this was because it provided time for the dc
component to decay and for steady state conditions to become established.
Unfortunately, the long fault clearance times resulting from the use of such relays
were found to be unacceptable and, for a time, biased systems were considered to
be the only solution.

13.5.5 Basic principles of high-impedance circulating current busbar protection:

stability

The requirements of high-impedance circulating current protection are, in principle,

the same whether it is applied to a two-ended zone, such as a short interconnector,
or to the multi-ended zones of busbar protection. The notable differences occur in
the stability level and the increased complexity of busbar protection required to
prevent the unwanted tripping of the several circuit breakers in one zone, e.g. due
to the inadvertent operation of a single relay.
The disadvantages of circulating current protection using low impedance relays
were described earlier, the main one being through fault instability due to c.t.
 104 Busbarprotection

saturation. It was found that this could be overcome by the correct choice of c.t.s
and relay circuit components, in what has become known as high.impedance
circulating current protection.
Consider a system to protect a zone have only two circuits connected to it.
The most onerous conditions for stability arise when:

(i) Maximum fault current enters the zone through one circuit and leaves through
the other to a fault just outside the zone, as shown in Fig. 13.5.5A.

High impedance relay Stabilising resistor

ALl ~ / LIB

C.T. secondary winding ItS" ~ _...L. I L~ C.T. turns

resistance = Rct ~ . ~ _ . _ i ~!~ ratio T

IF i1.,

Fig. 13.5.5A High impedance scheme

(ii) One c.t., say that on circuit B, saturates completely due to assymmetry
of the fault current, while the other c.t. does not enter saturation at all, and
maintains its output as a faithful reproduction of the primary current.

In Chapter 4 the equivalent circuit of a c.t. was described and this can be used to
illustrate the conditions existing in the system being considered, at which point it
should be noted that the calculated performance of the protection can only be
obtained with the desired reliability if the c.t.s are of the low reactance type as
defined in BS 3938. This being the case, the primary and secondary leakage reac.
tances can be omitted from the equivalent circuit, which is shown in Fig.13.5.5B.
Because the c.t. on circuit B is saturated by the d.c. component of the primary
current its magnetising branch may be assumed, pessimistically, to have zero
impedance. Hence it is valid to insert the connection PQ. In the absence of a
socondary e an.f., the c.t. on feeder B will behave as a resistor having a value equal
to the resistance of the secondary winding.
The equivalent circuit can be further simplified to that shown in Fig. 13.5.5C by
 8usbar protection 105

|11 ~ I
ii
II
i II
'A RCT RLA RLB RCT

5gZ 2~ 3~

/~ I Zero secondary
e.m.f, due to
wit ] I saturation

Non-saturated c.t. Magnetising and

core Joss branch I Qi '1
Saturated c.t.

Fig. 13.5.5B Equivalent circuit of high-impedance scheme

making the following assumptions:

(a) For the c.t. on circuit A, the impedance of the magnetising branch and the
resistance of the core loss branch are high relative to the other paths through
which the secondary current can flow (i.e. the c.t. is performing almost
perfectly).
(b) Since, in general, all c.t.s installed for the busbar protection of a given sub-
station will be identical, their secondary winding will have the same resistance,
shown in the equivalent circuit as Rct.

The secondary current from the c.t. on circuit A will divide between the relay
circuit and the secondary winding of the c.t. on circuit B. If, as we shall subsequently
find to be the case, the relay circuit resistance is high compared with the loop
comprising the saturated c.t. and its connections, the latter can be considered to
carry the whole of the current. The resultant voltage drop appears across XY and
therefore across the relay circuit and this must be insufficient to operate the relay if

RCT+ ~ ,
' _ _
p
4.5FZ 5.5~
SR
Perfect c.t. 24A Saturated c.t.

Fig. 13.5.5C Simplified equivalent circuit

106 Busbarprotection

the protection is to remain stable. The lead burdens will nearly always differ due to
variations in the distance from the relay to each three-phase of c.t.s and are shown
as RLA and RLB.
A similar calculation must be made assuming that the c.t. on circuit A saturates
and that on B behaves as a perfect c.t. Whichever voltage is the higher is referred to
as the stability voltage of the system. (In this instance, for the values of resistance
given in Fig. 13.5.5C, the former will be higher).
The relay circuit is shown as the relay operating coil, R, and a series resistor, SR.
If the relay is of the current-operated type with adjustable settings, its resistance
must be ascertained for the setting to be employed. The selection of the current
setting is discussed in Section 13.5.6. Almost invariably, the product of setting
current and the corresponding resistance is less than the stability voltage referred
to above and the value of the resistor SR is calculated so that the product of the
relay setting current and the sum of relay resistance and stabilising resistor (SR)is
equal to, or slightly greater than, the stability voltage Vs . It is apparent from this
how the stabilising resistor became so-called.
Summarising the stability calculation, the stability voltage must be determined
from:

Vs = 1F T(R L + Rot) 13.5.1

and, for relays with current calibration,

IS (RR + RSR) ") Vs 13.5.2

while, for relays calibrated in voltage"

setting ) Vs 13.5.3

where:

fault current in primary amps

setting of relay in amps

Rct = secondary winding resistance of c.t. at 75°C

RL =RLA ,RLB = resistance of leads between relay and the c.t. of

circuits A or B (total for the loop)

R R ~-" resistance of relay coil

RSR = resistance of stabilising resistor

 Busbar protection 107

T= turns ratio of c.t.

Vs = stability voltage

Let us take, as an example, the circuit of Fig. 13.5.5B with resistances as shown,
a c.t. ratio of 1•2000 turns and a maximum fault current of 50 kA.
Substituting in eqn. 13.5.1 to obtain the stability voltage, firstly assuming
saturation of the c.t. of circuit A:
50000
Vs = (2 +2-5)
2000

= 112-5 V

and then for saturation of the c.t. of circuit B"

50000
Vs = (3 + 2-5)
2000

= 137.5 V

If the relay to be used is of a type calibrated in current and having a burden at

setting of 0.5 VA we could select one having a setting range of 20 to 80% of 1 A
and use the 0.5 A tap, at which the impedance is:

VA 0.5
-212
Is ~ 0.52

Unless the relay has a low power factor, this may be taken on the coil resistance
RR ; otherwise the resistive component, Rn should be determined.
Rearranging eqn. 13.5.3 to give RSR ,

zs
RSR ~ - - - _ Rg
Is
137.5
...... 2
0-5

273 ~2
Alternatively, a setting range of 10-40% of 0-5 A could have been chosen and the
calculation based on the use of the 20% (0-1 A) tap, in which case the stabilising
resistor should have a minimum value of (1375-RR) I2.
 108 Busbarprotection

If a relay calibrated in voltage were used in this application, it would merely be

necessary to select a setting equal to, or greater than, 137.5 V. One design of such a
relay has settings in 25 V steps from 25-175 V, and in this case the 150 V setting
would be used.

13.5.6 Basic principles of high-impedance circulating current busbar protection:

operation

One of the main requirements to ensure reliable and fast operation of this type of
protection on internal faults is that all current transformers must have a knee-point
voltage which is at least twice the setting voltage of the relay circuit. For the
example just considered, therefore, the c.t.s should have a knee-point voltage of at
least 275 V (or 300 V in the case where the relay is set to 150 V).
The maximum internal fault current will usually be the same as that for the
stability condition and if we look at the current flowing in the secondary circuit
(as shown in Fig. 13.5.6A) it is seen that the c.t. attempts to drive 25 A through the
relay circuit. Since, in one solution to the previous example, the resistance of the
relay circuit was 1375 I2 (and could be higher for a voltage calibrated relay) it
would appear that the voltage across the relay circuit could be 34-4 kV or more.
Due to the limitation imposed by its core entering saturation on each half.cycle, no
practical c.t. could produce a voltage of this value but it could produce spikes of
very high voltage around the instants of zero flux as explained in the chapter on
c.t.s and v.t.s.

A formula due to Mathews (see Bibliography) gives a reasonable approximation to

the peak voltage produced under these conditions:

Vpk = 2 ~ 13.5.4

where:

V.k is the peak value of the distorted voltage waveform

v: is the r.m.s, value of the voltage that would appear if the c.t.
did not saturate

vk is the knee-point voltage of the c.t.

For example, Vf is the voltage of 3 4 4 kV already referred to and Vk is say 275 V,

just meeting the requirement of twice the setting voltage.

Hence,

Vpk = 8.7 kV
 8usbar protection 109

To protect the c.t.s, the secondary wiring and the relay from damage due to such a
high voltage, a nonlinear resistor is connected across points XY as shown in Fig.
13.5.6A if the peak voltage would exceed 3 kV. This resistor is of the type which
reduces in value as the voltage across it increases and is selected with characteristics
which limit the voltage to between 1 kV and 2 kV. The nonlinear resistors used for
this purpose contain a largo proportion of silicon carbide or similar material manu-
factured as discs. The properties of the material, coupled with the dimensions, give
the required characteristics.

ii

[]
50 kA j k0

It,s,
Af
Non-linear
resistor
~25 A
R 50f~
! 3 2 $ £~
l.18
,, ,, I ,

Fig. 13.5.6A Connection of voltagedimiting device

The voltage is applied between the two circular faces of the disc by connections
which ensure contact with the material over the whole of each area.
The resistor current and applied voltage are related by the following expression:

V = klO 13.5.5

where/3 varies from 0.2 to 0.25 and k is controlled by the dimensions.

For nonlinear resistors used for high-impedance protection, k usually has a value
between 200 and 1000. Because the manufacturing process is such that there are
large tolerance on both g and k, it is an advantage to specify the limits of voltage at
the maximum current and also the maximum value of/3.
The relationship in eqn. 13.5.5 only applies to instantaneous values of current
and voltage and therefore if the current is sinusoidal the voltage waveform will not
be. For this and a number of other reasons (thermal rating, for example) calculations
for the selection of suitable nonlinear resistors for busbar protection should be
supported by the results of tests on a typical installation.
A necessary part of the design of high-impedance circulating current protection
is the calculation of the fault setting, sometimes referred to as primary operating
current (p.o.c.).
110 8usbar protection

An internal fault, whose magnitude is just sufficient to cause operation of the

relay, will result in a voltage between X and Y (Fig. 13.5.6A) equal to the setting
voltage of the relay circuit. The currents which, at this voltage, pass through each
component and all the c.t. magnetising and core loss branches will, when added
together and referred through the ideal c.t., equal the fault setting in primary amps.
The relevant components and parts of the c.t. equivalent circuits are shown in Fig.
13.5.6B and it will be seen that an additional resistor, Rail, has been added in
parallel with the relay circuit. This shunt setting resistor is fitted when it is required
to raise the fault setting above that which would otherwise obtain. Reasons for
doing this are given in Section 13.6.10.

[] ...~"
,,¢-
]
RLA RLB

Turns t 2_ Rs
rati~ T

,,I L I
Fig. 13.5.6B Equivalent circuit for in ternal fault condition

The primary fault setting is calculated from"

& = ( h i e + rlR + [SR "t"]NLR + IsH ) + T 13.5.6

where"

n= number of c.t.s in parallel

le = exciting current of each c.t. at relay circuit setting voltage

IR = relay current at setting

INLR = current in nonlinear resistor at relay circuit setting voltage

1Stt = shunt setting resistor current at relay circuit setting voltage

ISR = supervision relay current at main relay setting voltage

 Busbar protection 111

r = number of relays in parallel

T= current transformer turns ratio.

If the relays have a range of current settings, it is frequently possible to avoid the
need for a shunt setting resistor by selecting the relay setting to give the required
fault setting. Where, however, there is a variable number of relays in service depend-
ing on the running arrangements of the substation, for example for busbar protection
discriminating zones, it is usual to select as low a relay current setting as possible in
order that the fault setting does not vary too widely and to fit a shunt setting
resistor.
It is usual to assume that:

(i) le is in phase with the resistive currents

(ii) INLR is sinusoidal with a peak value calculated from eqn. 13.5.5.

These assumptions simplify the calculation and tend to give a slightly pessimistic
result, that is a higher setting than will be found by injecting current through one
c.t. and increasing the current until the relay operation. The calculated setting is
adequate for most purposes. The c.t. exciting current is usually obtained from the
manufacturer's estimated curve but can be obtained from test certificates or
measurements made on site.

13.5.7 Extension of the basic principles to busbar protection

The basic principles just described are applicable to busbar protection with few
changes.
In the stability calculation, there is a larger number of c.t.s to consider and the
lead resistance between the relay and each c.t. in turn must be considered in order
to find the circuit which gives the highest stability voltage. If the c.t.s do not all
have the same secondary resistance, this should also be taken into account; this
occurs chiefly when a substation is extended with a different design of switchgear
or separately mounted post-type c.t.s. Modern practice is to connect the c.t.s by
buswire rings as will be described later and the resistance between relay and c.t.
must take account of the parallel paths formed by the two routes provided by the
ring.
When calculating the fault setting there will also be a larger number of c.t.s in
parallel corresponding in general to the number of circuits connected to a given
zone. Also, as can be seen from Fig. 13.5.9A, the number of c.t.s in parallel at
double busbar substations can depend on the number of circuits selected to a given
busbar. With busbar protection in particular, it is now common practice to fit more
than one relay per zone and the current passed by the additional relays must be
included when calculating the fault setting. For the reasons given in Section
13.6.10, control over the fault setting is also required and once the relay type has
112 Busbarprotection

been decided and the c.t.s designed, this is achieved mainly by the selection of the
value for the shunt resistor.

13.5.8 Types of high-impedance relays

A common form of relay used is the attracted armature type whose current setting
may be around 30 mA, and it has a relatively high stabilising resistor connected in
series with it to give it the voltage setting determined by the stability requirements.
Its operating time will be less than 60 ms at twice the setting current.
Another form of attracted armature relay circuit incorporates a capacitor in
series with the operating coil, which renders the relay insensitive to any d.c. voltage
which may be present during the first few cycles of fault current. The relay is given
the required voltage setting by the addition of linear and nonlinear resistors, and
its circuit is shown in Fig. 13.5.8A. The use of the latter ensures high operating
speeds at small increments above the setting voltage, since the current rises more
rapidly than the voltage. With this relay it is not necessary to take into account the
possibility of almost twice the calculated sinusoidal a.c. current flowing during the
first few cycles.
Developed from this is a similar relay in which the nonlinear resistors are
replaced by ohmic resistors.
Because of the distorted waveform of the current through the relay during
stability and operation, only certain relays are suitable for high4mpedance
circulating current protection. For this reason, relays should be selected from types
which have been tested in conjunction with typical c.t.s to demonstrate that they
give the performance predicted by calculation. In those tests the current or the lead
burden will have been increased until the system becomes unstable, showing that
there is a factor of safety in the calculations varying, for typical relays, from 1-4 to
2. From the earlier calculation of stabilising resistor value, or relay voltage setting,
it appeared that the relay would be just on the point of operating for a fault of
maximum severity and with the most adverse of other conditions. However, be-
cause of the pessimistic assumptions made in the calculation, this factor of safety
ensures that the required stability is achieved without introducing other factors to
increase the setting voltage above the value given by eqn. 13.5.1.

Volts I*/5 150 125 IO0 75 50 25

C- - Relay
Plug A ~'~ A ~ ~'~ Linear °Pcer°~ing
setting
bridge ,,
,','II"i-._o
I
Non-linear resistor Capacitor I
k

Stabilising resistor
j
r- 7
Fig, 13.5.8A Relay with capacitor and nonlinear stabflising resistors
 8usbar protection 113

13.5.9 Practical hish-impedance installations

For the reasons already given, most of the high-impedance schemes which have
been installed are of the phase and earth-fault type, employing three-pole relays
instead of the single-pole relays which would be required for earth-fault only
schemes. The three-pole relay scheme has the additional advantage of enabling
lower earth-fault settings to be achieved for the same size of switchboard.
The early installations employed only one relay for each check and discriminatin8
zone, as shown in Fig. 13.5.9A. The disadvantage of busbar protection schemes
employing relays controlling the tripping of more than one circuit breaker was that
the protection tended to be taken out of service more often than was necessary,
because of the fear of causing maltripping while testing. It is now recognised that
the consequence of a busbar fault, when the busbar protection is not in service, can
be considerably more serious than that of a possible maloperation while testing. It
was from the desire to reduce the risks of maloperation to a minimum and to
eliminate the difficulties associated with taking the busbar protection out of service
for normal commissioning and maintenance purposes that the scheme of Fig. 13.5.9B
was developed for 400 kV substations. Maintenance of this type of installation,
employing separate a.c. relays for each circuit breaker, is very much easier than it
is with installations employing only one relay per zone, and can be carried out with
the protection in service. The absence of auxiliary switches in the d.c. circuits
results in a simpler scheme and therefore one which is less likely to fail to operate
when required to do so. The scheme is designed in accordance with the principles
described in Section 13.5.5 and 13.5.6, and although there are now several relays in
parallel per zone a satisfactory fault setting can still be obtained by using relays
with a low current at setting so that the total current required for their operation
is still low.
Since individual relay rooms are provided at British 400 kV double busbar
substations using open-terminal switchgear, that is to say one room adjacent to each
circuit breaker, the per circuit high-impedance relays and the associated trip relays
are accommodated therein. It will be seen that bus section and bus coupler switches
require three high-impedance relays, namely one for the check zone, and two for
discriminating zones. Each relay room is equipped with a 1 l0 V battery and this
supplies the local relays.
Since the introduction of this system with its per circuit relays, the need to
achieve reliable, fast clearance of busbar faults has assumed even greater importance.
Although relay failures have been small in number, it was considered necessary to
cater for such a situation by connecting to the a.c. buswires per zone relays, the
contacts of which are connected to d.c. buswires (now known as back-trip bus-
wires) in order to trip each circuit breaker selected to the faulty zone so providing
a second independent tripping route for each circuit breaker.
A simplified diagram, Fig. 13.5.9C shows how the per zone relays are incorpor-
ated with separate trip relays and with the use of three tripping batteries, i.e. ' l '
and '2' in the relay rooms of circuits 1 and 2, respectively, and 'STN', the station
common battery. Apart from the c.t.s and their associated bus-wirin 8 there is
 Busbar protection 115

Reserve b u s b a r
/ / I
Rl/ R3 R4
R2/
ml/ Bus section M2 /
Main I Main 2
.,,...

busbar busbar
Bus c o u p l e r

I I LI

i ?
_

?,
Check
buswires

l* "

Relay Relay I
room room_..j

Reserve zone
(rl) (r4) (r2)
buswire

(m2) n °1
I)iscrim in ating I) iscri minat ing
main ! z(~ne busx~ire main 2 zone hus~vire

N~te: (I) All c.t.s arc I/2000 turns ratio

(2) Relays are high i m p e d a n c e t y p e s h u n t e d by
n o n - l i n e a r resistors (n()t s h ( ~ n )

Fig. 13.5.9B Simplified a.c. and d.c. circuits o f busbar protection for a 400 k V substation
with individual relay rooms
 116 Busbar protection

1
2
, ~o : ~ o_...o-G>
I

Back tripping
buswires

I I I

C.T. hus~vires

STN

B T C H - B a c k t r i p c h e c k receive r e l a y
B T D - B a c k t r i p d i s c r i m i n a t i n g receive relay
C H - C h e c k z o n e high i m p e d a n c e r e l a y
I) - D i s c r i m i n a t i n g z o n e high i m p e d a n c e relay
T - T r i p p i n g relay

Fig. 13.5.9C Simplified arrangement of Mk l1400 k V busbar protection

 Busbar protection 117

duplication of the fault-detecting and trip-processing arrangements. Even the failure

of one of the d.c. supplies could not, by itself, be the cause of a circuit breaker not
tripping when required to do so for a busbar fault. This reliability is aided con-
siderably by the existence of duplicate trip coils - a standard feature of 400 kV
circuit breakers. The back-trip discriminating relays would be connected to the
buswires via auxiliary switches on the bus selector disconnectors, but these have
been omitted in this diagram to illustrate the basic principle more clearly.
The arrangement in Fig. 13.5.9C has, in England and Wales, become known as
Mk II 400 kV busbar protection and that in Fig. 13.5.9B as Mk I.
i A

1 2

! t,'l'J t'l"J 2

('H - Check zone high impedance relay

I) - Discriminating zone high impedance relay
T - Tripping relay

Fig. 13.5.9D Simplified arrangement of Mkl1275 k V busbar protection

A large part of the 275 kV system was complete before the principle of per
circuit high-impedance relays was adopted for 400 kV busbar protection so that,
although similar arguments applied, the practice of depending on per circuit relays
had become well established. This factor, coupled with the exclusive use of
common relay rooms led to the required improvements being achieved by duplica-
tion of the high4mpedance relays and the busbar protection trip relays. Since the
118 Busbarprotection

i
i """I

f:
_ ("

bSTN ~"
CH - C h e c k zone high i m p e d a n c e relay
D - D i s c r i m i n a t i n g zone high i m p e d a n c e relay
T - T r i p p i n g relay

Fig. 1 3 . 5 . 9 E Simplified arrangement of Mk.I 132 k V busbar protection

station batteries have been very reliable, this single source of d.c. was considered
adequate but duplicate supplies have been derived, where practicable, from the
battery distribution board via separately cabled and fused ring mains or radial
feeders. The simplified diagram of Mk II 275 kV busbar protection is shown in
Fig. 13.5.9D.
At 132 kV the measures requiring duplication of relays and associated circuits
are not justified and the simplified diagram of Fig. 13.5.9E shows a typical
arrangement.

13.6 Practical considerations

13.6.1 Factors affecting the position of c.t.s in busbars

Although it is common for one or more bus section circuit breakers to be installed
in the main busbar, provision for sectionalisation of the reserve busbar is, with a
few exceptions, by means of disconnectors.
To obtain correct performance of busbar protection, it is necessary to consider
at the outset where the c.t.s are to be located. Previous references to the zones of
busbar protection have implied that there is only one check zone. For the present
this assumption will be maintained. Check zone c.t.s are only required in outgoing
 Busbar protection 119

or incoming circuits since the summation of the currents in all such circuits will
show whether or not there is a fault within the overall busbar zone shown in
Fig. 13.2.1A. Bus section and bus coupler circuit breakers do not have check zone
c.t.s since the circuit breakers are only carrying current between discriminating
zones within a given switchboard. This is shown in Fig. 13.5.9A, from which it is
also seen that the discriminating c.t.s are situated on the side of the circuit breaker
remote from the zone to which their secondary winding is connected. This ensures
that the fault of F I shown in Fig. 13.6.1A, anywhere between the circuit breaker
contacts and the c.t.s on the Main 1 side of the circuit breaker, would be detected
as within the Main 1 Zone and result in the tripping of all circuit breakers selected
to the Main 1 Busbar. A fault in the position just mentioned would also be detected
by the Main 2 Zone, perhaps unnecessarily since opening the bus section circuit
breaker may clear the infeed from the Main 2 Zone. If, however, the fault is within
the circuit breaker itself, it might not be interrupted when the circuit breaker
opened and the tripping of all circuit breakers selected to Main 2 Busbar would be
essential to the interruption of the fault. Similar arguments apply for tripping both
of these sections for a fault at F2. This overlapping of the discriminating zone c.t.s
at a bus section (or bus coupler) circuit breaker is therefore necessary for the rapid
clearance of such faults, not withstanding the fact that the provision of c.t.s on
both sides of bus section and bus coupler circuit breakers is, with some types of
switchgear, an expensive item.

~ m m ..,..m ~ m m ~ m m m m m

/ /

bl~ i'2
Main ! _ ~ Main 2
hushar L~ busbar

._. ~ INN~ . Common_,tiscriminating

I)iscriminating main _1 Discriminating main 2

zone buswire- zone buswire
Fig. 13.6.1A Discriminating main zone connections at a b~s section

It will be noted from Fig. 13.6.1B that c.t.s are not required in the reserve
busbar between the Reserve 1 and Reserve 2 Zones when these are only separated
by disconnectors. In the absence of a circuit breaker between disconnectors R3 and
R4 it is necessary, when these are both closed, to connect the Reserve I Zone and
Reserve 2 Zone buswires through auxiliary switches on R1 and R2. In the absence
of such a connection, a fault at F 1 shown in Fig. 13.6.1B would result in instability
of both reserve zones since the current flowing through R3 and R4 would not
 120 Busbarprotection

appear as a secondary current in the buswires and therefore each zone would be
unbalanced by the corresponding amount, which would flow through the high-
impedance relays and any parallel paths. If only one of the disconnectors is closed,

Reserve I ~.. R~¢¢ R~ Reserve 2

busbar ' I t" ' > i busbar

- . . . . . . -[3- . . . . . .

~ 1 . !
~,'vmmon discriminating zone buswire

-- High impedance discriminating r e l a y s - - - - - -

i)iscriminating Discriminating
reserve ! z~)ne
o.._~ o.....o - reserve 2 zone
r3 r4
buswire buswire
C

Note" I)isc~mnectors R l, R2, R3 and R4 closed

Fig. 13.6.1B D i s c r i m i n a t i n g reserve zone c o n n e c t i o n s at a bus section

say R3, the section of busbar between R3 and R4 becomes part of the associated
discriminating zone.

13.6.2 Effect of c.t. location in outgoing circuits

The position where the current transformers for the circuit and the busbar protec-
tion are fitted on outgoing circuits varies according to the type of the switchgear.

(a) C.t.s overlapping the circuit breaker: Fig. 13.6.2A(a) shows the arrangement
of c.t.s fitted on both sides of the circuit breaker which is common in outdoor bulk
off types and metalclad SF6 insulated switchgear. The performance of the protec-
tion will now be considered for the faults Fz to F4.
A fault at F I is a busbar fault which should be cleared by the busbar protection.
Fault F2 is a circuit fault and should be cleared by the circuit protection. Fa is a
busbar fault but because of its position would cause both busbar and circuit
 Busbar protection 121

protection to operate and the fault will be cleared, although the circuit breaker at
the remote end of the circuit may also be tripped. Although F4 is a circuit fault, it
may be detected by both circuit and busbar protection depending on their relative
operating times. Thus, circuit breakers selected to the busbar may be opened
unnecessarily for a circuit fault. This disadvantage is acceptable in view of the low
incidence of such faults.
(b) C.t.s on the circuit side o f the circuit breaker: Fig. 13.6.2A(b) shows the
most common arrangement of current transformer location for airblast switchgear
and other designs employing post type c.t.s. Faults F~ and F2 should be correctly
clearedas before, but F3 will only cause operation of the busbar protection,
because the fault is outside the circuit protection zone. Thus, the fault may remain
fed from the remote end of the circuit. Arrangements must, therefore, be made to
cause the circuit breaker at the remote end to trip in these circumstances.

(i) Direct tripping of the remote circuit breaker(s) without introducing any time
delay is the fastest of the available alternatives but, for feeder circuits, depends on
the existence of intertripping equipment which, on some circuits, would not
otherwise be required.

_ i I | -

FI
.,¢-

1
=¢,, S
"Lj - [

(a) C.T.s o v e r l a p p i n g (b) C.T.s on (c) C.T.s on

circuit b r e a k e r circuit side b u s b a r side

Fig. 13.6.2A Effect o f c.t. location on intertripping requirements

122 Busbarprotection

(ii) Each feeder circuit can be fitted with a time-lag relay set at about 100 ms,
this being started by operation of the busbar protection having its contacts con-
nected to unstabilise unit protection or to accelerate or unblock distance protection.
This gives a measure of discrimination in that only the circuit on which the fault
exists will have its remote end circuit breaker tripped.

(iii) If circuit breaker fail protection is installed, it would be connected so that it

is initiated by the busbar protection trip relay as well as by the circuit-protection
trip relays. Circuit breaker fail protection is described later in Section 13.7, here it
is sufficient to state that any fault current infed to Fault F3 from the remote end
would, provided the initiation signal has been received, maintain a current check
relay in the operated position. If this condition still persists after a predetermined
time the circuit breakers at the remote end of the circuit would be tripped by
unstabilising unit protection or by accelerating or unblocking distance protection.
It should be noted that there is no failure of the circuit breaker in the situation
just described but the features of the circuit breaker fail protection are being
used to advantage.

(iv) An additional relay called an interlocked overcurrent relay can be fitted to

detect any power infed at F 3 after the circuit breaker has opened. When circuit
breaker fail protection is not fitted, i.e. at 132 kV and lower voltages, this method
can be applied to feeders having unit protection and to generators and transformers.
This relay can be either a three-pole, or a single pole with summation winding,
induction-disc overcurrent type having an extremely inverse time characteristic with
an operating time of about 0-3s. Its lower magnet circuit is only completed while
the busbar protection keeps operated, and so if the fault persists at F3 after the
circuit breaker opens, the busbar protection will remain operated, so permitting
the interlocked overcurrent relay to function and send an intertripping or un-
stabilising signal to the remote end of the circuit.

(v) Where grid transformers feed busbars associated with metalclad switchgear, the
busbar protection covering these busbars is arranged to trip directly both the h.v.
and the l.v. circuit breakers of the transformer so eliminating any prolonged infeed
to a fault at F3.
Two grid transformers feeding different sections or voltages of busbars may be
banked on a common h.v. circuit breaker. In this case the low voltage busbar
protection would first trip only the grid transformer l.v. circuit breaker associated
with the faulty busbar. If the fault were at F3, then the stage 2 of the standby
earth-fault protection or stage 2 of the overcurrent protection would later operate to
trip the other transformer 1.v. and the common h.v. circuit breakers.

(vi) If the remote ends of the feeders connected to the busbars are fitted with
distance protection, the infeed to the fault at F 3 will be cleared by the operation
of the distance protection on the faulty feeder in second zone time (0.4-1.0 s)
and no special arrangements are needed for 132 kV and low voltage switchgear.
 8usbar protection 123

(c) C.t.s on the busbar side o f the circuit breaker: In Fig. 13.6.2A(c) the faults
at F~ and F2 will be correctly cleared. A fault at F3 will cause the circuit protec-
tion to trip the circuit breaker, but the fault will remain fed from the busbars. The
busbar protection will not operate as F3 is outside its zone. Again, an interlocked
overcurrent relay is used, but in this case, since it is the circuit protection which
remains operated for the fault at F3, it is this protection which is used to initiate
operation of the interlocked overcurrent relay. If F3 persists for about 0.3s, the
interlocked overcurrent relay then operates the tripping relays of the protection
of the section of busbar to which the circuit is selected.

13.6.3 Multiple check zones

Until now we have considered busbar protection systems with a single check zone,
an arrangement having the advantage of relative simplicity and economy. At the
larger stations, however, there may be strong reasons for departing from the
concept of a single check zone and using instead two or more check zones. One
consideration is that the total number of circuits may be such that it is not
practicable to obtain a suitable primary fault setting from a single check zone
covering the whole substation because of the number of c.t.s and, if applicable, per
circuit relays in parallel. Furthermore, the provision of more than one check zone
leads to greater security of the busbar protection during construction, commis-
sioning and maintenance and may be specified for this reason alone.
At substations where only bus section circuit breakers or permanent breaks are
provided to sectionalise the main and the reserve busbars, the splitting of check
zones is a simple matter, only requiring the addition of check zone c.t.s either side
of the bus section circuit breakers. The overlapping zones that this creates are
shown in Fig. 13.6.3A. However, where there is neither a circuit breaker nor a
permanent break (as is the case with the reserve busbar at the majority of double-
busbar switching stations) it is necessary, in order to sectionalise the check zone,
to adhere to the following rules:

(i) The check-zone demarcation points shall in all cases coincide with discrimina-
ting zone demarcation points.

(ii) Discriminating and check-zone currents shall each be measured by separate

c.t.s at the demarcation points so as to satisfy the reasons mentioned earlier for
providing two or more check zones.

(iii) There shall be no auxiliary switches in the c.t. circuits of the check zone.

(iv) With both of the section disconnectors open in the absence of a circuit
breaker (or with the single disconnector open if only one is fitted) a fault on any
part of one busbar must result in the immediate tripping of all circuit breakers
124 Busbarprotection

Check zone I
r" . . . . . . . . . . . . . . . . . . "I
I I
I
I
I-I . . . . . . . . . . . . . "-1
I I Check zone 2
I I r- . . . . . . I- --I . . . . . . . . . . "1
I I I r- "----" - -f- --I- . . . . . . . . -1 II
I I Main zone I l I~ [ r - - I .... ~ [ ~ I M a i n ,z o n e __2 l| l
I I , I 1 s I i-.i I I I , II I I
I I t I i t j
I I Discriminating I t_ . . . . ]" "- r- - - I
I Discriminating I
I L m._ainzonel --..-- -.. L- . . . . . . I main zone 2
I I I I
II [- . l
. . . . . .
Discriminating
reserve zone I
It. . . . . .
r- . . . .
~1
. . .
II
. .
Discriminating
reserve ~.nJL2
. . -.--I
ii

I I I I
I I I ~ r-1 r~-, ,-~
i
,,A-v-~ .... I I
I I Reserve zone 1 '1' ''1 I-,-I '[ ' ' ' ' Reserve zone 2 1 ]I
, , ' L ___,__', .........
, , ' - l _, . . . . . . . . . .
I I L . . . . . . I'- I
I L. . . . . . . . . . . . . . . J I
L. . . . . . . . . . . . . . . . . . J

Fig. 13.6.3A C.T. arrangements wth section switches in both busbars

selected to that busbar but must not result in the operation of the protection or in
the tripping of circuit breakers associated with the adjacent busbar.

(v) If both (or the single) section disconnector(s) referred to in (iv) are closed, a
fault on any part of one busbar must result in the immediate tripping of all circuit
breakers selected to that section of busbar and to the section connected by the
disconnector(s) in question.

(vi) On installations having duplicate section disconnectors, when one section

disconnector is closed and the other open, a fault on any part of one busbar,
including that part connected to it by the closed disconnector, must result in the
immediate tripping of all circuit breakers connected to the faulted busbar, but must
not result in the operation of protection or in the tripping of circuit breakers
associated with the healthy section of the busbar.
An arrangement which meets all these requirements for the case of a single section
disconnector is shown in Fig. 13.6.3B. The main busbars and bus selector discon-
nectors have been omitted for clarity. A fault occurring at F I with the disconnector
open is within check zone 2 and discriminating zone R2; c.t.s on circuits through
which current flOWS to the fault will energise the check zone 2 and discriminating
reserve 2 zone a.c. buswires, operating the per zone and per circuit relays connected
thereto, so tripping all circuits connected to the reserve zone. Neither check zone 1
nor discriminating reserve 1 zone a.c. relays will operate for this fault whether the
section disconnector is open or closed, but in the latter case circuits connected to
 Jojoauuoos!p uo!=oas aAJasaa al6U!S e pue sauoz
~loaqo o ~ qj!a~ uo!joaJoad Jeqsnq A~I OOP II'~IIN j o juama6uea~e pa!/flduJ!s 81¢'9"1CI, "8!dl
,(elaa llu!dd!a.L - I
XelaJ aauP.p;)dtu! q~!q a u o z Su!leu!tu!a:)s!Q - Q
XelaJ a o u e p a d t u ! q11!t~ ~uoz ~laaq,') - H,~
Xel;~a aA!a:)~a llu!l~u!m!aos!p d!a! ~lae8 - ( ] J . 8
/~lBl;tj ~^.la:)aa ~l:)aqa d!al~l:~el] - H,~.I.8
N¢S
',r
[~.~ Et,t,I
A ,
saJ!~snq
Su!dd!al
OI C ~
I
I
r m
I
I I
I
I I I
I I I
I I I
I I I
I I
I
I
'[
I
I
1• Z | i
o,o.- G ~-o I o
I I
~o.-@ @-.o'-~ • • !
!
,ii, =i,, , i I /¢._ u
L-7,,
I
Jeqsnq Z ~hJ':lS'rIH
1 zeqsnq I a^a~sa~!
1,4
H~ H~
.,q~ l uo!~3a3oJd J e q s n R
126 Busbarprotection

F2 CH D D CH
Reserve 1 busbar (2) (R2) (Rl) (1) Reserve 2 bushar

m ~. u . - J
|
,--1/ ~--,
I I 2 I , I 2
@-o__o
@_.o_#,o I

• i 2 ,
I
I
I
I
I I
I I
I I
I
I I
I I
I
I
I
I

Back
trippifig
huswires

2 / I

I : -EJ

1--1

~1 BTCH
BTD
-
-
~STN
Back trip c h e c k receive relay
Back trip d i s c r i m i n a t i n g receive rtAay
CH - C h e c k zone high i m p e d a n c e relay
I) - D i s c r i m i n a t i n g zone high i m p e d a n c e relay
T - T r i p p i n g relay

Fig. 1 3 . 6 . 3 C Simplified arrangement o f M k . I I 400 k V busbar protection with two check

zones and duplicate reserve section disconnectors
 Busbar protection 127

reserve 1 busbar will be tripped from the zone 2 p e r z o n e relays via the back-tripping
system.
If we now consider the fault to be at F2, again with the section disconnector
open, the check 1 zone relays will operate but check 2 zone will be stable. Since the
discriminating zone c.t.s in the busbar are disconnected by the disconnector switch
their output will not be available to stabilise the reserve 2 zone and so its p e r z o n e
and p e r circuit relays will operate. The back-trip receive relays of Section 2 will be
operated by the p e r z o n e relays of check 1 zone and discriminating reserve 2 zone.
A fault in this position with the disconnector closed would be detected by check 1
zone and discriminating reserve 1 zone which would result in the direct tripping of
circuit breakers in Section 1 while those of Section 2 would be opened via the
back-tripping system.
If two disconnectors are installed between sections of busbar, the c.t. secondaries
must be connected as shown in Fig. 13.6.3C. With the left-hand section discon-
nector open and the right-hand one closed, a fault at F2 would be cleared in the
same way as has just been described for F2 in Fig. 13.6.3B. Conversely, a fault at
F1, with the left-hand section disconnector closed and the right-hand one open,
would be cleared in a similar manner, but with the back-trip receive relays of
Section 1 operated by the p e r z o n e relays of the check 1 zone and the discriminating
reserve 2 zone.
With both disconnectors closed, a fault in either position would be detected by
the check and discriminating zones within which it has occurred, resulting in the
direct tripping of circuits selected to the corresponding section of busbar, while
those of the adjacent section would have their circuit breakers opened via the back-
tripping system.

13.6.4 Busbar selector auxiliary switches

In a limited number of busbar installations it is not possible to carry out on-load

transfers of a circuit from, say, the main busbar to the reserve. In this case it is
necessary to have auxiliary switches fitted in only one of the busbar selectors
(say the main busbar), and they are connected so that when a circuit is not selected
to the main busbar, it is assumed that it is selected to the reserve, Fig. 13.6.4A
shows this. With the more usual on-load selection possible with outdoor switchgear,
the condition of both selectors simultaneously closed must be allowed for, and so
auxiliary switch contacts are fitted in each busbar selector, as shown in Fig.
13.5.9A.
The relative times of opening and closing of busbar selector contacts to their
auxiliary switches is important. When closing a busbar selector, its auxiliary switch
must close before the main contacts close or pre-arc. When opening the selector,
the primary contacts must all open before the auxiliary contacts open. The need for
this can be seen in Fig. 13.6.4B where an on-load changeover of a feeder is being
made from the main to the reserve busbars. Let the current distribution in the
primary and secondary circuits be as shown and assume that the c.t. ratio is unity.
 128 Busbarprotection

In Fig. 13.6AB, the current of three units is entering the main busbar from feeder
A, passing through the bus coupler and leaving the reserve busbar via feeder B. The
c.t. secondary currents circulate as shown in Fig. 13.6.4B(a) and both main and
reserve busbar discriminating relays remain unoperated. If now the contacts of
the reserve busbar selector of feeder A close before its auxiliary switches, then a

i 'l Main bar selector

auxiliary switches

_ _ m i , ,,,

Reserve
M R Discriminating relays

, v , i . ,. .,,, , ., ,

Fig. 13.6.4A C.T. selection by only one busbar selector

BI3
B A3

~2
Reserve bar Reserve bar

A
....>
3
J i f l
I
I
I
L
.,.,,.=

Main bar Main and reserve Main bar

zones ultimately
paralleled by
reserve busbar
selector auxiliary
(a) (b)
switches

Fig. 13.6.4B Instability due to reserve busbar disconnector contacts closing before auxiliary
switches
 Busbar protection 129

i |

p i m m .-., m w ~., m m m i ~

i I I 'I
I
I I
'
, iL ] I
I , = ., R PI I
I |
II
!
'
Discriminating
I R zone c.t.s
I ' E
I
1 % ~'i . . ,, ,,,.

I I zone
I
I R B ,
c.t.s .

-~lL.-'lk.~~t_. ~ jF,~-dr~-..~r o~
L
v

:-:l -i--il ...................... J

L
.

Fig. 1 3 . 6 . 4 C Busbarselector auxiliary switch connections

130 Busbar protection

N~te: All d i s c o n n e c t o r s ~ p e n

,:
-- Mi:~ _
I I
II /-

-_L
E a r t h c u r r e n t fr~)m -->--d~ . . . . . . . . . ~,. T,, transfornler
p o i n t of f a u l t neutral c()nnecti()ns

MI- - l. i x e d ~ r p ~ r t a b l e m a i n t e n a n c e earth

Fig. 13.6.4D Possible fault current path through c.t.s, o f earthed circuit breaker

current distribution, as shown in the Fig. 13.6.4B(b) causing instability may arise
before the auxiliary switches parallel both main and reserve discriminating zones, as
shown by the dotted lines.
By consideration of similar current distribution during opening a busbar selector,
it can be seen that instability will again occur if the auxiliary switches do not open
after the main contacts, that is, auxiliary switches must make early and break late.
Measures must be taken to ensure that, with both busbar selectors for a given
circuit open, the c.t. secondary connections are not left open-circuited. Where
per circuit high-impedance relays are fitted these would still be connected to the
c.t.s as can be seen from Fig. 13.5.9B. In other cases auxiliary switches must be
connected to short-circuit the discriminating c.t.s as shown in Fig. 13.6.4C. If they
were left open-circuited there is the risk of damage to the insulation of the
secondary winding due to, say, maintenance earths on either side allowing current
in the substation earth mat to be diverted through the c.t. primary winding during a
fault elsewhere in the substation, or even externally to it, as shown in Fig. 13.6.4D.

13.6.5 C.T. test links

To enable c.t. secondary currents to be readily measured without disturbing any

secondary wiring, c.t. test links are fitted in every set of discriminating and check
c.t.s. These can be seen in Fig. 13.6.4C. They also have a changeover position
whereby the c.t.s are disconnected from the scheme wiring and short-circuited, and
this position may be used when a feeder is out of service and having current
injected through its c.t.s for test purposes, as it permits the busbar protection to
safely remain in service to cover all the live busbars.
 Busbar protection 131

13.6.6 Precautions against maloperation of busbar protection

As described earlier, the use of duplicate lines of defence, requiring the simul-
taneous operation of separate discriminating and check systems before tripping
takes place, is one of the most important methods of avoiding maloperation of
busbar protection. Whereas maloperation of the protection for an individual circuit
normally results in the wrong tripping of one or two circuit breakers, maloperation
of busbar protection may completely shut down a whole section of busbar. Also,
busbar protection d.c. circuits are arranged so that the inadvertent operation of any
one relay will, at the most, only trip a single circuit breaker.
Reliability of busbar protection is improved by avoiding the use of auxiliary
switches in the check system. Where they occur in the discriminating system British
practice follows one of the following alternatives:

(a) When disconnector auxiliary switches have been specifically approved for
switching current transformer circuits, single contacts are used throughout.
(b) When the auxiliary switches have not received such approval, two silver-
plated switches connected in parallel are used where contacts are open when
the disconnector is open and one silver-plated switch is used where contacts
are closed when the disconnector is open.

Further precautions can be taken by fitting relays to constantly supervise the

integrity of the c.t. circuits. The present method is to fit a sensitive supervision
relay S in parallel with the main protection relay E. The system primary load
currents will cause an out-of-balance c.t. secondary current if any connection or
c.t. becomes faulty. This method depends on there being sufficiently high load
current in the primary circuit which has the defective c.t. (or wiring) to operate
relay S.
With this in mind, it is usual to specify that the primary operating current of
the supervision relay lies between 1 and 5% of the rated current of the switchgear;
for discriminating zones this current will vary with the number of circuits selected to
the zone (because of the number of c.t.s connected to the relay a.c. circuit) but the
minimum and maximum values should be within the foregoing limits. Circuits with
a low full load current relative to other circuits in the same zone present a problem
since the normal load current may, in the event of an open circuit in one of its c.t.
secondary circuits, be insufficient to operate the supervision relay. The only action
that can be taken in such circumstances is to set the supervision relay to give as low
a primary setting as possible within the limits mentioned previously.
Inevitably there will be situations where, for part of the time, the supervision
relay would not operate on the occurrence of an open circuit in the secondary
circuit of one of the c.t.s because the current flowing at the time is below the
primary setting. The risk of instability due to such an open circuit remaining
undetected is accepted in view of the low incidence of open circuits in the c.t.
secondary wiring coupled with the coincident low probability of a busbar fault.
132 Busbarprotection

The operation of the supervision relay starts a time lag relay which after, say,
3-5s brings up an alarm that the particular discriminating or check zone is defective.
Originally an auxiliary relay, which operated after the time delay, was arranged to
switch out this defective zone by short-circuiting the c.t. bus wiring and inter-
rupting the d.c. supplies. The supervision relay would always have operated for a
busbar fault, and the time lag was provided to ensure that operation of the
protection would not be interfered with.
A reassessment of the priorities led, with the advent of the 400 kV system, to
the abandonment of buswire shorting. Retaining the busbar protection in service so
that it will operate for a genuine busbar fault is now considered essential. The in-
correct operation of the relays in the faulty zone on the occurrence of a fault
external to that zone, should one happen before the open-circuit connection is
located and repaired, presents a tolerable risk particularly since it would require the
complementary zone (check or discriminating, as appropriate) to become unstable
before circuit breakers were tripped. The absence of buswire short-circuiting relays
does require special consideration of the rating of relays and associated components
as explained in Section 13.6.10.

13.6.7. Tripping and alarm circuit arrangements

The tripping sequences of the d.c. relays in busbar protection installations are
arranged such that the operation of any relay in error, or by vibration, does not
cause more than one circuit to trip. Figs. 13.5.3E and F show the two most
common arrangements. The arrangement comprising one tripping relay per circuit,
as shown in Fig. 13.5.3E, permits the trip circuit selection to be achieved in the
relay operating coil circuit instead of in the actual circuit breaker trip coil circuit
as is done in the multi.contact tripping relay method shown in Fig. 13.5.3F. In
large substations with circuit breaker trip coil currents of 30A or more, the voltage
drops in the leads may well prohibit the use of multicontact tripping relays. Trip
testing from busbar protection can also be more safely carried out from schemes
using individual tripping relays.
Whenever practicable, the d.c. wiring for busbar protection should be segregated
from other wiring to reduce the possibility of multiple inadvertent tripping.
Fig. 13.6.7A shows one possible arrangement of d.c. isolating finks fitted in the
contact circuits of the individual tripping relays on the busbar protection panel.
The advantage of the duplication of links, as shown in the figure, is that when it is
desired to make every wire dead in the busbar protection panel, as, for instance,
when making alterations or extensions to the panel, the links are removed at every
circuit relay panel. Similarly, the links are removed at the busbar protection panel if
it is desired to make the relay panel wiring dead. If secondary injection is being
carried out to prove the busbar protection relays, it is necessary to remove the links
only at the busbar protection panel. To prove the final trip testing of the circuits,
the links are then replaced one pair at a time.
 Busbar protection 133

With the increased complexity of protection on circuits at the higher system

voltages, coupled with the use of duplicate trip coils, the number of links for all
purposes has tended to become unacceptably high, both from consideration of the
space required and the appreciation by the field engineer of the purpose of them.
As a result other arrangements will frequently be found such as the fitting of links
or disconnecting points in only one wire of a pair of trip relay contact connections.

Busbar protection panel

Individual trip relays

I I

Circuit trip links

! - -i -i I I
_/
I
I
To c~ther circuit
I
relay panels
I
I
I
I

protection L ]J 'J
trip links ¥ y I
/ I S ] Trip coil

Circuit protection I Auxiliary

, ~ _~=..~trip rela~, . , ' i switch . . . . .
,,. , , ,,,=
J
Circuit relay panel
Fig. 13.6.7A Arrangement of tripping links for busbar protection

A disconnecting point is usually an integral part of a relay panel terminal block,

such as a removable plug connecting the two wires to a given way.
As described earlier, a 'busbar protection defective' alarm is given after a time
delay of 3-5 s when a sensitive supervision relay operates for a fault in the c.t.
circuits, leads, buswiring etc. The operation of intermediate manually reset d.c.
relays, if any, can also be arranged to give an alarm.
The d.c. supplies for busbar protection can be monitored by a normally
energised protection supply supervision relay which gives a 'protection supply fail'
alarm when it drops off.
On the occurrence of a busbar fault an indication is given showing exactly which
zone of busbar protection has operated, this being achieved by contacts of the d.c.
repeat relay of the check relay in series with contacts of the d.c. repeat relay of the
appropriate zone discriminating relay.
134 Busbarprotection

~I ~i',-- I~I -" ~I,,,

I ]A/2A

__i'_---~':_: " TI)

"I,~ c.t.s t -

('11 Y ~'
M
R
''
i
BT ~.-_
('H)
11"|"

('H
I)
'
•
!
- ~ii,~ --

? l J . / 2 i~

('ire uit-hrua kc r

I.ine
dist:~) ntle t: t~)r
l,
i INT I 2 INT
tJ qJr A U q~r A

Fig. 1 3 . 6 . 8 A Circuit diagram of back-tripping, busbar protection and circuit breaker fail
protection (feeder circuit)
 Busbar protection 135

(.'() m m()n e q u i p m e n l

| i.,.i
CH
-'- ~ L Bushar
M - " - - " ~ ' - ' ~ ~ I~P r ° t ~ c l i ~ ' n

'~ - - - - ' 1 - - _ ' 1 - - I "'~ ,u.~,,ir~.~

Comm,)n (LIT-- ",.J
.,._,,....,....,.,..i ,,i,,ii. Z +
.........,.....,.--. ,.,..

I
relays ~

k.
4.,
--- ('H
M ...3
R

relays ~

,
l,fl ....
,
ii i' "' ' (.'H
I)
J

' ";R A --. A c c e l e r a t i o n

BI.('('K - ( ' i r c u i t - h r e a k u r f a i l c u r r e n t c h e c k relay
l("rP. - l n t e r l ~ c k e d c u r r ~ n t t r i p relay
IN; --intertripping
"! -- T r i p relay
TI)IA/2A - T i m e dulav relays I A and 2 A ' ~ Se~.. n~I~.'
I1)IB/2B - T i m e delay r~lays II~ and 2B )
tl - Linstahilisc,

I'~r ~ t h c r aI~I~r~viati~n.~ sc~.' I![z. 1 3.6.I~B

T(" T("
I~I I 2
N~t~..' 1"~ rt.'lay.~ inl parcllul h u t
i J c.~r A
~llly c ) n e s l l ~ ) ~ n
l"ecd~.'r c i r c u i t
136 Busbarprotection

To main bus section disconnector

auxiliary switches and c.t.s
~ _ ,,

. . . . . .

Busbar
protection I SD SD 1 _ ,,, I II I

a.c. buswires 1- 0 0-----~ 0

,,,,,,, I

i
+-
:
I c.'k r.,I p,ot~tio.
,i_
,,l
i
-o
m
c !
BBF BBTR

M~tl BB1 ~ BBF

I I To c.t. wiring
AR
MI
AR
,I oi M2
1il
Main section

?IM ....
2 side

1 I

,o
~<
M

R
=

i,,I = - T ' OO- O_O C

"F. ~ " I I "~~,n'

section
BT F ..... t Iside II "
CH "-
BI-~
CH.~' •.
q

" "r'i ' I

BT
,

I I ~,~: ,~,. J
I
Cl-i
D--
- "
T T -
s"~± s-'D

~ T(.,I
II TC2
TM

T TC2
BB
BTR
TR

~TTCI ~ T TC2 _ _

Bus coupler 1

Fig. 1 3 . 6 . 8 B Circuit diagram of back tripping and per circuit busbar protection relays for bus
coupler I and bus section I-2 (400 k V)
 Busbar protection 137

I i

~ CH

M2
--R2
C o m m o n (D)

BBI.'AR - Busbar fault auxiliary relay

BBBTRTR - Busbar back trip receive trip relay
To c.t. wiring BBTR - Busbar trip relay
BTCH - Back trip check receive relay
BTI) - Back trip discriminating receive relay
CH - Check zone (buswire or
high impedance relay)
D - Discriminating zone ( b u s w i r e or
high i m p e d a n c e relay)
M - Main zone ( b u s w i r e , auxiliary
switch or high i m p e d a n c e relay)
R - Reserve zone ( b u s w i r e , auxiliary
switch ,,r high i m p e d a n c e relay)
SD - S e c t i o n d i s c o n n e c t o r (reserve)
TCI - Trip coil I
TC2 - Trip coil 2

BBI'" BBF
AR AR
MI M2

BBv
TR '1

TC I TC2
-V
Bus S e c t i o n I - 2
138 8usbar protection

13.6.8 Back-tripping

Where per circuit high-impedance relays are used, as at 400 kV substations, there
is, at a first glance, no need for d.c. buswires since each circuit breaker in a dis-
criminating zone containing a busbar fault, would receive a trip signal initiated by
its per circuit relay.
However, a facility known as back-tripping is required to fulfil the following
functions"

(i) To trip all circuit breakers which must be opened to clear a busbar fault
detected by the imgh-impedance per z o n e check and per z o n e discriminating relays
as a safeguard against the failure of a per circuit relay.

(ii) On the occurrence of a fault in a section of the reserve busbar at a substation

with more than one check zone, to trip the circuit breakers of adjacent sections
when the section disconnectors are closed.

(iii) To trip all circuit breakers connected to the section of busbars to which is
connected a generator circuit breaker of the pressurised head air-blast type in the
event of a sudden loss of air pressure. This is to guard against the consequences of
an unwanted reclosure or internal flashover of the circuit breaker which could cause
reconnection of the generator out of synchronism.

(iv) Where circuit breaker fail protection is fitted, to trip all the local circuit
breakers necessary to disconnect a circuit breaker which has failed to clear a fault
which has been detected by the appropriate protection.

A standard back-tripping system has been developed for the British 400 kV double
busbar substations. It is designed to effect the high-speed tripping of all circuit
breakers selected to a particular busbar and to adjacent busbars through busbar
selector disconnectors or busbar section disconnectors.
The back-tripping facility comprises a discriminating system and a check system
arranged so that both systems must maloperate before incorrect circuit breaker
tripping due to secondary equipment defects can occur. Both back-tripping systems
have double-pole switched initiation in order that the back-trip receive relays do
not have to meet the more difficult requirements applying to trip relays.
One back-tripping system discriminates between sections of busbar, employing
auxiliary switches on busbar section disconnectors for this purpose, and is known as
the discriminating system. The other back-tripping system is known as the check
system.
Receipt of back-tripping by the discriminating relays associated with individual
circuit breakers is qualified by auxiliary switches on the appropriate busbar selector
disconnectors in order that the required circuit breakers, and no others, are tripped.
The basic connections of the back-tripping facility are shown in Fig. 13.6.8A
and Fig. 13.6.8B.
 Busbar protection 139

13.6.9 Test facilities

Because of the importance of busbar protection, some installations have built-in

testing facilities. Current can be injected into the secondary circuits to simulate
in-zone faults, and in a few installations external faults can also be simulated.
Fig. 13.6.9A shows an arrangement in which instruments are built-in so that the
whole testing procedure can be carried out by first switching-out by the 'protection
Reserve

' T ' ' '

/l<

/M Main | /*~liU Main 2 /.i)rl¢

l)iscriminating c.l.
Check ~.
~.'.t,
A
-41:]t-

rcsl husss ire

,..

, -

('
t ( "(J iii ii ii ) i1

b
M2

(r)~ -1o(,,,~ i ,_
('"

~T |,~
I
Check

I - Main I Main " _

Rcscsve

U u.~ irc~,

Spring return
test sss itches

e- Injected corrczll
ctil-qfff
~',itcl,cs
"~ ()ulpul curreal[

u_.>
<
o o
A.(.'. su|~pJ.~

Fig. 13.6.9A Built-in test facilities for high-impedance busbar protection

 140 Busbarpro tection

cut-off" (p.c.o.) switches and then switching on the test supplies, the latter being
interlocked through the p.c.o, switch 'off' contacts. There is either a separate
cut-off switch for each discriminating zone or a single switch for cutting-off the
whole installation. Usually, indicating lamps on the busbar protection panel
operated by the p.c.o, switches show whether the protection is in or out of service,
and in large substations remote indications are also given in the control room.
With the main test switch in position 2 and all the protection cut-off switches
open, an output from the test transformer can be injected to check each dis-
criminating and check zone relay in turn by operating the appropriate test switch,
the setting current and voltage being noted on the instruments A and V. By opening
link D, the settings of individual relays can be checked. When on load the relay
spill current is read on the ammeter A by changing over the main test switch to
position 1 and link D to the dotted position. The current from the buswire under
test, M1 in this case, then flows to the test buswire, through the ammeter to return
to the common wire as shown by the dotted arrows. The output from the c.t.s can
be similarly read by changing over link B. Before the link selections, described
above, can be carried out, it is necessary for the shorting links A and C to be closed
and opened again as appropriate, to avoid open-circuiting.
To obtain the magnetisation characteristics of any c.t. the main test switch is
turned to position 2 and link B changed to its dotted position. In this case the
primary circuit must be off load.
Other facilities to help trip testing are described in the section covering tripping
circuit arrangements.

13.6.10 Fault settings

(a) Earth faults: With a resistance earthed system, the fault current will be
controlled by the value and number of neutral resistors in service. If possible one
neutral earthing resistor circuit should be connected to each section of the busbars.
The earth-fault setting of the busbar protection should be from 10-30 per cent of
the smallest expected fault current, to ensure high speed relay operation.
If it is not possible to obtain such a low relay setting, then limitations may have
to be placed on the system operating conditions, whereby the requisite number of
neutral resistors are maintained in service.
With systems having solid multiple earthing, such as the 132 and 275 kV grid
systems, the earth-fault current is comparable to the phase fault value, and so earth-
fault settings are usually not difficult.

(b) Phase faults: The value of the current which flows for two- and three-phase
busbar faults depends on the number of infeeds to the busbar and the amount of
generating plant running. The lowest value may be when the majority of the
generating plant is shut down during the night, but lower values of current may
arise if the only infeed to the busbar is a single long feeder. It may not always be
possible to obtain a setting to meet this latter condition but that may well be an
 Busbar protection 141

acceptable risk if it is known that it is a situation that will occur very rarely. Other-
wise, back-up protection will have to be provided to clear such a fault if it is not
part of the protection already required.
In general, the overall fault setting should be between 10 and 50% of the mini-
mum fault current available, a range which caters for long term changes as the
power system develops and for short term changes due to switching and the genera-
tion capacity required at any particular time. The overall fault setting is determined
by the highest of the individual check and discriminating zone fault settings.
For uniformity of application and to avoid certain problems which will be re-
ferred to shortly, the fault settings of individual protection zones should normally
meet the following requirements, always bearing in mind the foregoing comments
so that only exceptionally will the setting exceed 50% of the minimum fault current
available from the circuit providing the lowest infeed:

(a) The fault setting of the check zone and the minimum fault setting of the
individual discriminating zones shall normally be equal to approximately 50%
of the full load current rating of the associated busbar.

(b) The maximum fault setting of the individual discriminating zones shall not
normally exceed the full load rating of the associated busbar.

(c) When two or more sections of busbar are connected together via section
disconnectors, the corresponding discriminating zone buswires are parallel con-
nected through the bus section disconnector auxiliary switches. Where the combined
discriminating zone thus formed would otherwise give rise to a fault setting in
excess of the full load current rating of the associated busbar, the minimum fault
setting of the individual discriminating zones shall be reduced below the value
specified in (a)above.

Reference has been made to the reduction in primary fault setting that can result
from adopting more than one check zone. A point that must be borne in mind is
that at those substations where the reserve busbar contains no permanent breaks, the
reserve discriminating zone (if there is only one for the station) or the combined re-
serve discriminating zones(where there are section disconnectors and these are closed)
can also be high. In fact, if all circuits are selected to the reserve busbar, there will
usually be more c.t.s in parallel than would be connected to a single overall check
zone because there will be bus coupler c.t.s connected to the reserve discriminating
zones but not to the check zone, resulting in the former having a slightly higher
primary setting than the latter. Although this would appear to detract from the
advantages of employing more than one check zone, it will nearly always be the
case that, with all circuits selected to the reserve busbar, the fault level will exceed
the maximum primary fault setting of the (combined)reserve discriminating zones
by an adequate margin. With fewer circuits selected to the reserve busbar, the fault
level will be lower but so will the setting. This is in contrast to the check zone
 142 Busbarprotection

where the setting is constant and therefore must be less than 50% of the minimum
fault infeed to any one busbar section.
The abandonement of buswire short-circuiting mentioned earlier imposes
limitations on the fault setting and current rating of secondary equipment. The
secondary current flowing in the relay circuit with an open-circuit current trans-
former secondary connection is related to the primary current and therefore can
correspond to full load. The relay(s), series and shunt resistors including nonlinear
resistors must therefore be rated to carry this current continuously since it may be
an appreciable time before the faulty connections can be located and repaired; if
the fault setting is half the rated current of the largest circuit these components
must withstand twice the setting voltage continuously. This can usually be achieved
but, since the power dissipated by the resistors for the condition considered will be
very large, adequate ventilation must be provided. Where the fault setting is less
than half the rated current of the largest circuit or of the busbars, whichever is the
greater, the solution is more difficult and the installation must be designed with all
these constraints taken into account.
Ideally, the fault setting of busbar protection should be greater than the full
load current of the largest circuit, so that if a.c.t, becomes open-circuited no relay
operation would occur. This is not always possible with resi3tance earthed systems.
The sensitive supervision relays fitted to detect open-circuited c.t.s usually have a
setting of 1-5% of the switchgear rated current.
In the high-impedance circulating current scheme of busbar protection, because
the relays require only 10-30 mA, the idle c.t. magnetisation current decides the
fault setting of the scheme. Thus, for schemes covering only earth faults, where
each circuit's c.t.s are paralleled, the fault setting will be approximately three times
higher than that of a scheme covering both phase and earth faults, since in this
case only one third of the c.t.s are paralleled on to a relay element, as it now only
covers one phase.

13.6.11 Stability limits

Busbar protection schemes are designed to remain stable on external close-up

two- and three-phase faults up to the rated current breaking capacity of the switch-
gear being protected. For earth faults on multiple solidly earthed systems, at very
large substations, the earth-fault current may be even larger than for phase faults,
and due allowance must be made for this. On resistance earthed systems, particularly
where liquid earthing resistors are fitted, allowance must be made for the possibility
of their flashing over, so permitting much larger earth-fault currents to flow.

13.7 Circuit breaker fail protection

13.7.1 Principle of operation

This protection was first introduced onto the British Supergrid System in the mid
 Busbar protection 143

1970s. Its purpose is to deal with the situation in which a circuit breaker fails to
interrupt the current which it is carrying in spite of the operation of a trip relay.
Among the possible reasons are:

(i) Failure of the trip command to reach the circuit breaker trip coil. This is an
unlikely cause where duplicate trip coils are fitted and Mark II standards of
d.c. circuitry are employed to ensure tripping in the event of the failure of
any single device or supply.
0i) Failure of the circuit breaker mechanism due to an electrical or mechanical
fault.
(iii) Failure of the circuit breaker current interruption device due to a defect or
the inadvertent operation of the circuit breaker outside its limits of per-
formance.

The policy of providing two fault-detecting systems is applied to feeder protection

and, where possible, plant protection. The arrangements are such that, as far as is
practicable no single secondary electrical failure should result in an uncleared
system fault. In general, duplication ceases at the circuit breaker trip coils. The
possibility that a circuit breaker may fail to perform its function, when instructed
to do so, is higher than for the protection and trip circuits. This is due to the fact
that mechanical, pneumatic or hydraulic tripping mechanisms and interruptors can
not be duplicated for practical and economic reasons. The available methods of
system back-up protection are generally inadequate to deal with the situation which
would exist due to the failure of such components and therefore circuit breaker fail
protection has been developed for this purpose.
The basic principle of this protection is the measurement of the duration of fault
current from the instant at which any one trip relay operates to trip the circuit
breaker. If, at the end of a preselected time delay, current is still flowing, it is
considered that the circuit breaker has failed to trip and the tripping of all other
circuit breakers connected to the connections on both sides of the circuit breaker
will be initiated. At double busbar substations this is effected by means of the
back-tripping system, while at mesh and other types of substation, circuit-breaker
fail d.c. tripping circuits must be provided.
Circuit breaker fail protection is fitted to each circuit breaker and comprises two
current check relays and four timer elements (two at mesh substations).
A simplified circuit diagram for circuit breaker fail protection is shown in Fig.
13.7.1A from which certain duplicated relay elements have been omitted for
simplicity as have the back-trip check buswires and associated receive relay.
Detection of the circuit breaker fail condition is governed by the current check
relays, which are static instantaneous overcurrent relays; these are only permitted
to operate following operation of one or more of the circuit trip relays and hence
coincidentally with the energisation of the associated circuit breaker trip coils.
Although the c.t. secondary current passes through the relay whenever primary
current is flowing, the relay only operates on the application of the auxiliary d.c.
supply.
 144 Busbarprotection

T CBTC

II TD
I O.15 sec

I I I
A

,o c I ®
I "I'" BTRTR I
I I
I I I
Back trip I
I I
.d,-, • I !

I '
I toO, ,
BTD

T Back trip
c
discriminating
buswires
1_1 m .

BTRTR - Back trip receive trip relay

w BTD
TD
-
-
Back trip d i s c r i m i n a t i n g relay
T i m e delay relay
T - Trip relay
CBTC - C i r c u i t - b r e a k e r trip coil

Fig. 13.7.1A Simplified diagram of breaker fail protection for a 400 kV double busbar
substation

Assuming that the circuit breaker has failed to clear a fault following operation
of a trip relay, the following sequence of events will occur:
The d.c. circuit of the current check relay will be energised, and if the secondary
current is in excess of the setting, the relay contacts will close, in turn energising
the time delay relay from the d.c. supply to the trip coil.

At the expiration of the time delay, the current check relay is still energised by the
uninterrupted fault current. Positive and negative supplies are connected to the back-
trip discriminating buswires associated with the busbar to which the failed circuit
breaker is selected. Energisation of these bus-wires trips all other circuits connected
to the same busbar. Receipt of back-tripping signals by individual circuit breakers
is effected via the same busbar selector disconnector auxiliary switches as would be
used by that circuit for back-trip initiation.

13.7.2 Precautions against maloperation

At a double busbar substation where circuit breaker fail protection operates into
the back-tripping system the consequences of an incorrect operation are com-
parable with a maloperation of busbar protection. The following precautions are
 Busbar protection 145

therefore adopted to achieve a high security against such incidents:

(i) Two current check relays are employed and both must operate before the
time delay relays are energised and both must remain operated for the back-
trip buswires to be energised.
(a) Two time delay relays must operate before the back-trip buswires are
energised.

These points are shown in the diagram of the d.c. connections for a feeder circuit
in Fig. 13.6.8 A.

13.7.3 Current check relay settings

The setting for the current check relay must be selected so that positive operation
is ensured for faults which occur at the electrical extremity of the protected circuit
at minimum plant conditions.
Ideally, in order to obtain minimum operating times for circuit breaker fail
protection, the current check relay setting should be above the resistor current of
the circuit breaker (where resistors are fitted), so that the relay resets as soon as the
main arc is extinguished. This shortens the time allowed to elapse before the
decision is made that the circuit breaker has failed. The disadvantage of this policy
is that one possible failure mode of some designs of circuit breaker is that in which
the main contacts interrupt the fault current but the resistor contacts do not break
the resistor current, which continues to flow causing the resistors to burn out. Such
damage to the circuit breaker is a hazard to the system since it may lead to an
internal flashover to earth (either as a busbar fault or circuit fault or both) and even
if this does not occur the circuit breaker will be unlikely to interrupt subsequent
faults satisfactorily. Therefore, it is preferable that the current check relay setting
is not more than two-thirds of the resistor current (to provide a margin for reliable
operation) even though this entails the adoption of timer settings giving longer
overall operating times.
Another factor must be considered in the case of circuit breakers which switch
generators. In such installations, the relay setting must be low enough to cater for
the situation in which the circuit breaker fails to trip following the detection of a
boiler, turbine or other nonelectrical fault condition. The resultant current drawn
from the transmission system is small in the period immediately following circuit
breaker failure when there is no steam input and the excitation has been suppressed.
The setting should be sufficiently below this current (which varies with machine
size and design) to ensure relay operation.

13.7.4 Circuit breaker fail timer settings

The setting applied to the time delay relays is governed by the following factors:
(a) The minimum circuit breaker trip operating time either to main arc extinc-
146 Busbarprotection

280~

26.';
260, V HS I ' 0 l R e m o t e c.b.s.

...

p. (.'arrier
; via i n t e r t r i p

T~tal c l e a r a n c e times
> fi~r 40 nts c.b.s,
240-

220- Local c.b.s, via back trip

2 O0

180 Back trip receive trip relay , J p e r a t e d

B,ack trip receive relays & interl,~cketl c u r r e n t trip relay ~ p e r a t e d

160 ('.B. fail t i m e r ~ p e r a t i ~ n c=~mpleted - b a c k trip i n i t i a t e d

140

120
,

I00 C.B. fail current check rela~ d r ~ p - , K f " c.b. fail ti,llcr st,~pped

C.B. trip (~peration c,)mpleted

v- £
80
d •

60
-C.B. fail timer started - current check relay contacts operated
Circuit pr,~ection trip relay ,~perated - c.b. trip initiated

40 - ' - - " C i r c u i t prcJtecti,Jn operation

20 "

0 S y s t e m fault

Fig. 1 3 . 7 . 4 A Time sequence for c i r c u i t breaker fail p r o t e c t i o n - 400 k V with two-cycle

circuit breaker
 Busbar protection 147

tion or resistor contact separation, depending on the criterion adopted for setting
the current check relay. This time must be that applicable under maximum short-
circuit conditions at normal trip coil voltage and, where applicable, normal air
pressure.

(b) The minimum allowable margin of time necessary to ensure discrimination

between the correct circuit breaker tripping operation and initiation of back-tripping
(or equivalent action at mesh substations etc.) due allowance being made for the
scatter of operating times of relays and circuit breakers.

(c) The maximum fault clearance time acceptable from local back-up protection
(in this case the circuit breaker fail protection). A common figure at the present
time is 300 ms. Fig. 13.7.4A shows a typical operating time sequence for the circuit
breaker fail protection and the resultant local back-trip and remote intertrip
functions. A discriminating margin of 60 ms is typical and has been allowed in this
case, resulting in a time setting of 105 ms when, as here, the nominal operating time
of the circuit breaker is 40 ms.
The time delay relay setting is derived from the following constituent times:

(i) The minimum circuit breaker trip operating time (either to main arc extinction
or to resistor contact separation as discussed earlier) for the particular type of
circuit breaker, plus
(ii) the current check relay maximum drop.off time for the particular relay type
(assumed as 10 ms for the purpose of Fig. 13.7.4A), plus
(iii) a discriminating margin of 60 ms to allow for scatter in circuit breaker and
relay operating times, less
(iv) The current check relay minimum pick-up time for the particular relay type
(assumed as 5 ms for the purpose of Fig. 13.7.4A).

Should the estimated fault clearance time be in excess of the allowable maximum, a
reduction in the discriminating time margin may be accepted provided due con-
sideration is given to the possible variation in the operating times of the circuit
breaker and associated protective relays.

13.8 Terminology

The separate functions of busbar protection, circuit breaker fail protection and
back-tripping have been described. While, in many installations all three will exist
and will be closely related both in terms of accommodation and electrical con-
nections, care is required to avoid the use of loose descriptions which ignore the
separate duties performed by each of these facilities.
148 8usbar protection

13.9 Bibliography

Books
The protective gear handbook by F E Wellman (Pitman, 1968)
Protective relays: their theory and practice (Vol.1) by A R Van C Warrington
(Chapman & Hall, 1962)
Protective current transformers and circuits by P Mathews (Chapman & Hall,
1955)
Protective relays application guide (GEC Measurements, 1975)

Articles
'Busbar protection' by I A Reid, (Electr. Rev. June 1957)
'Instantaneous balanced current protection' by J Rushton and F E WeUman
(Metrovick Gazette, May/June 1951)
'Busbar protection' by F L Hamilton (Reyrolle Rev. Spring/Summer 1958)
'Recent developments in busbar protection' by H D Nunney (IEE Colloquium
Digest No. 1968/19, p. 69)
 Chapter 1-4
Protection of motors, reactors,
boosters and capacitors
by P.M.Dolby

14.1 Introduction

This chapter deals with the general characteristics of motors, reactors, boosters and
capacitors, with the application of such plant to a closely interconnected power
system and with the selection and application of suitable automatic protective
equipment for the plant concerned. To specify adequate protection it is necessary
to have an appreciation of the way in which the plant is constructed, its character-
istics and how it is incorporated in the system, and so a significant proportion of
the chapter is devoted to these aspects.
The main function of automatic protection is the detection of a fault condition,
and, through the opening of the appropriate circuit breakers, the disconnection of
the faulty item of plant from the remainder of the system. Coupled with this is the
need to limit to an absolute minimum the damage caused to the affected equipment.
Whilst discrimination is, without doubt, the most important requirement of any
protective system, the need for rapid operation is frequently a consideration of
almost equal importance, bearing in mind the need to minimise damage, to safe-
guard system stability and to reduce as much as possible the risk to life and limb.

14.2 Motors

14.2.1 Characteristics of d.c. and a.c. motors

D.C. Motors

D.C. motors are classified as series, shunt or compound motors depending upon
the way in which the field winding is connected with respect to the armature wind-
ing, that is the method of excitation used (see Fig. 14.2.1A). The characteristics of
the motor are determined by the method of excitation.
In a d.c. motor the armature current la adjusts itself to produce a torque balance,
so that the torque developed is equal to the total opposing torque, this opposing
torque consisting of the sum of the torque losses of the rotating armature and the
150 Protection of motors, reactors, boosters and capacitors

V V
l 4 -- -"

(a) Series excitation (b) Shunt e x c i t a t i o n

(c) Compound excitation

Fig. 14.2.1A Methods of exciting d.c. motors

torque of the load. Developed torque, as the name implies, is the torque delivered
to the shaft and available to do useful work.

If the flux q~ is sensibly constant, as with the shunt motor, la will vary as the
torque T. If • varies with I a, as with the series and compound motors, the
relationship between la and the torque is no longer linear.
The equation for the back e.m.f, induced in a d.c. armature winding is

E - CN~b

where N is the speed, • is the field flux and C is a constant for the particular
machine concerned. The back e.m.f., however, is also given by

g = V-Ia(R a +Rsf )- Vb

where V is the suppl~, voltage, usually assumed constant, Ra is the armature resist-
ance, Rsf is the resistance of any series field winding and l/'b is the brush voltage-
drop, usually of the order of two or three volts.
Since the voltage-drop in the armature circuit including any series field circuit is
usually quite small (at least for normal values of la), the back e.m.f, is sensibly con-
 Protection of motors, reactors, boosters and capacitors 151

stant and hence the product of N and • is also reasonably constant.

The developed torque is proportional to the product of the flux • and the
armature current Ia.

[-

que
E

z"

0 Ill ' 11 '

Armature current Ia

Fig. 14.2.1 B Speed and torque characteristics o f a d.c. series m o t o r

(a) The series motor. In the series motor, the flux • is produced by the arma-
ture current la, the speed and torque characteristics of the motor, plotted as func-
tions of the armature current, being as shown in Fig. 14.2.1B. It will be seen that
when la and therefore ~, is small, the speed of the motor is veryhigh. At low
values of la, ~ is directly proportional to Ia and the torque increases as the square
of the armature current. At higher values of Ia the effects of saturation and
armature reaction cause cI, to become relatively constant, the torque then increases
proportionally with the armature current. At still higher values of Ia, the
pronounced effect of armature reaction results in an actual reduction in • for an
increase in la, with the consequence that the torque then increases at a lower rate
than the armature current.
Because of its very high speed at low loads, it is important that a series motor
should never be started without load. For the same reason, a motor of this type
should not be used if there is a possibility of the entire load being lost while
running.

