Chapter 10

Feeder protection" pilot-wire and

carrier-current systems
by F.L. Hamilton, revised by L. Jackson and J. Rushton

10.1 General background and introduction

Unit protection, with its advantage of fast selective clearance, has been used for
interconnections within power systems from the early stages in the development of
generation, transmission and distribution. In these early stages, generation was local
to the load, and distribution was the main function of the network. At this period,
we find unit protection of cable interconnections an example of the first applica-
tions of the principle of unit protection. As the main network was underground, it
was natural that the secondary interconnections necessary for unit protection
should be by means of 'auxiliary' or 'pilot' cable laid at the same time as the
primary cables required to be protected. Also, as the 'Merz-Price' or 'differential'
principle was developed in this period, it was understandable that this principle
should be used as the main basis of these early forms of unit feeder protection.
Since that time, unit feeder protection has developed considerably and many
variations and complexities have resulted, but it is well to remember that its origin
was basically simple. It should be noted also, that Great Britain has been one of the
most active users of the unit principle of protection. The USA and certain English-
speaking countries overseas such as South Africa and Australia have also used it
widely. The problems of adequately protecting modern power systems, with their
high degree of interconnection and with the need for fast clearance times have
lately led Continental countries to consider adopting unit protection (compared
with distance protection), and its use in these countries, therefore, is likely to
increase.
The simple concept of unit protection for underground feeders using pilot wires
was, by necessity, extended to overhead interconnectors as systems developed. At
the same time the problem of providing economically an auxiliary channel by
means of pilots directed attention to carrier techniques, which use the main
primary conductors, and radio links which use aerial transmission.
 Feeder protection: pilot-wire and carrier-current systems 165

10.2 Some basic concepts of unit protection for feeders

It will be appreciated that unit protection required the interchange, between the
several terminations of the protected zone, of information about the local con-
ditions existing at each termination. Feeder protection which uses information
from one point only of a network is inherently non-unit, even though the
protective system may be very complex in form and sophisticated in its action. An
essential feature of unit protection therefore is the provision of a channel over
which information can be passed between ends. The variety of unit systems of
feeder protection which have been developed over the years is due to the several
different methods of providing a suitable channel and the problems which each
kind of channel has presented in its application.
A B
• . . . . . . |

T -T
A i

Ill I Im.l~

I / I
I I I
I o I c : Trip B
I lip-- I
I I
I I
TripA : t 01~ i (~)
u I
I. . . . . . .j
Fig. 1 0 . 2 A Simple local differential protection with short interconnsetions between c.t.$

When dealing with the zone of protection of, say, a generator or transformer,
the circuit-breakers controlling the terminations of the zone are relatively close
together and therefore usually require only a single 'relaying point'. The need for
an information channel still exists, however, but it is relatively simple and is
formed by direct interconnection of the current transformers with the single relay-
ing point, the tripping of the associated circuit breakers being effected from that
point (Fig. 10.2A).
If the terminations and their associated circuit breakers are separated by greater
distances, say thousands of metres, the use of a single relaying point is no longer
practicable. This means that a relaying point must be provided at each circuit
breaker, and the information channel becomes more complex, each relaying point
being provided with information about all the others in order to be able to assess
this information together with its own and also to trip or stabilise as the case may
be (Fig. 10.2B). Attenuation, distortion and delay in transmission are all important
to some degree or other because significant distances are involved.
166 Feeder protection: pilot-wire and carrier-current systems

A B

__ogo
Trip A 1~.~=~.; iI'-'Om~"(~ ) Trip B

........ & L_C_ t_ 2 ..... J

Supplementary intertripping
channel sometimes required

Fig. 10.2B Unit protection with large distance between c.t.s.

10.3 Basic types of protection information channels

There are three main types of information channel used for protection at the
present: (a) auxiliary conductors, for example, pilot wires: (b) the main conductors
of the protected circuit or some other circuit running in parallel; (c) aerial transmis-
sion, for example, radio links.
The choice of a particular type of information channel depends upon many
factors such as economics, availability of channels, the type and length of the
protected line and other factors such as the possible uses (other than protection)
for which the information channel may be required.
Some of the features and parameters of these types of information channels have
been described in Chapter 7, together with associated terminal equipment,
particularly where this is used in digital information systems. The protection
systems covered in this Chapter are based mainly on analogue information and,
although most of Chapter 7 is still relevant, some features related to analogue
information are referred to in the following Sections.

10.3.1 Pilot wires

There is considerable variety in the types used, and the subdivisions of these
different types cover such factors as core size, insulation level of the cores, whether
the pilot is underground or overhead, and particularly whether the pilot is privately
owned or rented, Some of the main details of a representative range of pilots are
 Feeder protection: pilot-wire and carrier-current systems 167

given in Table 7.2.2.1A of Chapter 7.

It will be noted that the intercore capacitance values are subject to some
variation, the actual capacitance obtained for any given cable being dependent on
the design of the cable and on the type of insulating material employed. Thus,
telephone cables using plastic insulation (polythene or p.v.c.) generally have a
higher value of in tercore capacitance than paper-insulated cables, this being
particularly so in the case of p.v.c, insulation.

10.3.2 Main conductors

The main conductors of the power system may be used in either of two ways. With
a suitable h.f. coupling equipment, high-frequency signals may be injected and
received using the main conductors as the actual communication link. There are, of
course, many ways of superimposing information on such a basic h.f. cartier signal,
and these will be described later. The frequency range available for this purpose is
generally in the band of 70-500 kHz. The second way of using the main conductors
is to disturb the system deliberately by placing a fault on it, for example by a fault-
throwing switch, causing fault currents to flow and so causing operation of pro-
tective relays at remote points. This method is described in Chapter 17.

10.3.3 Radio links

Although not widely used for protection in the UK, their use has increased con-
siderably in other countries over the last 10-15 years. They have particular ad-
vantages where large amounts of information are required for control and
communication, and for some areas of difficult terrain. They depend on line-of-sight
transmission, and beamed radio signals generally in the frequency band of 1000-
3000 MHz, or even above. Again, various forms of information can be superimposed
on the basic carrier for different types of protection.

10.4 Types of information used

There are different types of information which may be derived from the primary
circuit and transmitted over the information channel. These range from complete
information about both the phase and magnitude of the primary current to very
simple information as, for example, when the action of some local relay is used to
switch on or off a communication signal. These various forms of information and
brief comments on their application are described below. They will be dealt with in
more detail in the relevant Sections of this Chapter concerned with specific forms
of protection. Those systems based on command signalling have been covered in
detail in Chapters 7 and 9.
 168 Feederprotection" pilot-wire and carrier-current systems

10.4.1 Complete information on magnitude and phase of primary current

The use of full phase and magnitude information automatically implies that the
system is a true differential one. In practice, this is applicable mainly to pilot wires,
and even then to special cases such as multi-ended circuits. It has also been used
in carrier systems for special applications on important transmission lines abroad.
The use of a differential system presents some difficulty in obtaining full infor-
mation linearly over a wide range of current, but, in theory, the differential system
has very few limitations.
Practical difficulties may impose the need to deviate from a true differential
system. A particular example is when the secondary signals derived from the
primary current are only directly proportional to the magnitude of the primary
current over a limited range. This has the effect, at high currents, of losing the
magnitude part of the information leaving phase-angle information only, as des-
cribed in the next Section.

10.4.2 Phase-angle information only

Information as to the phase angle of primary current is relatively simple infor-

mation, that is it is not dependent upon current magnitude, and therefore it is
relatively easy to superimpose on and transmit over any type of channel. Pilot
wires, carrier current over the primary conductors and, more recently, radio
links, have all been used for this purpose. There are certain limitations in protection
performance when phase-angle comparison is used as will be described in more
detail later.

10.4.3 Simple two-state (off/on) information

This form of information can be used with all three types of channel. The function
of the channel may be either to block tripping (secure stabilising during external
faults) or to secure tripping under internal faults. Such tripping may be direct, in
which case it can be classed as intertripping, or it may be permissive through some
relay at the receiving end. The method in which the channel is used, and the type
of channel which is used, will govern the various features and requirements of the
protective system. The technique used and their application are described in
Chapters 7 and 9.

10.5 Starting relays

These are necessary where the communication channel used for unit protection is
not continuously employed. This may be the case, for example, in:
 Feeder protection: pilot-wire and carrier-current systems 169

(a) Pilot wire systems, where the pilots are normally supervised and such super-
vision is not compatible with the discriminating action of the communication
channel. Alternatively, the channel may be used for general communication
under normal system conditions.
(b) Carrier systems where there are restrictions on continuous transmission of
carrier power.
(c) Any unit system where the principle used for discrimination is not valid at
low-level fault conditions.
Starting relays are usually simple devices and are nondiscriminative because they are
actuated by quantities existing only at the termination where they are located.
They are sometimes called 'fault detecting' or 'sensing relays'. They do not do any
tripping on their own account. However, they will usually decide fault settings, and
their design and choice is important in this respect. They will also affect the overall
tripping time of the protection, and this is also an important feature.
By way of illustrating their use consider a simple scheme, with a communication
channel not normally in service (Fig. 10.5A). On the occurrence of a fault
condition on the system, this channel must be changed over to protection use in
order to permit the discriminating feature of the protection to decide whether the
fault is internal or external, that is to trip or to stabilise. This switching is done by a
starting relay at each end, which detects the fault condition.

I I
I I
I I

C o n t i n u o u s d u t y ()f
i n f o r m a t i o n channel
P- Protection

S - S t a r t i n g relay
Fig. 10.5A Channel switching using starting relays

If it is an internal fault, fed at one end only, the starting relay at that end will
operate. As a result, the communication channel is connected at only this end and
tripping will therefore occur at this end. This, however, is acceptable under these
fault conditions.
170 Feederprotection: pilot-wire and carrier-current systems

If the fault is external, then stabilising is essential, and fault current is present
(of approximately the same value) at both ends and we must then ensure that the
starting relays operate at both ends.
This is easily done at high fault currents, in excess of the settings of the relays,
but not easily at low level marginal fault conditions because of relay errors. The
effect of this is illustrated in Fig. 10.5B.
It could give operation of relayA and not relay B, present the protection with an
internal fault condition and so produce a maloperation. Such marginal fault con-
ditions often arise on modern power systems.
The normal procedure for overcoming this difficulty is to fit two starting relays

Tolerance band

~"----. Set ting

of B

~' Setting I I For through currents I

era I I in this band I
[ [ real-operation [
I I ~ould occur I
I I I
0 I a I I B,I

Fig. 10.5B Maloperation owing to starting relay tolerances

at each end, one called the 'low set', which performs the above switching duty on
the communication channel, and the other the 'high set', which acts on the tripping
circuits of its associated end in a sense to prevent tripping by the discriminative
protection when it is not operated and to permit such tripping if it is. The relative
settings of such relays take into account errors and variations, as shown in Fig.
10.5C. In assigning relative settings it is often necessary to take other factors into
account, but these are mentioned in relation to particular forms of protection.
The high-set relays, of course, def'me the effective fault setting of the protection,
and the settings of such relays have to be chosen with some care.
As three-phase conditions are balanced, it is difficult to distinguish low-level
three-phase faults from load conditions. The three-phase settings of starting relays
must be chosen to exceed the maximum load conditions to be expected on the
circuit. (There is an exception to this in the case of impulse starting of phase-
 Feeder protection: pilot-wire and carrier-current systems 171

Tripping permitted with

~ . c e r t 2 i n tJ abovL thi_s v a l u e

I ' I
I Discrimination with I I
, certainty above this value I
_.2.'/ . . . . _J.__~
,
-t . . . . . I t

I l
I Nominal setting I I

I I I (LS) I I I I
I I II I I I I I
I I t t I I I I
I I I I I I I I
I I I I I I I I
0 -
i Ls I I HS I I LS I I HS I

"I p
Communicatio
Channel

Fig. 10.5C Application of high-set and low-set starting relays

comparison carrier protection Section 10.10.9).

