Chapter 3

Fault calculation
by [Link]

3.1 Introduction

3.1.I Purpose of fault calculation

Fault calculation is the analysis of power system electrical behaviour under fault
conditions, with particular reference to the effects of these conditions on the power
system current and voltage values. Together with other aspects of system analysis,
fault calculation forms an indispensable part of the whole function and process of
power system design. Correct design depends essentially on a full knowledge and
understanding of system behaviour and on the ability to predict this behaviour for
the complete range of possible system conditions. Accurate and comprehensive
analysis, and the means and methods of achieving it, are therefore of essential
importance in obtaining satisfactory power system performance and in ensuring the
continued improvement in performance which results from the development and
application of new methods and techniques.
The applications of power system analysis cover the full range of possible system
conditions, these being divisible into two main classes, namely conditions in which
the power system is operating in a normal healthy state, and others in which it is
subjected to one or more of a wide variety of possible fault conditions. The analysis
of these conditions and their effects on the power system is of particular relevance
R) such considerations as:
(a) the choice of a suitable power system arrangement, with particular reference
to the configuration of the transmission or distribution network
(b) the determination of the required load and short-circuit ratings of the power
system plant
(c) the determination of the breaking capacity required of the power system
switchgear and fusegear
(d) the design and application of equipment for the control and protection of the
power system
(e) the operation of the system, with particular reference to security of supply
and economic considerations
 54 Fault calculation

(f) the investigation of unsatisfactory performance of the power system or of

individual items of power system plant.
The present chapter is concerned principally with the analysis of system fault
conditions, these conditions being of direct and particular relevance to the design
and application of power system protection. The methods of analysis employed,
however, are essentially applications of general analysis and, as such have equal
application to a wide range of other problems whose solution is dependent on
electrical network analysis.
3.1.2 Types of fault

In the context of electrical fault calculation, a power system fault may be defined
as any condition or abnormality of the system which involves the electrical failure
of primary equipment, the reference to primary (as opposed to ancillary) equip-
ment implying equipment such as generators, transformers, busbars, overhead lines
and cables and all other items of plant which operate at power system voltage.
Electrical failure generally implies one or the other (or both) of two types of
failure, namely insulation failure resulting in a short-circuit condition or conducting-
path failure resulting in an open-circuit condition, the former being by far the more
common type of failure.
The principal types of fault are listed and classified in Table 3.1.2A, and are
discussed in greater detail below.
Table 3.1.2A Types of fault

Three-phase fault clear of earth

Three-phase-to-earth fault
Short-circuited Phase-to-phase fault
phases Single-phase-to-earth fault
Two-phase-to-earth fault
Phase-to-phase plus single-phase-to-earth fault

Open-circuited Single-phase open-circuit

phases Two-phase open-circuit
[Link] open-circuit

Simultaneous A combination of two or more faults at the same time, the faults
faults being of similar or dissimilar type and occurring at the same or
different locations. Typical examples are the cross-country earth-
fault and the open-circuit-with-earth-fault condition

Winding-to-earth short-circuit
Winding faults [Link]-winding short-circuit
Short-circuited turns
Open-circuited winding
 Fault calculation 55

Short-circuited phases: Faults of this type are caused by insulation failure

between phase conductors or between phase conductors and earth, or both, the
result being the short-circuiting of one or more phases to earth or to one another,
or both. The full range of possible fault conditions of this type is illustrated in
Fig. 3.1.2A. The three-phase fault, which may be to earth or clear of earth, is the
only balanced or symmetrical short-circuit condition, the presence or absence of
the earth connection being normally of little or no significance unless the fault
occurs simultaneously with a second unbalanced fault involving earth. The three-
phase short-circuit is commonly used as a standard fault condition as, for example,
in the determination of system fault-levels, these levels being normally quoted as
three-phase short-circuit values.

Three-phase clear ~1"earth "l'hree-phase-tt~-earth

- - - t I
i
Phase-to-phase Single+phase-to-earth

ii
Tw,o-phase-to-earth Phase-to-phase plus
single-phase-to-earth

4 . . . . I

- --- i
Fig. 3.1.2A Shorpcircuited-phase faults

Open-circuited phases: This type of fault, illustrated in Fig. 3.1.2B, is the failure
of one or more phases to conduct. The more common causes of this type of fault
are joint failures on overhead lines and cables, and the failure of one or more
phases of a circuit-breaker or isolator to open or close. The single-phase and two-
56 Fault calculation

phase conditions are of particular interest because they both tend to produce un-
balance of the power system currents and voltages with consequent risk of damage
to rotating plant.
Simultaneous faults: A simultaneous fault condition, sometimes termed a
multiple fault condition, is defined as the simultaneous presence of two or more
faults of similar or dissimilar types at the same or different points on the power
system. Such conditions may result from a common cause, from different but
consequential causes or, extremely rarely, from quite separate and independent
causes. The commonest simultaneous fault condition is undoubtedly the double-
circuit overhead-line fault in which a common cause (for example lightning or
accidental contact) results in a fault on each of the two circuits concerned. These

Single-phase
()pen-circuit

~,pcn-circui!

Three-phast.
t~l~t.n-circuil

Fig. 3.1.2B Open-circuited-phase faults

two faults, although possibly geographically coincident, will be electrically separate

to an extent determined by the point of fault and the particular power system
configuration. A simultaneous fault condition of particular interest is that
known as the cross-country earth-fault, in which a single-phase-to-earth fault at
one point in the power system occurs coincidentally with a second single-phase-to-
earth fault on another phase and at some other point in the system. This condition
is most commonly experienced on impedance-earthed systems where the second
earth-fault may be initiated by the increased healthy-phase voltage resulting from
the neutral displacement produced by the first. As already stated, a simultaneous
fault condition may consist of two different types of fault at the same point, and
one example of this is the open-circuit-with-earth-fault condition in which two
faults, namely a single-phase open-circuit and a single-phase-to earth fault, occur
coincidentally on the same phase and at the same point in the power system. Such
a condition can occur on an overhead line for example, due to a phase conductor
breaking at a point near to a tower, the conductor on the tower side of the break
being held by the suspension insulator and that on the other side falling to ground.
The fault conditions described are divisible into two distinct classes, namely
 Fault calculation 57

balanced or symmetrical fault conditions and unbalanced or unsymmetrical fault

conditions, the former class comprising all conditions which are symmetrical with
respect to the three phases and the latter class the remainder. Of the faults listed.
only the three-phase short circuit (to earth or clear of earth) and the three-phase
open-circuit are balanced fault conditions, the normal balanced three-phase load
condition being a further example of the balanced or symmetrical condition.
Winding faults: The types of fault which can occur on machine and transformer
windings are illustrated in Fig. 3.1.2C and consist mainly of short-circuits, from one
phase winding to earth, from one phase winding to another or from one point to
another on the same phase winding. The last mentioned condition is known as a
short-circuited turns fault, and is of particular interest from the protection stand-
point in that the fault current in the short-circuited turns may be very large and
that in the remainder of the winding very small. The open-circuited winding condi-
tion is quite rare in practice and is usually the result of damage to the winding as a
consequence of a preceding winding short-circuit at or near the point of fault. Open
circuits in transformers may also occur as a result of failure of the tap-change equip-
ment.