(b) The shunt motor: In the shunt motor, the flux • is produced by the current
If in the shunt-field winding, the current If being sensibly constant for a constant
supply voltage V. The speed and torque characteristics of the motor are shown in
Fig. 14.2.1C, the motor having a slightly drooping speed/current characteristic.
Assuming constant terminal voltage, an increase in Ia will result in increases in
armature reaction and total voltage-drop, and since these latter two quantities have
opposite effects on the motor speed, they will tend to cancel each other to give an
almost constant speed characteristic. In the normal case, the voltage-drop effect
predominates so that the speed current characteristic has the gently drooping form
shown.
It will be noted that the torque varies almost linearly with armature current,
except at the higher values of armature current where the effects of armature
reaction cause a reduction in ~.
 152 Protection of motors, reactors, boosters and capacitors

ed

i • i i l|

Armature current I a

Fig. 14.2.1C Speed and torque characteristics o f a d.c. shunt motor

A method of controlling a shunt motor is shown diagrammatically in Fig.

14.2.1D from which it will be seen that one end of the field winding is
connected direct to the supply and the other end connected via a field rheostat.
When it is required to increase the speed of the motor this is achieved by weaken-
ing the field by increasing the amount of resistance connected to it.
An alternative method of controlling a shunt motor is to employ a variable
resistance in series with the armature. The resistance must, of course, be suitable
for continuous energisation, but the method is not a particularly satisfactory one
because of the relatively heavy losses in the resistance.

O
1I _O

Field Field
winding rheostat
Fig. 14.2.1D Simple method o f speed control o f a d.c. shunt motor

(c) The compound motor: Fig. 14.2.tA shows that the compound motor is
provided with both a series and a shunt winding and that the characteristics of the
motor depend chiefly upon the relative strengths of the m.m.f.s produced by the
two field windings. If the series winding is connected so that it produces an .m.m.f.
in the same direction as the shunt winding the motor is known as a cumulatively
compounded motor; if the connection is such that the series winding m.m.f.
opposes that of the shunt winding the motor is known as a differentially com-
pounded motor. The cumulatively compounded motor may be connected in 'short-
shunt' or in 'long-shunt' as in Fig. ]4.2.] E.
 Protection of motors, reactors, boosters and capacitors 153

(a) 'Short-sh unt' connection (b) 'Long-shunt' connection

Fig. 14.2.1 E Alternative connections o f a cumulative compound motor

By adjusting the series field of the differentially compounded motor so that the
speed at full-load is equal to the speed at no-load then the speed at any inter-
mediate load will be approximately constant. This is not as advantageous as it may
seem since there is a tendency for this type of motor to start up the wrong way.
This is possible because of the initial rush of current through the series field and
armature, the shunt field taking rather longer to build up than the series field. When
both fields are fully established there may be only a relatively small nett field, and
hence torque, to drive the motor.
Another disadvantage is that if overloaded, the resulting decrease in flux will
tend to force up the speed so that the motor is overloaded still more. If the current
reaches a critical value the motor will begin to race in much the same way as a series
motor deprived of load. Thus the differentially compounded motor should be used
only in circumstances where there is no likelihood of overloading.
In the cumulatively compounded motor an increase in the load applied produces
an increase in flux and a corresponding decrease in speed, although the constant
flux produced by the shunt winding prevents this decrease being as rapid as in the
plain series motor. Thus its speed/torque characteristic will be approximately as
shown in Fig. 14.2.1F, its position relative to the characteristics of the series and

I
Differential compound

Shunt
¢)
Cumulative compound
= I Series
1
1
1
1
1
0 ,1 .....
Load current
Fig. 14.2.1F Comparison o f speed/current characteristics o f d.c. series, shunt and compound
motors
154 Protection of motors, reactors, boosters and capacitors

shunt motors depending upon the number of turns on the series winding. An
advantage of the cumulatively compounded motor is that the speed falls when a
sudden, heavy load is applied, thus helping to offset the burden of the increase in
load.

(d) Comparison of series, shunt and compound motors: A comparison of speed/

current and torque/current characteristics of the series, shunt and compound
motors is shown in Figs. 14.2.1F and 14.2.1G. It will be seen from Fig. 14.2.1G

r J//

G
Load current
Fig. 14.2.1G Comparison of torque/current characteristics of d.c. series, shunt and compound
motors
that the shunt motor characteristic lies between those of the differentially and
cumulatively compounded machines. The significant difference in characteristics
obtained by connecting the series winding of the compound motor either to assist
or to oppose the shunt winding should also be noted.
A summary of the characteristics of the three types of motor is given below.

Series motor: (j) High starting torque varying as the square of the
current at low current values.
0i) High speed at low values of load.
(iii) Decrease in speed as torque increases.

Shunt motor: (j) Practically constant speed from noqoad to full

load.
(ii) Practically constant flux and therefore torque
proportional to the armature current.

Compound motor: Differential

(i) Approximately constant speed from no-load to
full4oad.
(ii) Overloading tends to cause rapid increase in speed.
Application of motor is strictly limited.
 Protection of motors, reactors, boosters and capacitors 155

Cumulative
(i) Steeper decrease in speed with increasing torque
than in the shunt motor.
(ii) Speed/current and torque/current characteristics
lie between those of the series and shunt motors.

A. C. motors
Polyphase induction motors: Induction motors fall into two distinct categories,
namely the squirrel cage motor and the wound rotor, or slip-ring, motor. As the
name implies, both operate on the induction principle, only the primary stator
winding being connected to the supply.
The squirrel cage motor is often identified with its inherently poor starting
qualities, but it will be seen later that measures are taken to improve its perform-
ance in this respect. In a squirrel cage motor the rotor conductors are copper rods
laid in slots in the rotor core and riveted or brazed to a short-circuiting ring at each
end, giving the appearance of a cage. The absence of slip-rings ensures that sparking
will not occur and the cage construction gives a cheap and very robust motor. Its
disadvantages are low starting torque, high starting current and difficulty in provid-
ing an easy means of control.
The slip-ring motor has a 'wound' rotor, that is a group of coils forming the
windings is carried on the rotor itself. This makes it more expensive than the
squirrel cage motor, but it has the advantage that it develops the same starting
torque as a comparable squirrel cage motor with a considerably lower starting cur-
rent. Unlike the squirrel cage rotor, the total resistance of the wound motor is not
f'Lxed but can be varied by adjustment of the external resistors connected in the
rotor circuit during starting. An incidental advantage of the wound rotor is that
much of the heat generated during starting is dissipated in the resistors. Smooth
acceleration is achieved if the resistance can be reduced in small steps from the
'all in' to the 'all out' condition.
(a) The squirrel cage motor: Fig. 14.2.1H shows the torque/speed curves of a
squirrel cage motor. During starting the developed torque follows the upper curve
until the motor attains a speed at which the developed torque is equal to the load
torque. The motor will reach its rated speed if the load torque is not in excess of
the developed torque.
The torque is proportional to the terminal voltage squared, and although for
some drives the squirrel cage motor can be connected direct-to-line, it is often
necessary to reduce the terminal voltage at starting as a means of reducing the start-
ing current and the voltage drop in the connections. This is particularly applicable
when only a small starting torque is necessary since the effect is to reduce the
starting torque available. The methods used include the insertion of

(i) series resistors or reactors;

(ii) autotransformers;
(iii) star-delta switching.
 156 Protection of motors, reactors, boosters and capacitors

b-

o
b-

Running speed
0
Speed
Synchronous
speed
Fig. 14.2.1H Torque/speed characteristics o f a squirrel cage induction m o t o r

(i) Series resistor or reactor: By using this method the voltage applied to the
motor at starting is reduced by the simple expedient of introducing a series resistor
or reactor in the supply to the motor, the resistor or reactor being short-circuited as
soon as the motor has reached its normal running speed. The starting current is
thus reduced in proportion to the terminal voltage. The starting torque, however, is
reduced as the square of the terminal voltage, so that the use of a series resistor or
reactor to drop the voltage by 20% will have the effect of reducing the starting
current by 20%, the starting torque being thereby reduced to (0"8) 2 x 100%, that is
64% of the full voltage value.
(ii) Autotransformer: In this case (see Fig. 14.2.1I) the line current is equal
to the transformer primary current and the motor current to the transformer
secondary current. Similarly, the line voltage is equal to the transformer primary
voltage and the motor voltage to the transformer secondary voltage. Thus

line current = motor voltage

motor current line voltage

If, for example, the transformer secondary voltage is 80% of the line voltage, so
that the motor current is 80% of the full voltage value, then the line current is
0.8 x 0.8 x 100%, that is 64% of the motor current at full voltage. As with the
series resistor or reactor method, an 80% voltage value gives a motor torque of 64%
of the full voltage value.
(iii) Star-delta switching: By means of a three-pole changeover switch the stator
windings are star-connected during the initial starting period and delta-connected
when the motor has run up to speed. In this way the voltage per phase at starting
is reduced to 1/~/3, that is 58% of the supply voltage, and is restored to full supply
voltage for normal running. Switching from star to delta should be carried out at a
speed corresponding to fuU-load torque to avoid an abrupt change in torque as the
 Protection of motors, reactors, boosters and capacitors 157

Au,o-,rans ormer

~ Starting

k Running

Supply

Motor
stator

Fig. 14.2,11 Autotransformer starting o f a squirrel cage m o t o r

delta connection is made. This difficulty can be avoided by maintaining the circuit
through a resistor during the transition from star to delta connection.

(b) The wound-rotor motor: The total value of the rotor resistance of a wound-
rotor motor consists of the fixed resistance of the rotor winding plus the added
external resistance which is variable.
As the load torque changes so does the ratio of the total rotor resistance r t to
the slip s. To overcome the load torque, the appropriate values of secondary
current and flux, which together determine the motor torque, must be produced
and the motor's ability to achieve this is governed by the ratio r/s. The slip, for a
given value of torque, is proportional to the total rotor resistance. It will be seen
that for each value of external rotor resistance a speed/torque curve can be drawn,
and a set of typical curves is shown in Fig. 14.2.1J. Curve (a) represents the
condition when the external resistance is zero and curves (b) and (c) the conditions
with different values of external resistance connected.
By maintaining the ratio r/s constant, or nearly so, by a step by step reduction
of the external resistance the torque of the wound-rotor motor can be keptreason-
ably constant during the running-up period.
This contrasts with a squirrel cage motor in which the rotor resistance is fixed
and therefore only one speed/torque characteristic is obtainable.

Single-phase induction motors: There are many types of single-phase induction

motor and a representative selection will be considered here. The majority of such
motors have rotors of the squirrel cage type.
Unlike the polyphase induction motor, the single-phase version has no starting
torque and starting is achieved by the introduction of an auxiliary starting winding
in the stator to produce a rotating field (as with the split-phase motor), or by using
158 Protection of motors, reactors, boosters and capacitors

rs = Total secondary resistance

= r 2 plus external resistance
(c) (h) (a)

Speed

Synchronous
speed
Fig. 14.2.1J Torque/speed characteristics of a wound-rotor motor

a commutator and brushes on the rotor (as, for example, with the repulsion-
induction motor).

(a) Rotating field method: To obtain a rotating field, the axis of the starting
winding must be displaced in space with respect to the main stator winding, and the
currents in the two windings must be out of phase. The first requirement is
satisfied by placing the starting winding in empty or partly filled slots of the main
winding and the second by the choice of suitable resistance and reactance values for
the two windings, or by the addition of resistance, reactance or capacitance in series
with the starting winding. Some typical applications are given below.

t A.C. supply
!R
Running
winding ~ Starting

l i - -..° ° R
I|

Fig.14.2.1K Connections and phasor diagram of resistance-start split.phase induction motor

 Protection of motors, reactors, boosters and capacitors 159

In the split-phase motor, the resistance and reactance values of the two windings
are usually such that an angle of approximately 30 ° between the two currents is
obtained. To increase this angle and so produce a stronger rotating field, and
consequently a larger starting torque, a resistor is inserted in parallel with the
starting winding (usually categorised as a resistance-start split-phase motor as shown
in Fig. 14.2.1K). Again an increase in angle is achieved by using a reactor
(reactance-start split-phase motor). A further increase in the angle, making the
current in the starting winding lead the terminal voltage, is obtained by using a
capacitor in place of the resistor (capacitor-start split-phase motor, as shown in
Fig. 14.2.1L). To prevent overheating of the starting winding it is necessary in all
these cases to disconnect it as the motor approaches synchronous speed and this is
done automatically by a centrifugal switch on the rotor.
A.C.supply

(,' "t
/
T !R
I !R
apacl or T ~Running
~i:3tcr2fugalls~

Starting
~vinding 0 approaches 90°
Fig. 14.2.1 L Connections and phasor diagram of capacitance-start split-phase induction motor

(b) Commutator method: This method of starting is used for the repulsion
induction motor which is a form of induction motor with commutator and brushes
and with an additional squirrel cage winding in the rotor. This type of motor may
have constant-speed or variable-speed characteristics depending upon the design of
the windings. It differs from the plain a.c. series motor in that no current is led
into the armature, the brushes being short-circuited.
If the brush axis is perpendicular to the main flux, as shown in Fig. 14.2.1M(a),
the e.m.f.s induced in the two halves of the winding will cancel each other out, and
the armature will not carry any induced current nor produce any torque.
If the brush axis is in the same direction as the main flux then a current will be
induced in the armature. The four parts of this current will in turn produce four
torques which, as shown by the arrows in Fig. 14.2.1M(b), will cancel one another
out.
If now the brush axis is inclined at an angle to the direction of the main flux,
then torques T1 and 7'2 are produced, as shown in Fig. 14.2.1M(c). Thus resultant
torque T = 2(7'2 - TI ) and it can be shown that T is a maximum when the angle ¢
between the brush axis and the horizontal axis of the armature is 45 .0.

Three-phase commutator motors

The Schrage motor; Of several types of three-phase commutator motor, the rotor
160 Protection of motors, reactors, boosters and capacitors

(a)

(h)

(c) T! \

Fig. 14.2.1M Effect o f different brush positions on the action o f a repulsion m o t o r

excited shunt commutator motor (the Schrage motor) is by far the most commonly
used, and this section will be limited to a discussion of the salient features of such a
motor.
The outstanding feature of the Schrage motor is that the commutator allows
good speed control and power factor correction to be obtained in the one machine.
The speed of an induction motor can be controlled by impressing on its secondary
(stator) winding a voltage in the same direction and of the same frequency, that is
the slip frequency, as the e.m.f, induced in this winding by the rotating flux.
Selection of the correct phase and magnitude of the impressed voltage gives control
of the speed above and below synchronous speed. The three-phase shunt commu-
tator motor, or Schrage motor, is in effect a combined three-phase rotor excited
induction motor and a three-phase commutator machine, as shown in Fig. 14.2.1N.
Of the two rotor windings, one is connected to the slip-rings and the other to the
commutator segments. The stator winding consists of three separate phase windings
each connected to a pair of brushes which can be moved individually in either
 Protection of motors, reactors, boosters and capacitors 161

Compensating ~,.:.,a;.,~,
Three-phase supply
Rotor winding
(primary)

Stator windinl
(secondary)

Commutator
Fig. 14.2.1N Connections of the three-phase Schrage motor
direction round the commutator. The commutator voltage varies as the angular
distance between the brushes of each pair, being greatest when the distance is
180 electrical degrees, and decreasing to zero when they are both on the same
commutator segment, thus short-circuiting the stator winding. In Fig. 14.2.1P(a)
the brushes are shown in this latter position and the machine will run as an
ordinary induction motor. In Fig. 142.lP(b) the connections are such that the
potential difference between brushes x and y will have a component in direct
opposition to the e.m.f, induced in the stator, reducing the motor speed to a value
below synchronous speed. By reversing the brushes, as shown in Fig. 142.1P(c),
the commutator e.rn.f, will have a component in phase with the rotor induced
o.m.f., thus increasing the motor speed to a value above synchronous speed.

y i~i ,x
Ilil !11'1 I ! ! I~lkl ! I 1 11 i

I I I
(aj (h) (c)
Fig. 14.2.1P Alternative brush positions of Schrage motor

The phase of the commutator e.m.f, between brushes is determined by the

positions of the centre point of the brushes relative to the centre of the rotor
winding, so that by moving the brush system as a whole round the commutator the
phase of the e.m.f, injected into the secondary circuit will be varied relative to the
rotor induced e.m.f.
Thus the angular separation of individual brush sets determines the speed varia-
tion above and below synchronous speed, the angular position of the brush system
as a whole controlling the power factor of the motor.
From the typical speed/torque curves shown in Fig. 14.2.1Q it will be appreci-
ated that the speed is reasonably constant up to approximately 100% torque,
particularly at the higher speeds.
162 Protection of motors, reactors, boosters and capacitors

I I I
I I I I II
= 200- --I- - - -L - - --I - ----11 - - -- -I"
. . . . . . . . ~ ~ I
/
I I J_ . . . I . .
~ 160 - ~ . . . .l . 1-

= I I I-----
',, 12o . - -. - . . -I" I r -- I T

=~:~
c~
80 _...___..i.___4____.1____,.1.._ - II
"5 I I i I I
I " ; I t I
~- 4o --i ' l. . . . 1--- i "-t:
l t t
I
I ,
0 50 !00 1 50 200 250

% o f full load t o r q u e

Fig. 14.2.1Q Typical speed~torque characteristics o f a three-phase commutator motor

14.2.2. Application o f d.c. and a.c. m o t o r s

In choosing a motor for a given application, consideration must be given to the type
of load it will drive, and from this the speed and torque requirements can be deter-
mined. For loads such as cranes, hoists and electric traction, motors with speeds
varying with the load are required and the d.c. series motor or the a.c. repulsion
motor are most often used. Fans, blowers and unloaded compressors demand low
starting torque while pulverised fuel mills and loaded compressors start under load
and demand high starting torque to break away from standstill.
Constant speed applications include motor-generator sets, concrete mixers and
constant speed conveyors and in these an increase in load torque, for example the
addition of coal to a conveyor, requires a proportionate increase in power output,
the power output being proportional to the load torque. Of the motors dealt with
in this chapter, the d.c. shunt motor and the induction motor would be suitable for
such loads.
Variable speed applications may be classified as (i) constant torque, (ii) variable
torque, (iii)constant power-output.

(i) A good example of constant-torque application is the automatic machine

tool where the output is directly proportional to the speed. The wound-rotor
induction motor is the obvious choice for this application, but a d.c. shunt motor
provided with means of speed adjustment by field control could be used.

(ii) Fans and blowers are typical of the variable-torque application. With such
loads the torque varies as the square of the speed, and a d.c. shunt motor provided
with means of speed adjustment by field control would be suitable.

(£ii) Machine tools fall into the constant power-output category requiring, as
 Protection of motors, reactors, boosters and capacitors 163

they do, torque varying inversely with speed. Again, the d.c. shunt motor as for
(i) and (ii) or the wound-rotor induction motor could be applied.

When fluctuations in speed with varying load have to be catered for a 'shunt'
type of three-phase commutator motor is used. The best known of these, the
Schrage motor described earlier, is used extensively in the speed-control systems
for large three-phase induction motors providing power factor improvement at
the same time. It is applied widely in the printing and textile industries and for
driving some conveyor equipment.

14.2.3. Motor control

It will be seen later that the type of protection used, in particular the protection
against short-circuits in a motor or its connections, is to some extent dependent
upon the type of switchgear used to control the motor. This falls into two classes:

(a) a contactor, with high-breaking-capacity fuses providing the short-circuit

protection; and
(b) a circuit breaker, with short-circuit protection provided by instanteneous
attracted-armature type relays.

In general, the choice of switchgear will depend upon the size of the motor, for
example contactors and fuses for up to approximately 150 h.p. and circuit breakers
for larger motors.
Control by contactor and fuses: A contactor consists of a moving contact operated
by an electromagnet attracting an iron armature against the action of a spring. The
pressing of the 'start' push button energises the electromagnet, thus completing the
operating coil circuit. When the 'stop' button is pressed the coil circuit is opened,
de-energising the electromagnet and allowing the moving contact to isolate the
motor from the supply. Contactors may be of the air insulated or oil insulated
type, the latter having the advantage that the motor current is broken in oil which
generally permits a smaller overall size of the control unit.
Contactors can be used where the current to be interrupted is limited to
approximately six times their rated current, the rated current being of the same
order, or a little higher, than the normal full load of the motor. Direct-acting over-
load trip devices may be incorporated in the contactor, protection against short-
circuits being provided by fast operating high-breaking-capacity fuses. The fuses
should blow at currents in excess of those which can be handled by the contactor
but should not blow at currents within its capacity.
Since the contactor coil will hold the armature against the spring only when
there is sufficient magnetic flux, any severe drop in or failure of the supply will
result in the opening of the contactor. Thus an inherent 'no-volt' release feature
is provided ensuring that the motor cannot re-start after a partial or complete
failure of the supply.
164 Protection of motors, reactors, boosters and capacitors

When the contactor unit is small it is generally located adjacent to the machine
being driven, so that control of the machine is facilitated, in such cases the 'start'
and 'stop' push buttons are an integral part of the contactor. When the contactor
is too large to be located close to the machine, local controls are provided at the
machine.

Control by circuit breaker: Large motors are controlled by circuit breakers of

either the air-break or off-break type, which should be capable of interrupting the
highest current which can flow under the most severe fault condition. Oil-break
gear is seldom installed in generating stations chiefly because of the additional
fire hazard when oil is present under fault conditions.
A circuit breaker may be arranged for hand closing, electromagnetic solenoid
closing or spring closing. Hand closing is usually provided for small, infrequently
operated circuit breakers, solenoid closing for those operated frequently or from a
remote point, and spring closing for circuit breakers operated only occasionally
and where hand operation is undesirable. Overload trip devices operating direct
on to the tripping mechanism usually form an integral part of the circuit breaker.

14.2.4 Types of fault

In general it is necessary to protect a motor against abnormal running and fault

conditions arising from:

(a) prolonged overloading as a result of the application of excessive mechanical

load:

(b) single-phasing caused, for example, by the rupturing of a fuse or by the

open circuiting of a connection in one phase of a three-phase motor. If one phase
is open-circuited when the motor is running it will continue to run and provide
power even though it is connected to what is, in effect, a single-phase supply. If the
load on the motor is of the order of its rated output, the current drawn from the
supply will be appreciably higher than the current for which the windings are
designed and if the condition is allowed to persist, severe damage may be caused:

(c) short-circuits between phases or between phase and earth in the motor
winding or its connections. Short-circuits may be caused by the chafing of connec-
tions, accidental shorting of the motor terminals or cable sealing ends or by cable
faults:

(d) partial or complete collapse of voltage.

14.2.5 A.C. and d.c. motor protection

The protection of motor plant is based on the same essential considerations

 Protection of motors, reactors, boosters and capacitors 165

whether the motor is driven from an a.c. or a d.c. source. In some instances, for
example, the thermal overload relay, a modified single-phase version is applied to
the protection of d.c. motors. Any dangerous or potentially dangerous condition in
either an a.c. or a d.c. motor, its control or connections, must be detected and
action taken automatically to disconnect the affected equipment. Such conditions
are classified broadly as low or falling supply voltage and overloading beyond a pre-
determined safe value for an excessive time. To these conditions must be added the
open-circuiting of one phase of a three-phase a.c. motor and a short-circuit in either
an a.c. or d.c. motor.

7
I00 - Speed
"O
¢b
~-
u~ 80 ="

E
~ 60- 4-

3-
~, 4o-

20--
Current

o-

5 I0 15 20

Time in seconds
Fig. 14.2.5A Typical starting characteristic for "direct-on" started induction motor

Many motors draw a starting current from the supply of several times their
normal full-load current, and it is essential that the protection should be unrespon-
sive to this starting surge provided that the motor current returns to its running
value within the time determined by the design of the motor. On the other hand,
the protection must not be given a setting greatly in excess of full load or it will be
unable to safeguard the motor against overcurrent conditions. Figs. 14.2.5A and
14.2.5B show typical starting characteristics for an a.c. 'direct-on' induction motor
and for a d.c. shunt-wound motor, respectively.

Thermal relays: The two opposing requirements referred to are met in a relay
having an inherent time-lag characteristic, an example of which is shown in
Fig. 14.2.5C. The essence of the thermal relay is the specially designed element
which simulates as closely as it can the changing thermal conditions in the motor,
allowing the motor to be retained in service up to the point beyond which damage
would probably be caused. Essentially, the relay consists of three single-phase
elements, each element comprising a heater and an associated bimetal actuated
movement. The three bimetal spiral elements are mounted axially in line and
respond to a rise in temperature of the heaters which in turn produces movement
166 Protection of motors, reactors, boosters and capacitors

!!

t
"3
¢>
1
¢=
¢)
Current
i.
i.
Or}

12 !
l i l \
0
I L ....
0 ..... I- T I
I I I
I I
I
. . . . . ~
I I
.,=,- - . . . m . r

I I I
I I I
I I I
I I !
I I
I I I
I I I
I Speed I I I
I I I I
I I I I
0 -
I I I I

0 tI t2 t3 t4
Time

Fig. 14.2.5B Typical starting characteristic for a d.c. shunt-wound motor

l.oad indicating Adjustable

scale an,I p o i n t e r o v e r l o a d contact

Overload setting
scale and pointer

Heat shield
I

Amhient compensation bimetal

Fig. 14.2.5C Construction o f typical thermal relay for motor protection

 Protection of motors, reactors, boosters and capacitors 167

of the contact assembly (Fig. 14.2.5D). A change in ambient temperature rotates

the two outer ends of an element through the same angle without causing move-
ment of the contact arm. Characteristic curves for a setting of 125% under
'starting from cold' and 'running' conditions are shown in Fig. 14.2.5E.
Overload v "

contacts ,~
Phase-unbalance ~, '

contacts

/,%

f f

I y Trip B I
t /.
Fig. 14.2.5D Contact arrangement of typical thermal relay
28

18
14 ! ov~do~J
i ' L'-setting [25% } " '
•-. lO
= 8
E
7
.E
6
E
•,7. 5
\ ~ ~, S t a r t i n ~
~o 4
•,7. 3

~. 2
O
! Running ~ ""

|Jl

1 2 3 4 5 6

M~tor current as m u l t i p l e of rated current

Fig. 14.2.5E Characteristic of typical thermal relay for motor protection

Single-phasing protection: This is necessary in view of the ability of a three-phase

motor to continue to run with an open-circuit in one phase, although it may well
suffer damage as a consequence. The advent of motors designed on a maximum
continuous rating (m.c.r.) basis providing very limited continuous overload
capacity, has emphasised the importance of protection against single-phase running.
The single-phase feature is conveniently incorporated in the three-phase thermal
simulating relay, the operating time-lag depending to some extent on the load
being carried by the motor when the open-circuit occurs. With the motor on full
168 Protection of motors, reactors, boosters and capacitors

load the operating time is usually 20 - 30 s with correspondingly longer times at

smaller loads. The single-phasing contacts will close also if the out-of-balance
between any two phase currents exceeds about 12% at full load.

Short-circuit protection: Protection against short circuits in the motor winding,

or the connections to it, is often provided by incorporating in the thermal relay
case separate high-set, instantaneous overcurrent or earth-fault elements or both.
The attracted-armature type element operates against the action of a return spring
and in one design rests against a leaf spring to reduce the effects of vibration and
shock. In another design these effects are guarded against by using a statically
pivoted beam, the armature being attached to one end of the beam and the moving
contacts to the other end. Typical settings are 4 to 8 or 8 to 16 times full load for
the instantaneous overcurrent elements and 0.2 to 0.4 times full load for the
instantaneous earth-fault element. Three such overcurrent elements or two over-
current and one earth-fault element can be incorporated in the standard three-phase
thermal relay case.
It will be appreciated that when the motor supply is earthed via a resistor, the
earth-fault current may be less than that needed to operate positively the over-
current elements connected in the phases. In such cases it is necessary to employ
the residually connected earth-fault element.
A frequent cause of single-phasing is a 'blown' fuse and care is needed in the
choice of fuse ratings. During starting, the current approaches fusing current,
causing the fuses to run hot for an appreciable time. This may give rise to oxida-
tion and deterioration of the fuses and ultimately a fuse may blow during starting.
The motor will continue to run with considerably increased current flowing in the
other two phases which, if allowed to persist, will damage the windings.
The use of high-breaking-capacity fuses for short-circuit protection has the merit
that the circuit is opened during the first quarter-cycle of fault current thus
preventing the current from attaining its theoretical maximum or 'prospective'
assymetrical peak value which would flow in the absence of the fuses. The current
at which the fuse 'blows' is known as the 'cut-off current'. The fusing time may be
extremely short, for example between 0.001 and 0.002 s compared with a
minimum operating time of an overload trip of about 0.1s. Thus the fuse is well
able to completely interrupt the fault current before the overload trip can operate.

Smiling relays: The stalling relay is designed for use in conjunction with the
thermal overload and single-phasing relay. Basically the stalling relay consists of a
control contactor and a thermal overload unit fitted in the same case. The thermal
overload unit is energised via the contactor which closes only during the starting
period or, if the motor stalls, while starting or running (see Fig. 14.2.5F). During
a normal healthy start the contactor closes, switching the thermal unit into circuit,
but the motor current falls to normal before the thermal unit operates and the
contactor opens, de-energising the thermal unit. An optional extra feature is
shown in Fig. 14.2.5G and consists of an auxiliary contactor having a 'twice-in'
 Protection of motors, reactors, boosters and capacitors 169

Control
resistor

Control
contactor I

_ 7 i unit

Trip supply

Fig. 14.2.5F P and B type L IA stalling relay

Control
resistor

Control
contactor
Thermal
unit
Alarm

Alarm
common
I.imiting
To c . b .
resistor
trip coil

Trip supply

Fig. 14.2.5G P and B type L 1A stalling relay incorporating a "twice-in" feature

feature. This permits one attempted restart of a motor after a stall but is so
constructed that it latches in on the second stall, thus preventing repeated attempts
to restart a defective motor.

In one design the 'twice.in' feature consists of a shunt-connected attracted-

armature relay with a follower-pin attached to the top of the armature, and moving
in a labyrinth slot in a hinged metal plate (see Fig. 14.2.5H). As the armature
moves in and out on successive stalls the pin works through the labyrinth slot,
allowing the hinged plate to fall until, on the second stall, the pin is trapped at the
end of the slot, thus latching the armature. Flags are incorporated to indicate the
number of times the motor has stalled since the 'twice-in' feature was last reset.
Resetting after two stalls is by hand only.
170 Protection of motors, reactors, boosters and capacitors

i'ol!ower pin

Labyrinth
slot ,~
Hinge,
Red
plate

Core (;reen
I
Flag

J
\
Armature

Fig. 14.2.5H Arrangement of mechanical "twice-in" feature o f type L 1A stalling relay

Electronic overload relay: A modern alternative type of thermal relay, of solid-

state design rather than of the bimetallic element design protects the motor against
overload by deriving an a.c. input from, and related to, the motor supply current.
This current transformer derived input is rectified, adjusted to an appropriate level
by a current setting circuit and then applied to the overcurrent detection circuit.
When the reference level is exceeded by the input the time-setting circuit, incor-
porating a resistor/capacitor network, followed by an amplification stage and
closure of the tripping circuit, is triggered. The relay has a characteristic similar to
that of a bimetallic element relay, i.e. an inverse-time characteristic such that higher
levels of overload cause faster disconnection of the motor and lower levels are
tolerated for longer periods. The block diagram of a relay of this type is shown in
Fig. 14.2.5I.

Protection against an open circuit in one phase is provided by a separate circuit

which detects the higher ripple content of the full-wave rectified current derived
from the current transformers. After passing through a waveform and gating circuit
into the power circuit it provides a signal for operation of the tripping circuit.
The more sophisticated versions of this relay may incorporate means to adjust
the model characteristics so that an analogue closer to the actual temperature rise
of a motor may be obtained. Outputs may be used for purposes other than
tripping, e.g. regulating the motor load. The closer control provided by such
devices is particularly useful in the protection of maximum continuous rating
(m.c.r.) motors and others with an arduous duty cycle. Appropriate modelling will
overcome the problems encountered when a motor is required to start against a
stalled load, to 'inch' or to start and stop frequently.

Thermal tn'ps and electromagnetic overload protection: For the smaller size of
motor with a full-load rating of up to approximately 25A it is usually adequate
 Protection of motors, reactors, boosters and capacitors 171

R
i ( ~ Power
r - - Y source

T°l,°rsuoc°lac'°r !
Rectifier
circuit

Test
button
•----'------0 C
--1
setting
-- circuit

l
Openp.as: ]_ _!
I_
Overcurrent
detecting I- q detecting
circuit circuit

I
,,

Waveform ,'1
shaping
circuit circuit
1

V~ltage
regulating OR circuit
circuit

~o~r I____~ L____q

supply Amplifier
circuit circuit

Latching

Pov~er
circuit

• t rigger
R

Fig. 14.2.51 Block diagram o f a typical electronic overload relay

172 Protection of motors, reactors, boosters and capacitors

technically, and attractive from an economic standpoint, to employ a simpler form

of protection than the sophisticated thermal overload relay described at the
beginning of this Section. Recourse is then made to either the thermal overload
trip (in essence a much simplified form of the thermal overload relay)or the
electromagnetic overload and associated time-lag device. The thermal trip is
operated either directly or indirectly by the heating effect of the current on a
thermal element and the electromagnetic trip is operated by the increase of
magnetic flux with current.
(a) Thermal trips: In an early type of thermal trip a bimetal strip was used as
the sensing element, the motor current passing directly through the element. As
the thermal strip heated up it lifted and so raised, against spring tension, a
horizontal trip rod positioned above it. If the overload condition persisted, further
movement upwards of the rod, to a point predetermined by the setting, caused the
mechanical trip to operate to open the associated contactor. A second rod,
positioned so that the bimetal strip was between the two rods, followed the move-
ment of the bimetal and upper rod until held by a stop of the same material and
therefore having the same thermal characteristics as the thermal strip. Ambient
temperature compensation was thus obtained, and the motor current at which the
relay operated was substantially unaffected by temperature changes. These relays
were applied to d.c., single-phase a.c. and three-phase a.c. motors. When used with
the latter, inherent protection against single-phase running was obtained as any
appreciable unbalance in the phase currents caused the two rods to move in
opposite directions and so operate the tripping mechanism.
In a later version of this type of relay the bimetal elements do not carry current
but are operated indirectly by heaters. In the case of small motors, the heaters are
direct connected to carry the motor currents and are fed from current transformers
when used with larger motors. Two slotted trip bars or slides are coupled together
at one end and in normal, healthy conditions the two bars move together in a
horizontal direction corresponding with the balanced deflection of the three
bimetal thermal elements, as shown in Fig. 14.2.5I. After a predetermined amount
Bimetal thermal Upper
elements ~ trip bar

A
.,mO.----=-~-----
Lower
trip bar
Trip switch
contacts

Fig. 14.2.5,1 Simple thermal overload and single-phasing preventor for motor protection
(E/ectrica/ Apparatus Co.)
 Protection of motors, reactors, boosters and capacitors 173

of movement during an overload condition, one of the bars strikes a stop which
prevents it following the other bar any further. Continued movement of the other
bar under the action of the bimetal elements operates the mechanical trip to open
the motor contactor. If one phase is open-circuited while the motor is running
the thermal element in that phase will cool and deflect in the opposite direction to
those still carrying current. These two elements will remain deflected and differen-
tial movement of the two bars will again cause the mechanical trip to operate. Part
of the contact arrangement of this relay is shown in Fig. 14.2.5K.
Upper
trip bar
Trip s~'
contac

Bimeta
thcrma Vie~ looking
elemen in d i r e c t i o n
o f a r r o w A in
Fig. 14.2.5J
Lo~er
trip ba~

Fig. 14.2.5K Contact arrangement o f thermal overload and single-phasing preventor

The simple thermal trip devices just described should not be confused with the
much more sophisticated and comprehensive thermal relay described at the
beginning of Section 14.2.5. The latter relay is used extensively for the protection
of the larger three-phase a.c. motors driving power station auxiliaries, and a single-
element version of that relay is available for use with single-phase a.c. motors and
d.c. motors.

(b) Electromagnetic trips: These overload trips are instantaneous in operation

and, in order to take advantage of the motor's permissible overload capacity, are
usually employed with a time-lag device. The complete assembly consists of a
series-wound coil surrounding a vertical iron plunger, to form a solenoid, and an
associated time-lag in the form of an oil or silicone fluid f'tlled dashpot or air vane.
Their application thus depends on the current/time characteristic which can be
obtained to ensure the necessary retarding action to prevent tripping on the
occurrence of heavy overloads of short duration. The oil dashpot time-lag is the
type most commonly used.
The required time-lag is obtained by adjustment of the rate at which oil is
permitted to pass from the upper side to the lower side of the piston in Fig. 14.2.5 L.
Oil flow is governed by a small hole B in the rotatable plate C and a large hole D in
the piston. Hate C has a number of holes graded in diameter and by changing the
size of hole in use a range of time-lag settings between, for example, 10 and 30 s
can be obtained. Overload current calibrations are marked on the outside of the
plunger casing, the overload setting being changed by releasing the setting screw E
174 Protection of motors, reactors, boosters and capacitors

Trip pin

~Ovcrload coil

Plunger

Plunger easing

Setting
screw E

Hole

Plate Piston A

Hole D

Fig. 14.2.5L Cross-section o f a typical oil dashpot time-lag for an electromagnetic overload
trip (Allen West & Co. Ltd.)

and raising or lowering the dashpot to change the position of the plunger in relation
to the magnetic field of the series overload coil.
A more sophisticated design of oil dashpot is available in which a restraining
device, designed to operate on the occurrence of heavy transient or short duration
overloads, prevents unnecessary tripping under such conditions. When a heavy
overload occurs the sudden rush of oil causes a flap to close over the hole B. The
heavy current causes the plunger to lift slightly, creating a vacuum beneath it, and
when the current falls the plunger returns to its earlier position ejecting oil through
 Protection of motors, reactors, boosters and capacitors 175

hole B, forcing the flap off the hole. The piston is then able to function as a time-
lag device under normal system conditions.
The restraining feature has applications with direct-on starters for squirrel cage
motors and where motors may be subjected to severe peak loads of limited duration,
enabling the overload trip to be given a normal setting without fear of operation
under short-duration, heavy-current conditions.
When an overload condition has ceased or the trip pin has operated, the plunger
and piston sink back to the bottom of the dashpot taking typically some 10 - 30 s
to do so: in special designs this resetting time can be extended. This delay should
be considered a necessary feature preventing the motor from being restarted
immediately after the clearance of an overload condition. Even with the special
designs it is not generally possible to extend the resetting time beyond 100 - 150 s.
As the reset position of the plunger is determined by the position of the dashpot,
the further this is screwed down the more current is required through the solenoid
to raise the plunger and operate the trip pin. Thus the position of the dashpot can
be calibrated in terms of motor line current and if one such device is used in each
line, three-phase protection can be provided.
Fig. 14.2.5M shows a typical current/time curve for a two-rate dashpot the
function of which is to allow the device to clear a motor start condition but to
operate as intended under overload conditions. Time/current curves vary depending
800 -

700 -

600 -

u~
500 -

i.
>

t,- 4O0 -
O
4..

~J
i.
300 -

2OO

I00 =
I
=
I i
i
i I I I

5 10 15 20 25 30 35

Time lag (seconds)

Fig. 14.2.5M Typical current/time curve for an electromagnetic overload relay with a two-rate
dashpot
 176 Protection of motors, reactors, boosters and capacitors

4000 t
3000

2000

1000 -
800
600

~' 400
2 300
.~.
¢d
2oo

8 ioo i
> 80
6o !
40-
30-

20"

| 0 I - I ' 1' " I ' I I ' '"]

0 IO 20 30 40 50 60
Temperature (°C)

Fig. 14.2.5N Variation of the viscosity of the dashpot fluid with temperature

upon the viscosity of the dampening fluid used to fill the dashpot. The viscosity of
mineral off decreases significantly with temperature (Fig. 14.2.5N) and a point may
be reached when the timelag is unacceptably short. The use of silicone fluid instead
of oil considerably improves the performance but the change of time delay with
temperature will still have some significance (Fig. 14.2.5P).
To obtain optimum performance and take account of seasonal changes in the
viscosity of the fluid the dashpot setting should, theoretically, be adjusted to take
account of temperature changes between summer and winter working. This is
obviously difficult to ensure in a large installation and the need for it has to be
considered a limitation of the device.
It is not permissible for the dashpot to be filled with a fluid having a much
higher viscosity than that for which it was designed in an attempt to prevent the
device from tripping during starting: such an expedient would inevitably lead to
much slower operation than is indicated by the published current/time curves and
probably result in damage to the motor.
As the time delay is produced by utilising the effect of a piston moving in a
viscous fluid, the operating current/time curve cannot match, except in a very
 Protection of motors, reactors, boosters and capacitors 177

140-

120-

too

~ so

I-

60 I-- "! ! ' ' ' I '! I

0 I0 20 30 40 50

T e m p e r a t u r e ((~(')

Fig. 14.2.5P Change in dashpot time delay with temperature using silicone fluid in the
dashpot

approximate manner, the heating and cooling characteristics of the motor and this
limits the degree of protection which can be afforded the motor.

Comparison o f magnetic and thermal protective devices

As described in the earlier parts of this Section, motor protection falls into two
broad categories based on the detection of and response to either electromagnetic
or thermal effects. Although simpler and essentially of one basic type the former
clearly has a number of applications, but the range of thermal devices available is
more readily matched to the motor characteristics and to the conditions which lead
to motor failures. The performance comparison which follows will serve to
illustrate and summarise the essential features of the two categories.

Elec tro magn e tic trip s:

(i) These devices are relatively insensitive to small degrees of overload which,
if sufficiently frequent and allowed to continue for extended periods, could shorten
the life of the motor insulation. They cannot easily be set accurately enough for
close protection which is a prime requirement for m.c.r, motors.

(ii) The devices perform more effectively on medium to heavy overloads

particularly the latter when currents approaching stalling level are reached and the
required rapid operation under the conditions is achieved.

(iii) The current/time curve is not related to the actual conditions within the
178 Protection of motors, reactors, boosters and capacitors

motor and the device therefore has no thermal memory.

(iv) Under single-phasing conditions the performance is completely dependent

upon the degree of overload which this effect produces in the two healthy lines.
If the increase in line current is only small the operation of the device is doubtful.

(v) Changes in the viscosity of the dampening fluid with temperature will
cause variation in effective overload settings.

Thermal overload devices:

(i) All types provide some degree of protection at all motor overloads: the
more sophisticated relay types can be set more accurately and are thus better suited
to the protection of m.c.r, motors.

(ii) The relay types give a consistency of performance which is often as

important as inherent accuracy and a running load scale can also be incorporated.

(iii) As the operation of the devices is based on thermal effects the protection
provided is better related to the needs of the motor than that given by the
electromagnetic type: even so the thermal time constant of the relays is much
shorter than that of the motors being protected.

(iv) Single-phasing performance is good and with the more comprehensive

types of relay, operation on small degrees of load unbalance is achieved.

(v) Bimetallic devices and particularly relays with coiled bimetals have an
appreciable resetting time. This is a desirable feature in that the motor has some
time in which to cool down before a restart is possible. Even if the relay allows a
restart, it will trip again if insufficient cooling has taken place.

Thermistors

The forerunner of this type of device as used in motor protection was the thermo-
couple or thermostat embedded in the motor end-windings and applicable with
equal facility to a.c. and d.c. machines. Commonly referred to as 'motor overheat
protection' they provided protection against most motor internal conditions which
give rise to excessive temperatures including sustained overload; high, low or
unbalanced voltages; locked rotor; blocked ventilator; and single-phase running of
a polyphase a.c. motor.
The main advantage of this type of protection is that the temperature sensitive
elements are located in the motor windings themselves and thus give an accurate
measurement of temperature at the points where the overheating is likely to occur.
This feature also allows a motor to be run nearer to the limit of maximum
 Protection of motors, reactors, boosters and capacitors 179

permissible temperature at times of necessity.

The modern version of this type of device is the thermistor (positive-
temperature - coefficient resistor). Several such devices are embedded in and
bonded to the enamel-covered wires forming the stator windings during manufac-
ture and connections brought out to a solid-state control unit and interposing relay
mounted separately for small motors and usually built into the motor terminal box
of motors above 10 h.p. The relay is inoperative until the thermistors indicate that
the winding temperature and the current flowing in any phase exceed their
stipulated limits. High winding temperature and excessive current are thus needed
before tripping takes place and tripping is avoided on starting and on the
occurrence of thermal overloads of short duration.
The thermistor need be little larger than a match head with a correspondingly
low thermal inertia. As these small sensors can tolerate only a few milliamperes
the control unit includes an amplifier which, in conjunction with the relay (now
frequently solid-state) would initiate tripping and, if required, an alarm. The unit
is designed to fail to safety, i.e. to the tripping condition. The characteristics and
testing requirements are laid down in BS 4999 which deals with the thermal protec-
tion of electric motors.
Thermistor characteristics are now amenable to close manufacturing control and
any appropriate value of reference temperature may be selected from the nominal
110 - 160°C range of motor-trip reference temperatures. The use of positive-
temperature-coefficient (p.t.c.) thermistors gives a resistance/temperature charac-
teristic as shown in Fig. 14.2.5Q, from which it can be seen that over a relatively
narrow critical-temperature range the resistance increases very rapidly. Typically,
this increase is from about 100£Z to over 1000~ over a temperature range of only
about 20°C in the band between 100 and 200°C.
3000" p.t.c. 130

2500" 140
160
E
2000"

u I 500"
C

•~ ! 000"

5oo"

._.~- ! !
,
! ! ! !

0 20 80 I00 120 140 150 160

Temperature (cJt')
Fig. 15.2.5Q Resistance/temperature characteristics of posi~j've-temperature-coefficient
thermistors for operating temperatures of 130, 140° and 160°C

The characteristics are stable and the steep rate-of-change of resistance continues
well beyond the 1000~ level. The characteristics are controllable during manufac-
ture and the reference temperature (defined in BS 4999 as 'the nominal
temperature of the thermally sensitive device at which it is required to cause the
 180 Protection of motors, reactors, boosters and capacitors

protection system to operate') can be selected to provide a warning or a tripping

signal at any temperature required. In practice the required temperature will
generally lie between the l l0°C recommended in BS 4999 as the early warning
reference for Class E windings and the 160°C trip reference for Class F insulation.
The application of thermistors has been extended certainly up to motors
operating at 3.3 kV and they are now generally accepted as a useful, relatively
inexpensive form of close protection of particular value for motors running at or
near full output. A disadvantage is that they can only be installed at the manufac-
turing stage and later replacement is difficult. Their response time is rather slow so
that there is a tendency for them to lag behind the temperature of the winding with
the consequent risk that they may operate only after the insulation has been over-
heated.
A future development may be that a a unit incorporating direct thermal
protection for attaching to a motor after manufacture thus moving towards the
universal application of thermistor type devices.

Undervoltage protection: The chief reason for employing undervoltage protection

of either a.c. or d.c. motors is to ensure that the motor circuit-breakers or
contactors are tripped on a complete loss of supply, so that when the supply is
restored it is not overloaded by the simultaneous starting of all the motors. This is
particularly important with a.c. motors in a power station where the simultaneous
starting of a large bank of motors would probably overload, and result in the
tripping of the feeding transformer. It is necessary also to avoid danger to
operators when the supply is restored without their prior knowledge and, for those
motors which are not direct-on-line started, to prevent full voltage being applied to
a stationary machine. Undervoltage protection may take the form of:

O) a plain undervoltage release coil fitted with an oil-dashpot time-lag and

operating direct on to the circuit-breaker trip bar or contactor mechanism;
(b) an undervoltage no-close relay with contacts interlocked with the motor
starter and connected in the d.c. trip coil circuit, the time-lag being
obtained by oil-dashpot;
(c) an undervoltage relay with an auxiliary time-lag relay; or
(a) a simple contactor with electrically held-in coil.

The releases or relays are of the single-pole type connected between the two
poles of a d.c. supply or between two phases of a three-phase a.c. supply, the
voltage and time-settings being variable. To reduce the risk of unwanted tripping
of whole groups of motors, for example, on the occurrence of a fault on the trans-
mission system, it is necessary to ensure that the relay will not operate when
voltage depressions of short duration are experienced. A sufficiently low voltage
setting and an adequate time-setting must therefore be applied to this protection to
allow it to remain inoperative during transient disturbances. In practice this means
that a voltage setting of about 50% and a time setting of several seconds may be
 Protection of motors, reactors, boosters and capacitors 181

required, the settings chosen being compatible with the characteristics of the
associated plant. In the particular case of pulverised-fuel-boilers, special consider-
ation should be given to the time setting of the undervoltage protection to guard
against the risk of boiler explosion.
The electrically held-in contactor has an inherent undervoltage feature in that
when the coil is energised the contactor closes but remains closed only as long as
the coil remains energised. Face-plate type starters, particularly those used with the
small d.c. motors, incorporate an undervoltage release coil in the starting box. The
spring loaded moving-contact arm is held in the 'fully on' position by the electro-
magnetic coil which is energised when the control contactor is closed. If the
voltage falls to a level at which the strength of the electromagnetic field is
insufficient to hold the arm, the spring returns the contact arm to the 'off' position,
disconnecting the motor from the supply.

14.3 Reactors

14.3.1 The place of reactors in a power system

The reactors which may be used in a power system can be generally classified into
two main groups"

(a) series reactors for short-circuit current limitation

(b) shunt reactors for reactive compensation.

The purpose of the series reactor follows from the fact that the fault current which
flows for a fault at any given point in a power system is determined by the
impedance of the power system as seen from the point of fault. The maximum
fault current can, therefore, be limited to an acceptable value by the provision of
series reactors of appropriate value at suitable points in the power system. The
provision of such fault-limiting reactors can avoid the necessity of providing larger
or specially braced conductors or circuit breakers of higher rating, capable of with-
standing the short-circuit currents which would otherwise occur. The use of series
reactors may thus show appreciable advantages in capital cost and space require-
ments, a disadvantage being the increased regulation of the system, particularly
when supplying low power-factor loads. However, this increased regulation can be
compensated by the use of voltage regulators (see Section 14.4) where necessary.
The second type of reactor referred to, namely the shunt reactor, finds its
application in the compensation of capacitive reactance, the lagging current taken
by the shunt reactor being used to reduce or cancel the leading current taken by the
shunt capacitive reactance of the system at the point concerned. Thus, shunt
reactors are commonly used to compensate for the large capacitance currents which
tend to be present in cable networks.
182 Protection of motors, reactors, boosters and capacitors

14.3.2 Types of reactor

Series reactors may be classified according to their construction:

(a) air-insulated, cast-in-concrete or concrete clad types, as in Fig. 14.3.2A,

and
(b) off-immersed, as in Fig. 14.3.2B.

Fig. 14.3.2A 50 Hz single-phase air insulator reactor (GEC Transformers Ltd.)

They may be classified also according to their application, namely generator,

busbar or feeder reactor.

Air-insulated reactor: In this type, the winding is wound on supports and then set
in concrete so that the vertical and horizontal spacers are solid concrete, or the
winding is damped firmly between concrete end-rings. A cast-in-concrete reactor is
inevitably bulky, and because of the strong magnetic field surrounding it, it must be
sited well away from metal work and with adequate space for heat dissipation. The
use of a tank is thus ruled out making it necessary to provide some other form of
accommodation; this is often in the form of a special cell or chamber in, or
 Protection of motors, reactors, boosters and capacitors 183

Fig. 14.3.2B 50 Hz three-phase oil-immersed reactor (GEC Transformers Ltd.)

adjoining, the main building. This type of reactor is used generally at voltages up to
and including 33 kV.
Oil-immersed reactor (magnetically shielded)." In the oil-immersed reactor the
difficulty of the external magnetic field is overcome by providing a laminated iron
path around the coil through which the flux can pass easily and without significant
losses. Thus there is no appreciable external field to cause heating of nearby steel-
work or affect adjacent electrical measuring equipment. This type therefore lends
itself to both indoor and outdoor service. The magnetic shield, as shown in
Fig. 14.3.2C, consists of a 'bird cage' of iron laminations built up in the form of a
barrel, the end clamping bars and the hoops around the body having insulated joints
to prevent the formation of short-circuited turns in which the flux would otherwise
set up a circulating current. The presence of the magnetic shield reduces the
reluctance of the magnetic path causing the reactance under normal operating
conditions to be, perhaps, 10 - 15% higher than under short circuit conditions.
184 Protection of motors, reactors, boosters and capacitors

F ig. 14.3.2C Coil assembly of 1980 k VAr, 11 k V, 50 Hz three~ohase oil-immersed reactor

(GEC Transformers Ltd.)

During the short circuit, saturation of the iron occurs and the reactor behaves
substantially the same as an air-cored reactor. However, the rating is based on the
short circuit reactance, the increase in reactance at normal load currents being of no
great importance.
 Protection of motors, reactors, boosters and capacitors 185

Oil-immersed reactor (nonmagnetically shielded): The alternative of nonmagnetic

shielding with copper, or sometimes aluminium, sheets bent to form a cylinder is
often used. Typical shields are shown in Fig. 14.3.2D. The ampere-turns induced
in the shield have the effect of reducing the reactance by an amount equal to the
percentage of the coil ampere-turns induced in the shield. The larger the diameter
of the shield the smaller will be the 12R loss, but this advantage has to be weighed
against the disadvantages of a larger tank and a greater volume of oil.
The circulating currents induced in the shield result in a counter m.m.f, which
confines the flux to a h between the winding and the shield.

/.i
i.,.~:.i',ii I ~!~i!i:.
......

~...~

Fig. 14.3.2D Aluminium shields

Since no iron is introduced into the flux path, the nonmagnetically shielded
reactor can be considered as being of a constant ohmic reactance over very wide
limits of current.
This is the important advantage of the copper-shielded reactor but against this
is the relatively high loss in the shields when compared with the magnetically
shielded type. Again, compared with the magnetically shielded reactor, the copper-
shielded type gives a lower reactance for the same size of coil and requires a larger
clearance between coil and shields, with the result that it is usually bigger than an
equivalent reactor with magnetic shielding.

14.3.3 Reactor rating

The overcurrent factor of a reactor is defined as the ratio of the symmetrical r.m.s.
through-fault current to the rated current. When this ratio is small the thermal
rating of the reactor is determined by the normal throughput conditions, calculated
usually as for a transformer but making allowances for more widely spaced
186 Protection of motors, reactors, boosters and capacitors

conductors giving a less steep gradient between oil and winding temperatures and
permitting the use of a somewhat higher current density.
When the overcurrent factor is large, the current density is determined by the
short circuit conditions. There is no British Standard exclusive to reactors but
B.S. 171" 1970 should be used where it applies. It specifies that reactors shall be
designed to withstand, without damage, the electromagnetic forces due to a current
having a peak value equal to 2.55 times the r.m.s, value of the rated current
multiplied by the overcurrent factor; and also the thermal effects corresponding to
the specified overcurrent conditions. Table 13 of that Standard lays down the
maximum current densities depending upon the duration of the overcurrent.
=

SERIES CURRENT-LIMITING REACTOR

T O B R I T I S H S T A N D A R l ) 171 • 1970
u _

Reactive kVA T y p e ~f" cooling

System voltage Frequency m

Rated ~urrent Amp Reactance/phase o h m

Overcurrent factor Oil
Overcurrent time sec Weight ~)f (ill
Number of phases t'()re & wdgs.
I)iagram drg. No. T()tal ~'eigh t
Maker's serial No. Year of manufacture

Fig. 14.3.3A Rating plate for series curren t-limiting reactor

A2 B2 C2

I I 1' l
AI B1 CI

Fig. 14.3.3B Terminal marking plate for series current-limiting reactor

A typical rating plate and terminal marking plate for a series current-limiting
reactor are shown in Figs. 14.3.3 A and 14.3.3 B.

14.3.4 Reactor application

The most commonly used positions for the application of current-limiting reactors
are in the generator connections to the circuit breaker, between adjacent sections
of busbar, and in series with the feeders. Their effectiveness as regards the mainten-
ance of busbar voltage and the safeguarding of switchgear under fault conditions is
the dominant factor in deciding their location.
 Protection of motors, reactors, boosters and capacitors 187

Generator reactors: Most modern generators are designed and built with sufficient
inherent reactance to enable them to withstand a symmetrical short circuit across
their terminals. This was not always so and in older stations it is not unusual to
find a current-limiting reactor connected in series with each generator. The reactor
limits the current which can flow to a generator fault from the other machines
connected to the same section of busbar, and in doing so reduces the damage
sustained by the faulty generator as well as effectively reducing the short circuit
MVA to be handled by the switchgear. Thus the effect of the reactor is to confine
the disturbance and provide relief to healthy apparatus and feeders. Generator
reactors are effective whenever the machine is running and therefore involve a small
but continuous energy loss under running conditions. Fig. 14.3.4A shows a typical
arrangement.
l"eeders I,'ceders

A A
Busbar v ,w

( ; e l l e r a t o r react()rs 2

(;enerators 2

Fig. 1 4 . 3 . 4 A Connection o f generator reactors

Busbar reactors. (a) Series connected: A typical example of this method of

connection is shown in Fig. 14.3.4B and it will be seen that the reactor is inserted
directly in the busbar run. Preferably the connection of the feeders to the various
sections of busbar should be such that at full load the current circulating between
the sections is a minimum. Absolute balance between sections is unlikely to be
obtained but if this ideal can be approached undesirable losses and voltage drops
in the reactors will be kept to relatively insignificant values. They differ from
Feeders I"eeders

Busbar Busbar Busbar

section I R e a c t o r 1/2 section 2 R e a c t o r 2/3 section 3
v ,-

(;enerators I 2 3

Fig. 1 4 . 3 . 4 B Series connection of busbar reactors

188 Protection of motors, reactors, boosters and capacitors

generator reactors in that such units incur losses whenever the machine is connected
to the system whether it is operating on light load or full load.
This method divides the station capacity into several sections so that nominally
generator 1 supplies the load on busbar section 1, generator 2 the load on busbar
section 2, and so on. When a feeder fault occurs, for example on one of the
section 2 feeders, reactors 1/2 and 2/3 have the effect of limiting the current flow-
ing to the fault from generators 1 and 3.

Busbar reactors. (b) Tie bar connected: Two methods are commonly adopted and
may be classified as

(i) the star connection, as illustrated by Fig. 14.3.4C and

(ii) the ring connection, as illustrated by Fig. 14.3.4D.
Feeders

.=
Busbar v l v

2 3 Reactors

Generators 1 2 3

[ Tie-bar
Fig. 14.3.4C Star connection of tie-bar busbar reactd

Reactors
v 1. ..

(;enera tors 1 2 3

Tie-bar
, • , ,, ,, ,

Fig. 14.3.4D Ring connection of tie-bar busbar reactors

In (i) each section of busbar is connected via a reactor to a common star point, and
if the feeders and generators are suitably arranged little or no current need flow
through the reactors. If one section of busbar is out of service the other sections
remain in parallel through the reactors. This method has the obvious disadvantage
that an additional busbar, the tie-bar busbar, is required. It will be noted that there
are two reactors in series between sections and the ohmic value of each reactor will
thus be generally less than that of those required in the ring formation.
 Protection of motors, reactors, boosters and capacitors 189

In (ii) the reactcrs are connected in series between sections of busbar and closed
by a tie-bar to form a ring. The ring is broken when one section is out of service,
and there will be considerable reactance between remaining sections resulting in
poor voltage regulation. As in the star connection an additional busbar is required.

Feeder reactors: The function of the feeder reactor is to localise the voltage drop
to the feeder on which the fault has occurred and also to enable smaller and
cheaper feeder circuit breakers to be used. Compared with generator reactors, the
advantages of connecting a reactor in series with each feeder are that a feeder fault
will not seriously affect the busbar voltage with consequently less tendency for the
generators to lose synchronism and the effect of the fault will be localised. The
connection of feeder reactors is shown in Fig. 14.3.4E.

Feeder reactors

Bushar - - v

(;enerators

Fig. 1 4 . 3 . 4 E Connection o f feeder reactors

t tIt__I lt t
(;roup feeder
reactors

Busbar : '

(;enerators

Fig. 1 4 . 3 . 4 F Connection o f group feeder reactors

190 Protection of motors, reactors, boosters and capacitors

The disadvantages are that a continuous power loss and adverse effect on voltage
regulation must be accepted, and the reactor affords little relief against busbar
faults. If the total generating capacity of the station is increased it may be
necessary to increase the size of the feeder reactors to limit the higher short circuit
current, but this inherent reactance will of course have a similar effect on system
regulation as the deliberate introduction of a reactor unit.

Group feeder reactors: An alternative to the arrangement with a reactor connected

in each feeder is shown in Fig. 14.3.4F. Such units are called group feeder reactors
since several feeders are grouped on to a single reactor. Economy in the number,
and therefore the cost of reactors is the only merit of this method of connection,
and it has the disadvantage that the disconnection of the reactor for either fault
investigation or maintenance purposes deprives the system of not one feeder but
three. It is used occasionally at voltages up to and including 33 kV.

14.3.5 Reactor protection

General: Although a reactor is often similar in appearance to a transformer and

similar winding methods are employed, they differ in several other respects. Chief
among these is the single winding of the reactor and the absence of magnetising
inrush current when the unit is switched in. It is thus unnecessary, except for the
iron-cored reactor, to go to the expense and complication of biased systems with
the result that reactor protection is usually simpler than transformer protection.
In some cases the reactor may be included in the protected zone of another
system element, such as a feeder, and little, if any, additional protection would be
necessary. Examples of this are a reactor connected in a busbar run and embraced
by the busbar protection, or a reactor connected in series with a feeder and
embraced by the feeder protection. The additional protection for any important
off-immersed reactor would include a Buchholz gas- and off-actuated relay and a
winding temperature indicator, both of which are described later in this Section.
The types of fault which may occur include flashover of external bushings,
earth faults on the windings or connections (phase faults if the three phases are
mounted in one tank), core faults, interturn faults and overheating of the winding
resulting from excessive loading or failure of the cooler equipment. It is usually
considered unnecessary to protect against interturn faults since experience has
shown that such faults very quickly spread to earth and are then detected by
the earth-fault protection.

Series-connected reactors: overall differential protection: In deciding upon the pro-

tection required for a given reactor, the size and importance of the unit must be
taken into account and, as will be seen later, the arrangement and number of
terminals brought out from the winding has a bearing on the type of protection
which can be applied.
For the larger three-phase series reactors with all six ends of the windings
 Protection of motors, reactors, boosters and capacitors 191

brought out to terminals, an overall system of protection is justified, and, as already

mentioned, no biasing is required for this application. The emphasis on the number
of ends brought out stems from the need for an overall system to be driven from
two sets of current transformers, one set on each side of the reactor. The current
transformers are connected by pilot wires and the relay connected differentially, as
shown in Fig. 14.3.5A, giving protection against phase and earth faults. The six
current transformers have identical ratios and may be housed in the reactor
bushings, in the case of high-voltage reactors, or alternatively, in the associated
switchgear. In reactors for lower voltages, bushings may not be fitted and the ends
of the windings are brought out to a terminal chamber on each side large enough to
accommodate the current transformers.
Three-phase

I I C.T. reactor
C.'F.s

Relays

Fig. 14.3,5A Overall differential protection o f a three-phase series reactor

The modern relay used for the differential protection is a three-pole relay (or
three single-pole relays) of the instantaneous, high impedance, attracted armature
type. A tapped plug-bridge enables the setting to be varied, and in view of its high
impedance it is necessary to connect a non-linear resistor in parallel with the relay
and its associated components to limit the pilot voltage to a safe value during
internal faults.
The operation of this balanced system of protection is based upon the well tried
and reliable Merz-Price principle of circulating current protection. It compares the
currents flowing in on one side of the protected equipment with those flowing out
on the other. In such a system a fault occurring within the protected zone will
cause primary fault current to flow through both sets of c.t.s, in the directions
shown by the heavy arrows in Fig. 14.3.5B, the resulting secondary currents
circulating as indicated by the light arrows, the summated currents flowing through
the differentially connected relay to cause operation. When a fault occurs outside
the protected zone, primary fault current flows in the same direction through both
sets of c.t.s, as shown in Fig. 14.3.5C, and the secondary currents circulate in
directions such that there is negligible current through the relay and no relay
operation. Negligible current and not zero current, since in a balanced system of
protection there is inevitably a little 'spill' current resulting from slight inequalities
192 Protection of motors, reactors, boosters and capacitors

C.T. C.T.

~ D
w

• ~ ~

~~elay
_ , , ,, i

Fig. 14.3.5B Overall differential protection: operation on internal fault

(..'. "r.
C~.T.

<---
. ("V'%
A
F'ault~
.,j,
, . A

I Relay
, , ,, _ ,, _ |

Fig. 14.3.5C Overall differential protection: non-operation on external fault

in the c.t.s, at the two ends and, despite the use of pilot compensating resistors,
some mismatch of the pilot burdens between the point at which the relay is
connected and the c.t.s, at either side. Tests must be done during commissioning to
check that the spill current is indeed negligible, since its presence in any significant
quantity is tending towards instability of the protection under normal running or
external fault conditions.

Overcurrent and earth-fault protection: The system of overall-differential

circulating-current protection just described provides protection of the reactor
against both phase-to-phase and phase-to-earth faults, but failure to operate would
probably have serious consequences and it is prudent to add a back-up feature to
guard against such an eventuality. This usually takes the form of a conventional
inverse definite minimum time (i.d.m.t.)overcurrent relay driven from separate
c.t.s, and connected as shown in Fig. 14.3.5D. Such relays have already been
described in detail in Chapter 6.
A variation of this form of back-up protection to include an earth-fault feature
is illustrated in Fig. 14.3.5E, and consists of residually connecting the centre
element and applying a somewhat lower earth-fault setting to it. This element is
virtually identical with the overcurrent element but would have plug bridge settings
variable in seven steps between 10 and 40% compared with 50 and 200%.
 Protection of motors, reactors, boosters and capacitors 193

On small reactors, where an overall differential system is not justified, the

reactor may be protected by separate overcurrent and earth-fault systems, and this
is dealt with more fully in the following section on shunt reactor protection.
R Y B Three-phase
reactor
C.T.s

JL I.I
C.T.s for overall
O v e r c u r r e n t relay differential pr~tection

Fig. 14.3.5D Overcurrent protection as back-up to overall differential protection of three-

phase series reactor

R Y B
Three,phase
r~ac t or
C.T.s

.J'i. "~ ---- B

• C.T.s for overall

Overcurrent rela' differential p r o t e c t i o n

E a r t h - f a u l t relay

Fig. 14.3.5E Overcurrent and earth-fault protection as back-up to overall differential

protection of three~hase series reactor
The Buchholz relay: An oil-immersed reactor is not completely protected unless it
is fitted with a gas and oil operated relay. This is because a very slowly developing
core fault or incipient fault within the reactor tank will not be detected by the
other forms of protection. It is well known that an internal reactor fault is accom-
partied by gases which the heat liberates from the oil. This phenomenon is utilised
in the Buchholz relay which is fitted in the run of the pipework from the tank to
the oil conservator. Fig. 14.3.5F shows a typical relay.
With an incipient fault producing gas the upper float, or in the illustration the
upper bucket, operates when a specified volume of gas has been collected and
causes an alarm to operate. With a heavier fault requiring immediate disconnection
of the reactor, a surge of oil from the tank to the conservator operates the lower or
 194 Protection of motors, reactors, boosters and capacitors

]ampling cock

(closed)
T~~'ipcircuit
(open)

Alarm

To oil Fro m
conservator transformer

Mercury Drain
switch plug

Fig. 14.3.5F Double float gas- and oil-actuated relay (GEC Measurements Ltd)

surge float to complete the tripping circuit. Falling oil level also is detected by this
relay giving first an alarm and, if the loss continues, initiating the disconnection
before damage can occur.
One modern relay consists of a cast housing containing two pivoted buckets each
counterbalanced by a weight. Each assembly incorporates a mercury switch which,
owing to the weight distribution of bucket and counterweight, is normally inthe
open position. When an incipient fault occurs small bubbles of gas will be gener-
ated and these, in attempting to pass to the conservator, will be trapped in the relay
housing. As the gas accumulates the oil level in the relay will fall and eventually the
top bucket will be left full of oil. The bucket will not then be fully immersed and
the weight of the oil left behind will cause the whole assembly to tilt, closing its
mercury switch and completing the alarm circuit.
With a heavier fault the gas is generated much more rapidly and the displaced
oil surges through the relay causing the lower bucket assembly to tilt closing its
mercury switch and completing the tripping circuits to the h.v. and 1.v. circuit-
breakers.
Loss of oil from the tank is also detected by the Buchholz relay since the
 Protection of motors, reactors, boosters and capacitors 195

buckets will be left full of oil. This will cause first the alarm element and then the
surge element to operate, completing the alarm and tripping circuits.
The relay is mounted in a straight run of pipework sloping up at about 10 ° from
tank to conservator, as shown in Fig. 14.3.5G. The arrow on the relay must point
in the same direction as the oil flow to the conservator or the relay will not
function properly. It must be appreciated that the Buchholz relay is not a high
speed relay compared with, say, a modem attracted-armature relay. The surge float
may take as long as 0.5 s to operate if the magnitude of the fault current is limited,
and on all but the smallest units it fulfils a role supplementary to the other protec-
tion arrangements already described.
It should be noted that although the relay described is of the double element
pattern a single-element pattern is available, and in this the contacts would be
connected to give a gas alarm only.

Winding temperature protection: A reactor winding, like that of a transformer, can

withstand short periods of overloading without damage but overheating caused by
prolonged overloading or failure of cooler equipment will, if allowed to persist,
result in premature deterioration of the insulation and so shorten the useful life of
the reactor. To obtain a warning of overheating of the winding, a hot-spot tempera-
ture indicator is fitted in the manner shown in Fig. 14.3.5H. This device indicates
the temperature of the top oil and takes into account also the temperature of the
reactor winding. The first is obtained by the thermometer immersed in the top of
the tank, and the second by feeding a heater in the same pocket from a current
transformer located in the reactor winding. To obtain maximum advantage from
the reactor's ability to tolerate reasonable overloading for short periods, the
thermal time-constant of the thermometer is made to match as closely as possible
the thermal time-constant of the reactor winding.
In addition to serving as a winding hot-spot indicator the instrument incorpora-
ates mercury switches: one to initiate an alarm if the temperature should reach a
predetermined figure, 100°C is typical, and the other to complete the reactor trip
circuit if the temperature should increase by a further fixed amount, say 20°C.

Shunt-connected reactors: overall differential protection. If all six ends of the

shunt reactor winding are brought out an overall differential system of protection
can be applied using an unbiased circulating-current system as for the series-
connected unit.

Restn'cted earth-fault protection:In a shunt reactor it is not strictly necessary to

bring out all six ends, and, for reasons of economy, the star connection is some-
times made internally and only four connections brought out. It is not then
possible to apply an overall differential system of the type referred to above and
recourse is made to a system requiring only four c.t.s. This is known as a restricted
earth-fault system and, as the name implies, protects the equipment against earth
faults only. Since approximately 75% of all faults are earth faults or at least begin
196 Protection of motors, reactors, boosters and capacitors

Buchholz
relay
..=:=. /

Reactor
tank L rt~

Secondary wiring /
to trip and alarm circuits

Fig. 14.3.5G Buchholz relay mounted in pipework

as such, this is not such a serious drawback as it may seem at first, and phase faults
can be catered for with a separate relay.
The application of restricted earth-fault protection is illustrated in Fig. 14.3.5I
and employs a single pole, high speed, attracted armature relay of the type
described in Chapter 6. It is a balanced current system restricted to operation for
faults within the zone spanned by the c.t.s, and therefore unresponsive to external
faults. The setting applied should be the lowest possible compatible with stability
and should have regard to the need to protect as much of the winding as possible.

Overcurrent protection: Phase faults on reactor installations are a considerably

lower risk than earth faults and can be protected against by using a three-pole
i.d.m.t, overcurrent relay of the type described in Chapter 6, and connected as
shown in Fig. 14.3.5I. This relay is by its nature relatively slow in operation and it
must be set to discriminate with other protective equipments.

Buchholz and over-temperature protection: A shunt-connected oil-immersed

reactor would qualify for the fitting of a Buchholz relay and a winding temperature
indicator, just as in the case of a series-connected unit.
 Protection of motors, reactors, boosters and capacitors 197

-"1
('.T. I .'actor
top plate I[

L II
/ _~ Therm°meterll
/
Reactor
winding

Reactor tan
Oil

Uapil|ary
tube

Wintlillg h~t-sp~t - = - . .

ten1 pcrat ure intlic~,t~r

Trip

Mere ur.v sxvitches

Bt~urdt~n tube
Mercury s t i t c h carrier illecha nism

Fig. 14.3.5H Winding hot-spot temperature indicator for oil-immersed reactor

lntertripping: As the name implies, intertripping is a means of effecting the

tripping of a circuit breaker at a point remote from that at which a fault has
occurred. T o d o this it is necessary to transmit intelligence from the point at which
the fault has been detected to the other end or ends of the circuit and to arrange
for its reception to initiate the tripping of the appropriate circuit breaker or
breakers without reference to any other conditions prevailing at the time. Inter-
tripping finds its main application as part of the protection scheme of a transformer
connected directly to the end of a feeder, that is without an h.v. circuit breaker
local to the transformer, and as part of the protection of a reactor similarly lacking
in circuit breakers which could be opened to isolate a reactor internal fault. It
 Next Page
198 Protection of motors, reactors, boosters and capacitors

Three- pllase
reactor

C. T.

( )ve rc u r r e n t ,i ca r t h-t'a u l t
relay (i.d.m.t.) r relay

Fig. 14.3.51 Restricted earth-fault and back-up overcurrent protection o f a three-phase shunt
reactor

consists of sending a signal, initiated by the reactor protection, over a rented Post
Office pilot channel, or privately owned pilot channel if available. A carrier signal
transmitted over the power system conductors may be used in certain circum-
stances. Fig. 14.3.5J illustrates a typical application.

Overc urren t O ve rc u r r e n t
relay relay

Pilot w i r e f e e d e r p r o t e c t i o n

'
132 kV
I
II , £-I I' bINT
s
_~,.... INT
b; 132 kV

'I -fD "-11 O v e r a l l protection relay

I
I
Shunt
~actor (~ .~ I

I "~ I Buchholzrelay
I f_ I
i __ I

II "=" -- _I_~______~~
L Winding temperature indicat~r

INT Intertrip
R receive r e l a y

INT llntertrip
S send relay

Fig. 14.3.5J Diagram o f protection and intertripping for a 132 k V feeder and oil-immersed
shunt reactor
Previous Page
Protection of motors, reactors, boosters and capacitors 199

For further consideration of intertripping and protection signalling facilities and

their application reference should be made to Chapter 7.
To obtain a figure for the total time required to isolate a fault from such a
circuit it is of course necessary to add the intertripping time to the protection and
circuit breaker operating times. Thus the absence of a circuit breaker has the effect
of appreciably increasing the length of time a fault remains on the system with
consequent increase of damage at the point of fault and a greater tendency towards
system instability.
An alternative to conventional intertripping techniques for transformer or
reactor faults uses a single-pole fault-throwing switch. Operation of the overall
differential protection or the Buchholz protection trips the local circuit breaker and
initiates the closing of the fault-throwing switch to apply a deliberate system fault
between one phase and earth. The substantial fault current flowing under this
condition is detected by the protection at the remote end and tripping takes place.
It is simple and robust and functions independently of signalling channels and equip-
ment but has disadvantages in that its operation will subject the power system to
another, possibly more severe, fault condition than the one it is about to clear.
Designs have been developed suitable for application on systems having a
three-phase fault level of up to 2500 MVA at 132 kV.