The unbalanced condition cannot be anything other than a fault condition, and
the choice of setting is not therefore influenced by load conditions.
For a normal maximum load of 1.0 p.u., and possible overload of 1.1 p.u. as
shown in Fig. 10.5D, then under no circumstances, must the discriminating channel
be connected at this latter value. This means that the low-set relay must have a
r e s e t value greater than 1.1 p.u., say 1.25 p.u., to allow for relay errors. This
immediately draws attention to the need for a high reset ratio (reset)/(pick-up) on
the relay, if the overall settings of the protection are to be as low as possible. For an
80% reset, the pick-up of the low-set relay becomes about 1.5 p.u.
The pick-up of the high-set relay must not overlap that of the low-set, with the
practical variations to be expected (and any other requirements such as capacity
current), say a 20% margin, the pick-up of the high-set now becomes 1-2 x 1-5 = 1.8
172 Feeder protection: pilot-wire and carrier-current systems

1.8 i I I i I I 1 I I i i I iiii i I I I I

HS
D() Relay current
i 1.5

1.25 i i i I I / i / i I I i I / i i I

Overload
1.1

1.0
Normal maximum load
I

Fig. 10.5D Relationshipbetween starting relay settings and load current

and its reset 1.44. It may even be desirable to keep the reset of the high-set relay
above the pick-up of the low-set, and this would further increase the high setting.
It can be seen, therefore, that the need for such relays can increase the three-
phase setting considerably above the load, and that such relays should be accurate,
consistent and of high reset ratio, if this increase is to be kept to a minimum.
With regard to operating times, it has been noted that they can affect the overall
operating time of the protection, so they must be short, that is, 10-20 ms. They
should also be fast in resetting, as it is possible for a condition to arise, following
the clearance of an external fault, where the starting relays, both high-set and low-
set have not yet reset at each end, and there is a load current existing on the circuit
in the presence of capacitance current. If this condition is not compatible with
correct discrimination spurious tripping may occur before the relays have reset.
Certainly the high-set relays should be fast to reset and preferably faster than the
low-set. In some cases the reset times are deliberately controlled to be so. The
problem of fast resetting remains, however, and is greater the shorter the operating
time of the discriminating link becomes.
It can be seen that the choice, design and application of starting relays need
care, and present a different picture from the elementary idea of a simple electro-
magnetic relay with a nominal setting. Structural aspects such as high insulation,
elaborate contact assemblies, and special reliability requirements for some contacts
expand the problem further.

10.6 Conversion of polyphase primary quantities to a single-phase secondary

quantity

10.6.1 General philosophy

With few exceptions, it is general practice to use a single-phase quantity as the

 Feeder protection: pilot-wire and carrier-current systems 173

information to be conveyed over the communication channel; this reduces its

complexity and cost. This technique is generally justified because the primary
fault currents bear definite relationships for the various types of faults, and be-
cause the same summation technique is used in an identical manner at the several
ends of the protected feeder. Some reference to these summation techniques will
have been given in previous Chapters. They are summarised here with particular
reference to unit protection of feeders, and more details of individual circuits will
be given when particular forms of protection are described.

10.6.2 Interconnections of current transformers

The secondary windings of current transformers may be directly interconnected to

provide single-phase output currents. For example, the parallel connection of
current transformers will provide a single current output which is related to the
zero-sequence primary current. This is, in effect, a simple form of zero-sequence
output circuit (Fig. 10.6.2A). By providing different numbers of secondary turns
N

J'~ ,_,jw t ~ ~ . ~ Ir : I I + ! 2 + ! 0

l y : a211 +al 2 + I0

! b = al I +a212 + !0

I
! =~(! r + ly + !1, ) = 310
N
Nr

!r = ! I + ! 2 + !0

l y = a211 + al 2 + !0

Nb
1h = al I + a 2 1 2 + ! 0

I = l r +/.,V + i h
- .

N r Ny Nh

giving I= I + I + i !0 + 1 + + a ii + I +am+ 12
Ny

Fig. 10.6.2A Parallel connection o f current transformers

174 Feeder protection: p/Iot-wire and carrier-current systems

on each of the phases an equivalent output is obtained which will depend upon the
type of fault and the relative ratios of the secondary turns. These methods of
current summation are not used frequently for unit feeder protection as the
secondary current levels of the current transformers are not really suitable for
comparison over the information channel. Once it is necessary to transform the
current level, a current summation device is possible, as described in the next two
Sections.

10.6.3 Summation transformers

These devices were frequently used for feeder protection, particularly those
involving pilot wires. The general arrangements are shown in Fig. 10.6.3A, from
which it can be seen that it is possible to obtain a single phase output for all the
common types of faults. It is also possible to control independently the outputs for
earth faults and phase faults.
RC ,

y~ •
ID

N ~-
="

ID,

.
Output

1 /' R

Y
I.:quivalent (~utput for
equal fault - currents

R-N=n+2 3o,, \ /
Y-N=n+I nVn+!
B-N=n
R-Y=I
Y-B=I Y
B-R=2
'l'hree-phase : f'3

Fig. 10.6.3A Summation transformer

The output on earth faults is usually considerably more than on phase faults,
which means that the protection is more sensitive to earth faults which generally
satisfies the requirements on the system, particularly if this is resistance-earthed
with a limited earth-fault current.
As with all current summation circuits, the summation transformer is not perfect
and under certain special fault conditions it is possible for the output to be too
small to be significant. Two examples of this are shown in Figs. 10.6.3B and
10.6.3C. In general, however, it is possible to choose tappings which make such
 Feeder protection: pilot-wire and carrier-current systems 175

Fault

!e : I r + ! b

i.~h !h

!
r
RO

YO,
_

ib ~.
BO NR

|.i r B

le f.,, l le, I
NO'

/
I': v Ir

l<~tal a m p e r e - turns AT =- (N R - NB) I r + NBI e

~- (N R - N B ) ( 1 e - Ib) ÷ NB! e

f r o m ~ h i c h AT = N R I e - (N R - NB) ! h

Assuming ! b and i e t~ he in phase, the c o n d i t i o n

l'~r z e r ~ a m p e r e - turns i s : - Ib _ NR
Ie NR - NB

Fig. 1 0 . 6 . 3 B Showing zero o u t p u t f r o m a summation transformer under a two-phase-to-

earth fault condition

cases very unlikely, especially on multiple solidly earthed systems. There is always
the occasional combination of system conditions which might give rise to difficulty,
and these must receive special analysis.
 176 Feeder protection: pilot-wire and carrier-current systems

, ,

I.a ult

I p
m
qw
2" I" I" I ) i s t r i h u t i o n

Fig. 1 0 . 6 . 3 C Showing zero o u t p u t from a summation transformer owing t o 2" I" 1 current
distribution in. the power transformer

10.6.4 Phase-sequence current networks

This class of device is much more complex than a summation transformer and is
consequently more complicated to design and expensive to produce. It uses inter-
connection of the various current transformer secondaries and the inclusion of
complex impedances, the combination having the ability of filtering out one or
more of the sequence components of the fault current and so give an output which
is related to particular chosen sequence components. A simple example is shown
in Fig. 10.6.4A, but there are many variations.
Most sequence networks have been based on passive components, e.g. trans-
formers, reactors, capacitors and resistors, but increasing use is being made of
active components to amplify and/or combine the electrical quantities and some-
times to produce the required phase shifts.
As sequence networks use phase-shift circuits, they are really steady-state
devices. When used for instantaneous types of protection therefore, their response
to transient conditions and non-power frequency conditions must be considered in
the design. As will be shown later, it is usually necessary to build in special
frequency filters to avoid spurious outputs.
As in the case of summation transformers, there is the possibility that certain
system conditions will result in a reduced or negligible output. The characteristics
•of such networks, therefore, have to be chosen very carefully in relation to the
system conditions in order to minimise the danger of such failures.
The expense and complication of sequence networks generally prohibits their
use for pilot wire protection, and they are usually reserved for more elaborate
forms of protection using carrier techniques.

10.7 Elementary theory of longitudinal differential protection

Before proceeding to a detailed study of the particular forms of unit protection for
 Feeder protection: pilot-wire and carrier-current systems 177

1R

Iy

1B
I ! l

N<~te t h a t r and C have a 60 ~ phase-

angle so that, for positive s e q u e n c e
c u r r e n t s , V l a n d Vo cancel.
i c V l + V2= (). N c ~ , ~ u t p u t i s ~ b t a i l l c d
for zero s e q u e n c e c u r r e n t s s,nce
these cancel o u t in the primary w i n d -
ings o f the a u x i l i a r y t r a n s f o r m e r s .
. . _ _ _
. . _ . _

v 2 V I

V I +V 2
, , . . _ _
v

Protection ]

!B

Y /
f/

/
/
!R - Iy

/i/// Y -
i
1B

/
Iy

60 ° !R

V14V2-O

\
\
\
\
/
Iy - !B
JR - iy
(a) Positive s e q u e n c e c u r r e n t s (b) Negative s e q u e n c e c u r r e n t s

Fig. 1 0 . 6 . 4 A Typicalnegative-phasesequence current network

feeders this section considers some elementary principles which are necessary for a
better understanding of the problems involved and the techniques used.

10.7.1 Longitudinal differential protection with biased relays

With perfect c.t.s and symmetry of connection the basic differential system gives
178 Feederprotection: pilot-wire and carrier-current systems

perfect balance on through fault conditions and thus perfect stability. This is only
true of an unbiased differential system using a single relay, that is, where the ends
of the protected zone are fairly close. It will be seen that this perfect theoretical
balance does not exist on longitudinal differential protection on a feeder where the
ends are well separated.
Even for a differential system with a single relaying point stability on external
faults becomes a practical problem when instantaneous relays are used. The
characteristics of unbiased differential protection, Fig. 10.7.1A, show a single
current setting, and since this single current setting must cater for the minimum
internal fault conditions that can occur on a system, it will be well below the
maximum through current on external faults for which stability is required.
It is fairly easy to provide steady-state balance in a differential system but there
will be some upper level of through fault current at which the unbalance, both
steady-state and transient, will increase rapidly. The setting of the relay will
therefore be exceeded at the 'stability limit' of the scheme. The stability limits
obtainable with an unbiased low-impedance relay in practice are larely sufficient
and additional features are necessary, particularly with instantaneous schemes. One
way of securing stability is to use a high-impedance relay, as discussed in other
chapters, but this solution is not applicable to feeder protection. Another way is to
use a relay which has two actuating quantities, that is a biased relay, various forms

Relay
operat-
ing coil Relay setting
current characteristic External
fault
Stability limit I

Fault current
t.a ult setting

Fig. 1 0 . 7 . 1 A t,, .... ;ased-differen=Jal protection

of which have been described in Chapter 6. The relay would be connected in a

differential circuit, as shown in Fig. 10.7.1B, the operating winding being in the
usual residual connection and the other winding, the restraining winding, connected
in the circulating path between the c.t.s. The current required to operate the relay
will be increased by some controlled amount with increasing circulating current and
so extend the stability limit for through faults. The setting of the relay will be
increased on internal faults, but this will be relatively small provided the bias
characteristics are chosen correctly.
The presence of a through current at the same time as an internal fault current
will, of course, increase the setting further, as shown in Fig. 10.7.1C. Such a con-
 Feeder protection: pilot-wire and carrier-current systems 179

Relay setting characteristic - external faults

Relay setting characteristic - internal faults

,,,,,,

=
.= ©',"y f i "~1 !I [ ! J _ ~ Operating
; ,,o"/ ~ ~ "* / / !.. I ! F Jl I winding

o ~"I ~ .~ _ - ~ ~" / ~ External ~ winding

Relay

0
f ~ I"ault current
Fault Stability
setting Unbiased relay limit
setting characteristic
Fig. 10.7.1B Simplebiased-differential protection
dition can occur for a low level fault which did not collapse the system and reduce
the load. By delaying the growth of the bias characteristic, as shown in Fig. 10.7.1D,
it is possible to ease the problem of increased marginal settings at the same time as
being able to provide the necessary stability. Here, it must be emphasised that the
main problem with biased relays is to choose a bias characteristic which is
sufficiently in excess of the out-of-balance current on through faults to ensure
stability, and yet sufficiently below the fault current available on internal faults to
ensure fast positive action. The ease of doing this depends on the separation
between the curves of current in the operating winding for internal and external
fault conditions.

=_

.,.., ult)
o
¢j

°,..,

0
I I
~J

I I
I !

t Fault (or loud) current

Load Effective
fault setting
Fig. 10.7.1C Relay with linear bias characteristic
 180 Feederprotection: pilot-wire and carrier-current systems

,b.)

).=
L

,,,,. . .. t,): ,~,~e ~ " ~ ~

=_

e~

0 V I . . . .
I'ault (or load) current
Effective
fault setting

Fig. 10.7.1D Relay with nonlinear bias characteristic

10.7.2 Phase-comparison principles

As previously noted, the principle of phase comparison is particularly applicable to

the protection of feeders with either pilot wires or carrier. There are some
elementary principles which are worth appreciating, although these will be amplified
later. Considering a simple system where it is possible to compare the phases of the
currents at the two ends of the protected circuit, as shown in Fig. 10.7.2A, it is
seen that irrespective of the magnitudes of these currents the output signal of a
phase comparator would vary as shown.
It is seen, therefore, that according to the setting of this comparator, the
characteristic can be set to have a particular angle between the currents at the two
ends at which tripping takes place and a complementary angle where stability takes
place (Fig. 10.7.2B). The justification for using this principle of protection com-
pared with the full differential protection (which takes account of both phase and
magnitude) is that, with the exception of capacity current, the current entering the
line is the same as that leaving it for conditions of external fault. Provided the
characteristic angle is chosen with care, stability on external faults and tripping on
internal faults can be secured. Having chosen a stabilising angle, it is only safe to
permit comparison at current levels on through faults or through load when the
capacity current will not cause encroachment on the tripping region. The starting
relays referred to in Section 10.5 are often used to prevent this erroneous com-
parison. The phase angle between the two ends on internal faults may be
considerable and difficult to predict. It is normal, therefore, to arrive at the phase-
angle requirements from the stabilising condition which is more easily defined, and
to adopt as large a tripping angle as possible.
 Feeder protection: pilot-wire and carrier-current systems 181

I . ff ~j

e2

12~

il,i2 pr,m ry curr n,s

el ,e2 - voltages used for e I -e 2 i i
cot
c~mparison ..~a ~ ~ v

Ii/
....- cL il lib

"-il

~ el

iR iR

e I .... e 2

"- I I
_ , I . I
- I 80 ~ -0 0 0 180 °
Angle between ! 1 and 12
Fig. 1 0 . 7 . 2 A Basic conep t o f phase comparison pro tection

+. 0 - s t a h i l i t y angle

Fig. 1 0 . 7 . 2 B Characteristic o f phase comparison p r o tec tion

 182 Feederprotection: pilot-wire and carrier-current systems

10.7.3 Nonlinear differential systems

A system may begin as a differential system taking account of both phase and
magnitude. This means that the secondary quantities which are derived from the
primary quantities vary proportionally at low current levels. As the current level
increases the amplitude of the quantities may be limited by magnetic saturation or
non-linear resistance, and the comparison deteriorates to the only other remaining
information, which is of phase-angle difference between these limited signals (Fig.
10.7.3A). This principle is frequently used.