Phase-to-earth fault Phase-to-phase fault

__ t'Y'V'V'g'v"r~ . . . . ~ ¢'V"WY% _.

Short-circuited turns Open-circuited winding

Fig, 3.1.2C Windingfaults

Changing-fault conditions: The types of fault which have been referred to can all
be regarded as fixed fault conditions, in that the type of fault remains unchanged
for the duration of the fault. The great majority of fault conditions are of this type
but there are bthers, known as changing-fault conditions, in which the type of fault
changes during the course of the fault. Such changing-fault conditions can result
 58 Fault calculation

from a number of causes, the most common being the spreading of a fault arc, or
of the ionised gases from a fault arc, to other phases and even to other circuits. A
typical example is a single-phase-to-earth fault which develops into a two-phase-to-
earth fault and possibly, later, into a three-phase fault. The analysis of a changing
fault condition presents no particular difficulty, since the condition can be
considered as a succession of fixed fault conditions, each of which can be analysed
individually.
3.1.3 Factors affecting fault severity

The severity of a power system fault condition may be assessed in terms of the
disturbance produced and the fault damage caused, the magnitude of the fault
current and its duration being of particular interest, especially in relation to the
design and application of power system protection. The factors which affect fault
severity must therefore be given due consideration in all aspects of power system
analysis in order to ensure results which are truly representative of the conditions
which can occur in practice. The factors which normally require to be considered
are:
(a) Source conditions: These relate to the amount and disposition of all connected
generation (including all other power sources such as interconnections with
other systems), the two extremes of minimum and maximum connected plant
being of particular interest. The minimum and maximum plant conditions are
normally those corresponding to the conditions of minimum and maximum
connected load.
(b) Power system configuration: This is determined by the items of plant, namely
generators, transformers, overhead-line and cable circuits etc., assumed to be
in service for the particular condition being investigated and by such other
factors as have a bearing on the topology of the equivalent system network.
The system configuration may change during the course of a fault with conse-
quent changes in the magnitude and distribution of the fault current, typical
causes being the sequential tripping of the [Link] at the two ends of
a faulted transmission line and the sequential clearance of multiple fault
conditions.
(c) Neutral earthing:Faults which involve the flow of earth current (for example,
a singie-phase or two-phase fault to earth, a single-phase or two-phase open
circuit, etc.) may be influenced considerably by the system neutral-earthing
arrangements, particularly by the number of neutral earthing points and the
presence or absence of neutral earthing impedances. Power systems may be
single-point or multiple-point earthed and such earthing may be direct (that
is, solid earthing) or via impedance, typical examples being the direct multiple-
earthing employed on the British 132 kV, 275 kV and 400 kV systems and
the single-point and sometimes multiple-point resistance-earthing commonly
employed at 66 kV and below. Earthing impedance can be used to limit the
earth-fauit current to a very low and even negligibly small value, as in the case
of a system earthed through a Petersen coil or a generator earthed through a
 Fault calculation 59

voltage transformer.
(d) Nature and type of fault: From what has already been said, it will be evident
that the type of fault and its position in the power system may have a con-
siderable effect on the magnitude and distribution of the system fault current,
this being particularly the case in respect of earth-faults as compared with
phase faults, open-circuits as compared with short-circuits and faults within
machine and transformer windings as compared with similar faults at the
winding phase-terminals. Similarly, the effects of a given fault condition may
be considerably modified by the simultaneous presence of one or more other
fault conditions as, for example, in the combination of a short-circuit and an
open-circuit phase condition. A further factor which may require considera-
tion is the possible effect of fault impedance (for example, fault-arc resistance
and the ohmic resistance of any metallic or non-metallic fault path, etc.), this
being of particular importance in matters relating to the design and application
of distance protection.
The wide range of possible system fault conditions and the many factors
which influence them result in a wide range of possible levels of fault severity,
ranging from extremely low levels up to the maximum levels possible for the
system being considered. It is therefore of value, in referring to fault severity
generally, to be able to refer to a standard fault condition, namely the three-
phase short-circuit, and to the level of fault severity produced by this
particular fault condition, namely the three-phase fault level. This level may
be expressed in amperes or, as is more usual, in three-phase MVA correspond-
ing to the rated system voltage and the symmetrical value of the three-phase
fault current. The three-phase short-circuit can normally be regarded as the
most severe condition from the point of view of fault severity, and it is
accordingly the maximum possible value of the three-phase fault level which
normally determines the required short-circuit rating of the power-system
switchgear. A factor which may also have to be taken into account is the
maximum value of the single-phase-to-earth fault current which, in a solidly-
earthed system, may exceed the maximum three-phase fault current.
The three-phase fault levels experienced in this country range up to 35
MVA at the lowest distribution voltage of 415 V up to some 35 000 MVA
at the highest supergrid transmission voltage of 400 kV, the maximum fault
current in the latter case being of the order of 50 000 A for a three-phase fault
and 60 000A for a single-phase-to-earth fault. Fault clearance times range
from less than a tenth of a second to one second or more depending on the
protective arrangements employed, low clearance times being of particular
importance at the higher fault levels.
3.1.4 Methods of fault calculation

The information normally required from a fault calculation is that which gives the
values of the currents and voltages at stated points in the power system when the
latter is subjected to a given fault condition, the fault location and system operating
60 Fault calculation

conditions being specified. Fault calculation is therefore essentially a matter of

network analysis and can be achieved by a number of alternative methods, namely:
(a) direct solution of the network equations obtained from the mesh-current or
nodal-voltage methods,
(b) solution by network reduction and back-substitution and,
(c) solution by simulation using a fault calculator or network analyser.
The choice of method will normally depend on the size and complexity of
network and on the amount of information required from the analysis, a further
important factor being the availability of suitable computing facilities. Direct
solution of the network equations is now commonly employed using suitable
digital-computer facilities and appropriate computer programs, such use of the
computer making it possible to study a wide range of system and fault conditions
speedily and economically, particularly in the case of the larger networks.
Solution by network reduction using manual (that is slide rule or desk-calculator
computation) is widely used for such problems as involve, or can be represented by,
a network of limited size and complexity, there being a large number of fault calcu-
lations which fall into this class. Its use in more complicated cases, however, is
limited only by the amount of time required to obtain a solution, the cost of such
time being a not altogether unimportant consideration. Solution by simulation
using a fault calculator or network analyser has the advantage of simplicity of
application due to the one-to-one correspondence between the real and the simu-
lated system, but although widely used in the past, prior to the advent of computer
methods, its use is now generally limited to smaller networks and to situations
where computer facilities may not be readily available.
An essential part of power system analysis and fault calculation is that which
concerns the determination of the equivalent system network for the system
operating conditions and fault conditions under consideration. As already seen, the
fault conditions to be analysed normally fall into one or other of two classes,
namely balanced or symmetrical fault conditions (for example the three-phase
short-circuit) and unbalanced or unsymmetrical fault conditions, the latter class
being normally analysed by the symmetrical-component method. As will be seen
later, both classes of fault are analysed by reducing the power system, with its
fault condition, to an equivalent single-phase network.
3.2 Basic principles of network analysis