14.4 Boosters

14.4.1 The place of boosters in a power system

It is the responsibility of the Supply Industry to maintain within declared limits,

the voltage at every consumer's premises. This is no small task, but, as will be seen
later, it is one which is tackled in various ways depending upon the requirements of
the system concerned. The practice has been to concentrate an increasing propor-
tion of generating stations, with the exception of nuclear stations, in a few coal
producing areas, resulting in long transmission lines and the inevitable voltage drops
between the generation areas and the load centres. Voltage adjustments are usually
made at transforming stations, but in a scattered distribution area supplied from
one supply point adjustments may be made at the remote ends of the distribution
lines.
In this chapter, the term 'booster' is not restricted to the booster transformer
but embraces other forms of voltage regulating equipment. These include trans-
former tap-changing, the moving coil regulator and the induction regulator, and
together with the booster transformer these are discussed in more detail in the
following Sections. It should be noted also that although the term booster is most
commonly used, these equipments provide facilities for reducing or 'bucking' as
well as for boosting the voltage.
A transformer with tap-changing gear performs a double function; voltage
transformation and voltage boosting or bucking by varying the setting of the tap-
changing gear. The booster provides only the regulating facilities, but has advantages
200 Protection of motors, reactors, boosters and capacitors

in size and cost and can be installed at any point in the system where voltage
regulation is required. On the other hand, the lossses in the booster make the
combined transformer and booster arrangement less efficient than the transformer
with integral tap-changing. Features of the moving coil regulator and the induction
regulator are the continuous regulation obtainable and the absence of switches and
moving contacts, thus reducing the amount of maintenance required.

14.4.2 Transformer tap-changing

In this, the most obvious method, the transformer secondary voltage is varied by
changing taps, that is by varying the number of effective turns in one of the wind-
ings, thereby changing by a small amount, and without phase-shift, the transforma-
tion ratio.
This is done by means of selector switches, and it is essential that there is no
break in the winding circuit while the selector is passing from one tapping to the
next. Since this means that there will be a short period when the selector is
connected to two adjacent tappings, it is necessary to introduce resistors, or some-
times reactors, to limit the flow of short circuit current during the changeover
period. One phase of the winding and tap-changing arrangements of a typical
275/13 2 kV autotransformer is shown diagrammatically in Fig. 14.4.2A.
Transformer tap-changing gear does not form part of this chapter and is
mentioned here only for the sake of completeness. The tap-changing equipment is
included in the zone embraced by the transformer protection and no separate
protection is required, although the presence of the tap-changer does influence the
design of the protection applied to the transformer unit as a whole.
Series t'~mm~n

275 kV -

/ Boo,st/buck s~ itch

Diverter
switch

132 RV - - - ' " ' " - - - - ' " " - ~ Transfer

resistors
] ~ Tap winding

Fig.14.4.2A Winding arrangement of one phase of 275/132 k V three-phase autotransformer

14.4.3 Booster transformers

The booster is regarded as an adjunct to the main transformer, and consists

 Protection of motors, reactors, boosters and capacitors 201

essentially of an auxiliary transformer with its secondary winding connected in

series with the main transformer, so that the output is the resultant of the
secondary voltages of the two transformers. The secondary winding of the auxiliary
transformer is provided with a number of tappings to give a variable voltage over a
limited range. Thus the boosting voltage is governed by the number of secondary
turns actually in circuit, and the regulation is adjustable in steps, rather than
continuously as with the moving coil regulator and the induction regulator. The
kVA rating of a booster may be defined as the product of the maximum change in
voltage it can produce and the load current it can carry.
A simple form of booster for single-phase circuits consists of an autotransformer
with tappings, the tap position being selected by a switch of the face-plate type. At
voltages above, say, 3.3 kV and for high outputs on-load tap-changing gear is
incorporated.
Tapped , ,

Fig. 14.4.3A Diagrammatic arrangement of simple voltage booster

If the equipment were connected as shown in Fig. 14.4.3A the tap-change gear
would have to carry the full line current. Apart from other considerations this
would make the cost prohibitive, and the usual method is to employ an untapped
boosting winding with its secondary connected in the line and its primary energised
from a tapped regulating or 'teaser' winding, as shown in Fig. 14.4.3B and more
fully in Fig. 14.4.3C. It will be seen from these diagrams that the on-load tap-
changing is performed in a circuit remote from the line whose voltage is being
controlled. Bucking facilities also are available with this equipment, and are obtained
simply by reversing the connections to the primaries of the regulating transformer,
using the switch provided (not shown in diagram). The boosting and regulating
transformers usually form a combined unit mounted in a common tank.
For lines having high power factor loads, for example distribution circuits to
domestic consumers, a simpler and cheaper arrangement can be applied. It consists
Boosting transformer

Regulating
transformer

|,

Fig. 14.4.3B Connections of single-phase booster with regulating winding

202 Protect~onof motors, reactors, boosters and capacitors

Boosting winding

Output
Y i •

Regulating
win di ng

Fig. 14.4.3C Connections of three-phase booster with regulating winding

of an autotransformer with one tapping only and a reactor connected in series with
the incoming line, as shown in the single-phase diagram Fig. 14.4.3D. Reference to
Fig. 14.4.3E will show that under no-load conditions the no-load current Io, that is
the exciting current of the autotransformer, lags nearly 90 ° behind the incoming

.
Reactor
~ ,,

A uto4 ra nsfor m e r
T
V2
V!
VII I
Fig. 14.4.30 Single-phase booster (fixed boost)

V !

V~ ! o<: V 2

Vx

10

0
Fig. 14.4.3E Phasor diagram of single-phase booster on no-load
 Protection o f motors, reactors, boosters and capacitors 203

voltage VI. The small reactor voltage drop Vx is in quadrature with Io, and the
resultant primary voltage V~' is directly proportional to II2.
V X

V !

Vt I oc V 2

V x 0

Fig. 14.4.3F Phasordiagram of single-phase booster on high power-factor load

Fig. 14.4.3 F illustrates the conditions with load power factor approaching unity.
The primary current 1is now almost in phase with the incoming voltage V, and the
voltage drop Vx across the reactor will be increased. With the increase in power
factor this drop will be almost in quadrature with VI; the resultant primary voltage
V,' being considerably higher than before but again proportional to V2. Thus by
judiciously choosing the turns ratio and reactor impedance the increase in the
voltage 112 can be made equal to any required value. This method provides up to
about 6% boost and finds application in distribution systems supplying high power
factor domestic loads.

In-phase boosting." The boosters so far described inject a voltage in phase with, or
in phase opposition to, the line-to-earth voltage, that is they add or subtract a
voltage in phase with the line voltage. This can be illustrated by considering two
points A and B in an interconnected network, the voltages at these two points being
VA and VB. If, as shown in Fig. 14.4.3G(a), the voltages are equal and in phase, the
feeder between A and B will carry no current, but if an in phase voltage booster is

VA VB VB

(a) (h)
Fig. 14.4.3G Voltage phasors (a) unboosted (t)) boosted interconnected system
 204 Protection of motors, reactors, boosters and capacitors

connected at A, increasing the voltage there to VA, a current I will flow in the
feeder. Fig. 14.4.3G(b) shows this current lagging behind the increase in voltage
V~t - VB by an angle determined from the expression tan-t X/R where X and R
are the feeder reactance and resistance values, respectively. The value of X is usually
large compared with R so that, for most practical purposes, the current will lag by
an angle approaching 90 °. Thus the effect of the in-phase booster is to introduce a
lagging current, and so control the flow of reactive kVA (kVAr) in the system. The
power component of current is thus not significantly affected by in-phase boosting.

Quadrature boosting: It is sometimes desirable to modify the phase of the line volt-
age without substantially affecting its magnitude, and this is achieved by combining
with the line voltage Y~ another voltage II2 lagging or leading it by about 90 °. The
resultant voltage has undergone a phase change compared with the original line
voltage, the direction of the phase-shift depending upon whether the injected
voltage is lagging or leading. Little change in the magnitude of the voltage will occur
for small angles of phase-shift. This is known as quadrature boosting and the
resultant change in the current will be nearly in phase with the line voltage, so pro-
ducing a change in the transmitted power. Thus the effect of the quadrature
booster is to provide control of the power flowing in the system.
Considering Fig. 14.4.3H, the diagram for one phase, it will be seen that the

Booster transformer
R , , ,, -- -- -~A,,,AmI~j

Regulating
transformer

B ,

Fig. 14.4.3H Connections of quadrature booster (only one phase shown)

primary winding of the red phase boosting transformer is energised from the
tapped secondary winding of the regulating transformer connected between the
other two phases. As the voltage between these two phases is 90 ° out of phase with
the red 'phase-to-neutral' voltage, the effect of the boosting voltage is to change the
phase of this 'phase-to-neutral' voltage. For a complete three-phase quadrature
boosting equipment, similar connections for two additional sets of transformers
would be required, as indicated in Fig. 14.4.3 I.
 Protection of motors, reactors, boosters and capacitors 205

B()oster- transformer
- ~ . . . .

Y
i ] . . . . . .
..,,J"Yv'v'~_,

Regulating
transformer

1
• i

Fig. 14.4.31 Connections of three-phase quadrature booster

14.4.4 The moving-coil regulator

Introduced many years ago by Ferranti Ltd., the moving-coil regulator has its
windings wound on a two-limbed core much as in standard transformer practice.
The laminated iron core carries the fixed coils a and b, one at either end of one
limb of the core, and a short-circuited moving coil s which is free to move up or
down the limb, over the fixed coils, as shown in Fig. 14.4.4A. The moving coil is
isolated electrically, that is short-circuited on itself, and so does not require connec-
tions or slip rings.

! I

Fig. 1 4 . 4 . 4 A Ferranti rnoving¢oil regulator

206 Protection of motors, reactors, boosters and capacitors

The effective impedance of coils a and b is governed by the position of coil s. As

coil s is moved closer to coil a the lower the effective impedance of coil a will
become until coil s is in the position shown in Fig. 14.4.4B, when the effective
impedance of coil a will be a minimum and that of coil b a maximum.
If a voltage is applied across coils a and b in series, the greater part of this voltage
will appear across coil b, as shown in Fig. 14.4.4B. Conversely, with coil s in the
position shown in Fig. 14.4.4C, the impedance of coil a will be a maximum and coil
b a minimum, and the greater voltage will appear across coil a.

"

Fig. 14.4.4B Moving coil in upper position

t
E al
I 95
V~Jlts

I
ii
+.
II
1! 'li- I Volts

Fig. 14.4.4C Moving coil in Iower position

Since the core iron is run at a very low flux density the iron loss is small, and in
order to reduce the current in the moving coil, coils a and b are connected in
opposition, so that the currents they induce in the moving coil are to a large extent
neutralised. The resultant force on the moving coil is therefore very small and at
no-load is zero for any position of the regulator. The current in this coil when the
regulator is on-load is proportional to the load and remains independent of the
regulator position. Thus the power required to operate the regulator in very small
and even large regulators can be operated by hand.
The voltage values quoted in the figures are for illustration purposes, and for use
in a practical system a voltage variation of 25% would be adequate in most cases.
Variations of this order can be obtained by using the regulator with an appropriately
 Protection of motors, reactors, boosters and capacitors 207

tapped transformer, as in Fig. 14.4.4D, or by adding windings on the regulator, as

in Fig. 14.4.4E and Fig. 14.4.4F.
1 ~ov[ ----i
I1() V I M(~vingc()il "" i I
i I
"-L I a 20 V I
"'2 I

'l
I d
I

'D l
I _ q
4
I l
9

ql I
q h 20vl 1
I I .
ql , IliOV I t~O V
,q
. 1 .J out ut
91 V I ;)tltpUt p90V
.._./
l r ; i i1s I'() r !11e r

1
Fig.. 14.4.4D Movh~g~:oil regulator connections when used with fixed-tap transformer

,of,
0 a 90 a
I00 V 100 V
output output
1O0 V 90 V

100 b 1 sl0 b

, w v

(a) (b)
Fig. 14.4.4E Connections of moving-coil regulator for boosting

In the first case the regulator is connected across part of the transformer winding,
and by a careful choice of tapping the required voltage variation can be obtained.
The range shown gives a variation from 90 V to 110 V for one tapping position:
anther tapping would give a different range.
The additional transformer is undesirable and the more usual method is to fit
additional coils on the regulator. In Figs. 14.4.4E (a) and (b), coil r has been
added at the top of the limb over coil a, and we will assume that coil a has nine
times as many turns as coil r. It will be seen that the output voltage is maintained
at a constant value of 100 V when the input voltage varies between 90 V applied in
the maximum boost position and 100 V applied in the other extreme position.
Reversing coil r would have the effect of subtracting its secondary voltage from the
input voltage instead of adding to it.
208 Protection of motors, reactors, boosters and capacitors

I
E:a
100 V 100 V
ut output
IIOV 90 V ,

if'io
(a) (b)

Fig. 14.4.4F Connections of moving-coil regulator for (a) bucking and (b) boosting

To provide both buck and boost, another coil I is mounted at the bottom of the
limb beneath coil b, as shown in Figs. 14.4.4F(a) and (b). Assuming that coil b is
wound with eleven times as many turns as coil 1, then, with 110 V input and the
moving coil in the minimum boost (maximum buck) position, the output voltage
will be 110 - 10 = 100 V (Fig. 14.4.4F(a)). If the moving coil is moved to the
maximum boost position, the input of 90 V will be boosted to give an output of
100 V (Fig. 14.4.4F(b)). Thus with the turns ratio quoted for coils r and l, the
output voltage can remain constant at 100 V over input voltage variations of -+10%.
One significant advantage of the moving coil regulator over the induction
regulator is that by choosing appropriate turns ratios for coils r and I any value of
buck or boost, including all boost and no buck, can be obtained. The induction
regulator provides facilities for equal values of buck and boost.
One application of the moving coil regulator is as an alternative to on-load
tap-changing gear on a distribution transformer, partly because it is economic and
partly because it is more effective to apply the regulation at some point along the
distributor. When used for this purpose it is usually automatic in operation. A
voltage operated relay or contact-making voltmeter is used to initiate operation of
the regulator in the correct direction when the voltage varies by, say, +-1% from the
required value.
A relatively recent development of the moving coil regulator for control of pro-
cessing work incorporates the means of releasing the moving coil instantly, so re-
ducing the voltage to zero in under a second. Thus loads can be switched on or off
without recourse to switches or contactors.
Figs. 14.4.4G, 14.4.4H and 14.4.4I show views of typical single-phase and three-
phase regulators.
 Protection of motors, reactors, boosters and capacitors 209

Fig. 14.4,4G Core ~nd coil assembly Fig. 14.4.4H Core and coil assembly of
of lOOk VA, single-phase, 4 1 5 - 3 7 3 / 4 ! 5 V 1 0 0 0 0 k V A , three-phase, 11000/10345 - 11 6 0 5 V
automatic moving-cc~il voltage regulator automatic moving-coil voltage regulator
(Ferran ti Ltd.) (Ferran ti L t d )
2 I0 Protection of motors, reactors, boosters and capacitors

Fig. 14.4.41 Core and coil assembly of part of 14 O00kVA, 13 O00V moving~:oil reactor
bank for h.v. cable testing equipment (Ferranti Ltd.)

14.4.5 Tile induction regulator

An induction regulator is designed in much the same way as a three-phase, wound-

rotor induction motor, with the exception that the rotor is locked to prevent it
turning under the electromagnetic forces acting on it. The angular position of the
rotor with respect to the stator can be varied by hand. Fig. 14.4.5A shows the
method of connection, from which it will be seen that the stator or primary
winding is connected across the supply and the rotor or secondary winding is in
series with the line whose voltage is being controlled.
When the stator is connected to the supply a rotating magnetic field is set up
which induces an e.m.f, in each phase of the rotor. This induced e.m.f, is of constant
magnitude and, since it depends only on the strength and speed of the rotating field,
it is independent of the rotor position. Although the position of the rotor does not
 Protection of motors, reactors, boosters and capacitors 211

Statl)r x~illding_.,~ ~~_ Rc:t~Jr,v~indillg

l"x
In p
tit
A ~ .,,
( )utput
1 ,,,,,,,,____

Fig. 14.4.5A Connections of three-phase induction regulator

V I/

()

Fig. 14.4.5B Phasor diagram of one phase of three-phase induction regulator

affect the strength or speed of the rotating field it does affect the phase angle
between the rotor induced e.m.f, and the stator voltage.
The connectior, of the secondary windings in series with the line results in a
total line voltage equal to the vector sum of the incoming voltage and the voltage
induced in the secondary. This is illustrated in Fig. 14.4.5B, in which V1 represents
the supply voltage, V2 the boosting voltage, and V the resultant voltage. Altering
the position of the rotor has the effect of rotating the phasor 112, which remains of
constant magnitude whatever the position of the rotor. A phase-shift is introduced
between the supply voltage V1 and the resultant voltage V for all positions except
when the boosting voltage II2 is exactly in phase with V, the maximum amount of
phase-shift depending upon tile maximum boost of the regulator concerned. From
Fig. 14.4.5B it will be seen that, when the rotor position is such that the secondary
voltage is in phase with and in the same direction as the supply voltage, the resultant
voltage is equal to the sum of the voltage VI and I/'2. With the rotor moved to the
opposite position, so that the secondary voltage is in antiphase with the supply
voltage, the resultant is equal to the difference of the voltages VI and V2. Inter-
mediate positions of the rotor will, of course, give intermediate resultant voltages,
so providing continuous unstepped adjustment of the resultant voltage.
212 Protection of motors, reactors, boosters and capacitors

The phase-shift of the line voltage is not of great practical importance when the
induction regulator is associated with a line feeding an isolated section of the
system, but is an embarrassment if the line forms part of a network interconnected
with another source of supply. This is because the phase-shift gives rise to an
exchange of current between the sources and to differences in the magnitude of the
resultant voltages. Thus the two sources of supply which normally have equal
operating voltages will not exchange current as long as these voltages remain in
phase, but will do so if a phase-shift is introduced, for example, by the installation
of an induction regulator. In such cases a double regulator with two rotors is used,
each unit contributing one half of the total voltage added to each phase. Of these
equal boosting voltages one is arranged to lead and one to lag so that the combined
phase-shift is zero and the resultant voltage is always in phase with the supply
voltage. Fig. 14.4.5C shows the main connections for this equipment and
Fig. 14.4.5D the operating principle in phasor form. VI is again the supply voltage
and I:2 and V3 are the individual boosting voltages. Mechanical coupling of the
two rotors ensures that I:2 and V3 are inclined at the same angle to V1 and the
boosted voltage V is always in phase with VI.
Stator R~)tor Stator Rotor
winding i winding I ~vinding 2 winding 2

Input m - , Ji eYV'v'~ ! , , ~ - Output

~ ,

I C""'-
Regulator I Regulator 2
Fig. 14.4.5C Connections of double three~ohase induction regulator
v

Vs(

0
Fig. 14.4.5D Phasor diagram o f one phase o f double three-phase induction regulator
 Protection of motors, reactors, boosters and capacitors 213

14.4.6 Protection of boosters

It was mentioned in Section 14.4.1 that the term 'booster' is not limited to the
booster transformer but includes also transformer tap-changing, the moving-coil
regulator and the induction regulator. Of these the tap-changing equipment is
invariably included in the zone covered by the transformer overall differential pro-
tection, and for this reference should be made to Chapter 12. The protection
requirements of the remaining types of booster can best be described by consider-
ing a typical scheme of protection for a booster transformer.
Let us consider a booster installed at one end of a 132 kV overhead line for the
purpose of modifying the phase-angle of the line and improving the load sharing
between this and other parallel-connected feeders. If the line is assumed to have a
phase-angle of 70 ° lagging, the booster will have been designed to have a phase-
angle of 70 ° leading, so enabling the overhead line to carry its full wattful load. A
typical rating of such a booster would be 90 MVA at 132 kV, and the primary
connections would be as shown in Fig. 14.4.6A. In Fig. 14.4.6B the location of the
necessary current transformers is shown in, or adjacent to, the local 132 kV circuit
breaker, on the line side of the booster and in the booster earth connection.
Additional currenl transformers are shown for feeder distance protection (Z)and
the back-up overcurrent protection.
The protection of the booster may consist of restricted earth-fault protection
driven from current transformers in the circuit breaker, in the line and in the earth
connectit.a. The restricted earth-fault relay would be of the high speed, attracted
armature type described in Chapter 6. With such protection, sensitive to earth faults
only, it is necessary to add protection against phase faults. This can best be

70 ~

L=J L..J

132 kV I 32 k V

Fig. 14.4.6A Connection o f a booster transformer

provided as shown, by utilising the overcurrent relay which would be required, in

any case, for feeder back-up protection. This relay will therefore provide back-up
to the high-speed distance feeder protection and will afford protection against
phase-faults on the booster or on the connections between the circuit-breaker and
the booster.
When a booster is earthed via a neutral earthing resistor it is necessary to protect
the resistor against the effects of uncleared earth faults elsewhere on the system.
214 Protection of motors, reactors, boosters and capacitors

I _L .L I I I

I : I
I I

L_I
I
I
132 kV I I 32 kV
I
I
I
I
I
I
I I
I I
I -- I
L . . . . . . . . J
Fig. 14.4.6B Diagram o f protection and intertripping for a 132 k V feeder and oil-immersed
booster transformer

This is conveniently and adequately achieved by a standby earth-fault relay driven

from a current transformer located in the resistor earthing connection, as shown in
Fig. 14.4.6B.
As in the case of the off-Idled reactor and, of course, the conventional trans-
former, the booster requires protection capable of detecting faults within the tank,
preferably before they develop to the stage in which severe damage will be caused.
The only relay capable of detecting incipient booster internal faults of this type is
the Buchholz relay of which one type is described in Section 14.3.5 of this chapter
and another type in Chapter 12.
Again, as with reactors and transformers, a winding hot-spot indicator would be
fitted to provide a warning of any overheating of the booster windings. The two
mercury switches incorporated are arranged to provide first, an alarm to indicate
that the winding temperature has risen to a certain value, and second, a signal to
initiate tripping of the circuit-breakers if the temperature rises by a further fixed
amount. The instrument is described in Section 14.3.5 and shown in Fig. 14.3.5H
of this chapter.

Intertripping: Since the booster is connected solidly to the feeder, faults within the
booster can only be cleared by opening the local circuit breaker and the circuit
breaker at the remote end. The local end can obviously be tripped direct from the
 Protection of motors, reactors, boosters and capacitors 215

booster protection but intelligence has to be sent from the booster protection to
'instruct' the remote end to trip. As explained in Section 14.3.5, this is known as
'intertripping' and is usually effected by signalling over a Post Office line rented for
the purpose. The application of intertripping is shown in Fig. 14.4.6B, the restricted
earth fault, Buchholz and distance protection all being connected to initiate the
sending of an intertrip signal in the event of the operation of any one of them.

14.5 Capacitors

14.5.1 Capacitors in an interconnected power system

The transmission system in England and Wales is a closely interconnected one

providing facilities for the bulk transfer of power from centres of generation in the
coal producing areas to other large load centres, for interconnection between
centres with both generation and load, and for the supply of power in bulk to the
distribution companies for distribution to their consumers. Although an unloaded
transmission or distribution line is capacitive in character a fully loaded line has
inherent inductive and resistive characteristics. On main transmission lines
operating at near unity power factor, the power which can be transmitted is
determined largely by the system stability limit (the system will remain stable
provided the phase angle between the voltages at the sending and receiving ends
does not exceed a certain critical value). On distribution networks, especially those
operating at low power factors, the inductive reactance makes a major contribution
to the voltage drop, and it is voltage drop considerations which frequently dictate
the amount of power which can be distributed.
The effect of series inductance in a power network is to cause

(a) a phase shift between the different parts of the network tending towards
instability at an undesirably low power level,
(b) excessive voltage drops between the ends of feeders, and
(c) unequal sharing of the load between parallel feeders, thus limiting the total
power which can be transmitted.

Of these the stability problem affects primarily the transmission system, and the
regulation problem the distribution system. Inequalities in load sharing of parallel
feeders pose problems for both systems. When such problems arise it is sometimes
possible to postpone costly system reinforcement by installing reactive power
compensation equipment, although it is necessary to study each case on its merits.
Overall improvements in operating conditions will be brought about if means
are introduced of reducing the system reactance or of reducing the phase angle
between the system current and voltage. This can be done by installing static plant,
namely series or shunt connected capacitors, shunt connected reactors or rotating
plant in the form of large synchronous machines capable of being over-excited to
generate reactive power or under-excited to absorb reactive power, that is operating
216 Protection of motors, reactors, boosters andcapacitors

as shunt capacitors or shunt reactors respectively, depending upon the system

requirements at the time. These machines are operated as capacitors (overexcited)
during periods of heavy active load transfer, and as reactors (underexcited) during
periods of light load transfer.
With a static capacitor or reactor bank the reactive power can be varied only by
switching sections of the installation in or out as required, thus altering the
capacitance in steps. With a synchronous machine immediate variation is obtained
by automatic variation of the excitation, but disadvantages are complexity and cost,
the latter being approximately three times that of static plant of equivalent rating.
The remainder of this Section is devoted to consideration of the application of
static series and shunt capacitors and synchronous shunt compensators together
with measures necessary to protect them.

14.5.2 Series-connected capacitors

The simplified equivalent circuit of a transmission line with series connected

capacitor is shown in Fig. 14.5.2A and the associated phasor diagram in Fig.
14.5.2B. Fig. 14.5.2C shows how the voltage drop increases with the distance from
the sending end and how the voltage rise is concentrated at the capacitor.

1 RL XL XC
II
o;=_. ...... ,, ..... + - ,

Fig. 14.5.2A Simplified equivalent circuit of transmission Ih~e with series capacitor connected
at receiving end

IX L
IR L

Voltage rise /
: VC sin 0 j j ~ / ~ / Vs

IX C - V C

Vr

Fig. 14.5.2B Phasor diagram of equivalent circuit

 Protection of motors, reactors, boosters and capacitors 217

I . . . . . l.)istaticc | l_~ad

ii ,

Fig. 14.5.2C Diagram showing voltage drop with distance and effect of series capacitor

In the Figures,/~ and XL are the total series resistance and inductive reactance
in ohms, X c is the ohmic reactance of a series capacitor, Vs and Vr' are the sending
and receiving end line voltages, ~ is the voltage on the load side of the capacitor,
and I is the load current.
Neglecting the effects of line charging current, or shunt capacitance, and
considering lumped values for the line constants, as shown in Fig. 14.5.2A, the
phase-to-neutral w)ltage drop between I/s and I/r can be expressed as

voltage drop = I [RL cos 0 + (XI. - Xc) sin 0 ]

The conditions XL < Xc, XL = Xc and XL > Xc are referred to as overcompensation,

full compensation and undercompensation, respectively.
Overcompensation is not used in transmission systems, but applications of this
type are sometimes found in distribution systems. Examination of the regulation
formula above shows that by varying the quantity Xc the regulation can be
increased or decreased at will, provided that sin $ :# 0. As $ is the phase angle, the
condition sin $ = 0 represents unity load power factor.
The compensation of the line reactance obtained when a series capacitor is
applied has the effect of
(a) increasing the line carrying capacity,
(b) improving the load sharing of two or more parallel connected feeders, and
(c) improving voltage regulation chiefly at the point of installation.
The voltage change obtained takes the form of a sudden rise at the capacitor
terminals so that its beneficial effects are felt on the load side of the capacitor.
Thus the load voltage l/r will be larger than the received voltage ~'. Since the
voltage rise is dependent on the load current and power factor (that is voltage rise
VC sin 0 = IXc sin 0) a change in I produces a change in Vc and the capacitor acts
automatically as a voltage regulator.

14.5.3 Shunt-connected capacitors

The simplified equivalent circuit of a transmission line with shunt connected

capacitor is shown in Fig. 14.5.3A and the associated phasor diagram in Fig.
218 Protection of motors, reactors, boosters and capacitors

14.5.3B. Fig. 14.5.3C shows that, unlike the effect with a series capacitor, the
voltage rise is distributed uniformly along the length of the line.
RI. XL IL

XC Vr

Fig. 14.5.3A Simplified equivalent circuit of transmission line with shunt capacitor at
receiving end
v.. IC_ XL
IL'R 1.

Vr
irXL

- - I

Voltage drop due

I(, to load current

Fig. 14.5.3B Phasor diagram of equivalent circuit

Load
Capacitor ~ point
voltage rise ~ . ~ . , J
I
I .
J

0 -~,~ voltage drop e

I

.7o!tag.e
... drop due / "~_ I
to h)ad current

Fig. 14.5.3C Diagram showing voltage drop with distance and effect of shunt capacitor

The reduction of the phase angle between the voltage and the current, which is
obtained when a shunt capacitor is applied, has the effect of
 Protection of motors, reactors, boosters and capacitors 219

(a) reducing line current losses owing to the generation of reactive power,
(b) reducing the transmission line current to a value less than the current in the
load (Fig. 14.5.3 B),
(c) improving the power factor of the transmitted power, and
(d) reducing the voltage drop uniformly along the length of the line (Fig.
14.5.3C). This should be compared with the step-function voltage rise across
the terminals of the series capacitor.

The response to voltage dips is not as rapid as with series capacitors since the
switching of sections of the bank is initiated by the change in voltage, and at best,
the regulation is in steps.

14.5.4 Series or shunt connection

From what has already been said, it will be appreciated that series and shunt
capacitors serve a broadly similar purpose in improving operating conditions, the
series connection by compensating for line reactance, that is reducing it, and the
shunt connection by compensating the phase angle, that is reducing the phase
diplacement between system current and voltage. Clearly, there are several factors
governing the choice of connection and these may be summarised as follows.
With a series capacitor the regulation, or reduction in voltage drop, achieved
depends mainly on the reactive power of the load, and it follows that this type of
capacitor is of little use unless the conditions are such that reactive power is
consumed by the load. With a shunt capacitor the regulation achieved depends
mainly on the reactance of the system, and a useful increase in voltage will be
obtained only if the reactance is substantial.
If continuous and automatic voltage regulation is the main objective, a series
capacitor installation is most likely to satisfy the need. On the other hand a shunt
capacitor bank with means of automatically increasing or decreasing the number of
sections of the bank in service will provide a measure of regulation, albeit in steps
and with some delay.
A capacitor connected in series with a line must have a current rating equivalent
to that of the line, the output of reactance power from the capacitor being
dependent upon the line current
PXc
(output = - - - - - k V A r per phase).
10 3
With a shunt connected capacitor the output of the bank is independent of the
current, compensation being determined by the applied voltage.
On transmission circuits there are technical advantages in siting a series capacitor
at some point in the line route, but this introduces the need to acquire a suitable
site and problems of maintenance. A shunt capacitor, however, can conveniently be
connected to the low voltage side of a grid transformer situated at a substation and
need be insulated only for the lower voltage level. The cost per kVAr installed of
220 Protection of motors, reactors, boosters and capacitors

either series or shunt capacitors varies little over a wide range of capacity, and
recent figures for losses indicate the negligibly low level of 3 W per kVAr.
In the mid 1960s, the CEGB embarked on a programme of providing reactive
compensation at appropriate points throughout the transmission system and in
the short term this took the form of hydrogen-cooled synchronous machines of up
to 60 MVAr rated output at 13 kV, operating as referred to in Section 14.5.1 of
this chapter. As an immediate measure, a few such machines rated at 40 MVAr
were commissioned at 275 kV substations to be followed soon after by shunt-
connected static capacitors at other 275 kV sites (see Fig. 14.5.4A for a typical
example of the static shunt capacitor arrangement). An important feature of these
equipments, whether rotating or static, was their transportability and the value of
this facility was demonstrated when it subsequently became necessary to deploy
them at other sites as dictated by the operating requirements of the system. Since
that time all new autotransformers have been supplied with 13 kV tertiary windings
suitably rated for the possible connection of compensation equipment of up to
60 MVAr.
2 7 5 kV
a

[1

Earthing
switch
.¢.

180 M V A
2 7 5 / I 32 RV
a uto-transformer
P~

20 M V A r shunt
capacitor

I 32 kV

Fig. 1 4 . 5 . 4 A Primary arrangement of typical shunt capacitor

Several series connected capacitors have been installed in Area Board 11 kV
circuits for the purpose of reducing 'flicker', particularly in the vicinity of steel-
works and arc furnaces. The duty cycle during the melt at an arc furnace is
particularly onerous, and it is in such areas that applications for small series
capacitors are most likely to be found in this country.
Abroad, notably in Sweden and the USA, it is the practice to install large banks
of series capacitors in the very long, high voltage transmission lines common in
those countries.
 Protection of motors, reactors, boosters and capacitors 221

14.5.5 The capacitor unit

A capacitor unit (See Fig. 14.5.5A for illustration of typical unit)consists of a

number of small capacitor elements in a single case suitable for assembling with
other similar units to form a capacitor bank, or tank-type capacitor complete with
terminal bushings and protection.
Initially it was the practice in Great Britain to make capacitors of the tank type,
and examples of these are to be found at Kenfig (950 kVAr, 33 kV, installed in
1942) and Uanelly (7500 kVAr, 33 kV, installed in 1951). Following the establish-
ment of a need for large high-voltage capacitors on the transmission system, the
unit type built up to form a capacitor bank has found favour and such installations
are in service at Fleet, West Weybridge and Elstree. These are all shunt connected
in the 132kV side of 275/132kV autotransformers and are of 40, 20 and 20 MVAr,
respectively.
Oil i m p r e g n a t e d Hermetically
paper sealed c o n t a i n e r

Metal foil Metal fi~il

connec~,ed to ct~nnected t~
container terminal

Fig. 14.5.5A Arrangement of typical ASEA foil cooled capacitor unit

Each element consists of tile capacitor plates or foils with interleaving tissues of
special paper, the whole wound up to form a cylindrical roll. In a tank-type
capacitor a number of such elements are mounted on wooden supports and
connected in series-parallel to give the required kVAr output. The tank is then
Fdled with an impregnating fluid to ensure that all the interstices of the paper
tissues are filled and moisture excluded. Unit type capacitors are usually of
rectangular metal-box construction and employ similar elements in flat form, a
relatively small number being housed in each unit.
One interesting design of the unit type is made by the ASEA Company of
Sweden. The container is circular about 1 ft in diameter and 6 inches deep,
excluding cooling fins, and contains seven rolled elements mounted in star form,
 222 Protection of motors, reactors, boosters and capacitors

Fig. 14.5.5B General view of 20 MVAr, 132kV capacity bank (British Insulated Cal/enders
Cables Ltd.)

that is, six arranged around one centre element. The impregnant used is a mineral
oil, under a low pressure, to improve the dielectric strength of the impregnated
paper.
The unit construction has the advantage of greater flexibility, in that units can
be assembled on open-type outdoor construction as in Fig. 14.5.5B, and to any
size required.

14.5.6 Protection of capacitors

The need for adequate protection against damage is evident when consideration is
given to the vulnerability of such equipments to system faults, for example lightning
flashovers, and the loss of compensation which would follow the removal of a
damaged capacitor from service. The loss of a capacitor would, of course, result in a
 Protection of motors, reactors, boosters and capacitors 223

return to the very conditions it was intended to prevent.

The protection may be looked upon as comprising two parts: protection of the
individual capacitor elements within a unit or tank ('internal' protection) and
protection to prevent excessive stressing of the dielectric by system disturbances
('external' protection).
Internal protection consists of detecting any abnormal unbalance of the
nominally balanced impedances of the sections of each phase bank. It will be
appreciated that because of manufacturing tolerances in the capacitor elements a
small standing ubalance is inevitable, but this is kept to an absolute minimum by
careful selection and matching of elements and, where relays are used, is allowed
for in setting the relay. Requirements of the protection are:

(a) disconnection of a faulty element before damage or case rupture occurs, and
with a minimum loss of kVAr,
(b) indication of the location of a faulty element, and
(c) non-operation when the capacitor is subjected to non-injurious system
disturbances.

External protection is necessary to protect against the heavy currents which will
flow when a short circuit occurs on the capacitor itself, or its connections, or on
that part of the system to which it is connected. The effect of such currents is to
produce correspondingly high voltages across the capacitor which, if allowed to
persist, will overstress the dielectric. External protection may be provided in several
differing forms depending upon the size and connection of the capacitor.
For small series capacitors in distribution networks, spark gaps in one form or
another are generally used and are described in more detail later in this Section.
It might be supposed that a non-linear resistor would find an application as a simple
protector for a series capacitor, but the thermal characteristics of the resistance
material used in its manufacture are such that the resistor would have to be very
large, and therefore costly, to be able to carry the fault current for times of the
order of seconds.

14.5.6.1 Series capacitor internal protection: The choice rests between the use of
an individual fuse for each element within the unit or tank, or sometimes one fuse
for a group of elements, and the use of current or voltage transformer driven relays
connected to detect any significant out of balance between the two associated
sections. Sometimes a combination of fuses and instrument transformer driven
relays is considered desirable.
The practice of using fuses for element protection was adopted many years
ago in shunt connected capacitors and is now used extensively in series capacitors
also. The fuses are of the high breaking-capacity type incorporating a pin which
projects when the fuse has blown, or the expulsion type, from which a tail hangs
down to indicate a blown fuse; in either case the element is disconnected
permanently. The disconnection of an element in one section results in a small
224 Protection of motors, reactors, boosters and capacitors

change in the capacitance of the bank and therefore a small out of balance, of the
order of 1%, between that section and the other sections. This in itself is relatively
insignificant, but assumes importance if it is followed by the failure of further fuses,
thus increasing the stress of the remaining elements. A capacitor is designed so that
the failure of, say, one element will not result in the rest of the elements being
overstressed, that is the voltage across the remainder will be within the continuous
rating of the elements. Care must, of course, be taken to ensure that the fuses
remain intact when the external protection operates.
The introduction of current or voltage transformers brings with it the need to
divide each phase bank into two equal parallel halves, so that the detection equip-
ment can be connected across the midpoints. Under normal conditions no current
flows in the relay (other than the very small amount due to the minor difference in
individual elements), but the failure of, say, two elements would be arranged to give
an alarm and the failure of any further elements to trip the capacitor from the
system.
The cost of providing current or voltage transformers, and the space required
to accommodate them, renders this method of protection most suitable with large
installations where a case can be made out for providing a capacitor short-circuiting
switch.

14.5.6.2 Series capacitor external protection: As already mentioned, small series

capacitors such as are used in distribution circuits for voltage regulation purposes
are usually provided with spark-gap protectors to prevent damage to the capacitors
on the occurrence of faults on the capacitor or the network to which it is
connected. There is a wide variety of gaps available embracing the extinguishable
and nonextinguishable, the triggered and the nontriggered and those employing
electronic techniques. Detailed consideration will now be given to a selection of
these gaps:

(a) Nonextinguishing, nontriggered gaps: The electrodes of these gaps are of

copper or stainless steel and their useful life can be extended by adding a switch to
automatically short circuit them immediately after operation. One example of such
a gap is the English Electric Type HH device which consists of a spark gap connected
in parallel with a normally open spring-operated isolating switch (Fig. 14.5.6.2A).
The current transformer operated inverse-time relay determines the time for which
the gap is permitted to arc, and if this is exceeded the isolating switch is closed to
short circuit the gap and the capacitor. They will remain short-circuited until the
isolating switches on all three phases are re-set by means of an insulated pole.
Since these simple gaps do nothing to assist in extinguishing the arc they are
unlikely to survive more than one or two operations, and so their use is limited to
small series capacitors in places where the through-fault incidence is low. They are
sometimes applied to such capacitors in low-voltage cable networks.

(b) Extinguishing, nontn'ggered gaps: There are at least two devices in this general
 Protection of motors, reactors, boosters and capacitors 225

Spring loaded
~h(Jrt-circuititlg
~ i t c h c(~ntr~llcd by
illw,'r,~c time rcla)
___IZ "1 ./
I 1 /
/
/
/
J
l-I a n tl
Z~ resetting
lever
Inverse
time relay It--J
I
.-\d.iustab Ic
,,park g~lp

Fig. 1 4 . 5 . 6 . 2 A
C.T.

Type HH series capacitor pro tector

I
classification and these may be considered as the open type and the sealed type. A
desirable feature of any self-restoring gap is that the voltage at which it will flash
over after resetting shall be sensibly the same as the original setting, and it is
claimed that both the gaps described approach this requirement.

(i) Open type: Designed and first applied by Electricit6 de France on the
Ayergues Valley 10 kV network, this type of gap consists of two concentric copper
electrodes, the inner one carried at the top of a vertical copper spindle (Fig.
14.5.6.2B). The electrodes are so shaped that the space between them is in the form
of an annular gap slightly larger at the top than at the base. The size of this gap
determines the setting and is adjusted by raising or lowering the spindle. A coil
wound round the outer electrode carries the short-circuit current. When subjected
to a through-fault, the electrodes flashover at the base, the magnetic effect of the
coil causing the arc to rotate. As it does so it climbs in a spiral to the top of the
electrodes and continues to burn until it extinguishes itself when the fault is cleared.
The advantage of rotating the arc and moving it away from the point at which
flashover occurs lies in reducing the amount of burning at that point and helping to
maintain the setting. In service these gaps have withstood up to 50 operations
without requiring polishing or re-setting.

(ii) Sealed type: An example of the sealed type of self-restoring gap is one
designed by BICC Ltd. and developed by the then AEI Ltd (Fig. 14.5.6.2C).
Essentially it consists of a B1CC gap sealed in a glass envelope f'tlled with hydrogen
at low pressure. Adjustment of the gap is obtained by varying either the spacing of
the electrodes or the gas pressure. A coil is wound round the glass envelope and
under fault conditions carries the short-circuit current. Its purpose is to protect the
226 Protection of motors, reactors, boosters and capacitors

Electrodes

Insulation

Fig. 14.5.6.2B Open type, self-extinguishing spark gap for series capacitor protection

k gap

locating coil

Hydrogen filled
glass envelope

Series resistor

Insulator

Fig. 14.5.6.2C Sealedtype, self-extinguishing, gas-filled spark gap

 Protection of motors, reactors, boosters and capacitors 227

glass envelope by restricting the arc to the centre of the device during flashover
conditions, thus permitting the gap to be used at higher levels of fault current.
Approximately four times the fault current rating is obtained with the arc control
coil compared with the rating without the coil. Typical values are 250 A for 1 s or
500 A for 0.25 s without the coil, and from 250 A for 2 s to 2000 A for 0.2 s with
the coil. It is considered that further development up to 3000A fault current rating
is possible.

(c) Nonextinguishing triggered gaps: By 'triggered' is meant the method by which

the arc is struck on a calibrated, or trigger, gap and very quickly transferred to more
substantial electrodes where it continues to burn until the fault is cleared. An
example of this type was made by A. Reyrolle & Co. Ltd. and known as the
Spinner Gap (Fig. 14.5.6.2D). Multiple intermediate gaps in series are connected

In t e r m e d i a t e g a p s

Resist(Jr,,

Main gap

Arc -
spinner

S u p p , ~ r t inlstalatt~r

Fig. 14.5.6.2D Diagrammatic arrangement of "spinner'gap

together at one end through resistors to form the trigger element. The arc is struck
initially at the calibrated gap, the current being limited by the resistors. It spreads
to other electrodes of the intermediate gaps and then to the main electrodes when
the current ceases to be limited by the resistors. From the main electrodes it is
transferred by magnetic effects to the 'spinners' which cause the arc to rotate until
the fault is cleared by the feeder protection. The setting at which the gap flashes
 228 Protection of motors, reactors, boosters and capacitors

over can be varied within the range 3 kV to 6 kV and the current rating at
approximately 6000 A for 0.25 s is higher than with the other gaps described.
Another example of the triggered gap is that produced by the ASEA Company
and used in their large, high voltage series capacitor installations, for example at
Alfta and Djurmo in Sweden. Flashover takes place across the gap between the
spherical surfaces of metal electrodes (Fig. 14.5.6.2E) and is forced up rapidly to
the graphite electrodes by the action of electromagnetic forces and the upward flow
of air due to the heat generated. It is claimed that this design ensures rapid and
reliable ignition and that the flashover voltage remains unchanged after a number of
operations of the device.

Graphite
electrodes

Metallic
electrodes

,,4

Fig. 14.5.6.2E ASEA triggered spark gap

In these large Swedish installations this spark-gap forms but part of the equip-
ment provided to protect the capacitor. The arrangement for one phase is shown
diagrammatically in Fig. 14.5.6.2F, and in addition to the spark-gap comprises a
bypass circuit breaker, isolating switches, damping resistors, discharge reactors,
current transformers and a protective capacitor.
The protective capacitor is connected in parallel with the spark-gap to eliminate
any risk of flashover by steep-fronted overvoltage waves. This amounts to an added
safeguard since the main capacitor provides some protection against such flashover.
The damping elements reduce the stresses in the gap and the bypass breaker and
also serve to moderate the stresses in the capacitor when the gap flashes over. The
damping provided reduces the first current amplitude to about 70% of the
undamped oscillation and the second to about 40% of the first.
The discharge reactor is fitted to discharge any remanent energy in the capacitor
on the disconnection of the compensated line, thus reducing the risk of gap flash-
over when the line is reconnected. The discharge is accomplished in approximately
 Protection of motors, reactors, boosters and capacitors 229

0-3 s to permit the use of high speed autoreciosing facilities. The purpose of the
bypass circuit breaker is to short circuit the series capacitor and spark-gap in the
event of a fault in the bank or an inadvertent flashover of the gap.
Line . . . . . -

/ / I
/
capacit,Jr

I)amping [~ ~ ] Damping
cir cu it ~ (.l~se cir cui t
q

By I)a.ss , ()pen i
circuit [ "~"="~[ I l 1
breaker [ I o/c' u/c"

l'r~lcctiw: it I t
L___ --J
capacit~br
Spark gap

I)i,,charge I)i,scllarge
rcact~r rt.,~;ist4~r
()/1. ()~,erc'urret~! reJa~ U/C I_;ndercurrent rela.~
Fig. 14.5.6.2F Protection of/arge series capacitor for Swedish State Power Board
(only one phase shown)

Faults in the capacitor installation causing unbalance are detected by a relay in

each phase driven from a current transformer connected across equipotential points.
Operation of the relay initiates, via a timing relay, the closing of the bypass circuit
breaker, the timing relay having a setting of 1 sec. This delay is necessary to ensure
that operation of the relay does not cause closing of the bypass breaker on the
occurrence of faults in the compensated line, or by discharge current on bypassing
the capacitor.
The three-phase current transforn~ers drive time-lagged current relays which
initiate closing of the bypass breaker only if the spark-gap re-ignites after the
tripping, and automatic reclosing, of the circuit breakers at the ends of the
compensated line. Timing relay settings and the waiting period or 'dead time'
before the line breakers reclose are carefully chosen to prevent the bypass breaker
being closed when the line fault occurs and the spark-gap ilashes over the first time.
A continuation of the automatic sequence reopens the bypass breaker after several
seconds and if the spark-gap re-ignites again the breaker recloses and remains
closed.

(d) Electronic devices. A development by BICC Ltd. consists essentially of a

230 Protection of motors, reactors, boosters and capacitors

gas-filled spark-gap and two ignitrons, and is shown in Fig. 14.5.6.2G. The ignitrons
are connected in opposition across the capacitor terminals and are triggered by the
common spark-gap. If a fault on the load side of the capacitor is severe enough to
lgnitron

Igniter
circuit Spark
I~1 .. r ~
N

lgnitron
tl
IF '1
Discharge
resistor
CI ('2

Fig. 14.5.6.2G
i To capacitor
terminals i
Diagrammatic arrangement of electronic device for series capacitor protection

cause the spark-gap to flash over, a pulse of current flows through the igniting
circuit of one ignitron or the other, depending upon polarity. The ignitron operates
and carries fault current in addition to a capacitor discharge current during one
half-cycle of the fundamental power frequency. The sequence is repeated with the
other ignitron during the next half-cycle, the capacitor being temporarily returned to
service at each fault-current zero and permanently returned to service at the first
current-zero after clearance of the fault. The discharge resistor is provided to limit
the value of the capacitor discharge current carried by the ignitrons, and the
igniting circuit resistor to limit the ignitron igniting current.
A device of this type has been laboratory tested up to 1000 A for 1 s and is in
service at the Alston 11 kV series-capacitor installation.

14.5. 6.3 Shunt capacitor internal protection: As with series-connected capacitors,

the choice rests between the use of fuses, current transformer or voltage transformer
driven relays, or both, for the detection of unbalance between the nominally equal
capacitances of two sections of the bank. In the case of large shunt capacitors with
which this section is chiefly concerned, it is usual to fit fuses and relays to provide
complete protection against the consequences of element failure.
The fuse will, of course, merely disconnect the faulty element or group of
elements and provide an indication that it has operated. The relay operated by an
 Protection of motors, reactors, boosters and capacitors 231

instrument transformer in the midpoint connection between two sections will

usually be a 'two-stage' device, the first stage giving an alarm for the loss of one or
two elements, that is the blowing of one or two fuses, and the second stage
disconnecting the capacitor bank from the system in the event of further failures.

Relay protection: The use of relays for the detection of capacitor unbalance brings
with it the need to provide current or voltage transformers with consequent increase
in overall cost, particularly if voltage transformers are used. For this reason the
arrangements described below are usually applied only to the larger capacitor banks.

(a) Star-connected bank: (Fig. 14.5.6.3A). A voltage transformer is connected

across part of the series capacitance of each phase of the bank with its secondary
windings connected in broken delta to a sensitive relay. Any redistribution of
voltage caused by the failure of capacitor elements will be detected, but this
method, applied to a bank with earthed neutral, is liable to be affected by system
disturbances and by third harmonics.

R _1
O' I

C I

Relay

B I1 . iI_ !
II

Fig. 14.5.6.3A Protection o f star-connected shunt capacitor bank with earthed star p o i n t

(b) Delta-connected bank: (Fig. 14.5.6.3B). An arrangement of overvoltage

protection similar to ( a ) m a y be applied to the unearthed delta-connected bank
shown in Fig. 14.5.6.3 B. Again the single relay will detect any abnormal unbalance
between sections, but will be unaffected by harmonics.

(c) Split star-connected bank: (Fig. 14.5.6.3C). When an unearthed star-connected

bank is large enough to be split into two sections it is convenient to protect it
against capacitance unbalance by connecting a voltage or current transformer
between the neutral points of the sections. The two sections must be designed to
have identical capacitances or unacceptably high out-of-balance currents will flow
in the neutral. It has the advantage of low cost and is unaffected by harmonics.
232 Protection of motors, reactors, boosters and capacitors

i,t J il _ II _ II
or,, - ,, ~ ,I- I

Relay

Fig. 14.5.6.3B Protection of delta~connected shunt capacitor bank

, II
II ,II
II | _. 'I I
1
il.
II I
,I -
l
RO '

YC Relay
BC i!

II II . . . .
I| II

II II
II II

il li _
II II
Fig. 14.5.6.3C Protection o f split star.connected shunt capacitor bank
(d) Star-connected bank (two limbs per phase): (Fig. 14.5.6.3D). When each
phase is split into two equal sections the current transformers may be mounted
at either end of the phase connections, but since much less insulation is required
at the neutral end, there is a clear economic advantage in mounting them there.
This basic method was employed at the 132 kV shunt capacitor installations at
West Weybridge and Elstree, the current transformers being mounted at the neutral
end. It is very sensitive and can be applied to delta or star connected, earthed
or unearthed banks. The currents in each phase are individually compared
 Protection of motors, reactors, boosters and capacitors 233

rendering the protection inherently stable under unbalanced system conditions and,
assuming equal values of third harmonic in each section, unlikely to be affected by
such currents.
M

R
C-
) 1
_ II _1 t
.... ]
II I

I
I1-----I, " ' I

Y I
C h
i I'

I II

----II---II
B
C
l
,I-----I|
I
Fig. 14.5.6.3D Protection of star connected shunt capacitor bank

(e) Star-connected bank (two limbs per phase): (Fig. 14.5.6.3E). An alternative
to (d) is shown in Fig. 14.5.6.3E. This method of detecting unbalance may be used
when each phase is split into two equal parallel limbs, each limb consisting of a
number of units in series. The current transformer driving the relay is then
connected across equipotential points of the two limbs. It has the advantage over
(d) in requiting only one current transformer per phase.
It will be apparent that all the methods described above suffer from the disad-
vantage that they rely on the comparison of the capacitance of one section of the
bank with that of another. Thus, if elements fail simultaneously in the sections
being compared, symmetry of capacitance will be maintained and the protection
will be unable to detect the failures.
234 Protection of motors, reactors, boosters and capacitors

R
¢

II
I
A

II ' T
!
II.

BC

,, l
11 v
r
Re~la
,U
Fig. 14.5.6.3E Protection o f star,connected shunt capacitor bank

Thermal overload protection: A two-stage feature per phase is used to achieve

protection against overloads in excess of the level allowed in BS 1650:1955. These
overloads may arise from system conditions producing:

(i) increased voltage (kVAr loading is proportional to voltage squared).

(ii) increased frequency (kVAr loading is proportional to frequency),
and
(iii) harmonics, that is frequencies other than the power frequency.

The first stage consists of a thermal element having a minimum setting of 130%
nominal bank current and an inverse time characteristic matching that of the
capacitor up to 300%, and is arranged to sound an alarm. The second stage uses an
instantaneous overcurrent element which will initiate tripping if the overload
 Protection of motors, reactors, boosters and capacitors 235

exceeds 300%. The two-element relay is operated from a current transformer

located in the 13 kV capacitor circuit breaker.

14.5.6.4 Shunt capacitor external protection: The early shunt capacitors in the
British transmission system were directly connected to the 132 kV side of the
275/132 kV autotransformers but it is now usual practice to connect them, via
their own circuit breakers, to the 13 kV tertiary windings of the autotransformers.
Thus in the early installations the capacitors were controlled by the autotransformer
h.v. and 1.v. circuit breakers so that a capacitor fault resulted in the temporary
disconnection of its associated transformer. However, there is clearly some merit in
providing protection which enables operational staff to determine, from relay
operations and indications, whether the fault is on the capacitor, the transformer
or the connections between them. To this end subsequent installations have been
provided with discrete protection systems covering each of these main elements so
that if it is established that the capacitor alone is faulty, it can be isolated and the
transformer returned to service.
The remainder of this Section is devoted to a description of the protection
arrangements usually applied to star connected static capacitor installations
supplied from the autotransformer 13 kV tertiary windings. A diagram showing all
these arrangements is given in Fig. 14.5.6.4A, and Fig. 14.5.6.4B shows the d.c.
circuitry associated with them.

13 k V connection protection

The connections between the transformer tertiary winding bushings and the 13 kV
capacitor circuit breakers are protected by a differential circulating current system.
This takes the form of instantaneous high-impedance relays in the differential
circuit formed by balancing, phase by phase, a set of three current transformers in
the transformer bushings against similar sets of c.t.s of identical turns ratio in each
capacitor circuit breaker or its associated c.t. housing. The setting of the protection
should be capable of adjustment in excess of the full load current of the circuit.

Capacitor circulating current protection

The individual capacitor banks are each protected by a separate differential system.
This consists of two sets of three current transformers one set located in the
capacitor bank circuit breaker or its associated c.t. housing and the other set in the
capacitor neutral end connections, all the c.t. turns ratios being the same. The
setting of the high impedance circulating current relays is of the order of 20% of
capacitor bank full-load current. Operation of this capacitor differential protection
would trip only the 13 kV capacitor circuit breaker associated with the faulty bank,
leaving the transformer and the remainder of the capacitor bank in service.

Restricted earth-fault protection

With unearthed capacitor banks an earthing transformer and resistor are required to
236 Protection of motors, reactors, boosters and capacitors

tie the 13 kV equipment down to earth and this is teed to the 13 kV connections
between the tertiary winding and capacitor circuit breaker. A differential system of
restricted earth fault protection consisting of a single pole high impedance relay
driven from c.t.s located in the bushings or associated c.t. housing of each 13 kV
capacitor circuit breaker, and in the earthing transformer neutral connection, is
applied to cater for faults producing very low values of earth-fault current in the
earthing transformer or autotransformer tertiary winding. The relay is connected
to trip the autotransformer h.v. and 1.v. circuit breakers.

13 k V standby earth-fault protection

The neutral earthing resistor is designed to limit the earth fault current to 300 A
and it is important to ensure that it is not loaded beyond its rating. A single stage
of standby earth-fault protection is therefore provided and this also affords earth-
fault protection to the capacitor banks and back-up protection for 13 kV earth
faults. A relay with an i.d.m.t, characteristic is used and is supplied from a c.t.
located in the connection between the earthing transformer and the earthing
resistor. It is set to operate at 15% of the rated earthing resistor current with an
operating time of 2 s at this setting. The trip circuit is designed to open the
autotransformer h.v. and l.v. circuit breakers.

Overcurrent protection
To cater for the possible failure of the main differential protection an i.d.m.t, back-
up overcurrent relay is provided driven from c.t.s in the transformer tertiary
winding bushings. The relay is connected to trip all circuit breakers capable of
providing fault current infeeds.

Overvoltage protection
Under light system loading conditions the capacitor banks may be subjected to
symmetrical three-phase overvoltages. The value of the maximum permissible
overvoltage and its allowable duration will depend upon the overvoltage character-
istic of the capacitors and upon the system operating requirements.
Overvoltage protection is provided by two overvoltage relays with i.d.m.t.
characteristics one connected for alarm and the other for trip and both supplied
from the same single-phase 13 kV voltage transformer. The trip contacts are
arranged to trip sequentially the capacitor 13 kV circuit breakers at intervals of 30 s.
Subject to system operating requirements and to the characteristics of the capacitor
banks, the overvoltage alarm relay is set to operate at 130% and the overvoltage trip
relay to operate at 140% of capacitor rated voltage. Both relays would have an
operating time of about 30 s at their respective settings.

Other protection
Overloading of the capacitors may result from increases in system voltage or system
 Protection of motors, reactors, boosters and capacitors 237

frequency, or from the effects of harmonic currents. Two-stage thermal

overload protection for each capacitor bank is provided by a two-element relay
supplied from c.t.s located in the 13 kV capacitor bank circuit breaker, with
settings as described in Section 14.5.6.3. The first i.d.m.t, stage is set to alarm only
and the second, instantaneous element is connected to trip its associated 13 kV
circuit breaker.
Protection to detect capacitor out.of-balance resulting from the failure of
individual capacitor elements or the blowing of their fuses, takes the form of a c.t.
and relay scheme, the choice depending upon the size and configuration of the
capacitor bank. A number of typical arrangements are described and illustrated in
Section 14.5.6.3.
The earthing transformer is provided with a double float gas and oil operated
Buchholz relay generally of the type described and illustrated in Section 14.3.5.
The relay has alarm and trip functions, the trip contact being connected to trip all
circuit breakers capable of supplying fault current infeeds.
The tertiary winding of the autotransformer can withstand short periods of
overloading but prolonged overheating will hasten insulation deterioration and
eventually shorten the life of the transformer. To prevent this happening a winding
temperature indicator (w.t.i.), previously described in Section 14.3.5., is fitted.
Appropriate alarm and trip settings are applied and the trip contact is connected to
trip the 13 kV main oil circuit breaker.

14.5. 6.5 Protection of synchronous shunt compensators: Synchronous

compensators like the modern versions of their static capacitor counterparts are
connected to the 13 kV tertiary windings of the autotransformers. In addition to a
main 13 kV circuit-breaker two similar circuit breakers, 'start' and 'run' are
required in the synchronous compensator starting and control circuits. Again
discrete protection systems are provided for the machine, the 13 kV connections
between it and the autotransformer, and for the transformer itself. Each of these
systems is described below together with the appropriate back-up protection
features and these composite arrangements are shown diagrammatically in
Fig. 14.5.6.5A. The d.c. circuitry associated with them is shown in Fig. 14.5.6.5B.

13 k V connection protection

The connections between the transformer tertiary winding bushings and the 13 kV
main oil circuit breaker are protected by a differential circulating current system.
This takes the form of a set of current transformers in the transformer bushings
balanced phase by phase against another set in the main oil circuit breaker or its
associated c.t. housing. High-impedance relays are used and the setting of the
protection should be capable of adjustment to above the full load current of the
circuit. The c.t. turns ratio must be the same for each location.
 Protection of motors, reactors, boosters and capacitors 239

i 50AI C ( ) N I R ( ) L DIA{ R .-\ M

. I - I('S !
• " I- _. "1

(.."BI"
I%1A-O O--K31..% ~ K 3 . - \ - - ;.,¢
L'apacitq~r ~ , (':t pacifier
pr();ecti()n (:A I' Bl pr¢)tecti()n
O-- KI 3 1 A - - n - " l ~
I !)+ I)-
• (5( ~.'1] ! ~ J 2 --~
; 0 C - " K33.~
Sell rcset

c A P OB I

OV Ti)
"
C T
OV AUX ()V 1i)
- - O C K.~s .\
ii L

250AI t'()N I R O L I)i-\L;RAM

CBT
~"KIB ""0 O--K31B .--O--1~--K3 B-
1
(.'apacit(~r Lapacit(~r
protection CAP BT
~-------.-,.0., 0"- KI31B.--(3::::33
~Jl -Z ]'
~• !)+"
[ L;C ~ .t_. B_ _l _ . _
' OTFI,C i K331i Sell" reset I

•------.---O @

CAP OB [
C
- !
OV TD

OV ALIX ()V TI)

--"-------0 C 'K55B '

Fig. 14.5.6.4B Protection d.c. circuits for static compensators associated with autotransformers
240 Protection of motors, reactors, boosters and capacitors

. . . . . . . . I

protectionCapacit°rbank ~ I'SS =J - K2 II (.'apacit~)rpr(~tecti()nbank

1)+

ov
l~OO~/~ I K~ ~ u x I I
Sell" resu'l I
I

13 kV S B E I .

. . . . o_._o 1
R I.:i.
' O C ....
" i

Or.'
O C ...... KS35

13 k V L'C

IAcp+ IAcpt

, ~

('A P BT -- ('apacit~r hank trip relay

( ' A I ' O1:1 -- ('apacitance out-of-balance
CBT - t'ir~.'uit-hreaker t r i p relay
CC - Circulating current
13 kV (.'(' -- 13 k V c i r c u l a t i n g c u r r e n t
ETB - t':arthing t r a n s f o r m e r h u c h h o l z
Or' •- Overcurrent
OV -- Over voltage
OV ALIX - ()ver voltage a u x i l i a r y relay
OV T I ) Over voltage t i m e delay relay
I'SS --- Pr()tecti(~n s u p p l y supervisi(~n
R F:I" - Restricted earth-fault
13 k V SBI-:I. - 13 k V s t a n d b y e a r t h - f a u l t relay ( I D M T - 2 see.)
T('S - Trip circuit supervision
TH -- Thermal
TWT - Tertiary ~inding temperature
 Protection of motors, reactors, boosters and capacitors 241

Machine circulathzg current protection

The connections between the machine neutral, the starting transformer neutral and
the 13 kV main oil circuit breaker are protected by their own differential
circulating current system. This consists of a set of c.t.s in the 13 kV main oil
circuit breaker, or its associated c.t. housing, balanced phase by phase against two
further sets of c.t.'; located one each in the netural connections of the machine
and the starting transformer, respectively. High-impedance relays are used here also
and the differential protection setting should be approximately 20% of full load
current. Again, all the c.t. turns ratio must be the same.

Restricted earth-fault pro tection

As the machine is unearthed, an earthing transformer and resistor are required to
tie the 13 kV equipment to earth and, as shown ill Fig. 14.5.6.5A, this is teed to
the connections between the tertiary winding and the machine main oil circuit
breaker. A differential system of restricted earth-fault protection is applied
consisting of a set of c.t.s in the machine main circuit breaker, or associated c.t,
housing, and one c.t. in the earthing transformer neutral connection, it is designed
to detect low level earth-fault currents in the earthing transformer or autotrans-
former tertiary winding. The d.c. circuit of the higla-impedance relay used is
connected to trip tl~e autotransformcr h.v. and l.v. circuit breakers.

13 k V standby eartlz-fault protection

Two-stage standby earth-fault protection is provided. Stage 1 affords earth-fault
protection for the machine and stage 2 provides back-up protection for uncleared
13 kV earth faults and ensures that the neutral earthing resist3r is not loaded
beyond its rating. Both relays are supplied from the same c.t. located in the
connection between the earthing transformer and the earthing resistor. The stage 1
relay is instantaneous in operation, usually has a setting of 10% and is connected
to trip the 13 kV main oil circuit breaker only. The stage 2 relay which trips all
the h.v. and l.v. circuit breakers associated with the autotransformer as well as
the 13 kV machine main oil circuit-breaker is of the induction type with an i.d.m.t.
characteristic. This relay should have a setting of 5% and an operating time of about
2 s at twice the setting.

Overcurrent protection

To guard against tire possible failure of the 13 kV connection protection an i.d.m.t.

back-up overcurrent relay is provided supplied from c.t.s located in the bushings of
the 13 kV main oil circuit breaker or its associated c.t. housing. Tile relay is
connected to trip the 13 kV main oil circuit breaker only.

Reverse power protection

If the supply to a synchronous compensator were to be lost while its associated
244 Protection of motors, reactors, boosters and capacitors

circuit breakers remained closed, subsequent restoration of supply could result in

the machine being re-energised out of synchronism. Although this situation is
unlikely to occur at busbar type stations it is a possibility at stations with a limited
number of infeeds.
To cater for this eventuality, which could result in serious damage to the
machine, reverse power protection is usually considered necessary at stations other
than those of the busbar type. The trip circuit is arranged to trip the 13 kV main
oil circuit breaker only.
The low internal losses of the synchronous machines necessitate the use of a
highly sensitive relay for the detection of loss of power following an interruption of
supply. The voltage supply to these relays must therefore be of high accuracy if
malfunctioning is to be avoided. The phase angle errors resulting from the internal
regulation of the earthing transformer preclude its use for this purpose if there is
any likelihood of its being used to supply varying loads, such as motor starting
loads. In such cases a separate voltage transformer of high accuracy should be used.
The relay is, of course, designed to measure power and thus requires a current as
well as the voltage supply just described. This current source is provided by a set of
c.t.s located in the neutral end connections of the machine.

Other protection

To prevent the machine suffering damage from excessive thermal effects it is usual
practice to provide a thermal overload relay supplied from a c.t. in one phase of the
machine neutral end connections. Ideally, the characteristic curve of the relay
should match the thermal characteristic of the machine over its whole range: in
practice this may not be possible but it is essential that it does so over the working
part of it. The relay setting adopted will thus be based on the thermal characteristics
of the machine and the relay contacts are connected in the machine control circuit
so that following relay operation the machine excitation is reduced.
Temperature indicators with facilities for providing alarm and tripping functions
are provided on the machine bearings. The trip contact is connected to trip the
13 kV main oil circuit breaker.
The earthing transformer is provided with a double float gas-and off-operated
Buchholz relay generally of the type described and illustrated in Section 14.3.5.
The relay has alarm and trip functions, the trip contact being connected to trip
all circuit breakers capable of supplying fault current infeeds.
The tertiary winding of the autotransformer can withstand short periods of
overloading but prolonged overheating will hasten insulation deterioration and
eventually shorten the life of the transformer. To prevent this happening a winding
temperature indicator (w.t.i.), previously described in Section 14.3.5., is fitted.
Appropriate alarm and trip settings are applied and the trip contact is connected to
trip the 13 kV main oil circuit breaker.
 Protection of motors, reactors, boosters and capacitors 245

14.6 Bibliography

British Standards

BS.587:1957, 'Motor starters and controllers'

BS.4941:1973, 'Motor starters for voltages up to and including 1000 V a.c. and
1200 V d.c.'
BS.4999:Part 72 (1972) 'Thermal protection for electric motors rated at 660V
a.c. and below
BS.5000:1973, 'Rotating electrical machines of particular types for particular
applications'
BS.171 : 1970, 'Power transformers'
BS.1650:1971, 'Capacitors for connection to power frequency systems'

Books

Motor selection and application by C Libby Charles (McGraw-Hill)

The protective gear handbook by F E WeUman (Pitman)
Automatic protection ofa.c, circuits by G W Stubbings (Chapman & Hall)
Protective relays." their theory and practice by A R Van C Warrington (Chapman &
Hall)
Application of power capacitors (BICC Ltd. Publ.)

Papers and articles

'Short circuit rating and testing of reactors' by J G Wellings and R V Wheeler

(J. lEE, 1942, Pt. 2, p. 473)
'Technical and economic aspects of the supply of reactive power in England and
Wales' by W Casson and tt J Sheppard (Proc.ll;'l:', 1961, 108, Part A)
Report on the use of series capacitors in power systems (ACE Report No. 3, 1956)
'The protection of shunt capacitor banks: a critical review of published information'
by J A Nicholson (ERA Report No. Q/T136, 1953)
'Protection of series capacitors against external disturbances and internal faults' by
R Pellissier (BE Translation No. 391)
'380 kV series capacitors in Sweden' by G Jancke, K S Smedsfelt and P ttjertberg
(CIGRE Paper 322, 1954)
'Self-extinguishing gaps in large series capacitor stations' by 1. Ahlgren and B
Grundmark (CIGRE Paper 317. 1956)
'The use of series capacitors on high voltage transmission systems' by S Lalander
and L Norlin (CIGRE Paper 330. 1958)
'Series capacitors for distribution networks' by K S Smedsfelt and P Hjertberg
(ASEA Paper 7407E)
246 Protection of motors, reactors, boosters and capacitors

'Series capacitors for high power transmission' by I Herlitz (ASEA Paper 7251E)
Report on the protection of series capacitors (up to 33 kV) (Protection Sub-
Committee M2, ad hoe Panel on Capacitor Protection, 1961)
 Chapter 15
The application of protection to rural
d istribijtion systems
by J. Harris

15.1 Introduction

The growth in demand for electricity supplies in rural areas has led to rapid
increases in the extent of rural high voltage networks, and consumers in these areas
have come to demand supplies comparable in reliability to those provided in urban
districts. In view of the type of construction employed for rural electrification, this
demand has called for considerable skill and ingenuity on the part of electricity
supply engineers.
Because of the savings achieved thereby, rural feeders are almost entirely of over-
head construction, backbone lines radiating from step down transformer sub-
stations fed from the primary system and from these tappings are made to afford
supplies to small groups of consumers. A typical system is illustrated in Fig. 15.1A.
In order to take advantage of the impulse strength of the wood poles, these over-
head lines are generally of 'unearthed construction', that is the pole top steelwork is
earthed only at transformer and section switch positions and where pole type cable
boxes are installed at junctions between lines and cables. The performance of lines
with this type of construction is improved under certain conditions of fault liability
compared with the older type with all steelwork earthed, but overhead lines
generally are susceptible to faults due to lightning and other climatic conditions,
birds, cattle and a variety of additional causes. A large proportion of these faults,
amounting to approximately 80% are, however, of a transient nature and bring
about the operation of the protective device without causing permanent damage to
the system. As will be seen later, this fact enables automatic restoration of supplies
to be employed with consequent improvement in the continuity of supplies to con-
sumers fed from such lines.
Manual restoration of supplies in rural areas after interruption is often rendered
difficult by inclement weather and rough terrain and frequently involves journeys
of considerable length.
The fault power level on a rural network may be as high as 250 MVA at the
s o u r c e substation. Such a condition may cause burning through of small section
line conductors if clashing occurs during gales or storms and it may be necessary
248 The application of protection to rural distribution systems

Primary transmission line

Primary substations

Additional Additional
feeders feeders

'

Section

Section point

int (~
Section poin t

ily-open
switch

Fig. 15.1A Diagram of typical rural system

to restrict the fault level by the isolation of infeeding transformers to discrete

sections of switchgear. In such cases automatic sequence closing of busbar section
circuit breakers may be employed to restore supplies in the event of a transformer
failure.
Due to the comparatively high reactance of overhead lines, the attenuation of
fault power along the line is high and the current values resulting from faults at
points remote from the substation are considerably reduced as will be seen from
Fig. 15.lB.
 The application of protection to rural distribution systems 249

250

C~nductor,~ -- 3 2 mm 2 C~pper

150
<
>

-C

-- I00

50

O't
l !

,
Rq)ute
i !

2
length fr~in s~urce
;
- km
,,

,;

Fig. 1 5 . 1 B

In the case of earth faults, a t'urther reduction may occur due to high resistances
caused by difficult earthing conditions. Protective devices for rural lines must
therefore be capabte of withstanding the thermal effects of heavy current whilst
being sufficiently sensitive to operate on restricted currents. Furthermore,
protective equipment, such as fuses or circuit breakers, which is installed on line
supports must perform reliably after prolonged exposure to varying climatic con-
ditions. It is the provision of devices meeting these requirements, at a cost in
accord with the overall economics of rural electrification, that poses the problems
of rural system protection.
250 The application of protection to rural distribution systems

15.2 Fuses

15.2.1 Types employed

One of the oldest and simplest methods of protecting rural systems is the use of
fuses. Early designs consisted of little more than an enclosed wire and their per-
formance was somewhat erratic. A large amount of research and development has,
however, been carried out and the modern fuse is a much more reliable device.
Two types of high-voltage fuse are now generally available for the protection of
rural systems, namely

(i) Expulsion type

(ii) Powder filled (high breaking capacity) type.

The construction and characteristics of these fuses have already been discussed in
Chapter 5.
The liquid f'tlled fuse was once widely used but troubles were experienced due
to breakage of the glass and leakage of liquid. Furthermore, its application was
somewhat restricted by its limited breaking capacity as system fault levels increased.
The expulsion type is cheaper in capital cost and fuselink replacement and with
available ratings being suitable for the fault power levels normally encountered in
rural networks, its use has now superseded that of the liquid tidied fuse.
For fault levels in excess of 150 MVA at 11 kV, an adaptation of the expulsion
fuse carrier to incorporate a cartridge type powder-filled fuse link may be employed.
The construction of the expulsion type fuse readily permits the removal of the
fuse link to be used as a means of isolating the protected equipment.

15.2.2 Application

Early applications tended to fuse each spur line and individual transformer separ-
ately, as in Fig. 15.2.2A(a), the fuse rating being related to the full load current
rating of the plant protected. Whilst fuse blowing due to transformer failure was
infrequent, surges during lightning storms often resulted in widespread fuse opera,
tion without permanent damage to apparatus. This led to the concept of 'group
fusing', that is the protection of a number of transformers on a spur line by a single
set of fuses at the spur tapping point, as shown in Fig. 15.2.2A(b). This safeguards
the main line from disturbances on the spur and is the principle generally employed
today. It is usually accepted that h.v. fuses be employed for short circuit rather
than overload protection, and it is usual, therefore, to install 'group fuses' of
comparatively high rating to prevent unwanted operation due to switching surges
and l.v. faults. This, however, has the disadvantage that low current faults, such as
transformer interturn failures and certain broken conductor faults, may not cause
the fuse to operate. The number of such incidents is, however, negligible in com-
parison with those caused by lightning and, with the high reliability of modern
 The application of protection to rural distribution systems 251

distribution transformers and present day overhead line design, the principle of
h.v. group fusing is undoubtedly justified. Overload protection of individual trans-
formers is provided by 1.v. fuses.
Current flowing to a fault on the l.v. side of a transformer is supplied via the h.v.
group fuse and, under these conditions, the l.v. fuse protecting the faulty circuit
should operate first in order that the h.v. supply to healthy transformers is not

substation
Iiackh,lllv
1, i

liric

Sour~
substati,~n ~] l
l , II ]Ill I.' k I~,, I1t.'
I - line

(l>---

(b)

Spur lincs

Fig. 15.2.2A Diagrams of systems protected by fuses

252 The applicat/on of protection to rural distribut/on systems

affected. The use of high breaking capacity powder filled fuses, with their inherent
high speed of operation, on the 1.v. side, combined with the comparatively high
current rating of the h.v. group fuse, materially assists in obtaining this
discrimination.
This is shown in Fig. 15.2.2B where the h.v. and l.v. fuse protection of three-
phase 11 000/440 V transformers is illustrated. A 100 kVA transformer having
an impedance of 4-75 per cent and connected to an h.v. system having a fault

l0 • I

\,\

1 ~r," 400 amp H.B.C.l.v. fuse

1.0 ,

u~
•o 0.5
O
¢)
Oq

I
¢)

E
.,.,,
p.,

300 amp H.B.C 2S a m p fast

I.v. fuse /blowing h.v. fuse
0.1

\
0.0S
\
\
\

0.01 i. . . . •
500 1000 SO00 10000 $0000

L.V. current - Amps

Fig. 15.2.2B Discrimination between h. it. a n d I.v. fuses protecting a 11 0 0 0 / 4 4 0 V transformer

 The application of protection to rural distribution systems 253

level of 150 MVA will pass a maximum l.v. fault current of 2760 A. For all values
of fault current up to this figure discrimination will be obtained between the h.v.
expulsion type fuse of 25 A rating and the high breaking capacity l.v. fuse of 300 A
rating, tolerances on fuse characteristics being neglected. The ratio between l.v. and
h.v. fuse ratings is 12"1, although the transformation ratio is approximately 25"1,
discrimination with fuse ratings of this ratio being made possible by the high speed
of the filled type of fuse relative to that of the expulsion type.
An l.v. fuse of 400 A rating will not discriminate with the 25 A h.v. fuse at fault
currents below the maximum that can occur on the l.v. side of the 100 kVA trans-

2.0
I I 1 I
2B5 a m p
A
,s°""",,' BTS~' ip.
i

1.0

\
0.5

0.4
\
0.3 • .