_i ..... °
imiter V

I : °l
/
z
/
/

_ i i i

!
P

Angle
between
V and I
P
0
lp

Fig. 10.7.3A Magnitudelimiting

In a pure phase,comparison system, the amplitude of the currents is ignored and,
hence, if there is a capacitance current fed into the protected circuit at a low level
of through-current, the resulting phase-shift between the currents at the two ends
can vary, as shown in Fig. 10.7.3 B.
If a fixed tripping angle is assigned to such a system of protection, there will
obviously be a lower limit of through.fault current below which the phase angle
between the currents at the two ends of the line will exceed the stabilising angle,
thus tending to produce unwanted operation of the protection.
Amplitude-sensitive relays such as starting relays, must therefore be provided
 Feeder protection: pilot-wire and carrier-current systems 183

at each end to prevent such unwanted operation. The problem is rather more
complicated than this simple treatment suggests and is treated more fully in the
Section on carrier protection.
! A ! + 1C 1B i
. . . . r > m--t
A ~Ic B
Line capacitance

IA 1

I-- l~

I(, 0 0

J Angle between ! A and I B

90 t) _ /

Trip-angle setting of protection

J Starting-rela,,
[ setting
i I

0 m

Range of currents
~.~hich ~ould cause
tripping on through
fault
Fig. 10.7.3B Incorrect comparison under external faultconditions owing to line capacitance
In a hybrid scheme, namely one which operates as a simple differential system at
low current levels and as a phase-comparison system at higher current levels, we can
arrange for limiting to take place at a level which automatically safeguards against
maloperation at low levels. The capacitance current would then appear as an out-of-
balance in the unlimited working range of the protection, and we would ensure that
the setting of the differential relays was sufficiently in excess of this to prevent
maloperation. Such a system, therefore, can be considered as containing its own
amplitude-sensitive feature which, in a pure phase-comparison system, would have
to be provided by additional starting relays.
True phase comparison is really permissible only for a two-ended protected
zone. Where the protected zone has three or more ends, such as tapped or tee'd
circuits, the principle fails or is very dependent upon particular system conditions
184 Feederprotection: pilot-wire and carrier-current systems

(Fig. 10.7.3C). It is seldom applied to such circuits.

A
[fA If B
= .=P I

[3--
iI
fB G
!| I'B I [ ~|i fB

If = IrA + lfB + i l f B

F a u l t c u r r e n t leaves p r o t e c t e d z o n e on i n t e r n a l fault -
a m o u n t d e p e n d s on relative p r i m a r y i m p e d a n c e s to fault
Fig. 10.7.3C Incorrect comparison on three-ended feeder

10.7.4 Directional comparison systems

A pure phase-comparison principle requires the transmission of information of the

actual relative phase angle between the two primary currents. A variation on this
theme is obtained by using directional comparison, in which the phase of the
current is compared with respect to its associated primary voltage, the same thing
happening at both ends. Directional relays do this comparison. Assessment as to
whether the fault is internal or external is obtained by transferring information
between ends as to the operation or non-operation of the relays, the communica-
tion link acting in a simple switching sense. The principle is not so elegant as phase-
comparison and chronologically preceded it. It is subject to similar limitations with
respect to capacitance current and is similarly not very suitable for multi-ended
circuits.

10.7.5 Current sources and voltage sources

It is important in protection to recognise the difference between a 'current source'

and a 'voltage source' as both are met with in practice. A current source is shown in
Fig. 10.7.5A and has the characteristic that the current through a connected load
circuit is relatively unaffected by the impedance of this load circuit. Looking into
the terminals from the load it can be imagined as consisting of a very high voltage
driving into a very high impedance. A current-transformer is a practical case and
the two forms of equivalent circuit, namely the current-source equivalent and the
voltage-source equivalent, are shown in Fig. 10.7.5B.
If two currents are to be compared, they may be fed into a single point some-
times called a 'junction' and the summated current leaving this junction monitored
 Feeder protection: pilot-wire and carrier-current systems 185

I
f

Current
z v
s()urce

C I

_I

V - ZI. F o r c ( ) n s t a l l t current, V varies w i t h Z

E i~ large and Z s )) Z

S~,tl,.t I " E .~ E
Zs + Z Zs

Fig. 10.7.5A Current source and its equivalent circuit

lp ¢
1 lp/N !
, i ,
r -- ,C
J
f

N Z
I

"l
.. O

Z S !

I
m
m

lp
E=_ Zs
N
lp
i=_ (, Z s _ )
N Zs+ Z

A c c u r a c y d c p ~ : n d s , ~ n Z s )) Z. wl,~,:cc
i- E = !,.,Z. 1 .t~ I
Zs+ Z ~ " Zs+ Z "--" "NP

Fig. 10.7.5B Current transformer equivalent circuit

186 Feeder protection: pilot-wire and carrier-current systems

by a relay (Fig. 10.7.5C). It is easy to see the practical equivalent to this in a

differential connection of current transformers.

kl 2
.,. ,

ll/N I~I~N
-N N-

--I v
I-
Summing
junction

= (il + ! 2
lR ---__- )
N

where ! 1 a n d i 2 art' p h a s ~ r s

Fig. 10.7.5C Comparison of two currents

! I' ~
v

i ' ..............
Z>) z s I=____E_E .,-,. I.:
Zs+Z Z
Z

Fig, 10.7,5D Voltage source equivalent circuit

1,:I + 1-2

Fig. 10.7.5E Summation of voltages

 Feeder protection: pilot-wire and carrier-current systems 187

A 'voltage source' is defined as one in which, if a load impedance is connected,

then the current flowing in this load impedance is directly proportional to the
impedance value. If two voltages are to be compared they are connected in series
and either the total voltage or the current flowing in a connected impedance
(Fig. 10.7.5E) is measured.
Both forms of source have errors in practice, and it can be seen that this error
(or a measure of the goodness of the source) is directly dependent on the relative
magnitude of the source impedance and the load impedance.
It is rather arbitrary as to whether the secondary quantity related to a primary
current is derived as a voltage source or a current source. The deriving of a secondary
current is more familiar, but in feeder protection it is sometimes more convenient
to derive a secondary voltage source.
A current source is relatively easily converted into a voltage source by feeding
the current into an impedance which is low compared with the impedance to be
connected to it (Fig. 10.7.5F). This is not a perfect voltage source but sufficient
for practical purposes.

ip~f lp/N

_J
( '-- --~ " i 0
--N Z Zs Z

-'1

! i

0
Z
if Z s ) ) Z then circuit reduces to:

Fig. 10.7.5F Converting a current source to a voltage source

10.7.6 Nonlinearity and limiting

As has already been mentioned, this is an important feature of many forms of

feeder protection. To appreciate the mode of operation of some systems it is neces-
sary to understand the two principal forms of nonlinearity used.
(a) Limiting due to saturation o f iron circuits: Some of the features of magnetic
saturation have been dealt with in connection with current transformers in Chapter
4, but they are repeated here for the sake of clarity. Taking a simple iron circuit,
shown in Fig. 10.7.6A, which is excited by an a.c. current, then there are a number
of quantities which can be plotted to a base of exciting ampere turns, namely flux
density, average voltage, peak voltage and r.m.s, voltage. If a simplified BH charac-
teristic of the material is assumed then at some specific value of ampere-turns the
material will be incapable of developing any appreciable increase in flux density,
that is, the iron becomes saturated. This means that the average excursion of flux
188 Feederprotection: pilot-wire and carrier-current systems

I
! I, ,

I (Ignoring
hysteresis)

~" i iI • ii

I H (u IN)
lsN
Fig. 10.7.6A Saturation of an iron circuit

in the winding becomes constant and the average value of voltage, irrespective of
its waveform, will be limited, as shown in Fig. 10.7.6B. It will be clear from Fig.
10.7.6C that at high currents the whole of this flux change will occur around the
region of current-zero and that the duration of voltage output will be short, but that
the amplitude may be high. Ideally, the area of the voltage waveform will be

I
|

Average
I
voltage

o
I I'N
lsN
Fig. 10.7.6B Limiting o f average output voltage

/ . . . . ~ .wv~'ed__.ror__m
~Current

lS

-I s

wavef°rm~ V

Fig. 10.7,6C Voltage output with severe saturation

 Feeder protection: pilot-wire and carrier-current systems 189

constant, which is the same as saying that the average value of voltage is constant.
It is clear from this that the peak value of voltage will continue to rise, and in an
ideal system would be proportional to the slope of the current input which, of
course, would be proportional to the peak of the input current, assuming it is a sine
wave. In practice, hysteresis and eddy-current losses together with the effects of
connected impedance will limit the rate of increase of peak voltage (Fig. 10.7.6D).
/

Pea k va I ue
()f v(~ltage
/
/~ / Effect of hysteresis
and e d d y loss

| ii

IN
Fig. 10.7.6D Variation o f peak voltage with saturation

In practice, this peak voltage rise is some fractional power of the input current
(0.2-0.5). Nevertheless, an iron circuit, which may have a limited average voltage of
say 50V, may well develop peak voltages of the order of 1 or 2 kV at high currents.
The r.m.s, value of voltage, which would often be measured, also would be limited
but not quite so effectively, as it would rise gradually after saturation. This is
because the r.m.s, value of a quantity is dependent upon its waveshape even though
its average value may be constant. There are a number of important aspects of iron
saturation to take note of:
(i) Peak voltage values are not limited.
(ii) The waveform is considerably distorted and peaky, containing many
harmonics.
(iii) The peaky output voltage is located around current zero and can therefore be
considered as defining the phase of the input current by a short duration
pulse.
(iv) The value of voltage that we will measure under such conditions will be de-
pendent upon the type of instrument that is used.

(b) Nonlinear resistors: If a current transformer is loaded by one of the nonlinear

resistance materials, for example, Metrosil, then limiting takes place with an entirely
different action. The d.c. characteristic of the Metrosfl material shows that the
voltage and current are related by a power law, as in Fig. 10.7.6E. For all practical
purposes, the d.c. curve applies for instantaneous values of an a.c. condition. If a
high level sine wave of current is applied to such a combination as shown in Fig.
10.7.6F, the voltage developed across the non-linear resistor will be a flattened
waveform, and the peak value will be limited and will only rise slightly as will also
the average and r.m.s, values. Some important observations about such a system of
 190 Feederprotection: pilot-wire and carrier-current systems

V = ('1/3

0 I
' ' ' ' '

C is a c o n s t a n t e x p r e s s e d in v o l t s d e p e n d e n t on the
d i m e n s i o n s a n d m a t e r i a l o f t h e m e t r o s i l u n i t a n d is the
v o l t a g e d e v e l o p e d a c r o s s t h e m e t r o s i l at a c u r r e n t o f 1 a m p .

/3 is a c o n s t a n t , r a n g i n g f r o m 0 . 2 5 t o 0 . 3 , a n d is d e p e n d e n t
on the material only.

Fig. 10.7.6E Characteristic o f metrosil

! !

Fig. 10.7.6F Voltage waveform with limiting by nonlinear resistance

limiting may be made.