3.2.1 Fundamental network laws

The great majority of fault calculations are concerned with the behaviour of the
power system under steady-state conditions or conditions which, from the point of
view of analysis, may be regarded as steady-state conditions. It can also usually be
assumed that all the power system currents and voltages vary sinusoidally with time
at a common constant frequency and can therefore be treated as vector quantities
and be expressed, together with the power system impedances and admittances, in
complex-number form.
 Fault calculation 61

The relationship between the currents, voltages and impedances in any linear
network is governed by the three basic network laws, namely Ohm's Law and the
two laws of Kirchhoff, a formal statement of these laws in terms of vector quantities
being given below.
Ohm's Law: Ohm's Law states that the vector voltage drop V produced by a
vector current I flowing through a complex impedance Z is given by the vector
equation
V = IZ [Link]
An alternative form is
I= VY [Link]

where Y is the reciprocal of Z and is the complex admittance. The law is illustrated
in Fig. 3.2.1A from which it will be noted that the sense of the voltage-drop Vis in
opposition to that of the current I.
Kirchhoff's First Law: Kirchhoff's First Law states that the vector sum of all the
currents entering any junction or node in a network is zero or, stated in equation
form
Z Ii = 0 [Link]
i

where I i is the vector current flowing into the node from branch i, the summation
extending over all the branches connected to the node. Outflowing currents with
respect to the node are simply treated as negative inflowing currents.
The law, also known as the Junction Law is illustrated in Fig. 3.2.1B.
Kirchhoff's Second Law." Kirchhoff's Second Law states that the vector sum of
all the driving voltages (that is source voltages) acting round any closed path or
mesh in a network is equal to the vector sum of the voltage drops in the impedances
of the component branches of the path. Thus, in equation form

L Ei = ~. llZl [Link]
i i

! z
,.~ z I ; o -
v
v =lZ
Fig. 3.2.1A Ohrn'$ Law

where Ei is the vector driving voltage in branch i, Ii the vector current in the
branch and Z i the complex impedance of the branch, the summation extending
over all the component branches of the path or mesh. The driving voltages and
currents must all be measured in the same direction round the path. Expressed in
another way, the law simply states that the vector sum of all the voltages (that is
driving voltages and voltage drops) acting round a closed path or mesh is zero.
62 Fault calculation

The law, illustrated in Fig. 3.2.1 C, is also known as the Mesh Law.
These three basic network laws applied to the branches, nodes and meshes of
any linear network (that is, a network in which the impedances are constant and
independent of the currents through them), enable the branch currents to be found
if the branch driving voltages and branch impedances are known.

Ih

Ia + Ib + Ic + Id : 0
Id

\
Fig. 3.2.1B Kirchhoff'$ First Law

~ I."2

j'/~
El
E3

Zs
Z3

E5

I
i
!

E l + E 2 + E 3 + E 4 + E 5 = I I Z I + 12Z 2 + 13Z 3 + 14Z 4 + 15Z $

Fig. 3.2.1C Kirchhoff'$ Second Law

As a simple example consider the elementary two-machine system shown in

Fig. 3.2.1D and comprising a generator and a synchronous motor, the former
represented by a driving voltage of 118 + j 2 4 V behind an impedance of 1 +]4 [2
and the latter by a driving voltage of 100 +/0 V behind an impedance of 2 +/'5 f2.
The current flowing from the generator to the motor is 3 - j 1 A, and this satisfies
Kirchhoff's Second Law because the total voltage drop in the path is
 Fault calculation 63

(3 - / 1X3 +/'9) v , that is 18 + j 24 V, this being equal to the total driving voltage
acting round the path, namely 118 ÷/24 V minus 100 + / 0 V.
118 + j 2 4 V 100 +j0V

3 -jlA

I11 + j I 3 V

N
Fig. 3.2.1D

Assume now that a short-circuit occurs at the generator terminals, this short-
circuit path having an impedance of 1 + ]0 I~. The conditions are then as shown in
Fig. 3.2.1E where 11 is the current from the generator, 12 the current from the
motor (now acting as a generator) and 13 the current through the impedance of the
short-circuit path.
It is required to determine 11,12 and 13, and hence applying Kirchhoff's First
Law we obtain the equation

11 + I 2 - 13 = 0 3.2,1.5

while application of the Second Law to each of the two meshes in turn gives the
equations
(1 +j4)11 + (1 +j0)I3 = 118 +j24 [Link]
(2 +j5)I2 + (1 +/0)13 = 110 +j0 [Link]

the first of these equations relating to the generator mesh and the second to the
motor mesh.

118 +j24V I00 +jOV

!3

I + jO~

Fig. 3.2.1E
64 Fault calculation

Eliminating 13 from eqns. [Link] and [Link] by use ofeqn. [Link], we obtain
the two equations

(2 +/4)11 +(I +/0)12 = 118 +/24 [Link]

(I +/0)11 +(3 +jSY2 = II0+/0 [Link]

these being a pair of simultaneous equations in the two unknowns 11 and 12.
Eliminating/2 from these equations so as to solve for 11 we obtain

[(3 +jSX2 +]'4)- (1 +/oX1 +/0)] I,

= [(3 +j5XI18 + j 2 4 ) - (1 + j 0 ) ( l l 0 + j 0 ) ]
which reduces to
(-15 +/22)I~ = 134 +/662
giving
I1 = 1 7 . 7 - / 1 8 . 1 A
The value of I1 can now be used in eqn. [Link] to obtain 12, thus
(100 + j 0 ) - (1 +/0X17-7 -i18.1)
I2=
3 +/5

82.3 +/18.1
3 +/5

giving
la = 9 . 9 - / 1 0 . 5 A
Finally, substituting the values Ofll and 12 in eqn. 32..1.5 gives
Is = 27.6 - / 2 8 . 6 A
The voltage at the point of fault, namely the product of I3 and Za, is seen to be