.~ 0.2

¢.¢q

E
i
\
•- 0.1

0.05

0.04

0.03

0.0 2

0.01
100 200 300 400 500
_L
1000 2000 3000 4000 5000
Symmetrical current - Amps

Fia. 1 B.2.2C Discrimination between h.v. fuses in series

254 The application of protection to rural distribution systems

former but will do so at higher values of fault current which may be experienced oll
the l.v. side of larger transformers.
In order to check the discrimination between fuses in series, it is necessary to
plot the total clearing-time characteristic of the fuse remote from the source and
the prearcing-time characteristic of the fuse nearer the source to a common current
base. In the case of transformer h.v. and l.v. fuses, the current values must be
converted to a common voltage base. Discrimination will result if the characteristics
are separated by an interval of at least the sum of the permitted characteristic
tolerances for the fuses in question. Prior to the advent of BS 2692, tolerances on

1000

500

100

50

10 -
0

5
I

.E

1.0

0.5

0.I0

O.OS

0.01 - Ill

I 5 10 50 I00 $00 1000 5000 10000

Current - Amps

Fig. 15.2.2D Time/current characteristics of fast and slow blowing h.v. fuses
 The application of protection to rural distribution systems 255

the time/current characteristics of high-voltage fuses were not specified and a

simple rule-of-thumb method of ensuring discrimination between h.v. fuses was to
employ a major fuse having a current rating twice that of the minor fuse. The
tolerance on the characteristics of fuses to BS 2692 is + 20% and in the absence
of more specific information from the fuse manufacturer, this tolerance must be
allowed when comparing fuse characteristics. With certain designs of expulsion fuse,
however, greater accuracy is possible and a discrimination factor, that is minor fuse
rating/major fuse rating, of 0-75 is obtainable. Fig. 15.2.2C shows the
discrimination limits for three commonly used sizes of expulsion fuse. The lines AA
represent for each rating the published characteristic whilst the lines BB have been
plotted with the current ordinates reduced to 0.75 of the values represented by AA.
These fuses will discriminate since the shaded area between the lines tbr any rating
does not overlap that of the other ratings. Fuses of intermediate rating will not
discriminate if overlapping of their characteristics, plotted as above, occurs.
In some instances, the use of fuse links with a slower operating characteristic at
the higher currents may be required to obtain adequate discrimination. Fuse links
with such characteristics are available in the liquid filled and expulsion types but
are not available with the powder-filled h.b.c, type which is inherently fast in
operation. A comparison of tile relative prearcing times of fast and slow blowing
h.v. fuse links is illustrated in Fig. 15.2.2D.
Some types of fault on the low-voltage side of transformers employing certain
winding connections will give rise to unbalanced currents on the high voltage side
and currents in excess of those carried by a balanced three-phase fault may occur.
It is essential therefore to ensure that adequate discrimination can be obtained with
these particular fault conditions.
Experience has indicated that the time/current characteristics of overhead line
h.v. fuses change after a length of time in service and maldiscrimination may occur
between fuses in series, even though correct discrimination is indicated by a com-
parison of their published characteristics.
For this reason and because of the wide tolerances on the characteristics
sometimes encountered, the use of such fuses in series is not recommended.

1-5.3 Automatic circuit reclosing

15.3.1 Principle

As mentioned in the Introduction, approximately 80% of the faults occurring on

rural systems are of a transient nature and cause no permanent damage to lines or
plant. Such faults, however, require the operation of a protective device, for
example a fuse or circuit breaker, for clearance, and an interruption of supply
results until the fuse has been replaced or the circuit breaker reclosed. In rural areas
a long journey may be involved often at an inconvenient time and in inclement
weather, and in consequence the supply may be interrupted for a very long time. It
256 The application of protection to rural distribution systems

will be apparent therefore that an improvement in the continuity of supply to

consumers and a reduction in operating costs will result if the supply can be
restored automatically. Automatic circuit reclosing is now extensively applied to
rural networks and has resulted in a high standard of security of supplies to rural
consumers.
Of the factors bearing on successful automatic reclosing, the most significant
is the speed of operation of the circuit breaking device. If the opening of the circuit
is delayed more than a few cycles, the thermal and other effects of the fault current
may cause permanent damage at the fault position thus creating a persistent fault,
preventing immediate supply restoration. High speed fault clearance, however, will,
in most cases, prevent such damage and permit successful restoration of supplies.
The time interval between tripping and subsequent reclosure is usually referred
to as the 'dead time' and may vary over a wide range. It is influenced by the
characteristics of the circuit-breaking device, the nature of the faults likely to be
encountered and the type of load supplied. In this latter connection, differing types
of motors have conflicting requirements, synchronous machines requiring a dead
time sufficiently long to ensure operation of their undervoltage protection whilst
for induction motors the period needs to be short to enable them to coast through
the open-circuit period. The dead time must in all cases however, be of sufficient
duration to ensure deionisation of the fault path and stabilisation of the breaker
mechanism. In practice times ranging from 0-4-120s are employed but evidence
from field trials suggests that a time of 10-15s affords the best performance under
typical system conditions.
An important characteristic of reclosing schemes is that known as reclaim or
resotting time, i.e. the time following a successful reclosing operation after which
the scheme will provide a full operating sequence in the event of subsequent faults.
It is of particular significance under repetitive fault conditions, e.g. lightning storms
or conductor clashing in high winds, when rechim times in excess of the interval
between the incidence of successive faults may cause unnecessary lockout and
interruption of supply. Service experience indicates that good results are obtained
by the use of reclaim times of the order of 5s, although in the case of spring actuated
mechanisms it may be governed and extended by the spring winding requirement.
The use of single phase interruption and reclosing produces a tendency for the
voltage on the faulty phase to be maintained by the sound phases via the magnetic
circuits of three phase transformers connected to the system. Whilst this may assist
induction motors to continue to run, it may delay the disconnection of synchronous
machines. Since induction motors are the more commonly used in rural areas,
however, the scheme has been used with considerable success.
In a large proportion of non-persistent faults, estimated to be some 80% of the
total, the circuit may be successfully re-energised at the first attempt, but with the
remainder a second or third reclosure may be necessary. Considerable improvement
in the continuity of supply can therefore be obtained by the use of 'single shot'
reclosing, and any further improvement which may be achieved by multishot
reclosing must be judged against the cost of the additional equipment required to
provide this feature.
 The application of protection to rural distribution systems 257

15.3.2 Repeater fuses

Where protection of the h.v. system is by fuses, automatic restoration of supplies

may be achieved by the use of repeater fuses. These are of the drop out expulsion
type and one, two or three restorations may be obtained by the installation of the
appropriate number of fuses per phase. Only one of these is normally in circuit, and
following its operation on fault the energy of the falling carrier in isolating the
spent fuse is utilised to actuate a spring loaded changeover contact to bring the next
fuse into circuit. By the incorporation of a time lag device this action may be
delayed to ensure that any fault arc path has become sufficiently deionised to
permit the circuit to be successfully re-energised.
The disadvantages of the use of repeater fuses are the amount of equipment
which it is required to accommodate on an overhead line support and the number
of replacement elements required after clearing a persistent fault.

15.3.3 Pole-mounted automatic circuit reclosers

(a) Construction: The repeater fuse permits only a limited number of operations
before resetting becomes necessary and periodical inspections are therefore
required to ensure full availability.
A form of reclosing circuit breaker which overcame the disadvantage of this
limitation, by virtue of an operating mechanism which is self resetting if the fault is
removed before the completion of its operating sequence, was developed in
America about the mid 1940s. This type of equipment offered other advantages for
the protection of rural overhead lines, and designs are now available in this country.
Units of this type have been commonly termed 'high speed reclosers' but this
term is something of a misnomer since, although the open circuit interval is reduced
to approximately one second, their main attribute is a very high tripping speed
which minimises fault damage to apparatus and prevents unnecessary operation of

t Flexible Series
I connecti~n ~

M Alternative ~_1 ~' N~,n linear

c,,ntact 7 link I ]~~ I " resist,,r
I positions ~- ~1 I ~ f I'~. I

_ i I
'' I l
| Fixed c,,ntact
I
]
I ,
Fig. 15.3.3A Internal connections of recloser (AEI Ltd.)
258 The application of protection to rural distribution systems

associated fuses. For certain applications, however, some delayed tripping operations
may be required and units with variable tripping characteristics have therefore been
produced.
The first recloser of this type produced by a British switchgear manufacturer was
a single phase per tank design which utilised the energy of the fault current to open
the circuit-breaking contacts, and, at the same time, to charge springs which subse-
quently reclosed the contacts. It was essentially a series solenoid device, the
eletrical arrangement being as shown in Fig. 15.3.3A. The thermal characteristics
I0

8
7
6

\
\ / Delayed tri p

| m
"0 JJJ
¢:
0
0.9 - - !11!
¢J 0.8 - -
0.7 - -
¢) 0.6 - -
E
...,,
0.S - -

0.4 - -
¢)
.,,,.,
0.3 P - - -

I,,.
.m
0.2 i'--------
\
c..)

..~ Instantaneous trip

0.|
Jjjil
0.09
0.08 ---------
llllll
0.07 ----------
IIIII
0.06
0.0S ---------

0.04 --------

0.03

0.02
C~ C) C~ C~ C~ C~ C~
C~ Q O C~ C~ C~ C~

,m ¢,q ¢1"

Current (Percentage of normal current rating of series coil)

Fig. 15.3.3B Recloser time/current tripping characteristics (AEI Ltd.)

 The application of protection to rural distribution systems 259

of the solenoid imposed some limitation on the maximum breaking capacity,

particularly on units for the lower normal current ratings.
The timing mechanism incorporated an oil dashpot which provided a full
operating sequence of two high speed and two delayed trips, or alternatively up to
four high speed tripping operations. On the delayed trips it imparted an inverse
time characteristic modified to improve discrimination with time fuses fitted to the
source circuit breaker. This characteristic is indicated in Fig. 15.3.3B.
If a fault was cleared before completion of the full operating sequence, the
recloser remained closed and the mechanism reset. When, however, the fault
persisted for the duration of the full sequence, the recloser locked in the open
position after the final trip and then had to be reclosed manually after the fault
had been removed.
Following a manual reclosure, the recloser had only one delayed trip to lock
out in order to prevent tripping due to inrush currents and to prevent unnecessary
reclosing, if a fault still existed on the line. If no such tripping occurred however,
the mechanism reset within a short time to provide a full operating sequence.
This type of recloser was manufactured with ratings up to 200 A in single phase
units with mechanical interphase coupling. It was arranged to provide tripping and
reclosing on the faulted phase or phases only, but for locking open or manual
operation all three phases were actuated simultaneously.
The minimum operating current was approximately twice the normal current
rating and thus tripping on low values of fault current, such as may obtain on some
types of earth fault, was not possible. This type of recloser is therefore no longer
in production having been superseded by the current design which affords this
latter feature.
The design now available encloses three phase units in one tank and is arranged
for three-phase tripping and reclosure. The energy for charging the reclosing spring
is derived from a multistroke electromagnet, the coil of which is connected across
two of the incoming terminals through auxiliary switches coupled to the main
contacts which ensure that the coil is in circuit only for the time necessary to
charge the spring. The electrical arrangement is as shown in Fig. 15.3.3C. It should
be noted that the act of manual closing merely closes the coil auxiliary switches,
and closing of the main contacts is by the action of the electromagnet and closing
springs.
Under certain circumstances, such as when maintenance or repair work is being
carried out, it may be necessary for the current flow through a recloser to be in a
direction reverse to that normally employed. When operating under this condition,
a recloser with the closing arrangement described above will trip once to clear a
fault occurring on the side remote from the new source of supply, but cannot be
reclosed, either automatically or manually, since the reclosing coil is then isolated
from the supply by the main contacts. To avoid unnecessary interruptions due to
this cause, a 'lock-closed' feature has been introduced which is brought into action
by the operation of an external handle and which prevents the recloser from
tripping. Protection against faults on the recloser protected section is then provided
by another recloser or a circuit breaker nearer the source.
260 The application of protection to rural distribution systems

Incomingsupply terminals

"--'J ~ transformers
Current
'7 - , .

Earth
fault
relay Series overcurrent
trip coils
sAwUiic;liha~Y "[ /1
Spring 0 - Maincontacts
II
Charging E
solenoid
Outgoing supply terminals

Fig.15.3.3C Internal connection diagram of recloser with earth fault feature

(A Reyrolle & Co. Ltd.)

Overcurrent protection is provided by conventional series trip-coils with a

dashpot type mechanism to give an inverse time characteristic shown in Fig.
15.3.3F. For instantaneous tripping, the dashpot action is bypassed by the opening
of a valve in a cylinder, the control of this valve being by means of a cam which
thus provides a ready means of sequence control.
As a consequence of the fine mechanical clearances employed in the construc-
tion of the dashpot, some increase in operating and resetting times occurred in cold
weather due to the increased viscosity of the oil. Under repetitive fault conditions,
e.g. conductor clashing during winter storms, this had the effect of cumulative
progression to the lockout position due to incomplete resetting between consecutive
operations. One method of alleviating this situation was to use a special low
viscosity insulating oil which produced operating characteristics at 0°C approxi-
mating to those at 20°C with normal oil to BS 148.
This cold weather problem is now overcome by the incorporation in the design
of a modified timing and sequencing mechanism. In the modified tripping dashpot,
the clearances are controlled by a bimetallic strip and thus afford delayed over-
current tripping times which are comparatively independent of temperature. The
resetting mechanism is now noncumulative and resets to provide a full sequence of
operations in a constant time irrespective of the number of trip operations which
 The application of protection to rural distribution systems 261

have occurred. The standard reset time is 90s, but this time can be reduced to some
5-10s by drilling a small hole in the controlling dashpot.
Provision is made for an instantaneous earth-fault trip coil operated from current
transformers in turrets at the base of the terminal bushings. This feature is operative
on earth fault currents down to 20 amps.

Spur line

So urce
substation

=
Recloser

Spur line

Fig. 15.3.3D Recloser and fuse protected line

A range of trip coil ratings up to 200 A is available with this design and the full
breaking capacity of the unit is maintained with all ratings throughout the range.
Owing to the single tank construction, the unit may be mounted on a single line
pole.
Both types of recloser described above have nominal dead times of one second.

(b) Application: In order to protect as large a section of network as possible, a

recloser should be installed as near to the source of supply as limitations of normal
current and breaking capacity will permit, the actual location being determined
largely by these factors. Reclosers in series may be justified on extensive or
important networks and in some cases individual spurs may be protected by a
recloser.
The method of application of reclosers based on their use in conjunction with
h.v. fuses is indicated on Fig. 15.3.3D. The recloser and fuses are so co-ordinated
that transient faults are cleared by the high speed initial tripping of the recloser.
The fuse, or fuses, operate during the delayed tripping period to isolate persistent
faults and minimise the section of network without supplies. This requires a
recloser with an operating sequence of instantaneous followed by delayed tripping
operations, as shown in Fig. 15.3.3E. The fuse should remain intact during the
instantaneous trips and suffer no deterioration due to partial melting so that its
262 The application of protection to rural distribution systems

performance on subsequent faults may be in accordance with its published

characteristic. On the delayed tripping operations, however, the fuse should melt
and isolate the faulty section before lockout of the recloser occurs.
In order to determine the current range over which these requirements may be
achieved it is necessary to compare the total heat input to the fuse during the
relevant portion of the recloser operating sequence with the time/current character-
istic of the fuse. Over the portion of the sequence during which the fuse should
remain intact, however, some modification of the published fuse curves is required,
in order to take into account the effect of preloading and progressive deterioration
of the fuse element with consequent loss of characteristic after carrying repeated
transient fault currents. This is done by derating the melting time curve of the fuse
by the application of a factor of 0-75, a figure which may be used in the absence of
specific information on this point relating to the particular fuse employed. On the
other hand, some degree of cooling of the element takes place during the open
circuit interval of the recloser, and the correction time to be added for this effect
is given by the formula

t=---T

where t is the correction time to be added to the fuse melting time characteristic,
Im is the minimum melting current of fuse, T is the total open-circuit time for the
portion of reclosure sequence considered, and I/is the circuit fault current.
In practice the error introduced by neglecting the cooling effect is so small as
to be of little consequence.
Fault
inception
/
Closed ' ' ' ~ ~ -- , _

Open Locked

__.._j open

¢j

= =~ E=
;~ o o o Time ~ o o
• Z~ Z,Z

Fig. 15.3.3E Travel/time diagram for instantaneous and delayed tripping recloser
 The application of protection to rural distribution systems 263

Fuse t()tal clearance time

2 inst. + 2 delayed trips

2 inst. + 1 delayed
trip

i inst. trips

inst. trip

I- C()-ordinati(}n v \ \

(.'urrent

Fig. 15.3.3F
264 The application of protection to rural distribution systems

The modified fuse characteristics are then superimposed upon the cumulative
time/current curves of the recloser, as shown on Fig 15.3.3F. The upper limit for
correct co-ordination is given by the intersection of the derated fuse melting
curve and the last instantaneous tripping curve of the recloser (shown as A), whilst
the lower limit is obtained from the intersection of the fuse total clearing time
curve with the recloser final delayed tripping curve (shown as C).
With fault currents between A and B, co-ordination will be obtained only if the
fault is removed before the second instananeous trip, and currents in excess of B
will cause the fuse to blow before any operation of the recloser. Point A is there-
fore regarded as the upper limit since above this value the full advantage of a second
instantaneous trip is not obtained. At currents below the lower limit, locking open
of the recloser will occur before the clearing of the fuse, provided that the current
is in excess of the minimum value required to trip the recloser, normally twice its
rated current.
The range of co-ordination is obviously influenced by the slope of the fuse
characteristic curve, the steep slope of the fast blowing fuse restricting the range in
comparison with that of the slow blowing type. The latter type are therefore
employed with reclosers, and it has been found that the size of fuse is dictated by
co-ordination requirements rather than load currents, and fuse ratings somewhat
higher than those in use on systems without reclosers are generally employed.
The dual mode fuse referred to in Chapter 5 is designed to afford improved co-
ordination with reclosers by having fast blowing characteristics at low currents
and slow blowing characteristics at high currents.

(c) Development of alternative operation cycles: Operational experience of

reclosers and fuses in combination initially indicated that the number of supply
interruptions on the protected line were reduced by some 90% of their original
number, and that the operation of a fuse was a definite indication of a persistent
fault. After two or three years, however, a sharp increase in the number of fuse
operations on transient faults occurred inferring that fuse deterioration in service
had, in fact, taken place.
The effect of this deterioration can be offset, and the initial performance of a
recloser/fuse combination maintained, by the replacement of all fuse elements at
intervals of approximately two years. Alternatively, the fuse deterioration factor of
0-75 may be reduced, but this is attended by a reduction in the range of co-ordina-
tion and may not, in any case, be the complete solution. In view of the small
number of persistent faults occurring on recloser protected systems, however,
consideration was given to the elimination of fuses entirely. In addition to reducing
installation and operating costs and avoiding unnecessary interruptions due to
transient faults, this course permitted an improvement in the overall system per-
formance since delayed recloser tripping operations could be eliminated and dis-
crimination with back-up relays thereby facilitated. It was appreciated that all
persistent faults would result in the locking open of the recloser causing a greater
number of consumers to be affected by interruption of supply, but it was felt that
against the reduction in interruptions on transient faults this could be tolerated.
 The application of protection to rural distribution systems 265

Reclosers having all instantaneous tripping operations, as indicated in Fig. 13.3.4J

were therefore introduced.
Reclosers with delayed trips have been used without associated fuses, but they
impose unnecessary strain on the system due to the long clearance times for per-
sistent faults. In addition, the problem of discrimination with back-up relay pro-
tection still remains.
Operating experience of reclosers on systems from which fuses have been
eliminated indicates that the incidence of supply interruptions is reduced to a
consistently low level. The main criterion to be employed when comparing this
method of application with that employing fuses, however, is that of the product
of the number of consumers and time for which supplies are interrupted, taken
over a period of time rather than for an isolated incident.

(d) Fault indicators: On systems protected by reclosers with the all instantaneous
tripping sequence, damage due to persistent faults is greatly restricted due to the
very short duration for which the fault current is allowed to persist. At times this
renders fault location difficult, and adequate line sectionalising facilities are
necessary to assist the location of faults. Where fuse mounts already exist, these
may be converted to sectionalising points by the fitting of solid links in place of the
fuses.
The installation of fault indicators at strategic positions on an overhead-line
circuit aids the location of faults by indicating the passage of fault current. One
such device provides a fixed phase fault setting of 100, 200 or 400 A and a
minimum earth fault setting of 5% of the phase fault setting. Indication is displayed
whilst the setting current is exceeded or the line is dead after fault clearance, reset-
ting taking place automatically on restoration of supply.
The device is activated by three detector coils located inside special insulators
mounted on the line crossarm. The magnetic flux surrounding each line conductor
supported by the insulators induces a voltage proportional to the line current in
the associated coil and these voltages are applied via flexible leads to solid-state
phase and earth-fault measuring circuits contained in a weatherproof housing
mounted on the pole below the crossarm.
At the set current value, a signal is applied to a trigger circuit which discharges
a capacitor into the indicator operating coil. Whilst the line is energised the device
receives a pulse every few seconds from a second capacitor to retain it in the reset
state but the operating circuit is arranged to override the resetting circuit.
The device derives the energy necessary for operation of the measuring and
indicating functions from the line itself by means of capacitors formed b y con-
ducting glazes on the upper shed and internal coil housing of the insulator. Since
there is no stored energy available prior to energising the line, the device takes
approximately two seconds to operate when closing a line on to a fault and this
time may be increased if the source impedance is high and the voltage reduced. If
the line protection clears the fault within that time, no indication is made.
For close-up faults when the line voltage may fall to zero, sufficient energy is
stored to allow normal operation for up to 4 s after the loss of voltage.
266 The application o f protection to rural distribution systems

Fault
inception

Closed ,

Open Locked
open

¢¢ ¢¢ ¢¢
•, - 112

oq oq
Z- z,-

Time .--------e.

Fig. 15.3.3G Travel/timediagram for all-instantaneous recloser

Trip rod

Successive u u
positions of ~ ...... d Series coil
moving piston
. . . . . 'd
Reset position
of moving piston ~1~
Moving piston stop
Non-return Captive piston
valves

Iron core ~ Oil intake

Fig. 15.3.3H Linesectionaliser timing device

 The application of protection to rural distribution systems 267

The indicator housing and the detector coils may be installed or removed by the
use of live line techniques.
Another type of line fault indicator may be used as a portable device for attach-
ment to poles at strategic points on a line during fault finding operations and
subsequent removal for use on later faults on other circuits. Alternatively, a version
suitable for permanent installation is available.
The device is fixed to the pole near to the base and is responsive to changes in
the magnetic field produced at this level by currents flowing in the line conductors,
this field being zero under normal balanced load conditions since the vector sum of
the phase currents is zero. Under the unbalanced conditions obtained during faults
however, a resultant field is produced and an earth fault current of 20 A is
sufficient to cause operation of the indicator. The phase to phase fault current
required to operate the device is dependent on the conductor configuration, the
minimum value for lines of horizontal construction to BS 1320 being 900 A and
somewhat less for lines of triangular tormation, although operation may not occur
in the case of faults which cause currents to flow in the outer conductors only.
The operating sequence of the instrument is initiated when a system disturbance
causes the rate of change of the magnetic field to be sufficient to induce a voltage
of the set value in the detector coil. Irrespective of its state, the device immediately
resets to the start of its sequence and for 50 ms operation is inhibited. This delay
is introduced to avoid unwanted operation on magnetising inrush transients but if
at the end of this period the field is still adequate to produce the required voltage
in the detector coil, the device will indicate after a further 20 ms. Following this,
further operation, i.e. resetting, is prevented for one second so that the device is
unaffected by the change in field caused by the clearance of the fault by a circuit
breaker. The detector circuit is then reinstated and any further system disturbance
will reset the device and restart the operating sequence. If, however, no such signal
is received, the device resets automatically after 8 h. The times quoted above are
preset during manufacture but may be varied to suit particular requirements.
Operating power is derived frown internal batteries, a ten year life being claimed
for those used in the permanently installed version.

(e) Sectionalisers." A method of isolating faulty sections of a recloser protected

network without employing fuses is the use of automatic line sectionalisers. These
devices are pole mounted oil disconnectors which are arranged to open automatical-
ly during a predetermined 'dead time' of the associated recloser, thus isolating the
fault before lockout of the recloser occurs. This characteristic ensures co-ordination
with the recloser at all values of fault current up to the full rating of the device.
The sectionaliser functions by counting the passage of pulses of fault current let
through by the recloser, these pulses flowing through the operating solenoid causing
a captive piston in an oil filled cylinder to be pulled down against a spring. A non-
return valve in the crown of the captive piston allows oil from the underside to pass
through to, and be trapped in, the space above the piston, forming an oil column
268 The application of protection to rural distribution systems

which is forced upward by the spring when the recloser operates to cut off the fault
current in the solenoid. The oil column in turn lifts a moving piston. After a preset
number of pulses, the off column formed is sufficient to lift the moving piston
against the trip bar of the sectionaliser, the contacts then being opened by the
action of a spring which has been charged during manual closing of the device. The
basic arrangement of the sectionaliser is shown in Fig. 15.3.3H.

(f) Grading of reclosers with other forms of protection: In a typical recloser

protected system, as depicted in Fig. 15.3.3D, the operating characteristic of the
recloser must be between those of the fuse and the protective relay on the source
substation circuit breaker. Any maldiscrimination, in addition to causing unnecessary
loss of supplies, may also delay the restoration of these supplies by presenting a
false indication of the location of a fault to the operating staff.
A method for determining the co-ordination between recloser and fuse has been
set out in Section 15.3.3(b), and it is essential to ensure that the size of fuse
installed permits a range of co-ordination adequate to cover fault currents likely to
be encountered on the protected spur.
The protective relay associated with the substation circuit breaker will usually
be of inverse definite minimum time type, and when a persistent fault occurs
beyond the recloser this relay may be energised for two very brief periods followed
by two somewhat longer time intervals.
Since little resetting of the relay disc can take place during the very brief dead
times, cumulative forward movement of the relay will occur. In order that
discrimination between the circuit breaker and the recloser is obtained, it is
therefore necessary to take this into account when choosing the time setting of the
relay. An operating time of approximately twice that of the time delay tripping
characteristic of the recloser, as in Fig. 15.3.4L, is usually satisfactory.
Compliance with this requirement may at times be difficult against the back-
ground of the overall scheme of protection further back towards the source of

Source
substation
! Recloser 1 Recloser 2
12] 121 '
I Locks open Locks open
after 4 operations after 3 operations

N.B. I f o n l y one recloser is fitted with an earth fault relay this should
be employed as recloser No. 2.

Fig. 15.3.3J Grading of all-instantaneous tripping reclosers

 The application of protection to rural distribution systems 269

supply. Where, however, full discrimination cannot be obtained, the installation

of reclosers will still show a considerable improvement in the quality of supply
since some 90% of the faults occurring will be cleared in the nonpersistent stage by
the instantaneous trips of the recloser which permit discrimination. The disad-
vantage of this arrangement lies in the fact that when maldiscrimination occurs, the
operating staff may be presented with a misleading indication of the location of the
fault and valuable time lost in restoring supplies.
With all-instantaneous tripping reclosers the difficulty of obtaining
discrimination between the recloser and circuit breaker is greatly reduced. When
reclosers of this type are operated in series, however, the differences in the
operating times of units of various ratings are so small as to render time grading
unreliable. Recourse must therefore be made to grading by the number of
operations to lockout, as in Fig. 15.3.3J. The recloser most remote from the supply
source is set to lock open after two or three unsuccessful reclose operations, whilst
the one nearer the source is arranged to complete a full sequence of four trips
before locking open. Thus, if the current to a fault beyond the more remote
recloser is of such value as to cause simultaneous operation of both reclosers, the
more remote recloser will lock open to isolate the fault whilst supply to the healthy
section is maintained through the recloser nearer the source.
When a recloser with an earth-fault tripping feature is employed in series with a
recloser not so fitted, as may occur due to piecemeal growth of a network, it must
be borne in mind that all earth-fauh tripping operations are instantaneous. The
recloser with the earth-fault feature must therefore be employed as the more
remote unit of the pair; otherwise false indication of fault position may be given,
since the magnitude of the fault current may be high enough to operate the earth-
fault feature but insufficient to cause overcurrent tripping of the recloser not so
fitted.

15.3.4 Substation circuit breakers

The location on a rural system of the pole-mounted reclosing devices previously

mentioned may be governed by considerations of normal current rating and
breaking capacity, and these may demand that the recloser be installed some
distance from the primary substation. In such cases, the section of line between the
substation and the device is without the benefit of high speed automatic reclosing
and faults on that section, whether persistent or nonpersistent, will affect the
continuity of supply to the remainder of the line. The advantage of providing
automatic reclosing of the substation circuit breaker to minimise interruptions of
supply on this section will therefore be apparent.
Control relays for circuit breaker reclosing schemes have been available tbr a
nulnber of years. The simplest of these provides a single closing impulse after a
predetermined time delay, whilst another torm permits a preselected number of
reclosures with selected time intervals between them. This latter relay, known as
a periodic reclosing relay, consists of a motor driven disc around the periphery of
 270 The application of protection to rural distribution systems

which are spaced a number of tapped holes. Into selected holes are screwed pins
which project through the disc, and as this is rotated the projections in turn actuate
a contact to energise the closing circuit, the time between reclosures (the dead
time) being determined by the spacing of the pins. If the circuit breaker remains
closed after a reclosure, the relay is reset by the energising of the motor reverse
field through an auxiliary switch on the circuit breaker. The final pin has a some-
what longer projection than the others and if the reclosure initiated by it is
unsuccessful, continued rotation of the disc causes it to operate a lockout contact
which disconnects the supply to the foward field of the driving motor, preventing
further reclosure. The electrical circuits of this type of relay are indicated in Fig.
15.3.4 A.
The time interval between tripping and reclosure generally employed with the
above relays is comparatively long and may vary from 10 to 120 s. The number of
reclosures permitted is dependent on the nature of the system protected
and may be two or three where it consists almost entirely of overhead lines, and
somewhat less where the proportion of underground cable is appreciable. Operating
experience indicates that with the use of these relays, in conjunction with inverse
time protection, successful reclosure occurs in approximately three out of four
faults, the number depending on the speed of operation of the protection.
The improved quality of rural supplies obtained by the use of the pole-mounted
automatic circuit recloser indicated the advantages of high speed fault clearance and
the desirability of providing a similar operating sequence on substation circuit
breakers. This course removes the normal current and breaking capacity limitations
of the recloser and extends the reclosing facilities to the whole of the protected
line.

D i v e r t er R e s e t t ing
Reverse limit Auxiliary
resistor field switch switches

d ---- ,-,~a

Auxiliary Lockout !
[ ~ relay contact
, ,

I I Closing I
i '
I I

Changeover ---O.J.o----
contact operated
by peg in disc . Auxiliary supply . . . . . .

a - m a d e w h e n o.c.b, closed
b - made when o.c.b, open

Fig. 15.3.4A Connections for periodic reclosing relay

 The application of protection to rural distribution systems 271

I.I).M.I. Auxiliary
rel:.l.v,, l r i p ccJil s~,~ it oh,as
/ " ~ ,
0 C- " ~-30----
Inst.
rela ~s
T'
1 Spring I
release
~.,~il
I
c~rl/~J ft
0 C 0 O
I i --r-,
,\ u ,~iIiar v
iI ch
,',;,,,,
~v'i s p r i n g l"iuling
ch)sing rel:l.~
nllc ch;,lllJXnl

a i n a d e ,,,,hen ~ . c . b . c h , s e d

b made ~ e n ~ . c . h , ~,pen
',.'h

Fig. 1 5 . 3 . 4 B Basic reclosing scheme for substation circuit breaker

To provide this feature, the circuit breaker is fitted with a combination of high
speed and inverse-time protection, each having approximately the same range of set-
tings. On the occurrence of a fault, the high speed protection operates to trip the
circuit breaker, which is reclosed by an auxilliary timing relay after a brief time
interval. The tripping circuits of the high speed relays are then held open, either by
an auxiliary switch on the discharged spring closing mecl+anism or by an auxiliary
relay, so that if the fault persists after reclosure, Further tripping of the circuit
breaker is accomplished by i.d.m.t, relays. The control scheme is arranged to lock
open the breaker on the operation of the inverse tinm relays, but if the reclosure
is successful, the relays reset to provide a full operating cycle on subsequent faults.
The basic circuit arrangement of a simple spring closing scheme is shown in Fig.
15.3.4B and the operating sequence diagram in Fig. 15.3.4C.
i"ault High speed Retl,,,,e ,II)MI i rip
incepti()n trip

Closed .

l.c,~, kcd
Open
l)ead t i m e

I'inle I~a.,e

Fig. 1 5 . 3 . 4 C Travel/time diagram for reclosing circuit breaker with high speed and i . d . m . t .
p r e tec t i e n
272 The application of protection to rural distribution systems

This sequence is known as repetitive single-shot reclosing and was adopted to

limit the number of breaking operations at the higher source substation fault levels
to a figure within the capability of standard circuit breakers. Subsequent investiga-
tions and service experience have shown that breakers currently available in this
country are capable of carrying out many more breaking operations before
maintenance becomes necessary than those required by certification tests, particu-
larly at the lower duties.
In the early 1960s, a number of Area Boards introduced individual auto
reclosing schemes for use with ground mounted circuit breakers but in 1966 a
report recommending standard reclosing schemes was issued by the Electricity
Council. The schemes, all incorporating the single shot sequence with optional short
or long dead times were for:

(i) Single-shot reclosing for circuit breakers having hand charged spring mechan-
isms.
(ii) Repetitive single shot reclosing for breakers with motor wound spring closing
mechanisms. This scheme utilises an auxiliary switch on the spring charging
mechanismto use the rewinding time as the reclaim time of the sequence.
(iii) Repetitive single-shot reclosing for circuit breakers with solenoid closing
mechanisms. In this case the dead and reclaim times are determined by a
timing relay.
(iv) Repetitive single shot reclosing for breakers with solenoid or motor wound
spring mechanisms and incorporating precumulative lockout and cumulative
lockout alarm features.

In all these schemes reclosing is initiated only by the operation of the high speed
protection thus ensuring that it takes place only after a fault and not following
manual tripping of the breaker. The repetitive schemes incorporate cumulative
operation counting relays which limit the number of tripping operations on faults
to an extent determined by the maintenance requirements of the breaker. When the
preset number of operations has been achieved, the inverse time protection only
is effective and further reclosing is inhibited.
The schematic diagram for scheme (iv) is shown in Fig. 15.3.4D.
A subsequent report concerning protection against lightning discussed the use
of automatic circuit reclosing in this context and recommended the adoption of
long dead times and short reclaim times, the figures quoted being 10-15s and 5s,
respectively. These times enable the system to recover fully from a fault and
minimise the possibility of unnecessary lockouts on repetitive faults. This report
also suggested that, with improved knowledge of circuit breaker capabilities, multi-
shot reclosing schemes might be employed with advantage. Modifications of the
original standard schemes have therefore been made to permit the optional use of
an additional counting relay to extend the operating sequence up to a maximum of
three reclosures, (i.e. four tripping operations) on any one fault.
Comprehensive relay equipment providing both protection and reelosing
 The application of protection to rural distribution systems 273

Control switch
tnc
Cumulative operation
counting relay
[
-.-o I
I [
o . . . . .
Cot. bkr.
~

- o:.t.',
= , ; ~; _ 1_ . . R ecl,,~i.g
. . . . . ,, a~x.
. .~w. ......

] O/C & E/F ~ II ] isolati . . . . . lay ~ Trip coil

d.c. bus
wire wire
I ~ I ~ -lr~ '

l
o',j
~ Sensitive ,,I
/~--- -~....o--- -k~_
~ . .e /.r ., e., a .y . . I, ........ --- .....

I Bus zone Bus zone I

I relay . . . . . . lay_L'~ I
I~ - O o . . . . . . . . . . . . -.,';r ~ o / c, ~, r:/ l.
I 1 ~ _LI I instant, relay / relay
~ ~ : oyio-----~r-o, .._
Lockout II . . . . J Lockout
relay a na
autl)/non-auto
switch J I Aux. SW. on oj
, o
I closing-spring m cot. bkr.
l mL,chanism ~ .....

I II e
0 O-

! ~ L..ol . . . .
I ~,~ Pre-lockout { Reclosing
I [ I alarm relay i,~IP isol. relay
I
i I oo.--~ Solenoid/spring
release coil
contactor relay
Reclosing I 0~0.-]
relay T ' ] '
internal l I I Solenoid/s~rinl~.reiease
bridging ~ | Pre-lock()ut I , co~tactor
contact "---0 1 C
z I " I ' • -- '
I I I I a arm relay I Cumulative .per. Pre-lockout
Control switch ~ rely a,I. . . . . lay / ~ _

Cumulative <)per.
c~)unling relay
reset coil
-0 C

Reset push button

To reclosing circuit T(~ oct. hkr,

r_L~-'~pre l , )ck o u,. ~ ' ~ lock(,ut alarm

I'ru-lockout alarm relay Lock~ut relay

Auxiliary switch (~n
closing.spring
~ d mechanism @ ~ ~

MxJtor supply Motor supply

(may he d.c.) (may be d.c.)

Mechanical counter to record total number of opening operations.

Auxiliary switch d on closing spring mechani.~m is ()pen when spring is charged. Auxiliary switch e on closing spring mechanism is
closed when spring is charged.
"riming relay to have a normally open time delay contact A closing after a preset time delay of 0.5 - 1.0s or 5 -20s l',)r a period . f
not less than 0.6s and not greater than 2s and a normally open time delay c~mtact B closing after a preset time d~:lay of 30-35s
or 3.~-60s. The remaining contacts c)perate instantaneously.
Auxiliary relay must operate in not more than 8ms. The normally (~pen c(mtacts mu.~t have a lOOms delay on drop c~ff when the
relay is shorted.
Normally <)pen contact on the cumulative operati()n-c~unting relay must not close in less than I Sins.
Reclosing relay internal-bridging contact is chtsed ~hen the relay assembly is withdrawn from) the case.

Fig. 15.3.4D C i r c u i t d i a g r a m o f r e p e t i t i v e s i n g l e . s h o t reclosing s c h e m e f o r c i r c u i t b r e a k e r

with motor-wound spring o r s o l e n o i d - c l o s i n g m e c h a n i s m a n d h a v i n g p r e l o c k o u t
a n d l o c k o u t a l a r m facilities
 274 The application of protection to rural distribution systems

facilities is now available in solid-state form. These relays afford high speed pro-
tection with characteristics shaped to prevent operation by transformer magnetising
inrush currents together with a choice of various inverse time overcurrent and earth
fault characteristics by means of interchangeable plug-in modules. Sensitive earth
fault options are also available.
Flexible control of the auto reclosing sequence and protection operation
selection is provided by changeover links mounted on the face of the relay. Up to
3 or 4 reclosures per fault incident may be obtained and the instantaneous
protection can be inhibited for any trip operation of a multiple-reclose sequence.
Both dead and reclaim times are adjustable but whilst the former may differ for
each reclosure, the preset reclaim time is applicable following all reclosures. An
operations counter and circuit breaker inspection and lockout alarms are provided.
In addition to greater flexibility in application, the solid-state protection and
reclosing control relays have the merit of occupying little control panel space com-
pared to their electromagnetic counterparts.
The suitability of a circuit breaker for autoreclosing duties depends upon its
total break time, the mechanical stability of its operating mechanism and, in the
case ofoil circuit breakers, the effect of any residual gas pressure in the tank when
successive breaking operations occur within a very short time. In addition, taking an

!i r I i i i
1
I!,0/ , v S O
25 .amp lain.so

|.0 -
~ll
MI
M
nap
•
i
[

IX
!~,
7

1i i l i
\ !
I II Envelope of O C B --
\k I
Envelope of
O C B opening \
\ \
\
opening time plus - -
inst. relay time for __
\
time plus inst. \ 3 0 0 A setting
relay time for L I /I a !

60 A setting

0.I - fJ'lgJ lJ
1 ! [~,fJ//~4,,'t~ I ..4P'I f J J
]~lli I \ 1 I_ .J'l"fTJ
! P¢~! ! \1 A'/JJl
\1 \l. illl
~Xl Xlill
Envelope of vacuum k Envelope of vacuum
-CB opening time CB opening time
plus inst. relay time plus inst. relay time

0.01
for 60 A setting
_Illl~ I l
! for 3 0 0 A setting
I I I I I f

100 1,000 10,000

Current - Amps

Fig. 15.3.4E Limits of co-ordination between circuit breakers and fuses

 The application of protection to rural distribution systems 275

overall view, the breaker's ability to perform a number of operations before re-
quiring maintenance must be taken into account.
The vacuum interrupter is ideally suited to this duty. This device possesses high
speed of fault clearance, the current being interrupted at the first current zero after
trip initiation unless opening of the contacts takes place within 3 ms of that point
when clearance is effected at the next zero point. Such a characteristic permits a
high degree of co-ordination between the vacuum breaker and any associated fuses.
This is illustrated in Fig. 15.3.4E which shows the co-ordination between both oil
and vacuum circuit breakers and fuses. The fuse characteristics have been plotted at
75% of the published data to allow for deterioration of the fuse link and the inter-
section of the fuse and circuit breaker characteristics determines the limiting
current beyond which co-ordination will not be obtained.
It is not possible to employ sectionalisers in series with substation circuit
breakers having repetitive single shot reclosing facilities. Since the sectionaliser has
no breaking capacity and must open during a dead time interval in the operating
sequence, it would isolate the line section before the single reclosure irrespective of
whether the fault was of a transient or permanent nature. Reclosers having all
high-speed tripping sequences may be used in the role of sectionalisers in con-
junction with substation breakers, however. Whilst the first trip of both breaker and
recloser may occur simultaneously, the subsequent high-speed trips of the recloser
will ensure discrimination with the inverse time protection of the circuit breaker
and final clearance by the recloser.
The application of multishot reclosing schemes to substation circuit breakers
permits the use of associated sectionalisers at a lower cost than that of reclosers.
Interruptions of supply on lines protected by this form of reclosing are reduced
to a level similar to that obtained by the use of pole-mounted reclosers and circuit
breaker maintenance requirements have been found to be light.
The provision of any form of automatic reclosing on a circuit breaker requires
some form of power closing mechanism. If full advantage is to be derived from this
feature, this mechanism must be of the solenoid, motor wound spring or other self-
charging type.

15.4 Sensitive earth-fault protection

It has been mentioned earlier that earth-fault currents on rural networks may be of
very low magnitude owing to long line lengths, the use of neutral earthing resistors,
and difficult earthing conditions. Such circumstances arise when an overhead
conductor breaks and falls on ground of high resistivity or across a hedge or
haystack.
The 1947 edition of the Overhead Line Regulations required that such a
conductor should be rendered dead but compliance with this requirement was in
some cases not possible since the fault current under such a condition might be less
than the minimum setting of the earth fault relays then available. This particular
276 The application of protection to rural distribution systems

requirement is not included in the current (1970) edition of the Regulations but in
view of the implications of a broken conductor remaining alive, some means of
automatic disconnection is clearly desirable.
To avoid a dangerous condition arising on broken conductor faults, a sensitive
form of earth-fault protection, responsive to primary earth-fault currents of the
order of 5-10 A, is required. Such values represent the minimum setting which can
be satisfactorily employed, since normal system unbalance currents, for example
due to the unequal line to earth capacitance currents of varying lengths of single
phase spur lines, may approach the lower value. Even so this sensitivity may be
insufficient to detect broken conductor faults in which the conductor on the side
of the break remote from the source falls to the ground. The conductor in this
instance remains energised via the high-voltage windings of any transformers con-
nected to the line beyond the point of fault and the fault current may be limited to
a very low value by the impedance of these windings. Its magnitude will be
determined by the number of transformers connected and the load connected to
them, but in the extreme case of transformers on open circuit will consist of the
magnitising currents only.
Whilst possessing a low fault setting, the components must be sufficiently robust
to withstand the effects of much heavier fault currents which may occur from time
to time.
With a relay which operates at such low primary currents, the effect of the
magnetising current drawn by current transformers in idle shunt has an appreciable
bearing on the value of primary current required to cause operation of the relay.
For this reason, therefore, early installations utilised a moving-coil relay element
energised by a core balance current transformer mounted on the cable termination.
The ratio of such a current transformer may be independent of the circuit rating
and fault level and can be arranged to provide the optimum setting, generally of the
order of 5 A.
When applying such a scheme to existing switchgear installations, however,
considerable expense and inconvenience may be involved as it may be necessary
to break down the cable end box in order to mount the core balance transformer
and to fit the insulated glands essential for such a scheme, if the current transformer
is mounted externally.
The development of relays having extremely low burdens, of the order of
0431 VA or less, has enabled sensitive earth-fault protection to be provided by
connection of the relay in the residual circuit of the current transformers employed
for the conventional earth-fault protection. Such relays of the polarised element
type enable settings as low as 2% or 3% of the primary current rating to be obtained
whilst the corresponding static version provides a setting range of 1% to 16% of
the rated current.
High-speed clearance of the low fault currents in question is unnecessary and
indeed may be undesirable, since the high sensitivity of the protection might then
give rise to operation on momentary system abnormalities. It is usual, therefore,
to apply a time-lag to the tripping operation and values of approximately 10
seconds are in common use. A typical co-ordination of sensitive earth-fault pro-
 The application of protection to rural distribution systems 277

tection and reclosers having delayed tripping and earth-fault features is shown in
Fig. 15.4A.
The operation of sensitive earth-fault protection applied to an overhead line is
frequently an indication of the existence of a dangerous condition on the circuit,
e.g. a broken conductor lying on the ground. For this reason, some earlier auto-
reclosing schemes are arranged to inhibit the reclosing feature after tripping
initiated by sensitive earth-fault protection. Experience has shown however that the
condition to which this form of protection is designed to respond arises frequently
during normal operation of an overhead system and some undertakings now employ
automatic reclosing in such circumstances. A reasonable degree of successful re-

100

Fault currents cleared by

Range of earth fault
recloser with earth fault
currents cleared by
tripping feat ure
sensitive E/I" relays
,., 1__~ Fault currentsL
v ! cleared by
I recloser with
10 ips

Delayed f .
overcurrent trip

~ _ _ maximum .
0.1 ////////////.i/7"

Earth fault trip ~ minimum

.01
/ 10 I00 1000
Pick up current Current - Amps
of sensitive
earth fault relay

Fig. 15.4A
 278 The application of protection to rural distribution systems

closing has been achieved and the more recent schemes include an option for
selection of autoreclosing following the operation of sensitive earth-fault
protection.
Sectionalising devices on overhead lines protected by sensitive earth-fault
protection should be of the type which breaks all three phases simultaneously,
otherwise the unbalance caused by the sequential opening of individual phases may
exceed the set time delay and cause the relay to operate. In some cases, switches
are provided in the tripping circuit to render the protection inoperative whilst
carrying out switching operations.

15.5 Are-suppression coils

Prior to the introduction of the unearthed crossarm form of overhead line con-
struction, faults on rural systems were predominantly to earth. To minimise
interruptions due to these earth faults some systems were designed to operate
with the neutral point unearthed, but it was found that under fault conditions
the capacitance currents between the healthy phases and earth were sufficient
to permit arcing at the fault position. The capacitance currents were invariably
of low magnitude and consequently were difficult to detect by the normal
protection. Furthermore the arcing could be prolonged and dangerous overvoltages
might occur. Such systems are not now permitted by statute.
To overcome this, the arc-suppression or Petersen coil was introduced. As
described in Chapter 1, this takes the form of a reactor connected between the
system neutral point and earth, the coil impedance being such that under earth
fault conditions the capacitance current to earth of the sound phases is neutralised
by the inductive current of the coil. The system line-to-earth capacitance will, of
course, vary with the amount of line in service at any one time, and tappings must
therefore be provided on the coil to cater for different operating conditions. With
a limited number of tappings, however, it is virtually impossible to neutralise
exactly the capacitance current for every system combination, and in practice the
current at the fault is the difference between the system capacitance and coil
currents.
The difference current is normally very small and can be tolerated by the
system. If the coil is continuously rated, an earth fault may be allowed to persist
for a period sufficient to enable switching operations to be carried out to isolate
the faulty section with the minimum loss of supplies. This facility is invaluable
where no standby supplies exist, but suffers from some disadvantages. Under earth-
fault conditions the voltages to earth of the sound phases of the whole system are
raised to the full line-to-line voltage and this increase may cause insulation
breakdown on these phases at points on other lines of the system, as indicated in
Fig. 15.5.A. The resultant condition is termed a 'cross-country' fault and
constitutes a phase-to-phase fault against which the arc-suppression coil is inef-
fective. For this reason, the transfer of existing networks to arc-suppression coil
working is generally done in small sections, as the number of breakdowns of
 The application of protection to rural distribution systems 279

R Y B
Secondary
winding of
supply transformer - Feeder 1

_~ Subsequent
flashover

I1 -:-
%
on phase B

',,, Cross country fault

Continuously
ll
lil
~ Original fault
on phase R

rated arc 1 !,-, l.'eeder 2

suppression
coil

Fig. 15.5A Cross~ountry faults on system protected by an arc-suppression coil

Automatic
switch l L
Earthing Short time rated
resistor arc suppression coil

Fig. 15.5B Automatic short~:ircuiting scheme for short time rated arc-suppression coil

insulation weak spots occurring during a fault is thereby reduced and their location
facilitated.
Where alternative supplies are available or the persistence of the voltage increase
cannot be tolerated, a short time rated coil fitted with automatic short-circuiting
equipment, as in Fig. 15.5.B, may be employed. With this arrangement, if the fault
condition persists for more than a few seconds, the system neutral point is
automatically connected to earth, either direct or through a resistor in parallel with
the coil, thus permitting the passage of sufficient earth-fault current to be detected
and the line isolated by the normal feeder protection. In some installations, the
short circuit applied to the coil is automatically removed after a time-lag sufficient
to permit the isolation of the faulty line. This ensures that the system is without
the benefit of the arc-suppression coil for the minimum of time but may lead to the
system being overcompensated if the line to earth capacitance current of the
280 The application of protection to rural distribution systems

isolated faulty feeder is large. To overcome this, coils with automatic tap changing
may be employed, but this naturally adds to the cost of the coil.
As an arc-suppression coil is effective for a fault anywhere on the complete
system, it is desirable to provide indication of the faulty feeder when the coil is
brought into operation. With automatic short circuiting, the tripping of the feeder
circuit breaker is a clear indication. Where a continuously rated coil is used,
indication of a faulty feeder may be obtained by the use of a wattmetric type relay
sufficiently sensitive to detect the small difference currents referred to earlier in
this section. To energise such relays, however, it is necessary to provide a voltage
transformer on each feeder circuit breaker and the arrangement is therefore costly.
The adjustment or tuning of a coil in relation to the protected system is carried
out by applying an earth fault to the system and varying the coil tapping to give
maximum voltage across the coil with the normal amount of system connected. The
reactance is then reduced slightly to ensure that a reasonable balance will exist
when a section of the network is removed for any reason. Some degree of mistuning
is permissible without seriously affecting system operation, and for 11 kV systems
this tolerance is sufficient to allow sections of networks to be transferred from
other substations when necessary to maintain supplies. For systems where frequent
variations in length of line connected occur, automatic on-load tuning equipment is
available but is normally not considered economic.
When viewed in the light of the amount of line which it may protect, an arc-
suppression coil is a comparatively cheap item of equipment. This advantage may
be somewhat offset by the cost of strengthening insulation weak spots and the
possible transfer of single phase spur lines or line transposition in order to achieve
a phase balance of capacitance currents. Its operation is, however, independent of
fault current levels and the co-ordination of existing protection is not affected by
its use. Statistics indicate that a reduction of some 40% in the number of supply
interruptions compared with fuse protected systems is achieved on systems
protected by arc-suppression coils, although a greater reduction might be expected
on older lines with a greater number of points having earthed metalwork.

15.6 Performance/cost comparison of protective equipment for rural systems

It will be appreciated from the preceding sections of the chapter that in the
application of protection to rural systems, the primary function of the clearance of
faults is allied to the rapid restoration of supplies. In comparing the performance of
various forms of protection available, therefore, some cognizance must be taken of
this factor. Unfortunately, in such an assessment, an infallible criterion is difficult
to find and the influence of particular local conditions must be weighed.
The quality of rural supplies may be judged in terms of

(a) Interruptions per 100 km of line per annum.

(b) The product of the number of interruptions and the km of line affected per
100 km per annum.
 The application of protection to rural distribution systems 281

(c) Total consumer hours lost per annum.

(d) Consumer hours lost per consumer per annum.

None of these takes into account all the factors which cause irritation or incon-
venience to the consumer, such as time of day or a spate of interruptions of supply
to a particular consumer, but that of interruption per 100 km per annum is usually
most readily available and is therefore most commonly used.
The cost of each form of protection must enter into any comparison which is
made between them, and since the amount of network protected by each form
varies, the cost per kilometre of line protected may serve as a useful basis. This cost
should take into account operating costs such as maintenance charges and cost of
fuse replacement, in addition to capital cost, and may well be expressed in terms
of total annual charges.
Operating experience with the types of protection covered by this Chapter
indicate that the performance figures shown in Table 15.6A may be expected.

Table 15.6A Performance figures for various types of protection

Interruptions per
Type of protection 100 km per annum

H.V. fuses 31
Arc-suppression coils 16
Pole-mounted recloser 2
Substation reclosing circuit Periodic reclosing 7
breaker Single-shot high-speed reclosing 3

The significance of these figures should be viewed in the light of the following
observations.

(a) Fuses: Individually these call for only a small capital outlay, but for adequate
protection of a network a large number may be required. The incidence of a fault
requires the replacement of the fuse element and this may involve high labour and
transport costs. Deterioration which occurs in service may cause unnecessary supply
failures.

(b) Arc-suppression coils: A single coil will protect the whole of the system to
which it is connected and it is therefore low in terms of capital cost per mile of line.
In addition, very little maintenance is required. It is, however, effective only against
single phase-to-earth faults, which amount to some 50% of the total number of
faults on overhead lines of unearthed construction, and, in addition, requires that
the line insulation be in sound condition in order to withstand transient over-
voltages as well as the rise in line-to-earth voltage during fault operation. Some form
282 The application of protection to rural distribution systems

of remote alarm system is desirable when the coil is installed at an unattended

substation.

(c) Pole-mounted reclosers: Each main line of the network requires a separate
recloser for protection and, as the capital cost of each installation is some 40% of
the cost of an arc-suppression coil, the total cost may be higher than that of a coil
protection system. The high degree of protection is provided only on that portion
of the line on the side of the recloser remote from the source of supply, and rated
current and breaking capacity limitations may restrict the extent of this portion.
Nationally, the average length of line protected per recloser installed is of the order
of 12 kilometres. The cost of periodical maintenance must be taken into account
but, due to the reduced number of supply interruptions with the use of reclosers,
system operating costs are greatly reduced.

(d) Substation reclosing circuit breakers: The cost of providing automatic

reclosing features on substation circuit breakers is that of additional relay equip-
ment only and is generally but a small percentage of the cost of the circuit breaker
itself. There is merit, too, in the fact that such features may also be readily applied
to existing switchgear. High-speed tripping may, with considerable advantage, also
be provided at low cost.
The use of reclosing circuit breakers affords the benefits of automatic
restoration to the whole of the associated feeder in contrast to restrictions imposed
by the limited breaking capacity of the pole-mounted recloser. The average length of
line protected by substation reclosing circuit breakers is nationally some 20 km per
circuit breaker.
Rapid fault clearance combined with automatic reclosing provides the lowest
incidence of supply interruptions and, because of its performance, low cost and
other advantages, the provision of this feature on substation circuit breakers has
considerable advantage.
Multiple reclosure sequences offer some improvement in system performance
compared with single-shot schemes and also permit the use of sectionalisers to limit
the extent of supplies interrupted. The cost of the sectionaliser at present available
is however some two-thirds of that of a recloser and largely on this account they
have not been widely used in this country.
Fault passage indicators, whilst greatly facilitating fault location, are generally
too expensive to permit their permanent installation in the profusion required to
derive their full benefit whilst the portable form involves delay and expense in
setting out and collection for each fault location.
The elimination of h.v. fuses from networks utilising automatic reclosing offers
economic and operational advantages.

15.7 Primary networks in rural areas

Economic considerations play a large part in the design of the 66 kV and 33 kV

 The application of protection to rural distribution systems 283

primary networks in rural areas. Ring main schemes are commonly employed to
afford alternative supplies and use is also made of feeder transformers and teed
feeders. Overhead lines are generally supported on wood poles and switchgear is
kept to a minimum, single circuit breaker substations being common practice even
though involving some loss of operational flexibility.
Pilot channels for protection are seldom available. Sometimes combined with an
earthing conductor, they may be carried on the overhead line supports but such
circuits are expensive and liable to faults caused by climatic conditions and other
external influences. In addition, it is necessary to guard against the effects of
induction from the primary conductors.
On some 33 kV systems, particularly those having a predominance of steel tower
lines, arc-suppression coils are sometimes used for protection against earth faults.
With the increasing use of wood pole unearthed type construction, however, their
use is rarely extended to new systems.
On simple ring main systems, current and time graded directional overcurrent
and earth-fault protection has been employed for many years. Such protection,
however, has long operating times for some types of faults and is limited in applica-
tion to ring mains having not more than five substations. These limitations together
with a general lack of pilot channels for unit schemes, have provided the incentive
for the development of the cheaper forms of distance protection incorporating
switched measuring elements, as described in Chapter 9.
The protection of feeder transformers has been discussed in Chapter 12 but on
rural systems overall schemes utilising pilots are rarely used for the reasons stated
above. Instead simpler forms of protection are usually employed. One such scheme
for a single circuit is shown in Fig. 15.7A. At the source end of the line the circuit
breaker is equipped with instantaneous earth-fault and inverse definite minimum
time overcurrent relays, the latter incorporating high-set instantaneous elements.
The rapid clearance of line faults by the use of relays without time lags minimises
damage at the fault point and consequently increases the possibility of successful
reclosure of the circuit breaker.

• I. . . . . . . "1
I , I
I -"#

i -- -J I

T '

'
I
I
I i
'
'
I I
mw I
"J" I
L . . . . . . .~

Fig. 15.7A Schemeof protection for single feeder transformer

284 The application of protection to rural distribution systems

H.V. L.V. H.V. L.V.

N.P.D. relay N.P.D. relay

IILv
(a) Voltage transformer energised
H.V. L.V. H.V. L.V.

i l
~, •

T~ " "T" bushings TI__~ nser hushing

Z~--.
°
~.P.D. relay
~
,,g,
D. relay

(b) Condenser hushing energised

Fig. 15.7B Neutral point displacement protection

Faults on the h.v. winding of the transformer are detected by the protection at
the source end whilst restricted earth-fault protection is applied to the secondary
winding. In addition, overcurrent and earth-fault protection is fitted on the
secondary side to guard against busbar faults or the failure of circuit breakers to
clear faults on the outgoing feeders. Protection against interturn and other internal
faults which cause gas evolution is provided by a Buchholz gas and oil relay.
As an alternative the transformer may be protected by an overall scheme, but
such an arrangement requires current transformers on the h.v. side, usually accom-
modated in the terminal bushing turrets.
Intertripping of the source circuit breaker is necessary on the occurrence of a
transformer fault, and this is accomplished by a fault throwing switch connected
between one phase and earth or, in the case of arc-suppression coil systems,
between two phases. The fault throwing switch is closed on the operation of the
transformer protection by an electrically released manually charged spring mech-
anism and will ensure the tripping of the remote h.v. circuit breaker.
Where duplicate feeder transformers afford supplies to a substation, similar
schemes of protection to those used for single units may be applied. An earth fault
on the line, however, will cause the appropriate circuit breaker at the source end to
 The application of protection to rural distribution systems 285

;----I ..... ' --~

I I
, : .... .~
L_
..v. I
I !

I>- - -

I ". I I
I I I
I I - .. " I i
L -~ "- "J .L
B---1
r---f" - t "I I~l
' i ,_.L.~~ I I
:' :' L-=_
! I ~ ~ I I "-'-
; ,~ ~ , ou,~,,i.~
[~ L ~ l~----"# feeders

<

N.P.I). - Neutral p,)int displacement relay I I

I
I -• _]

Fig. 15.7C Scheme of protection for duplicate feeder transformers

H.V. L.V.

C.T.

Single
m
element
directional
overcurrent
re l a y

Fig. 15.7D Single element directional overcurrent relay for protection of duplicate feeder
transformer

trip but the fault will still be back-fed through the transformer. Since the neutral
point of the transformer h.v. winding is not usually earthed, the voltage
to earth of the two healthy lines will rise to the full line-to-line voltage. This
condition may be detected, and the transformer I.e. circuit breaker tripped,
by the use of neutral point displacement protection at the receiving end. This form
286 The application of protection to rural distribution systems

Protective
relay contact
Timing or o.c.b. Auto/non-auto
relay aux. switch switch

]Aux. relay ~~~ Clis: [

(hand reset)

contactor ~ t gr switch
coil
Trip

Fig. 15.7E Simple nonrepetitive single-shot reclosing scheme

of protection uses a relay connected to respond to a residual voltage, the relay

being energised by a voltage transformer or condenser bushings, the latter method
being somewhat cheaper. The arrangements for differing transformer connections
are shown in Fig. 15.7B. A time-lag relay having a setting of approximately 5s is
normally employed in conjunction with the neutral displacement detecting relay to
prevent operation when faults occur on other sections of the network. A typical
scheme for the protection of duplicate feeder transformers is shown in Fig. 15.7C.
Another form of protection which has been successfully employed experi-
mentally to bring about tripping under the condition described above uses a
sensitive reverse power relay to detect the reverse magnetising current of the
transformer. With modern transformers utilising cold rolled steel cores, however,
the loss and magnetising currents are very small and special relays and current
transformers may be necessary. Compensation for phase angle errors on the current
transformers may be required and is provided as an integral part of the relay, this
feature being cut-out during heavy through fault currents. The application of this
scheme is generally limited to unteed parallel feeder transformers since tees intro-
duce the possibility of incorrect operation.
Protection against the back-feeding of phase-to-phase faults occurring on the
lines of duplicate feeder transformers may be afforded by directional overcurrent
protection installed on the lower voltage side of the transformer. For this purpose
the 30 ° relay connection is unsatisfactory and the 90 ° (or 90 ° minus 45 ° ) con-
nection should be used. Where the transformers are delta/star connected, a single-
element relay, connected as shown in Fig. 15.7D, may be employed for this
purpose since a phase/phase fault on the delta side will produce a 2:1:1 current
distribution on the secondary side. (See Chapter 12).
The principles of automatic circuit reclosing discussed in Section 15.3.1 apply
equally to rural 33 kV and 66 kV systems. Feeder circuit breakers on such systems
are therefore generally fitted with reclosing facilities, usually of the nonrepetitive
single shot type, that is the type in which only one reclosing operation is provided,
lock out of the circuit breaker occurring on subsequent tripping. For this purpose
 The application o f p r o t e c t i o n to rural distribution systems 287

II.V. suppl~

/ / /
|
! _I , !
II

,!
(.

/'3
\"~ ~ It ;i~'.c
t ra 11% f< ~r111 L'r

""-,
~ ~-A i~',':~',~ ,~.,,.,).
! ~i,~.
L.V.
circuil
hrca kL'r
.... ~~,,~,~,,,~.~, [ I~° .

~i iI ItC | ()|l-~lIJ[~)
k)
~.,,,~,,,~
circui t
,4,, 2,,
I1~
,.\ u x .
~uppl.v

L.V. hush~lrs

Fig. 15.7F Initiatior., of circuit breaker reclosing by voltage relay

288 The application of protection to rural distribution systems

a simple reclosing scheme may be used, a typical example being shown in Fig.
15.7E.
The use of automatic reclosing features permits the use o f instantaneous pro-
tection in addition to the more usual i.d an.t. forms, as described in Section 15.3.4
and such schemes are being increasingly used on rural primary systems. As a result,
fault damage is minimised and the number of prolonged interruptions o f supply
due to line faults is reduced.
With some control arrangements it is necessary to reclose automatically the
circuit breakers on the lower voltage side o f some system transformers. This should

---1

I I
\ I
£- g
I I I ex
I e~
I I E:
I I
I I
e~

= I I

q
e-

x~
J O.
E

e.
-]'~ .O
--q
,- .Q
I
I g.,
,¢
I
I
I
I I I
I i Y

~ °~o~
I I
I I
E
q, I
j i
1. . . . 4---- - -> --I
T I
I I
I N-- -4
I I

i I
I
I '~ L9

tL
 The application of protection to rural distribution systems 289

only be carried out if the primary circuit is healthy, consequently tile reclosing
feature is initiated by a voltage relay energised from a voltage transformer on the
secondary side of tile main transformer. A typical scheme is shown in Fig. 15.7F.
Extensive use is made of automatic reclosing in a single circuit breaker sub-
station, the schematic diagram of which is shown in Fig. 15.7G. In considering
the functioning of this scheme, let it be assumed that line A is normally the in-
coming supply, whilst line B is tile outgoing supply to an adjacent substation.
A fault on line A is cleared by tile tripping of the local and remote h.v. circuit
breakers together with the intertripping of transformer A lower voltage circuit
breaker to prevent a back feed to the fault. The line is subsequently re.energised
by the automatic reclosing of the remote breaker and this is then followed by the
reclosing of the local h.v. circuit breaker if tile voltage relay controlling the re-
closing feature of this breaker remains energised for a period exceeding its time-lag
setting. This time-lag is set to a value in excess of the operating time of the
protection at the remote end of the line so that on persistent fault no reclosing of
the local breaker occurs. The automatic reclosing of the transformer lower voltage
circuit breaker is governed by restoration of supply to the transformer as
mentioned in the previous paragraph.
Operation of the scheme in the event of a fault on line B is similar to that
described above, except that the local b,.v. circuit breaker recloses automatically
after a predetermined time lag without voltage restoration control. It thus acts to
clear persistent faults on tile outgoing feeder in a manner similar to that in which
the circuit breaker at tile remote end of line A clears such faults on that line.
On transformer faults the transformer protection is arranged to trip the local h.v.
and appropriate transformer l.v. circuit-breakers and to close the fault-throwing
switch to ensure the tripping of the remote h.v. circuit breaker. After a time-lag
sufficient to ensure operation of this latter circuit breaker, the faulty transformer is
isolated by the opening of the power operated disconnector, following which the
local and remote h.v. circuit breakers are reclosed automatically.

15.8 Bibliography

Regulations
Overhead line Regulations, HMSO

Articles
'Line protection by Petersen coils" by H Willott Taylor and P F Stritzl (].1l:7"." 82,
page 387, April 19 38)
'Automatic circuit reclosers' by Peirson. Pollard and Care (].11£l:', 102 Pt.A, (6),
1955)
'Auto-reclosing switchgear in distribution practice' by S H Money and J Harris
(Proc. lEE, 115, (2), 19 68)
'Discrimination with expulsion fuse switches' (k;'lec. Times, January 1962)
290 The application of protection to rural distribution systems

'High speed reclosing on 11 kV rural networks' by S H Money (Elec. Times, 18

June 1959)
'High speed tripping and reclosing on rural networks' by S H Money (Elec. Times,
15 & 22 June 1961)

Reports
'Arc suppression coils and auto-reclosing switchgear', The Electricity Council,
(DRP Report No. 1,1962)
'Report on standardisation of auto-reclosing facilities on 11 kV ground mounted
metalclad distribution switchgear', The Electricity Council, (ACE Report No.11,
1966)
'Lightning protection of distribution networks', The Electricity Council, (Report
Ref. ECR/R 566, 1972)
'Report on auto-reclosing schemes for 11 kV distribution networks The Electricity
Council (ACE Report No. 54, 1977)
 Chapter 16

The application of protection

to urban and metropolitan systems
by K.A.J.Coates

16.1 Introduction

In the early days of electricity supply, many small generating stations supplied their
own local areas and there was little intcrconnection. As the load grew, larger
stations were built and transmission networks were superimposed which intercon-
nected the larger stations and augmented supplies to areas where the local generation
was inadequate. A complex system of generation, transmission and distribution
evolved having a variety of different voltage levels for generation and transmission,
although distribution was usually at 6-6 kV or 11 kV. Much of this still exists but
with the siting of new stations away from large concentrations of population and
the introduction of 275 kV and 400 kV transmission a new pattern is emerging.
Typically, energy will be received at a 400 kV supergrid point where it will be
transformed down and distributed by systems operating at 132 kV and 33 kV to
primary substations which in turn will transform to 11 kV. At the higher load
densities, for example, in city centres there are economic advantages in having only
one system level between 400 kV and 11 kV and if the 33 kV level is dispensed
with, as may be the case, the whole system needs only two stages of transformation
namely 400/132 kV and 132/11 kV.
At the higher system levels, the units, e.g. transformers, busbars, cable circuits,
are large and repair and replacement times are lengthy. Failures at this level tend to
affect large numbers of consumers tor long periods. To safeguard against such
extensive outages, plant and circuits are duplicated and usually operate in parallel.
Protection is required to identify correctly and initiate the disconnection of any
faulty unit at the same time remaining stable against faults at lower system levels
for which other protection exists. The provision of adequate, often complex,
protection at high system levels is expensive but its cost is usually a small percentage
of the capital value of the main plant.
At the lower system levels the units become progressively more numerous,
duplication ceases, fewer consumers are supplied until finally only the consumers'
292 The application of protection to urban and metropolitan systems

service remains to be protected. To correspond with these conditions the

protection becomes simpler and cheaper but must remain reliable so that the scale
of servicing involved is contained within reasonable limits.
Economic considerations demand attention to the fault susceptibility of the
various classes of equipment and if faults are rare then only minimal protection
may be justified and some risks accepted. Protection at some other point is then
relied upon to give disconnection.
Protection may be necessary to comply with an explicit or implied statutory
requirement; the provision of protection on a consumer's service is a requirement
under the Electricity Supply Regulations (superseded).
Although the primary function of protection is usually to minimise loss of
supply due to faults it may also be valuable in reducing damage to capital plant
by limiting fault current duration.
Thus the extent and type of protection afforded at any point may depend upon
a number of factors. The availability of comprehensive fault statistics enables fault
probabilities to be calculated which serve as a guide to the value of protection at
a particular point on a system. Because protection affects the way in which a
system operates under fault conditions it is rarely that protection arrangements can
be assessed in isolation from broader system design considerations, however.

16.2 Characteristics of urban and metropolitan areas

Towns and cities usually have a concentration of commercial premises at the

centre, surrounded by predominantly residential areas which in turn merge with
surrounding semi-rural areas. Industrial premises may be present, scattered in small
pockets or concentrated in a particular area.
Typically, the distribution system comprises a 415/240 V network supplied
from 1 lkV/415/240V substations with a capacity ranging from 300 kVA to 1000
kVA. 1 lkV feeders each supplying perhaps a dozen such substations, emanate from
primary substations which in turn may be supplied at 33 kV, 66 kV
or 132 kV. In these primary substations there are usually two or more transformers
ranging in capacity from 5 MVA to 60 MVA. This type of distribution system com-
monly supplies the whole range of load densities from, say, 100 MW/km 2 in the
centre of the largest cities to the 5-12 MW/km 2 typical of residential areas down to
the much lower densities of semirural areas. At the highest load densities substations
may be only 30m apart compared with 200m apart in residential areas. In the latter
areas voltage regulation in the 415/240 V network may affect the spacing of
substations, a consideration which does not apply in the high density areas where
spacing is determined entirely by the load and substation transformer capacity.
In the central areas substations are of necessity often situated in the premises of
the larger consumers. As an alternative to 415/240V supply, some large consumers
may be afforded supply at 1 l kV but this arrangement is more common where the
load is industrial. The 415/240V and l lkV distribution is carried out almost
invariably by underground cable.
 The application of protection to urban and metropolitan systems 293

Primary I ~,-,~),,,,,,,, ! I k V
substation

H.V. network

--oZ~z"o ...........
~d~ H.V. fuse-switch

l:t Transformer

L.V. busbars

415/240 V

Service
to building

l L__i_ . . . .

"---'i
l
I
I

I..V. network

Srv,cfuse
Link box

Services

Fig. 16.3A Elements of a distribution network

294 The application of protection to urban and metropolitan s/stems

16.3 Distribution system protection - radial I.v. systems

The elements of a distribution system are shown in Fig. 16.3A, and will be used to
discuss the protection aspects.

16.3.1 Services

The first part of the distribution system to be protected is the service cable, and this
is achieved by a service fuse at the termination of the distribution company's
service cable on the consumer's property. For special reasons this may take the
form of a circuit breaker, but in the majority of cases the best and cheapest
protection is a high breaking capacity (h.b.c.) fuse. Usually, the size of the fuse
should be related to the current-carrying capacity of the service cable and not to the
load applied for or taken by the consumer.

16.3.2 L.V. cables

The cables which make up the l.v. distribution network require protection against
faults and it is common practice to fit cartridge fuses in the phase conductors of
each outgoing cable at the l.v. busbar in the substation. Fuses may also be inserted
in the underground network at cable junctions where link boxes exist.
In choosing the size of fuse the following three points need to be taken into
consideration, bearing in mind that they are not always compatible and a compro-
mise is often necessary.

(a) The fuse provides protection for overcurrent as a result of faults and not
overload. Overloads can occur on distribution cables through increasing demands,
but these are unlikely to be of such a magnitude that they cannot be carried safely
for several hours without damaging the cable. Overloads can also occur through
alteration due to load transfers, and in these cases the increase in load may be much
greater. Nevertheless there is still an interval of time before the cable may be
damaged. The size of fuse must be chosen so that it does not blow at times of peak
load, but at the same time safeguards the cable against overcurrent due to break-
down of insulation. The size of fuse is therefore chosen more in relation to the
short-term rating of the cable than the size of the load. Thus a cable considered
appropriate for a 300A load could be fused with a 600A or 800A fuse. A further
factor which needs to be considered is the degree of discrimination which should be
obtained between the l.v. fuse on this cable and the protection on the transformer
in the substation.

(b) Faults on the consumer's premises which are not cleared by the consumer's
own fuses or which occur between the distribution company's cutout (fuse) and the
consumer's fuses, must be cleared by the distribution company's fuse, and
therefore the fuse at the substation or in the network must be such as to discriminate
with the largest fuse in the Board's cutout.
 The application of protection to urban and metropolitan systems 295

It should be noted here that the remark made previously, that is that the size of
fuse in the distribution company's cutout should be related to the current-carrying
capacity of the service cable, requires no modification by the points made above.
That is to say, where the distribution cable and service cable are of the same size,
the fuse on the distribution cable should be the larger. This fact will help
discrimination.
(c) The minimum fusing current of the fuse must be less than the minimum
earth-fault current. This is often difficult to achieve when the density of the load is
low, or in distribution systems where the size of distributor is chosen to suit the
current at every point (that is a tree or tapered system), and therefore where the
conductors become smaller and smaller as the distances from the source become
longer.
In order to ensure that an earth fault at the end of a long length of small section
cable is cleared, a fuse could be used in a link box where the cables are joined, the
fuse rating being chosen to discriminate with a larger fuse further back in the
network. Fuses can also be used in this way in an attempt to improve the
continuity of supply by trying to limit the number of consumers which will be
without supply in the event of a cable fault.