(i) Average values are limited.
(ii) Peak values are limited.
(iii) The duration of the voltage wave is the same as the input current.
(iv) The output voltage def'mes the phase relationship of the primary current by
flattened voltage wave of the same duration.
Generally it is found that systems of the older type use saturation of iron and pilots
with an adequate insulation level to withstand the high peak voltages on the pilots,
whereas more modern systems developed for use with telephone type pilots will
favour the use of nonlinear resistors. The use of either method, however, has the
effect of removing the amplitude relationship to the input current and retaining
only the phase relationship.
 •uo!~oo, oad aapool ,.tun ~o suo!~eo.qdde ,{[aea oql m. posn se,~
aId!ou.ud s.rq,L "uo!~oouuoo ~m.,~eIa:t aql jo tuoIqoad o!seq aq~, atuoo:ta^o ol aaoo loF.d
pa.nO e ~Ju!sn 'tua~s£s ~uoaano I~!~ua:taj3!p ona~ ~ s! paq.uosop ~,uotuo~Ju~aa~ OtLL
"pas.~aouaun
s! as pu~ 's~lnu3 l~Uaa~xa uo dool ~oI!.d aq~ uo s~u!od Ie!~uo~od!nbo ssoaa~ s!
uo!~oouuoo £~IO:t aq~ '£~!o~d~o ~oF.d ou ~od "spua q~oq :to pua ~aq~!a tuoa3 pa3 s~[n~j
l~Uaa~u! ~ma.np pos#aaua aae q~oq pue 'aaoo p2t!t0 s.tlt~ q~,!~ sa!aas u! aa~ s£~iaa Oq,L
"it I"8"01 "/~!d
m. u,~oqs se 'aaoa s,oF.d pa.rq~, ~ ptm sX~ioa o,~n, ~u!sn Aq pou.re~qo aq ueo uo!l
-oouuoo £~IO~ aldug.s e 'aO~leq ~ua~ma l-e!~uo~aj3t.p jo old!ou.ud o!s~q oq~ ~Ju!doo)I
•spua q~oq ~ arums aq~ 's! ~q~ 'I~O.u~atutu,(s
aq ~sntu uo!~oauuoo S!tLL "s~ln~J l-ema~xa ptm ~uaa~m. uaaA,,~aq uo!~u!tu!aos!p
jo £~!I~nb oql m.e~aa ol ,(aessaoau s! uo!~oatmoo aaqlo atuos pue aiqeo.qdde aa~uo I
ou s! ,(ela~ :~m.odp.nu Fmo!~uaAuoo al 0 ~q~ su'~atu s~u!od ~m.g~la~ o t ~ aoj paau s.ns,,L
•£1IeBua~ajj.rp pa~oatruoo ~Jm.aq s~u!od ~u!Aeia~ asaq~ 'aa~leaaq ~,!noa!o pa~e!ooss~ s~!
~u.qIoa~,uoo ~,u!od ~Ju!geIaa e pua qoea ~,e a^eq o~, paau e s! aaaq~, 'aapoaj e jo spua o,a~,
aq~, ~u.qloaluoo saa~i~aaq l!noa!o aq~, uaa/~J,aq uo!leaedas alq~:tap!suoo at 0 q~!A~
AelaJ auo 6u!sn Lua.~sAsle!iuaJatl!p o!se8 VL'8'OL "6!,.I
,,l=lll
8 d!JJ.- Oi 0 B~
-i-
ale - V d!J.L
XP.laH
•sia~leaa q ~.mo:t!o q~oq [o:quoo
u~o qo.rqA,, pu~ spgaI ~u!~oouuo~aa~u! aq~ jo ~u!od l-e!~ua~od!nba aq~ ~g po~oo[
s! qo.nt,a ~u!od ~u!A~loa auo opp, oad ol alq!ssod s! 1! luatua~Ju~aae oldtu!s ~ qans q~!A~
"A~oI aq os~ ~ aou~!oed~o oaooao~u! I~O~ oq~ ~aoqs oar suo!~oouuooao~u! oq~ sV
•,~oi £1~Jmp.uodsaaaoo s! uapanq VA oq~ pu~ 'aou~s!soa ~Jm.p.u~ aatuaojsu~a~ ~,uaa:mo
aq~ q ~ pa:tedtuoo ,~,oI s! sPeoI 8u!]oouuoo aq~ jo aouepodtu! aq~ 'V['8"O[ "~!d
m. u,aoqs tuo~s~s l~!~uaaojj!p o!seq oq~ q~!,a ~ut.la~S "po~!ooadde oq X~tu 'u~J!sap
a!aq~, ~Jm.ouanIJm. sao~,oej oql pu~ 'stuo~,sAs sno!ae^ jo sop,S.tla~,oeaeqo aq~, ~,~q~, as uop,
-~aaptsuoo slt:totu uo!~oalo~d .~apooj ol aid!area, d Iep,uaaojj!p oq~, jo uo!l~o!Idd~ OtLL
sold!au.ud aisufl I'8"01
uo!$aa$oJd aJ!~-~Ol! d 8"01
16l suJa,~sAs~uaJ./no-Ja./jJeo pue aJ./M-,]ol./d :uo!joa,~oJd Japaa_I
192 Feeder protection: pilot-wire and carrier-current systems

A B

--12..-MI. . . . . . .. d

|
Trip A Trip B

Relay-connection voltage
h d
E E
A m B

V(,Itage diagram for external fault

Fig. 10.8.1B Two relaying points connected by a third p i l o t core

I r
!
I

v
v

Z Z

Fig. 10.8.1C Principle o f voltage balance

It was realised early in the history of feeder protection that an alternative solu-
tion to the problem of a suitable relaying connection was the use of a 'voltage
balance' arrangement. In such an arrangement the secondary currents are replaced
by, or converted to, an equivalent secondary voltage source of fairly low impe-
dance, and these are then compared round a pilot loop of two cores, as shown in
Fig. 10.8.1C (see previous Section on current source and voltage source.)
 Feeder protection: pilot-wire and carrier-current systems 193

Any comparatively low-impedance load across the current transformer would

serve the purpose of converting a secondary current into a secondary voltage and
of providing the necessary low source impedance. In early systems, an air-gap
introduced into the current transformer core was used for this purpose. In later
systems, air gaps in auxiliary summation transformers or loading impedances pro-
duced the same effect. It should be noted that the voltage balance system is still
fundamentally a differential system and it is only the secondary mechanism of
balance which has changed.

10.8.2 Practical relay circuits

In practice, the presence of pilot capacitance and current transformer inaccuracies

require some modification to the basic circuits previously described. In simple
differential protection where the two ends are close together, pilot capacitance is
negligible and inaccuracies are overcome either by using biased relays or high
impedance relays (see Section 10.8.1). High-impedance relays are not generally
applicable to feeder protection; biased relays are and appear in many practical
systems. The application of biased relays can be sprit into two broad classes:
(a) Those having true current source secondary energisation such as in Fig. 10.8.1B
(b) Those having secondary voltage source energisation such as in Fig. 10.8.1C.
In the first class, bias is derived from the current circulating round the pilots
during through faults. It is therefore equivalent to a simple biased differential
system.

'l_
i~ IB I
..... / \ %j

(a) Balanced voltage on exlernal l'aulls

I I

r-L-l r--L~
lOP I IOPI
t..Tj t-r-J

OP - Operating winding
(h) Balanced current on externul faults
B - Biasing winding
Fig. 10.8.2A A Iterna tive arrangements for biased-differen tial feeder pro tec tion
194 Feederprotection: pilot-wire and carrier-current systems

In the second class, we can take a different approach to understanding the func-
tion of a biased relay. The connection of such a biased relay is entirely local and
there are thus only two quantities which can be used for operation or bias, namely
the current in the pilots and the voltage across them. The relay receives information
about the conditions at the far end due to the effect that the far end secondary
voltage has on the voltage and current at the relaying point.
With this general approach to the second class, there are two possibilities, as
shown in Fig. 10.8.2A. In one case, (a), the pilot voltages oppose around the pilot
loop during external faults. In this case the operating quantity on the relay is pilot
current and the biasing quantity is pilot voltage. In the other case, (b), the pilot
voltages are arranged to be additive during external faults thus reverting to a circu-
lating current principle but with voltage source. The pilot current is then the biasing
quantity, and the pilot voltage the operating quantity.
The various basic principles described have advantages and disadvantages when
applied to practical systems using practical pilot circuits. These have led to con-
siderable variety of detail in proprietary forms of protection and some change in
fashion as the needs of power systems have changed, or the introduction of new
components or techniques have favoured one arrangement or another.

10.8.3 Summation circuits

These have previously been described and it will be sufficient to say that the device
generally used is a simple summation transformer or its equivalent.
In addition to the function of current summation to provide a single-phase
quantity, the summation transformer may have other uses as follows:
(a) Ratio changing: The high values of pilot loop impedance, for example, up to
2000 I2 or more, make the direct connection of pilots and current-transformer
generally unsuitable. It is necessary therefore to change the current level
between the pilot loop and the current transformer secondaries, and this is
usually done by stepping down the current level on the pilot side of the
summation transformer. Turns ratios of the order of 1:10 or more may be
used thus reducing the burden that the pilot circuit imposes on the current
transformers to an acceptable level (Fig. 10.8.3A).
(b) Isolation of pilots: The two windings of the summation transformer can be
used to isolate the pilots from the current-transformers. Specifications usually
require current transformer secondaries to be earthed. Pilots are unearthed.
The isolation between the pilots and the current transformers will therefore
prevent a spurious earth connection between ends. Such a connection would
be very undesirable. The provision of high insulation level on the summation
transformer can also isolate the current transformer circuits from possible
overvoltages induced in the pilot, or earth potential differences between the
two ends.
(c) Conversion from current to voltage: It has previously been noted that one
 Feeder protection: pilot-wire and carrier-current systems 195

__i
Say 2000[]
20~. t
-l
I'10
Fig. 10.8.3A Use of summation transformer to give change in impedance

method of doing this is to introduce an air gap into the magnetic circuit of
the summation transformers, the latter being fed from solid core current
transformers. Impedance loading of the secondary winding of the summation
transfomer can be used as an alternative method (Fig. 10.8.3 B),

Air gap
,

J V ip
Zs . . . . . . . . v -- ~ z~
7. . . . . . 0 where Zs is ()pen circuit
I'n impedance ()f s u m m a t i o n
transformer
(a) Air gap transformer

lp

t
I:n
I Z~
-

•w ' ,
VI Ip
Vl=-- Z
Nn

(b) L o w value secondary i m p e d a n c e

Fig. 10.8.3B Conversion from current source to voltage source

10.8.4 Basic discrimination factor

The discriminating quality of a differential system can be assessed in terms of a

factor defined by the ratio of the degree of correct energisation of the relay under
internal fault conditions to that which occurs (and is unwanted) under external
fault conditions at the same primary current. In a perfect system, this factor would
196 Feeder protection: pilot-wire and carrier-current systems

S
.I
-N
II,,
A

N -
L = ~urrent I

1 F
B,-q
v

~lf

lf/N Stability
limit

(a) Unhiased relay (Discriminating factor AC

~1-~)

A
Current necessary .x'~~ 1 ~
to operate relay ~,4~ !
c~n external fault

J
(-N
(
-1_
If

| ii -

o lf/N C ~
|

Stability
(b) Biased relay (Discriminating factor -- ~AC ) limit

Fig. 10.8.4A Discriminating factor

be infinity and this is theoretically approached in a simple differential system. In

practice, differences in the characteristics and loading of the current transformers
reduce this factor. This is shown in Fig. 10.8.4A which gives the characteristics for
(a) an unbiased relay, and (b) a biased relay. In the case of feeder protection, even
if these practical limitations are ignored, the maximum discriminating factor
possible is governed by the characteristics of the pilot.
Part 1, Section 10.7.1 of the chapter, showed the characteristics of a differential
system by plotting the operating current in the relay for both external and internal
 Feeder protection: pilot-wire and carrier-current systems 197

fault conditions, together with the setting or bias characteristic of the relay, against
a base of fault current. If a discriminating factor is assigned, then it can be seen that
the ease with which a relay characteristic can be fitted to satisfy both the internal
and external fault conditions (with an adequate margin for both conditions)is
directly dependent on the value of the discriminating factor. With simple
differential systems the choice of relay characteristic is relatively easy, since the
theoretical discriminating factor is inherently high. Practical inaccuracies increase
the problem but a satisfactory characteristic can usually be chosen.
In differential feeder protection, the theoretical discriminating factor is limited
by the pilots and in extreme cases can create a problem of choosing a satisfactory
relay characteristic. It is useful to consider the relationship between pilot capaci-
tance and discriminating factor for the two basic types of circuits as follows:
(a) Current-balance system: With a simple current-balance system, as shown in
Fig. 10.8.4B, the capacitance of the pilots does not introduce unbalance on
external faults but will slightly reduce the current available under internal
faults. The basic discriminating factor is still infinity, and the pilots do not
play a great part in this case especially as the pilot lengths are short.

Fig. 10.8.4B Effect of pilot capacitance with two-core pilot system

(b) Voltage source systems: Where secondary voltage sources are used to energise
the pilots, the effect of pilots on discriminating factors may be examined in
general terms. Two schemes are possible depending on whether the voltages
are in opposition or in addition round the pilot loop under external fault
conditions. In both cases the relationship between the pilot voltage and pilot
current is the criterion which governs discrimination between internal and
external fault conditions. With the basic circuit, as shown in Fig. 10.8.4C,
three simple conditions of energisation can be used, namely:
(i) from one end only;
(ii) from both ends, the voltages being in opposition round the pilot loop;
and
(iii) from both ends, the voltages being in addition round the pilot loop.
The first condition must correspond to an internal fault fed from one end
only. The second and third conditions will correspond either to an internal
fault fed from both ends or to an external fault, according to the polarity of
the interconnection. In all cases, the condition governing the operation or
non-operation of the relay will be the ratio of pilot voltage to pilot current or
its reciprocal.
198 Feeder protection: pilot-wire and carrier-current systems

C "~
i .L ; , < O

A Z EA I z 13

c
I- I
I
I"" o

t - • - v
l,
= r
T t
I
Main c o n d i t i o n s :
(a) E B = 0 . . . . Internal fault fed f r o m end 'A'.
(b) E A = E B . . Voltage balance: pilots can be considered
open-circuited at m i d - p o i n t .
(c) E A = -E B . . Circulating current: p i l o t s can be
considered short-circuited at m i d - p o i n t .