/:3 = (27.6 -/28.6X1 +/0)

giving
v3 = 27.6 -/28.6 V

This simple example has been solved by treating the network branch currents as the
unknowns in the network equations, and while this is obviously the most direct
approach, it nevertheless suffers from certain disadvantages if the number of meshes
in the network is at all large. These disadvantages stem from the fact that the
number of unknowns, namely the branch currents, will generally be larger than is
necessary for the solution of the problem and that the equations containing these
unknowns will not generally be amenable to a systematic method of solution. To
avoid these difficulties, network analysis is better carried out using mesh-current
analysis and nodal-voltage analysis. These methods are briefly described in the
following two Sections.
 Fault calculation 65

3.2.2 Mesh-current analysis

This method of analysis can be understood by considering the simple network

shown in Fig. 3.2.2A. It will be noted that each mesh of the network is assumed to
carry a circulating current, and it is these so-called mesh currents which are treated
as the unknowns in the problem, the current in any given branch being readily
obtained once the mesh currents associated with that branch have been determined.
Because the mesh currents associated with any given node flow through that node it
can easily be seen that they satisfy Kirchhoff's First Law. We can therefore proceed
to write down the equations which result from application of the Second Law,
there being one such equation for each mesh of the network.

ZC

Zaf Zb

Ea() ( ) Eb

Fig. 3.2.2A Mesh-current notation

Thus, remembering that the current in any branch is the vector sum (taking due
account of direction) of the mesh currents flowing in that branch, we obtain the
following equations for meshes 1, 2 and 3, respectively:

z J , + z a ( , q - I2) + z N , - I3) --~,

Za(I2 - 11) + ZcI2 + Ze(I2 - la) =-Ec [Link]
Z.f(13 - 11 ) + Ze (I3 - 12 ) + Zb 13 = Ec - El,

Rearranging these equations in a more systematic form, we obtain

Z~111 +Z1212 +Z1313 = E l ]
Z211~ + Z~212 + Z2ala = E2 [Link]
Za111 + Za~]2 + Z3ala = Ea
where
Z~I = Z , +Z,~ + Z r
Z22 = Zc + Za + Ze [Link]
Z33 = Za + Z~ + Z I
66 Fault calculation

ZI2 =Z21 ='-Zd /

Z23 = Z32 =--Ze [Link]
Z31 = Zl3 = -Z:

E 2 -- -Ee [Link]
Es - - E c - El,

It will be noted that Zl i is the sum of the impedances of the branches forming
mesh 1, Z22 the sum of those forming mesh 2, and Z33 the sum of those forming
mesh 3. It will also be noted that Z~2, equal to Z21, is equal to minus the branch
impedance common to meshes 1 and 2, that Z23, equal to Z32, is equal to minus
the branch impedance common to meshes 2 and 3, and Z3~, equal to Z13, is equal
to minus the branch impedance common to meshes 3 and 1. Finally it will be noted
that El is the sum of the driving voltages in the branches forming mesh 1, E2 the
sum of those in the branches forming mesh 2, and E3 the sum of those in the
branches forming mesh 3. It should be noted that the branch driving voltages are
treated as positive when they act in the same direction as the mesh current in the
mesh to which the equation refers, namely, in a clockwise direction, and are treated
as negative when they act in the opposite, anti-clockwise direction.
Eqns. [Link] are a set of three simultaneous equations in the three unknown
mesh currents l t , / 2 and/3 and it will be evident that the general case of a network
of n meshes will result in a similar set of n simultaneous equations in terms of the
n unknown mesh currents. Knowing the branch driving-voltages and impedances,
this set of equations can be virtually written down by inspection from the rules
already given. The solution of these equations to obtain the mesh currents can be
achieved by a number of different methods, each branch current being then
obtained from its component mesh currents.
3.2.3. Nodal-voltageanalysis

In this method of analysis one of the network nodes is chosen as the reference node
and the voltages of the remaining nodes, measured with respect to the reference
node, are treated as the unknowns in the problem. The voltage across any branch of
a given mesh is equal to the difference between the node voltages at the two ends of
the branch in question. It is seen, therefore, that the summation of the component
branch voltages round the mesh must be zero because the node voltages, whose
difference constitutes any given branch voltage, will be cancelled out in the summa-
tion by the contributions which the same node voltages, but reversed in sign, make
to the voltages across the adjacent branches of the mesh. The node voltages thus
satisfy Kirchhoff's Second Law and we can therefore proceed to apply the First
Law to each node of the network, in turn, with the single excepton of the reference
node.
Thus, for the simple network shown in Fig. 3.2.3A, we obtain the following
equations for nodes 1, 2 and 3, respectively:
 Fault calculation 67

1 Yc 2

Ya Yb

v 3f~3

Eb

) I ~ (
o

Fig, 3.2.3A Nodal-voltage notation

( O + E a - V,)Y. + ( V 2 - V , ) Y c + ( V 3 - V,)Ya =0 /
(0 + Eb - V2 ) Yt, + ( V, - V2) Ye + (V3 + Ec - V~) Ye = 0 [Link]
(0- V3)Yf+(V 1 - V3)Ya + ( V 2 - Ec- V3)Ye = 0

It will be noted that these equations are written in terms of the branch admittances
as compared with the branch impedances used in the mesh-current method of
analysis.
Rearranging the equations in a more systematic form, we obtain

Y l l V t + YIiV2 + YI3Va =11

Y21 gl + Y22 V2 + Y23 V3 =12 [Link]
Y31 VI + Y32V2 + YaaVa =Ia
where
YIl = Ya + Ye + Yd ]
Y22 = Yb + Ye + }re [Link]
Y33 = Yd + Ye + Yf

YI2 = Y21 = - Y¢ |
Y23 = Y32 = - Ye [Link]
Y31 = Y13 = - Yd

I1 = Ea Ya ]
I2 = Eb Yb + Ee ]re I [Link]
13 = -Ec }re
It will be noted that Yt i is the sum of the admittances of the branches connected
to node 1, Y22 the sum of those connected to node 2, and Y33 the sum of those
connected to node 3. It will also be noted that Y12, equal to Y2~, is equal to minus
68 Fault calculation

the admittance of the branch connecting nodes 1 and 2, that Y2a, equal to Yaz, is
equal to minus the admittance of the branch connecting nodes 2 and 3, and that
Fax, equal to Yla, is equal to minus the admittance of the branch connecting
nodes 3 and 1. Finally it will be noted that I~ is the sum of the products of driving
voltage and admittance for each of the branches connected to node 1,!z the sum
of these products for each of the branches connected to node 2, and la the sum of
these products for each of the branches connected to node 3, the branch driving
voltages being treated as positive when they act towards the node in question and
negative when they act in the opposite direction, namely away from the node.
Eqns. [Link] are a set of three simultaneous equations in the three unknown
node voltages VI, V2 and Va and it will be evident that the general case of a
network of n nodes, excluding the reference node, will result in a similar set of n
simultaneous equations in terms of the n unknown node voltages. Knowing the
branch driving voltages and admittances, this set of equations can be virtually
written down by inspection from the rules already given.
The set of simultaneous linear equations obtained is identical in general form
with that obtained using the mesh-current method of analysis and is amenable to
the same methods of solution. Having solved the equations to obtain the node
voltages, each branch current is then readily obtained from the node voltages at its
two ends, the branch current being given by the product of the branch admittance
and the voltage across this admittance. In obtaining this latter mentioned voltage
from the node voltages, due account must, of course, be taken of the driving
voltage, if any, in the branch in question.