16.3.3 Substation transformers

Consider now the protection of the transformers in the substation. Due to their
large number, the relative unimportance of each unit to the system as a whole,
their good record of reliability, and the high percentage cost of giving complete
protection, economics play a large part in determining the type of protection to
be used.
In the simple case of a single transformer, plain h.v. fuse protection is the
cheapest, but careful consideration must be given to the choice of fuses. In urban
areas it is not unusual to give a supply to a single consumer in the order of 250 kVA
at 415/240 V. While it might be felt desirable to supply such loads direct from
separate substations, this is not always economical or practicable, and they will
have to be supplied from the 415/240 V networks. For such a load, the service
fuse should be rated at 400 A, which necessitates an 800 A fuse on the outgoing
distributor at the substation in order to maintain a ratio of 2.
On an 1 lkV system, the ratio of transformation would be something under 25,
and therefore to maintain a ratio of 2 with an 1.v. fuse rating of 800 A, the h.v.
fuse rating should not be less than 64 A. The discrimination which would be pro-
duced by such an arrangement is illustrated in Fig. 16.3.3A which shows typical
curves for the fuses mentioned. The curves used, however, must be those for the
actual fuses employed. It must be remembered that the reactance and size of the
transformer sets an upper limit to the prospective 1.v. fault current, and that any
impedance in the fault itself will further reduce this current. There is no point in
trying to achieve discrimination at currents greater than those likely to be found in
practice.
 296 The application of protection to urban and metropolitan systems

-~ |
Maximum short circuit
current for
o.s
.E 500 kVA transformer

._

S0 I00 500

C u r r e n t in p r i m a r y amps at I I k V

Fig. 16.3.3A Typical fuse characteristic for discrimination

No mention has yet been made of relationship between the size of the h.v. fuse
and the size of the transformer, but in considering this aspect it must be
remembered that simple fuse protection may give little or no protection for earth
faults at the middle of the h.v. winding and that, as the ratio of the size of fuse to
full load current increases, less of the winding is covered for earth faults.
Furthermore, fuse protection is not intended to guard against overcurrents due
to overloading of the transformer, but primarily against overcurrents due to faults.
These facts, coupled with a limited choice of h.v. fuses, makes for a certain lack of
relationship between the full load current of a transformer and the rating of the
h.v. fuse.
In the UK, the ratings of h.v. and l.v. fuses appropriate to the various sizes of
distribution transformers in use have been standardised and an Electricity Supply
Industry Standard No. 12-8 deals comprehensively with this subject.
Where economic considerations justify a greater degree of protection, then the
single transformer can be protected by a circuit breaker with overcurrent and
earth-fault protection. If relays are used instead of direct-acting trip coils, then the
protection can be set more accurately, but this introduces the problem of providing
the necessary source of supply for tripping the circuit breaker. If only two-pole
overcurrent relays are installed, the setting must be chosen to ensure that the relay
will operate for a phase-to-phase fault on the secondary side of a delta star
transformer which produces a 2:1:1 distribution of current in the primary circuit.
Whatever the type of protection chosen for a substation transformer it is necessary
to bear in mind that it is required to operate and clear the fault before operation
of the protection on the outgoing feeder at the primary substation from which it is
fed.
 The application of protection to urban and metropofitan systems 297

When two transformers feed a common busbar, steps may need to be taken to
ensure discrimination when the healthy transformer feeds the faulty transformer
through the l.v. side. In such a case neither fuses nor overcurrent and earth-fault
relays will provide this discrimination and furthermore circuit breakers must be
provided on both the h.v. and l.v. sides of the transformers. The protection of two
such transformers could take five forms:

(a) Suitable unit protection, for example Merz-Price, which will provide dis-
crimination for phase and earth faults on both sides of the transformers.
(b) Overcurrent and earth fault on the primary side, restricted earth fault on
the secondary side, and intertripping, which will provide discrimination for
earth faults on both sides of the transformer.
(c) H.V. and l.v. overcurrent and Buchholz, which will provide discrimination for
faults inside the transformers.
(a) Overcurrent and earth fault on the primary side, directional protection on
the secondary side, and intertripping, which will provide discrimination for
phase and earth faults on both sides of the transformers.
(e) Frame leakage, Buchholz, h.v. overcurrent and intertripping which will
provide discrimination for internal faults and earth faults on both sides of
the transformers.

In practice, in radial l.v. systems, transformers are rarely operated in parallel on a

common l.v. busbar because their failure rate is not high enough to justify the cost
of the additional switchgear and protection equipment involved.
The question of whether to apply unit busbar protection to either the l.v. or h.v.
busbars cannot be seriously considered. The cost is out of all proportion to the risk
of failures and the small advantage to be gained by installing it.

16.3.4 H.V. cables

For the protection of the h.v. cable feeding the substations, the same considerations
apply as in the case of the protection of the substation transformer. If there is only
one h.v. cable and the possibility of back-feed need not be considered, then plain
overcurrent and earth-fault protection at the main substation is sufficient. If the
h.v. cable is run either directly or indirectly in parallel with other h.v. cables, some
form of unit or directional protection is necessary.
Again, the relatively heavy cost of providing protection in such circumstances
has to be considered, and therefore a layout is often chosen which avoids h.v. cables
running in parallel, yet gives facilities for alternative means of supply. Such a layout
uses ring mains, as illustrated in Fig. 16.3.4A and the ring is normally open at a
convenient point. Under such circumstances the protection at the main substation
would again be overcurrent and earth fault. The setting of the earth fault relay
should be as low as possible consistent with sufficient discrimination against the
earth-fault relays or fuses for the transformers in the substations. The overcurrent
298 The application of protection to urban and metropofitan systems

setting should be as high as possible consistent with the size of cable and its short-
term rating, as already described in connection with fuses for 1.v. cables, and
provide discriminaticn with relays closer to the source of supply.
If grading with overcurrent and earth-fault protection at the primary substation
is possible, there is advantage in installing similar protection at a point on the ring
midway between the primary substation and the normally open point. For those
consumers nearer the primary substation this gives a probability of a 50% reduction

t°
¢
T

!1 kV

Normally
open

ff---q
Note" ] EFI] signifies Earth fault indicator
1 1

Fig. 16.3.4A H.V. ring main system

 The application of protection to urban and metropolitan systems 299

in outage time for h.v. faults.

When a fault occurs on an h.v. cable in a radially operated ring main system
substations are shut down until the fault is located and supply is restored by
resectioning.
The process of fault location is facilitated if earth-fault indicators are installed
at appropriate points on the network as shown in Fig. 16.3.4A. The majority of
faults are earth faults, or commence as earth faults before changing to phase faults,
and as such can be detected by a core-balance current transformer fitted over the
h.v. cable. The current transformer operates an attracted armature type relay having
a manually reset flag indicator. Inspection of the indicators immediately after a
fault enables the extent of the path of earth-fault current to be determined, leading
to quicker identification of the faulty cable section. The relay has a brief time delay
to prevent response to transient unbalance and is sensitive to earth fault currents of
50A or more in the main conductors. It is usually necessary to ensure that the flux
due to earth-fault current in the phase conductor is not nullified by the return of
the same current, in reverse direction, in the cable sheath. The electrical continuity
of the cable sheath is therefore broken, usually at the point at which the current
transformer is fitted. Successful use of this particular indicator is dependent upon
all indicators being reset soon after the event in readiness for any future fault. An
alternative type is available with an automatic voltage operated reset feature. The
voltage source is the l.v. supply in the substation in which the device is installed
and resetting therefore takes place at the instant when l.v. supplies are restored.
If it is essential to maintain a firm supply on a given busbar under fault con-
ditions on one of the feeding circuits, two or more h.v. cables must be run in
parallel to these busbars, each suitably equipped with protection to ensure dis-
crimination and automatic isolation of the faulty cable. In general, the distances
are short in urban networks, and therefore a unit protection scheme with pilot
wires can be used.
An h.v. network which affords firm supplies to four substations in this way is
shown in Fig. 16.3.4B. Each cable circuit is equipped with unit protection using a
pilot wire system, thus there are a total of seven protected zones. A fault in any
protected zone causes tripping of the circuit breakers at each end; the affected
cable is lost but supply is maintained by the other cables. It is necessary to ensure
that the different current distribution after a fault at any point on the network,
does not give rise to overloads.
The method affords effective protection against cable faults but there are other
effects which need to be taken into consideration. At each substation the protected
zone extends only to the point on the main circuit conductors at which the current
transformers are installed, so that the busbar, switchgear and most of its internal
connections are outside the zone. There is a possibility of pilots being open-
circuited, a trip circuit failure or failure of a circuit breaker to trip for mechanical
reasons. A cable fault coupled with failure of the unit protection, or a switchgear
fault, are similar in that they can only be cleared by circuit breakers further back in
the system, resulting in the loss of all supplies to the network under consideration.
Although this could be alleviated by the provision of additional safeguards, e.g.
300 The application of protection to urban and metropolitan systems

directional protection, busbar protection, pilot-monitoring etc., the cost of such

additional equipment is difficult to justify and the method usually adopted is to
install overcurrent and earth fault i.d.m.t, relays at selected points to sectionalise
the network and limit the extent of outage. In Fig. 16.3.4B, the i.d.na.t, relay

:-(03_...
, , , = , , ,,, ,,,,

D @ E3

lng rt,);l ~

{
F ig, 16.3.4B Unit~protected h.v. ring main system

installed in the ring separates the network into two groups each of two feeders and
only one group is shut down for the type of fault under consideration.
The shut down of the group involves operation of a total of three relays, two on
the feeders at the primary substation and the ring relay. Until the fault is cleared by
operation of all three relays, fault current is sustained by the transformer at the
 The application of protection to urban and metropolitan systems 301

primary substation which will be tripped if its relay operating time is exceeded. It
is instructive to consider how faults at different positions affect the operating
sequence of the three network relays and the total clearance time. If a fault occurs
in a feeder cable-box at the primary substation virtually the whole of the fault
current will flow directly through its own circuit breaker which trips at the relay
setting. The fault current path is then via the other three circuits, feeding the
fault from the remote end. The fault is now seen by a total of four relays, the three
on the feeders plus the ring sectionalising relay. If the feeder impedances are
roughly similar and the ring sectionalising relay has a lower time setting it will be
the first to trip thus cutting off fault current in two feeders. The remaining feeder
is the only circuit now feeding the fault and its relay operates to complete the fault
clearing sequence. The total fault clearance time is therefore a combination of fully
sequential and simultaneous relay operating times. By comparison, a fault at the
remote end of the network, say on the busbars at a substation, produces significant
current in all relays simultaneously. The ring sectionalising relay will be the first to
operate thereby clearing the fault from two feeders and final clearance comes when
the two remaining feeder relays have both operated. Under these circumstances the
total clearance time will be different from the previous case and almost certainly
shorter.
With paralleled feeder networks, therefore, the position of the fault has a pro-
nounced effect upon the relay operating sequence and it is necessary to find the
condition which gives the maximum total clearance time. This is done by selecting
a fault position, performing network analysis to ascertain fault current distribution,
calculating relay operating times to find the first relay to operate and how far the
others have moved toward their operating points. The network condition after
the first relay has operated is then analysed to ascertain the revised current distribu-
tion and relay calculations performed again to ascertain the next relay to operate
and so on until the fault is cleared. The whole process is then repeated for faults at
different positions until the maximum total clearance time has been established.
The operating time of the transformer protection at the primary substation must
then discriminate with the maximum total clearance time.
The calculations are not so onerous as would appear because the factors
contributing towards the longest time become apparent after only a few fault
positions have been tried. Overcurrent and earth fault relays need to be considered
as two separate groups. Phase faults appear throughout the system as overcurrent
and as a result all overcurrent relays are involved in the time grading considerations.
The maximum time delay is fixed by protection considerations on the primary
substation transformer and the minimum by the necessity to discriminate with
protection on any distribution transformer. The preferred minimum discrimination
interval of 0-4s between stages usually severely limits the number of stages that can
be employed between the upper limit at the primary substation and the lower
limit at the distribution substation. Because earth-fault current does not appear as
such throughout the entire system there are usually fewer stages of earth-fault
protection to be graded between upper and lower limits and no difficulties arise.
302 The application of protection to urban and metropolitan systems

The method of affording firm supplies using unit protection by pilots with back-
up overcurrent and earth-fault protection has the merit of relative simplicity and
is economic if the pilots are laid with the main cables. A number of other arrange-
ments have been used. Unit protection by pilots may be modified to include the
switchgear busbars in the protected zone so that each successive protected zone
overlaps the next to this extent. As a result back-up protection is no longer required
to cope with busbar faults but other risks, such as pilot circuit failure, remain. Time
graded directional schemes as described in Chapter 2 eliminate the pilot difficulty,
are responsive to busbar faults and other failures but require a source of polarising
voltage usually derived from voltage transformers which add to switchgear costs.
Distance protection is inapplicable to high density urban distribution because of
difficulty in achieving sensitivity to the short distances involved and again a voltage
source is necessary.

16.3.5 Primary substations

Primary substations are the points in the system where supply is received at e.la.v.,
transformed to h.v, and fed via busbars and circuit breakers to the h.v. distribution
network. The h.v. distribution voltage commonly in use in the United Kingdom is
1 lkV; at e.h.v, the voltages are typically 33kV, 66kV or 132kV, the trend being
towards 132kV. There are usually two or more transformers ranging in size from
about 5 MVA to 60 MVA, the very large transformers having double secondary
windings to limit the fault level on the 11 kV side and the normal current rating of
the lower voltage switchgear to an economic value. In addition to the variety of
voltages and transformer capacities in use there are variations in the method of
neutral earthing, the arrangement of busbars and the facilities for voltage control.
There may be local e.h.v, switchgear controlling the transformers and possibly other
circuits or the transformers may be controlled by remote e Ja.v. switchgear in which
case each transformer and its associated e.h.v, cable is known as a transformer
feeder. The running arrangement is usually with pairs of transformers in parallel
to afford a firm supply to the 1 l kV busbar but schemes of separate transformer
running arrangements exist under which loss of supply to a section of 1 lkV busbar
is restored by coupling automatically to a healthy section.
The requirements of the protective gear are primarily:

(a) to clear a faulty 1 lkV feeder from the system before higher level protection
operates,
(b) to clear a faulty transformer or transformer feeder before the protection of
any parallel healthy circuit operates and before any higher level protection
operates,
(c) to limit the effect of busbar faults on other parts of the system.

In addition to these primary requirements there is a secondary requirement to limit

the effect of the failure of a particular part of the protective system to operate.
 The application of protection to urban and metropolitan systems 303

Requirements (a) and (b) are clear cut, but (c) poses problems as to the extent of
the provision to be made against busbar faults. At one extreme it may be argued
that busbar faults are rare and that provided higher level outage is protected against
then only a single busbar with a nonauto section switch between transformer
positions, is justified. On the other hand the magnitude of an outage, with possibly
lengthy restoration time may be considered as justifying a duplicate busbar arrange-
ment with unit protected sections. As one might expect, an arrangement some-
where between these two extremes is usually adopted but the busbar arrangement
may not rest on security considerations alone.
The various schemes of protection appropriate to primary substations are dealt
with in detail in Chapters 12 and 13.

16.4 Distribution system protection - interconnected l.v. systems

Various interconnected 1.v. systems are in use in large cities in this country and
abroad and such systems involve protection considerations over and above those
of a radial system.
Interconnected l.v. systems are particularly suited to city distribution because
the l.v. cables required to supply the load naturally form a practically continuous
grid over large areas. Thus the substations which supply such a network may easily
be operated in parallel through the l.v. network via solid links or cartridge fuses.
There are savings in both capital and running costs. The savings take place because,
to a greater degree than on a radial network, an interconnected network couples

I 1
18 18
d',
-" r----1

I L
6

!
m 1 I 1
6 2 2 6
I""1 ) . J J
'" L...._]

_1
L
6 6
Fig. 1 6 . 4 . 1 A Current distribution in an interconnected network
 304 The application of protection to urban and metropolitan systems

together the loads virtually all of which have some diversity one with another. Thus
the maximum demand supplied by the transformers is less and fewer are required or
their capacity is less. The percentage loading of transformers tends to equalise
because they are effectively in parallel, giving the condition for minimum copper
loss and since the total transformer capacity is reduced the total iron loss is less.
Losses in the l.v. network are also reduced because cables are fed from both ends
and because the number of parallel paths is increased.
The other important feature of interconnected l.v. systems is the generally
adopted practice of coupling together the l.v. networks supplied from substations
on different h.v. feeders and arranging that the l.v. network is immune to the loss
of a single h.v. feeder, the load being taken up on the remaining feeders in the
same group. This is a valuable advantage because the greatest single cause of loss
of supply to consumers is h.v. feeder faults.

16.4.1 The l.v. network

There are variations in the method of dealing with the 1.v. network. Fig. 16.4.1A
shows a fully fused arrangement, cartridge fuses being fitted in the phase
conductors at all link boxes. If the fault current distribution followed the ideal
pattern shown the fuses would discriminate and only those at each end of the
faulty distributor would operate. In practice irregularities in the pattern of im-
pedances between junctions upset the distribution of fault current and unwanted
fuse operations takes place. This situation is complicated by a further effect which
is that some faults clear themselves leaving fuses intact, particularly with the larger
size of fuse. The form of construction of the cable, the conductor material and the
prospective fault current all have a bearing upon the probability of fault self-
clearance. If all these factors are favourable then very nearly all faults will clear
which leads to the proposition that the fuses are unnecessary, or relatively so. On
this basis some interconnected l.v. systems have operated with large blocks of load
solidly interconnected, ultimate disconnection for a non self-clearing fault resting
upon l.v. protective gear at the substations. A disadvantage of this degree of solid
interconnection is that there are operational difficulties in restoration after a non-
self-clearing fault. Between the two extremes of fully fused and solidly intercon-
nected arrangements there are hybrid schemes which perform well.

16.4.2 Substations

In an interconnected l.v. system, the transformers are effectively connected in

parallel across the h.v. and l.v. networks and any faulty unit is required to be
rapidly isolated from both networks. The usual protection considerations for the
parallel operation of transformers apply and it follows that protection is required to
intorrupt quickly the backfeed to a faulty transformer from the l.v. network.
Although unit protection of the transformer could be employed, economic con-
siderations favour the use of directional protection on the l.v. side because this
protection is responsive also to the back-feed current from the l.v. network to an
 The application of protection to urban and metropolitan systems 305

h.v. cable fault. L.V. protection for faults in the forward direction, e.g. on the l.v.
network, is usually by direct-acting trips shunted by time fuses, the switchgear
being in the form of an air circuit breaker.
Protection on the h.v. side of the transformer is required to respond to internal
faults and faults on the h.v. and l.v. connections. This requirement may be met in
different ways. If the loss of an h.v. feeder due to a transformer fault is an accepted
risk then local protection can be omitted, reliance being placed upon the feeder
protection at the primary substation to give fault clearance. With such an arrange-
ment however there is difficulty in discriminating between the current due to heavy
load on the feeder and that due to a fault on the l.v. side of a transformer. Local
protection of the transformer overcomes this difficulty and at the same time
enables sensitive earth-fault protection to be provided economically for the trans-
former primary. The switchgear is usually an oil circuit breaker and the protection
is by direct acting trips with a time fuse delay for overcurrent faults and,
instantaneous operation for earth faults, the latter having a sufficiently low setting
to protect the whole of the primary winding. An intertrip to the l.v. circuit breaker
ensures that the l.v. back-feed is broken when the h.v. circuit breaker operates. H.V.
fusegear is a possible alternative to the h.v. circuit breaker, again with an intertrip
to the l.v. circuit breaker, but this does not provide the sensitive earth-fault pro-
tection.

16.4.3 The h.v. network

It is usual for the h.v. network supplying an interconnected l.v. network to

comprise a group of from four to eight h.v. feeders, each feeder operated radially
and supplying typically about ten 500kVA substations. There may be provision for
coupling feeders together at their extremities and elsewhere to the extent that if a
section of h.v. cables is lost due to fault, supply to all substations may be restored
by switching. Alternatively, such coupling facilities may be absent in which case
supply to an area which has suffered the loss of an h.v. feeder can only be
maintained through the interconnected l.v. network. A disadvantage of this arrange-
ment is that a fault at the primary substation end of a feeder may impose undue
stress on the l.v. network since the latter is required to support the whole of the
load of the feeder until the fault is located and repaired.
As an alternative to radial operation and at a rather higher capital cost, feeders
may be equipped with unit protection and operated in parallel. Supply to sub-
stations is then firm against any single h.v. cable fault and under this condition
load is not required to be supported through the l.v. network, other than that due
to the loss of a substation transformer.
The h.v. network may be interleaved or blocked. In an ideal interleaved network,
the h.v. cables are routed so that neighbouring substations are supplied from dif-
ferent feeders. In a blocked network one feeder supplies a number of neighbouring
substations which is termed a block and this block adjoins blocks supplied by other
feeders. If the number of substations in a block is small then typically the number
306 The application of protection to urban and metropolitan systems

of blocks is large and the blocks may be interleaved, giving a distributed block
arrangement. Interleaving reduces the burden of current across the l.v. network
which occurs under feeder outage conditions.

16.4.4 Protection on a distributed block l.v. interconnected system

Fig. 16.4.4A shows a distributed block system supplied from a group of four
1 lkV feeders fed from one section of busbar at a primary substation. Each feeder
on leaving the primary substation supplies an l.v. interconnected block through four
substations and this block is coupled to similar blocks on other feeders each by four
sets of fuses, termed fringe fuses. The feeder then continues to a more remote area
where it supplies another interconnected block similarly arranged. The diagram
shows a totally symmetrical formation for simplicity of presentation but a number
of variations are possible. For example, any feeder may supply more than two
blocks and the number of transformers feeding a block may be more or less than
four; this in turn affects the number of fringe fuses between blocks. The extent to
which such variationsmay exist is determined by a full knowledge of how the net-
work operates under various fault conditions but, in general, symmetry and uni-
formity ease protection discrimination considerations. Although not shown in the
diagram the configuration of the 1.v. network is frequently such as to allow direct
fusing between diagonally opposite blocks, e.g. A1 and C1, thereby improving the
uniformity of loading when a feeder is lost due to fault.
Three features of the way in which the network is required to operate under
fault conditions may be noted at this stage. An 1.v. network fault in an intercon-
nected block, if it does not self-clear, results in shut-down of the block, clearance
being obtained by operation of all the 1.v. circuit breakers on transformers feeding
that block and by operation of the fringe-fuses which connect it to other blocks.
An h.v. cable phase fault is cleared by feeder protection at the primary substation
and by reverse-power protection on the l.v. circuit breakers on all transformers
supplied from that feeder. Both the near and remote blocks on the faulty feeder
are therefore involved in this operation. An h.v. cable earth fault is cleared by
feeder protection at the primary substation but the cable may remain alive back-fed
from the 1.v. network since the reverse-power protection will not necessarily
operate under this condition. The cable can therefore, via the transformers
connected to it, provide paths in parallel with l.v. cables and assist in the transfer of
power across the supported 1.v. blocks.
Fig. 16.4.4B shows the protection arrangements. The 1 lkV feeder protection is
a two-pole overcurrent and single-pole earth-fault relay with a normally inverse
characteristic on all elements.
The transformer h.v. circuit breaker in each substation is equipped with direct
acting trips to give time fuse delayed overcurrent and instantaneous earth fault
protection. An auxiliary switch intertrips the l.v. circuit breaker.
Protection associated with the l.v. circuit breaker is current.transformer operated
and comprises direct.acting overcurrent trips with time fuses and a reverse-power
 The application of protection to urban and metropolitan systems 307

II k V -

Fig. 16.4.4A Distributed block interconnectedsystem

308 The application of protection to urban and metropolitan systems

D
11 kV
A "" B rl" C

I E! [ ~l E] 1 Ell

TTwo 5A trips
o/oTo%
r . ~ T w o 5A trips
L ~ with time fuses fuses
one IA EF trip F trip

f tu with time fuses

Z
[ ~ Three 5A OC trips
with time fuses

,n
415/240 V
_!
.. A

Service _~
to building
F~use

,w l , i

L.v. n e t w o r k X Y
Fig. 16.4.4.B Distributed block system protection
relay operating a shunt trip. The shunt trip also responds to tripping current from
the intertrip circuit and supply is taken via fuses, from the transformer l.v. con-
nections. Supplies to the voltage coils of the reverse-power relay are afforded in a
similar way.
 The application of protection to urban and metropolitan systems 309

The reverse-power relay has to satisfy a number of requirements. Because of

transformer reactance in the fault path the power factor of the fault current is
low and the relay is therefore designed to have its maximum sensitivity at about
55 ° lagging. Voltage depression may be considerable at the time when the relay is
required to operate and the relay, together with the l.v. circuit breaker shunt trip
coil, is designed for operation with the phase to neutral voltage reduced to 30V, to
suit typical worst conditions. Rapid disconnection of an h.v. cable phase fault is
necessary, otherwise fringe fuses will operate, and this gives a requirement for very
fast operation, typically 30ms. Subject to fast operation and stability under
transient conditions, the operating current is somewhat less critical, but 200% is a
practical value. The relay must be stable for faults in the forward direction and a
three-phase design rather than three separate single-phase elements overcomes the
problem of incorrect response in one phase which occurs under certain
asymmetrical fault conditions.
The characteristics of the fringe fuse used in the l.v. network is determined
having regard to the fact that under h.v. fault conditions the current experienced by
fringe fuses seldom exceeds 5 000A. In order to provide a 200ms delay at this
current to ensure discrimination with the l.v. reverse-power protection (including
circuit breaker operating time) a 400A fuse with a slightly modified characteristic
is used.
Referring to Fig. 16.4.4B and considering the chain of protection from the
primary substation down through feeder B to a single substation and thence to the
l.v. network at point X, it is evident that discrimination is unlikely to present
problems except possibly between the time fuses either side of the substation trans-
former. These fuses are effectively in series, however, so that correct discrimination
serves only as a guide to fault conditions and the performance that may be obtained
by grading the time fuses is adequate for this purpose. Correct discrimination may
also be anticipated if consideration is extended to point Y on the other side of a
fringe fuse and thence to point Z on the remote side of reverse-power protection at
a substation on feeder C. The reverse protection at Z operates, all other protection
protection on feeder B remaining stable. Correct discrimination at point Y is
however dependent upon there being only one fringe fuse in the circuit and this
gives rise to a fundamental principle which relates the number of fringe fuses
between interconnected blocks to the total transformer capacity in the block. The
relationship which has been established is that the number of fuses should not
exceed one fuse per 750kWA of transformer capacity.
It is necessary to consider the stability of h.v. feeder overcurrent protection
when an l.v. block supplied by the feeder experiences a non-self-clearing fault. At
each substation supplying the faulty block the l.v. circuit breaker overcurrent
protection is subject to that substation's individual contribution but the feeder
overcurrent protection is subject to the sum of these contributions. In addition
the feeder carries the load of a healthy block which it also supplies. There is in
consequence a possibility of failure of discrimination between the feeder protection
and the substation l.v. protection. If the feeder protection setting is determined by
 310 The application of protection to urban and metropolitan systems

100 .. . ~
!
I! '~
]
=
:1
I I .I J , _

50 • 120 A

H.V. feeder
I L ) M T L 5 A relay
4 0 0 / 5 c.t.
10 1 5 0 % 0 . 2 T M 11 k V
Scaled for M V phase-
5 phase fault relationship

/
.5
/ p

:E
/
/

-o 30
C \
10 k ,
\ \ % .....
.

_ ~
_ ~\\!\ .....

\X~ k
, .

- ri _-~..% ." ' ~

,.., ._-~ . . . . _...,,~ _
. ,, ,

0.5 - r .. ~
_ \

4 I.v. c i r c u i t - j#" ~N -
breakers 1200/5 / N~ ....
i ,, _
c.t. 7.5 A t i m e f u s e s

0.! i i
\
I I L
[
" •. .
i.
. . . I N " . , " .' . . . .

L 1 N~
0.05 -~\
f -.. - -_ , _.

_ ~..

"\
0.01
400 500 1000 3000 5000 I0000 20000
I I kV Line Current, Amps
11000 22000 66000 110000 220000
I I |m I i mn i • | • i I, ni

MV Line Current, Amps

(i.e. 11 k V C u r r e n t x 2 2 )

Fig. 16.4.4C Protection characteristics

H.V. feeder and I.v. circuit breakers
 The application of protection to urban and metropolitan systems 311

I00 l m
I m

50 I
/
/
I
IO l
5

.E \
H.V. feeder
Load
I D M T L 5 A relay
240 A
4 0 0 / 5 c.t.

r -i 150% 0.2TM 11 kV
Scaled for MV phase-
30 phase fault r e l a t i o n s h i p
J
~D

!0 '
[!
•

"-,,,.
\
0.5

/
il
6 x 400 A
fringe fuses
O.I

0.05 "

0.01
200 500 1000 3000 5000 10000
11 kV Line C u r r e n t , A m p s
4400 I I Q00 22000 66000 110tO0 22o9oo
• i II i

MV Line C u r r e n t , A m p s
(i.e. l 1 kV C u r r e n t x 22)

Fig. 1 6 . 4 . 4 . D Protection characteristics

H.V. feeder and frinae fuses
312 The app/ication of protection to urban and metropolitan systems

other considerations, the need to maintain stability places an upper limit on the
total capacity of substations which may be coupled together to form an intercon-
nected block. The worst condition is for an l.v. phase/phase fault giving a 2:1:1
distribution of current in the h.v. phases under which the most heavily loaded phase
has an effective transformation ratio of 1:22. Fig. 16.4.4C shows a comparison of
the protection characteristic of the h.v. feeder protection adjusted for load on the
healthy block against the characteristic of the l.v. overcurrent protection of four l.v.
circuit breakers in parallel. The proximity of the curves indicates that the limit of
feeder protection stability is reached with four l.v. circuit breakers fitted with
120015 current transformers and 7.5A time fuses. This protection is appropriate
to a 750kVA transformer so that the upper limit of transformer capacity in an
interconnected block is four 750kVA transformers.
In addition to the condition of feeder stability for a non-self-clearing fault in
either block directly connected to the feeder it is necessary to consider the effect
on stability when the fault is in a remote block under which condition fringe fuses
are required to operate to isolate the faulty block. The feeder carries the load of its
own two blocks plus the load due to the fault current in the fringe fuses. As in the
previous case the worst condition is the 1.v. phase]phase fault. Fig. 16.4.4D shows a
comparison of the protection characteristic of the h.v. feeder protection adjusted
for the load on its two blocks against the characteristic of six fringe fuses in parallel.
It is seen that discrimination fails if the number of fringe fuses exceeds six.
The protection discrimination requirements described above are limiting con-
ditions which apply to a theoretically regular network and they therefore afford a
guide to design. Beneath these limits irregularities, which occur principally in the
l.v. network, may impair current sharing and cause instability under fault or load
current support conditions. In practice, these conditions are checked by a computer
programme which by network reduction converts the numerous impedances making
up the 1.v. network into equivalent impedances between substations and fringe fuse
positions. Currents are then calculated at these positions for three conditions:

(a) Current flow in the l.v. network and transformers due to an h.v. feeder phase/
phase fault, after the primary substation circuit breaker has opened.
(b) The same condition but after reverse-power relays have operated, to
determine the l.v. current flows to support the peak load in the blocks which
have lost h.v. supply.
(c) The support condition again but for an h.v. feeder phase/earth fault when the
h.v. cable remains in circuit as a link between transformer primaries and
forms a parallel connection across the supported blocks, giving a different
distribution of l.v. current.

Under condition (a), current in the fringe fuses is checked against the 5000A limit
at which discrimination with reverse-power relays fails. The l.v. phase/neutral
voltage is checked to establish that a minimum of 30V phase to neutral is available
for the operation of reverse-power relays. Under the two support conditions (b) and
 The application of protection to urban and metropolitan systems 313

(c), the method of analysis simulates load current by the application of a limited
fault of appropriate magnitude. The transformer and fringe fuse currents thus
derived are subsequently superimposed upon the normal loads at these points to
give a true representation of the condition, enabling points of excess loading to be
identified. Since the h.v. system is arranged in the form of a ring main and supply
to all substations may be restored, under single outage conditions, by switching, the
safe loading is that which can be sustained until switching is complete, a period of,
say, two hours.
Analysis of the network in the way described enables its performance under
fault conditions to be determined with considerable accuracy and the few sources
of instability which may exist are readily identified. They can usually be remedied
by modifying the impedance pattern of the network, i.e. by changing fringe fuse
positions or switching substations to other blocks. If this is not possible the pro-
cedure adopted is to forfeit some of the support capability of the network to
ensure that the protection discrimination requirements are met under fault
conditions.

16.4.5 Supply to large point loads

Large point loads due, for example, to major office premises frequently produce
demands in excess of 1000kVA. Such loads may be supplied from substations
incorporated in the normal way into a distributed block l.v. interconnected system
but the arrangement tends to be uneconomic in respect of the aggregate capacity
of connections to the l.v. network required to support the internal load of the
premises when h.v. supply is lost. Also there may be difficulty in locating any
nearby fringe fuses so as to avoid excess loading under the support condition. These
drawbacks are overcome by installing two or more transformers each equipped with
l.v. reverse power protection and supplied from a different h.v. feeder. Fig. 16.4.5 A,
shows such an arrangement using two 750kVA transformers and the protective
gear is similar to that already discussed. It should be noted that the l.v. busbar
forms a direct link between transformers on different feeders. Since this direct link
exists in the position normally occupied by fringe fuses its presence may affect the
number of fringe fuses permissible elsewhere between the same two feeders.
Stability considerations may also limit the number of fuses that can be installed
between the l.v. busbars and the external network.
Where a number of large point loads are grouped together the network may take
the form shown in Fig. 16.4.5 B. Small loads interposed between the major, directly
fed, loads are supplied by single l.v. distributors laid from point to point and fused
at each end. Ultimate clearance of an l.v. cable fault now rests on two fuses only
and the feeder stability considerations which relate to a similar fault under the
interconnected block condition no longer apply.
The electrical installations of large point loads may include plant driven by large
induction motors. Voltage depression on the system supplying such motors causes
the motors to act as generators, feeding into the system a current which decays
314 The application o f protection to urban and metropolitan systems

Primary
substation

Normally
,

Two 5 A OC trips Two 5 A OC trips

L,.J with time fuses Lir.J with time fuses
~ one I A El," trip ~ o n e 1 A EF trip
_

H.V. network
"/50 kVA transformers

C Three 5 A OC trips Three 5 A OC trips

with time fuses with time fuses

®
- - A A _ A

Services Services
to building to building

H.V. network External

network

Fig. 16.4.5A Method of supply to a point load

 The application of protection to urban and metropolitan systems 315

1 1 kV feeders

L.V. network ~ -14.r---"l -

Fig. 16.4.5B Supply to a group o f four p o i n t loads

Note: h.v. and I.v. switchgear omitted for clarity

with time over several cycles. This current may cause unwanted operation of local
reverse-power protection if the voltage depression is due to a fault on feeders other
than those affording the supply. The effect is dependent upon both the size of the
induction motor load and the characteristics of the reverse power relay and is
unlikely to cause problems if the motor load is less than about 400 h.p. The
characteristic of the relay may be compared with the calculated characteristic of
the generated current and it is possible to determine the extent to which the relay
needs to be delayed to prevent the effect from taking place. Alternatively,
additional equipment may be installed to block relay operation while motor current
persists. The additional delay which is introduced however is an undesirable feature
316 The application of protection to urban and metropofitan systems

in respect of a fault on one of the feeders affording supply, under which condition
fast clearance is desirable for motor stability reasons.

16.4.6 Supply to h.v. consumers

If supply is afforded to a consumer at high voltage from a single h.v. feeder and the
h.v. feeder suffers a phase fault, the feeder is shut down and the consumer loses
supply; reverse-power relays in all substations connected to the feeder operate and
isolate any back-feed from the l.v. network. If the fault is an earth fault, however,
the reverse-power relays do not operate and the consumer's load is back-fed from
the 1.v. network. The load therefore appears as an additional commitment on the
l.v. network and is supplied from a healthy source via fringe fuses. Additional fringe
fuses to cater for this load cannot be provided because the total number of fringe
fuses is limited by the number of transformers feeding the l.v. network, as
previously discussed.
The difficulty may be overcome, at the same time affording a higher degree of
security of supply to the consumer, by taking supply from two h.v. feeders and
using directional protection. Fig. 16.4.6A shows such an arrangement. The i.d.m.t.
relay is set to discriminate with similar relays at the primary substation and detects
fault current in either direction corresponding to a fault on either feeder. The
reverse-power relay which has a similar characteristic to the relay used with the
interconnected l.v. systems detects the direction of fault current and its changeover
contact directs tripping action to the circuit breaker facing a faulty h.v. cable. Thus
the supply to the consumer is firm against an h.v. cable fault and the burden on the
interconnected l.v. network is removed under feeder earth-fault conditions. It may
be noted that a fault on the h.v. busbars supplying the consumer results in the
opening of one local circuit breaker only, fault current is sustained on the other
local circuit breaker but cleared by operation of the feeder switch at the primary
substation end.
An alternative method of integrating the supplies to h.v. consumers with the h.v.
network supplying an interconnected l.v. system is to supply the h.v. consumer or
consumers from a unit protected h.v. network and to provide a spur feed to sub-
stations supplying the interconnected l.v. system from a circuit breaker fitted with
i.d.m.t, protection.

16.5 Private generation

In systems which supply urban and metropolitan areas there may be consumers
who operate private generating plant. The machines are usually of the synchronous
type with capacities in the range 300 to 3000kW, the smaller sizes operating at
415/240V and the larger usually at l lkV. They may operate continuously in
parallel with the public supply or as standby equipment with changeover or
facilities for infrequent parallel operation. Since the consumer may be afforded
supply at either voltage the layout of the consumer's installation in respect of the
 The application of protection to urban and metropolitan systems 317

Primary s u b s t a t i o n A
I lllll II A . "l I,I

l Ez ~
r-
I ""1

I I I I I
I i
I I I I
I I I
I
I I I I I I
I I I I I I
I ! I I I I I
I I I I I I I
I I I I I I I
I I I I I I I
L . . . . .... J t..... .... J I I !..... . . . . ..J

\
Interconnected
I.v. b l o c k s

, ,,, @

L_
I ,~
i r
I
I
I
I

H.V. services to
consumer

F ig, 16.4.6A Supply to an h.v. consumer in a system which also supplies an. interconnected
I.v. n e t w o r k
318 The application of protection to urban and metropolitan systems

interface between the public supply and the generation may take a variety of
different forms, too numerous for consideration here in detail. There are however
protection problems peculiar to the interface which can be considered in general
terms and which are separate from the specific generator protection considerations
dealt with in Chaper 12.
It is important that the private generation should not jeopardise the public
supply and equally that faults on the public supply should not interfere with the
private generation. The presence of generation increases the fault level on the
system and if generation is at high voltage this increase particularly affects the
whole of the network on the primary substation concerned. Practically the whole
of the generator's fault current contribution appears at the primary substation
busbars with the result that this is usually the point of highest fault level on the h.v.
system. The effect needs to be determined to ensure that the fault rating of switch-
gear throughout the system is not exceeded.
The neutral point of the consumer's system is required to be earthed. If the
consumer's generation is at high voltage, the system earthing of the public supply is
available during parallel operation but the consumer must provide his own earth for
periods when independent operation is taking place. This may be by means of
switched star point earthing but a preferable arrangement is the provision of an
earthing transformer connected via a nonautomatic circuit breaker to the
consumer's busbar and kept permanently in circuit. The transformer is required to
have normal protection against internal faults under which condition isolation is
by tripping the switches controlling sources of supply to the busbar to which it is
connected. System earth faults may be detected by earth-fault protection in the
connection between the transformer star point and earth. Normal overcurrent and
earth-fault protection at the supply terminals is provided by i.d.m.t, relays to safe-
guard the public supply against faults on the consumer's installation. This will also
respond to phase and possibly, earth faults, in the reverse direction, i.e. on the in-
coming public supply but may be supplemented with reverse-power protection,
polarising voltage being obtained from the voltage transformer provided for
synchronising purposes. The same source provides reference voltages for neutral
voltage displacement protection which is the principal protection for earth faults on
the incoming public supply. The objective in respect of faults on the incoming
supply is to clear such faults without affecting the stability of supply to the con-
sumer's load from his generation but problems may arise due to the protection
being sensitive to voltage depression caused by unrelated faults on the public supply
system.
Where generation is at low voltage the generator star point is connected directly
with the consumer's earth electrode system and then to the earthing facility associa-
ted with the incoming supply. The latter normally has a direct connection with the
star point of the secondary winding of the incoming supply transformer. With this
arangement, restricted earth-fault protection may be applied to the generator itself
but for faults outside the restricted zone and at the interface with the public
supply, overcurrent protection by i.d.m.t, relay is usual. Where there is more than
 The application of protection to urban and metropolitan systems 319

one generator three-wire machines may be used together with a static balancer
connected to a busbar common to all machines, thereby providing a single star
point which is earthed and effective for the complete system.

16.6 Future trends

In the application of protection techniques there is, in common with other aspects
of electricity distribution, a need to practice due economy in affording a reliable
supply to consumers.
Using data based upon comprehensive fault statistics available in the UK it is
increasingly common practice to apply methods of cost benefit analysis and system
reliability evaluation to system design and protection in order to meet this
objective, and the trend can be expected to continue. Coupled with these processes
exists the facility of carrying out analysis of networks for all types of faults using
computer programmes which may be readily updated to suit changing network
conditions. The precise requirements of protection schemes may therefore be
determined and their performance predicted.
So far as equipment is concerned the present trend is towards the development
of solid state devices which duplicate the performance of their electromechanical
counterparts; a radical change is therefore taking place in equipment design, as
distinct from its application. The solid state devices are more complicated and have
more elements than their electromechanical equivalents but a particular feature is
that their operating currents can be very much smaller. If advantage is to be taken
of this feature then a reduction in the size of instrument transformers is a probable
trend followed possibly by the introduction of new methods of detecting the mag-
nitudes and phase angles of currents and voltages in the main conductors.
A logical extension of the application of solid-state techniques in complex
systems could lead to substitution of individual solid-state relays by integrated
systems which receive all necessary data and perform the appropriate protective
functions and, in addition such functions as indication, alarm and metering.

16.7 Bibliography

Standards
ESI Standard 12-8: The application of fuse-links to 1 lkV and 6-6kV/415V distri-
bution networks
ESI Standard 37-2: MV distribution fuseboards
ESI Standard 41-5: Indoor metalclad switchgear; ratings up to 250 MVA at 6-6kV
and 350 MVA at 1 lkV
ESI Standard 41-12: Non-extensible ring main equipments incorporating an
automatic fuse-switch and switches; ratings 250 MVA and 1 lkV and 150 MVA at
6.6kV
ESI Standard 48-2: Fault passage indicators for 6-6kV and 1 lkV underground and
overhead distribution systems
320 The application of protection to urban and metropofitan systems

Engineering Recommendation G26" The installation and operational aspects of

private generating plant (The Electricity Council)

Article
'Application of test results to the calculation of short-circuit levels in large industrial
systems with concentrated induction motor loads' by Cooper, Maclean and Williams
Proc. IEE, 1969, 116, (11)
 Chapter i7

The application of protection to

transmission systems
by J . C . W h i t t a k e r

17.1 General principles of application of protection to transmission systems

17.1.1 Introduction

This chapter describes the application to transmission systems of the various

protection systems described in previous chapters, and gives the reasons for
choosing a particular protection system for a given application. The chapter is based
on the practice followed in the United Kingdom. This practice, which is constantly
changing as new requirements arise and as new protection systems become available,
is influenced not only by technical and economic considerations but also by
historical factors; it is also based on the protection requirements for solidly earthed
transmission systems, such as exist in the United Kingdom. The detailed protection
requirements for other types of power system, for example for a transmission
system earthed through arc suppression coils, would be appreciably different from
those described in this chapter, although many of the basic requirements would be
similar. Comparison of the practice in the United Kingdom with practice in other
countries brings to light many interesting similarities and differences, but it is
beyond the scope of this chapter to attempt to explain the reasons for the dif-
ferences, or to describe in detail the many different techniques used overseas.

17.1.2 System design considerations

It is essential that there should be close liaison between the system design engineer
and the protection engineer at a very early stage in the design of a power system.
With high fault levels, for example 35 000 MVA at 400kV, one of the main limita-
tions in system design is the speed at which faults, in particular three-phase faults,
can be cleared from the system. This fault-clearance time governs the stability limit
of the power system which in turn governs its design. For example, if there are teed
feeder circuits, fault-clearance times may be appreciably increased, and the saving in
the cost of circuit breakers as a result of tee-connecting circuits may have to be
322 The application of protection to transmission systems

Station Station
A B

Station C

Fig. 17.1.2A Transformer tee,connected via a c i r c u i t breaker to a feeder

offset by the system design engineer against the cost of reinforcing the power
system to ensure that it can withstand the longer fault-clearance times.
Normally, the system design engineer assumes that the operation of all pro-
tection on the system will be fully discriminative; but it is essential that he checks
with the protection engineer that discrimination is feasible with the particular
circuit configuration he proposes. The importance of any protection difficulties can
then be assessed together with the cost of overcoming these difficulties. A simple
example of this is a 120 MVA transformer tee-connected via a circuit breaker
midway along a 257 kV line (Fig. 17.1.2A). As will be apparent later in this
chapter, it can be very difficult and costly to provide a fully discriminative high
speed fault clearance for such a circuit. The particular faults which are difficult to
clear discriminatively are those occurring between the transformer h.v. circuit
breaker and the transformer h.v. bushings, as these faults may be within the first
zone reach of any accelerated distance protection installed at stations A and B.

17.1.3 Factors which influence the choice of protection

The following are some of the factors involved in choosing a suitable protection
scheme for a given application.

17.1.3.1 Plant to be protected: The first factor to be considered in choosing a

suitable protection scheme is the nature of the plant to be protected, for example
generator, transformer, feeder etc.
The detailed characteristics of the plant to be protected must then be examined
to determine if any special considerations apply in choosing a suitable protection
scheme. As an example, a feeder may consist of a length of overhead line con-
nected in series with an underground cable. Not all protection schemes suitable for
protecting overhead lines are necessarily suitable for a combination of overhead
line and cable, for example the relatively low impedance of the cable may be such
 The application of protection to transmission systems 323

that the effective length of the feeder is too short for the satisfactory employment
of distance protection.

1Z1.3.2 Probability of various types of fault: The various types of fault which
may occur on the plant to be protected should next be considered, together with
the probability of such faults occurring.
For example, on a transformer circuit the majority of faults within the trans-
former tank are earth faults, the risk of a phase-to-phase fault clear of earth within
the tank being low. Breakdown of the transformer core insulation, and interturn
faults also occasionally occur.
Earth-fault protection is therefore essential, and Buchholz protection is required
to detect core faults and interturn faults which are clear of earth.
Fast and fully discriminative phase-fault protection, for example biased
harmonic restrained overall differential protection, can often only be justified to
provide phase-fault protection for the connections to a transformer. These con-
nections, unless they are phase segregated, for example consisting of single-core
cables rather than open copperwork, may present an appreciable phase fault risk.
Another example is on a generator not only can earth faults and phase faults
occur, but the generator can be severely damaged by negative phase sequence
currents. These currents arise from either uncleared unbalanced faults on the
system or from single or two-phase open circuits. The latter may result from a
circuit breaker failing to make or break on all its three phases, or from broken
jumpers on a line. These risks, and their resultant effects on the generator, have to
be assessed and it is usually found to be necessary to provide negative phase
sequence protection for the generator.

1 Z1.3.3 Load and fault currents: The magnitude of the minimum and maximum
fault currents at appropriate points on the system must be established in order to
determine the sensitivity and stability requirements for the protection. The maxi-
mum expected possible continuous load current must also be known as this will
determine the required continuous thermal rating of the protection. It may also
affect the minimum permissible phase-fault setting, and the settings of any over-
current, rate of change of current, or distance starting equipment.
The charging currents of overhead lines and cables must also be taken into
account, as they may influence the minimum permissible protection settings.

17.1.3.4 Voltage and current ratings of protected plant: Not only should the
normal continuous voltage and current ratings of the protected plant be taken into
account in choosing the protection required, but any special requirements such as
short time overload ratings, must be considered.
As an example, it may be permissible to overload a power transformer for a few
minutes with safety, say to 50% overload, thus necessitating a minimum overcur-
rent protection setting greater than 150% of the rated current of the transformer.
Another example of a short-time overload rating required for a few minutes
324 The application of protection to transmission systems

occurs in the case of certain overhead-line circuits where it is desirable to utilise the
short-time thermal capability of the overhead line to enable quick starting gas-
turbine or pumped-storage plant to relieve the overload condition.
For some unit systems of protection, particularly those utilising Post Office
pilots, it is highly undesirable that continuous comparison of the currents at each
end of the protected circuit be permitted under expected load conditions,
otherwise a momentary interruption of the signalling channel may result in inad-
vertent tripping and starting (fault.detector) relays are therefore included to control
transmission. For power line carrier systems the continuous transmission of carrier
signals may also cause unacceptable interference to be radiated in frequency bands
reserved for normally quiescent carrier. Furthermore, some power line carrier trans-
mitters may have no continuous thermal rating but only a short-time rating of a few
minutes.
In addition to the short-time overload rating, the protection system i.e. both the
relays and associated current transformers, must be designed to withstand, both
thermally and dynamically, the maximum fault current which can flow in the
primary plant it is protecting. This short time fault current rating is normally for a
3 or 1s rated duration. The choice of time depends on the expected maximum fault
clearance times. Where only l s rated equipment is available, it may be necessary
to investigate the thermal time constant of the protection bearing in mind that it is
very unlikely that the magnitude of the fault current can in practice remain at its
maximum value for the total duration of the fault.

1 Z1.3.5 Necessity or otherwise for high-speed operation: The maximum permis.

sible duration of fault current is governed by two factors:

(a) limitation of damage to plant

(b) effect of long fault-clearance times on the system which is feeding the fault
current.

The effects of long fault-clearance times must be assessed, not only with regard to
the damage at the point of fault, for example an oil fire may be unnecessarily
started in a faulted transformer, but the plant feeding the fault may also be
damaged by the fault current, for example a cable through which the fault current
passes may be damaged.
Fast fault-clearance times especially for three-phase faults may be essential to
maintain system stability. On the British Supergrid system stability studies indicate
that at some 400 kV stations system stability will be endangered if a close-up three-
phase fault is allowed to persist for more than 80ms. Fast clearance times are also
required on plant feeding large induction motor loads (e.g. at steelworks or oil
refineries). The reason for this is that a prolonged dip in voltage may result in the
induction motors slowing down to such an extent that, when the fault is cleared
and normal voltage is restored, the motors draw such a high current attempting to
accelerate that it produces a correspondingly large voltage drop between the supply
 The application of protection to transmission systems 325

and the motor terminals. This reduction in voltage may result in the motors being
unable to produce enough torque to regain their normal operating speed.
On the other hand, unnecessarily fast protection should not be specified as, with
certain exceptions, the faster schemes are more expensive, more complex and, in
many cases, more liable to maloperate. For example, distance protection which has
an operating time of under 10ms usually requires more faithful voltage inputs than
are normally obtainable from most designs of capacitor voltage transformers.
In quoting protection operating times for assessing their effect on system per-
formance it has to be borne in mind that the times quoted for a given protection
system can usually only be approximate, as so many factors influence the actual
operating time obtained on an individual fault.
Some of the principal factors involved are:

(a) the magnitude of fault current (usually expressed as a multiple of the setting
of the protection)
(b) the type of fault, e.g. phase-to-earth, phase-to-phase or 3-phase
(c) the ratio of the impedance of the source of the fault current to the impedance
between the relaying point and the point of fault
(a) the magnitude of the d.c. component present in the fault current, and the
prefault magnetisation state of the current transformers
(e) the magnitudes and phase angles of harmonic components and nonsinusoidal
waveforms
(,0 the magnitude and direction of the load current flowing immediately prior to
the fault
the resistance of the fault itself
(h) the characteristics of any signalling channels used between the local and
remote circuit breakers. The characteristics of the channel may be appre-
ciably modified by transient electrical interference in the channel at the time
of the fault, including interference from adjacent signalling channels.
With an increasing amount of generation being connected into the transmission
system at higher voltages, i.e. the generator/transformers being switched at 400 and
275kV instead of 132 or 33kV, the ratio of reactance to resistance of the source
impedance for faults on the transmission system is steadily increasing. This results
in a higher probability of a large d.c. component in the fault current when the
fault occurs at other than at the maximum point-on-wave of the voltage. Hence
item (d) in the list is becoming of increasing importance because, if the main or
interposing current transformers saturate, very long operating times may occur.

17.1.3.6 Importance of security of supply: To obtain fully discriminative pro-

tection for all types of fault may not always be economically justified, depending
on the importance of the load being supplied. For example, in the arrangement
shown in Fig. 17.1.3A, the l.v. busbars are supplied solely from the two trans-
formers and the only form of phase fault protection is inverse definite minimum
time (i.d.m.t.) overcurrent protection. If a phase-to-phase fault clear of earth occurs
326 The application of protection to transmission systems

- , , , ,,, - H.V. busbar

I!
f" i ~0(,!

T l

i e to phase fault

L.V. busbar

Load

Fig. 17.1.35o, Nondiscriminative tripping for transformer I.v. fault

i 1 " H.V. busbar

TI T2

e to phase fault

! ~o~!
1 T
- i ' .... , , _ , L.V. busbar

!
Load

Fig. 17.1.3B Discriminative tripping for transformer I.v. fault

 The application of protection to transmission systems 327

on the 1.v. connections of transformer 2, the fault current is supplied equally by

both transformers and hence, neglecting load currents, the l.v. overcurrent relays on
transformer 1 and transformer 2 will operate in the same nominal time, tripping the
l.v. circuit breakers of both transformer 1 and transformer 2. Any load current will
tend to cause transformer 1 relay to operate before transformer 2. In addition, the
h.v. overcurrent protection will operate on transformer 2 thus clearing the fault.
Any load connected to the l.v. busbar would thus lose its supply.
If, on the other hand, in addition to or in place of the l.v. inverse definite
minimum time overcurrent protection, directional l.v. overcurrent protection or
transformer overall differential protection were provided, either of these forms
of protection would operate before the i.d.rn.t, overcurrent protection and only the
faulted transformer would be tripped, thereby maintaining the supply to the load
(Fig. 17.1.3B).

17.1.3.7 Compatibility with existing protection: When choosing a protection

scheme and its settings due regard must be paid to the characteristics of any existing
protection on the power system. If the setting of a distance protection on the trans-
mission system has to be such that it can 'see through' the step-down transformers
and thus respond to faults on the distribution system, its time setting must be high
enough to discriminate with the protection on the distribution system.

17.1.3.8 Availability of signalling channels: In transmission networks where

substations are separated by only a few miles, as in London, the most satisfactory
and economic protection for two~nded feeders is often to employ private pilot
wire protection utilising pairs of wires in a network of private pilot cables. It is
important to arrange that sufficient spare pairs are provided in the pilot cables
when they are originally laid to ensure that an adequate number of pairs is available
to meet all expected future needs.
Where private pilots are not available, it is usually economic in the UK to hire
pilots from the telecommunication companies for the protection of two-ended
feeders. However, since 1978, the telecommunication companies have in general
withdrawn the facility for the hire of metallic circuits having maximum declared
resistance and capacitance values. Hence many of the differential and phase-
comparison protection systems designed to compare fundamental frequency
quantities over telecommunication company metallic pilots are not suitable for
most applications, and protection designed for use with voice frequency signalling
systems is used in these cases. Power-line-carrier channels can often be
economically justified, but the frequency bands available to the Supply Industry
for this form of signalling, are rapidly used up and difficulty may often be
experienced in allocating suitable frequencies without the risk of causing inter-
ference in adjacent channels, although better utilisation of the available frequency
bands is achieved by installing carrier equipments which require narrower signalling
bandwidths than hitherto. In allocating frequencies, not only must allowance be
made for an adequate separation between signalling frequencies on lines at the same
328 The application of protection to transmission systems

voltage, but cognisance must be taken of any mutual coupling of carrier signals
between lines of different voltages, for example where carrier signalling is used on
275kV and 400kV lines which follow the same route.
Microwave signalling channels may be justified economically and technically,
but the requirement for microwave links that repeater stations must be within line
of sight of each other severely restricts their application. Even greater restrictions
in the United Kingdom are the requirements to obtain a license to operate a micro-
wave link and to obtain planning permission for the erection of microwave aerials
and towers. The use of fibre-optic signalling channels, which incidentally are
immune to electrical and magnetic interference problems, are likely to be used in
future as protection signalling channels, particularly if the fibre-optic cable can be
satisfactorily incorporated as part of the earthwire of an overhead-line circuit.

1 Z1.3.9 Cost: Where several protective schemes meet the specified requirements,
the cheapest scheme which meets the specification would normally be chosen. In
costing the schemes, care has to be taken to include all the relevant costs, for
example capitalised costs of pilots, maintenance cost, etc. (See Section 17.5).
The cost of a particular scheme may be influenced by the need for other
facilities such as control, metering or telecommunication requirements. The instal-
lation of private pilot cables for example, may not be economic for protection
purposes only, but may be justified if also required for intertripping, control,
metering and telephones.

17.2 Main and back-up protection and location of current transformers

17.2.1 Main and back-up protection
Main protection is defined as that protection which is normally expected to initiate,
quickly and discriminatively, the tripping of appropriate circuit breakers to clear a
fault from the system. Back-up protection is the protection which is intended to
operate in the event of a failure or inability of any main protection to clear the
fault.
The number and types of main and back-up protection systems to be installed on
a given feeder or item of plant depends on technical and economic considerations.
As an example, on the UK electricity system it has been found, based on cost-
benefit analyses, that the provision of some measure of redundancy for the main
protection on the 400 and 275kV systems can be economically justified; but not,
except in exceptional circumstances, for the 132kV system. These analyses take
account of the estimated cost to the nation of a loss of supply, i.e. not only the loss
to the supply authority of its sales revenue but also the cost to consumers. For
protecting the 400 and 275kV systems the general principle adopted is that no
credible defect in a single component or secondary circuit of the protection or
tripping schemes should result in the nonclearance of a primary fault from the
system.
Back-up protection for the 400,275 and 132kV systems is usually of the non-
unit type, e.g. definite minimum inverse time overcurrent or earth-fault protection,
 The application of protection to transmission systems 329

or sometimes in specific applications distance protection. In order to reduce ap-

preciably the risk of the back-up protection being called on to operate a con-
siderable measure of redundancy is provided in the main protection arrangements.
Also most 400kV and 275kV circuit breakers are equipped with circuit-breaker fail
protection.

17.2.1.1 Main protection: In order to ensure fast, discriminative and reliable main
protection where more than one main protection system is provided they should
be as dissimilar as possible.
For example, in the case of feeder protection:

(a) The first and second main feeder protection, where practicable, should utilise
entirely different principles, preferably a differential system and a distance
system.
(b) The first and second main protection should operate into independent
tripping systems. (Fig. 17.2.1A).
(c) If two distance protections are used, where possible one should be of the
blocked type and the other should either accelerate or intertrip.
(d) The first and second main protection should preferably employ different
types of signalling channel, e.g. power-line-carrier and Post Office pilot, but
where this is not possible and pilot circuits must be used, the two channels
should be separately routed.
(e) Two different manufacturers should provide the first and second main
protection.

m
I I
I
lil I
ch I
I

I
I
No. I d.c. I
supply
_I ::L:ao'n
---I
I
I
I I
l No. 2 d.c. I
I Second main supply
I protection -- ,,,
I
I I
! I I
I I I
I No.]
No. i d.c.
supply
alz- I
I
,,
'@
I Trip I
I coils I
No. 2 d.c. I No. 2
ry'y~
supply e
Fig. 17.2.1A Provision o f two discrete tripping systems
 330 The application of protection to transmission systems

For main plant and its associated connections full duplication of the protection
may not be technically or economically justifiable in which case, e.g. for trans-
formers or generator~,, th=~ protection of each item of plant can be grouped into two
groups of duplicate or supplemental relays, each group being fed from separate d.c.
supplies into two segregated tripping systems.

17.2.1.2 Back-upprotection: Inverse definite minimum time overcurrent protection

is usually provided as back-up protection for feeders, transformers and generators,
but it suffers from the disadvantage of having relatively long operating times, which
may be unacceptable from system stability or large induction motor load con-
siderations. I.D.M.T. relays of electromechanical design are cheap, simple and
reliable but relays of static design may have more flexible characteristics, e.g. a
choice of various inverse time characteristics.
Fault clearance may also be nondiscriminative with back-up overcurrent pro-
tection. For example, Fig. 17.2.1B represents a small part of an interconnected
network. In such a network, as the fault current may flow in any direction de-
pending on the fault location, it is impossible to grade satisfactorily the back-up
overcurrent protection settings on feeders, and it is usual, in the absence of specific
operational requirements, to give the relays the same nominal settings.

x, x /
J
x /
1 f
I I
I I
I i
I Station Station Station II
I A B C I
| ,= _ _

ILA~ ILA~ ~---.~=~---Load

FA lt.'A
11."A 1 FC

ILA = Load current from station A

IFA = Fault current from station A

IFC = Fault current from station C

Fig. 17.2.1B Nondiscriminative clearance of fault by Ld.m.t. overcurrent protection

 The application of protection to transmission systems 331

Assume in the example chosen fault current can only be fed from stations A and
C and an earth fault occurs on feeder BC close to end C. If the main protection on
this feeder fails to operate, the back-up overcurrent relays at stations A, B, and C
measure the following currents on the faulted phase:

At station A, feeder AB overcurrent relay measures fault current (IRA) + load

current fed from end A(ILA ).
At station B, feeder BA overcurrent relay measures fault current (IFA) + load
current fed from end A(ILA ).
At station B, feeder BC overcurrent relay measures fault current fed from
end A(IFA).
At station C, feeder CB overcurrent relay measures fault current fed from
end C(IFc).

It is apparent that the fault will be cleared at station C by the relay on feeder CB.
It will also be cleared at station A or B, or both, by the relays on feeder AB. Both
of these relays measure the same fault and load current and hence will have the
same nominal operating times. There will therefore be a total loss of supply at
station B.
The current setting on a back-up overcurrent relay is governed by

(a) the minimum fault current for which the relay is required to operate
(b) the maximum load current the circuit concerned is required to carry under
emergency conditions.

It may happen that the setting required by condition (b) is greater than the setting
required by condition (a), in which case the risks involved may have to be accepted.
In the United Kingdom the usual policy, in order to avoid the risk of cascade
tripping in the transmission system, is that the overcurrent relays are set above the
maximum load currents which are to be expected under emergency conditions.
One method of alleviating this limitation of overcurrent protection, at least as
far as earth faults are concerned, is to connect residually one element of a three
element overcurrent relay (Fig. 17.2.1C). This residually connected element can
then be given a lower setting than the phase connected elements as it does not
respond to balanced load currents. The choice of a suitable setting for the residually
connected element is, however, difficult as the setting has to be graded not only
with any other residually connected overcurrent relay elements on the system, but
also with any distance protection on adjacent circuits. Furthermore, the minimum
fault current to be measured by the relay may vary between wide limits over a
period of 24h owing to changes in the system zero-sequence impedances. A
common cause of these variations is the switching of generator transformers in and
out of service at generating stations.
Distance protection with its 2nd zone and 3rd zone reaches inherently provides
a back-up protection, but the effective reaches of the protection are much reduced
332 The application of protection to transmission systems

B, ~ryy~

()vercurrent
relay
N
i • | i ii i

Normal 3 phase c o n n e c t i o n of 3 element overcurrent relay

Y , ,, 1

Overcurrent
relay

Residual c o n n e c t i o n of one element o f 3 element overcurrent relay

Fig. 17,2.1C Overcurrent element connections

by any fault current infeeds between the measuring point and the point of fault.
Fig. 17.2.1D illustrates this effect. A fault in this typical example is assumed to
have occurred near to station F on a feeder between stations E and F and the feeder
circuit breaker at E has failed to trip. The apparent distance measured by the
distance protection on the various feeders is indicated by the figures enclosed by
boxes. It will be noted that on feeder D-E at station D the fault is apparently 376krn
instead of 91 + 12 = 103 krn away, and at station A the fault is apparently 1195kin
 The application of protection to transmission systems 333

I ,95t-
i i 1

-18 51
1 58 "1376]
1
55

]
91
Fault

D
i |°'

Fig. 17.2.1D Apparent fault distance as measured by remote distance protection

away instead of 149km. In practice, more adverse situations can easily occur and in
general distance protection is unlikely to provide any appreciable measure of back-
up protection for the system when it has to measure beyond the busbars of a major
power station or any other substation with a major infeed. Occasionally, distance
protection has been installed as system back-up protection in its own fight, in place
of inverse-definite minimum time overcurrent protection but its predicted per-
formance has not been satisfactory owing mainly to the reach shortening effect.
Another form of back-up protection applicable to resistance earthed systems is
standby earth-fault protection. This protection system consists of a current trans-
former, the primary of which is connected in series with the neutral earthing
resistor, and a relay which has a long operating time, usually up to 30s. The standby
earth-fault protection operates in the event of any uncleared earth faults on the
system and, where operationally advantageous, its tripping can be made in two
stages. For example, the first stage trips the 1.v. circuit breaker of the transformer,
and if the fault current persists, the second stage, after a further time, trips the h.v.
circuit breaker.
334 The application of protection to transmission systems

The maximum operating time of the standby earth-fault protection should be set
below the rated time of the neutral earthing resistor.
It should be noted that the standby earth-fault protection should trip all the
fault current infeeds to the affected part of the system, otherwise part of the
system may be left still energised but without its neutral point earthed. The latter
condition is liable to lead to 'arcing ground' conditions which may cause very severe
damage to plant.

17.2.2 Effect of location of current transformers in determining protection to be

provided

It is essential that there should be no plant or connections in the transmission

system which are left entirely unprotected, and it is usual to ensure that the various
zones of protection into which the system is divided overlap each other. To obtain
the required overlapping without thereby introducing some lack of discrimination
is impossible. In determining a protection scheme one of the most difficult tasks
is to reduce this lack of discrimination to a minimum and to assess the extent to
which a scheme should be complicated and money spent to cater for faults occur-
ring on perhaps a few inches of connections between an open disconnector, or
circuit breaker, and a current transformer housing. In this respect the location of
the current transformers in the transmission system is of paramount importance,
especially with regard to circuit breakers.
Considering first circuit breakers with current transformer accommodation in
the bushings each side of the circuit breaker, e.g. dead tank bulk oil or SF6 circuit
breakers, it is easy to arrange that the zones of protection overlap within the
circuit-breaker tank (Figs. 17 2..2A and 17.2.2B).
The risk of a fault occurring within the tank is very remote, and if such a fault
should occur, from safety considerations it would normally be desirable to make
both zones A and B dead. There is, therefore, little difficulty or disadvantage with
this design of circuit breaker in arranging suitable overlapping of the protection
zones.
With some forms of circuit breaker, e.g. live tank air-blast and SF6 circuit
breakers, for economic and technical reasons current transformers are often located
on one side only of the circuit breaker (Figs. 17.2.2C and 17.2.2D). Arrangements
to obtain discriminative tripping may become complex, and full consideration must
be given to the risk of faults occurring between the circuit breaker contacts and the
O v e r l a p zone
A~
f
Zone A , r"t Z o n e B_
ii
1,,,8 ! !

Fig. 17.2.2A Zones overlapping circuit breaker

 The application of protection to transmission systems 335

//'
t e 0

C.T. a c c o m m o d a t i o n C.T. a c c o m m o d a t i o n

Fig. 17.2.2B Dead tank circuit breaker

Zone b e t w e e n c.t.
housing and circuit
breaker contacts
--'t
Zone A Zone B
~, ,

I I v

Fig. 17.2.2C Zones not overlapping circuit breaker

c.t. housing. The effects of such faults both on the plant itself and on the sur-
rounding system must also be assessed. For example, in Fig. 17.2.2D the risk of a
fault occurring at the insulator supporting the blast heads on outdoor switchgear
is not negligible and measures must be taken for fast, and, as far as possible, dis-
criminative clearance of such faults. It should be noted that such a fault (Fig.
17.2.2E) would still be fed from station B when the circuit breaker at station A
and its associated busbar selectors were open, provided that the circuit discon-
nector were closed.
Usually at busbar stations, faults between the c.t. housing and the circuit-breaker
contacts will be detected by the busbar protection, and cleared by the tripping of
all the circuit breakers connected to the appropriate busbar and by the tripping of
the circuit breaker at the remote end of the circuit concerned.
Various arrangements can be made for tripping the remote circuit breaker, the
choice depending on the type of protection on the circuit and on the degree of
discrimination required. The arrangement adopted also depends on the extent to
336 The application of protection to transmission systems

which increases in the complexity and cost of the protective scheme are acceptable.
Typical methods by which the remote circuit breaker may be tripped are"

(a) by unstabilising unit forms of protection on the outgoing circuits, for example
by contacts on the busbar protection trip relays. If a delay of approximately
O.15s is introduced before unstabilising the unit protection, thus enabling
the circuit breaker at station A to open first, discriminative intertripping can
usually be achieved.
(b) by operation of the main or back-up protection at station B if this protection
is of the non-unit type.
(c) by means of an intertripping signal which is transmitted by a pilot wire or
carrier current channel from station A to station B. This signal can be initiated
by either the busbar protection trip relay or a fault-current detecting relay
such as the current check relay in circuit-breaker fail protection.

The use of a current check relay has the advantage that the fault clearance is
discriminative, but it has the disadvantage that the clearance is deliberately delayed
and this delay may be unacceptable for fault-damage or stability reasons.

.•, ~. .nc,. .?~'. ,nq,,;?

,!iI !!,, !!

Fig. 17.2.2D Live rank air-blast circuit breaker

Station A
Reserve
bar
| Main
i bar / c.t. Station B

,-- ,-- . . . . . . . - ~

Circuit

Fig. 17.2.2E Fault between c.t. housing and circuit breaker

 The application of protection to transmission systems 337

It would appear that there is a strong technical case for mounting current
transformers on both sides of the circuit breaker. It must be borne in mind,
however, that for circuit breakers of other than the dead tank type, mounting
current transformers on both sides may entail having an undesirably large overlap
zone. Furthermore, the insulation of busbars should be kept as reliable as possible
and the introduction of current transformer housings may slightly increase the risk
of busbar faults. The provision of room for c.t. housings on both sides of the circuit
breaker may also increase the size of the switchgear bay and hence the size of the
substation.

17.2.3 Two-stage overcurrent protection

If the main fault infeed is only from the h.v. side of the transformer, it is some-
times advantageous, both technically and economically, to replace the transformer
l.v. overcurrent protection current transformers and the associated inverse definite
minimum time relay by a two-stage relay comprising an inverse definite minimum
time element, energised from c.t.s on the h.v. side of the transformer, and a d.c.
timelag element (Fig. 17.2.3A). The first stage, the i.dJn.t, element, is arranged to
trip the 1.v. circuit breaker only, and the second stage, the d.c. timelag element, the
h.v. circuit breaker. This two-stage relay enables faster discriminative overcurrent
protection to be obtained for the transformer as the discriminating time can be
reduced from the usual minimum time of 0-5s for grading two i.d.m.t, relays to a
fixed time of, say, 0.2s, that is, the maximum circuit breaker tripping time plus a
safety margin.

H.V. b usbar

2nd stage trip

2 stage
overcurren t
relay

i
I
I
/ I
I
I
. I
1st stage trip

L.V. b usbar

Fig. 17.2.3A Two,stage overcurrent protection

338 The application of protection to transmission systems

IDMT Instantaneous
o/c relay o/c relay
(1 per phase) (I per phase)
rvrrx..___.
I I
I I
I I L.V.e.b.
! I trip relay

I
I I L._ _.1
I I
I I Timer
o ~0 °_L.°'-" I
I
I H.V.c.b.
I trip relay
OI'c
L__J
Fig. 17.2.3B Two-stage overcurrent protection d.c. circuits

Typical d.c. tripping circuits for a two-stage overcurrent protection are shown
on Fig. 17 2.3B. Most i.d.m.t, relays, having operated, take an appreciable time
(over 100 ms) for their contacts to disengage. Hence, if the tripping of the 1.v.
circuit breaker causes "the cessation of fault current, there is a danger that, unless
the timer is immediately de-energised, the timing relay will complete its opera-
tion, energise the h.v. circuit breaker trip relay, and the unnecessary tripping of
the h.v. circuit breaker would result. For this reason, unless the i.d.m.t, relay
contacts have a disengaging time of less than 100 ms, an instantaneous current relay
is provided with contacts having a very fast disengaging time. The contacts are
connected to open the coil circuit of the timing relay immediately the fault current
ceases.

17.3 Intertripping and protection signalling

17.3.1 General

The various types of intertripping and protection signalling equipments and associa-
ted communication channels are described in detail in Chapter 7.
The choice of the method to be used for initiating the tripping of the remote
circuit breakers usually lies between the following:

(a) D.C. signalling, normally only used for intertripping, in which an uncoded
d.c. signal is transmitted over privately owned pilots. The receive relay may or
may not be surge-proofed, depending on the risk of appreciable induced a.c.
in the pilot wires.
(b) Voice-frequency signalling, in which a coded voice-frequency signal is
transmitted, normally over a hired telecommunication company channel, but
sometimes over private pilots or other types of channel.
 The application of protection to transmission systems 339

(c) Carrier signalling, in which a carrier signal, normally coded, but may be
uncoded for certain simple signalling applications, is transmitted over the con-
ductors of the primary system.
(a) Fault throwing, in which a fault is deliberately initiated on the primary
system by closing a fault throwing switch. The fault is detected by the pro-
tection at the remote station which in turn initiates tripping of the appropriate
circuit breakers.

For a given application, in addition to technical considerations, the main factors

which govern the choice of method of tripping the remote circuit breakers are the
availability and cost of providing the signalling equipment and channels.
Where more than one intertripping or signalling channels are required for
security of operation reasons they should be as diverse as possible with regard to
both the design of the equipments and the types of communication channel used,
as mentioned in Section 17.2.1.1.

Pilot cable N o . ! .,,'1

Ii . . . .

Transformer feeder 1

Transformer feeder 2 ~ :

Pilot cable No. 2

I, f "M

Fig. 17.3.1A Routing of protection and intertripping channels for double,circuit transformer
feeder

Problems may arise, however, as a direct result of providing this diversity. For
example, if it is assumed that the only communication channels which can be
economically justified for the protection and intertripping for the double circuit
transformer feeder shown in Fig. 17.3.1A are two separately routed private pilots,
and it is also assumed that each feeder is equipped with two main protection
systems and that the intertripping for each feeder is duplicated, then in order to
increase the dependability, i.e. to reduce the risk of a failure to trip, it is evident
that the first main protection for TF1 should be associated with pilot cable No. 1
and the second main protection with pilot cable No. 2. Similarly, the first inter-
tripping channel for TF1 should be in cable No. 1 and the second intertripping
channel for TF1 in cable No. 2.
340 The application of protection to transmission systems

However, some forms of main protection will maloperate on load or through

fault current in the event of the conductors in the pilot cable becoming short
circuited. Similarly, with a simple d.c. intertripping system a live conductor in the
pilot may, through insulation breakdown, be inadvertently connected to another
conductor resulting in maloperation of the intertripping equipment.
One of the main causes of damage to pilot cables, particularly in urban areas,
is building or road construction work when, for example, sheet steel piling may
accidentally result in severing with consequent short-circuiting of conductors in
the pilot cable. In this case, in the event of the communication channels for TF1
and TF2 being routed in the manner described, both TF1 and TF2 might be inad-
vertently tripped. On the other hand if all the protection and intertripping associa-
ted with TFI were routed in pilot cable No. 1 and that for TF2 were routed in
pilot cable No. 2 the risk of inadvertent tripping would be confined to a single
primary circuit.
Where the protection or intertripping for more than one primary circuit is
routed in one pilot cable the risk of more than one circuit being tripped in the
event of damage to that pilot can be reduced in various ways. For example, the use
of starting relays for the protection will reduce the risk of protection maloperation
and the risk of inadvertent maloperation of the intertripping can be eliminated by
using a voice frequency intertripping system instead of a d.c. system.
In practice, an engineering assessment has to be made of the consequences
of the loss of supply owing to damage to a pilot cable or equivalent signalling
channel and, within the economic constraints, the appropriate measures taken to
provide reliable high-speed fault clearance.