Fig, 10.8.4C Effect of pilot capacitance with voltage balance system

10000

6000 "

4000 -

E
,- 2000 "
O
¢)

¢~ 1000 "
E
.,.,.

~ 6o0 -
e ~ .

< 400 -

J 7 / 0 . 6 7 m m s/c

too - i I ~ i ......
i ~ I' -
2 4 6 8 ,0 20 40 6D 8~) ,DO
Pilot length (km)
Fig. 1 0 . 8 . 4 D Open-circuit and short.circuit impedance characteristics of typical pilot
circuits
 Feeder protection: pilot-wire and carrier-current systems 199

This ratio is the effective pilot impedance or admittance seen by the relay
at the pilot terminals under the various fault conditions. This is strictly true
only if we ignore the unbiased setting of the relay, and this is permissible in
practice on most relays, particularly at high fault currents.
This useful generalisation enables the discriminating factor to be expressed
in terms of the impedance characteristics of the pilots, making allowance for
the impedance with which the pilots are terminated. Typical open-circuit and
short-circuit characteristics for two types of pilots are shown in Fig. 10.8.4D.
From Fig. 10.8.4C it will be appreciated that the open-circuit and short-circuit
impedances are important because:
(a) The short-circuited impedance of the whole pilot length represents the impe-
dance seen by the relay on a single-end fed internal fault (assuming zero
termination impedance).
(b) The open-circuit impedance of half the pilot length is the impedance seen by
the relay under external faults, on a voltage balance system.
(c) The short-circuit impedance of half the pilot length is the impedance seen by
the relay under external faults on a current-balance scheme using voltage
sources on the pilots.
The separation between the two curves (open-circuit and short-circuit) of a pilot
is therefore a measure of discriminating ability. This is only true if the relay

-R R

ZI = ~ (With voltage 0
II at one end
only)

Z2 = i ~ (With voltages
12 assisting)

Z3 - ~ (With voltages
13 in opposition)

Z3

Fig. 10.8.4E Complex impedance diagram showing effective input impedance of a pilot
circuit for various conditions of pilot energisation
200 Feederprotection: pilot-wire and carrier-current systems

concerned is responding only to the magnitude and not phase-angles of the

quantities.
If the pilots are long enough the impedance seen at the terminals will approach
the characteristic impedance, and the relay at one end will not be able to detect
either short-circuit or open-circuit condition at the other. In other words, the
system may lose all its discriminating ability unless some additional feature is
introduced.
Various discriminating factors are possible with practical lengths of pilot, and
the case of the 5.68kg/km conductor pilot shows the order of difficulty
encountered.
Fig. 10.8.4E shows approximately to scale the complex impedances which occur
at the pilot terminals under the three basic conditions. The lowest discriminating
factor for each type of circuit would be obtained by taking the single-end internal-
fault impedance Z1 and comparing it with the corresponding external-fault impe-
dance. For the system where voltage opposition corresponds to an external fault,
the discriminating factor would be Z3/ZI. For the case where voltage addition
(current balance) corresponds to external faults, the discriminating factor would be
ZI/Z2. Both these quantities may have the same approximate scalar value but in
the former case there is a bigger phase angle difference between the two quantities.
The fact that the two quantities may have this phase difference is important as the
effective discriminating factor may be increased using a relay device which is
sensitive to the phase angle of the impressed quantities as well as their scalar ratio.

10.8.5 Typical pilot circuits

(a) 7/0.67 underground pilots: These were among the earliest to be applied and
grew up with cable systems where they could be economically laid at the same
time as the main cables, thus avoiding additional trenching costs. The core section
was 7•0.67, similar to that used in local auxiliary wiring. Insulation was about the
same level, that is 2 kV test. The cores used for protection are usually wormed to
reduce inductive interference. To suit some types of protection, separate screening
of individual protective cores may be required.
(b) Overheadpilots: For a differential protection applied to overhead lines where
the pilots would have had to bear the whole laying costs, overhead pilots strung
either on separate poles or on the overhead line towers are sometimes used. In the
latter case they are frequently combined with the earth wire for the circuit to form
a composite cable known as an overhead pilot earth wire. In the case of overhead
pilots, and especially those strung on the same towers or poles as the primary
circuit, particular care is necessary as high longitudinal voltages may be induced in
the pilots because of mutual inductance between the primary conductors and the
pilot. This condition is severe in the case of earth faults because of the high zero-
sequence coupling. A wormed arrangement for the protective cores is necessary to
avoid unbalance of these induced voltages causing a voltage around the pilot loop.
 Feeder protection: pilot-wire and carrier-current systems 201

Where open cambric-covered pilots are used (generally abroad), frequent trans-
posing is necessary. A high degree of insulation is required, and for overhead pilot
earth wires special types have been manufactured with both magnetic and electro-
static screening in order to reduce the induced voltage. The induced voltage
depends upon the length and construction of the line and the value of earth fault
current, so that it is usually expressed in V/A/kin for earth fault conditions. A
value of 1.9 is generally used for pilots without special screening. Even with this
special screening the induced voltages may be relatively high (0.6), for example a
16 km length line with an earth-fault current of 5 kA would give an induced voltage
of 5 kV. The value would be 15 kV for the unscreened type. Relay equipment
connected to such pilots will thus require either isolating transformer coupling or
an adequate level of insulation. In the latter case adequate labelling warning that
the relay equipment and pilots are subject to high induced voltages is advisable,
particularly if this equipment is mounted at the relay panel.
(c) Telephone type pilots: For overhead lines, the provision of special pilot
cables for protection may often be uneconomic and underground telephone cables
are frequently used instead. These may be privately owned by the supply company
and form part of their general communication and control network or they may be
rented. The type of cable is similar in both cases and presents problems in the
design of suitable protection. Where pilots are rented, certain additional difficulties
exist because of the limitations of voltage and current which are laid down. In
addition inductive loading coils included for communication reasons create
difficulties. The problems in the use of telephone pilots, particularly those of the
rented type, have resulted in the design of special protective systems which are
described later. More recently, the growing nonavailability of solid copper-pilot
channels has created further problems which have resulted in the development of
phase comparison systems of protection, using voice frequency carriers, which can
be applied to a variety of telecommunication company channels. This is described
in Section 10.8.9.

10.8.6 Typical systems for privately owned pilots

There are a number of forms of pilot.wire protection in use and it is not possible
in this chapter to cover every type. A representative range will serve to illustrate the
problems involved and their solutions. Some of these systems were designed many
years ago and are still in use. In general, these are suitable for use with 7/0.67 pilots
with their relatively high insulation level, whereas the more recently developed sys-
tems or modifications of systems are often suitable for use with privately owned
telephone pilots.
(a) Translaysystem, Type H (GEC Measurements): This is a well known system
which has been used for many years and installations will be found in many parts of
the world. In its basic form it is suitable for 7/0.67 pilots, although some modified
arrangements have been used with privately owned telephone pilots. It is a balanced
202 Feederprotection: pilot-wire and carrier-current systems

voltage system with the addition of a directional feature which enhances the
discriminating factor. The other feature which is worth noting is that at higher
currents, the nonlinearity of the iron circuits causes it to become a phase-comparison
system.

Induction-type
_ - relays

Shading ring

v-------x / - ~
\/
/x
o----- J \ ~ -----c

Fig. 10.8.6A Translaysystem, type H

(GEC Measurements Ltd)
The general arrangement is as shown in Fig. 10.8.6A, the terminal equipment
consisting of a special wattmetrical directional relay. Summation of secondary line
currents is effected in the top winding of the upper magnet. As this winding is
coupled to an output winding which in turn is effectively across the pilots, the
upper winding can be considered as acting as a summation current transformer. This
upper magnet has an air gap, and it can be considered therefore that the output
voltage has the characteristics of a voltage source; the upper magnet therefore
converts from a current source to a voltage source. In addition, the flux in the
upper magnet will be related to the voltage on the output winding and so to the
pilot voltage. This flux, which is one of the applied quantities on the relay, is
therefore related to the pilot voltage. The lower magnet is energised by a winding
which is connected in series with the pilots, and since the flux in this coil is the
second quantity applied to the relay, we have a relay which is effectively responsive
to the pilot voltage and the pilot current.
This type of relay develops a torque which is proportional to the product of the
two fluxes and the sine of the angle between them. It therefore develops a maxi-
mum torque when the fluxes are at 90 ° and zero torque when they are in phase. In
 Feeder protection: pilot-wire and carrier-current systems 203

this particular application, the fluxes would be in phase if the pilot impedance were
purely capacitive, and the relay would be insensitive to pilot capacitance current
flowing under external faults. The pilot impedance will be generally resistive on
internal faults so that a positive tripping torque would be developed provided both
fluxes existed, that is fault current existed at the relay terminal concerned.
The action of this relay may be considered in terms of the complex impedance
diagram developed for various fault conditions in Section 10.8.4(b). In plotting
the tripping zone of a directional relay, such as the Translay relay, on this diagram
(Fig. 10.8.6B) it will be seen that it generally includes those impedance vectors
occurring under internal faults but would exclude that impedance occurring on
external faults.
X
I
I
I
I
I
I
I
I
Stabilise I Trip
, R
0

I z~
1.:
I Z I - -i" w i t h v¢~ltage at
one end only
I
i.:
I Z2 = -i" w i t h wJItages
I assisting
z31 E
Z 3 --- ] - with voltages
I opposing
-X

Fig. 10.8.6B Simplified directional impedance characteristic of Translay H system

For a practical scheme, there would be conditions where the errors of current
transformers could produce unbalance in the pilots and consequently tripping
tendencies on the relays. It should be remembered that the existence of one flux
alone does not theoretically produce any torque on the relay. To give the relay
some of the qualities of a biased differential system, a shading ring is included on
the upper magnet which enables the relay to develop a restraining torque due to the
presence of the upper flux alone. This shading ring has an effect on the complex
impedance representation of the relay which is to distort the basic linear polar
characteristic of the directional relay, as shown in Fig. 10.8.6C. This enhances the
ability of the relay to distinguish between the impedance factors for internal and
external faults as well as providing this bias feature. Another practical modification
204 Feederprotection: p/lot-wire and carrier-current systems

/ / ~ " Characteristicv~ith
shadingring only
/
/
I
Stabilise I Trip
-R R

~ ~ ZI l'inal characleristic
\ ~ j with phaseshift as
"\k ~ I asshadingring

-X
Fig. 10.8.6C Curvature of relay characteristic produced by shading ring in Translay H
protection

to the relay which is used to further improve the stability on the external fault
impedance factor is to shift the angle of maximum torque slightly as shown, thus
increasing the separation between Z3 and the characteristic curve and the tripping
zone of the relay.
The performance of this system is generally as follows

Fault settings
Earth fault 22-40%
Phase fault 45-90%
Three-phase fault 52%
Speed of operation is about 6 - 7 cycles at five times fault setting. Being an induc-
tion pattern relay, the times vary somewhat inversely with the fault current.
It is generally suitable for 7/0.67 pilots with 2 kV insulation level and can cope
with pilot inter-core capacitance of up to about 3 oF, corresponding to a pilot loop
resistance of about 300 [2. The scheme is also suitable for 3/0.67 pilots provided
the pilot loop resistance does not exceed 1000 ~2. Because of the product nature of
the relay, i t i s a single-end tripping system, that is, only the end carrying fault
current is tripped on an internal fault. In cases where it is necessary to trip both
ends, d.c. intertdpping relays on the same pilots may be used.
(b) Solkor A system (Reyrolle): This is another system which has been success-
fully applied for a number of years. It is essentially a voltage-balance system
 Feeder protection: pilot-wire and carrier-current systems 205

~o

R coil

Bias coil

Fig. 10.8.6D Solkor A system (Reyrolle)

intended for use with 7/0.67 pilots. The general arrangement is shown in
Fig. 10.8.6D, from which it can be seen that the line current transformers are
summed in a solid core summation transformer, the output of which energises the
pilots in a voltage-balance sense.
The design of this summation transformer is such as to deliberately introduce
saturation of the core at about 1.5 times the c.t. secondary rating for an earth fault
energising the whole summation transformer primary winding. This use of a
saturated characteristic requires accurate standardisation of the summation trans-
formers during manufacture. The relays are of the rotary moving iron type, a.c.
operated and of high sensitivity, for example 22 mVA. They are of the biased type,
the bias winding being in effect energised by the pilot voltage through a winding on
the summation transformer. The operating winding is fed from a relay transformer,
in series with the pilots, which is shunted by a capacitor. At fault currents below
saturation, the summation transformer output is sinusoidal and the system is purely
differential, taking account of both the magnitude and the phase angle of the
primary currents at the two ends of the feeder. At high currents when saturation
takes place, the output voltage waveform is distorted and peaky as previously
described, and comparison is mainly on the basis of phase angle. Because of the
peaky output voltage the effect of pilot capacitance current on the relay would be
exaggerated, and the shunt capacitance is used to tune the operating circuit to the
fundamental frequency component. This component, of course, reaches a limiting
value and does not increase, and is related to the phase angle of the primary current
in a similar manner at both ends on an external fault. With such an arrangement at
the high current levels we have a stabilising angle of the order of +--60°. This arrange-
ment is very effective in that it enables fairly high stability limits to be achieved
with relatively low performance current transformers, provided their characteristics
206 Feederprotection: pilot-wire and carrier-current systems

Internal fault

t..

i
tm

.F.
t~
t..

tx Setting of Solkor relay

°

External fault

0
Fault setting Primary fault current

Fig. 10.8.6E Discrimination characteristics of Solkor A protection

and loadings are within acceptable limits.