3.2.4 Application of mesh-current and nodal-voltage analysis

The mesh-current and nodal-voltage methods of analysis have both been shown to
result in a set of simultaneous linear equations, the solution of which leads to a
complete determination of all the network branch currents. As already stated, the
rules for the formulation of the equations are such that, with either of the two
methods, the resulting set of equations can be virtuaUy written down by inspection
once the network meshes or nodes have been numbered.
The solution of a set of simultaneous linear equations is a standard computa-
tional procedure and may be achieved by a number of methods including elimina-
tion, determinant and matrix methods and iterative methods. Details of these
methods will be found in a number of the references given in the bibliography. With
all methods of solution, however, the use of manual computation, even with the
aid of a desk computer, is not normally a practical proposition if the number of
equations is at all large, since the computation time tends to be considerable and
increases rapidly with the number of equations. Solution by high-speed digital
computer is however, very much a practical proposition, such computers being
capable of handling very large sets of equations (that is, simultaneous equations
representing a large number of meshes or nodes) and of providing a rapid and
accurate solution.
 Fault calculation 69

The choice between the mesh-current and nodal-voltage methods will normally
be determined by the need to reduce computing time to a minimum. It is of
interest to note, therefore, that the great majority of networks have fewer nodes
than meshes and are therefore more suited to solution by the nodal-voltage method.
This is readily demonstrated by the simple two-machine network of Fig. 3.2.1E
which, it will be noted, has two meshes but only one node, apart from the reference
node. Thus, applying the mesh-current method we obtain the two equations

(2 + j 4 ) / , - (1 +/0)/2 = 118 + / 2 4 /
=(I +]O)/l +(3 +/5)/2 = - I00-]0 I [Link]

where I~ and/2 are the left-hand and right-hand mesh currents, respectively and are
assumed to act in a clockwise direction.
Applying the nodal-voltage method, on the other hand, gives the single equation
[ 1 1 ,
l ] +j24 100+/0 l
l+j4 1+/0 2+/5 +/4 2
where Vx is the voltage at the point of short-circuit, and is denoted by V3 in
Section 3.2.1.
It will be left to the reader to verify that Equation [Link] and Equation [Link]
both give the same results, these having already been obtained in Section 3.2.1.

3.2.5. Network theorems and reduction formulas

The mesh-current and nodal-voltage methods of analysis are both a means of ob-
taining the complete analytical solution to any linear network problem. They are
being used to an increasing extent, their use being limited only by the availability
of suitable computing facilities and the fact that an analytical solution can also be
achieved by other means, which in many cases may be simpler and more direct.
This alternative means of analysis, known as the network-reduction method,
depends on the reduction of the network to a simpler equivalent form, the method
being particularly useful where the solution is required to give the values of only a
limited number of branch currents.
Network reduction is the process of combining network branches so as to reduce
the given network to an equivalent network with fewer branches, this equivalent
network being amenable to direct and simple solution. Thus, a complete and
complex network, as viewed from any given pair of nodes, may be reduced in this
way to its simplest equivalent form, namely a single equivalent branch. The network
theorems and reduction formulae most commonly required are described below.

Combination o f series branches: Taking first the general case of any number of
branches connected in series, any given branch i comprising a driving voltage E t and
series impedance Zl, the equivalent single branch comprises a driving voltage E r in
70 Fault calculation

series with an impedance Zr where

Er= ~. EI= Ex + E2 + . . . + En [Link]
i

and
Zr = ~. Zt =ZI +Z2 + . . . +Zn [Link]
i

and n is the number of branches. The driving voltages are all measured in the same
direction with respect to the end nodes of the series combination. These rules,
applied to the combination of three series-connected branches, are illustrated in
Fig. 3.2.5A.

Z Z ~ Z3
p o q

_~ p~ Zr")---oq
Fig. 3.2.5A Combination o f series branches

Combination o f parallel branches: Taking, again, the general case of any number
of branches but now connected in parallel, any given branch i comprising a driving
voltage E i and series admittance Yi (that is series impedance l[Yi), the equivalent
single branch comprises a driving voltage Er in series with an admittance Yr (that is,
an impedance 1/Yr), where

Er=(l/Yr) ~ EiYt=(I[Yr)(EIY~ +E~Y2 + . . . +EnYn) [Link]

i
and

Yr = Z Yi = Y1 + Y2 + . . . + Yn [Link]
i

and n is the number of branches. The driving voltages are all measured in the same
direction with respect to the common nodes of the parallel combination. These
rules, applied to the combination of three parallel-connected branches, are illustrated
in Fig. 3.2.5B.
Star-to-delta transformation: The star-to-delta transformation permits any set of
three star-connected branches with isolated star-point to be replaced by an equiva-
lent set of three delta-connected branches. Thus, let a, b and c denote the three
terminals of the star and its equivalent delta, and let the star-connected branches
comprise driving voltages Ea, E~ and E c and impedances Z a, Z b and Z c respect-
ively, the voltages being measured in the direction away from the star point. Then
denoting the driving voltages and impedances of the equivalent delta-connected
 Fault calculation 71

Yr
P o <) q "---p I I,~oq

y3l_. ~

Fig. 3.2.5B Combination o f parallel branches

branches by Eab , Et, c and Eta and Zab, Z b c and Zca respectively, the latter values
are given in terms of the former by the equations

Eab = E a - E b + IZab
Ebc = .Eb - E c + .[Zbc [Link]
E c a = E e - E a + 1Zea

Za

co /

Fig. 3.2.5C Star and delta circuits

and ZaZb + ZoZc + ZcZa

Zat ~ -
Zc

Z a Z b -I- Z b Z c ..I-Z c Z a
Zbc - [Link]
Za

ZaZb + ZbZc + ZcZa

Z ea -~-
Zb
72 Fault calculation

where I, equal to (Eab + Ebc + Eca)/(Zab + Zbc + Zca), is any arbitrarily chosen
current assumed to be circulating in an anticlockwise direction through the
branches of the delta. It is normally convenient to assume this current to be zero.
The driving voltages Eab, Ebe and Eta act from b to a, c to b and a to c, respect-
ively, in the branches concerned. These rules are illustrated in Fig. 3.2.5C.
Delta-to-star transformation: The delta-to-star transformation permits any set of
three delta-connected branches to be replaced by an equivalent set of three star-
connected branches with isolated star-point, the relationships between the star and
delta quantities, using the nomenclature of the previous section, being given by the
equations