17.3.2 D.C. signalling

D.C. signalling, which utilises privately owned pilots between the stations con-
corned, has the great advantages of simplicity and reliability, but if surge-proof
relays are used, the operating time may be unacceptably long (about 150 ms). The
operating time may increase appreciably with some types of relay if induced
voltages are present in the pilot wires. Where faster operating speeds are required a
voice.frequency signalling system has to be used.
Where the pilot circuits are very long, it is advisable to use a separate insulated
battery for energising the pilots for the following reasons:

(a) With some designs of two-way two-core intertripping schemes, if the intertrip
send relays are operated at both ends simultaneously, the station tripping
batteries at the local and remote stations may be connected in parallel via the
intertripping pilots. This can lead to difficulties if the normal voltages of the
two batteries is appreciably different.
(b) It is undesirable to increase appreciably the capacitance to earth connected
to the station tripping battery; excessive capacitance of the d.c. wiring can
cause relays to maloperate in the event of a single earth fault.
342 The application of protection to transmission systems

(c) A fault in the pilot circuit, which may not be detectable until a signal is sent,
might affect the security of the battery supplies at the station from which the
intertripping signal is being sent; for example, tripping of circuit breakers at
that station may be delayed until the appropriate fuses have cleared the fault
in the pilots and the tripping supply voltage has been restored.
(d) Any rise in earth potential at the remote station due to the fault current may
be transferred through the pilot circuit and imposed on the local station
tripping battery, with resultant insulation failure (Fig. 17.3.2A). It is usual
to provide 15 kV insulation to earth for batteries used to energise long pilot
circuits.
17.3.3 Post Office signalling

When private pilots are not available voice frequency equipments are often used in
conjunction with hired telecommunication company channels for protection
signalling and for inter-tripping. Typical operating times for equipments at present
in common use are between 25 and 40 ms.
The main disadvantages of using telecommunication company signalling
channels are:

(a) the equipment is complex and costly.

(b) the hired pilots may be subjected to unauthorised human interference. They
may also be out of service when a signal is required to be transmitted. In
practice in the UK these risks have been low. However, the new designs of
amplifiers and digital communication equipment incorporating solid-state
devices may have to be protected by overvoltage protection with settings of
a few hundred volts on their external terminals. There is therefore an
increasing risk that the communication channel may be momentarily
interrupted at the instant of fault by the effects of induction or rise of earth
potential caused by primary system faults in the vicinity of the
telecommunication company equipment. Also, although most equipment
used on channels hired by the supply companies is independent of 240 V
mains operated supplies and hence not affected by voltage dips in that supply,
in a few instances some is not and the changeover arrangements to a standby
supply may not always be fast enough to ensure reliable protection and
intertripping.
(c) the signalling code may be simulated by disturbances in the hired circuits thus
causing incorrect operation. Again this is a low risk, but the risk may increase
appreciably if the telecommunication company transmission circuit includes
carder or microwave links, which may introduce unwanted frequency or
phase changes in the received signal.
(d) relatively frequent routine tests are required to prove the equipment is in
proper working order.
(e) the propagation time of the signal over a long telecommunication company
link may be significant.
 The application of protection to transmission systems 343

17.3.4 Carrier signalling

Carrier signalling is widely used where the cost of the power line coupling equip-
ment can be justified. In the UK, its use is practically universal as one of the com-
munication channels for the protection or intertripping of long feeders, unless
technical restraints, e.g. a length of underground cable prevent its application. It is
particularly useful as an alternative to telecommunication company signalling
channels where these channels may be suspect, e.g. alternative routing for two
independent channels may not be available, or where there has been a history of
insulation problems due to flooding.
The main technical limitation in the application of carrier signalling is that in
general it is unsuitable for circuits containing cable, owing to the loss of signal
resulting from the mismatch in the characteristic impedances of the carrier
frequency path at the cable sealing ends. Attenuation of the signal also occurs
within a cable, the magnitude of the attentuation varying appreciably with cables of
differing construction. These limitations can sometimes be overcome by providing
line traps, or coupling equipment, or both, in the primary circuit at the appropriate
points but it is rarely economic.
Another technical limitation in the application of carrier signalling is the at-
tenuation of the carrier signal at any tee-points in the primary circuit. The resulting
loss at the tee-point can normally be reduced to an acceptable value, typically
3.5 dB, by carefully matching the impedances of the coupling equipments at each
end of the tee-connected circuit.
Carrier signalling has the advantage that the signalling channel, that is the power
line itself, is not subject to unauthorised interference by human agencies in the
same way as hired pilot circuits. On the other hand, carrier signals are liable to be
severely attenuated under certain line icing and hoar frost conditions, and it is
advisable on certain vital circuits to back up carrier signalling by telecommuni-
cation company signalling. Also, if the signal has to be transmitted along a power
line which is faulted, satisfactory reception of the carrier signal cannot always be
guaranteed. Carrier intertripping would therefore be unsatisfactory if total reliance
had to be placed on it for the correct clearance of line faults.
Great care has to be taken in the design of a carrier signalling equipment to
ensure that it will not maloperate when subjected to the severe interference by the
opening and closing of disconnectors. This interference is particularly severe when
disconnectors energise or de-energise conductors having a very small capacitance,
for example short sections of busbar.

17.3.5 Fault throwing

Fault throwing, where the fault level permits, is a simple and reliable method of
tripping remote circuit breakers. Fault throwing switches are at present available in
344 The application of protection to transmission systems

the UK for fault levels up to 3500 MVA at 132 kV.

The main disadvantages are:

(,1) when a fault throwing switch operates the system is subject to a disturbance.
(b) the operating time of a fault throwing switch is long (approximately 0-3s) and
may be influenced by weather conditions such as ice in the mechanism.
(c) there is a risk of inadvertent operation.
(a) when fault throwing switches are installed on transformer feeders directional
overcurrent protection may have to be used to trip the transformer 1.v. circuit
breaker for feeder faults. The restrictive limitations in the application of
directional overcurrent protection are given in Section 17.7.5.

17.4 Automatic switching

17.4.1 Design and application of automatic switching equipment

Insofar as this Section is concerned, the term automatic switching equipment refers
to the equipment which initiates the post-fault automatic operation of circuit
breakers and disconnectors in a prearranged sequence. The equipment initiates:

(a) Permanent isolation from the system of any faulted plant which must not be
re-energised, e.g. a faulted transformer.
(b) Re-energisation of overhead lines and similar plant which may be only
transiently faulted and which may be safely re-energised. If the fault on re-
energisation is found to be permanent, the overhead line or similar plant may
be automatically isolated to permit any banked or tee-connected healthy
plant to be put back into service automatically.
(c) Automatic reclosure of all circuit breakers which were closed prior to the
fault and which may be safely reclosed.
(a) The switching into service of standby plant, e.g. a transformer, which was not
in service prior to the fault, because, for example, the associated switchgear
fault current ratings would have been exceeded.

This Section deals only with the United Kingdom practice of three-pole tripping
and reclosure of circuit breakers. Single-pole tripping and reclosure has not been
installed in the United Kingdom because, in general, the transmission network is
so strong and interconnected that little if any operational benefit would be gained
by single-pole tripping and reclosure; hence the cost and complication of providing
single-pole tripping and reclosure facilities would not be justified. It should be
pointed out that, particularly with multiple earthed systems with double circuit
lines and often using protection systems incorporating summation transformers,
it may be difficult and costly to arrange for unambiguous identification of the
faulted phase or phases. Also in the United Kingdom most circuit breakers at
 The application of protection to transmission systems 345

present installed on the transmission system have been electrically and mechanically
designed solely for three pole operation.

17.4.2 High-speed automatic reclosing

Single-shot three-phase high-speed automatic reclosing of circuit breakers was

widely applied in the 1950s to 275 kV feeder circuits. In this context 'high speed'
implies reclosure of the circuit breaker within one second of it being tripped,
typically in about 0-4s. In the event of reclosing onto a permanent fault initiation
of further reclosures was prevented.
In a high-speed automatic reclosing sequence, once tripping has been initiated no
time is available for subsequent checks to be made of the voltage conditions on the
system. In fact, owing to the inertia of the circuit breaker's pneumatic or oil and
mechanical systems, it is necessary to initiate reclosing before the tripping
operation has been completed.
The permissible time for a successful high speed automatic reclosing operation is
govered mainly by"

(a) A minimum time, dependent on the time taken to extinguish the arc at the
point of fault and for any ionised particles at the point of fault to disperse.
(b) A maximum time, dependent on the rate at which the voltage vectors at the
stations at each end of the circuit swing apart. This change in phase angle
between the voltages is governed by many factors, such as duration, location
and nature of the fault, the design of the system, in particular the impedance
of the alternative routes between the stations concerned, the load on the
system, the load on the feeder concerned immediately prior to the fault, the
generator plant in service at the time of the fault and the transient character-
istics of that plant.

It is apparent that in order to obtain the maximum circuit dead time for a mini-
mum circuit reconnection time it is desirable that the circuit breakers at each end of
the particular feeder should as far as possible trip simultaneously. Unit forms of
protection are therefore ideal for feeders which are required to be equipped with
high-speed automatic reclosing. Distance protection can however be used provided
that sufficiently fast acceleration or blocking facilities are installed.
It was found during the 1960s that on many feeders the disturbance to the
CEGB system which would result from high speed automatic reclosure on to a
permanent fault was unacceptable, and for this and other reasons, such as the
simplification of the design of circuit breakers, high-speed automatic reclosing was
replaced by delayed automatic reclosing.

17.4.3 Delayed automatic reclosing

Delayed automatic reclosing, sometimes referred to as low speed automatic re-

346 The application of protection to transmission systems

closing, can be defined as reclosure which is interlocked to ensure that specific

switching and system conditions are satisfied before reclosure takes place in
upwards of 2s from tripping. The minimum reclosure time is governed by the inter-
lock system and by the design of the circuit breaker concerned; for example some
types of oil circuit breaker should not be reclosed within 10s of tripping in order
to enable the arc-control device to refill with oil, and any gas and carbonised oil to
disperse.
The longer time taken to perform a delayed automatic reclosing operation, as
compared with high speed automatic reclosing, permits checks to be made, after
tripping has taken place, of such factors as whether the circuit concerned is dead,
or, if it has been re-energised from a remote point, whether the voltages on each
side of the circuit breaker which is to be reclosed are in synchronism or sufficiently
close in phase and magnitude. In particular, delayed automatic reclosing has the
advantage compared to high-speed automatic reclosing of enabling faulted plant to
be isolated automatically before reclosure takes place. For example, in Fig. 17.4.3A
a faulted transformer in a banked pair can be automatically isolated by operation of
the motorised disconnector M, permitting the healthy transformer to be restored to
service within a few seconds. Delayed automatic reclosing can also be applied to

A
i

Trans rans
A

Qll Lla

Fig. 17.4.3A A u t o m a t i c isolation o f a faulty transformer in a banked pair

Fig. 17.4.3B Faults on transformer feeder

 The application of protection to transmission systems 347

Trans A _ I

1"° I
Trans B

-M ~F 1

F ig. 17.4.3C F a u l t on one transformer o f a banked pair in a transformer feeder

transformer feeder circuits as there is adequate time to determine, and to signal to

each end, sufficient information concerning the fault to allow the correct action to
be taken. For example, for the primary circuit shown in Fig. 17.4.3B if the fault
occurs only on the line at F l, the circuit breakers can be automatically reclosed,
but in the event of simultaneous line and transformer faults F t + F2, reclosure
must not take place.
If two transformers are connected at the end of a transformer feeder (Fig.
17.4.3C) and a fault F~ occurs on one of the transformers, reclosure can take place
when the faulted transformer has been isolated.
Typical sequences for delayed automatic reclosing of circuit breakers are as
follows:

(a) A transformer fault on a banked pair of transformers (Fig. 17.4.3A).

Time Operation
s ¢
0 Fault occurs on transformer B at FI
0.2 Fault detected by transformer B protection and circuit breakers P, Q
and R tripped
443 Automatic opening of the h.v. disconnector M for transformer B
initiated

12+ Transformer B h.v. disconnector M fully opened

17+ Circuit breaker P reclosed thus re-energising transformer A
19+ Provided voltages across circuit breaker Q are in synchronism or if the
busbar is dead, circuit breaker Q is reclosed
348 The application of protection to transmission systems

(b) A line fault (but no transformer fault) on a transformer feeder (Fig. 17.4.3B).

Time Operation
s
0 Line fault at F~ occurs
0.2 Fault detected by feeder protection and circuit breakers P and Q tripped
15 Circuit breaker P reclosed thus energising the line
15-2 If the fault is a permanent fault circuit breaker P trips and no further
reclosure takes place
17+ Provided the line has remained energised for 2s, that is, the original
fault was a transient fault, and if the voltages across circuit breaker Q
are in synchronism or if the busbar is dead, circuit breaker Q is reclosed.

(c) Simultaneous line and transformer faults on a transformer feeder (Fig.

17.4.3B).

Time Operation
s
0 Line and transformer faults F I + F2 occur simultaneously
0.2 Faults detected by feeder and transformer protection and circuit
breakers P and Q tripped. Automatic reclosing locked out by operation
of transformer protection

(d) A fault on one transformer of a banked pair at the end of a transformer

feeder. (Fig. 17.4.3C).

Time Operation
s
0 Fault occurs on transformer B at FI
0-2 Fault detected by transformer B protection and circuit breakers Q and
R tripped. Circuit breaker P tripped by intertripping signal from trans-
former B protection.
443 Automatic opening of transformer B h.v. disconnector M initiated
15+ Transformer h.v. disconnector M fully opened. Intertripping signal
removed, permitting automatic reclosing of circuit breaker P to proceed
17+ Circuit breaker P reclosed thus re-energising the line
19+ Provided voltages across circuit breaker Q are in synchronism or if the
busbar is dead, circuit breaker Q is reclosed

17.4.4 Equipment design and programming

In the years prior to the late 1970s most automatic switching equipments consisted
of electromechanical relays, mainly all-or-nothing relays and timers, mounted in
relay cases with up to fifteen relays in each case. The relay cases were interconnected
 The application of protection to transmission systems 349

by panel wiring, as required, to provide the requisite facilities at a particular sub-

station. Some of the relays were of standard design to provide, in conjunction with
other relays, delayed automatic reclosing facilities for certain standard substation
layouts, e.g. double busbar, four-switch mesh, three-switch and single-switch
stations. There were, for example, groups of relays in one or more cases which
comprised standard designs for 'mesh corner units', 'mesh circuit-breaker units' and
'transformer l.v. circuit-breaker units'.
With the increasingly high costs of switchgear in the 1970s, it was found more
economical to depart from standard designs of substation, and to tee-connect or
bank primary circuits in nonstandard configurations. This entailed designing the
protection and automatic switching requirements very much on an individual
circuit and substation basis, and the standard circuit diagrams for automatic
switching schemes and equipment were no longer appliable, except possibly in the
case of double busbar stations without tee-connected primary circuits. A typical
example of a delayed automatic reclosing scheme for a double busbar station
using electromechanical relays is given in Fig. 17.4.4A. This simple scheme is for
the automatic reclosing of a feeder circuit breaker after a feeder fault. The corres-
ponding sequence chart is given in Fig. 17.4.4B.
A further difficulty which arose from using standard designs of automatic re-
closing equipments was that of modifying existing schemes to deal with the addition
of new banked primary circuits at an existing station. Extensive changes in ex-
isting standard relays and panel wiring were involved which were both costly and
time consuming.
Furthermore, the use of conventional electromechanical relays to perform a
large number of simple logic and timing functions required a relay panel suite
typically 4 m long for the relays associated with the automatic switching of the h.v.
circuit breaker at a four switch mesh substation. This was reduced to about 2 m by
using static relays.
Another problem which arises from the use of electromechanical relays is that
some logic circuits require many relay contacts to be connected in series which, if
not regularly maintained, introduce a risk that the scheme may fail to operate
owing to contact failure. The risk of electromechanical equipments failing to
operate when required to do so has been found in practice to be about 7% of the
total number of required operations.
These problems can be overcome to some extent by installing suitable designs
of sequence controllers using programmable wired logic arrangements, such as a
diode matrix.
However, the logic requirements for automatic switching equipments have to
take into account not only the conditions obtaining at the start of the automatic
switching sequence but also any changes in these conditions whilst the sequence is
in progress. As an example, if there is a thunderstorm in the vicinity of a substation
with a complex primary circuit configuration, a line may be struck by lightning
resulting in tripping of circuit breakers at that substation. If, before there is time to
reclose the circuit breakers, another line into the substation is struck further circuit
350 The application of protection to transmission systems

Line ,
Part of voltage
selection and
synchronising
scheme Neutral ='~~S-Y Y..~ . , [

Trip reJsys

T RO/4
m

Line disconnector
~,..-,~ aux iliar~ switch B- 1 L- 1 VT F/2
1 . . . - - - .

Circuit-breaker (2 sec. release)

auxiliary switch V.T. (~ [ _ . ,~O _' _- _P_ ~- 3
ID-I~ ~ Buchholz~ TRO-I_ VTF-I
Busbar
protection
VTF-2
B-2 LKI L-3
O O~ID~'~--~r--! (Delayed
O~ I I °perate
O C-
m
-
~ 2
o
P--2
o----
LK3 I I 5-6~0sec.)

Circuit-brea ker
low air pressure
B- L-4 D/!
repeat . . . . . . . . ....

C:i= 2 sec. operate---- C/4

-'l~r,v,y~ ]

_ 2 sec. release " ' " ~ ' I ' Y ' ~

4 ' ~ -
m

P-4
! MC/1
To indication ( =
circuit O.,m..O--~
C-4
! : °°-3
(60 sec. operate)
_ l'Circuitlbrea k e r - !
/closing circuits . [

Circuit-breaker Circuit-breaker
closing +ve
closing -ve

Fig. 17.4.4A Delayed autoreclosing for feeder at a busbar station

 The application of protection to transmission systems 351

KEY TO USE OF" LINKS 1 2 3 4 $

. . . . . .

i. Full synchronising facilities:

1.1 Synchro. check only O O O O O
1.2 Dead line charge or synchro, check I O O O O
1.3 Dead bar charge or synchro, check O O O O i
1.4 Dead line or bar charge
or synchro, check O O O 1

2. Limited voltage selection

(synchro. check n o t available):
2.1 No line reference v.t.
2.1.1 Assumed dead line charge
bar alive . . . . . . . . . . . . . . . . . . . . . . . . I O O I O
2.2 No busbar referencev.t.
Part o f voltage 2.2.1 Assumed dead bar char~
selection and line alive . . . . . . . . . . . . . . . . . . . . . . . O 1 O O !
synchronising scheme 2.2.2 Dead line charge bar
assumed alive . . . . . . . . . . . . . . . . . . . . O 1 1 0 0

O - Link o u t I - Link in
~ R

S Busbar side voltage m o n i t o r i n g relay

SYA C Reclosing relay
CP Close pulse timing relay
D Dead line reclose timing relay
ID Persistent intertrip detection timing relay
IRTRX - lntertrip receive trip relay
L Line side voltage m o n i t o r i n g relay
MC - Manual close repeat relay
P Reclose start relay
PY Line dead check start seal relay
RC - Reclose cancel relay
SY&SYB- Synchronising selection relays manual
SYA - Synchronising selection relay auto-reclose
SYN - Synchronising relay
TRO - Trip relay o p e r a t e d relay
VTF - V.T. m o n i t o r i n g and switchgear repeat relay
352 The application of protection to transmission systems

SEQUENCECHART. I.'OR DELA.yED AUTOMA. TIC RECLQSING OF A F E E D E R

CIRCUIT BREAKER AT A BUSBAR STATION

Feeder protection Feeder circuit-breaker was

trip relay operated closed, and line v.t. was alive
up to 2 sec. before fault

No operation of line
t
Feeder protection
v.t. Buchholz trip relay operated

I
Time delay
I
C.B. opened
( I 0 sec.)
I
I Line v.t. dead
Feeder protection
trip relay reset
I
No operation of line v.t.
Buchholz or busbar protn.

I
Busbar v.t. alive

I
I _
Line v.t. alive

I
l-nergise
Time delay
(variable 5 - 3 0 sec.)
synchronising relay
I I
No operation of line
Synchronisin~ relay v.t. Buchholz or
operated after a busbar protn, during
minimum delay of 2 sec. above sequence

I
No operation of line
I
Line v.t. dead
v.t. Buchholz or busbar v.t. alive
busbar protn, during
above sequence

Reclosure
by hand

. . . . . .

if fault persistent ('h)se C.B. Inhibit fu~rther

trip relay reo~erates (2 sec. pulse) initiation of
auto-reclose sequence
I
C.B. tripped C.B. closed
I Time ~delay
(4 sec.)
Signal sent to lockout I
'Auto-reclose Operated' I
reclosure of remote end Reset
indication
circuit-brea ker (usually auto-reclosing
a prolonged equipment
intertripping signal)

Fig. 17.4.4B
 The application of protection to transmission systems 353

breakers will be tripped, necessitating in some cases a change in the sequence of the
reclosure requirements. It is therefore essential that all the appropriate input data is
scanned by the automatic switching equipment before each sequential step is taken.
In practice, this results in relatively simple controllers such as are sometimes used
for controlling simple sequential processes being unsuitable for automatic switching
applications.
It also results in it being very difficult to prepare fully comprehensive sequence
charts or flow diagrams to show all the logic requirements for complex automatic
switching schemes, as so many options have to be shown. If such a chart or diagram
is prepared it is essential that all the logic is included, i.e. the requirements for
nonoperation as well as operation of the automatic switching equipment.
Another problem is to ensure that the various proposed sequence operating
times are correct under unusual primary system switching conditions and that,
when necessary, the times are automatically changed within the equipment.
In practice, these difficulties result not only in complex logic requirements for
the automatic switching equipment but, equally important, the need for compre-
hensive testing facilities to enable the proposed requirements to be checked for
various assumed fault locations and system switching conditions.
Dealing first with the automatic switching equipment itself for complex primary
circuit configurations the interactive logic the equipment must provide has to be
based on some form of computer technology, usually incorporating microproces-
sors in the design. A microprocessor can be defined as a digital processing unit
constructed as one or more integrated circuits using large-scale integration (l.s.i.)
manufacturing technology. It is an extremely compact programmable electronic
component from which a digital computer can readily be constructed by the ad-
dition of memory, clock, input and output components. For an automatic
switching equipment application suitable interfacing devices e.g. reed relays, opto-
isolators etc., are required between the computer and the substation equipment.
A great advantage of a microprocessor based equipment is its flexibility. The
hardware, can be ordered and installed largely independently of specifying the
detailed requirements of the automatic switching scheme. This greatly eases con-
struction problems. Furthermore, most changes in scheme requirements either
before or after installation can be catered for by the relatively easy process of
changing the software. The changes are usually accommodated by removal and re-
placement of the appropriate memory modules followed by the necessary retesting.
The microprocessor-based equipments are much smaller than their electro-
mechanical relay predecessors, typically one 19 inch rack provides for a complete
substation, most of the space being occupied by external cable terminations.
The equipments are also in practice more reliable than electromechanical equip-
ments, particularly as by using suitable programs they can both monitor themselves
continuously and check for discrepancies in the input information.
The main disadvantage of microprocessor-based equipments is that extensive
precautions have to be taken to ensure that interference from the substation
environment does not reach the integrated circuits. This interference can easily
354 The application of protection to transmission systems

cause damage and malfunctioning unless the equipment is specifically designed to

operate in a substation environment. One of the main sources of interference is
caused by the opening and closing of disconnectors in the substation. These discon-
nectors can generate high energy interference over a very wide frequency band, in
some instances up to tens of MHz, and the automatic switching equipment may
itself be controlling the opening and closing of these disconnectors.
The preparation of the programme for the microprocessor-based equipment (the
software) requires great care. One successful technique is to use in the design office
a computer as a design aid and to program this computer in such a manner that it
can satisfactorily simulate the input logic and also the equipment in the substation
being controlled. In this way the proposed logic can be checked to prove its suita-
bility under various system switching conditions for various assumed fault locations.
It should be noted that the computer model of the substation and associated
primary system must include simulation of remote substations and also of the
relevant secondary equipment at those substations, e.g. intertripping send and
receive equipments with their appropriate operating times. In addition, in order to
follow and analyse the sequential operation of the proposed logic and of the
simulated plant in the system model, all relevant changes in inputs, outputs and the
necessary internal memories and timers have to be logged, as shown schematically
in Fig. 17.4.4C.
Having proved the logic requirements for the automatic switching scheme, the
next step is to program the automatic switching equipment in the required manner.
For electromechanical and diode logic equipments this is a relatively easy task but
for microprocessor based equipments the program has to be prepared in a suitable
format by a computer and fed into the equipment by suitable programming devices
e.g. for coupled-charge p.r.o.m.s (programmable read-only memory) components
pulses of electrical energy for each bit of information.
A considerable amount of effort can be saved in programming microprocessor
based equipment if the substation is split into defined functional groups of plant

! 1
I System & I
"1 fault model I
L . . . . . . J
! !
I Auto-switching I
"qL e2 u i j m j n t m,.._~deljI

-L l'Zvent logger

Teletype or V.I).U. 1_._ .__ r- "" " "" "-" --- "-" -i

interactive ! -- r "1 Executive I Recording

I_ . . . . . . _1
control
Printout

Fig. 17.4.4C Computer used for con trol logic design

 The application of protection to transmission systems 355

Substation
Nodes
Remote ~ =,11, =,= m ......... C Group !

Substation

C.B.'s ,~ ,, ~.. Group 2

[~l ~m mm

Adjacent ¢ / 1~ .~ Group 3

Nodes in
i'-1_
Substation ~- I._i / ~ Group 4

~. Group 5

Groups for Assembly of Model

Node Node
L-J
-- / ---O

/
Node

I
I
I
I Model constructed from Groups
I

Fig. 17.4.4.D Substation model

(Fig. 17.4.4D) which can then be interconnected to form a substation and standard
software designed to control these functional groups, in a standard manner e.g.
'This line has a persistent fault. Open the disconnector'.
Using a library of standard software modules, programs can be built up to
perform standard automatic switching functions, e.g. 'After transformer fault on
tee connected transformer feeder, open transformer disconnector and close line
circuit breakers at mesh station'. This technique has not only the advantage that the
program can be prepared quickly but also, once a standard module of software has
356 The application, of protection to transmission systems Next Page

been proven for a particular grouping of primary plant, the standard module can be
confidently incorporated in programs for future substations, even though the
automatic switching equipment may be of a different type.
Assemblies of the standard software modules can thus be used to simplify the
determination of the logic requirements for a complex automatic switching scheme
without the need to prepare complex sequence charts or flow diagrams. This results
in considerable savings in engineering and drawing office manpower. It is, of course,
essential that the program documentation incorporates the facility for 'plain
language' comments on the tasks being accomplished at each step.
17.4.5 Commissioning

A major problem with complex automatic switching schemes is to commission the

equipment once it is installed in the substation.

Disconnection points
S u b s t a t i o n plant
Test points
'.p IX' ,i l
------- I~-- I
a I

Substation
"----
~
1 ~ I O
'
I
1 O'--- I
auto-switching I
equipment ~ D IO----
(micro-pr,,cess,,r i ~ ~ I
based) ! --I. I

l
1

~ T e s t lead

('~pnlputcr ~ ~ ~. __I '__o_L L / I

J System & I
J fault model I
L . . . . . . . .J

1 Event logger j ~ [V.D.U.

T e l e t y p e ,,r V I).
interactive
con trol l
/-- v L . . . . . . .J cor d;
ng

Printout
Fig. 17.4.5A Computer used for on-site testing
Previous Page The application of protection to transmission systems 357

Operation of the primary plant itself by the automatic switching equipment,

apart from a few checking operations, is unacceptable from the point of view of
requiring an unduly large number of operations of circuit breakers and discon-
nectors to prove all the required sequences. It would also be too costly in terms of
the outage times which would be required.
With the growth in complexity of the schemes it is also found that adequate
modelling of the substation plant by relays, switches and lamps becomes too
difficult.
For complex schemes, a fairly elaborate model of the substation is required
which can conveniently be provided by a transportable mini- or microcomputer
programmed to simulate the substation concerned. In some cases it may be
practicable, provided it is transportable, to use the same computer for the com-
missioning tests as used for programming the automatic equipment. An example
of this technique is shown schematically in Fig. 17.4.5 A.

17.5 Economic considerations

The cost of a protection system is influenced by such factors as:

(a) the complexity of the protection equipment,

(b) the voltage and current rating of the necessary ancillary equipment associa-
ted with the protection, for example current and voltage transformers, line
traps, pilot cables, etc. The cost of these items may be very many times the
cost of the associated protection relays,
(c) the development and engineering costs of the protection system,
(d) whether or not the protection equipment is being purchased in bulk, and
(e) the special requirements of individual purchasers.

It is therefore not possible to quote general prices for protection systems, but some
indication of the relative costs of protection relays is given in Fig. 17.5 A.
Type of relay Relative cost

Single-polo instantaneous differential current 1.0

Three-pole i.d.m.t. 2.5

Three-pole directional i.d.m.t. 5-0

Pilot wire protection operating over metallic pilots with

starting and supervision relays (per end) 124)

Switched high-speed distance relay for phase and earth

faults (per end) 20.0
Fig. 17.5A Typical approximate relative costs of protection relays
358 The application of protection to transmission systems

In costing any scheme, full account should be taken of all the relevant costs
involved; for example, the capitalised cost of any hired pilots, the total installation
costs, and the probable maintenance costs, including the cost of any necessary
circuit outages.
The cost of the consequences of the proposed protection schemes failing to
operate should also be assessed, e.g. the cost of any consequential interruption of
supplies to consumers.

17.6 Typical protection applications in a major transmission system

In this Section, the factors involved in selecting protection systems for typical
applications in a major transmission system are considered but, as it will be seen,
in some cases the choice is not easy. The schemes described in this Section are
those employed for protecting high-security transmission systems requiring fast
fault clearance times for all fault locations. For less important transmission sys-
tems and e.h.v, distribution systems the requirements are not so onerous, and
various economies and technical relaxations can be made; some of these are
described in Section 17.7. It should be emphasised that the characteristics quoted
for various types of protection refer to the protection systems at present available.
New systems are continually being developed, and these developments modify some
of the factors involved in choosing a suitable scheme.
For example, the recent development of very-high-speed protection systems
measuring the direction of propagation of the step changes in current and voltage
produced at the point of fault ('travelling-wave' protection) have recently been
applied in some parts of the world.

17.6.1 Feeder protection

In order to provide fast and fully discriminative protection for feeders with the
requisite degree of redundancy in operation and diversity of design (Section 17.2.1)
two separate main feeder protection systems are normally provided. The back-up
protection is usually definite minimum inverse-time overcurrent and earth-
fault protection. In areas of exceptionally high earth resistivity a definite-time
earth-fault protection may also be provided with a very high sensitivity to operate
for high resistance earth faults.

17.6.1.1 Protection for a long overhead feeder: In this context a long overhead
feeder implies a feeder over about 30 km long but under 200 km, with a negligible
amount of cable in its route.
In order to obtain as much diversity in the protection as possible it is preferable
to have a current differential system for one main protection and an accelerated
distance protection for the other.
In view of the distance between the substations a power-line-carrier phase-
comparison system is normally used for the differential system, although up to
 The application of protection to transmission systems 359

40 km a voice frequency phase comparison system utilising a telecommunication

company channel can be used. The latter is considerably cheaper than power-line-
carder protection, as no coupling equipment and line traps are required. It is
probable however that the distance protection would normally use a
telecommunication company channel for acceleration, and hence the desirable
diversification of communication channels for the first and second main protection
is lost.

(a) Carrierprotection: Practically all the power-line-carrier protection at present

in service in the UK is static phase-comparison protection. This form of protection
replaces older thermionic phase comparison protection and directional comparison
protection.
The main advantages of phase comparison power line carrier protection are:

(i) It provides fast and discriminative protection along the whole length of the
protected feeder.
(ii) It does not require a voltage input, hence it can operate correctly in the event
of a voltage transformer associated with the other main protection on the
feeder being defective.
(iii) It is in general more sensitive than distance protection to high resistance earth
faults (e.g. faults caused by ionised and polluted air from grass fires).
(iv) The circuit breakers at each circuit end of the protected feeder are tripped
within approximately 20 ms of each other, irrespective of the fault location.
This feature is very important if high-speed automatic reclosing is provided; it
is not important with delayed automatic reclosing.
The main disadvantages are:

(a) high cost

(b) possible difficulties in allocating frequencies in areas where an appreciable
amount of carrier protection is installed
(c) relatively complex commissioning, testing and maintenance procedures are
required
(a) under certain conductor icing conditions the carrier signal may be so attenua-
ted that the protection has to be taken out of service to avoid the mal-
operation of the protection on through faults
(e) normally even short lengths of cable cannot be inserted in the path of the h.f.
signal. This limitation can be particularly difficult as it may not be known
until a relatively late stage in the planning and construction of an overhead
line, that for example for environmental reasons, a length of cable has to be
inserted.

The early designs of phase comparison power line carrier protection were of the
thermionic type and naturally required noninterrupted power supplies of many tens
of VA. This necessitated in most cases the provision of large d.c. motor-generator
360 The application of protection to transmission systems

sets driven from the 110 V station battery. Not only did these motor-generator sets
require a considerable amount of maintenance and room to accommodate them in
the substation buildings but, in order to provide as is customary 6 hours standby in
the event of the l.v.a.c, supplies failing, a larger station battery was needed. With
the advent of static phase-comparison protection and the replacement of thermionic
protection, a considerable reduction in these auxiliary power supply requirements
has occurred. Nevertheless, the provision of adequate non-interrupted power
supplies for the phase-comparison protection is still necessary. The present practice
is to use 48 V battery supplies for this purpose as they are subjected to much
smaller voltage fluctuations in service than the 110 V battery used for closing and
tripping circuit breakers.

(b) Distance protection: In UK applications the polar characteristics of distance

protection are usually the conventional circular mho characteristics, although
more complicated shaped characteristics are being introduced, where justified, to
reduce the problem of setting the third zone so that it will not maloperate under
emergency heavy load transfer conditions.
For both phase and earth faults separate or common measuring elements per
phase are used for zones 1 and 2, and separate measuring elements for zone 3. It is
considered that the extra reliability and speed obtainable with a multielement
distance relay compared with a switched distance relay are justified for protecting
feeders in a major transmission system.
Static designs of distance protection were introduced in mid 1960s and have
steadily taken over from electromechanical designs. The main advantages of static
distance relays are flexibility of characteristics, in particular the ability to shape the
characteristic in the required manner, low VA consumption and high speed of
operation.
The disadvantages of static distance relays are the need for auxiliary
power supplies and the greater susceptibility of the relays to maloperate or be
damaged by high frequency interference. The latter risks have in practice been
reduced to a negligible magnitude by careful design and extensive type testing.
Recent designs of distance protection incorporate microprocessors as embedded
components in their design. This has resulted in great flexibility in characteristics
and logic sequences which can be easily changed without changing the design of
the hardware. The characteristics themselves can also be improved, for example,
the operating time can be made very fast (7 ms) provided there is no distortion
in the input signals. In the presence of distortion the measuring time can be appro-
priately lengthened to ensure correct operation of the protection. Improved test
facilities are also available if required, e.g. built-in on-line testing. The latter facility
is included to improve the overall reliability of the protection in service.
If used in conjunction with phase comparison protection as the other main
protection, for diversification reasons the distance protection should preferably be
accelerated rather than blocked as phase-comparison protection is in this respect
fundamentally similar to a blocking scheme, i.e. the local end must not be allowed
 The application of protection to transmission systems 361

to trip until it has waited to receive information from the remote end and that
information has been compared with locally derived information regarding the
current zero crossing points.
The main advantages of distance protection are:
(i) It provides fast and discriminative protection for all faults occurring on the
feeder on which it is installed to within about 25% of the feeder length from
the remote substation.
(ii) It provides some measure of back-up protection (zone 2 and zone 3) for other
feeders and plant at remote stations and, often, on reverse reach of zone 3 at
its own substation.
(iii) It does not necessarily need an incoming signal from a remote station in order
to trip, even though the lack of an incoming signal may delay tripping.
The main disadvantages of distance protection are:
(a) It requires costly voltage transformers and if the protection is of the high
speed type, e.g. under 10 ms, it normally requires voltage transformers with
a high fidelity transient response; to obtain this response may be very costly.
(b) With some designs it may be difficult to obtain very fast clearance of close-up
three-phase faults due to lack of polarising voltages.
(c) The protection may trip on load current in the absence of a voltage input, e.g.
due to the voltage transformer fuses being inadvertently left out after line
maintenance. To overcome this problem very-high-speed voltage transformer
output monitoring relays may be installed to disconnect tripping before the
protection can maloperate, but the cost and complexity of this monitoring
equipment and the risk of the monitoring equipment maloperating and dis-
connecting tripping during a system fault has to be carefully assessed.
The available choice is between blocked or accelerated distance protection. Blocked
distance protection has the advantage that when the remote end circuit breaker is
open, faults are cleared quickly and discriminatively along the whole length of the
protected feeder. On the other hand, for an accelerated protection with the remote
circuit breaker open there is no operation of zone 1 protection at the end with the
open circuit breaker and hence no acceleration signal is sent from that end. Faults
on the feeder in zone 2, i.e. approximately 25% of the line length from the open
circuit breaker are therefore only cleared in zone 2 time, which may be too long to
be acceptable.
In the event of a known failure of the communication channel the blocked pro-
tection must be switched out of service as it would be unstable on a through fault.
An accelerated protection however has the advantage that in the event of a
known loss of the communication channel the protection can remain in service, but
faults occurring in zone 2 reach will only be cleared in zone 2 time instead of the
faster accelerated time.
Various designs of blocking systems are available from manufacturers with
differing performance characteristics and the most suitable system has to be chosen
for a given application. Many of these systems include a first zone tripping facility
362 The application of protection to transmission systems

independent of the receipt of any signal. The blocking or acceleration

communication channel for the distance protection should ideally be of a different
type from that used for other main protection on the feeder. On the other hand in
the United Kingdom for a feeder, say, 30 km long using a voice-frequency phase-
comparison protection over a telecommunication company channel for the first
main protection, it might sometimes be difficult to justify the use of a carrier
channel for blocking or acceleration of the second main distance protection. In fact,
if a carrier channel could be economically justified it would normally be better to
use the carrier channel for the phase-comparison protection (and thus eliminate the
inherently long propagation time over a telecommunication company circuit of the
complex phase comparison signal) and to use the channel for the simpler blocking
or acceleration function. The latter proposal would not apply, however, where, for
example due to a short length of cable in the cartier signal path, the attenuation by
the cable and, more importantly, the reflections of the signal by the cable sealing
ends result in a carrier signal which may be distorted and have an indeterminate
effective phase shift in its propagation. In such a case it is preferred to transmit and
receive the relatively simple blocking or acceleration signal over the carrier
channel and rely on the telecommunication company channel to transmit the more
phase sensitive phase-comparison protection signal.
Where a carrier channel cannot be used, e.g. lack of room in the available
frequency spectrum or excessive attenuation due to a short length of cable in the
middle of the long overhead feeder, double distance protection with
telecommunication company signalling channels may have to be used. In this case,
for diversity reasons, one of the main protections should preferably be of the
blocked type and the other accelerated. This ensures that with a remote end circuit
breaker open fast clearance times are provided by the blocked distance protection
on the feeder. Also if all communication channels fail in an extreme emergency,
e.g. due to widespread flooding of telecommunication company circuits, the three-
zone feature of accelerated distance protection could still be used to provide
discriminative protection for the feeder, albeit with some time delay.
In areas where high resistance earth faults are considered likely, the earth-fault
distance protection can usefully be supplemented by a directional comparison earth
fault protection module in order to provide the extra earth fault sensitivity; the
separate signalling channel required for this facility would normally be over the
same communication link which is provided for the associated distance protection.

7.6.1.2 Protection for a short overhead feeder: In this context, a shorter feeder is
under 30 km long with a negligible amount of cable in its route length. A wider
choice of protection is available than for long feeders since the use of private pilots
as well as telecommunication company channels become economically feasible.
Pilot wire systems are described in detail in Chapter 10.
The choice of protection is largely governed, in particular for very short feeders
a few kilometres long, by the cost of providing private pilot cables. As an extreme
example for protecting an overhead line between two substations, say 4 km apart
 The application of protection to transmission systems 363

in a city area, it may be possible to utilise spare pairs in an existing pilot network at
negligible cost. If, however, new pilot cables had to be laid, the cost of opening
trenches, crossing and resurfacing roads etc., solely to provide pilot cables for
protection could well be prohibitive. Overhead pilot cables suspended from the
towers or embedded in the earthwire have rarely been used in the United Kingdom
owing to unsatisfactory service experience. Fibre-optic cable embedded in the
earthwire may in future provide a more satisfactory communication link.
Often underground pilot cables for protection and control purposes are installed
in trenches on the same route as power cables, and hence are subject to induction
and rise of earth potential effects in the event of faults on the power network.
These induction problems can be one of the main limitations in using private
pilots for protection. In a few cases it has even been found necessary to install
isolation transformers every few kilometres in an existing pilot in order to reduce
the induced voltages impressed on the pilot cores. Induced voltages, in addition to
causing breakdown of the insulation of the pilot cable, in extreme cases may cause
maloperation of the differential protection to which the pilot cores are connected.
Details of the precautions which should be taken in choosing pilots are given in
Chapter 7.
Where for the first and second main protection two physically segregated cables
are not available for two private pilot protection systems, it is often economically
justifiable to provide one private pilot wire protection as the first main protection
and a telecommunication company pilot wire protection as the second main
protection. In the past, metallic pilots, with known resistance and capacitance
characteristics, have been rented from telecommunication companies which are
suitable for a longitudinal differential protection system and these are likely to
remain in use for many years, especially on very short routes. Where only voice
frequency circuits can be rented from telecommunication companies a protection
using a voice frequency signalling system can be used, such as voice-frequency
phase-comparison protection or blocked distance protection.
Where no private pilots exist nor can be economically justified, one of the main
protection systems may be a telecommunication company pilot system, and the
other should be of fundamentally different design such as distance or, for important
circuits, power-line-carrier protection. The latter two types of protection are costly
and in some cases cannot be justified; in this case two telecommunication company
pilot protections with separately routed pilots may be installed.
For very short overhead feeders, depending on the impedance of the source of
fault current and the impedance of the circuit to be protected (i.e. the system
impedance ratio) distance protection may be unsuitable; most distance protection
systems have a maximum s.i.r, of about 50. In practice, most lines in a transmission
system longer than about 6 km can be satisfactorily protected by distance
protection, but it may have to be of the blocked type.

17.6.1.3 Protection for an underground feeder: Most of the considerations which

apply to short overhead feeders apply equally to underground feeders. In view of
364 The application of protection to transmission systems

the high capital cost of e.h.v, cables, these feeders are normally under 30 km.
However, private pilots for protection purposes are much more likely to be
available than for at, overhead feeder as usually the cheapest way to provide the
necessary communication channels between substations is to lay pilot cables on the
same route as the e.h.v, cable, preferably in separate trenches to reduce induction
problems. The latter particularly applies where individual e.h.v, cables are provided
for each phase of the feeder.
On some feeders and networks, the high capacitance of the primary cables may
require shunt reactors to be connected to the cable circuits either permanently or
through a circuit breaker. This can lead to severe switchgear rating and protection
problems related to the high charging current of the cables and to the interruption
of highly inductive reactor currents. In general, separate fully discriminative pro-
tection systems are required for the cable and for any permanently connected
shunt reactor to provide adequately sensitive protection for the reactor and to
establish quickly which is the faulted item of plant.
The main problem in protecting cable circuits is the high charging current which
may be an appreciable fraction of the load current; this limits the choice of mini-
mum fault current setting for the protection. In fact, a maximum current setting of
several times the steady state charging current depending on the manufacturer's
recommendations may be necessary to ensure that the high transient discharging
and charging currents in the primary cable circuit under external fault conditions
do not cause the protection to maloperate. Similar transient currents flow when
cable circuits are energised or de-energised. The frequency and magnitude of these
transient currents depend not only on the parameters, i.e. capacitance, inductance
and resistance of the primary circuit being energised, but also on the characteristics
of the circuit breaker, e.g. the values of any switching resistors. These transient
currents have, in practice, a frequency of the order of a few kilohertz, and in view
of the low shunt impedance of the cable circuit at such frequencies appear as
internal fault currents. The protection has to be designed to be sufficiently in-
sensitive at these frequencies to prevent it from maloperating.
If distance protection is used for protecting the cable circuit the characteristic
angle of the distance relays should preferably be approximately the same low angle
as that of the cable circuit being protected in order to obtain optimum settings.
The setting of distance protection on cable circuits with crossbonded sheaths
presents a particularly difficult problem as the effective impedances of the cable
circuit are dependent on the return paths of the fault current, and may vary over
a wide range of values depending on the fault location, resistivity of the cable
trench back-filling, etc.
As mentioned in Section 17.2.1.1, where more than one intertripping or sig-
nalling channels are required for dependability they should be as diverse as possible
with regard both to the design of equipments installed and the types of communi-
cation channel used.
Problems may, however, arise as a direct result of providing this diversity. For
example if it is assumed that the only communication channels which can be
 The application of protection to transmission systems 365

economically justified for the protection and intertripping of the double circuit
transformer feeder shown in Fig. 17.3.1A are two separately routed private pilots,
and that each feeder is equipped with two main protection systems with duplicate
intertripping for each feeder, then in order to reduce the risk of a failure to trip it is
evident that the first main protection for TFI should be associated with pilot cable
No. 1 and the second main protection associated with pilot cable No. 2. Similarly
the first intertripping channel for TFI should be in cable No. 1 and the second inter-
tripping channel for TFI in cable No. 2.
However, some forms of protection may maloperate on load or through fault
currents in the event of the conductors in the pilot cable becoming short circuited.
Also with a simple d.c. intertripping system a live conductor in the pilot may be
inadvertently connected to another conductor through a short circuit, resulting in
maloperation of the intertripping equipment.
One of the main causes of damage to pilot cables, particularly in urban areas, is
building or road construction work where, for example, sheet piling may accidentally
result in the severing and short circuiting of conductors in the pilot cable. In this
event, if the communication channels for TF1 and TF2 were routed in the
above manner, both TF1 and TF2 might be inadvertently tripped. On the other
hand, if all the protection and intertripping associated with TF1 were routed in pilot
cable No. 1 and that for TF2 were routed in pilot cable No. 2 the risk of inadvertent
tripping would be confined to a single circuit.
Where the protection or intertripping for more than one primary circuit is
routed in one pilot cable the risk of more than one primary circuit being inad-
vertently tripped can be reduced in various ways. For example, the use of starting
relays for the protection will reduce the risk of protection maloperation and the
risk of inadvertent maloperation of the intertripping can be practically eliminated
by using a voice frequency intertripping system instead of a d.c. system.
In practice, an engineering assessment has to be made of the consequences of
the loss of, or damage to a pilot cable or equivalent signalling channel and the
appropriate measures taken within the economic constraints.

17.6.2 Protection for a transformer

Details of the protection systems available for protecting transformers are given in
Chapter 11.
Transformers connected to a transmission system vary widely in size from
several hundred MVA down to a few MVA, the latter being mainly for providing
auxiliary supplies at small power stations. It is usual, at least as far as the high
voltage connections are concerned, to provide similar protection for these small
transformers to that provided for the larger transformers, as a slowly or incorrectly
cleared fault on a small transformer may be as disastrous to the operation of the
transmission system as a similarly cleared fault on a large transformer. Hence all
transformers directly connected to a major transmission system require fast earth-
and phase-fault protection, although the risk of phase faults clear of earth occurring
inside a transformer tank is extremely low.
366 The application of protection to transmission systems

The risk of phase faults occurring on the h.v. or l.v. connections to a trans-
former is, however, not low, especially if open type connections are used. Flying
debris can lodge on the connections although even if the connections are completely
phase segregated, safety earths inadvertently left connected after maintenance work
can result in three phase faults. As mentioned in Section 17.2.1, discriminative
clearance of phase faults may be difficult if inverse definite minimum time overcur-
rent protection is the only phase fault connection provided for the transformer.

El.EEl

I , |

Id ' L

I m.
p

].

Interposing
transformer -tl l
•
i i.. ]
I-"-"1

o REI"

Interposing
transformer

Restraint coils
_ I

I
I
Operating coils

I
Fig. 17.6.2A Transformer overall differential protection
 The application of protection to transmission systems 367

With regard to the protection of the transformer itself, for double wound trans-
formers fast earth-fault protection is usually obtained by unbiased differential
protection separately covering the l.v. and h.v. windings, and fast phase-fault
protection by a biased overall differential protection system (Fig. 17.6.2A). For
economic reasons the earth- and phase-fault protection systems usually share the
same current transformers. In addition to some form of inverse-time back-up
overcurrent or earth-fault protection, Buchholz protection, and on some trans-
formers other gas pressure operated devices, are provided. The Buchholz protection,
as well as providing back-up protection for phase and earth faults on the trans-
former windings, also provides some measure of protection for interturn faults on
the windings, and for core faults, i.e. failure of the lamination insulation resulting in
high damaging circulating currents within the transformer core.
Most forms of transformer overall differential protection use a harmonic
restraint feature to prevent the protection maloperating in the presence of
B m
mmm

IJ'Z_ [] I

--" 1 _ I-"1

r .

Fig. 17.6.2B Typical example of phase and earth-fault differential protection for individual
winding of a transformer
368 The application of protection to transmission systems

harmonics such as occur, for example, in the transformer magnetising inrush

current. With loads having a very high harmonic content, e.g. arc furnace loads, this
harmonic bias might result in very slow operation or even nonoperation of the
transformer overall differential protection in the event of an internal fault occurring
in the transformer. An alternative way of providing phase fault protection for such
applications is shown in Fig. 17.6.2B where the protection for the l.v. windings
only has been shown. Each winding is provided with a separate differential
protection which avoids the transformer inrush and harmonic current problems but
it requires current transformers to be mounted in the delta winding connections.
This may be difficult and costly in practice and can rarely be justified, bearing in
mind the very low risk of phase faults, clear of earth, occurring within the trans-
former tank. A typical protection scheme for a double-wound transformer is
shown in Fig. 17.6.2C.
If the h.v. or l.v. connections are very short, the separate protection can be
omitted, but the main reason for providing this protection is to give discriminative
indication of the fault location which reduces appreciably the time required to
establish the exact location of the fault. If, for example, the h.v. connection pro-
tection has operated and there is an insulation co-ordinating gap in the protection
zone covered by the h.v. connection protection, there is a high probability that the
fault was caused by the gap flashing over and that it is safe to restore the trans-
former to service without any tests being made. In some cases it is permissible to
initiate automatic reclosiag from the connection protection, arranging of course
the prevention of automatic reclosing if the transformer differential protection"
has also operated.
For phase faults, the unbiased differential connection protection is also inherently
faster than the biased differential transformer protection, thus providing faster
clearance of phase faults than would be the case if the biased transformer dif-
ferential protection zone embraced the transformer connections.
The high set overcurrent protection in Fig. 17.6.2C fed from the current trans-
formers in the transformer h.v. bushings provides fast clearance of phase faults at
the upper end of the transformer h.v. winding. It also backs up the transformer
overall differential and restricted earth fault protection. The main difficulty when
applying high set overcurrent protection is in establishing a suitable setting. The
setting current must be sufficiently low to ensure that it will operate quickly
with the minimum fault current available, yet it must not be so low that it will
respond to faults on the l.v. side of the transformer or operate with transformer
magnetising inrush currents which, for some types of transformer, may be as much
as seven times the full load current.
The back-up inverse time overcurrent protection operates in two stages to
provide discriminative clearance of faults beyond the transformer l.v. circuit
breaker in the first stage, and faults within the transformer or on its 1.v. connections
in the second stage. Both the high-set and inverse-time overcurrent protection are
often energised from current transformers in the individual transformer h.v.
bushings to permit future banking of a second transformer without having to make
 The application of protection to transmission systems 369

t i

/
/

lnvc ---'r
- I
I
I
I
i It 1
CK
,
CK

I __, ......... I~!

I
I
I
I
I
I
I
~--I
, o~ l ¸
I I
I
~- - RES
I I HV
I

1
I
I l
t_.._

I
I

I
I

F-
I LT! l

I
I
I
. ,-g. I
B

g
t"--
I
I
] I
m

m I
_J

Fig. 17.6.2C Protection scheme for a double w o u n d transformer

370 The app//cat/on of protect~on to transmission systems

R
|.| R
7

......~ "~t_._ ¥
B 'L '
____] '

[
A

Delta t e r t i a r y
ind ing

,, i v , ,-

, k - ,

Fig. 17.6.2D Differential protection for an auto-transformer

major changes in the protection of the existing transformer.

The unbiased differential protection of an auto-transformer is applied on a
phase-by-phase basis as shown in Fig. 17.6.2D. The protection is equally fast and
sensitive for phase and earth faults.
A typical protection system for an autotransformer and its connections is shown
on Fig. 17.6.2E; it is basically similar to that for a double-wound transformer. One
of the differences between Fig. 17.6.2C and 17.6.2E is that the l.v. side in Fig.
17.6.2C is assumed to be connected to a resistance earthed system (e.g. 33 kV)
with only short connections between the transformer and the 1.v. circuit breaker
and hence the l.v. connection protection has been omitted. In Fig. 17.6.2E, the l.v.
side is assumed to be solidly earthed (e.g. 275 or 132 kV)and there is separate
protection for the l.v. connections.
The tertiary winding on an autotransformer may be used to provide, via a
h.v./l.v, transformer, the l.v.a.c, supplies in a substation; it is also frequently the
means adopted of coupling reactive compensation plant into the transmission sys-
tem. A typical scheme for-protecting such a primary circuit arrangement is shown
in Fig. 17.6.2F. It should be noted that the tertiary winding, when not used for 1.v.
a.c. supplies or reactive compensation coupling, often has one corner of its winding
solidly connected to the neutral of the autotransformer to ensure that it is properly
earthed and also included in one of the autotransformer differential protection
 The application o f protection to transmission systems 371

/
1
t I

]
~--T I
m

. . . . . . . C
I
I m
C,K
I
I
I m !

I
I
I
i
I
I
I
I
I
i
I
I
I
I
I
I
• -=- ==-- ~ -~IP

I
I
D
i

, 1 I
l I
l I
L. . . . . . 1

equipment
I~1 I,'
'

I I
I t

I
~"

-
___,~
U
[_~ .... ~ __~- --~.
I _
"1___~ .

[1

Fig. 17.6.2E Protection scheme for an autotransformer

 The application of protection to transmission systems 373

zones, (Fig. 17.6.2D). When the tertiary is connected to reactive compensation

equipment and is earthed as shown in Fig. 17.6.2F, the earth connection on one
corner of the delta tertiary must be removed to avoid multiple earthing, and the
earth-fault protection for the tertiary winding is then provided by the differential
earth-fault protection energised from the current transformers in the h.v. neutral
of the earthing/l.v.a.c, supply transformer and in the bushings of the circuit
breaker controlling the reactive compensation plant. Reliance is placed on Buchholz
and back-up overcurrent protection for the very remote risk of a phase fault in the
tertiary winding not involving earth.
It should be noted that, for economic reasons, there is no circuit breaker on the
l.v. side of the l.v.a.c, supply transformer. The l.v. winding and connections of this
transformer are protected by differential earth-fault protection, but fuses provide
phase and earth-fault protection beyond the disconnector.

17.6.3 Protection for banked transformers

The protection requirements for banked transformers are influenced by:

(a) the number and types of windings on the transformers to be banked, for
example star/delta/star, interconnected star, or auto
(b) whether the transformation ratio of each transformer in the bank is the same
(c) the relative ratings and impedances of the two transformers
(d) the extent to which each transformer is required, to have its own discrimi-
native protection for operational and maintenance reasons.
If transformers of similar types are to be banked, and other factors permit, it may
be permissible for a single transformer overall protection system to be shared
between more than one transformer (Fig. 17.6.3A).

,,, !
Overall - I
/ differential ~1

Trans.
A
l"t
~
Trans.
B

[1
F ig. 17.6.3A Common overall differential protective system for banked transformers
374 The application of protection to transmission systems

Care must be taken, however, to ensure, especially with biased transformer

overall differential protection systems, that not too many current transformers
are connected into the differential system, otherwise the c.t. magnetising currents
may lead t o an unacceptable increase in the setting of the protection. In general,
for biased overall differential protection systems not more than two power trans-
formers can be protected by a single system. It must also be borne in mind that
with biased systems the bias is usually obtained solely from current flowing
between the h.v. and l.v. current transformers, and if it is possible for high through-
fault currents to flow between the l.v. current transformers, no bias will be available
to ensure stability of the protection. In such an application it may be necessary to
derive an additional bias feature from the l.v.c.t, currents.
If the banked transformers are of differing ratio it may be possible to employ a
common overall differential system, but the magnetising currents of the additional
main and interposing c.t.s, if any, may adversely affect the sensitivity and stability
of the protection to an unacceptable extent. If a small transformer is banked with a
much larger transformer, it is essential that a separate differential scheme be pro-
vided for the smaller transformer in order to obtain adequate settings for faults on
the l.v. side of the smaller transformer. Alternatively, if the transformer is relatively
very small indeed, for example, a 0.3 MVA 33/0.415 kV auxiliary transformer
associated with a 90 MVA, 132/33 kV transformer, high breaking capacity fuses
mounted as close as possible to the transformer 415 V winding would usually
suffice. Faults in the auxiliary transformer which produce fault currents less than
the setting of the overall differential protection will be detected by its Buchholz
protection. However, for most transmission applications fully discriminative pro-
tection for each transformer and its associated connections is an operational
requirement, any common connections having their own protection; an example is
shown in Fig. 17.6.3B. It should be noted that very few changes are necessary to
convert the protection arrangements shown for single transformers (Figs. 17.6.2C
and 17.6 2E) into the arrangements for a banked pair.
Disconnector auxiliary switches are shown in the differential protection bus-
wiring to enable the h.v. connection protection which is common to both trans-
formers to remain safely in service when a transformer is out of service for
maintenance and its h.v. disconnector is open. A separate differential protection
relay energised from c.t.s in the transformer h.v. bushing which remains in service
when the disconnector is open, may be provided, to protect the h.v. connections
between the open disconnector and the transformer h.v. bushings, thus allowing the
transformer to be energised from the l.v. side only. This mode of operation may be
needed for example where reactive compensation plant is connected to the trans-
former tertiary windings and it is desired to retain this compensation plant in
service when the transformer h.v. disconnector is open.
If certain limitations are acceptable an arrangement such as is shown in Fig.
17.6.3C may be permissible, in which there is a common protection system which
initiates tripping, and tank earth-fault alarms indicate which of the transformers is
faulty. The setting of the tank earth-fault indication relay has to be high enough to
 The application of protection to transmission systems 375

I
/
it
I
/
l

p:cF~
I
I CK

-I
I ,-- BB !
I
I
[ 7
I
I
/ I
i
I [ I r - - J i)B
/
| 6 '

> ___l [ A
~---~'4 t (l),- RE'S
.v 1
I

~- -"~i'°<!,,.s
II
~ "
. . . .
_
1
I
I
, , ,ocll,oc!
I
I
2S
!
-
HS

To c o m p e n s a t i o n
~---~-,~---r~
_._~ I
. . . .

~
J
I-" "-- - -
I
~

'~1" I
I
~) '
I
,E

I , r--T---¢ i ~__, k-
.v
RES

• It E~s
I
I ---- SB
I - LTI

I ------,[ BB
I ~"
i.
----J . . . .
I

, i_
I

[
CK
H CK

2oc
!
1 I . . . . - - - - ' -I1
I
BB

I ] i
El
-
m
' [~_cJ
I
J

Fig. 17.6.3B Protectionscheme for banked transformers

 376 The application of protection to transmission systems

/ Overall
differential
Trans.A i Trans
B

I li~-

ii
II
! _
II

Fig. 17.6.3C Indication of faulty transformer by tank earth-fault relays

ensure that transient currents which may flow through the capacitance of the
transformer windings under external fault or switching conditions are insufficient
to operate the relay. A typical setting used for a 275/132 kV 180 MVA transformer
was 200 A. It should be noted that this arrangement does not permit discriminative
indication of faults on the transformer connections and hence would not normally
be acceptable for controlling automatic isolation of the faulted transformer. For
this, and other maintenance and construction reasons this arrangement is now very
rarely used. The arrangement may be acceptable if operating conditions permit a
visual inspection of the transformer and its connections to be made prior to
reclosure.

17.6.4 Protection for a teed feeder

Some of the most difficult problems the protection engineer has to solve are
concerned with the protection of teed feeders. It is often very difficult to obtain
fault-clearance times, sensitivity and stability comparable to those obtainable for
two-ended feeders, and in rare cases it may even be impossible to provide a satis-
factory protection scheme at all, thus necessitating a change in the design of the
power system.
For many applications e.g. relatively short teed feeders, some form of differential
or phase comparison protection system is preferred, using private pilots, power.line.
carrier or microwave communication links. In present practice fast (under 40 ms)
differential protection systems require a microwave link or similar wide bandwidth
 The application of protection to transmission systems 377

channel. Slower biased differential systems use private pilot wire circuits. In practice,
in order to protect teed feeders especially if each leg is under 5 km in length, it is
often necessary to split up the teed feeder into two or more protection zones by
inserting current transformers at the tee point as illustrated in Fig. 17.6.4A.
In the scheme shown, there are three separate differential protection zones,
and when a fault occurs in one of the zones the protection for that zone operates
and in addition the operated protection is arranged to unstabilise the protection
for the other two zones, thus initiating tripping at all three ends. Also, operation of
any one protection system is usually arranged to initiate intertripping signals to
all ends as the protection at the other ends may not operate even if unstabilised.
Also direct intertripping of the circuit breaker is often a faster means of clearing the
fault than waiting for operation of a local end o f a protection system which has
been unstabilised.
In addition to the relatively long fault clearance times obtained the scheme
shown in Fig. 17.6.4A is very costly, especially if overhead lines have to be termi-
nated in a compound solely to enable post-type current transformers to be installed
at ground level.

1 B

'i"i'
I
I, ,,otl
i
C

Fig. 17.6.4A Protection o f teed feeder with two sets o f current transformers at tee~ooint

One method of avoiding the need to install current transformers at the tee-point
is to install a power directional comparison system in which the direction of power
flow at each end of the protected feeder is measured and tripping is inhibited at all
ends if the power flows out of the feeder at any one end (Chapter 10).
This system has occasionally been successfully applied but the following con-
ditions must be met:

(a) The system configurations and fault infeeds must be such that for all internal
fault conditions fault current can never flow out of the protected feeder (Fig.
378 The application of protection to transmission systems

-12,
Blocking signal
~=~ sent to B and C A

I --'--
I
I

Fig. 17.6.4B Inhibition o f tripping bM false block

A 2 1 B

i ' '

l i .... /

v v

Fig. 17.6.4C Fault current distribution

17.6.4B), otherwise the protection at the end where the fault current is
flowing out of the feeder will send a blocking signal to the other two ends
and thus inhibit tripping.
(b) If overcurrent starting is used for l~he protection, the margin between the
settings of the high set and low set relays must be large enough to ensure that,
for an external fault, a high-set relay at one end cannot operate before a
low-set relay at another end has operated, otherwise the protection will be
unstable. If the through-fault current distribution is as shown in Fig. 17.6.4C,
it will be seen that the high-set relay setting must be at least twice the low-set
relay setting to ensure that a low-set relay at Station B or C operates before
 The application o f protection to transmission systems 379

the high-set relay at Station A. To ensure that the low-set relays do not
operate with normal load currents, a minimum fault setting of at least 200%
of the maximum normal load current is necessary.

Often a more satisfactory form of protection for a teed feeder, particularly for a
longer teed feeder, is accelerated or blocked distance protection. For diversity
reasons one main protection should preferably be of the blocked type and the other
accelerated, although it should be noted some modern systems use a mixture of
acceleration and blocking for different measuring elements in the same distance
protection.
There are very many factors to be considered in choosing the design and settings
of distance protection for teed circuits.
One of the most important is the absolute and relative lengths, and hence
impedances, of each leg of the teed circuit. The primary circuit configuration
shown in Fig. 17.6.4D is for example much easier to protect than that shown in
Fig. 17.6.4E as in the latter case the distance protections at stations A and C cannot
be set to operate for faults near station B without responding to faults in or beyond
stations C and A respectively.
Another important factor in choosing suitable forms of distance protection are
the magnitudes and variations in the fault current infeeds at each end of the teed
circuit. One aspect of this is illustrated in Fig. 17.6.4F. Assuming a fault occurs
near station C, it is apparent that the distance protection at station A will see this
fault further away than it actually is owing to the voltage being injected at the tee
point from station B. Similarly, the distance protection at station B will see the

A B

I
I
I
I

Fig. 17.6.4D Symmetricalteed circuit

380 The application of protection to transmission systems

A B

-D i • - D--
I
I
I
I
I

ii

F ig. 17.6.4E Teed c i r c u i t w i t h unequal legs

A B

. . . . . . - - - -

I
I
I

Fig. 17.6.4F Fault on a teed feeder

fault further away owing to the fault current infeed from station A. The relays at
stations A and B thus 'underreach' and, depending on the settings which can be
given to these relays, they may or may not operate in their second or third zones.
It is therefore frequently necessary to arrange on such a teed circuit that when the
protection at station B operates it initiates the sending of intertripping signals to
stations A and C. Incidentally, in general it is usually prudent even where it is not
absolutely essential to arrange that the operation of the protection at any one
station initiates the intertripping of the circuit breakers at all other stations on a
teed circuit. Some of the parameters which have to be calculated and checked for a
typical accelerated distance protection system applied to a teed circuit are:

(a) maximum permissible zone 1 settings

(b) minimum attainable zone 1 settings
 The application of protection to transmission systems 381

(c) with all circuit breakers closed effective reach of relays at each end as a result
of remote infeeds
(a) with one circuit breaker open, effective reach of relays
(e) with two circuit breakers open, effective reach of relays
09 minimum zone 3 settings
~) load transfer limitations.

The major underreach problem which occurs with an accelerated distance protection
system can be reduced by using a blocked form of distance protection system for
which the reach of the tripping zone can be appreciably increased.
However, choosing appropriate settings for the relays performing the blocking
function can be difficult. Not only do the magnitudes of the phase and earth fault
current infeeds affect the choice of setting but also the direction of the currents
can affect operation as shown in Fig. 17.6.4B.
Some of the calculations and checks that have to be made for a typical blocked
distance protection application are:

(a) Check that the minimum setting of the trip feature which is available after
blocking has been removed reaches beyond the remote circuit end under the
worst infeed conditions.
(b) Calculate the minimum setting for the block-remove feature.
(c) Check that the trip feature after blocking has been removed or the block-
remove feature do not reach beyond the reverse reach of the remote-end
block initiate feature.
(a) Check that the block-remove feature does not have a setting that will cause
operation on reverse earth faults.
(e) Establish the load transfer limitations.

With some designs of complex blocked distance protection systems an assessment

of its performance for varying fault locations and differing types of fault and
system conditions can be very time consuming and a difficult task requiring
extensive computation.~:In some cases in fact, the design of the protection system
has to be tailored, with the aid of the manufacturer, to make it suitable for a
particular teed circuit configuration, assuming certain maximum and minimum
infeed conditions at" each end of the teed feeder concerned and certain assumed
ranges of positive, negative- and zero-sequence impedances for that part of the
primary network which is effectively in parallel with the teed feeder to be
protected.
After completing the calculations, checks or tests, it may well be found that the
proposed teed circuit cannot be protected in a satisfactory manner to meet all
operational conditions, load transfer restraints being a particularly common unac-
ceptable limitation. The situation can be improved appreciably by inserting current
transformers at the tee point as shown in Fig. 17.6.4A. A cheaper arrangement, by
which acceptable performance may still be obtained is to insert post-type current
382 The application of protection to transmission systems

D I
-..t21 'f:,, i i,.v.,
ii
I

[]
_ ii C

F ig. 17.6.4G Protection o f teed feeder with extra current transformers in one leg

I I
--B _., &,? I

I
I ,/ I
I DA I
I

L ..- .m.
I I
I

Fig. 17.6.4H Protection o f teed feeder with individual protection for each leg
transformers in only one leg of the tee as shown in Fig. 17.6.4G; in this case it
may be technically better and more cost effective to protect part of the teed feeder
by a diffecential or phase-comparison protection.
If there is a requirement for automatic operation of disconnectors at the tee
point to permit automatic restoration of the healthy part of a teed circuit, it is
essential that each leg should be separately protected to enable the faulted leg to
be identified and isolated as shown in Fig. 17.6.4H.
In spite of all the techniques mentioned above there are some arrangements of
teed circuits which cannot be satisfactorily protected and the only option left is
to rearrange the primary system.
 The application of protection to transmission systems 383

17.6.5 Protection for a transformer-feeder

For very short transformer-feeders, less than approximately 2 km long, it may be
technically possible, current transformer lead burdens permitting, to include the
feeder in the associated transformer differential protection zone in which case the
protection could be identical to that for a transformer. However, in order that a
transformer-feeder may be switched back into service as soon as possible after a
fault, it is necessary, for major transmission system applications, for the feeder and
the associated transformer each to have its own individual discriminative protection.
The feeder section is in general protected in the same manner as described in
Section 17.6.1, two main protection systems being provided for those transformer-
feeders which form an important part of the transmission system.
The relatively high impedance of the transformer does, however, appreciably
affect the application of any distance protectioh. The application of distance
protection is eased where it can be set to look into the transformer winding, i.e. the
impedance of the transformer can be used to enable the zone 1 reach setting to be
longer than the feeder length, thereby permitting fast operation of the protection
for all feeder faults between the transformer h.v. terminals and the h.v. circuit
breaker. On the other hand, this advantage does not apply to any distance
protection installed at the transformer end of the transformer feeder. Furthermore,
the impedance of the transformer limits through fault currents to relatively low
magnitudes which necessitates low settings for any blocking relays. The transformer
is protected in the manner described in Section 17.6.2.
The transformer protection must trip both the h.v. and I.e. circuit breakers.
Also the feeder protection at the transformer end may be very sluggish or even fail
to operate if, as is often the case, the fault current infeed through the transformer
is low, and hence two-way intertripping is normally required (Fig. 17.6.5 A).
For very important transmission circuits, two independent intertripping systems
are provided, normally employing voice frequency signalling over private
or telecommunication company pilots. For dependability reasons, the

I.... INT
l---q
! INT ]~-------i-..
_. ]
I ~ I
I I I
I

'
I
I I 1I.'e ,,.. ..... -,=. - --~
'
I

'i' "' =i'-P

A
I
3oc .j
Lt i i
F '
Fig. 17.6.5A Protection for short transformer feeder
384 The application o f protection to transmission systems

/
protection

. . . . . lu ,. ,i i i

IsM I"
I"
_I----~"
'BB I -I_____1"
I
I
I

I ,o,,fl ,o~1

' I
I
I
~F----, ~-
I /
I
.j-- --~
T

LVC
t CK CK El !
t
. . . .

2
[]

Fig. 17.6.5B Protection for transformer feeder forming part of a major transmission system
 The application o f protection to transmission systems 385

intertripping and protection signalling equipments and communication channels

should be as different as practicable. In addition, the intertripping is backed up by
unstabilisation of any differential, phase-comparison or blocked distance feeder
protection, and acceleration of accelerated distance protection. For less important
transmission circuits, in view of the latter back up facilities, only one intertripping
system may be provided. The reason for the provision of this considerable amount
of redundancy is to reduce to a minimum, within economic constraints, the risk of
an uncleared, or slowly cleared fault on the transformer or its associated l.v. con-
nections. Severe and costly damage, particularly by fire, can result from slowly
cleared faults on the transformer or its l.v. cables; this may result in a very long
outage time before the transformer.feeder can be restored to operational service,
with a consequent possible risk to continuity of supply to consumers.
A typical scheme for protecting a transformer-feeder forming part of a major
transmission system is shown in Fig. 17.6.5 B.
The protection arrangements for transformer-feeders not forming part of a major
transmission system are described in Section 17.7.

17.6.6 Protection for a double-busbar station

Although the risk of a busbar fault occurring at most major transmission stations is

A ii

,EF--o -

. ,,,.,.

O O

nC+) (~12
Fig. 17.6.6A Simplified arrangement for busbar protection at a major transmission station
386 The application o f protection to transmission systems

i
/
' /i i
/
'i/ /
I

- t~°~1 i~1
,. ~
/

v
1 i
II [3 [] I~
~-t ~c~i ~: ~ "@ ~-i ~ !
,-t
_ _~o:,,~ "4'°"~ _ '°'~I

/ / / /

Banked
1 Teed
t t
Feeder
t
Transformer
transformer feeder

High speed distance

(blocked Mho) protection

Fig. 17,6,6B Typical protection for part of a double busbar station

 The application of protection to transmission systems 387

extremely low, perhaps once in the life of the station, the consequences to the
system of a slowly cleared busbar fault are so severe that the expenditure incurred
in providing costly and complex busbar protection can be fully justified. A similar
justification can be made for the provision of circuit-breaker fail protection at
major transmission stations.
It is therefore usual practice at 400 and 275 kV stations to provide the busbar
protection and breaker fail protection which is described in detail in Chapter 13.
It should be noted that at an important transmission station the design of the
busbar protection should be such that the failure of any one relay to operate should
not result in an uncleared busbar fault. It is therefore usual to provide some
measure of redundancy in the 'two out of two' relay logic, e.g. extra per zone relays
may be connected in parallel with both the check and discriminating relays as
illustrated in Fig. 17.6.6A.
A typical example of the protection arrangements at a 400 kV double busbar
station are shown on Fig. 17.6.6B.

17.6.7 Protection for mesh stations

The protection arrangements for a typical mesh type station are shown in Fig.
17.6.7A. The main differences between the protection arrangements at a mesh
station compared to those at a double-busbar station concern the protection pro-
vided for the busbars. At a mesh station the number of circuit breakers required to
be tripped in the event of a busbar fault is much smaller than at a double-busbar
station and the operational consequences of an inadvertent operation of the busbar
protection are very much less severe; hence there is no need for a 'two out of two'
logic in the tripping sequence. The busbar protection at a mesh station is normally
of the high impedance differential type and is arranged in a 'one out of two' logic.
Auxiliary switches on the feeder disconnector are arranged to disconnect the
feeder current transformers and associated differential relay from the mesh corner
protection buswiring when the feeder disconnector is open. Fast discriminative
protection for the connection between the open feeder disconnector and the
current transformer housing is thus provided when the feeder is energised or left in
service from the remote end, e.g. for a teed feeder circuit.

17.6.8 Protection for complex primary circuit configurations

With the high cost of switchgear there has been an increasing tendency to design
primary systems with the minimum number of circuit breakers, e.g. by banking and
tee-connecting circuits. A relatively simple example is shown in Fig. 17.6.8A. It
will be noted in this case that only a few changes are necessary to the protection
arrangements shown in Fig. 17.6.3B. Nevertheless, for some primary circuit con.
figurations the protection, and also the automatic switching arrangements can
become very complex. An example is shown in Fig. 17.6.8B and the associated
 388 The application o f protection to transmission systems

-~- [ ~ " Zone-3- i

_ protection BF BI."