The limitation of fundamental pilot voltage also limits the maximum value of
equivalent pilot capacitance current flowing in the relay under external faults so
that, with the addition of a moderate degree of bias, maloperation can be prevented.
The typical characteristics, which are shown in Fig. 10.8.6E, indicate the order of
discriminating factors obtainable. The discriminating factors in practice are
enhanced somewhat from the theoretical values because the tuned circuit is
resonant at the higher internal fault levels and gives an additional output of current
to operate the relay under internal faults. The general performance is as follows

Fault settings
Earth fault settings 40-60% (normal)
or 10-20%
Phase-phase faults 120-240%
Three-phase fault 140%

Operating time at five times fault setting is about 6 cycles. The system is suitable
for 7•0.67 pilot loops up to 400 I2. A modification to this scheme has been brought
out recently to overcome some of the problems of mechanical instability
experienced in some installations with the early design of rotary sensitive relay. In
this arrangement a transductor relay is used in place of the rotary sensitive relay, as
shown in Fig. 10.8.6F. Both the electrical and mechanical stability are improved
but the remainder of the performance is about the same.
(c) Solkor R system (Reyrolle Protection): This system was adapted from a more
elaborate system developed for use with rented pilots. By taking advantage of the
easier requirements of privately owned telephone pilots, a much simpler form of
 Feeder protection: pilot-wire and carrier-current systems 207

j Ad!usting resistors control shape of bias curve

Bias
_ .

I
i

" I1
d p

p
i

( p

(
-,-, ! | , p

¢ r

p d
d ]
t p

", i
i p
m
(

t
-, i .
i,
p

Operating

Tuning impedance I I ~"~ Electro

of shunt capacitor ~~---~-.-~~ ~" magnetic
-,, rela y
C- ~ - ~ . . . . . . . . . .

Fig. 10.8.6F Transductor relay for Solkor A protection

protection was obtained. However, the basic principles are the same and these will
be explained here rather than under the rented-pilot type system. The method of
operation may be explained in two different ways. First, the system can be
considered as a current-balance type with the relays shunt-connected at the pilot
terminals. The problem of imparting differential characteristics and obtaining
correct relaying connections for such an arrangement has been overcome by intro-
ducing half-wave rectifiers into the pilot loop and the relay connection, as shown in
Fig. 10.8.6G.
The pilot loop contains a half-wave rectifier at each end shunted by a resistance
R a and so arranged that conduction is in opposite directions at each end. The relays
are connected across the pilot terminals in series with the half-wave rectifiers as
shown. Taking the external fault condition and considering successive half cycles
we can see that if R a = R p , then the equipotential point on the pilots will alternate

l
Fig. 10.8.6G Solkor R protection (Reyrolle)
208 Feederprotection: pilot-wire and carrier-current systems

R a Rp
+ : • "~"~P"~ Ra ~ 4"

i-, ,, ,,,,,

© ® External fault 0 0
Ra R R Ra Rp Ra
.i.--~ p a .~-- ÷ _
~ ~ - :

. . . . ÷ ,,. '--

Q Q internal fault Q Q

Voltages at relaying points I and 2

Q , © Ra=R
I'

©
®

V R a > R I,

Fig. 10.8.6H Principle of operation of Solkor R protection

between ends, one relay being at the midpoint on each half cycle. The rectifier in
series with the relay circuit is so arranged that it prevents current flow under the
polarity of voltage that the relay experiences when it is not at the midpoint. In
this way each relay effectively experiences a midpoint connection.
An alternative description of this action is that the rectifiers act as switches and
 Feeder protection: pilot-wire and carrier-current systems 209

permit the use of a single pilot loop for a two-way signalling action using a system
of time-sharing switched at the power frequency.
In a practical scheme, the resistor R a is made greater than Rp and so on external
faults causes reverse voltage on the relay rectifier to occur on both half-cycles, as
shown in Fig. 10.8.6H. This effectively displaces the midpoint connection further
towards the relaying end as shown. This particular feature is equivalent to applying
a bias signal to the relay, and effectively stabilises the relay on through-fault condi-
tions even with considerable transient errors in the current transformers. In most
practical cases R a can be made a pure resistance, and it is not necessary to mimic
the pilot capacitance. The pilot voltage is unidirectional on external faults, although
the current from the summation transformers is alternating. This feature, together
with offsetting of the midpoint, has the effect of reducing the influence of pilot
capacitance on through-fault stability during external faults.
Under internal-fault conditions, the relay is fed by half-wave pulses of current,
although for a single-end fault the distribution between ends is such that only the
relay at the feeding end can be considered operative. The system is therefore of
the single-end tripping type. In order to limit the peak voltage, Metrosil devices are
included across the output of the summation transformers. These limit the pilot
voltage to about 350 V and give the system a phase-comparison action. With these
voltage levels and a maximum loop impedance of 1000 f~, it is possible to use a
simple attracted-armature relay with adequate contact performance for direct
tripping of circuit breakers. The summation transformers are insulated to a level of
5 kV to cater for applications where there might be longitudinal induction.

Fault settings
Earth fault R = 2 5 % Y =32% B=42%
Phase-phase faults R-Y = 125% Y-B = 125%
B-R = 62%
Three-phase fault 72%

Operating time is about 3 cycles at three times setting. The system is suitable for
use with 7•0-67 pilots up to 30 km, 5-68 kg/km telephone pilots up to 15 km and
11.36 kg/km telephone pilots up to 30 km.
(d) Type DSB7 biased differential protection (GEC Measurements): This system
is based on the voltage balance principle and, as it is primarily intended for applica-
tion to teed feeders, the derived quantities at each relaying point are not limited,
but vary directly with the associated fault current. Fig. 10.8.61 shows a simplified
arrangement, two pilot loops being formed by a 4-core pilot, one loop summating
vectoriaUy the derived voltages of the relaying points for the operating quantity,
and the other loop summating arithmetically these voltages to form the biasing
quantity. The derived voltages are obtained from tapped air-gapped (quadrature)
summation transformers at each relaying point.
The principle of operation is that under normal load or through fault conditions
the summated bias quantity greatly exceeds the summated operating quantity,
210 Feeder protection: pilot-wire and carrier-current systems
II °
,,,,,j
¢:
I II
r~
ILl
m
FY"I
¢:
~b
f
',y
'z-,
--3
.9
I
¢:
.0
,.,,.,.. ,P..,
,._ .;.:
:3.-¢ ,9
._- - i%
.-. ,..,
o~
m
(JD
O0
0
t,-,,,
o..
UL
 Feeder protection: pilot-wire and carrier-current systems 211

which is theoretically zero. Under internal fault conditions, the operating quantity
at each end will exceed the biasing quantity with an adequate margin. The
operating and bias quantities are compared at each end by means of two windings
on a permanent magnet moving coil relay.
The particular requirements for a system of this type applied to teed feeders are:
(i) The derived quantities should be linearly related to the primary current,
and, on through faults, balance will be required for differing values of fault current
at the various ends. In this scheme the quadrature summation transformers are
linear up to thirty times rated current.
(ii) The biasing level should be adequate to overcome the real errors in
vectorial summation on through faults. With a linear system, this imbalance, for
example, due to pilot capitance, increases with fault current.
(iii) Perfect balance between current transformers operating at differing
values of primary current is not achievable, particularly under transient condi-
tions. The biasing level needs to be capable of stabilising against operation under
such conditions of through fault. This should also take account of differing designs
of current transformers at each end.
Additional stabilising features are included in the complete scheme. Two second-
harmonic tuned circuits are included in the operating circuits at each end, one to
f'tlter the harmonic from the operating coil and the other to provide an additional
bias related to this harmonic. This feature enhances stability on magnetising inrush
currents and also on transient distortion of current transformers.
A further feature is the inclusion of a low-impedance stabilising resistor in the
star connection of the current transformer secondaries.
The following figures give typical performance data:

Fault settings Summation transformer tappings give nominal settings for

earth fault between 20-33%, for phase faults between 50-
100% and three-phase settings of 58%.
Operating times from 60-80 ms at ten times the fault setting.
Stability This is nominally thirty times rated current but achieved
values depend on pilot length and characteristics. This may
reduce to about 10-15 times for pilots of 4 mF capacity
and 1200 ~2 loop resistance. Note, this is acceptable in most
cases as the fault current reduces accordingly with increas-
ing feeder length.

(e) TranslayS (GEC Measurements): This protection is based on the circulating

current principle and uses static phase comparators as the measuring elements. The
basic circuit arrangement in Fig. 10.8.6J shows a summation current transformer
T1 at each line end the neutral section of which is tapped to provide alternative
sensitivities for earth faults.
The secondary winding supplies current to the relay and the pilot circuit in
parallel with a nonlinear resistor, RVD, which is nonconducting at load current
212 Feeder protection: pilot-wire and carrier-current systems

Lint. c.t.,. I.i,1¢ t.t.s.

Ait . . . . . . . . . . . . . . --

(' ~ "

I°r I'r

,, I - - I I - - I ,. o --o-o

_ Io ,. II.
• '~
It'rlllill;ll.~ '\1" lt2 tl~t'd |¢~!" |llll)l '~ll|)~.'l'~ INillll
I,.rllli~l:d,, :\1' \.l tl~,~.',l I~r 1111,~!;11~ili',illg;llld illlt'flrlpp1111~

Fig. 10.8.6J TranslayS protection (GEC Measurements Ltd.)

levels and which, under heavy fault conditions, conducts an increasing current
thereby limiting the maximum secondary voltage. At normal current levels the
secondary current flows through the operate winding To on transformer T2 and
then divides into two separate paths, one through resistor Ro and the other through
the restraint winding T r of T2, the pilot circuit and resistor Ro of the remote
relay.
The currents in windings To and Tr are summated and fed to the phase com-
parator to be compared with the voltage across winding T t of transformer T1. The
voltage across T t is in phase with that across the secondary winding T s which in
turn is substantially the voltage across Ro.
Taking into account the relative values of winding ratios and circuit resistance
values, inputs to the phase comparator are:

(/,4 + 2IB) and (21A + IB)

where I A and I B are the currents fed into each line end and for through faults
IA = -lB. By putting each input in turn to zero we find the values of IB in terms
of/,4 for which the systems is stable. Thus,

/.4 4" 2I B = 0 and I B = -IA/2

2IA + IB = 0 and IB = -21.4

so that stability is obtained for values of IB outside the above limits because the
inputs to the phase comparator will be of opposite sign.
The phase comparator has angular limits of -+90° giving a circular bias character-
 Feeder protection: pilot-wire and carrier-current systems 213

, ~ / - : ~ i : ,~ • .i~:.: ::~ ~i., ~ ~:~:::::~):/~.~?i:~)~

"::~:" ' ..... • " ~:~i:/::i:~/'.!•,~~:i:I/I!/ .~~::: •:" ' ~ .~,,~:.~ :.~ :~:(~ •:.:: :....... ~ ~i • iii: ~II i~'I :i!):~:~:!i~

• • . . i•i••i:i~i: • ::"~:::• ~i
:" ~, :~. .... ~i .~,

~!! ::?:] i: i~ii.:~

....
"
: d

~ .i :41
" iii:!:i~i!ii:':

: ~::!!'~:~i~i!i
~
~:~:i i ~~ i~i~:::ii!iiii~
~ .,~:::- ::i,/~:: - ======================
/ii':i~!ii'!/!~!ili~

:: .... ,:~ .i,( ¸

Fig. 10.8.6K Translay S protection (GEC Measurements Ltd.)

istic in the complex current plane and in fact the comparator is equivalent to the
traditional low impedance biased relays which are often identified with differential
protection. Such relays have an amplitude criterion for operation such that

IG + IB I >~ K IIA - 1B I

and the equivalent phase comparator inputs are

I A ( l - K ) + IB(1 +K) and Ia (l +K) + IB(I-K )

By putting K = l/3, the phase comparator inputs become

23 (IA + 2IB) and ] (21 A + IB)

i.e. of the form used in Translay S.