E a - E b = Eab - [Zab
E# - E c = Ebc - 1Zoc [Link]
E c - E a = Eta - 1Zca
and
Zat~Zca
Za =
Zab + Zbc + Zca

Z bc~ab
Z# = [Link]
Zab + Zb~ + Z ~

Z caZ bc
Zc =
Zab + Zbc + Z ~

where I, equal to (Eab + Ebc + Eca)/(Zab + Zbc + Zca), is the current circulating in
an anticlockwise direction through the branches of the delta.
It should be noted that any one of the three driving voltages E a, E b or E c can be
chosen quite arbitrarily, this value being then used to determine the remaining two
driving voltages. It is normally convenient to arrange that the sum of the three
driving voltages is equal to zero, in which case

E o -- '/3 [E~b - E , ~ - I ( Z , b - Z~)]

E b = t/3[Ebc - E a b - I ( Z b c - Z,,t,)l [Link]
E , = '/~ [ E ~ - E b , - S ( Z , ~ - Z b c ) ]

where I, as already stated, is given by

~a~ +Eb~ +E,~
I= [Link]
Zab + Zbc + Zca

For the special case in which Eab + Ebc + Eta = 0, giving I = 0, the star driving
voltages then become
 Fault calculation 73

Ea = ]/3(Eat~- Eca)
Eb = t/3(Eric- Eat,) [Link]
Ec = R/3(Eca - Ebe)

I * iO~Z --K ~ i3V ~ / ¢ ~ 6--.i2.2~Z

2- lOW

Fig. 3.2.5D Example o f equivalent star and delta circuits

The transformation rules are illustrated by the circuit shown in Fig. 3.2.5D.
Combination o f equal driving-voltages: If two or more branches have a common
node and identical driving-voltages with respect to this node, then the individual
branch driving-voltages can all be removed and replaced by a single external driving
q

/a

l~
'J q P

"~'~% r f

Fig. [Link] Combination o f equal driving voltages

voltage connected to the common node, the latter driving voltage being equal to the
common value of the original branch driving-voltages. The theorem is illustrated in
Fig. 3.2.5E and follows readily from the fact that equipotential points can be
joined together without any resultant change in the electrical state of the network.
The theorem is particularly useful in power system analysis in the reduction of a
multiple-source network to its single-source equivalent.
Superposition Theorem:The Superposition Theorem, applicable to any linear
network, states that the current which flows in any branch of a network as a result
of the simultaneous action of several driving voltages is equal to the vector sum of
74 Fault calculation

118 ~, i 2 4 V

© I ~" j4,~

15.6- j2 1.2A
I'
A ~

N 1.7
2 + j5~Z
. ]

- j4.2A

~ 17.3 jl7.0A

I + jO~

I00 + jOV

I "~ j4,~:'

--2.1 - j3.1A
I'
-- ~
2 + jSl2
"-

8.2 - j 1 4 . 7 A
)
@
10.3 -.il 1.6A

I + ,iOSZ

118 ~ i 2 4 V I00 +jOV

17.7 ilt~.lA 9 . 9 -- j l O . S A
~"-- 27.6 j28.6A

I + jO~

N
Fig. 3 . 2 . 5 F Illustration o f the superposition theorem

the currents which would flow in the branch in question with each driving voltage
acting individually and all the remaining driving voltages equal to zero, that is,
short-circuited. The theorem is illustrated in Fig. 3.2.5F using the simple two-
machine problem of Section 3.2.1.
Thevenin's Theorem: Thevenin's Theorem, similarly applicable to any linear
network, states that any such network containing driving voltages, as viewed from
any two terminals, can be replaced by a single driving voltage acting in series with a
single impedance. The value of this driving voltage is equal to the open-circuit
voltage between the two terminals, and the series impedance is the impedance of
the network as viewed from the two terminals with all the driving voltages equal to
zero, that is short-circuited. The theorem is illustrated in Fig. 3.2.5G and its appli-
cation to the simple two-machine circuit of Section 3.2.1 is shown in Fig. 3.2.5H.
 Fault calculation 75

Active network

,,0
q
TVoc
P
o
Same network with
a l l source voltages
removed and replaced
by s h o r t circuits
0
q

n°- t

i P
I ! O

V°c i
0
I
~.j q

Fig. 3.2.5G Thevenin's Theorem

!
II1 +jl3V I +j412 J
I
Ij P
J! 0

;
!
2 + jSs2
I
' I + j012
I
i
ii o
l N

Fig. 3.2.5H Thevenin's Theorem applied to the example o f Fig. 3.2.1E

Norton's Theorem: Norton's Theorem is the dual of Thevenin's Theorem and

states that any linear network containing driving voltages, as viewed from any two
terminals, can be replaced by a single driving current shunted by an impedance. The
value of this driving current is equal to the short-circuit current which will flow
between the two terminals when connected together, and the shunt impedance is
the impedance of the network as viewed from the two terminals with all the driving
voltages equal to zero, that is short-circuited. The theorem is illustrated in Fig.
76 Fault calculation

Active n e t w o r k lsc

P
0
Same n e t w o r k w i t h
all source voltages
•4 - ' - " Z
removed and replaced
by short circuits
~,~
q

ri I p
I ! o

0
i
i
lsc
i z i
i o

Fig. 3.2.51 Norton's Theorem

3.2.5I and its application to the simple two-machine circuit of Section 3.2.1 is
shown in Fig. 3.2.5J. It is important to note that the driving current is constant in
value irrespective of the voltage across it, just as a given driving voltage is constant
in value irrespective of the current through it. A constant driving current can be

F"
[ 19.S - j43.6A [
i "" i
i -@ i P

i
i i
I + j4F~ 2 + j51"I i I + j0D,
i
i i
iL.. i
.;
o
N
Fig. 3.2.5,) Norton's Theorem applied to the example o f Fig. 3.2. I E

regarded in physical terms, as the current produced by an infinite voltage acting in

series with an infinite impedance, the ratio of the voltage to the impedance being
the required driving current.
 Fault calculation 77

A number of other useful theorems and reduction formulae will be found in the
references given in the bibliography.