I ~_,..~" 1

, ,[] ,
'_~ , I~
,~- a ICK! -
I •
I I I I

, ,F~ , , - ,
'II, ~~ , ISM , ~ '~r,~l II' ~i l
CK !t ~
I"eeder '
--~" I I"I
~
I'I I "I
zone l
I "I L__II
Zone
X__
- #~- to I mI
, r-
'
!
" F--T--" I

I
I
,
[t'~" I
' --
'41
I ]
I
RES
HV

• I m t-12s I
--~ :I :L4-J
I l ,ii |3OCI
To c o m p e n s a t i o n T I _ T / ]
equipment ,t

--T-- , _ ,~L2y_I
' " I---7 "--~~ ' E~s
I - I LTI
I ~ "I
I ---- i
, I~F-- -J ~
t ' - I
I ~ . . . . "" I

[,~vc_ _ _,..' - . " ~ t cK .... ~ B8

r] ~L_EU '

'I-- .t3°c[
!
I
I
.I

Fig. 17.6.7A Protection at a mesh station

 The application of protection to transmission systems 389

"-+
System backup
protection

.4

I . . . . C
CK
, -. ~
I
I .r- . . . .
1,- --'--
, .,(, ', l
I , ~ . . . . .., I
I

, + -I i, ,,.:~ "'
= |.

<;- L. i_. _IHVCi

-:.'- i 141
"l"<Jcompensation ~N

,,~ t I'

l(-'v-i'-I ~ ,
'L_~" - -?- -'It_ ~ L
U
,, ~-____i~

I Bit,

t .L I...... B,."' I
' --- " I-.- ---I-~
1 ------,---:-~ i
[] }

l ,_+,+, l I" -'

Fig. 17.6.8A Protection for banked transformers and teed feeder
390 The application of protection to transmission systems

Station C

400 kV

F o r p r o t e c t i o n details []
see schedule
for Fig. 17.6.8B

9 k m to tee point

Station A

275 kV

To station E
275 kV

! _l_ 26 k m f r o m
station B
I
tee point

I
Station B /
1
L_. 12 km

50 MVA 2 x 750 MVA

12 km

4 0 k m t o tee p o i n t
I

240 MVA
To station i."
132 kV

Station D

- ,
400 kV

.... 400 kV circuits

"---- " "----275 kV circuits

Fig. 1 7 . 6 . 8 B Complexprimary circuit configuration

 The application of protection to transmission systems 391

Protection and automatic switching scheme for feeders shown in Fig. 17.6.8B

Station Voltage Circuit Protection

A 275 To stations C & D 1st & 2nd Main: between stations A and
via station B B differential pilot wire feeder protection
of two different types

A 400 To stations C & D 1st & 2nd Main: accelerated mho distance
via 275/400 kV protection of two different types
transformer at
station A

B 275 To station A 1st & 2nd Main: differential private pilot

wire feeder protection o f two different
types

B 400 To stations C & D I st & 2nd Main: accelerated mho distance

protection of two different types

C 400 To stations A & D 1st Main: mho blocked distance protec-

tion with PO signalling system to D
2nd Main: accelerated mho distance
protection

C 400 To stations B & D 1st Main: mho blocked distance protec-

tion with PO signalling system to D
2nd Main: accelerated mho distance
protection

D 400 To stations B & C 1st Main: mho blocked distance protec-

tion with PO signalling system to C
2nd Main: accelerated mho distance
protection

D 400 To stations A & C I St Main: mho blocked distance protec-

tion with v.h.s. PO signalling system to C
2nd Main: accelerated mho distance
protection

Notes
1. Stations C and D are major power stations whereas station A is a relatively weak infeed.
The distance protection reaches from C and D are limited by the transformers at A and B
and hence, there is no need to block the blocked mho distance protection from stations A
and B. Furthermore, it was found that reliable blocking could not be initiated from
stations A and B, if the circuit breakers at C or D were open.
2. lntertripping is provided between all stations using a mixture of private and post office
signalling channels diversely routed and cascaded where necessary. Operation of any
protection intiates intertripping of all circuit breakers on circuit concerned.
3. Automatic control facilities to avoid switching overvoitages and the effect of ferrores-
onance are arranged as follows:
(a} A-B-D circuits
Following the opening of the circuit breakers, the 400 kV switch disconnectors at B are
opened automatically. The autoreclose sequence is as follows:
(i) close D 400 kV circuit breaker to charge line
(ii) close B 400 kV switch disconnectors to charge transformers and A cables
(iii) check synchronise then close C400 kV circuit breaker and A 275 kV circuit breakers.
(b} A-C-D circuits
Following the opening of the circuit breakers, the 400 kV switch disconnectors at A and D
are opened automatically. The D switch disconnector may then be reclosed immediately.
The autoreclose sequence is as follows:
(i) close D 400 kV circuit breaker to charge line and D transformer
(ii) close A switch disconnector to energise transformer
(iii) check synchronise, then close C 400 kV circuit breaker, A 275 kV circuit breaker
and D 132 kV circuit breaker.
The above closing sequences are also followed for manual operation.
392 The app/ication of protection to transmission systems

schedule gives detailed protection arrangements for this particular primary system
configuration.
With such complex primary circuits it is practically inevitable that, for protec-
tion and other reasons, there are operational limitations, e.g. some circuit breakers
must not be left closed if certain other circuit breakers are open.
Also, the protection arrangements for such complex circuits inevitably result in
longer fault clearance times than are attainable for simpler circuits and these longer
times may not be acceptable as adjacent parts of the system are developed and
faster fault clearance times are required. These complex configurations may, further-
more, necessitate complex logic for the automatic switching equipments to avoid
switching overvoltages which might result from energising particular combinations
of overhead lines and transformers.
Nevertheless, especially for transitional schemes, there are occasions when these
complex primary circuits and their associated complex protection and automatic
switching arrangements can be justified on economic grounds.

17.7 Typical protection applications in a minor transmission system or major

distribution system

For less important transmission systems and major distribution systems where
longer fault clearance times are permissible and where the dependability require-
ments for the protection are less severe major economies can justifiably be made in
many cases by installing less elaborate protection schemes than those described in
Section 17.6. On the other hand, in the growth of a power network in a country it
is highly probable that a system which starts as a transmission system after several
years or possibly decades will become part of a distribution system, and a cost-
benefit analysis has then to be made of whether or not it is worthwhile to replace
existing protection schemes on a part of the network by simpler schemes. Major
factors in this analysis are the availability and cost of spares for the old protection
and cost of maintaining it.

17.7.1 Protection for a feeder

In most applications, two main protection systems cannot be justified, and one
main protection and overcurrent back-up protection is usually adequate. Similarly,
duplication of protection signalling systems is rarely applicable; and for some
feeders distance protection of the switched type without acceleration or blocking
may well suffice. The same factors apply in choosing a suitable protection system
as are described in Section 17.6.1 but for the majority of applications the cost of
power-line-carrier protection may well be prohibitive, depending on the cost of the
coupling equipment.
 The application of protection to transmission systems 393

17.7.2 Protection for a transformer

The protection scheme is often basically the same as described in Section 17.6.2,
but there may not be sufficient justification for providing overall differential
protection in all applications; h.v. and 1.v. differential earth fault protection, h.v.
overcurrent protection, and l.v. overcurrent protection, often directional, may
suffice.

17.7.3 Protection for banked transformers and dual secondary transformers

Some of the schemes described in the first part of Section 17.6.3 may be acceptable,
using for example a common overall differential system for both transformers in a
bank. In some substations dual secondary transformers, e.g. 132/11/11 kV may be
installed and a typical protection scheme for such a transformer is shown at the
transformer end of the transformer-feeder in Fig. 17.7.5B. It should be noted that
in this scheme instead of attempting to provide transformer h.v. overcurrent pro-
tection at the transformer end, which would introduce discrimination problems, a
relay is connected differentially with h.v. current transformers and current trans-
formers in the 1.v. circuit breakers to provide a plain balance scheme covering both
transformers.

17.7.4 Protection for a teed feeder

Apart from there being no need for duplication of the main protection, the same
considerations apply as described in Section 17.6.4. However, longer fault clearance
times are often admissible in minor transmission systems and major distribution
systems which, in general, eases the problem of protecting teed feeders.

17.7.5 Protection for a transformer feeder

Major economies can sometimes be made in the protection of transformer feeders

compared to the schemes described in Section 17.6.5; longer fault clearance times
are often permissible for feeder faults, and the cascade tripping situation illustrated
in Fig. 17.7.5A may be acceptable. In this example, it is assumed that the limited
fault current infeed at the transformer end is initially insufficient to operate the
unit type feeder protection at that end. After the feeder protection at end A has
operated and the circuit breaker Q has tripped, a higher voltage is available at
station A busbars to drive an increased fault current through the transformers at
station B, thus enabling the feeder protection at the transformer end to operate and
initiate tripping of circuit breaker R. If the fault current infeed at station B for
feeder faults is insufficient to operate the protection at that end and provided that
the transformer feeder is unlikely to be energised or left energised from its l.v. side
with the h.v. circuit breaker open intertripping can be provided to trip the I.v.
circuit breakers. Alternatively, provided the fault current infeed at the h.v. end is
394 The application of protection to transmission systems

Station ~ I-'P I Station

R
- - -

"

l.'ault current prior to circuit breaker Q tripping

Fault current subsequent to circuit breaker Q tripping
Fig. 17.7.5A Sequentialtripping

high enough, some forms of differential and phase comparison feeder protection
have an inherent intertripping facility which may be sufficiently fast. If the trans-
former feeder is protected by distance protection at its h.v. end and the h.v. system
is solidly earthed directional overcurrent protection on the 1.v. side of the
transformer can often provide adequate protection for feeder faults, thus saving, at
the transformer end, the cost of h.v. voltage transformers, current transformers and
a set of distance protection (see Fig. 17.7.5B).
The directional overcurrent protection must be set to discriminate between
faults on the h.v. system and faults on the transformer feeder (see Fig. 17.7.5C).
If fault current infeed through the transformer feeder 1.v. circuit breaker is
significant for faults occurring beyond the h.v. circuit breaker, either the directional
overcurrent protection must have an operating time sufficiently long to obtain dis-
criminative operation, or the directional overcurrent relay must be interlocked with
the distance protection at the h.v. circuit breaker end via a protection signalling
channel so that it cannot operate unless the distance protection has operated. This
is referred to as inhibited directional overcurrent protection.
If the fault current infeeds at the h.v. end of the transformer feeder are always
expected to remain within a restricted range, very simple and relatively cheap high
set overcurrent and restricted earth-fault protection may take the place of the h.v.
end distance protection as shown in Fig. 17.7.5D. (See Chapter 12, Section
12.9.2(d)). By choosing suitable settings for El, E2 and the h.s.o.c, relays operation
of the protection can be restricted to faults occurring between the feeder protection
current transformers and the transformer h.v. neutral connection.
The high-set overcurrent must be set so as not to operate for faults on the l.v.
side of the transformer. The star/delta interposing current transformer eliminates
zero sequence currents from the inputs to the E2 relays; the setting of these E2
relays must however take into account, for faults on the h.v. system beyond the
 The application of protection to transmission systems 395

132 kV

" i.-J
-u
-D
-%
/
"t
~ mm~
t:' l I
I
I I
I I
I I
I 'I
I I
I
I' I
I
I
I
I
I
-- i -I.

I
I
I
----, 41k-

i I ":" I
r I I I
I I I
,,_,, .
I
I
,,
', Ill
~---ff-~ ,, ,~-~ _; I
"
L
j ,.- -.,.
L_~ L. . . . . J

r i
Fig. 17.7.5B Protection scheme for transformer feeder with 132/11/11kV dual secondary
transformer
396 The application o f p r o t e c t i o n to transmission systems

1--O------ ||,"

I,o.~
Fig. 17.7.5C Directional overcurrent protection on a transformer feeder

132 kV
busbars

[HS l

J l Hs I
[-g]___[oc I

El E2
Trip

: 0 0 -

0 i
0"----'-'--

Fig. 17.7.5D High-set overcurrent with restricted earth-fault protection

 The application of protection to transmission systems 397

feeder protection c.t.s, any negative sequence current backfeed through the power
transformer, including the backfeed from stored energy in any rotating plant on the
l.v. network fed by the power transformer. Where necessary, the setting of both E1
and E2 must also take into account the current distribution when parallel trans-
former feeders supply a common l.v. busbar. The E l , E2 and h.s.o.c, relays should
all be designed to be relatively immune to the effects of d.c. transients to avoid
unwanted operation due to any d.c. component in the fault current and trans-
former magnetising inrush currents. To back up the intertripping on a transformer-
feeder, it is often the practice for faults on the transformer or its associated l.v.
connections to unstabilise any feeder unit protection or accelerate the h.v. end
distance protection.
In some cases, to back up the intertripping, the transformer h.v. disconnector
may be opened automatically after a suitable delay. In the event of the other means
of tripping the remote circuit breaker having failed if the current is very low, e.g.
from a Buchholz operation, the h.v. disconnector may successfully break the
current. If it fails to break the current a local fault will be produced by the arcing
at the disconnector which will result in operation of the feeder protection at the
h.v. end. The philosophy of this technique is that the consequences of an uncleared
fault on the transformer or its associated l.v. connections may be worse (e.g. a
transformer fire) than if an attempt were not made to open the disconnector.
For many applications, particularly for long transformer-feeders, instead of
providing hired or private pilot intertripping it may be cheaper and more dependable
to install a fault-thrower. At 132 kV, the present design of fault throwers is limited
to a maximum fault level of 2500 MVA. If fault throwers are installed, directional
overcurrent protection is normally used to trip the l.v. circuit breaker for feeder
faults.

17.7.6 Protection for a double busbar and mesh station

The protection for a double busbar and mesh station is usually identical to that
described in Sections 17.6.6 and 17.6.7 except that the amount of redundancy can
be reduced, e.g. there is in general no need for more than one discriminating zone
relay and check zone relay to initiate tripping at a double busbar station, i.e. it can
be a simple 'two out of two' relay logic scheme with no redundancy. For mesh
stations a 'one out of one' relay logic scheme is normally adequate.

17.7.7 Protection for complex primary circuit configurations

The same considerations which are mentioned in Section 17.6.8 apply for choosing
a protection scheme, but the longer permissible fault clearance times ease the
problem. Also the need for high emergency load transfer capability is usually not as
severe a problem as on a major transmission system.
398 The application of protection to transmission systems

17.8 Bibliography

General application

'Review of recent changes in protection requirements for the CEGB 400 and 275
kV systems' by J C Whittaker (lEE Conf. Publ. 125, 1975, pp.27-33)
'Recent developments in CEGB protection policy and practice' by J C Whittaker
(lEE Conf. Publ. 185, 1980, Discussion Record, pp.l-4)
'Cost of power outages - the 1977 New York City Blackout' by W T Miles et al.
(IEEE Industry Applications Society Annual Meeting, Seattle, 1979, pp.65-69)

Feeder protection

'Ultra high speed relay for EHV/UHV lines based on directional wave detection
principles' by R P Carter (lEE Conf. Publ. 185, 1980, pp.166-170)
'The application of phase comparison protection to EHV transmission lines' by
G Fielding et al. (UMIST Proc. of Internat. Conf. on Feeder Protection, 1979)
'Performance of distance fault detector relays' by W P Lewis (lEE Colloq. Dig.
1968/19, pp239-249)
'Analysis of complex distance-relay characteristics taking load into account' by
L P Cavero (lEE Conf. Publ. 185, 1980, pp.192-194)
'High speed protection for transformer feeders without pilot wires' by J Rushton
(Electr. Energy, April 1958, p.132)
'Protection aspects of multi-terminal lines' (IEEE Special Publ. 79, TH0056-2-PWR)
'An evaluation of the comparative performance of distance and differential feeder
protection systems' by W D Humpage and J Rushton (lEE Colloq. Dig. 1968/19,
p.267)

Automatic switching

'A coherent scheme for the design and testing of automatic switching equipment'
by C R Seymour (lEE Conf. Publ. 185, 1980 pp.69-73)
 Chapter 18

Testing, commissioning and

management of protection
by E.C.Smith, revised by D.Hay

18.1 Introduction

This chapter deals with protective gear throughout its life, from its being specified
on a scheme, through the testing and commissioning stages into routine maintenance
and fault investigation. The management of protective gear covers all those aspects
and more; for example, keeping records, assessing performance, and the avoidance
of mistakes in testing.
The chapter does not attempt to describe in detail how to commission any
specific type of protection, but seeks to establish principles which apply to all
schemes. Similarly the recommendations relate to the tests which need to be done,
without specifying who shall do them.

18.2 Contractual obligations

There is little doubt that, where it is economically possible, it is better for the user
of the equipment to commission it himself; by so doing he has a keen interest in
the equipment right from the start and has the opportunity to get thoroughly
familiar with it before it goes into commission. This knowledge will stand him in
good stead during subsequent maintenance, and especially during fault investigation.
However, for various reasons it is customary to specify that the equipment shall
be 'put to work' by the contractor. This immediately introduces such difficulties as
'marrying' extensions to existing schemes, and connecting d.c. and a.c. supplies to
new equipment.
Such difficulties can arise even when one contractor is entirely responsible for
all the new equipment; but the difficulties are worsened when several contractors
are involved. For example, there may be different manufacturers for the h.v.
switchgear, the 1.v. switchgear, the transformers and the multicore cables. Each
contractor is required to prove his own equipment, but he cannot always do so
because of its interdependence with other equipment. During the commissioning
400 Testing, commissioning and management of protection

tests something may fail to function; the cause of failure may not lie with that
contractor's equipment but may be due to crossed multicores or wiring on associa-
ted equipment supplied on another contract. It is here that some co.ordinating
influence is necessary.
This co-ordination can best be done by the user or the consulting engineers.
Very often, in the interests of expediency, the user will investigate the trouble him-
self, but this is contractually wrong. If a contractor is responsible for setting his
equipment to work, the user should refrain from sorting out his troubles for him.
He should, as far as possible, require each contractor to prove his own equipment.
When several contractors are involved on a project it is advisable to have a
group discussion beforehand to outline the tests which have to be done and to
consider the effect of those tests on the equipment as a whole.
Although many tests may have to be done to prove a comprehensive scheme,
some are usually more vital than others. The user may not be able to be repre-
sented at all the tests; he will have to decide which he must do himself, which he
will participate in or closely supervise, which he will let the contractor do to a
previously drawn-up programme (inspecting the test results afterwards), and which,
if any, he will accept verbal assurance from the contractor as having been done by
him.
Whether there is close participation or whether there is almost casual supervision,
the user must remember one thing very clearly: he, and he alone, is responsible for
ensuring that the new equipment has been correctly installed and that it can safely
be connected to the system.
The contractor must satisfy the user of this. It is common practice, particularly
for major installations, for the Board's specifications to call for 'testing and setting
to work'; the specifications also require a 12-month maintenance period, during
which any faults caused by poor design or bad workmanship are put right free of
charge by the manufacturer.

18.3 The mental approach to commissioning tests

A fundamental thought should be kept in mind when approaching any commis-

sioning test: the test is being done to prove the equipment is correctly installed. It
may not be.
Throughout the tests, critical attention should be paid to all the aspects of the
work; nothing should be 'glossed over' without a full explanation being sought. It is
only by paying minute attention to detail that a high standard of performance can
be achieved.

18.4 Commissioning tests

18.4.1 Reasons for commissioning tests

Commissioning tests are done to prove:

 Testing, commissioning and management of protection 401

(a) That equipment has not been damaged in transit and that it can be safely and
confidently connected to the system.
(b) That the specified equipment has been correctly installed.
(c) To prove characteristics of the protection which are based on calculations, for
example the primary settings of high impedance protection systems.
(a) To obtain a set of test figures for future reference if and when necessary.

Commissioning tests, as well as covering protective gear, control and alarm circuits,
also cover much of the primary plant with which such equipment is associated.
Since, however, this book deals with power system protection, nothing more will be
said of primary plant commissioning except that, on any primary circuit being
commissioned, it is a good plan to consider it as being built of units such as busbars,
busbar protection, circuit breakers, transformers, cables, disconnectors, protective
gear, intertripping, and so on. Each such unit should be proved before attempting
conjunctive tests on the complete circuit.

18.4.2 Planning of commissioning tests

It is desirable to plan commissioning tests in a logical order, ensuring that no equip-

ment is overlooked and, once tested, is not unnecessarily disturbed, thus requiring
a re.test. A commissioning programme should be planned as follows:

Study of Diagrams

(a) Block diagrams: It is helpful to have (or to prepare if none is available)a

block diagram of a substation tO show the lines, transformers and busbars and the
c.t. and v.t. positions. The various types of protective gear may be indicated in
block outline with thin solid lines between them and the c.t.s, and dotted lines
between the relays and the circuit b~eakers they trip.
A block diagram of this kind (Fig. 18.4.2A) presents a very clear picture of the
protection and control layout arrangements of a substation.

(b) Circuit diagrams: The manufacturers will often supply circuit diagrams, but
if not they should be extracted from the main wiring diagrams. Separate circuit
diagrams can be prepared for d.c. tripping, closing, alarm and indication circuits;
and a.c. diagrams for various types of protection, voltage and synchronising circuits.
Each class of circuit can be shown on a separate diagram so that each can be studied
and tested without confusion. A typical circuit diagram is shown in Fig. 18.4.2B.
Two warnings are necessary with regard to the use of circuit diagrams; both
warnings involve completeness of the diagram. If an engineer is extracting his own
circuit diagram, he would be wise to make quite sure that he has included the whole
of the circuits connected, say, between any given fuse and link; by doing a full
diagram he can be confident that no possible 'sneak' circuits are overlooked. If
using a manufacturer's diagram, he should satisfy himself that it is complete and
 402 Testing,commissioning and management of protection

/
Po..Open ~ __ _..
--A---"
'-v
2 sec
F ADO -1
I
|I aux. sw. Carriage sw. I
,=~-- - ~ - _.- ,'
I
L.V. sw.
aux. breaker H.V.
disconn.
aux. sw.
~l .11
~oo,, E[E Stage 1 ~ I
soo,, E E
I I
I I
' - - - " ~"~~ . ~ J /0.577---" ¥ I
[> i~~f~ L~ I
90 MVA Main II I
132133 kV
I, ~. . . . ~- - II
I-~ __1

earthing I I
trans.

i .... TI

~ l,_~/I
E
II
'3 E---~
Area
Board .
/,..
,~._ c
:ooo/,
: 1" Stage 2 ~ iI
BIB ....

Metering =

Inst.&DO("
--

E
,, ~ t E ~°°°~' I
iii f d Stage I
"II
tT
Sll .

" ~ooo/,
= - :" - - f:- t ~°cl is
14
c,,n,pensati~;n-' C - - - <
E--3 E '
,~ I1_

i 33 kV

Fig. 18.4.2A Typicalblock diagram

 Testing, commissioning and management of protection 403

Circuit Tripping
protn, relay
,O ~ I
' ""
L-.-J
Trip coil
supervision II'A"

38.~0. ~2 I l
'Trip coil C.B. Trip
I / isol'n .' aux. s~. coil

] Sw'. gr. ON(." RS Sw} Igr. I

'1) + '
1 Trip coil "D '
o oh,, o - - o oJ
: ~'Remote Cl, Ip I supervision

I10 V d . c .
:o1:
i I
'°

LSC close I I 'Standby

~: .,77.O'~ I I rUllH)tu
LSC I
C(,'C LR Clll~ie- ,
..... ,ol-lt-'t , '
~,o.~ / r ' - - ¢ I It
'Standby ONC
,
I
,, S ~ I Iigr.
close +' II I I
, Ii
lit I
'Local ONC C.P. I
'Closing close +' I I I
'Closing
coil+' ~ Ii Ii~l 0
~ S w . gr.

S ~ . gr. I
110 Vd.c.
CCC C£'("

KEY

BBPTR - Busbar protection tripping relay

¢" - c l o s e LSC - limited selective control
L'C - c l o s i n g N - normal
C C C - closing coil contactor O - open
t'P - control panel R - remote
L - local S - standby

Fig. 18.4.2B Typical circuit breaker control diagram

that it shows every hazard in the circuit. For example, at one time o.c.b, rack-in
plug and socket connections were not always shown on tripping and closing
circuit diagrams. If circuit diagrams are being drawn by the user, he should make
them as simple as possible, ignoring the physical layout of the equipment and
making his diagram a pure electrical diagram, avoiding unnecessary crossovers and
ensuring a smooth flow of the eye from side to side or top to bottom without any
unnecessary 'to and fro' motion. Circuit diagrams of a semiphysical type, which are
drawn so that the number of multicores necessary between two points can be
readily assessed, are almost as difficult to follow as a wiring diagram, through
various contacts and coils, until the circuit from positive to negative is complete,
by which time (unless the engineer has an unusually retentive memory) a state of
partial confusion will result.
Good circuit diagrams are invaluable during commissioning tests and fault
investigations in that they permit the details of the circuit to be assimilated quickly.
404 Testing,commissioning and management of protection

From a circuit diagram it is often possible to diagnose the reason tbr a circuit fault
before doing any confirmatory testing.

Planning the programme

(a) Overall review: The equipment to be commissioned may vary from a single
additional circuit connected to a low voltage busbar, to a complete 400 kV double-
busbar substation with transformers, generators and lines.
In either case the method of approach should be the same. First of all, a list
should be prepared itemising all the tests it is necessary to do; for example, insula-
tion resistance, secondary injection, primary injection, control, indication, inter-
locks, alarms, tripping, end-to-end, auto-reclose, synchronising, on-load tests etc. In
some cases it may be helpful to have sheets available headed, for example, 'discon-
nectors' or 'circuit breakers' or 'transformers', and on these sheets have a list of
items that have to be inspected or tested before that particular piece of equipment
can be said to have been checked.
Fig. 18.4.2C gives the schedules in rough outline of some of the tests usually
done on different pieces of equipment. The schedule may be elaborated as
necessary for a particular circuit configuration.

Fig. 18.4.2C Typical schedule of protection commissionirJg tests

Inspection prior to Voltage transformers Tripping tests

detailed testing Ratio Trip circuit supervision
Terminal tightness Back-energising Protection gear to tripping
Check of ferrule numbers Phasing relay
Insulation resistance tests Trip interlocks
Examination of auxiliary Transformers Tripping relays to breakers
switches Phasing at 415 v Intertripping
Winding temperature Acceleration
O'rcuit breakers indicators Autoreclose
Timing tests Buchholz device
Check of c.t. positions Alarm scheme
Current transformers Shorting of initiating
Disconnectors Magnetic curves contacts
Operating mechanism Polarity Proving from all initiating
Control circuits Ratio devices

Pro tec tio n

Secondary injection
Primary injection
On-load tests

Whenever possible there should be a pattern in the approach to commissioning

tests; for example, the a.c. circuits may be considered as feeding into the protective
 Testing, commissioning and management of protection 405

gear and the d.c. circuits as feeding out from the protective gear to the tripping
relays and circuit breakers they trip. This method of working breaks the job up into
easily manageable sections at an early stage and enables faults to be quickly cleared
without the risk of losing track if a complete sequence of tests has been attempted
altogether.
Having proved each section individually, tests to check the overall functioning
can be done fairly quickly. The cause of any failure to operate satisfactorily can
easily be pin-pointed, as the failure must be in the unproved connections between
the already proved sections.

(b) The commissioning log: A commissioning programme can be carried through

more efficiently and effectively if proper records are kept. One convenient way is
to make up a folder, sometimes called the commissioning log, containing stencilled
sheets for recording"

the programme of tests

brief diary of day-to-day progress
the various test results
temporary conditions
outstanding items.

The commissioning log should also include copies of the circuit diagrams to be used
during the tests, and any other sheets thought necessary.
Such a commissioning log enables a comprehensive record to be kept of the
whole job, which will be of value not only during the testing programme but for
any future fault investigation.

(c) Programming: Tests need to be carefully programmed, not only to ensure

that equipment is ready on time, but as regards plant outages, permits, allocation of
staff, etc. The tests should also be programmed to take place in a logical and
efficient order so that nothing already tested is disturbed during subsequent tests.
In preparing an overall programme of manufacturing time, delivery, erection,
inspection, testing and commissioning, adequate time must be included for the last
three items. Engineers engaged on the final commissioning tests are doing their part
of the job after most of the other people have finished theirs; there is often im-
patience to get the new equipment into commission, and the commissioning
engineers are liable (usually unjustifiably) to receive criticism for any delays in
making alive, and are sometimes urged to cut short their tests. Commissioning
tests may reveal several defects which need correcting, and a reasonable amount of
time for such rectification should be included in the overall planned time.

18.4.3 Inspection prior to testing

The inspection tests can be itemised as follows"

406 Testing,commissioning and management of protection

(a) Checking for tightness all connections in relay panels, control panels, marshal-
ling kiosks, c.t. terminal boxes, etc.

(b) Checking that fuses and links are properly wired. In the case of rewirable
fuses, the wire should be tinned copper - n o t a soft lead compound which can
deform and either blow at a current less than its rating or become loose after several
weeks. The fuse wire should not be taut, as it is liable to break when the fuse is
being inserted in its holder. Cartridge-type fuses should be inspected to ensure that
the correct rating is used for the circuit.
If the links are of the type which are wired with stiff wire they should be
inspected to ensure that the wire stiffness is not distorting the jaws and thus pre-
venting adequate contact pressure.

(c) Clearances should be checked between relay studs and the edges of the holes
in the steel panel, making suitable allowance for possible dust accumulation over
the years.
(d) Ensure that all the secondary equipment in outdoor substations is weather-
proof, but that there is provision for ventilation and heating (if required) to prevent
condensation, in accordance with the Board's specification.

(e) Particular attention should be paid to auxiliary switches. These appear in

almost every circuit - for example, tripping, closing, indication, alarm, synchronis-
ing, autoreclose, electrical bolt interlocks, busbar protection, and occasionally in
c.t. circuits other than busbar protection.
Points to look for are silver plating where specified, contact pressure, cleanliness,
adequate 'wipe' and follow through. The drive should be examined to ensure that
it is positive, not liable to jam or become disconnected from the main mechanism,
and that its action corresponds with the main switch. On auxiliary switches used for
busbar protection c.t. selection, it is essential that they are duplicated, and that the
auxiliary switches make before and break after the associated primary contacts, so
as to ensure that the secondary circuits are complete before any primary unbalance
can occur due to pre-contact arcing or a high resistance in one part of the primary
circuit. Tests should be devised to prove that the auxiliary switches meet this
requirement. The importance of correct operation of auxiliary switches cannot be
overemphasised, as failure of an auxiliary switch can defeat the best designed
schemes.

(.I") Before energising new secondary wiring it should be checked as far as

reasonably practicable that there are no bare, dangling, uncompleted leads any-
where on site which will be made alive at the same time as the equipment under
test. Any such leads should be terminated, disconnected or (as an absolute mini-
mum) well taped up.
 Testing, commissioning and management of protection 407

18.4.4 The tests

We can assume that the tests to be done have already been scheduled, primary
circuit by primary circuit, probably with target dates. Before beginning testing it
should be confirmed that the particular part of the circuit under test is complete
and that no one is likely to interfere with it. The tests to be done should be listed
in chronological order together with any special precautions to be taken. The tests
will fall under four broad headings:

(a) Tests before making secondary equipment alive.

(b) Tests before making primary equipment alive.
(c) Synchronising tests where the new equipment is alive, but not yet in parallel
with existing equipment.
(a) On-load tests.

Schedules for recording test results should be compiled in advance, when prepara-
tions can be made at a time free from testing stress. The initial and comprehensive
advance planning is a big help in making the tests run smoothly.
Below are listed some of the more usual tests, with comments.

(a) Insulation resistance tests: These tests should be done with a 500 V Megger,
care being taken that telephone type pilots are not included. To ensure that this
voltage does not appear across rectifiers in the event of a wiring error, or across
capacitors and solid state equipment, they should be shortened out until Megger
tests are complete. Each circuit should be tested in turn whilst all the others are
earthed to ensure that there is no connection between two supposedly separate
circuits. This point is particularly important and needs stressing.
It is difficult to advise what value of insulation resistance should be obtained. On
panel and other indoor wiring one would expect a figure in excess of 100 Mf~. On
outdoor wiring in a medium-sized substation, 10 M~ would be a fair figure; and on
a very large substation, where wiring loops in to many auxiliary switches and
junction boxes, 1 Mf~ would not be unreasonable. The weather can also affect the
results; a warm humid day tends to lower the value, whereas a cold dry day gives
high values.
Often the Megger pointer will swing towards zero when meggering begins on a
circuit. This is because the connected wiring has capacitance; the pointer will climb
towards infinity as the wiring builds up its charge. It is advisable not to touch such
wiring immediately after stopping winding!

(b) Secondary injection tests: These tests consist of a.c. injection into the relay
coils to prove that the relay calibration is correct. Even if the relays have been
thoroughly tested in a laboratory or in the manufacturer's works, some kind of
check should be made when they are mounted on the panel to ensure they have not
been damaged in transit or by handling.
408 Testing,commissioning and management of protection

In general, high-accuracy secondary injection tests on site are not necessary.

Laboratory standards are not being sought. What is needed is a set of figures, repro-
ducible later from time to time on routine tests, which bear a known relationship to
accurate figures previously obtained in a laboratory. If, however, the protective gear
contains tuned circuits which have to be calibrated by secondary injection on com-
missioning, due regard should be paid to the frequency and wave shape of the test
supply to ensure there are no harmonics present.
There are several types of relays, and these are discussed briefly under their
appropriate headings.

(i) General: Whatever type of relay is being tested, if it is fitted with a plug bridge
the tests should be done on the lowest tap so that the whole of the coil is included.
Calibration checks may be done on other taps if desired, but this should not be
necessary. It is, however, wise to check the continuity of each coil tap. This can be
done by moving the setting plug from tap to tap and checking that current flows
when the test supply is switched on. The plug bridge shorting contact should be
proved when the setting plug is withdrawn. This should be done with the test
current switched on, and it should be observed that withdrawal of the plug does not
open-circuit the relay coil.
Although the tests mentioned are of an electrical nature, due regard should be
paid to the mechanical condition of the relay. Dust and other foreign bodies should
be removed with a feather, or by a blower; not by the breath because this contains
many droplets of moisture. The relay movement should be checked for clearance of
gaps, contact alignment, and freedom of operation. If the relay is fitted with a
mechanical flag, the flag should operate as, or just before, the contacts make, and
the flag mechanism should not interfere with the operation of the relay.

(ii) Instantaneous relays: These will be actuated by current or voltage and will
usually be of the attracted-armature type. The current should be increased slowly
until operation occurs. The current value immediately prior to operation should be
noted because of the change in reluctance of the magnetic path of the relay (see
Chapter 6). The current or voltage should then be decreased slowly, and the value
recorded at which the relay resets.

(iii) Inverse definite minimum time (i.d.m.t.) relays: It should be checked that
the contact just makes on zero time multiplier (t.m.). Also the resetting time at t.m.
1-0 should be measured at zero current. For record purposes, the maximum current
which will allow the relay to reset fully should be ascertained.
It should not be necessary to prove the relay curve, but at least two points
should be checked on it; one to find the minimum current, with t.m. = 1 43, at
which the relay will just operate (usually referred to as the creep value), and the
other to check the timing at, say, four times current or voltage setting, and
t.m. --- 1.0.
 Testing, commissioning and management of protection 409

(iv) Biased relays: Such relays have at least two coils; one tends t o cause operation
whilst the other(s) restrains. The method of test will vary with the relay type and
application, but it is usual to check the bias curve by measuring the current needed
in the operating coil to cause operation when different values of separately injected
current are passed through the bias coil. Care should be taken not to overheat the
relay by the fairly large operating current necessary to overcome a strong bias
current. Only three or four points on the curve need be checked.

With some types of end-to-end protection, it may be necessary to carry out

simultaneous injection tests at both ends to check, for example, the bias character-
istics of a teed feeder pilot protection scheme. Several warnings should be issued
hero:
If the (dead) overhead line itself is used as part of the secondary injection
equipment, the greatest care should be exercised with the equipment because of
induced voltage from other lines on the same towers or in the vicinity. Such
voltages can be dangerously high, and the protective equipment and 'live' parts of
the secondary injection test equipment should not be touched unless temporarily
earthed. Such work should only be carried out with the authority of a sanction-for-
test card.
End-to-end tests using the primary line as an interconnector are deprecated
because of the danger, but are mentioned here for the record.
If simultaneous secondary injection tests are to be done at the two ends using
two sets of test equipment and separate supplies at each end, it is essential that
the two supplies are in phase or in anti-phase, depending on the type of protection.
If the phasing is known, well and good. If not, it will have to be established. Be
careful not to rely too much on telecommunication company pilots as a means of
phasing out test supplies (which should, of course, be transformed to a safe voltage
before applying them to the pilots), because the inductance and capacitance of the
pilots can introduce appreciable phase-shift. In one case, 30 V supplies at each end
were phased out over a pair of telecommunication company pilots by engineers
who had not appreciated the possibility of phase-shift. Their results, when plotted
out, indicated approximately 30 ° difference in phase between the supplies at the
two ends. To compensate for this 'difference' the tests were done phase-to-neutral
at one end and phase-to-phase at the other, when the two supplies were in fact in
phase. Needless to say, most interesting results were obtained!
This problem of phase-shift in pilots can be overcome once it is known, but it is
safer to phase out test supplies against a known voltage reference at each end.
With end-to.end injection there is wide scope for errors, and it is only too easy
to overlook a 180 ° shift through a transformer.
End-to-end tests by secondary injection should not be relied on entirely for
establishing protective gear polarity. At best, they are only a guide and the results
should always be confirmed later by on-load tests.
If a biased relay is used for the overall protection of a power transformer it may
have a harmonic.bias coil on it in addition to the load-bias coil. The harmonic-bias
410 Testing, commissioning and management of protection

coil stabilises the relay against unwanted operation on magnetising inrush currents
at the instant the power transformer is switched on to the system. The calibration
of the tuned circuit associated with this coil should be proved, and here the
frequency of the test supply should be confirmed as 50 Hz, otherwise inaccuracies
may be present.

(v) Distance protection relays: Certain distance relays are of the d.c. permanent
magnet moving coil type, in which the coil moves axially or radially. Sometimes the
relay has only one coil fed from the d.c. bridge which compares the magnitudes of
the current and voltage and so is a measure of the impedance to the fault. On
another type the relay possesses two coils on the one former; one coil is fed from
the rectified current and tends to cause operation, whilst the other coil is fed from
the rectified voltage, thus tending to restrain.
In secondary injection tests on these relays, current and voltage of varying
magnitude (to simulate different values of fault impedance) are fed into the pro-
tective gear to prove that the equipment is in calibration.

(vi) Directional relays: Directional relays are frequently of the induction type,
where the disc or cup is acted upon by a current coil and a voltage coil. Depending
on the phase angle between voltage and current, the relay will either operate to
close its contact or remain with its contacts open.
Directional relays usually act in conjunction with other relays. For example,
in a directional overcurrent relay the sensitive directional element controls the
operation of a low-set overcurrent element. In many types of distance protection
the directional relay acts as a starting relay and controls the operation of many
auxiliary relays associated with the measuring elements.
In testing directional relays it is usual to prove that they will not operate or even
creep in the presence of either actuating quantity alone. It is also necessary to find
the phase angle between the current and applied voltage to give zero torque - that
is, the point at which the relay changes direction. If the applied volts are then
moved 90 ° from that position, maximum torque should be obtained. In that
position figures may be taken of the minimum current required to cause operation
at rated volts, and of the minimum voltage required to cause operation at rated
current.
In the case of directional earth-fault relays, consider what happens when an
earth fault occurs on a solidly earthed system. This condition is depicted vectorially
in (a) of Fig. 18.4.4A in which a red phase-to-earth fault is assumed. The relay is
fed with residual current and with a voltage derived from open.delta transformers:
this is shown by VOD in (b) and the fault current is shown by IREs. The relays for
this application are designed to operate with maximum torque when the applied
current and voltage are in phase, and it is, therefore, necessary to move one vector
or the other to achieve this in-phase condition under fault conditions. The voltage
vector is the one most conveniently moved, and (c) and (d) show how the requisite
amount of compensation is achieved. The angle of compensation is usually of the
 Testing, commissioning and management of protection 411

order of 60 ° to 70 ° .
Measurement of the compensation angle may be made by single-phase injection
on, say, the red primary winding of the auxiliary v.t.s and by then measuring the
open delta (FOB), capacitor (I/c), and relay coil (I/REL) voltages.
The secondary injection tests on directional earth-fault relays should be carried
out in a similar manner to the injection tests on directional overcurrent relays, due
allowance being made for the difference in maximum torque angle and sensitivity.

VR VR IR ES
IF i,., f

VB
VB Vy~
VOD
(a) Voltage and current phasors (b) Open-delta voltage and
for Red phase earth fault on residual current in relay
solidly-earthed system
VOD ~ ~ 7 vC

- - , - --n
v' compensation

1__31 l
I_ ] - Relay -
VOD
(c) Auxiliary v.t.'s, open-delta and
voltage compensation circuit (d) I'has~rs showing angle of
compensation

- I B at unity power factor load

= IRE S
0
I (e) Red voltage fuse removed
J and limb shorted.
Blue c.t. shorted and disconnected.
-V R Note similarity to (b)
= VOD

Fig. 18.4.4A Proving of earth-fault directional relays

(c) Preliminary tests on current transformers: Many things should be checked

before the stage of heavy-current injection is reached. These are"

(i) Overlap of c.t.s: Where the c.t.s are arranged to overlap the circuit breaker so
that a fault within the breaker is covered by both adjacent zones of protection, it
should be proved conclusively that the c.t.s do overlap, otherwise a fault within
 412 Testing,commissioning and management of protection

Post-type c.t.'s
A
! v
¢ ,
I I
I I
z I
)

Test
transf()rmer

Heavy test current is possible

240 V supply

Test current severely limited by impedance of t r a n s f o r m e r

Fig. 18.4.4B Effect o f c.t. position on magnitude o f test current

the breaker could not be cleared by either set of protection.

The proving may be done by visual inspection, thought being given to the ever
present possibility of crossed ferrule numbers. If visual inspection is impossible or
difficult, then a Megger of continuity indicator may be applied between the ap-
propriate relay(s) and earth at the relay panels, whilst a roving earth is 'dabbed' on
and off the secondary terminals at the c.t.s to prove they are on the correct side of
the breaker.

(ii) The correct c.t.s: Often there are several sets of c.t.s in the one bushing and it
is vital to prove that these are respectively connected to the correct protection.
Sometimes all the c.t.s will be of the same ratio but will have very different
characteristics. At other times the c.t. ratio will be different but close enough to
cause confusion unless great care is taken.
With post-type c.t.s or c.t.s in an o.c.b, bushing the proving of the correct c.t,
ratio causes little difficulty because heavy test current can be passed through them;
but where such c.t.s are in the bushings of a power transformer, the position is
much more difficult because only small test currents can be passed. The two cases
are indicated in Fig. 18.4.4B.

(iii) Magnetisation current tests: These tests are done:

(a) to prove there are no shorted turns in the c.t.,

(b) to establish the knee point voltage (k.p.v.) of the c.t. (the knee point voltage
 Testing, commissioning and management o f pro tection 413

is that point on the magnetisation curve where a 10% increase in applied

voltage will result in a 50% increase in magnetising current),
(c) to establish the capability of the c.t.s. (This is allied to the point (c) (ii),
page 412 and is a way of proving that the different types of c.t.s can be
distinguished from each other.) Magnetisation curves are obtained by ap-
plying a sinusoidal voltage to the secondary winding of a c.t. and measuring
the magnetising current flowing for different values of applied voltage. A
typical curve is shown in Fig. 18.4.4C. There is no need to continue the test
beyond the knee point. These tests must be made with the primary of the c.t.
open-circuited. When carrying out magnetisation tests on c.t.s, it is most
important that the applied voltage is progressively reduced to zero before
interrupting the supply; otherwise the very high rate of change of flux may
induce a voltage sufficient to damage the secondary insulation.

Knee p o i n t voltage is t h a t voltage

at w h i c h a 10% increase in voltage
500 results in a 50% increase in c u r r e n t

I 0%,

400
I so% I
I I
I I
0
IJ I K n e e p o i n t voltage 4 2 0 V
>
~, 300 I ] c.t. ratio 5 0 0 / i
I I
I I
I I
200 I I
I I
I I
I I
I00
I I
I I
I I
I I
I I

0 50 i00 150 200

Current (mA)
Fig. 18.4.4C A typical c.t. magnetisation curve

(iv) Flick tests: Much preliminary checking can be done with the aid of a 1.5 V
battery and a voltmeter. All the c.t.s in a group can be checked in turn to prove
that they are connected to the protection in the same polarity. A voltmeter, on a
414 Testing,commissioning and management of protection

sensitive d.c. range, should be clipped across (and left across during the tests) the
relay or c.t. secondary whilst the 1.5 V battery is touched across the bar primary of
each c.t. in turn. Great care should be taken to ensure that the battery leads and
voltmeter are applied in the same way each time. Fig. 18.4.4D shows the arrange-
ment. It is important to note that the deflection of the voltmeter is the same each
time the battery is connected, and each time it is disconnected.
A flick test can be done profitably with c.t.s in a transformer bushing, because
the d.c. current can traverse the transformer windings with little opposition. It is
even possible to do rough operation and stability tests by excluding or including the
balancing neutral c.t. within the path of the battery current.
V o l t m e t e r left
clipped across relay
during test

IV2V "-
cell

,,,,,.

-- I . o r w a r d k i c k on a p p l i c a t i o n o f test l e a d .
T h i s lead f l i c k e d B a c k w a r d k i c k o n r e m o v a l o f test lead.
on a n d o f f

Fig. 1 8 . 4 . 4 D Arrangements for flick tests

(d) Primary in/ection tests: Primary injection tests are required to prove that c.t.s
are correctly connected to relays. These tests will usually include a check of the
polarity of the three c.t.s in a group, and often the comparison of the group with
another. The sensitivity of the protection, in terms of primary amps, may also be
determined.
At this stage a simple diagram should be drawn (to form part of the commis-
sioning log) showing c.t. locations, the testing transformer and the position of the
test leads in the circuit for each test of operation, and stability, phase-to-phase,
phase-to-earth, etc. Typical diagrams are shown in Figs. 18.4.4E, F, and G, but will
be referred to again later.
Primary injection tests can be a prolific source of error. The following facts
should be borne in mind:

(i) The primary equipment must be dead. The engineer conducting the tests,
although possibly not in charge of the safety aspects of the test, would be well
advised to satisfy himself that all equipment to be worked on has been isolated and
locked off from all live equipment.

(ii) The work will probably have to be done under a sanction-for-test car.

(iii) Safety earths may have to be removed.

416 Testing,commissioning and management of protection

Connection applied to
each phase in turn
. . . . . . . . . 0 L
240 V supply

/f~:O N
/ /
// [
Amm~
/ / I Here for
/ &stability
/ H e r e for
operation

Fig. 18.4.4F Proving of c.t.s in transformer bushing by injection from 240V a.c. supply

(iv) Test connections will have to be applied, and must make good contact in view
of the low voltages and heavy currents which are used.

(v) The test current will sometimes flow through c.t.s other than the ones in
which one is specifically interested at the time. Precautions will, therefore, be needed
to prevent the tripping of adjacent circuits and the remote ends of circuits.

(vi) It is possible to produce what is thought to be an operating condition, where-

as, in fact, the condition may be of one stability. This can occur by accidentally
passing current in the wrong direction through the c.t.; the sketch of the c.t.
positions and the test connections will help to prevent this.

(vii) Earth.fault relay coils, stabilising resistors, non-linear resistors and the like are
often short.time rated and so may be damaged unless suitable precautions are taken.

(viii) Multipurpose instruments may be damaged unless they are on the correct
range for the conditions being applied. Before switching on the test supply, think
about the whole job and make sure that the necessary precautions have been taken.
The precautions taken should be listed so that when the tests are complete every-
thing can be restored to normal without the risk of anything being forgotten.

It is advisable to switch on at a low value of current and increase it slowly in case

there are any c.t.s open-circuit or incorrectly connected. If the test results differ
appreciably from those expected, reduce the current and switch off while the matter
is investigated.
 Testing, commissioning and management of protection 417

w . . . ~

I
I
I
I Temporary
earth applied
I to each phase
l in turn

4 1 5 V 3 phase test supply

applied here
Fig. 18.4.4G Proving of polarity of c.t.s, in transformer bushings by injection of 415V three-
phase supply on L v. side of power transformer
A word of warning about the burdens of test instruments when connected in
circuit on the miUiamp range. On this range the instrument will have quite a high
impedance, and this may affect the distribution of currents in the secondary circuit
under test. This can be even more confusing if only one instrument is available and
it has to be transferred from phase-to-phase.
It is wise to prove operation first - that is, prove that the protection will operate
from at least one set of c.t.s. If the c.t.s are connected to form a circulating-current
scheme, stability can be proved by going through the two sets of c.t.s in series.
There should be no, or little, current in the relay.
Fig. 18.4.4H(a) shows some typical primary injection test equipment. The test
current will usually be between 100 and 400 A. For the sake of clarity on the c.t.
diagrams showing the different types of test, the injection equipment has not been
drawn again, but the two heavy.current leads labelled (1)and (2)are temporarily
connected to the c.t.s being tested.
Fig. 18.4.4H(b) shows three c.t.s residually connected to a relay. The ratio of
the c.t.s must be proved, also the correct connections of the c.t.s in the group. If
there is more than one group of c.t.s connected to a relay, as in busbar protection
(Fig. 18.4.4H(c)), then not only must each group be proved as a group but each
group must be balanced against the others.
Fig. 18.4.4H(d) shows how a ratio test is done. The test current is passed
through one c.t. (say Red) and primary and secondary currents are measured. If the
c.t. is correctly matched to the protection, the relationship between the two should
confirm the c.t. ratio.
418 Testing,commissioning and management of protection

Measuring c.t. and a m m e t e r

L
r, w , ~ , (~) R Y B _

a.~ Varia: ~ ; ~ testing transformer

• < Q - _
N
(a) Primary injection testing equipment (b) Group of c.t.s
c o n n e c t e d to a relay
R Y B R Y B R Y B

(c) Several groups o f c.t.s c o n n e c t e d to a relay

R~Y ~B RjY A
B
1 / f

Zero

(d) Ratio test (e) Balance test R to Y

Group A Group B Group C
,

.
v v _

(f) Balance test group against group

Fig. 18.4.4H Connections for primary injection tests

To prove the e.t.s in the group, inject into the Red c.t. and out of the Yellow as
shown in Fig. 18.4.4H(e). There should be little or no reading even though the same
test current is flowing as before. If these conditions are satisfactorily achieved,
current should be injected into Yellow and out of Blue, when similar results should
be obtained. The correct balance of the e.t.s in the group has now been proved, as
has also the ratio of all three.
Fig. 18.4.4H(f) shows several groups of e.t.s in parallel on a relay. One group
must be proved for ratio and balance as already described, and the three c.t.s in
each of all the groups must be balanced against each other. It is now only necessary
to prove group against group. This can be done in several ways and one is described:
test current is injected into (say) the Red c.t. of group A and out of the Red c.t. of
group B. If the polarities and ratios are correct there should be little or no spill
current in the ammeter. It has now been established that the ratio of the c.t.s in
 Testing, commissioning and management of protection 419

group B is the same as in group A; and since it has already been proved that the
three c.t.s in each group balance, group A must balance group B. In a similar way
group A (or group B) may be balanced against group C, etc. Where the relay is of
the high impedance type, the test ammeter (A) in Figs. 18.4.4H(d), (e), and (f)
should be connected in parallel with the relay instead of in series.
Reverting to c.t.s in transformer bushings where it is impossible to inject sub-
stantial test current, a test can be done by applying a 240 V a.c. lead to each h.v.
bushing in turn whilst the l.v. side of the transformer has a three-phase short-circuit
across it. If desired a 415 V three-phase supply can be applied to the three bushings
but this will not prove the neutral c.t. as no current will flow through it. The use
of the single-phase supply enables the neutral c.t. to be included in the test if
desired as indicated in Fig. 18.4.4F. It will be appreciated that the primary currents
flowing will be small so the secondary currents will only be of the order of milli-
amps. However, such a test can prove correct connections.
It is possible to obtain a larger test current by applying the three-phase 415 V
supply to the l.v. side of the power transformer. This is shown in Fig. 18.4.4G,
which also shows how a temporary short on the h.v. side can be applied in different
positions to prove stability on each phase. Relay spill currents must be checked
later when the transformer is on load.
Many transformers are protected by overall biased differential protection
schemes incorporating h.v. and l.v. restricted earth-fault relays. Although heavy-
current primary injection tests can be made from c.t.s to relays, it is difficult to
prove the h.v.c.t.s against the l.v.c.t.s because of the phase shift through the trans-
former. Such balance can be proved when the transformer has gone on load, but a
word of warning is necessary; if the transformer is fitted with a fault throwing
switch, temporarily remove the tripping from the overall biased differential pro-
tection until the protection has been proved correct. A case is on record where the
polarities were wrong, so the protection interpreted load current as an internal fault;
the protection operated and closed the fault throwing switch on to the live 132 kV
line.
On generator protection the zone of protection may include both the generator
and its step-up transformer. The connections from all the c.t.s to the relays must
be proved by primary injection, but the balance of the h.v.c.t.s with the l.v.c.t.s
has to be proved when current is flowing through the generator and transformer.
This is achieved by using the machine itself as a source of test current. Three-phase,
phase-to-phase, and phase-to-earth temporary connections are applied in turn to the
h.v. side of the transformer, and the machine is run on low excitation whilst correct
behaviour of the protection is checked. These tests, in conjunction with the
primary injection tests already done, prove the machine protection.
Because the generator transformer is fitted with tap changing gear, some spill
current may appear in the overall protection when the machine is run on the
external three-phase short-circuit. This test should, therefore, be done initially on
the nominal-ratio tap of the transformer, and repeated on the two extreme tap
positions. Because the primary currents are balanced in the three phases, there is
420 Testing,commissioning and management of protection

no need for this part of the test to be hurried.

Unbalanced currents in a machine cause rotor heating, so care must be taken to
ensure that this is not excessive. Before the unbalanced current is applied it is
necessary to know the amount of unbalance, in order to determine for how long
the unbalance can be tolerated, and what reading these conditions will give on the
machine ammeters. As an example, consider the case of a 60 MW, 0.8 p.f., 11-8 kV
generator connected to the 132 kV system through a 75 MVA delta/star trans-
former. On full load c.m.r, the 11.8 kV and 132 kV currents will be 3660 A and
327 A, respectively, and the currents in the delta windings of the transformer will
be 2120 A. This is shown in Fig. 18.4.4I(a).

Generator 60 MW IR 3660 R 1
2120 I R 327

o11.8
. , kV Iy 3660 Y 21 Y
i|

327 i|

75 MVA B 3660 B B 327

v

1 1 . 8 / 1 3 2 kV
(a) 3-phase full load .--'

2120

Iy 0 Y Y .

B 2120~.
~ B B ..~ ,
Z 327 327
(b) Phase-to-earth fault --" ---

2120
J

Iy 2120 Y y

2120 B

(c) Phase-to-phase fault

F ig. 18.4.4 I
If it is decided to pass the full load current of 327 A through to phase-to-
earth 'short' on the 132 kV side of the transformer, the currents flowing in the
three phases of the generator will be 2120 A, 0 A, and 2120 A as shown in Fig.
18.4.4I(b).
 Testing, commissioning and management of protection 421

If the same current of 327 A is now passed through a phase-to-phase 'short' on

the 132 kV side of the transformer, the currents flowing in the three phases of the
generator will be 4240 A, 2120 A, and 2120 A as shown in Fig. 18.4.4I(c).
To assess the effect of these unbalanced currents on the generator, reference
should first be made to the sequence components shown in Fig. 18.4.4J. From
this it will be seen that a phase-to-earth 'short' imposes a negative phase sequence
condition of 57.7 per cent of c.m.r.
1R

I,R I1R I 12 R

i l B ~ y
II B I Iy !2 y I2 B 12 ,,~'- J L - - .~ 12
I I B

Positive Negative Addition of positive

and negative

12 IR 4240 2120 : 5"/.7% CMR

R --
2 ~
2
Sequence components for phase-to-phase fault

I1 R IR

12 '~lRf "I
12y R 12y ~ 12 R

",, ,,"-/1,-",<
12B 1B
Positive Negative Addition of positive
and negative

12 = I.RR= 2120 = 1220 = 33.3% CMR

R ~ 7~
For phase-to-earth fault
Fig. 18.4.4J Sequence components of generator currents for faults on h.v.. side o f generator

Fig. 18.4.4K gives the n.p.s, withstand figures in percentage of c.m.r, rating for
2 2
machines of various 12 t values. A machine with an 12 t value of 15 could withstand
the earth-fault condition for 130s and the phase-fault condition for 47s, whereas
2
a machine with an 12 t value of 7 could only withstand these conditons for 60s and
2 Is, respectively.
Reverting to the phase-to-phase fault condition which imposes currents of
4240 A, 2120 A, and 2120 A on the machine, the manufacturer may be unwilling
to permit this slight overload on one phase (the c xn.r. rating is 3660A). If so,
the test current would have to be kept lower, which would give a longer time for
422 Testing,commissioning and management of protection

Curve A B C D E !- G H J

122t 20 15 13 12 10 7 3 2.5 2

200

100

50

20

10

J H G

10 20 30 40 50 70 I00

NPS as a p e r c e n t a g e o f C M R

Fig. 18.4.4K
 Testing, commissioning and management of protection 423

the readings to be taken. Whatever current value is taken and whatever the type
of machine, reference to the appropriate curve will show for how long the con-
dition can be withstood. It is then advisable to allow a factor of safety by halving
these times.

(e) Tests on commissioning of carrier protection (Refer also to Section 18.4.4(gj:

The band of frequencies at present employed in the UK for power-line-carrier
signalling is 70-700 kHz, although certain parts of this band are not available for
continuous carrier signalling. This restriction is necessary to avoid causing inter-
ference with radio service including those used for aeronautical and marine nagiva-
tion purposes. However, with the exception of a band designed to protect the
international distress frequency, normally quiescent power-line-carrier systems may
be used anywhere in the remaining available frequency range provided that test
transmissions do not exceed two seconds and that there is an interval of at least two
minutes between successive test transmissions. Carrier protection consists of the
starting networks, the electronic equipment, the line coupling equipment and line
traps, and the power supply unit and power packs. Tests are made to prove each of
these items.
The starting networks can be checked by secondary injection. Before starting
work on the electronic equipment the power supplies should be checked. These are
usually provided by d.c./d.c, convertors which invert a battery supply to provide an
a.c. supply which is then rectified and stabilised to provide the appropriate voltage
level(s) required by the equipment. Arrangements vary with the type of equipment
and the manufacturer's commissioning instructions should be consulted for detail,
particular attention being paid to fuse ratings, polarity of supplies and voltage
values.
Certain special test equipment is necessary for use on the electronic equipment
and on the line coupling equipment. This test equipment comprises a cathode ray
oscilloscope, a valve voltmeter, an oscillator and amplifier covering the range 70 to
700 kHz. The protection equipment consists of a transmitter and receiver at each
end of the feeder, and tests have to be done on these to adjust the outputs and
sensitivities and to check the performance of filters.
Line-coupling equipment provides a means of injecting h.f. carrier signals on to
an h.v. power line and it prevents these signals being dissipated in the substation. A
typical set of line-coupling equipment is shown in Fig. 18.4.4L. The equipment
comprises a line trap and an h.v. capacitance stack, stack tuning coil ('series arm')
and shunt arm associated with each of the two phases used for the transmission of
carrier frequencies. The 'shunt arms' are connected across the earthy end of the
series arms to provide a low impedance to earth for 50 Hz. The shunt and series
arms constitute a bandpass filter. A number of standard bands are available, typical
ones being shown in the table below.
On commissioning, the tuning of the line traps, the series arms and the shunt arm
has to be adjusted or checked. The series and shunt arms should tune to the mid-
band frequency, adjustment being provided on the series arm coils. The shunt arm
should have been supplied already tuned to the midband frequency.
 424 Testing,commissioning and management of protection

Band Frequency range kHz Midband (geometric mean) frequency kHz

1 70- 81 75
2 80- 95 87
3 90-110 99.5
4 100-125 112
5 110-140 125
6 120-158 138
7 130-175 151
8 150~14 179.5
9 180~80 225
10 250500 354
11 350G00 495

The tuning of the line traps is achieved by suitable choice of shunt capacitor.
One British manufacturer recommends tuning them to the carrier frequency,
whereas another manufacturer recommends that they be tuned to the midband
frequency. The method of tuning will be the same in both cases and is described
below.
For these tests the oscillator, amplifier and valve voltmeter are used. Signal levels
in excess of the maximum output normally given by test oscillators are required to
mask possible interference, thus the amplifier should have an open-circuit output of
about 50-100 V. For all tests the earthy side of the amplifier output and valve
voltmeter should be connected to the station earth.
When checking the tuning of the various resonant circuits in the line coupling
equipment, great care must be taken to see that electrical connections between
units do not introduce capacitance in parallel with the unit under test. Stray capaci-
tance should also be reduced to a minimum. With these points in view it is recom-
mended that isolation of the line traps should be achieved by earthing the line traps
on the line side with portable earths and disconnecting the copper connectors on
the station side, the disconnection being made at the line trap end of the connectors.
Reference should be made to Fig. 18.4.4M which shows the test connections re-
quired to tune the line traps, the series arms and the shunt arms.
Line traps: Because the line trap is a parallel-connected wide band resonant
circuit with high impedance at resonance, the valve voltmeter reading will be a
maximum over the bandwidth at resonance. To obtain good selectivity a high
impedance source is required. This is achieved by connecting the live side of the
amplifier output via a high resistance (a half-watt 30 k~2 carbon resistor is suitable)
to the station side of the line trap. The valve voltmeter is connected with its live
side to the line-trap side of the resistor.
Series arm: The series arm is a circuit with low impedance at resonance, and the
valve voltmeter reading will be a minimum at the resonant frequency of the series
arm.
 Testing, commissioning and management of protection 425

Line traps

Busbars Line

Coupling
capacitors

0-- 0- ,

i Series
tuning
!
unit
I
m

Capacitor
v.t.h.v.
chambers

"--I

t"ilter
cubicles
~ Shunt
L tuning
unit

-"2"
Coaxial cable
to indoor e q u i p m e n t

Fig. 18.4.4L Linecoupling equipment

 Testing, commissioning and management of protection 427

The lower end of the series arm should be isolated as it enters the capacitor v.t.
h.v. chamber. On a combined coupling capacitor/capacitor v.t. assembly, the
capacitor v.t. must be included in the circuit. To obtain good selectivity a low
impedance source is required and an amplifier output impedance of 75 S2 is satis-
factory. The live side of the amplifier and valve voltmeter are connected to the
lower end of the series arm.
S h u n t arm: The shunt arm is a parallel-connected resonant circuit. It should tune
to the midband frequency of the coupling filter.
All connections should be removed from the unit and the live side of the ampli-
fier should be connected to the unearthed side of the unit via a 30 k ~ resistor. The
valve voltmeter reading will be a maximum at the resonant frequency of the shunt
arm.
Since attenuation of the carrier signal occurs along the primary line an end-to-
end measurement of this attentuation must be made. Again use is made of the
oscillator and amplifier, which are connected to the h.f. cable at one end of the line
whilst the received signal level is measured at the other. It is usual to make this test
over a range of frequencies on either side of the intended working frequency. If
possible three sets of tests should be done: one with the line dead and earthed on
the station side of the line traps, one with the line unearthed, and one, later, after
the line has been made alive and is carrying load. A curve of some typical results
is illustrated in Fig. 18.4.4N.
Precommissioning tests of modern phase-comparison carrier protection should
include tests on transmitter levels, receiver sensitivities and trip angle settings.
These are best achieved by following the manufacturer's recommended commis-

32

30

28

26

-~ 24

.£ 22

= 20

< 18

16

14 Note" C u r v e d r a w n t h r o u g h
mean of points obtained
12

10
I00 120 140 160 180 200 220 240 260 280 300 320
I'requency (kHz)

Fig. 18.4.4N On-load attenuation figures for 275 k V line; designed carrier frequency 132 kHz
428 Testing,commissioning and management of protection

sioning instructions for the particular equipment, care being taken to ensure that
adequate margin exists between alarm and trip level sensitivities. The built-in testing
facilities which, by comparison of the injected currents at the two ends, permit
routine stability and tripping tests to be carried out, should also be checked over
and phased out.
The latest designs of normally quiescent power-line-carrier protection have had
their built-in test facilities modified to comply with the permitted limits of carrier
transmission time and are therefore capable of application throughout the band.
However, during commissioning tests whenever carrier transmission to line is
required care must be taken to comply with these conditions if operating at fre-
quencies within the restricted bands. Individual manufacturer's commissioning
instructions provide detailed information on this aspect specifically for their equip-
ment.
To assist in checking the trip angle and sensitivity of the protection, facilities
are provided for disconnecting the output to line and inserting a dummy load so
that the equipment is terminated in 7512, thereby enabling these tests to be carried
out without transmission of carrier to line.
When all these tests have been done satisfactorily the line may be put on load
(see Sections 18.4.5 and 6) after which an end-to-end polarity test must be made.
This can be done by the temporary disconnection and shorting of the same phase
c.t.s at the two ends of the line to obtain an output. Operation of relays must be
simulated if load current is below the setting of the equipment, and the phase of
the carrier signals checked on the cathode ray oscilloscope. Fig. 18.4.40 shows the
trip and stability zones of a typical phase-comparison equipment.
End A End B

End B current
as in conditi,~n (h; End A current

Uondition (a) Definite tripping

Find B leading End A

end A as in leading end B
End A End B conditi,~n (c)
Stability
....,Jldllllllllll
llillllilll lllllllllilillr'"'""°"'i
t ' o n d i t i o n (b) Definite stability
I
Tripping
zont~
F 30 ° gaps " 7

gllllillllil.. Iggllll(ll ,,...,. l

HIIlIlIIlilI I!llJilllil lie I'hasor representation
t.'ondition (c) Marginal tripping/stability of tripping and stability zones

Fig. 1 8 . 4 . 4 0 Typical pictures from ¢.r.o

 Testing, commissioning and management of protection 429

(f) Trip angle test unit: One make of equipment has built-in facilities for checking
the trip angle setting, but for another an external trip angle test unit is required.
The fundamental requirement is to modulate the carrier frequency signal in order
to produce variable gaps and so determine the trip angle setting.
This device produces pulses of variable duration which are used to modulate a
constant carrier signal from the local oscillator in order to establish the shortest gap
which will cause tripping.
D.C. pulses, adjustable in magnitude over the range 50 mV to 30 V and in
duration from 20 to 75 electrical degrees are produced. Fixed duration pulses of
180 electrical degrees are also available by switch selection. Reference should be
made to the manufacturer's instructions for suitable test points for connection of
the test unit output.
Fig. 18.4.4P shows the circuit diagram of a trip angle test unit consisting of the
following main parts:

(/) Power pack: This is a simple full wave push-pull rectifier circuit producing
approximately 30V across C1.
~) Phase shifter: This is composed of a power transformer Trl, resistors R1,
VR1 and capacitor C2. Resistance R2 is high compared with R1 or VRI:
consequently R1, VR1 and C2 can be considered to pass a common current
and the voltage across R1 and VR1 will be 90 degrees in advance of that
across C2. The vector relationship of these voltages is shown in Fig. 18.4.4Q
from which it can be seen that the locus of the voltage at point D is equal in
magnitude to one half of the secondary voltage of Trl while it advances in
phase relative to that voltage as VR1 is reduced.
Locus of point D

Q Common
Secondary voltage of Trl
Fig. 18.4.4Q Voltage phasor relationship
(iiO Phase comparator: This is a logical NAND circuit, where the collectors of
transistors T1 and T2 are both shorted to common if a positive voltage is
present on either base. Therefore, as illustrated in Fig. 18.4.4R a pulse is
produced on this common collector once per cycle, its duration being de-
termined by the phase difference between the voltages at B and D (i.e.
dependent upon the setting of VR1).
(iv) Current amplifier: This is an emitter follower circuit, the output of which
 Testing, commissioning and management of protection 431

is fed to the preset potentiometer VR2 which provides adjustment of the out-
put voltage level.
(v) Output switch: In some cases it is necessary to have 180 degree pulses and
these are obtained by removing the action of transistor T1 from the com-
parator circuit by shorting D3 via an output selector switch.
The components can be conveniently built into a small metal box with the potentio-
meter control and switches located on top for easy operation.

Voltage at @

/i
I i\ /~
t iX /~
I I ~ I

Voltage at ( ' ~ I \ I /¢ I
x,_) i \ t /

i I\I /i i
I I I I I
I I I
I I I I
: I I
Voltage on
T I base ,t ,/
I "1
I I
Voltage on
T2 base t /-
Voltage on
T l collector ....
I L

Voltage at
output
1 L
Fig. 18.4.4 R Voltagewaveforms

(g) Tripping and closing tests: It is not possible to prove the trip and close wiring
of a new circuit until the fuse and link have been inserted; yet if there is anything
wrong with the wiring the battery may be short-circuited when this is done. As
432 Testing,commissioning and management of protection Next Page

a precaution against anything untoward happening, it is wise to check with a

voltmeter across the top and bottom jaws of the holder before the fuse is inserted,
and again for the link. A zero reading on the meter proves that the fuse or link can
confidently be inserted. A reading on the voltmeter is an indication that there is a
path from positive to negative through the voltmeter and some part of the circuit.
Such a voltage may be quite healthy - for example, there may be a path through
some circuit supervision equipment or indicating lamp; but the reason for that path
should be established before inserting the fuse or link.
As mentioned earlier, a copy of all extracted diagrams should be included in the
commissioning log and as each contact, fuse, link, changeover switch etc., is proved,
it should be ticked off on the appropriate diagram. If the same diagram is used for
several circuits, differently coloured ticks may be used for the different circuits. If a
relay has two separate contacts, one tripping a circuit breaker direct and the other
tripping a tripping relay which then trips the circuit breaker, each should be
checked separately.
Most present-day relays are fitted with mechanical flags, and it should be
carefully checked that these are not likely to interfere with the operation of the
relay element. If series electrical flags are employed it should be proved that two of
them operate if their associated contacts are both closed in parallel. Under this
condition the flag current will be roughly halved, and it is desirable to check that
there is an adequate margin for operation.
Tripping tests should be proved at 100% and at least one lower value e.g. 60%
battery volts. The latter is not an entirely theoretical test because if a circuit
breaker is closed onto a fault requiring the breaker's immediate tripping, the volts
available from the battery will be reduced by the drain of the closing solenoid.
However, it should be noted that a requirement of British Standard Specification
5311 and 3659 is for the full fault breaking capability of circuit breakers to be
available when the voltage at the trip coil is within the range 80-120% of nominal,
and care should be taken in the design stage to ensure that 80% battery voltage is
available at the trip coil under the most onerous conditions of operation.
Trip circuit supervision relays should be proved, and it should be checked that
the alarm supply is not fed from the battery being supervised, otherwise no alarm
could be given if the supervised supply failed. Trip-healthy lamps and resistors
should be proved and the lamps should be shorted-out to prove that the series
resistance is adequate to prevent tripping. In doing this test, it is advisable to short-
out the lamp several times in quick succession whilst somebody listens adjacent to
the switchgear to make sure that the associated trip coil plunger is not moving. A
case has been experienced where the current passed by the resistor was sufficient
to lift the plunger and move the tripping mechanism slightly but not enough to trip
the breaker. After several operations, however, the tripping mechanism had 'inched'
so far that tripping occurred.
If, during any of these tests, it has been necessary to disconnect the drives to
banks of auxiliary switches to operate them in both positions, the drives must be
reconnected after the tests and any split pins, circlips, etc., replaced.
Previous Page Testing, commissioning and management of protection 433

(h) Indications and alarms: These tests involve checking that each indication and
alarm functions correctly and they are largely self-explanatory. They are, however,
important and the opportunity should be taken of instructing the operating staff in
the functioning of the various controls and of giving them the opportunity of
cancelling the alarms.

18.4.5 Phasingtest.,

It is necessary to prove correct phasing before a new piece of equipment can be put
in parallel with existing equipment. For example, a 132/33 kV grid transformer
is generally fitted with links in the 33 kV side which enable it to be connected to
vector group Yd I or Yd I 1, and tests should be done to establish that the new trans-
former phasing is the same as that with which it is to be connected in parallel.
This can be done before the transformer is made alive from the system by
applying 415 V three-phase to the h.v. side, measuring voltages between the 132 kV
and the 33 kV sides, and recording them in a schedule similar to the one below.

33kV Side 132 kV Side

Rt Yl Bl

R 2 " ' " ' ". . . . . ". . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Y2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I . . . . . . . . . . .

B . . . . . . . .2
............. .. . . . . . . . . . . . . . . . . • . . . . . . . .

Volts should be measured from R~ to R2, ]"2 and B2, from Y~ to R2, Y2 and
B2, and from B~ to R2, Y2 and B2. It is essential that the 132 kV and the 33 kV
sides are temporarily commoned at some point for this test, otherwise misleading
values will be obtained as there is no metallic return path for the voltmeter current.
When the results have been obtained, the vectors can be plotted from which the
phasing of the transformer can be established.
On lower voltage systems, for example up to 11 kV, it is possible to employ
phasing sticks. These are insulated sticks which can safely be inserted into switch-
gear spouts and used in the same way as a voltmeter to phase out the R, Y and B
phases of one circuit with the R, Y and B phases of a circuit already proved.
It is impracticable to use phasing sticks above this voltage, so for 33 kV, 66kV,
132 kV, 275 kV and 400 kV systems, phasing tests are usually carried out between
the 110 V sides of voltage transformers connected to such systems. Phasing tests
can, however, be done between a 110 V supply and one of 240 V or 415 V. Here
again there must be a common point of reference between the voltage supplies
which are to be compared. A 'null' method is insufficient - for example, if volts
were only measured R - R , Y - Y and B-B, no reading would be obtained if the
compared voltages were identical in magnitude and phase. Such a test could indi-
cate perfect phasing; but it could also mean that the readings were not being
434 Testing,commissioning and management of protection

V.T. ~lo. 1 V.T. No. 2

meter current Y

B E a r t h e d v.t. y e a r t h e d v.t.

B
Fig. 18.4.5A Danger of "null" method of phasing R-R, Y-Y, and B-B

obtained because the two voltage supplies had no common point of reference. It
would thus be possible for the supplies to be out-of-phase, yet no voltage reading
would be obtained to indicate that fact. Fig. 18.4.5A illustrates this.
If the two supplies have no common reference point, one must be arranged
either b y temporarily earthing the unearthed supply on one phase or at the neutral,
or by temporarily strapping the two systems together at one point.
To get a positive check, the red phase of one supply should be compared with
red, yellow and blue of the other; then the yellow phase of the supply should be
compared with the other three phases, and finally the blue. The results may be
recorded on a schedule as shown in Fig. 18.4.5B.
Measured voltages Pha.~l}r~ ph,tted Actual phasing
Example I
RIR 2 RI
R2
R2 Y2 B2

RI 2 109 I10
YI 109 2 III
BI I10 III 1.5 BI B2 Y I Y2

Example 2
R2 Hi RI k2
R2 Y2 B2

RI 34 90 123 BI ~ ~ y Y2
YI 124 33 90 B - YI
B I 90 12 .} 34 I B2
B2

RI RI
I-:xample 3
R2 Y2 B2
R2 R2
RI 210 240 310 Y2
YI 350 240 310
BI 210 240 130

<
Fig. 18.4.5B Voltage measurement, plotting of phasors, and determination of phasing
 Testing, commissioning and management of protection 435

When a circuit is being paralleled for the first time, the v.t. associated with that
circuit must be energised from a known source to prove that the v.t. connections
are correct. After this has been done the voltmeter should be left connected, and
then when the new v.t. is energised from an unknown source and synchronism is
indicated, the two primary supplies must be in phase.
Although a phase rotation meter may be used in the case of a generator, it is a
more definite check to employ two voltmeters connected red to red and blue to
blue between the generator v.t. (akeady proved) and a v.t. on the system. The two
voltmeter readings should beat up and down together at slip frequency, thus
indicating that the phase rotation is correct, and they should both be at zero when
the synchroscope needle is at 12 o'clock. Having checked this, the synchroscope
can then be relied upon for closing in.