214 Feederprotection: pilot-wire and carrier-current systems

The input circuits of the phase comparator are tuned to the power frequency so
that the threshold of operation increases with frequency. This desensitises the relay
to the transient high.frequency charging current that flows into the line when it is
energised and also reduces any distortion caused by saturation of the current
transformers.
In order to maintain the bias characteristic at the designed value it is necessary
to pad the pilot loop resistance to 1000 I2 and a padding resistor Pr is provided in
the relay for this purpose. However, when pilot isolation transformers are used the
range of primary taps enables pilots of loop resistance up to 2500 ~2 to be matched
to the relay. The equipment is illustrated in Fig. 10.8.6K.
The following gives typical performance data:

Fault settings Summation transformer tappings give nominal settings for

earth faults between 12-33%, for phase faults between
8-44% and three-phase setting of 51%. These settings may
be further varied between limits of 0-5 - 2-0 by a setting
multiplier.
Operating time This is adjustable by a special factor Kt so that the
operating time can be increased with a corresponding de-
crease in knee-point voltage requirements for the c.t.s.
Kt 40 20 14 6
Average time at 5 x
Setting current (ms) 30 50 65 90

10.8.7 Use of rented pilots

The significant application of rented pilots to feeder protection in the UK dates

from the post-war expansion of the transmission networks at 132 kV. At this
period the main protection for such networks had become phase-comparison
carrier-current protection which was both complex and expensive. Many cases
existed where the length of the overhead line was 50 km or less and in these cases it
was not felt that carrier protection was justified. Manufacturers of protective gear
were asked to investigate the possibility of extending pilot-wire protection to the
use of rented pilots. Most of the problems which have already been discussed are
emphasised in the case of rented pilots but, in addition, particular problems are
introduced by the conditions of renting from the telephone authority.
At this point it should be noted that it is feasible to apply differential principles
to extreme pilot conditions provided the protection is accurately matched to (and
compensated for) the pilot characteristics. This, however, was not the preferred
solution at the time as the need was for a standard form of protection which re-
quired the minimum of adjustment and had the minimum of dependence on the
pilot characteristics. Provided the many technical problems can be overcome, the
application of such systems is economically attractive compared with the alterna-
 Feeder protection: pilot-wire and carrier-current systems 215

tive forms of protection using distance relays or carrier-current equipment.

Before considering the particular systems resulting from this development it is
useful to review the special problems associated with the use of telephone pilots
as they are common to all designs.
(a) Discriminatingfactor: With the lengths of pilots involved, the characteristics
of high loop resistance and relatively low shunt capacitive impedances cause a
problem in respect of basic discriminating factor. This has already been referred to
but in this type of application it is particularly shown up by the fact that a 30 km,
5.68 kg/km cable gives a discriminating factor of two (without special compensa-
tion) this being reduced to 1.3 if the length is increased to 50 km. The acceptance
of the basic discriminating factor, which is common with 7/0.67 pilots, is no longer
possible with the lengths of rented pilots required, and additional design features
must be included to enhance this factor. There are a number of different ways of
doing this all of which involve some complication, limitation or loss of flexibility
in the overall equipment. Some of the methods which may be used are as follows.
O) Tuning of pilot capacitance: With a voltage-balance arrangement, the
effect of pilot capacitance current on external faults may be reduced by connecting
a chosen value of inductance across the terminals of each end of the pilots, as
shown in Fig. 10.8.7A. This inductance would normally be tuned to half the pilot
capacitance, the arrangement being symmetrical. The discriminating factor can be
improved up to about 6 by this technique, the distributed nature of the pilot
resistance and capacitance limiting the sharpness of tuning which may be achieved.

I Pilots I
Operate I I
R
Y
Restraint I .L
B

N
L tunes w i t h o p e n - c i r c u i t I I
i i
c a p a c i t a n c e of half pilot length

Fig. 10.8,7A Tuning of pilot circuit capacitance

In any case, it is necessary to limit this so that the effect of system frequency varia-
tions should not be pronounced. This improvement of discriminating factor is
further limited because in most systems the use of lirniting produces nonsinusoidal
waveforms and such compensation can only be effective at the fundamental fre-
quency. This method has been successfully used but it involves a compensation
which must be suited to the particular type of pilot and route length. If the pilot
is changed then the compensation must be altered accordingly.
(ii) Use of complex-impedance relay characteristics: As previously described,
improvements in discriminating factor can be achieved if account is taken of the
phase angle of the effective pilot impedance as well as its magnitude. This technique
 216 Feederprotection: pilot-wire and carrier-current systems

has already been used in the Translay H system with respect to 7/0.67 pilots, and
this system was developed in a modified form to suit telephone pilots up to pilot
loops of 1000 ~. With the considerable developments in the distance relay field
and sensitive relays with a variety of complex-impedance characteristics, it is
possible to adapt such relays to pilot-wire protection.
Oil) Compensation for pilot capacitance current: More exact ways of com-
pensating for pilot capacitance current are possible compared with the simple
tuning technique referred to above. The technique is generally to create a replica
impedance corresponding to the pilot under external fault conditions. Such impe-
dances are energised by the pilot voltage or some voltage corresponding to this, and
the current through them is used to counteract the pilot capacitance current or
create more exact balance points. This method does not involve tuning but again
is dependent on the type and length of pilots. One possible arrangement is shown in
Fig. 10.8.7B.

Bias coil

Operating coil
to-
\ /
\ /
\ I
\I
ix
IL
/ \
I \
I \
X-o-

Replica
impedance

Fig. 10.8.7B Pilot compensation by replica impedance

(iv) Use of current balance: As previously noted, the basic discriminating

factor of a current-balance arrangement can be fairly high if the problem of correct
relaying connection is solved. There are methods of overcoming this problem with-
out resorting to a three-core pilot. One method which is shown in Fig. 10.8.7C uses
a replica impedance at each end which is arranged to form one arm of a bridge, the
other arm being one half of the pilot loop. Under external fault conditions, the
pilots may be considered as short-circuited at the midpoint and each end finishes as
a balanced bridge, the relay being theoretically unenergised. A bias signal may be
obtained from the circulating current to ensure stability under practical conditions
of external faults. Under internal faults, each relay is energised equally, even when
 Feeder protection: pilot-wire and carrier-current systems 217

_~ P"L ~ ' -

-C •
p,2
l~J'4
Replica impedances Rp,S/~
llm

Bias c~}il

,I . • - - - l _ I.,{
L ~i 0 .... 0 •

resistance

t
,
!i
,,
**
<.__

Currents f(}r external fault

-....>
--.-> <___

tl't t
t
<----
Currents for single-end internal t'ault

Fig. 10.8.7C Pilot compensation in current balance system by replica impedance. Replica
impedance set to impedance of short~circuited half pilot section

the fault is fed from one end only. This system requires the adjustment of the
replica impedances to suit the type and length of pilot.
(b) Regulations affecting the use of telephone type pilots:
(i) l~'lot voltage and current: The conditions governing the use of telephone
type pilots in power systems are particularly stringent, and when the use of such
pilots was considered for protection these conditions had to be interpreted and
eased by negotiation with the telephone authority. The conditions generally related
to safety aspects and external interference rather than the electrical capabilities of
the pilot. At present the regulations require that the peak voltage developed on the
pilots under maximum fault-current conditions shall not exceed 130 V and that the
218 Feederprotection:-pilot-wire and carrier-current systems

pilot current shall not exceed 60 mA r.m.s. These values are in excess of the normal
values applied to Post Office circuits and were only permitted because of the short
duration of fault conditions.
(ii) Insulation and protective gaps: The present regulations require all equip-
ment connected to the pilots to be suitably insulated from power-system equip-
ment and circuits, for example current transformer, d.c. and a.c. auxiliary circuits.
The level of this insulation is 15 kV r.m.s, and screens (Fig. 10.8.7D) must be
provided. This is a high level of insulation when considered in relation to the
design of relays and auxiliary current transformers, and has greatly contributed
to the size and complexity of the overall equipment. Protective gaps on the protec-
tive gear equipment are also required and these are usually of the glow-discharge
type with a setting of about 750 V. The introduction of these gaps does not affect
the protection provided they do not fire under normal conditions (including
system fault conditions).

Insulation for 15 kV

Operate

I I
' ii ' Telephone
pilots
I I
[ I
± +I _ °---

750 V Protective gaps

Fig. 10.8.7D Use of insulation and protective gaps on system using telephone pilots

(c) Relays power and limiting: The values of permitted maximum voltage and
current on the pilots, taken in conjunction with high loop impedances of 2000
or more, represent a serious limitation to the power which it is possible to transmit
over the pilots. Taking the maximum peak voltage of 130 V and sinusoidal wave-
form, the maximum power possible in the relays for a voltage-balance system would
be about 4 W (in the local relay) for fault current fed at one end only. This would
be for optimum impedance termination which in practice may not be possible.
The actual power in the relay would be nearer 0.5 W and this would be at the
maximum level of fault current. With a linear system where the voltages are pro-
portional to the fault current, this would give a power of the order of 0.2 mW at
the relay setting, assuming an earth-fault setting of 50% and a maximum earth-
 Feeder protection: pilot-wire and carrier-current systems 219

Maximum pilot voltage

Perfect
I

"6

~l~ -- - - 7 " "-" Perfect limiter _

-" Practical limiter at operation

0 /~ Maximum
I.'ault setting fault current

Fig. 1 0 . 8 . 7 E Action of voltage limiting

fault current of 2500 %. This is a very sensitive relay, bearing in mind the contact
requirements and the insulation levels required. The general action of limiting is
shown in Fig. 10.8.7E, the most convenient form at present being the non-linear
resistor (Metrosil) which has a voltage/current characteristic defined by Y = CI~
where ~ is an index of about 0.2-0-3 and C is a constant. The limiting action of such
a device may be improved by the inclusion of a hybrid transformer and a resistor,
as shown in Fig. 10.8.7F, or by a multistage arrangement as in Fig. 10.8.7G.
The introduction of such limiting def'mes that the protective systems are
basically phase-comparison over the pilots. Zener diodes are also used for voltage

v - Cl~ '- ""

Metrosil only

1
v = cl~ ~ Rl c > A

Metrosil and hybrid

0 ] ~ ~ C u r r e n t

Fig. 1 0 . 8 . 7 F Improvement to limiting using hybrid transformer

220 Feeder protection: pilot-wire and carrier-current systems

Single I
metrosil I

!
C i o

V characteristic

C - - -0
0 11 i1 i i
!

Fig. 10.8.7G Two-stage voltage lirniter

limiting in modern equipments.

(d) Requirementsfor pilot supervision: The pilot link is essential for correct
discrimination. In systems using buried pilots of the 7/0.67 type, the reliability of
the pilot system is rated fairly high. When telecommunication company pilots are
used for protection it is generally considered that the reliability is reduced, due to
the possibility of unauthorised interference, since these circuits form part of a
communication system and will go through junction boxes and exchanges. Even
though special precautions are taken, pilots may be short-circuited, open-circuited
or reversed, and it is becoming standard practice to provide supervision equipment
so that these conditions may be detected and an alarm given. In this respect, careful
consideration should be given to the three-phase fault setting under the faulted pilot
conditions mentioned. It is not possible to prevent maloperation of the protection
under through-fault conditions with such pilot voltages, but it is possible to keep
the corresponding three-phase settings above maximum load rating. Most
supervision schemes are basically the same and involve the circulation of a d.c.
current of about 5 mA round the pilot loop as shown in Fig. 10.8.7H.
A polarised relay at the receiving end of the pilots is operated by the super-
vision current. Any of the faulty pilot conditions of short-circuit, open circuit or

I Low-set relay contacts (sometimes omitted) I

Protection
- [ ' T ~ F7 (s°metimes omitted) 1['--'1

o,a,'so"r0, ,J I failure relay • Contact ~ n

[. -
low-set relay

failure + -- " I ~ Supply

alarm

Fig. 10.8.7H Typical arrangemen t for pilot supervision

 Feeder protection: pilot-wire and carrier-current systems 221

reversal will cause the supervision relay to reset and an alarm to be given. A monitor
relay at the sending end is arranged to detect the failure of supervision supply and
give an alarm.
The supervision supply is disconnected during fault conditions so that the super-
vision current does not affect the discriminating action of the protection. All alarms
are provided with time delays of three seconds or more so that alarms are not given
due to disconnection under primary fault conditions. The supervision supply is
usually provided from an auxiliary a.c. supply, through rectifiers and filters. The
Telephone authority regulations require a high degree of smoothing in order to
keep a low level of interference with adjacent communication circuits.
(e) Requirements for starting relays: The philosophy of starting relays has been
explained previously and it is only necessary here to state that they are prime
requirements because of the supervision aspect. The provision of starting relays
which have close settings, particularly for the three-phase condition, is very
difficult and this is particularly emphasised in the case of two-stage starting. Under
all conditions the starting relay settings must exceed those of the discriminating
link, so that there is no condition of starting relays being operated without a corres-
ponding adequacy of sensitivity on the discriminating link. The starting equipment
may consist of separate relays energised by the secondary line and residual currents
or, alternatively, a separate summation network without output relays. In both
cases the relays must have a high reset ratio, a high operating speed and equality or
near equality of phase and three-phase fault settings. The earth-fault sensitivity is
largely independent of the phase-fault sensitivity and does not constitute a problem.
(f) Pilot circuit characteristics: The pilot link in the case of telecommunication
company pilots is liable to rerouting when the necessity arises. In order to facilitate
design and application, it is normal to use a standard pilot-loop resistance. Where
the actual pilot resistance is below this value the loop is padded up by the addition
of resistances at each end so that the effective loop is always the same. The pilot
capacitance will vary with the route, and where pilot tuning is used, this must be
adjusted as well as the pilot padding resistors.
The upper value of 3500 I2 based on a conductor of, say, 5.68 kg/km makes dis-
crimination difficult even with the additional compensations which are used. In
specifying the performance of protection in terms of pilots it is common to express
the upper value of pilot-loop impedance and total pilot capacitance. This value of
capacitance will generally be lower than the equivalent length of pilot defined by
the maximum loop resistance. The 3500 ~2 figure is based on the inclusion of
lengths of 2.5 kg/km aerial conductor at the terminations, for example, between
the substations and the nearest exchange and, of course, the aerial conductors have
negligible capacitance compared with the cables.