33 Calculation of balanced fault conditions

3.3.1 Single-phase representation

The component items of electrical plant which together form a three-phase power
system, namely generators, transformers, overhead-line and cable circuits, etc., can
all be regarded, for most practical purposes, as having electrical characteristics
which are balanced or symmetrical with respect to the three phases. Thus, the plant
impedance characteristics are usually such that the phase self-impedance values can
be regarded as the same for all three phases and the phase-to-phase mutual-
impedance values regarded as symmetrical with respect to the three phases. Simi-
larly, the driving voltages produced by the generators and any other synchronous
machines are balanced in that the three phase-to-neutral e.m.f.s produced by any
given machine are all equal in magnitude and symmetrically spaced, at 120 ° intervals,
in time-phase, the phase order beingthe same for all the machines. This phase order,
namely the order in which the instantaneous phase voltages attain their maximum
values in the cycle, is termed the positive-sequence phase-order. It readily follows
from this system phase-symmetry that conditions which are themselves balanced
with respect to the three phases (for example, balanced4oad and three-phase short-
circuit conditions) will preserve this phase symmetry in that the resulting phase
currents and similarly, the resulting phase voltages at any point in the system will
also be balanced with respect to the three phases. Thus, for the conditions referred
to, the three phase currents and, similarly, the three phase-to-neutral voltages at
any given point in the system will be equal in magnitude and equally spaced at
120 ° intervals in time-phase, the phase order being the same as that of the generated
phase-to-neutral voltages, namely the positive-sequence phase-order. These currents
and voltages are termed positive-sequence currents and voltages and are represented
in vector-diagram form in the manner shown in Fig. 3.3.1A.
The mathematical relationships between these balanced (positive-sequence)
phase values at any given point in the power system are the same for both current
.and voltage, namely

i~ = cao
and [Link]
Vb =a~Va
v~ =aVo

where the subscripts a, b and c denote the three phases in positive-sequence phase-
order and the symbol a (not to be confused with the phase reference) is a constant
78 Fault calculation

given by
1 .V3
a. . . . 1 [120 ° [Link]
2+/"--2 "-

and in known as the 120°operator.

IC
Vc

1h

~ h•V a
Fig. 3.3.1A Phasor-diagrarn representation o f positive-sequence voltages and currents

It will be noted that multiplication of any given vector by a gives a resultant

vector equal in magnitude to the given vector but advanced in phase (that is rotated
in an anticlockwise direction) by 120 ° with respect to the given vector. Similarly
multiplication by a a give a phase-advance of 240 °, because
( 1 ~_.3)( 1 . ~ / 3 ) i x/3
32 = 2"+1 - 7 +I 7 = - ' 2 - j 2 = 1[240° [Link]
J

Further multiplication gives

aS =~ - 7 +'i "2-; 2-1 ~ ' - = 1/360° = 1[0° [Link]

and so on. It should be noted that

a = 1 [120 ° = 1 / - 2 4 0 °
a 2 = 1 [ 2 4 0 ° = l [ - 120 ° [Link]
1 +a +a 2 =0

bearing in mind that a vector rotation forward through any given angle is the same
as a rotation backwards through an angle equal to 360* minus the given angle.
The balanced conditions referred to are illustrated by Fig. 3.3.1B which shows
an 11 kV generator with an impedance of 0 +]1 I2 per phase supplying a balanced
load of 20 + ]5 ~2 per phase through a transmission line of 2 +/3 I2 per phase, the
generator neutral point being earthed, as shown, and the transmission line having a
three-phase fault clear of earth at 0ae end F remote from the generator. Denoting
 Fault calculation 79

0 I jl$~ 2 ~ j3~l 20 ~ j 5 5 l

--0 L....-J I ]
; ")

@
,--4m,,
0 + illl 2 ~ j3~l 20 4 i55~

I ; C ; ]

© I i (;
2 + .i3$Z 20 ~ j5-~l

Ic Ec

~ c ~ Ea
z,-""V ~I

Vb

Ia

Fig. 3 . 3 . 1 B Illustration of a balanced three-phase fault condition

the three phases, in positive-sequence phase order, by the subscripts a, b and c, the
generator phase-to-neutral driving-voltages are given by

Ea = 6350 + / 0 = 6350/0 ° V
E b = - 3175 - / 5 5 0 0 = 6350/- 120 ° V
E c = - 3175 + / 5 5 0 0 = 6 3 5 0 / 1 2 0 ° V

the magnitude of the phase to neutral voltage namely 6350 V, being the line-to-line
voltage o f 11 000 V divided by ~/3. It will be noted that e.m.f. E a has been
arbitrarily chosen to have zero phase-angle.
Using the rule given in Section 3.2.5 for the combination o f parallel branches it
will be seen that the voltage to earth, at F, o f the three short-circuited phases is
80 Fault calculation

given by

IrF=-
2+i4[E. +eb o
3 2+14 .j=
because the three phase driving-voltages sum to zero and neutral current being
absent, the neutral point N and the earth are equipotential points. The impedance
voltage-drop in each phase must therefore be equal to the phase-to-neutral driving-
voltage in the phase in question and, hence, the three phase currents in the
generator and transmission line are given by

6350 +10
Ia = = 635 - 11270 = 1420/-63.4°A
2+14
-3175 -/5500 = - 1418 +i85 = 1420[176.6 ° A
/'b =
2+/4
-3175 +/5500
Ic = = 783 +11185 = 1420[56"6 ° A
2+/4

The phase-to-earth (and phase-to-neutral) voltages at the generator terminals G are

equal to the phase voltage drops in the transmission line and are therefore given by

Ira = (635 - / 1 2 7 0 ) (2 +i3) = 5080 - / 6 3 5

= 5119/-7.1"v
v b = (- 1418 +i85) (2 +/'3) = - 3091 - 14084 = 5119[ - 127.1°v
v c = (783 +11185) (2 +13) = - 1989 +/4179 = 5119[112.9°v

The three-phase short-circuit at F has been assumed to be clear of earth in this

example but it will be evident that, because V F has been shown to be zero, a three-
phase-to-earth fault at F would result in zero current in the earth connection and
would thus give precisely the same results as those already obtained for the fault
clear of earth. The three-phase vector diagrams of the currents and voltages are
shown in Fig. 3.3.1B.
Now, it will be evident from Eqns. [Link] that a knowledge of the phase
currents and voltages relating to any one chosen reference phase will enable the
corresponding values of current and voltage in the other two phases to be deter-
mined. Thus, if phaso-a is chosen as the reference phase then the phase-b values of
current and voltage are simply the corresponding phase-a values multiplied by a 2
and the phase-c values the corresponding phase-a values multiplied by a, where a is
the 120 ° operator. For the purposes of analysis, therefore, the power system of
Fig, 3.3.1B can be represented by the single-phase network shown in Fig. 3.3.1C,
this single-phase network and the associated single-phase vector diagram both
relating to the chosen reference phase, namely phase a.
The [Link] representation, although demonstrated for the extremely simple
system of Fig. 3.3.1B, is equally applicable to any balanced three-phase network
 Fault calculation 81

operating under balanced conditions, for example balanced4oad or three-phase

short-circuit conditions. The single-phase network obtained is termed the system
positive-sequence network, and the impedances represented in the network are the
impedances appropriate to positive sequence conditions, termed the positive-
sequence impedances. It will be noted from Fig. 3.3.1C that the three-phase short-
circuit is represented in the positive-sequence network by the short-circuiting
connection from the point of fault F to the zero-potential (neutral) bar represented
by the thick horizontal fine. With this short-circuiting connection absent, the
conditions represented are those corresponding to the balanced-load condition
existing before the application of the three-phase short circuit.