18.4.6 Closing up

The settings of overcurrent relays should be set down for paralleling or for switch-
ing in new equipment to ensure that if it is faulty, the fault will be disconnected
with the minimum disturbance to the system. In this connection it is more
important to set down the time multiplier than the plug setting. If the latter is set
down then it must be restored to normal before load in excess of the temporary
relay setting is picked up.

18.4.7 On-load tests

Having got the new equipment on load it is advisable to do certain tests. These have
been touched on earlier, but are here given in more detail.

(a) Currents in relays: A split plug and ammeter is a convenient way of checking
currents in relays fitted with plug bridges. The ammeter must be on a suitable range
and a split plug should be inserted on the same, or higher, tap as the relay setting -
otherwise there is the possibility that the relay will operate if the circuit is heavily
loaded.
It should be checked that there is negligible spill current in any residually.
connected relays and in any protection working on the differential principle. Before
measuring current in the c.t. circuits of differential relays it is advisable to remove
the appropriate trip link.

Directional relays
(b) Phase-fault relays: On4oad tests of directional relays are required to prove the
polarity of the v.t. connection with respect to the c.t. connection.
For these tests it cannot be stressed too strongly that the direction of the MW
and MVArs must be positively known, also the power factor. Any load variation
during the tests must be closely supervised - especially a change from a lagging to
a leading power factor.
436 Testing,commissioning and management of protection

Primary injection tests and secondary wiring checks should already have
established that the correct currents and voltages are associated with the
appropriate elements. If proven instruments are already available on the primary
circuit concerned, the load characteristics will be known; if not, it will be necessary
to do 'wattmeter tests'. In these tests a set of vectors is plotted by which the
position of the current relative to the voltage can be seen. A centre-zero wattmeter
is used and the current-coil inserted in each phase in turn.
The wattful component of the red current is measured with R - N , Y - N and, for
good measure, B - N volts. Some of the readings will be positive and some negative.
These readings, plotted along the voltage vectors in the appropriate direction, will
determine the position of the current vector. Fig. 18.4.7A shows a typical result.
vR

l Volts Wattt'ul comp()nent

+1 1.2 ~ IR R -- N ÷! i.2
Y 20

ii ~ .
Note:
....
( I ) Plot the three voltage phasc)rs
VR, V y , VB, and their continuation
in the reverse direction.
(2)With the wattmeter current coil
in the Red phase, measure the watts
for R - N. Y - N, and B - N volts.
(3) Plot these as shown' the point
VB Vy where the three projections meet
gives the position of the Red current

Fig. 1 8 . 4 . 7 A Wattmeter tests on directional relays to determine load a n d p o w e r f a c t o r

All that now remains is to check that each phase-fault directional relay is be-
having in the way it should. To check this it will be necessary to know the type of
directional connection (30 °, 90 °, or 90 ° - 45 °) applied to the relay. It will also be
necessary to know the polar characteristics of the relay; if these have not been
furnished by the manufacture they should have been obtained during secondary
injection tests by applying to the relay voltages of varying phase angles relative to
the injection current, and finding at what angle it has minimum torque.
Thus with all this information available, it will be possible to predict exactly
what the directional relay should be doing for any load condition; and it must now
be proved that it is doing the right thing for the particular load at the time. Normally
a phase-fault directional relay is connected so that it will operate for a flow of fault
power out from the busbar.
If separate phase-fault directional elements are used for each phase they should
all behave in the same way under the same conditions, but if the three directional
elements drive on to a common disc and spindle movement, each should be proved
separately whilst the other two elements are temporarily disconnected.
 Testing, commissioning and management of protection 437

It cannot be stressed too strongly that phase-fault directional relays have not
been proved unless the following conditions have been satisfied:

(i) The type of directional connection should be known.

(ii) The relay characteristics should be known - that is, its polar curve.
(iii) The direction and power factor of the load must be known without the
slightest doubt.
(iv) Each phase-fault element in turn must then be seen to operate in the ex-
pected way for that load condition. This establishes that the c.t. and v.t.
polarities are correct.

(c) Earth-fault directional relays: Directional earth-fault elements are more

difficult to prove in respect of c.t. and v.t. polarities because the polarising volts
and amps will not be there under healthy load conditions. Fault conditions, there-
fore, have to be simulated.
On the load test it must first be checked that there is no open-delta voltage or
residual current under this condition. The most convenient way of proving that a
70 ° relay will correctly respond to an earth fault on or beyond the protected line
(as in (b) of Fig. 18.4.4A), is by short-circuiting and disconnecting the blue phase
current transformer, removing the red phase v.t. fuse, and strapping the dead side
of the fuse to the neutral connection. Section (e) of Fig. 18.4.4A indicates the
resultant vectors and similarity to (b). The relay should operate for load-current
phase-angles between about 90 ° lag and 60 ° lead.
It is not strictly necessary to make the test cyclically round the phases, since
zero outputs for balanced input have already been checked; the other phases could,
however, be proved for good measure.

18.4.8 Modifications to existing substations

All the tests so far enumerated have applied to new substations. Although possibly
complex, the commissioning of a new substation is straightforward compared with
major modifications to an existing site. In the latter case, much of the equipment
has to be kept in commission during reconstruction which may involve reallocation
of multicore cables, the connecting in or reconnecting of c.t.s and alterations to
relay panels. Considerable ingenuity is needed here and much improvisation.
Essential tripping and closing supplies have to be maintained during wholesale
reorganisation, and much temporary cabling and wiring may have to be run to
effect a smooth changeover.
It is impossible to stipulate any procedure for this class of work beyond stressing
the need for particular care and advance planning.

18.5 Routine maintenance tests

Protective gear may stand quiescent for months and yet be required to operate with
438 Testing,commissioning and management of protection

precision if a fault occurs on its associated primary equipment. Perhaps a more

difficult requirement of the protection is that it must remain inoperative on nearby
external faults. Routine maintenance tests on protective gear are done to ensure that
the equipment is always ready to perform in a fully discriminative manner the duty
required of it.

18.5.1 Causes and effects of deterioration

Protective gear does not normally deteriorate with usage, but can be harmed by
adverse conditions. For example, a relay subjected to pronounced continuous
vibration can suffer damage to its pivots or bearings; dampness in junction boxes
and circuit-breaker kiosks can lower the insulation resistance of the multi-core
cables and wiring; a polluted atmosphere may have a detrimental effect on relay
ligaments, relay contacts and auxiliary switches; the heat produced by continuously-
energised coils can often age the insulation; electrolysis may cause 'green spot',
leading to open-circuited coils or contacts.
Measures to alleviate or prevent many of these causes are taken in the design
stage of modern equipment. Outdoor terminal boxes are ventilated, heated, and
proof against vermin. Access to terminal boxes is made easier by the provision
of hinged or slide-off doors, instead of bolted covers; new relays are subjected to
vibration type tests to prove that they will stand reasonably heavy ill-treatment
without maloperation or damage; insulating materials are chosen to stand heat or
damp without deterioration; and electrolytic corrosion can be prevented on d.c.
circuits by using biasing equipment on the battery, thus keeping the whole of the
d.c. wiring at a negative potential relative to earth.

18.5.2 Frequency of routine maintenance

When determining a programme, the frequency of routine maintenance inspections

and tests will depend on the fault history and fault liability of the equipment.
There is no point in overtesting equipment; the programme should be planned so
that faults in the equipment are anticipated, not precipitated. The frequency of test
will, however, vary widely with the type of equipment. Certain items are continu-
ously monitored, some are checked several times a day, whilst others may be tested
monthly, yearly or even every two years. Typical examples of the frequency and
types of equipment are"

(a) Continuous monitoring

Pilot supervision
Trip circuit supervision
Relay voltage supervision
Battery earth-fault supervision
Busbar protection c.t. circuit supervision
 Testing, commissioning and management of protection 439

(b) Daily
Relay flags inspected on every shift
Carrier protection tests -either manually every shift or by clock-test every 4 hours

(c) Monthly
Water level of liquid earthing resistances

(d) Two-monthly
Intertripping channel tests without tripping any switches

(e) Six-monthly
Tripping tests

(f) Yearly
Check operating levels, sensitivites, tripping angle and reflex test facilities of phase
comparison carrier protection
Secondary injection on the more complex forms of protection
Insulation resistance tests
Check of battery biasing equipment
Injection of gas and oil actuated relays with air or oil
Calibration check on liquid earthing resistances

(g) Two-yearly
Secondary injection tests on the more robust relays

18.5.3 Inspections and tests

The main features of the various inspections and tests are briefly indicated below:

(a) Inspections: On outdoor equipment, check that heaters are on and that vents
are not blocked. Check there is no evidence of vermin in relay panels and in
switchgear terminal boxes; check relay settings against schedules or cards; inspect
relay contacts and ligaments for correct positioning and corrosion; inspect relays
for foreign bodies such as dust, specks of iron f'ding in magnet gaps, flaking of
plated parts; voltmeter check to prove that battery biasing equipment is in order,
etc.

(b) Intertripping tests: These can be a prolific source of error especially on three-
or four-point circuits. Particular care is needed to be on the correct panel. A detailed
procedure should be established for carrying out these tests, the test links, if any,
being changed over at all points of test before hand. Care must be taken to restore
the links when the tests are completed.

(c) Tripping tests: The complete sequence of tripping should be proved from the
440 Testing,commissioning and management of protection

protective relays to the tripping of the circuit breaker. The relays are usually
operated manually, but with particularly delicate relays there is some advantage in
actuating them electrically. When relay covers are removed prior to these tests, any
dust should be wiped off the top of the cover before removing it so that dust does
not fall into the relay. The opportunity should be taken on these tests to prove that
the relay contacts are clean and secure, and to observe the contact wipe. The flags
should be checked for positive operation, and the appropriate alarms should be
proved at the same time.

(d) Insulation resistance tests: These may be done with a 500 V Megger, on one
secondary circuit at a time with all the others earthed. The results obtained may
vary with the weather, but it is the trend rather than the absolute value which is
important.

(e) Secondary injection tests: Any variation of operating or resetting times or

value could indicate a cracked or sticking bearing. Before secondary injection
maintenance on relays, the terminal boards and relay stems should be brushed
down with an insulated brush or insulated vacuum cleaner nozzle.

(f) Recording of test results and progress: It is important to keep a record of the
progress of the results of routine maintenance tests. The methods adopted will vary
to suit the particular organisation, but in principle should show the frequency of
any test, when it is due, when it was done, and if it becomes overdue. The tests
results can best be recorded on test sheets specifically made up for the different
kinds of tests and protection. One method in use has a complete binder of such test
sheets per station, and each sheet has enough spaces to cover several years' testing
so that the results of past tests can be seen at a glance and compared. The test
sheets are made out in schedule form, in such a way that the sheet is not only a
record but a guide to what should be done on the tests. Any defects found in the
equipment should be recorded. Defects of a design nature should be reported to
the manufacturers so that they can improve their design.

(g) General. In this section on routine maintenance tests, no attempt has been
made to describe any test in detail; the philosophy of such tests has, however, been
fully discussed. Several further important points can be made to conclude this
section: one should be familiar with the equipment on which work is to be done,
and full use should be made of appropriate diagrams and test schedules. It may be
helpful to prepare a list of items disturbed during the tests so that the equipment
can be quickly restored to normal without anything being overlooked. There is little
point in maintaining equipment in perfect condition if it is put back into commis-
sion on the wrong setting or with a connection left off. In this respect, it should be
noted that the disconnection points provided in certain wiring circuits are intended
for special fault investigation purposes only and should not be used as isolation
points during routine maintenance.
 Testing, commissioning and management of protection 441

18.5.4 Maintenance of busbar protection, back-tripping and circuit-breaker fail

protection at double-busbar type substations

The consequences of busbar protection failing to operate are very serious because
clearance of the fault would then be dependent upon the operation of remote
distance, overcurrent or earth-fault protection. These remote relays would be
unable to discriminate between the healthy and faulty busbar sections resulting in
complete disconnection of the substation with consequent disruption of the system.
Not only would any generating plant at the faulted station be lost, but it is also
possible that due to the long fault-clearance time involved other generators in the
close vicinity would lose stability and that low-frequency protection would operate
causing further interruption of supply. Therefore it is of paramount importance that
busbar protection is maintained regularly to prove its integrity. Owing to difficulties
involved in testing busbar protection, and the consequences of inadvertent opera-
tion during such testing, it has been the practice to take the protection out of
service when carrying out maintenance. However, for the reasons stated above long
outages of the protection can no longer be tolerated and techniques have been
evolved for testing the protection in service except when checking insulation
resistance of d.c. circuits.
From a study of Chapter 13 it will be seen that busbar protection, back-tripping
and circuit-breaker fail protection are closely interrelated and test procedures must
overlap all three facilities. It is usually permissible to take back tripping out of
service to simplify test procedures, because the risks to the system are not so great
as when busbar protection is out of service. Circuit-breaker fail protection can be
tested when a circuit is out for routine maintenance.
The test procedures which follow are based upon the 400 kV busbar protection,
back-tripping and circuit-breaker fail schemes shown in Chapter 13, but the
principles demonstrated can be adapted to suit other protection arrangements as
required.

(a) Basis of busbar protection & back tn'pping tests: The basis of the testing
procedure is to work on one zone of protection at a time relying upon the 2 out of
2 feature to prevent inadvertent tripping, and then ideally proving operation by
simultaneous secondary injection of the check zone and a discriminating zone to
trip each section of busbar in turn with its associated circuits. The trip outputs of
the back-tripping receive trip relays must be isolated to ensure that the tripping
is initiated from the circuit busbar protection relays and not from the zone relays
via the back-tripping system. Operation of the back-tripping system up to the
trip relay contacts is proved at this time and then the circuits are selected to the
opposite busbar and the zone relays manually operated to prove the correct selec-
tion of the back tripping system via the busbar selector auxiliary switches.
If system conditions are such that it is not acceptable to trip a section of busbar
together with its associated circuits a possible alternative is to trip only bus-coupler
and bus-section circuit breakers leaving the circuits connected to an isolated section
442 Testing,commissioning and management of protection

of busbar. As in the ideal method described in the previous paragraph all circuit
back tripping relays should be operated via both busbar isolators to prove the
complete scheme. For this test procedure tripping, intertripping and protection
unstabilisation or acceleration from the circuit busbar protection trip relays, and
interlocked current trip relays, must also be removed. Tripping of individual circuits
can then be proved by manual operation of the latching type circuit trip relays
during individual circuit maintenance.
In carrying out the tests it is recommended that the following precautions
should be observed:

(i) No switching, earthing or similar work should be allowed within the station
for the duration of the tests.
(ii) Weather conditions should be good at outdoor type substations.
(iii) An agreed testing schedule should be prepared and adhered to.
(iv) A minimum of two engineers experienced in the protection should carry out
the tests and each should check the action of the other.
(v) Autoreclose equipment at the remote feeder ends should be taken out of
service.
(vi) Test points should be clearly identified and caution notices displayed on
equipment not under test.
(vii) The test supply source should be isolated, earth free and of high impedance
with a substantial two pole switch between the injection supply and the test
point.

(b) Test procedures for busbar protection and back tripping

O) Insulation resistance: Insulation resistance to earth of the current transformer

circuits should be measured using a 500 V megger. The instrument should be
applied across the appropriate earth links before they are opened and should not be
removed until the earth link is reclosed. The insulation resistance of d.c. circuits
must be measured with the appropriate supply links and fuses withdrawn and trip-
ping circuits isolated. Wiring associated with individual circuits should be tested
during routine circuit outages.

(ii) Secondary in/ection: All a.c. and d.c. tripping relays should be visually
checked to ascertain that they have reset before and after injection of each a.c.
relay. Injection tests should be carried out on only one phase of a discriminating or
check zone at a time, increasing in small voltage steps and recording operating
values of supervision, circuit and zone relays in turn noting also that only the relays
associated with the injected phase and zone operate. The duration of current in-
jection should be kept to a minimum to prevent damage to relays and equipment.
It is not necessary to repeat these tests with the circuits connected to the opposite
busbar to prove correct selection of a.c. wiring via busbar selector auxiliary
switches, as these are monitored by the a.c. supervision relays.
 Testing, commissioning and management of protection 443

(iii) Tripping tests: The protection should be operated for each busbar zone in
turn, thereby tripping the associated circuits and operating the back tripping
receive trip relays. Operation should be by simultaneous a.c. injection of the appro-
priate discriminating and check zone relays. The injected voltage should be
approximately 150% of the setting value obtained in the earlier tests. It is not
adequate to operate the a.c. zone relays manually for these tests as the manual
effort achieved may exceed the electrical energy available for operation. This
procedure will have established correct operation of the busbar protection but the
circuits should be connected to the opposite busbar and the zone busbar protection
a.c. relays operated manually, to prove correct selection of the back tripping receive
trip relays by the busbar selector auxiliary switches which are not monitored.
Tripping of the circuit breakers from their back tripping receive trip relays can be
proved during circuit maintenance by manual operation of the latching type relays.

(c) Basic test procedure for circuit breaker fail protection: As described in
Chapter 13, circuit-breaker fail protection opens, via the back tripping system all
the circuit breakers selected to the same busbar as the stuck circuit breaker. After
proving the back tripping system by the test described in Section 18.5.4(b) the
circuit-breaker fail protection can be tested during individual circuit maintenance,
without risk to other circuits, by removing the links which isolate individual circuits
from the back tripping bus wires. With these links removed the circuit-breaker fail
protection can be tested down to its own back tripping receive trip relay. The tests
should include insulation resistance, a.c. secondary injection of current check
relays, d.c. sequence and timing tests.

18.5.5 Maintenance and testing of intertripping and protection signalling

equipment

Type T40 intertripping and T15 protection signalling equipment (see Chapter 7)
have achieved their fast operating times at the expense of simplicity, many more
components being used than in earlier equipments: the circuitry also is much more
complex. The quest for higher speed has also meant that the circuitry is much more
susceptible to air-borne or cable-borne interference and precautions have to be
taken to ensure that interference does not produce maloperation or inhibit or delay
tripping.
Although all modern equipments are not necessarily to T40 or T15 specifications
they tend to be based on the same design using similar techniques and requiring the
same precautions against interference.

(a) Commissioning: A full list of commissioning checks will usually be given in

the manufacturer's hand book but the following aspects, which may not be
included, should also be checked to ensure that:

(i) the equipment is bolted down and securely earthed to the station earth bar
444 Testing,commissioning and management of proWction

(ii) all connections are tight and of the correct polarity

(iii) the cabling is of the correct size, and that twisted pairs are used if specified
(iv) there are no spurious earths on the cabling, or on the cubicle battery supply
leads
(v) the intertrip send relay does not operate with less than the specified mini-
mum operate current
(vi) the equipment, including modules, has no dry soldered joints
(vii) the electrical noise fed back to the battery from the equipment is within the
specification limit. A typical value for this is one not greater than 5 mV
peak to peak over the bandwidth d . c . - 1 MHz. It is important to realise
that such noise on the battery can adversely affect equipment connected to
it, d.c.-d.c, power supply units being the most usual source of noise
(viii) if Post Office line isolation is required this has been fitted and has not in
any way been bypassed.
(ix) no spurious tripping operations occur when withdrawing or inserting the
power supply fuses. This should be checked by using a device which can
respond to pulses as short as 1-2 ms and a suitable device using a reed relay
is described in Section 18.5 ~(d)
(x) the output relay operation is as specified for the protection scheme; for
instance is the delayed autoreclose facility required? Check also that the
relay is operated for the minimum specified time, irrespective of the dura-
tion of operation of the sending relay
(xi) end-to-end signalling times are in accordance with the specification. It is
important to follow the setting up procedure as given in the manufacturer's
handbook for the commissioning tests. End-to-end tripping tests should be
carried out first with the battery voltage normal and then with it at the
specified extremes, i.e. typically on the 48 V battery at 44 and 56 V, in-
eluding tests with high volts at one end and low volts at the other.

(b) Routine maintenance and testing: The equipment handbooks give advice on
the frequency and type of routine tests but the recommended frequency of testing
varies between equipment types. For some equipments even weekly tests are
suggested, albeit level tests carried out with the equipment remaining in service. In
practice the frequency of testing is very much a compromise between the economic
use of resources and the need to keep the probability of an undetected fault as low
as possible.
A commonly accepted maintenance programme is for the equipment to be
subjected to routine tests every two months. However, if the increased reliability
anticipated from the more modern equipment materialises it should be possible to
extend the period between routine tests to six months without detriment. The
recommended programme would then be for routine tests to be carried out at the
end of the first, second, fourth and sixth months after commissioning and there-
after at intervals not greater than six months.
The types of tests to be carried out vary between the different forms of equip-
 Testing, commissioning and management of protection 445

ment but in any case are detailed in the manufacturers' handbooks. They cover
such items as the checking and adjustment of power supply voltages, oscillator
frequencies, monostable timers, signal levels at various test points etc. Finally the
equipment should be end-to-end tested and timed.
Signal transmission time is critically important for intertripping and protection
signalling equipment. For example when protection signalling equipment is used for
blocking, on the occurrence of a through fault a signal must be received from the
remote end in about 20 milliseconds from the detection of the fault, otherwise the
protection will maloperate. Consequently, it is recommended that the equipment is
timed during routine maintenance periods. Push button testing, whereby the send
button is operated and the received signal observed at the remote end, does not give
a real indication of signal transmission time.
Timing can most easily be achieved by a 'reflex' or 'round-the-loop' test. In this,
the send relay at end A is operated by a push key, a contact of which also starts a
digital timer . At the remote end B, the receive contacts are arranged to operate the
send relay and return the signal to end A, the receipt of which stops the timer. The
test is comparatively simple if switches are specially provided for this facility on the
intertripping/protection signalling equipment or on the relay panel. Absence of
such switching arrangements would make the test more difficult and entail the
disconnection of pilot cores from the equipment terminals.
The time recorded for the test is the total for the signal transmission in both
directions: half this time is normally a satisfactory estimate of the time for each
channel. If the recorded time is outside the equipment specification further investi-
gation is required to locate the trouble. It should be remembered that the measured
time also includes the line transmission time but this is unlikely to be greater than
about 1 ms for 22 km of cable length, which is the time for signal propagation
along standard loaded cable. The propagation time for carrier routes should be
much less.
(c) Fault investigation: The recommended fault investigation procedures for solid
state equipment as outlined in Section 18.6.3 are applicable to these types of equip-
ment.

(d) Trip monitor device: This device mentioned in Section 18.5.5(a)(ix)may be

used during the commissioning and testing of intertripping equipments, and is easily
constructed. The circuit diagram is shown in Fig. 18.5.5 A.
The trip receive contacts which can either be clean i.e. clear of power supply, or
supplied from 48 (or 110) V d.c. batteries are connected to terminals A-B. A 48 V
power supply is connected to terminals C-D. (In either case the polarity is im-
material because of diode bridges D1-D4 and D5-D8). If the trip relay contacts are
connected to a battery then key switch KA should be selected to the position
shown in the diagram. When the trip receive relay operates, relay RA is energised
and locks in over its own contact to the second coil. The audible alarm is given and
the light emitting diode (LED) is lit. When the trip contacts have reset the alarm
can be reset by momentarily operating the reset key.
446 Testing,commissioning and management of protection

i_--,.-,d OA
i
~ i)10

.k_a 4~/I Io v Clean I rip

,
__ COl|I~lC|S

4 8 / i l 0 V KA
KA

Clean

1
KC"
Reset
Or"
Alarm
silence
AI) ~CI d.c.
O D
I
r .L:
~ l ~oo I-I
° [ 772

Fig. 1 8 . 5 . 5 A Trip monitor device

If long term supervision of a trip is required the alarm silence key can be operated
to disconnect the audible alarm: a trip receive will still be indicated on the LED.
If the reset key is changed to the operated position, the alarm will automatically
reset when the trip contacts restore. With fleeting trip conditions the audible alarm
sounds and the l.e.d, lights only during the slow operate time of relay RB.
When the trip contacts are clean, key KA should be operated. The circuit to
operate RA is then: positive terminal C or D; diode bridge D5-D8; key KA operated;
terminal B; trip contacts; terminal A; diode bridge D1-D4; RA coil; key KA
operated; negative via diode bridge D5-D8 to terminal C or D.
The equipment is compact and can be fitted in a box typically 110x90 x 55mm.

18.6 Fault investigation

Fault investigation falls under two main headings: the investigation of primary
faults and the investigation of faults on the protective equipment.

18.6.1 Primary faults

In this context the cause of primary faults is ignored, and only the effects of that
fault on the protective gear and secondary equipment are considered.
Faults which are cleared correctly by the protective gear normally need little
investigation. The protective gear engineer should, however, satisfy himself that the
protective gear operations are commensurate with the nature of the primary
trouble. If there is any doubt he should also satisfy himself that the protective gear
 Testing, commissioning and management of protection 447

has not been damaged in any way by the fault conditions to which it has been
subjected,
A suspected maloperation of protective gear should, however, be investigated in
the greatest detail.
Greatest care should be taken to collect and consider all the relevant evidence
and to assess it objectively.
Much can be gleaned from recording voltmeters and ammeters, stopped clocks,
automatic fault-recording oscillographs, reports of flashes, vibration, noises, alarms,
and dips on the lights.

18.6.2 Faults on the protective equipment

Many faults on protective equipment are simple and self-evident, such as low
insulation resistance, badly adjusted flags, and relays out of calibration. Intermit-
tent faults, or failure of a switch to trip when tripping tests are being done are often
more difficult to diagnose.
It is important that the investigation of the fault should be carried out in such a
way as not to destroy any evidence.
The most likely sources of trouble should be looked for first. In the case of a
circuit breaker failing to trip, these should include low battery voltage, faulty plug
and socket connectors, auxiliary switches, burnt-out trip coils, local]remote selectors
in the wrong position, and some kind of mechanical failure. If a quick visual inspec-
tion reveals none of these failures, then electrical tests should be made. Such tests
should cover checks of the trip supply, fuses and links and the contact resistance of
the auxiliary switch and the local/remote selector, and the continuity of the trip
coil circuit. These points should be methodically checked with an ohmmeter or
voltmeter - n o t by pulling out fuses or disturbing any contacts. By checking
through the circuit with an ohmmeter or voltmeter the faulty part can be identified
with certainty. The trouble can then be remedied, and one is then certain that the
reason for the failure has been found and cleared.
Sometimes the investigation may be lengthy and involved. In this case keep a
chronological note of the items or parts of circuit tested and the conditions apply-
ing at the time those tests were done. This will help to avoid the confusion of think-
ing that a different result was obtained an hour earlier when the 'same' tests were
done. In fact, an auxiliary switch, selector switch, or indicating lamp may now be in
a different position from what it was before, thus altering the conditions
completely.

18.6.3 Faults on solid-state equipment

The procedures previously described for commissioning and routine maintenance

of conventional equipment are, in the main, also applicable to solid-state protection
but the following additional points should be borne in mind.
448 Testing,commissioning and management of protection

(a) Insulation resistance tests should not be carried out at a voltage in excess of
500 V.
(b) During such tests diodes and transistors which might sustain damage should
be shorted out.

Failures in solid-state equipment can usually be traced to a particular printed circuit

card or module with the aid of the manufacturer's test procedures. Replacement of
a faulty card or module, which usually has plug connectors, enables the equipment
to be restored to service with a minimum of delay. Tracing and replacement of
faulty components on the printed circuit card requires specialist knowledge and
equipment and is best carried out at a central depot or at the manufacturer's works.
Owing to difficulties in proving the card after repair unless a complete equipment
identical to the one from which the faulted card originated is available, the latter
course is to be preferred. A further point in its favour is that replacement com-
ponents used by manufacturers will usually have been selected by a process of soak
and heat tests before use. These tests eliminate most early failures and hence
improve the subsequent reliability of the equipment. If in an emergency
components from local suppliers are used they should be replaced by components
supplied by the equipment manufacturer at the earliest opportunity.
In order to achieve quick restoration to service of faulted equipment, a suitable
stock of spare cards and modules should be available. Holding such spares is ex-
pensive and this prohibits their provision at each equipment location. An
economical arrangement is to hold spares at a central location to serve a particular
area. As a general principle an overall spares holding, based on population, of 10%
or one of each item whichever is the greater should prove adequate to meet most
situations. This will obviously be subject to some variation in the knowledge of
experience with particular equipments. Certain cards and modules are frequency
conscious and in some cases it would be very difficult to justify holding one of each
item. In such circumstances, consideration should be given to holding stocks of
components to permit the adaptation of cards for a variety of frequencies.

18.7 The avoidance of errors when testing

18.7.1 General

The greatest care is necessary when making tests on site to ensure safety to life and
security of supply. Every effort must be made to avoid testing errors, but the list of
precautions can never be complete.

(a) The correct equipment: Make sure that the correct equipment is approached.
This sounds very easy, but on suites of panels or on panels which are badly laid out
it is quite possible to become confused. A common cause of error is for the correct
panel to be identified on the front, and the wrong one on the back; or vice-versa.
Ways of preventing such errors are the use of temporary screens or rope barriers,
 Testing, commissioning and management of protection 449

PROTECTIVE & CONTROLGEAR TESTSON SITE

REMINDER SHEET
Station : ................................... Equipment'.............................................................. Date'. ..........................................
Brief Descriptim of Work: .................................................................................................................................................
Reasonfor test' ....................................................................... Engineerin charge of test: .............................................
A tick sEould be placed against the items disfu,bed

EQUIPMENT Distu,bed NOTES Restored

..
Trip Link or Fuse
Intertrip Test Links
c.P.o. P,'O~L~nks
Battery Biasini] Link
C.T. r~ Terminal Board Links l L

VoitaEe Fuses
II
Test Switches
Wirinl' (a) Leads off
(b) T est le~s on'
(C) TemporaryStraps

Relays: (a) Settings

(b) Wedges
(c) Coils Shorted
(d) Stabilisinz Resistances
(e) Metrosils ....

/~uto Rec=o3e
Automatic Voltaie Control
Local/Remote Selectors
Standby/RemoteSelectus l
P,usbar Pwotection(a) DiscriminatinzZones
(b) Check Zone(s)
Adjacent Circuits ?
RemoteSubstations? II

GENERAL TESTING HINTS

Identify correct equipment Check separately and independently

Instructions - clear and concise Test leads (a) Continuous (b) Well-insulated
Diagrams up to dote ? Care with multi-range instruments
Special vigilance if programme is changed Danger of "un.progrommed extra"

THINK before taking action

Final check-up after tests
FURTHER DETAILS OF ABNORMALITIES MAY BE WRITTEN OVERLEAF
Fig, 18.7,1A Reminder sheet

the application of warning labels to be panels (back and front), on either side of the
panel to be worked on, and the locking of those relay cubicles not being worked
Oil.

(b) List of abnormalities: A list should be made of all things disturbed during
testing, such as wiring disconnected, test leads applied, relay settings altered, paper
wedges in relays, trip links removed, voltage fuses removed, and so on. Such a list
 450 Testing,commissioning and management of protection

helps restoration to normal to be made quickly and ensures that nothing is over-
looked. A standard form can be used for this purpose, a typical form being shown
in Fig. 18.7.1 A.

(c) Primary injection tests: During primary injection tests:

(i) It must be ensured that the primary equipment is dead and safe to work upon.
(ii) The diagrams should be studied and c.t.s through which primary current will
be passed should be noted. The necessary precautions should be taken to
safeguard the security of adjacent circuits (see section relating to primary
injection tests).
(i~i) List any abnormalities as in (b) above.
(iv) Ensure that no c.t.s are open-circuited.
(v) Check that the test ammeters are on the correct ranges.
(~) After tests are completed, check that all test leads are removed and that
everything is back to normal.

(d) Relay wedges: During tests it is often necessary to wedge relays in the
operated or non~perated position. In induction disc relays such wedges should be
placed under the disc (so as not to apply excessive pressure to the bottom bearing).
They should be conspicuous so that their presence cannot be overlooked.

(e) Test leads and clips: Test leads should be periodically checked for insulation
and continuity so that no mishaps occur due to faulty leads.
Test clips of the type which have a relatively large amount of bare metal exposed
should not be used in congested locations where one clip could short across two
relay studs.

(f) Instructions: Where it is necessary during tests to pass instructions from one
to the other, either directly or by telephone, care should be taken to ensure that
such messages are thoroughly understood. A telephoned instruction in particular
should be repeated by the recipient to make sure that the message has been
properly received and understood.

(g) Responsibility for tests: Where it is necessary for two or more engineers to
work together on a series of tests, it must be agreed beforehand who is in charge.
No action should be taken without the knowledge of that person.

(h) Multipurpose instruments: If a multipurpose instrument is being used, great

care should be taken to ensure that it is on the correct range for the measurements
being made. Failure to do this may result in a c.t. being effectively open-circuited,
or in an internal flashover which may wreck the instrument and injure the user.

(i) One source of supply: No work should be done on any equipment which
 Testing, commissioning and management of protection 451

controls the only source of supply to consumers.

(1") Final check: Before leaving the site get into the habit of having a careful
look round. Think back over the work that has been done to make sure that every-
thing has been restored to normal. Although advice can be proffered and various
devices used on site to reduce the chances of error, it is the continual practising of
self-discipline that will do most to keep testing errors to a minimum.

18.8 The test equipment

Provided one is prepared to operate relays manually on a tripping test, no equip-

ment is required for such a test. Only a megger is required for insulation resistance
tests. Secondary injection tests, however, are in a different category, and opinions
differ as to the best selection of equipment required for such tests.
Much can be done with a slider type resistor, an ammeter, a transformer, a
switch and a stop-watch. The assembly of these, however, possibly several times a
week, is tedious and inefficient and it is desirable to have such portable equipment
already assembled in boxes to which external connections only need be made.
The design of such a test set depends on the type of protection for which it is
needed, but considerable flexibility should be built into it to cater for all types of
protection and current ranges.
One principle should be kept firmly in mind; one is not trying to set up a
standardising laboratory on site, nor trying to reproduce the type approval tests.
The purpose of portable secondary injection equipment is to confirm as far as
is practicable that the protection is in order and that it keeps in calibration within
limits during the years which follow commissioning.
A supply is needed and a means of connecting that supply to the test equipment.
A supply may also be required for a hand lamp, soldering iron or an oscilloscope, so
some thought should be given to the provision of a test supply box to which one
supply can be fed but from which a number of supplies can be taken.
In the test equipment some current control is necessary, the ranges usually being
from 0 to 3 A on 'mains direct' or from 0 to 20 A through a transformer. Some
instrumentation is needed, together with switches for controlling the supply; and a
timing device to measure speed of operation. For distance protection or directional
relays a voltage supply is also necessary, again controllable. The secondary injection
test equipment is then completed by sundry test leads, alligator clips, split plugs
and a pear switch. Each assembly should be portable, stackable and free from
excrescences, and the requirements are discussed in more detail under the appro-
priateheadings below.

(a) Test supply: For full flexibility this should be three-phase, four-wire, 415
volts of a sinusoidal waveshape and adequately protected. The greatest care must be
taken when connecting to that supply to ensure that it is not accidentally earthed
or short-circuited.
452 Testing,commissioning and management of protection

. , J. .

, ° .

Fig. 18.8A Test g e a r - test supply box

A three-pole and neutral switch specially and permanently allocated for test
supplies should be provided in each relay room. Leads pushed into sockets and
wedged in by match sticks are strongly deprecated.

(b) Test supply lead." This should preferably be four-core flexible t.r.s or p.v.c.
cable of ample length to connect from the supply to the test supply box located at
the point of test.
The connection of the cable on the test supply box should be designed to
prevent strain on the terminations.
The cable may be accommodated in a reel for easy unwinding and winding.
 Testing, commissioning and management of protection 453

" F~'L R Y B N

- I:,( " -

Fig. 18.8B Test g e a r - t e s t s u p p l y box

(c) Test supply box: The 415 V, three-phase, four-wire supply entering the test
supply box is fused on entry and connects a four-pole main switch to four insulated
output terminals. The box also contains a 240 V pilot lamp (connected between
one phase and neutral), a 240 V 5 A socket, a 240/110/25/3 V transformer, and
sockets for a hand lamp and a soldering iron. A two-pin socket is provided for con-
necting a 3 V lamp used for relay inspection. A photograph of one such box is given
in Fig. 18.8A, and Fig. 18.8B shows its connections.

(d) Current box: The current box derives its supply from the test supply box;
this is achieved by an insulated bar containing four slotted brass 'spades' feeding a
length of four-core cable running to the current box. The slotted brass spades easily
slip under the insulated terminals on the output side of the test supply box.
Only one phase is used in the current box, but all three phases are brought to an
arrangement of three sockets to provide selection of the phase required. The voltage
box, described later, obtains its supply from the three phases of the current box.
The selected single-phase supply is fed through a switch or contactor to a resistor•
potentiometer network and thence through a 'fine control' resistor to the output
terminals for the 0-3A range. A two-way switch on the output side of the 'fine
control' resistor enables the supply to be switched to 'mains direct' or alternatively
to the primary of an external 240/30 V 20 A transformer. A pear switch on the end
of a long lead enables the current to be switched whilst a relay high up on a panel is
observed or timed. The current box is shown in Fig. 18.8C and a diagram of its
connections in Fig. 18.8D.
When using the 240/30 V transformer to test relays, it will be appreciated that
 454 Testing,commissioning and management of protection

.,:~:, • , • . . . . . . . .....,,,;:. ,.~

.................... - ~'.~ ,, • ...... ,,,.i. -.~,i

~I[.C~9.,$1F.~A.R°¢
W~'[~TIt~N T[~'[r ,~f,.,r. ~I,,I~IIIAT~ DIA~':~¢J@I~L :
.:,.-...:.~,,,,,~ .......: ~ ~.:.: .. :. ...................... : . .............. ,a~,ll:.tl I

Fig. 18.8C Test gear - current box

if too much resistance is introduced in series with the primary of the transformer
some distortion of output wave form will result. On the other hand, the resistor in
series with the relay on the 'mains direct' side will give a good waveshape - pro-
vided, of course, that the voltage supply is sinusoidal to begin with.
A small 2401110 V transformer is included in the current box to give a supply
for a timing clock.
 Testing, commissioning and management of protection 455

The external 240/30 V 20 A transformer draws its supply from two pins of a
four-pin socket mounted in the side of the current box, and feeds the 30 V output
back into the other two pins on the same socket, from where it runs to output
terminals for testing 5 A relays.

(e) Voltage box: This plugs into the current box on a multipoint plug containing
the three-phase four-wire 415 V supply and various contactor leads. Two selector
switches allow any pairs of phases or any phase-to-neutral to be applied separately

............ ;;7 . . . . . . . . . . . . . . . . . . . . . . 7: ............. -..,,,,- ;, . . . . . . . . . . . . . . . ::~:"::'''.

..

......... ~ < .~ " "'" ..... ~ . .,. x....: .,~ ~ 'i:

i . . ........ , ~ ; .:'.......... 1 ...... '!, 7.i~ ,.?..... ~,

i.:::::+.>F'.:.;;:;.ii '~ ' !-"-" ......... . {. ;;",.. ;.~ ....... ).../L:;:~...,Jl .....

.... ~,'--~'~ ~~.i:~"" - : .................................

;
;. i... ..... !:. ';~;:. ........................ " "
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . . . . . . . . . :>. . . . . . . . . . . . . . .

Fig. 18.8E Test gear - voltage box

456 Testing,commissioning and management of protection

O.c.b. control
O O
"I'"
I
If"--"
I I
I Local
= ==-__JJ [ test

I
TEST
SUPPLY L .....
BO X

i "~1:~ I I I Timinl~trlake
Clock
supply
switch

415V
Supply VOLrA(;II UNIT
switch
Restraint
' circuit
415 II0

~ 3Q
240 ~ l0
-- 0

o
I
f- off

[ T,...,o,~~, 1
m "' .... " I ,oo~^,,,,o~
' ~IL circuit 415 _ I I0 _

I~, ['Z. 30

--Jr- 240 0 Variac

off 115V/I35V

Fig. 18.8D Secondary injection test set circuit diagram

 Testing, commissioning and management of protection 457

~ Ohms Substitute meier

O- / 4800 O.S A
o- ~ _ 2,00 c/o o s ^ ~.,,,~k 9 s.cke. Q o.s ^
~-- ~
~,,oo ~
SWitch
_~_ ! ~ !
poin! ch..s~...... ~

o- I ~ ooo ~~._~o,~ ~3,,~

°'l. o 3oo I Live O-,,A ~'k ~" "l~r~
O,= in

~. itute
22 e ~r
SA
34 ,)hms S A check 9 suckers 9 SA
I..022-- 2 A I point 1 1,~., . . . . .
22 ~ " " - k;
22

I~ '2 I.~1¢:ol ~ '*

F:xternal
a~Sv~itch II L _ sockets
• I I

N~utral link

i
Restrain1
adjustment 600 IZ

600 l~

i Relay
restraint
circuits

Set
restraint

I -

VI Ranlte V2 Range " 1 ) ~ R. . . . . . d O

0-30 0-10 a ~ O ~ Relay
O-lSO V 0--50 V
Norm d i : : : : i s hal

,, r ~ ii ..... o
458 Testing,commissioning and management of protection

to internal transformers and Variacs for feeding:

(i) relay directional circuits,

(ii) relay restraint circuits (collapsible to simulate fault conditions).
Fig. 18.8 E shows the voltage box, and Fig. 18.8 D shows its connections.

(f) Pn'mary injection equipment: This equipment is designed to pass a heavy

current at low voltage through the primaries of current transformers.
One in common use is of 5 kVA intermittent rating; it has both the primary and
secondary windings arranged in sections in such a way that the primary can be fed
from a 240 V or 415 V supply, and the secondary can give an output voltage of 7 V
or 14 V at a current of about 500 A. The current drawn from the mains may, for
short periods, be as high as 30 A. Control of the primary circuit can be provided by
a 5 kVA Variac on the primary side of the transformer.
Heavy flexible test leads with clamps for fastening to the primary connections
are necessary, and a measuring c.t. and ammeter are needed to determine the
output current. Because of the low voltage and heavy current, it is essential that all
connections are clean and tight.
The equipment (f) above enables primary currents to be simulated by which it
can be proved whether the c.t.s are correctly connected to the protection. The
relays themselves can be injected with smaller currents from the secondary injection
test equipment and proved to be in calibration. Thus, by the use of this test equip-
ment the whole chain can be proved from the initial primary current to the final
closing of the protective relay contact.

(g) Instruments and other devices: In conjunction with the above test equipment,
a variety of instruments and other devices may be used. In considering instruments
one should consider what quantity they are required to measure and what they are
calibrated to measure. Instruments of the moving iron, dynamometer and hot-wire
type will read correctly on d.c.; they will also read correctly on a.c. because they
read the r.m.s, value of the current or voltage.
A permanent-magnet moving-coil instrument, however, will incorporate a
rectifier if it is to be used on a.c. and is consequently very sensitive to wave shape.
Such an instrument responds to the average value of a wave, but the scale is cali-
brated to indicate the r.m.s, value. Reference to Fig. 18.8F shows that a sine wave
has a ratio of r.m.s, value to average value of 1.11 (the 'form factor'). Provided,
therefore, the instrument is used on a sine wave it will measure the average but is
calibrated to indicate (correctly) the r.m.s. If, however, the wave shape is non-
sinusoidal because it contains harmonics, the form factor will be different from
1.11; the instrument will thus give a wrong reading. This emphasises the need to
ensure that the right instrument is used for any given application. It is advisable to
have the calibration of all test instruments periodically checked.

(i) Multipurpose instruments: Many of these instruments incorporate rectifiers

 Testing, commissioning and management of protection 459

M.S. value

I ~erage value

., | ,

0 _E.
n n n 5rr 3rr
2 2

I'{}rm factor = I . ! i

=
R.M.S. value
- = . . . .

-" i A i e r a g e value

t_ j L .... I I

0 ~
2
/
rr~rr --.-- 2rr 5_._.~~n
n2
I'{}rm factor = 1.3 3
Rectifier i n s t r u m e n t error
on RMS -: ~_1.34 - 1 . 1 1 ) x 100'
1.34
= 17.1s%

Fig. 1 8 . 8 F The effect o f waveform on measurements

which are in circuit when on the a.c. ranges, so consideration should be given to
waveshape as already discussed.
On these instruments there is often a 'divide by two' button which doubles the
reading if this facility is needed for any reason. Care should be taken to ensure that
the button does not stick down and so give too high a reading.
The instrument has different impedances depending on the range in use. It is
always desirable when using the instrument on the voltage ranges to have its
impedance high compared with any other resistance in the circuit, otherwise the
same voltage may be read differently on different ranges of the meter. The number
of ohms-per-volt is a measure of the current consumed by the instrument, and is
usually marked on the back of the instrument or is given in the literature.
An easy way of calculating the current drawn by a voltmeter at full-scale deflec-
tion is to apply the formula:
1000
mA for full-scale deflection =
ohms per volt
 460 Testing, commissioning and management o f protection

rJ~

C ,, . . . . . , . . . . ,, _ , _ _ ,

Fig. 18.8G Load Z connected to supply

~. V A

, ,

" ]

• _ . . . . . . . . . . . _ - . _ _

Current measured is correct.

Load voltage is less (by VA) than
the voltage m e ~ u r e d .

Fig. 18.8H Voltmeter on supply-side of ammeter

IZ + i V Q IZ

I "

iV

!
Volts measured are correct.
Load current is less (by IV) than
the current measured.

Fig. 18.81 Voltmeter on load-side of ammeter

 Testing, commissioning and management of protection 461

On the current ranges it should again be realised that the meter represents an
impedance in the circuit; this impedance is greatest on the lowest current range. A
high-burden ammeter can upset the distribution of current in c.t. and relay
secondary circuits and may give misleading results when measuring spill currents.

(ii) Ammeters and voltmeters: The points discussed in the previous paragraphs
really cover most of what need be said about the instruments. However, if an
ammeter and voltmeter are being used together to compute accurately the impe-
dance of some circuit it should be remembered that the ammeter has some resistance
and that the voltmeter takes some current. Consider Fig. 18.8G. If it is desired to
measure the impedance of the load circuit denoted by Z, it could be done by
applying a supply and measuring the voltage and the current taken, then dividing
the volts by the amps. But where should the voltmeter be connected if a really
accurate answer is needed? If it is connected as in Fig. 18.8H the ammeter will
read the true current, but the voltage across the load would be slightly less than
that read because of the volt-drop of the load current passing through the
impedance of the ammeter. If, however, the voltmeter is connected as in Fig. 18.8I
it will read the true voltage across the load, but the ammeter will read the load
current plus the small current taken by the voltmeter.
In most circuits it makes negligible different which method is used. If the
current is heavy then the method of Fig. 18.8I is better because the voltmeter
current is infinitesimal compared with the load current. If, however, the load
current is small, Fig. 18.8H would be the best method because the volt-drop across
the ammeter is negligible compared with the supply voltage.
These examples are quoted as a reminder that accurate resistance measurement
is not always as easy as it first appears. The way to connect the instrument is, how-
over, largely a matter of common sense, and the instrument position can if
necessary be allowed for in the calculations.
Finally, brief mention might be made of valve-voltmeters which have virtually an
infinite impedance and therefore give a true reading even in high-impedance
circuits.

(iii) Wattmeters: Wattmeters have a current coil and a voltage coil. Often the
voltage coil has a resistor in series with it, and if accurate measurements are
required it is important to see that the meter is connected up correctly in the way
already discussed for ammeters and voltmeters. If extremely accurate measurements
are required it may be necessary to allow for t h e / ~ R losses in the current and
voltage coils.
Another point to be watched is the potential of the current and voltage coils. In
Fig. 18.8J the current and voltage coils are at the same potential. If, however, the
connections were reversed, as in Fig. 18.8K, there is phase-to-neutral voltage
between the two coils. If accurate measurements are required, Fig. 18.8K should be
avoided because the electrostatic attraction between the two coils can introduce
minute errors.
462 Testing, commissioning and management of protection

C u r r e n t coil

L
C _ , A

y . . . . . .

Voltage
coil T h e s e t w o coils
are at t h e s a m e I I
potential -- correct

Resistance !
N
C .... w . . . . . . -- , L!

Fig. 18.8J Correct wattmeter connections for very accurate measurement

C u r r e n t coil

Resistance T h e s e t w o coils have

a potential between
them -- incorrect

Voltage
coil

N
O " v -- ,

Fig. 18.8K Incorrect wattmeter connections for very accurate measurements

(iv) Signal generator: The function of a signal generator is to produce a range of

frequencies. Such an instrument should be available when testing carrier equipment.
These instruments can be obtained with different frequency ranges, and a model
should be selected which will be most suitable for the intended application. For
tests on carder protection a frequency range between 70 kHz and 700 kHz is
required; the instrument should have low harmonic-distortion and good frequency-
stability.
The maximum output available from frequency generators is normally limited to
a few volts. Such outputs are adequate for testing carrier equipment at the pro-
tection frequency, but to test the performance of filters at frequencies outside their
pass-band and to check the line-coupling equipment, much higher outputs are
required. For these tests it is necessary to pass the output from the signal generator
through an amplifier with an output of approximately 10 W.

(v) Oscilloscopes: An oscilloscope is required when carrying out work on carrier

or v.f. equipment, and it can also be used to advantage when the wave form of a
50 Hz test supply is in doubt. Its usefulness arises from the fact that it provides
a visual representation of a voltage wave form at a speed limited only by the writing
 Testing, commissioning and management of protection 463

speed of the electron beam.

The deflection of the beam is proportional to the voltage applied to the deflec-
tion plates, the signal under test being applied to the Y plates (vertical deflection)
and an internal time base to the X plates (horizontal deflection). Internal ampli-
fiers are normally built into the instrument to improve sensitivity. These amplifiers
are calibrated, and the instrument can thus be used to obtain an estimate of signal
amplitude. In this respect it is essential to confirm from the label on the selector
switch that the band-width of the amplifiers covers the range of frequencies of the
signals under test. In general, the instrument should be suitable for use with signals
of frequencies up to 700 kHz, this being the upper limit of frequencies on carrier
systems.
Current wave forms may also be examined by passing the current through a
noninductive resistor and observing the voltage drop across the resistor in the
normal way. The value of the resistor should of course be small compared with the
impedance of the circuit under test, to minimise its effect on the circuit.
The oscilloscope can also be used for frequency comparisons. An approximate
estimate of frequency can be obtained from the time-base frequency if a stationary
trace is obtained and the number of cycles on the screen counted.
A more accurate method of frequency-measurement is possible by making use
of a signal generator. This is connected to the X plates in place of the internal time
base, and the unknown frequency is applied to the Y plates. When the two fre-
quencies are identical, a stationary trace, known as the 'Lissajous figure' is obtained
in the form of a straight line, ellipse or circle. The shape of the figure depends on
the phase relationship between the two signals.
The other points worth mentioning are that the oscilloscope can also be used to
measure d.c., and that the common terminal for a.c. and d.c. is usually earthed to
the frame of the oscilloscope, which in turn is earthed through the supply lead.
Great care should therefore be taken in connecting up a.c. or d.c. supplies to the
terminals when this earth is present.

18.9 Records

For the proper management of protective gear, good records are essential as they
enable the performance of protective gear to be assessed for long or short-term
periods. Typical records are discussed below under their appropriate headings.

18.9.1 Relay settings

For the proper calculation of relay settings, many facts have to be known. Apart
from the mathematical ability (including familiarity with symmetrical components)
required to do the actual calculations, much information is needed before the
calculations can even be started. This information includes:
(a) Current transformers: Ratio, errors, internal resistance, excitation charac-
teristics including knee point voltage.
464 Testing,commissioning and management of protection

(b) Voltage transformers: Ratio and errors.

(c) Protection: Types of relays, relay ratings and burdens, setting ranges, operating
times, bias characteristics (if any).
(a) Primary equipment: Lengths of lines and cables; impedances of generators,
transformers, reactors, lines and cables; types of line construction (as some
types of line have different impedance characteristics).
(e) Diagrams: Diagrams of the local system are required so that the requirements
for discrimination can be assessed.
(t) Fault levels
(i) Maximum fault levels need to be known to determine the degree of
stability the protection has to have on external faults, and to calculate
relay operating times.
(ii) Minimum fault levels are assessed to ensure that there is always enough
current available to operate the protection, and to calculate relay
operating times.
(iii) Data on the infeed from other points on the system are necessary at
times.
Relay settings, once calculated, have to be recorded and readily available at all
appropriate points - for example, not only at the location where they were calcu-
lated but also on site and in the Grid or Area Control Room.
One such system of recording employs typed sheets from which extra prints
can be taken. The sheets list the station, circuit, type of protection, c.t. ratios,
serial number and ratings of all relays, range of settings available, settings in use,
and the date on which the settings were applied.
It is vitally important that theserecords are always up to date, and the flow of
such information should be carefully monitored. Except in an emergency, the
passing of relay setting information by telephone is deprecated.

18.9.2 Test results

The recording of commissioning tests and routine maintenance tests has already
been described in earlier chapters. It is, however, worth listing the main points
again:

(a) Commissioning log: For recording all test results on commissioning and
keeping a running record of progress.

(b) Block and circuit diagrams: Block diagrams showing the single-line primary
connections of a substation together with c.t. positions, the protection those c.t.s
operate and the circuit breakers the protection trips are a ready means of enabling
an overall picture of the substation to be readily assimilated.
Circuit diagrams show c.t., d.c. and other circuits in more detail in a simple non-
physical form, so that, again, a scheme can be readily understood. Such diagrams
 Testing, commissioning and management of protection 465

are invaluable on commissioning tests, routing maintenance tests and fault investi-
gation.

(c) Test record schedules: Schedules are desirable for recording the results of
secondary injection, tripping, insulation resistance and other tests. These schedules
will vary in design depending on the type of test and whether it is a commissioning
test or a routine maintenance test. A commissioning test is usually only done once
and usually in great detail, whereas a routine maintenance test will be done many
times in less detail.
The schedules can be conveniently contained in binders for a particular station.
Commissioning logs are used for tests on new equipment, and routine maintenance
logs for tests done subsequently. These latter schedules were discussed more fully in
Section 18.5.3.

(d) Routine maintenance test records: As mentioned earlier, opinions differ as to

the frequency of routine maintenance testing. However, some scheme should be
devised of ensuring that the engineer responsible is reminded, well in advance, that
tests are due or overdue on a particular piece of equipment. This reminder can take
the form of cards, schedules, or even notes on calendars or in diaries. Regular pro-
gramming is desirable, no matter how recorded.

(e) Fault record sheets: To form an assessment of the need for routine main-
tenance tests, a 'fault history' of all equipment is a good guide. If engineers record,
on simple forms, everything found wrong with a particular scheme of protection, or
type of relay, auxiliary switch, pilot cable and the like, trends in performance can
be watched. In an ambitious scheme the information could be coded on to a
punched-card system so that the information, fully cross-referenced, could be
reviewed whenever necessary. Such records enable the frequency of routine main-
tenance tests to be varied to line up with the fault liability of a particular equip-
ment.
The sheets record brief details of the trouble and of the corrective measures
applied. If these two classifications are entered under headings on different parts of
the same sheet they enable outstanding faults to be seen at a glance in those cases
where it was not possible to clear the fault at the time it was found.
Or) Data storage using computers: The Central Electricity Generating Board have
evolved a central storage system for technical data. A national catalogue of power
system data has been produced, and a protection data scheme has been started.
Each piece of primary equipment is identified by a code and all known data
appears in coded form. Questions referring to any type of equipment can be
answered quickly by the computer.
The same codes are used in t h e protection data scheme, in which system faults
are recorded in code, operations of all relays are tabulated, and information on
incorrect performance i~ given in detail. The information fed into the computer can
be used to determine the performance of any type of protection equipment in use.
 466 Testing,commissioning and management of protection

(g) Protective gear operation reports: The justification for the careful testing of
protective gear is its high overall performance. There must be some method of
assessing whether it has behaved correctly; to this end a detailed record should be
kept of every fault and every protective gear operation, together with the circuit
breakers which trip. A careful note should also be made of any failure to trip.
An annual study of these figures enables the protective gear performance to be
assessed. The system fault performance can be expressed as the ratio of the number
of system faults correctly cleared to the total number of system faults.
It sometimes happens that during work on protective gear or control circuits
a mishap occurs and circuit breakers are tripped. This could be due, for example,
to secondary wiring or vibration. In assessing the overall performance of protective
gear, these unnecessary trips should also be recorded. Here the performance can be
expressed as ratio between the number of circuit breakers unnecessarily tripped and
the total number of circuit breakers at risk. On the UK electricity system, such
faults are known as non-system faults, because the trips occur at a time when there
is no fault on the power system.
The system fault performance thus gives the technical performance of the
gear, the non-system fault performance gives the performance influenced by the
human element, and the combination of the two gives the overal performance of
the protective gear and the staff.

(h) Lines of communication: When dealing with protective gear, attention to

detail is of paramount importance. It is therefore essential that all relevant in-
formation is fully and clearly recorded. This is not easy with a multiplicity of gear
and voltages and a staff which may be dispersed over several thousand square miles.
Day-to-day contact can be made via telephone calls or visits to site, but written
information needs further consideration. It is not always realised that the flow of
information needs to be two way - from headquarters to site and from site to
headquarters.
Relay settings, plant data, circuit diagrams, fault levels, commissioning logs,
routine maintenance records, fault record sheets and protective gear performance
statistics require a two-way flow of information, and to avoid chaos some system
by which this information is channelled to and fro should be devised. The method
used will vary to suit different conditions.
For information which usually flows in one direction only, some form of
memorandum or circular may be desirable. For communications on protective gear,
protection memoranda bearing serial numbers may be issued. In this way, Ides may
be devoted exclusively to these memoranda. Methods of test could form the basis
of a numbered series of test circulars to cover different types of protection, testing
hints, warnings and the like.
The volume of information so handled can be quite formidable but if several
files are used for different subjects, reference to a particular topic can quickly be
made.
 Index

Abnormalities, during testing, 448-450 principles, 328, 330,331,358

Air-insulated reactors, 182, 183 reactors, 192, 193
Alarms standby earth fault, 236, 241, 333, 334
testing, 433 Bank, capacitor, 221-223,230-235
Alston series capacitor, 230 BICC
Aluminium shield (of reactor), 185 electronic protection device, 229
Arc-suppression coil sealed spark gap, 225
cross country fault, 278 Bonding bar, switchgear, 87
operation, 278 Boosters, 199-205
tuning, 280 function, 200
Armature, motor induction regulator, 210-212
current, 149-152 intertripping, 214
reaction, 151 moving-coil regulator, 205-210
winding, 149-152 protection, 213-215
Attenuation transformer tap-changing, 200
carrier signal, 427 Booster transformers, 200- 205, 213, 214
Automatic switching bucking, 199, 201,208
application of equipment, 344, 382, 391 in-phase, 203
commissioning, 356,357 oil-immersed, 214
delayed automatic reclosing, 345-348 quadrature, 204
design of equipment and programming, Breaker fail protection, 50, 51
344,348-356 British Standards
electromechanical relay equipments, a.c. circuit breakers BS 5311, 432
348-353 air-break circuit breakers BS 3659, 432
microprocessor-based equipment, capacitors BS 1650, 245
353-356 reactor rating BS 171,186,245
high-speed automatic reclosing, 345 Bucholz relays, 54,193-195
Auto-transformer protection, 57, 58 booster, 214
differential, 58 earthing transformer, 245
tertiary winding, 58 reactors, 196
Busbar fault clearance
Auxiliary switches for busbar protection, 131 back-up overcurrent and earth-fault relays,
on busbar selectors, 127-130 83
on section disconnectors, 124 distance protection, 84, 85
Busbar protection, 81,385-387,397
Back tripping, 134-138, 441 basic principles, 81, 82
maintenance testing, 441-443 biased schemes, 94
Back-up protection combined phase and earth fault, 83
distance, 331-333 defective alarm, 132, 133
overcurrent, 325,327-331 earth fault, 83
468 Index

frame earth systems, 87-93 log, 405,414,432,464,465

influence of c.t. performance on through tests, 400-437, 443,444
fault stability, 103 planning, 401-405
maintenance, 441-443 procedure, 405,406
stability limits, 142 schedule, 404
supervision relay, 131,132,142 types of, 407-437
test facilities, 139, 140 Commutator motors, 159
tripping circuits, 88-93, 96-100, 130 Compound motors, 149-155
typical schemes, 113-118 Contactors
unit systems, 86, 93 in motor control, 163
zones, 82, 96-99 in stalling relays, 168
multiple check, 123 principle, 163
Bushings, transformer, 414, 416, 419 undervoltage release, 180, 181
Buswire shorting, 132 Copper shield of reactor, 185
By-pass circuit breaker, 228,229 Core and coil assembly of moving-coil
regulator, 209, 210
Cable gland, insulated, 89 Current balance tests, 418, 419
Capacitors, 215- 245 test box, 453,454,456,457
delta connected, 231 Current-limiting reactor, 186
external protection, 224-230, 235-237 Current transformers
function and use, 215,219, 220 balance tests, 418, 419
internal protection, 223,224,230-235 circuit supervision relay, 131,132, 142
series connected, 215-217,219, 220,223- connection and polarity tests, 413-423
223-230 effect of magnetising currents, 374
shunt connected (215-220,230-237 equivalent circuit, 105, 110
split star connected, 232 knee-point voltage, 108, 412, 413
star connected, 230-235 location, effect of, 334-337, 377, 381,
tank type, 221 382
unit type, 221,222 magnetisation curve, 412, 413
Carrier protection, 327,328,359,360,392 overlap, 411
commissioning tests, 423-433 position of, for busbar protection,
CEGB 118-127
reactive compensation programme, 220 ratio tests, 417, 418
synchronous compensators, 215 records, 463
Charging currents, effect of, 364 saturation of, 325
Circuit-breaker, fail protection, 142-147 test links for busbar protection, 130
tests, 441-443
Circuit breakers Data storage on computers, 465
control diagram, 403 Diagrams, types of
tripping and closing tests, 431,432 block, 401,402, 464
Circulating current principle, 94, 95 circuit, 401,403,456,457,464
for busbar protection, 95-103 Differential protection
high-impedance schemes, 103-127 biased, 419
fimiting overvoltages, 108, 109 capacitors, 235,238
operation, 108 generators, 7-11
stability, 103-108 generator, transformer units, 63
stability voltage, 106 reactors, 190-192, 195
Circulating-current protection (see also spill current, 435
Merz-Price principle) synchronous compensators, 237, 241,242
capacitors, 235,238 transformers, 39-45
reactors, 191 Directional overcurrent protection, 48-50,
Coil assembly of reactor, 184 79,436
Commissioning Disconnecting points for relay testing, 440
 Index 469

Distance protection stator, 2

advantages and disadvantages, 361,362 Generator protection systems, 7-38
applied to complex primary circuit, 391 back-up overcurrent and earth-fault,
applied to long feeders, 360-362 12-17
applied to short feeders, 363 biased differential, 9-11
applied to teed feeders, 379-382 interturn, 21
applied to transformer-feeders, 383, 394, loss of excitation (field failure), 22-24
397 low forward power, 29, 30
applied to underground cables, 364 mechanical and hydraulic, 31-33
reduction in reach, 331-333,379-381 negative phase sequence, 17-21
overspeed, 30, 31
Earth-fault protection pole slipping 24-27
boosters, 213 rotor earth fault, 27-29
capacitors, 235,236 sensitive power, 29
reactors, 192, 195 tripping supplies, 33, 34
sensitive, 275,276 unbiased differential, 7-9
tests, 410,411,415-418,437 underexcitation, 31
Economic considerations for selection of Generator transformer units, 60-66
protection, 357,392 biased differential protection, 63
Electronic overload relay, 170, 171 overfluxing protection, 65, 66
stator earth-fault protection, 64
Fault investigation, 446-448
tripping arrangements, 33, 34, 65
primary, 446,447
protective equipment, 447 Harmonic currents, 234,237
solid-state equipment, 447,448 Harmonic restrain, 43-45
Fault indicators for overhead lines, 265=267 High-impedance circulating current busbar
Fault settings for busbar protection,
protection, 103-127
140-142
H.V. systems, 297-302
Fault-throwing switches, 76, 199
parallel operation, 299-302
Flick test for c.t.s, 413,414
back-up protection to pilot-wire schemes,
Form factor in testing, 458
300-302
Fuses
pilot-wire protection, 299, 300
application, 250
radial operation, 297-299
group fusing, 250
ring main system, 297,298
capacitor protection, 223,230
earth-fault-passage indicators, 299
characteristics
overcurrent and earth-fault protection,
slow blowing, 255,264
297-299
cooling, 262
deterioration, 262,264 Ignitron, 230
discrimination, 251-255 Inductance, series, effect on power network,
expulsion, 223 215
high breaking capacity, 223 Induction motors
insertion, 431,432 polyphase, 1S5 - 157
inspection, 406 single phase, 157-159
pre-loading, 262 split phase, 158
repeater, 257 squirrel cage, 155-1S 7
short-circuit protection, 168 starting characteristic, 165
types, 250 wound rotor (slip ring), 155,157
Induction regulator, 210, 212
Gas-turbine-driven generators, 34-38 Inspection tests, 405,406,439
Generator faults, 2-4 Insulation-resistance tests, 439, 440, 442,
external, 3, 4 448, 451
mechanical, 3 Interconnected l.v. systems, 303-316
rotor, 2, 3 distributed block h.v. network, protective
470 Index

discrimination requirements, 306, fault types, 164

309-313 protection, 164-181
fringe fuses, 306,309 Moving-coil regulator, 205-209
intertripping, 306
Negative potential biasing, rotor earth-
reverse power protection, 306-309
fault detection, 27, 28
h.v. network, 305,306
distributed block h.v. network, 305,306
Oil-immersed reactor, 185, 190
interleaved h.v. network, 305
protection, 190-198
l.v. network, 304
On-load tests, 435-4 37
self clearance of faults, 304
Overcurrent protection
substations, 304,305
directional, 48-50, 79
transformer protection, 304,305
e.i.d.m.t.l., 17
Interlocking low forward power relays, 65
Intertripping high set, 368,394-397
boosters, 214 i.d.m.t.l., 12-17, 47, 48, 55, 325-331
choice of method, 338-340, 383-385, instantaneous, 70-73
397 interlocked, 50
commissioning tests, 443,444 reactors, 192,193,196, 198
d.c. signalling, 338,340-342 two-stage, 337, 338,368
fault throwing, 199, 339, 343,344 Overload protection, thermal, 165,166,234,
maintenance, 444-446 237
power-line, carrier, 339,343 Overvoltage protection for capacitors, 236
reactors, 197 Phase-sequence components, 421
voice-frequency signalling, 338,343 Phase shift during testing, 409,428
Line coupling equipment, 423,426, Phasing tests, 433-435
testing, 424,426,427 Pole-slipping protection, 24-27
Line traps, 424,425 Power directional comparison protection,
testing, 424,426 377-379
Links for testing Primary injection tests, 414-423
insertion, 431,432 equipment, 458
inspection, 406 Primary networks, rural
Lissajous figure, 463 auto reclosing, 286
protection
Magnetic shield of reactor, 183 duplicate feeder transformers, 284-286
Main protection, application principles, ring main systems, 283
328-330 single-feeder transformers, 283,284
Maintenance single-circuit-breaker substation, 289
frequency schedules, 438, 439 Primary substations in urban distribution,
records, 465 302,303
routine tests, 437-446 Private generation in urban networks,
Mercury switches, 194,195,197,214 316-319
Merz-Price principle, 191 Protection applications in a major transmis-
Microprocessors, use o f , for automatic sion system
switching, 353-356,360 banked transformers, 373-376
Motors, 149-181 complex primary circuits, 387-392
a.c. double busbar station, 385-387
characteristics, 155-162 long overhead feeder, 358-363
types, 155-162 carrier protection, 359,360
applications, 162 distance protection, 360-363
control, 163 mesh station, 387
d.c. short overhead feeder, 362,363
characteristics, 149-155 tank earth-fault, 376-376
excitation, 150 teed feeder, 376-382
 Index 471

transformer, 365-373 cold-weather performance, 260

transformer-feeder, 383-385 construction and operation, 258-261
Protection applications in minor transmission co-ordination with fuses, 261-264
and major distribution systems grading
complex primary circuit, 397 reclosers in series, 269
double busbar or mesh station, 397 with fuses, 261-264, 268
feeder, 392 with relays, 268
teed feeder, 393 Reclosing, automatic
transformer-feeder, 393-397 dead time, 256,272
transformers, 393 multishot, 256,272
Protection in urban n e t w o r k s - future principle, 255,272
trends, 319 schemes, 272
Protection operating times, 324, 325,361, single shot, 256,272
393, 394 Records, 463-466
Protection signalling relay settings, 463
choice of method, 338-340 test results, 440,464,465
d.c. signalling, 338,340-342, 362-365 Regulators
power-line carrier, 339,343,362 induction, 210, 212
voice-frequency signalling, 338, 342, 362, moving coil, 205,209
363 Relays
Protection signalling equipment Buchholz, 193-195
commissioning tests, 443,444 current check, 122,143-145
maintenance, 444-446 for c.t. circuit supervision, 131,132, 142
gas and oil operated, 193-195
Qualitrol device, 54 high-impedance, 112
Radial l.v. systems, 294-299 duplication of, 117
l.v. cables, 294 per circuit, 113
short-term ratings, 294 per zone, 113
services, 294 rating, 132, 142
substation transformers, 295-297 records, 464
parallel operation of transformers, 297 settings, 463
protection by circuit breakers, 296 single pole, 180, 191
protection by fuses, 295,296 stalling, 168-170
Reactors, 181 - 199 tests, 408-411
application, 186-190 attracted armature, 408
busbar, 187 biased, 409
feeder, 189 directional, 410,411,435-437
generator, 186 distance, 410
group feeder, 190 inverse definite minimum time, 408
current limiting, 186 thermal, 165-167
discharge, 229 thermal trip, 170-173
function, 181, 215 Relays, reclosing
series, 155, 181 periodic, 269,270
shunt, 181 solid state, 272-274
protection, 190-199 Remote circuit breaker, tripping of, 121, 122
overall differential, 190-192, 195 Resistors, use in busbar protection
overcurrent and earth fault, 192, 196 non-linear, to limit overvoltages, 108, 109
rating, 185 rating, 132, 142
types, 181 - 185 series, 155
Recloser, pole mounted, automatic, 257 shunt setting, 110
alternative operating cycles, 264,265 stabilising, 106,107
application, 261 Reverse power protection, 241,242
characteristics, 259,260 Reyrolle
472 Index

spinner gap, 227 non-extinguishing, non-triggered, 224

Rotating field, starting method, 158 non-extinguishing, triggered, 227
Rotor heating, 420-423 Spill current, 103
Rural system protection Spinner gap, 227
performance/cost comparison, 280 Squirrel-cage motor, 155-157
arc-suppression coils, 281 Standby earth-fault protection, 333,334
automatic pole-mounted reclosers, 282 capacitors, 236,238
fuses, 281 synchronous compensators, 241,242
reclosing substation circuit breakers, 282 Star connection (busbar reactors), 188
Rural systems Star-delta interposing c.t.s.
construction, 247 generator h.v. overcurrent protection,
design, 247 13-17
fault levels, 247,248 transformer-feeder instantaneous over-
current/earth-fault protection, 71-73
Sanction-for-test card, 409, 414 Starting characteristic, 165
Schrage motor, 159-161 induction motor, 165
Secondary injection tests, 407-411,440 shunt-wound motor, 166
biased relays, 409 Station transformer protection, 57
busbar protection, 44 2 Substation circuit breakers
directional relays, 410, 411 reclosing, 269
distance relays, 410 control schemes, 272
equipment, 451-458 operating sequence, 271
instantaneous relays, 408 requirements for reclosing duties, 274
inverse definite minimum time, 408 vacuum 275
Sectionalisers Substations
co-ordination modification to, 437
with reclosers, 267 Supplies to h.v. consumers in interconnected
with reclosing circuit breakers, 275 l.v. systems, 316
operation, 267,268 Supplies to large point loads in inter-
Selectors, tap-change, 200 connected l.v. systems, 313,314
Series arm tuning, 424-427 back-feed from large induction motors in
Series motors, 149:- 151,154 interconnected l.v. systems, 313-316
Series reactor, 181,182 Switching, star/delta, 156
protection, 190-195 Synchronous machines, 215,220
Short-circuit protection, motors, 168
Shunt-arm tuning, 426,427 Tank-type capacitor, 221
Shunt motors, 149, 155 Tap-changing, transformers, 200
starting characteristics, 165, 166 Tests
Shunt reactors, 181 back tripping, 441-443
protection, 195-199 busbar protection, 441-443
Signalling channels circuit-breaker, fail protection, 443
choice of, 327,329,338-340, 376 commissioning, 400-437,443,444
fibre-optic, 328 current transformers, 411-414
microwave, 328,376 indications and alarms, 433
Post Office pilot, 324, 327,328, 343 insulation resistance, 407,448
private pilot, 327 on-load, 435
routing of, 339,340, 362-365 phasing, 433-435
Single-phasing protection, motors, 167 primary injection, 414-423
Solid-state equipment, 447,448 secondary injection, 407, 411
Spark-gap protectors, 223-227 tripping and closing, 431,432
extinguishing, non-triggered, 224 equipment, 423, 424, 429, 430, 445,
open type, 225 451-463
sealed type, 225 error avoidance, 448-451
 Index 473

winding temperature, 52, 53

instruments, 450,458-463
intertripping, 443-446 Transformers, auto, 220, 235
maintenance, 437-446 Transmission systems
frequency of, 438, 439 factors influencing choice of protection,
insulation resistance, 440,448 322-328
secondary injection, 407-411,440,442 cost, 328
polarity, 435 load and fault currents, 323
protection signalling, 443-446 plant to be protected, 322
wattmeter connections, 462 protection compatibility, 327
Thermal overload protection, 165-173 rating and capability of protected plant,
thermal relays, 165-168,244 323,324
thermal trips, 170-173 security of supply, 325-327
thermistors, 178-180 signalling channels, 327,328
Tie-bar connection, busbar reactors, 188 speed of operation, 324,325
Time-lag trips, 173-177 types of fault, 323
Torque, motors, 151 system design considerations, 321, 381,
Transformer faults, 4 - 7 , 3 2 3 , 3 6 6 382,392
Transformer-feeder protection, 67-80, 383- Transient d.c. component, effect on busbar
383-385,393-397 protection, 103,104
directional overcurrent, 79 Triggered gaps, 227,228
high set instantaneous overcurrent and Trip-angle test unit, 429, 430
earth fault, 70-74 Trip monitor device, 445,446
intertripping, 74-76 Tripping tests, 431,432,439
neutral displacement, 77 Tripping of remote circuit breakers, 335,
overall, 67-69 336,383-385,397
separate, for feeder and transformer, Tripping supplies, 33, 34
69-74 Trips, motor-protection relays
Transformer protection electromagnetic, 173-177
application in a major transmission system, thermal, 172, 173
365-376
Tuning
application in a minor transmission system
series arm, 424-427
or major distribution system, 393
shunt arm, 426,427
application to banked transformers in a
Twice-in contactor, 169
major transmission system, 373-376
probability of various types of fault, 323, Undervoltage protection, 180
366 Unit capacitors, 221,222
protection of tertiary winding, 370-373 Unit transformer protection, 35-38
Transformer protection systems, 39-59 Urban distribution
biased differential, 42-45,367, 369, 374 load density, 282
directional overcurrent, 48-50 system voltage levels, 291
distribution-transformer protection
schemes, 54, 55 Windings
gas generation and oil surge, 54, 54 armature, 149-152
interlocked overcurrent, 50, 51 field, 152
overcurrent, 47, 48 reactors, 190
restricted earth fault, 45-47 regulating, teaser, 201
standby earth fault, 51, 52 rotor, 160
tank earth fault, 52 series, 152, 153
transmission transformer protection shunt, 152, 153
schemes, 55-59 tap changing, 200, 201
unbiased differential, 39-42, 368, temperature protection, 178, 195, 214,
370 237, 245