10.8.8 Typical systems for use with rented pilots

Two types of protection are described, these represent widely differing solutions to
222 Feeder protection: pilot-wire and carrier-current systems
,,o
.2 :.,
- ~ I'L .L .~
T _ "1|
.2 T~.... ~ '~
- ~ ~
--~ ..~
_@ - "~2
,
:2 \
I
I
, !e
, !
I I
I I
I I I
I
I I I
I I I
I I I I
E I
I I I E
I I I ;- ~ ~ ~: ~ ,~
..j
I I I I I I 1 I I t I 'Y
× --"~
I I I
I I I
I I I
I
I.u
t:
-- F-.
.,
C~
-- 1~. '
e..=
I.I.
 Feeder protection: pilot-wire and carrier-current systems 223

the problems involved and therefore make an interesting comparison.

(a) TypeDSC8 (GEC Measurements): This is a voltage balance system in which
the relays respond to the apparent impedance of the pilot circuit. The general
arrangement is as shown in Fig. 10.8.8A. Operation of the protection may be
considered in two parts, the 'restraint' circuit and the 'operate' circuit.
The 'restraint' circuit comprises a shunt connection of R6, C6-a and (a) R2 l-a at
frequencies below fundamental and (b) C3 at frequencies above fundamental. The
current derived from the voltage developed across R~ is transformed by auxiliary
transformer T1 and rectified by the diode bridge (D1-4) to produce the restraint
component of the d.c. current fed to the differential relay (87). Circuit L~ C~ is
tuned to fundamental frequency to reduce the effect of harmonics on the restraint,
operate and pilot circuits. Capacitor C2 provides additional restraint during
transient conditions occurring at the inception of through faults, or their clearance,
when pilot capacity discharge current may flow.
The 'operate' circuit comprises essentially, two circuits in opposition: the half-
winding of T2 in series with the pilots and the pilot loading resistor Rv I and the
other half winding of T2 in series with R3 C4 (representing a lumped, constant
equivalent of half the maximum pilot shunt capacitance). The pilot shunt circuit
C5 Rva allows compensation for varying values of pilot capacitance and Rv2 allows
compensation for the resistance of varying lengths of pilot and the combinations
serves to improve the discriminating factor. The pilot padding resistor Rvl is
adjusted to half the difference between the minimum permissible pilot loop
resistance of 3000 ~2 and the actual pilot loop resistance in use. The output current
from T2 drives the 'operate' diode bridge (D5-8) and is approximately zero when
the pilot compensation has been correctly set.
The d.c. outputs of the two diode bridges summate algebraically and the
resultant current is measured on the moving coil unit (87).
Under through-fault conditions, ends A and B present voltages of opposite sign
to their pilot ends, but since the pilots are crossconnected, only a small value
current is caused to flow in the apparently high-impedance pilot loop. The much
larger value of current circulated through T~, therefore, restrains the differential
relay.
Under internal-fault conditions, the primary system current at one end is re-
versed in direction and the pilot terminal voltages are then such as to drive current
around the apparently low impedance pilot loop. The d.c. current in T2 is then
greater than that in the diode bridge of T~ and the difference current circulates
through the differential relay to cause operation.
Typical protection performance characteristics are shown in Fig. 10.8.8B.
The pilot supervision unit injects a small d.c. current into the pilot loop. At the
end remote from the supervision unit, the circuit comprising R4 and D9 provides a
low-resistance path to the supervising d.c. current, capacitors C6-8 shunting the a.c.
in the pilot circuit. The supervision unit provides alarms for 'pilot circuit faulty'
and 'supervision supply fail' conditions.
The tapped summation transformer (C) provides the 15 kV insulation required
to comply with Telephone authority requirements and in conjunction with resistor
224 Feederprotection: pilot-wire and carrier-current systems

6
<
4000 ohms
s

4- 3000 ohms
/
r,.) 0 ohms

1000 o h m s
.'~0 0 0 o h m s
0 ohms
End A rela~

End B relay

0.2 0.4 0.6 0.8 1.0

C u r r e n t at end B ( a m p s )

Pilots: 5.68 kg l o a d e d
End A c u r r e n t in phase with end B

(a) Bias characteristic with

varying pilot resistance

2.o - tA/o~o ÷40 °

IB ~ I A leads i B
1.5 - +30 °

1.0 - +200
• A

0.5 - + I0 °

tg /0 o
0.5 l.O 1.5 2.0 2.5 IB

0.5 - ~ ~/ - 10 °

1.0 . _ 20 °

1.5 _30 °

--40 o 1A lags I B
2.0

Pilots: 2 0 0 0 o h m s 5.68 k g / k m loaded

(b) Polar characteristics

Fig. 10.8.8B Characteristics of type DSC8 protection

 Feeder protection: pilot-wire and carrier-current systems 225
,/.
._
-.2"
I I
I
I "=-
r-- L ~_ oj
.o
. . . . . . I II _
_ _I ~-- ~
.i,_,_~
_ A-+.-~
I
¢J
o
g.
226 Feederprotection: pilot-wire and carrier-current systems

R~ develops voltages for the restraint, operate and pilot circuits. The voltage
applied to the pilots is limited to 130 V (peak) by the shunt metrosil (D). To ensure
that tripping cannot occur because of defective pilots, operation of the differential
relay auxiliary tripping relay (87X)is controlled by overcurrent and earth-fault
check relays (50A, 50C and 64). The performance is generally as follows

Fault Settings (including starting relays)

Earth faults 40%
Phase-to-phase and
three-phase faults 130%
Operating time at
5 x fault setting 2-3 cycles
Stability 30 times c.t. rating
Pilots Suitable for loop impedances up to 400012 with total
capacitance up to 2.1 taF with normal distributed loading.

(b) Reyrolle high-resistance pilot wire (HRPW) scheme. Some of the advantages
of a current-balance system have already been referred to, and this principle forms
the basis of a rented pilot scheme. The basic principles of operation have already
been explained in Section 10.8.6C. The general arrangement of the scheme is shown
in Fig. 10.8.8C. A conventional tapped summation transformer is used, this being
provided with the necessary 15 kV insulation and earthed metal screen between
windings. As previously described, the current in the relay connection is
unidirectional on internal faults. This current is used to control a transductor which
is in series with a telephone-type relay, the combination being fed from an auxiliary
constant-voltage supply obtained through a network fed by the c.t. secondary
currents. This transductor gives a switching gain of about 20, and helps to overcome
the problem of power limitation of the pilots while still permitting the use of a
relatively robust relay (15-20 mW) capable of tripping the circuit breaker directly.
The transductor also provides a convenient means of isolating the relay circuits and
contacts from the pilots. Limiting is obtained by nonlinear resistors across the
summation transformer secondaries, but in this case their limiting action is supple-
mented by a hybrid transformer arrangement.
There are two stages of starting in this scheme because the pilots are normally
short-circuited at each end by the low-set relays. A single.relay (STA) is used for
low-set and another (STB) for high-set, these being energised by the rectified
output of a special summation network. This network provides outputs for various
types of faults in the same way as a summation transformer but gives better
equality of settings. The low-set relays have normally closed contacts which short-
circuit the pilots. This provides a path for the d.c. supervision current, and so pre-
vents this current from flowing in the windings of the summation transformers.
This feature also has the effect of making the single-end fault settings independent
of the through-load. The low-set relays have contacts which normally short-circuit
the control windings of the transductors, and thus ensure that the discriminating
 Feeder protection: pilot-wire and carrier-current systems 227

link is completed by the operation of the low-set relays on external faults before
permitting the comparison to be effective. Supervision equipment of the type
previously described is included.

Typical performance figures are as follows:

Fault settings
Earth fault 50-60%
Phase fault 190%
Three-phase fault 160%
Operating time 2-3 cycles at five times setting
Stability At least 25 times c.t. rating
l~'lots Suitable for loop resistance up to 3500 ~2 and pilot capaci-
tance up to 2/.tF with standard loadings

10.8.9 V.F. phase-comparison protection (Reyrolle Protection)

This system uses phase-comparison principles and is a development of the power-

line-carrier phase-comparison protection described later in Section 10.10 adapted to
operate over rented Post office voice frequency channels. In designing a phase
comparison system to operate over this type of communication link a number of
restrictions not normally encountered with power-line-carrier phase-comparison
protection need to be overcome. In the case of rented telephone channels the
continuous power transmission must not exceed -13 dBm; for short durations the
power may be boosted to 0 dBm, Neither the length nor the composition of the
channel can be guaranteed to be constant during the period of service and may be
changed by automatic route switching without prior knowledge of the user. Attenu-
ation in the link may vary by -+3dB with time and may also vary b y - 1 to +4 dB
over the usable frequency range. Isolation transformers in the link may also add
losses of between 0.5 and 1.5 dB per transformer and the group delay over the
usable frequency range may be in the order of 1 ms. As well as accommodating the
characteristics of the channel the equipment must cater for the presence of random
noise induced into the channel before, or at the time of, power system faults.
The main items of the v.f. phase comparison scheme are shown in Fig. 10.8.9A.
At each end of the protected line a terminal equipment is situated comprising a
power frequency section which interfaces with the power system and a v.f, section
which is connected to the channel, e.g. a four wire v.f. communication circuit.
The basic principle of operation of the protection is the same as for power-line-
carrier (p.l.c.) phase-comparison protection. In the v.f. equipment information is
transmitted in the form of a frequency shift keyed signal. In this case the signal is
transmitted continuously and the frequency of the signal is switched between two
values in accordance with the modulation signal. The low-frequency signal would
correspond to carrier transmission in the p~l.c, case and the high frequency would
228 Feederprotection: pilot-wire and carrier-current systems

correspond to no carrier or gap. The continuously transmitted signal with no modu-

lation is at the high frequency and is therefore directly analogous to the p.l.c, case.
Thus, in the v.f. equipment the condition which determines tripping is the effective
duration of the high-frequency signal at each end of the line after the received
signal from the remote end and the local signal have been combined in an equiva-
lent manner to the carrier blocks in the pJ.c. case.
In the p.l.c, case the carrier signal travels at almost the speed of light and the
transmission delay incurred is normally accommodated in the stability angle setting
of the equipment. In the v.f. case the transmission delay may exceed 5 ms and is
not constant due to possible rerouting and cannot therefore be catered for within
an acceptable stability angle setting. This problem is overcome in the v.f. phase
comparison by measuring the total transmission delay around the channel loop and
delaying the local signal by half of the loop delay before comparison with the
remote signal which has itself been delayed by the communication link. In this way,
even for variations in routing, the local and remote signals maintain their relative
phase difference except for communication errors. As the communication errors are
small and substantially constant they are catered for in the stability angle setting.
IA IB

Internal ~ [ I';x ternal

fault ~ fault

section Rx 4 wire ]:x sectic,n

End A I'nd B

Fig. 10.8.9A V.F. phase comparison protection

The principle employed in the automatic compensation for transmission delay is

shown in Fig. 10.8.9B. Under healthy power system conditions a measurement
pulse, consisting of a short burst of low-frequency signal, is transmitted by the
master end to the slave end and is reflected by the slave. The master measures the
time from transmission to reception of the pulse and stores the value. A short time
after receiving the master end pulse the slave transmits a measurement pulse which
is reflected by the master and the transmission time is stored by the slave. This
process is repeated continuously and each end of the equipment computes an
average value of the total transmission delay which is used for automatic compensa-
tion.
The main features of the v.f. section of the equipment in relation to the com-
munication link may be summarised as follows:
 Next Page
Feeder protection: pilot-wire and carrier-current systems 229

Slave end
Master end
I I
O- A-------- I
Stored tA ~
loop
delay
I
I TB Time
I deh~y

I
I
I Stt~red
h,op
delay
I
Time rA 1
delay

t--'--'-

rA .J

F -I
Data Tx
Vk_
Data Rx --FL ........ I-'-I

V.I-. Tx ~°°;.,,, Wv'

V.!-. Rx
~ru _!_ 'B ._1
Data Tx ...... n? i,i_ ~;~i -h, -1 ....
Data Rx

V.F. Tx

V. !-, Rx

Tim~

Fig, 10.8.9B Measurement of propagation delay

(a) Continuous evaluation of the tr~nsmission delay including terminal equip-

ments.
(b) Automatic compensation for phase error caused by transmission delay.
(c) Signal quality circuits employed to detect the condition of poor signal to
noise ratio and prevent the measured tr~aasmission delay from being affected by the
noise.