_•• O ~.il~
f ;
G 2 ~ i3fl
| I
F 21) ~ j S ~
r-'-I

Fig. 3.3.1C Single-phase representation of Fig. 3.3.1B

3.3.2 Use of a c o m m o n voltage base

The simple example of the previous Section was solved by direct analysis of the
given network using actual values of plant impedance together with actual values of
the power-system currents and voltages. This direct application of the network laws
was possible because the network interconnections were direct (that is, not through
transformers) or, put in another way, because the whole system had the same
nominal rated voltage for all its component items of plant. The power systems
which have to be considered in practice, however, almost invariably consist of a
number of sections operating at different nominal rated voltages, the interconnec-
tions between sections being by means of transformers. The analysis of such
multiple-section systems is achieved by representing the given network by an
equivalent network whose several sections all have the same nominal rated voltage.
This procedure allows the transformer interconnections to be replaced by equiva-
lent direct interconneetions, the resulting network being then amenable to direct
solution using the normal methods of network analysis.
82 Fault calculation

The derivation of this so-called common-voltage equivalent network can be

readily understood by considering the two network sections, P and Q, shown in
Fig. 3.3.2A, the two sections being interconnected by an ideal transformer of turns
ratio M (= Np/Nq) where
nominal rated voltage of network-section P
M= [Link]
nominal rated voltage of network-section Q
The two network sections are each represented, for simplicity, by a single branch
consisting of a driving voltage and an impedance in series, their values being Ep and
Zp for network-section P and Eq and Zq for network-section Q. The ideal-
transformer terminal-voltages and currents are similarly denoted by Vp and lp and
Vq and 1q, where It, is the current flowing out of network-section P and lq that
flowing into network-section Q. Now, because the transformer is ideal (that is has
windings of zero [Link] and takes zero exciting current), it follows that
Vp = MVq = V; say I
= Iq/M -
- I~' say I [Link]
and it is easily seen that these same equations are satisfied by the equivalent system-
network of Fig. 3.3.2B in which the ideal transformer is represented by a direct
zero-impedance interconnection. This equivalent network has the same nominal
rated voltage throughout, namely that of network-section P, and consists of the
given network-section P and the equivalent or referred network-section Q'. It will
be noted that the referred network-section Q' contains the referred values

E~ =MEq
=
3. ~:2.3
lq'= Iq/M
Zq = M 2Zq
P
F . . . . . . . . . "7
i

Ip M: I Iq Zq ~ [

i
!
'; tl i
VpI
_1
Vq
i
!

J
Fig. 3.3.2A Simplified representation o f two networks with ideal-transformer interconnection

in place of the actual network values, the expression for the referred impedance
following from the fact that the voltages in the referred network are all M times the
corresponding voltages in the actual network. Thus, the voltage-drop across Zq in
the actual network is lqZq and the corresponding voltage-drop in the referred net-
 Fault calculation 83

Q,

f 1 r
i /p Ii, [ i Itt/M M2Zq
I
i r--1 ; ~ I i v f i i

i li ' I MVq
i
i
! o
i i 0
k.. i i
Fig. 3.3.2B Common-voltage equivalent of Fig. [Link]

work is IqZq, so that

MIqZq
zi, = = M~ Zq

Now, eqns. [Link] will apply to all the voltages, currents and impedances in
network-section Q when the latter is considered as a general network rather than as
a single branch. For this general case, therefore, the voltages, currents and
impedances given in eqns. [Link] can now be regarded as general network values,
the equations given representing the relationships between actual values and their
referred equivalents. This general case is represented in block-diagram form in
Fig. 3.3.2C.
I' QI
r i F ]
i i i o ,, ,~o i
i % '" °i I < i
iI iI iI i
i i i i
i Vp Zil
o I
i,o v ,t I , "t, I ,
I
i J t J
Fig. 3.3.2C Generalisedrepresentation of Fig. 3.3.28

Consider, now, an ideal-transformer interconnection between network-section Q

and a third network-section S, the turns ration N (-- Nq/Ns) of this transformer
being given by
nominal rated voltage of network section Q
N= [Link]
nominal rated voltage of network section S

By similar reasoning to that already described, the two network sections Q and S
can be represented by an equivalent network consisting of the given network-
section Q and an equivalent or referred network-section S', the referred network-
84 Fault calculation

section S' containing the referred values

E; = NEs
v; =NG
[Link]
z; =N~Z~

in place of the actual network values. The ideal-transformer interconnection

between network-sections Q and S is now represented as previously by a direct
zero-impedance interconnection in the equivalent system-network.
The equivalent system-network is shown in block-diagram form in Fig. 3.3.2D.
Suppose, now, that all three network sections are interconnected through the
given ideal transformers. Then it will be evident that the equivalent network con-
sisting of Q and S' can itself be replaced by another equivalent network consisting
of Q' and S", this latter equivalent network being directly connected through zero-
impedance to the network-section P. Now the values in the network-section Q' are
(.) S~
r I r 1
I' I I, i
i i I
j I:,t I~l o . [ o ~ I~I e
I
! i i i
I ; i i
s
'
I i i v~ '~I ;
i
I v ,. i i o
! o ,
-._ J t ,,
Fig. 3 . 3 . 2 0
those already given by eqns. [Link] while those in S", are given by applying these
same equations to the values in S'. Hence, it follows that

E~' = [Link],
G' =M~VG
q' = ;,/Mlv [Link]
z;' = (.~v)~z~
Qi S"
| r I
i i if l
l'.p Ip
C , i~ oi i~ i
i ':, ':, i i ': '" i
io io io iI io
Vp /p I ~ vlq ,'~
' ! I
o
v," "~
' I
o
i i I !~ I
j L° °j t i
Fig. 3.3.2E
 Fault calculation 85

the resulting equivalent network for the full three-section system being represented
in block-diagram form by Fig. 3.3.2E.
From eqns. [Link] and [Link] it is seen that
nominal rated voltage of network section P
MN =
nominal rated voltage of network section S

and it thus follows that the rules for referring any given network section to a given
common nominal rated voltage, termed the common voltage base, are

E' = KE
V'=KV
I' = I / K [Link]
Z' = I ~ Z
where
common voltage base
K= [Link]
nominal rated voltage of given network section

and the primed and unprimed symbols represent referred and actual values
respectively. It will be noted that these equations are simply a more general state-
ment of eqns. [Link] and [Link].
Although these conversion equations have been derived by considering network-
sections interconnected as a single point it is not difficult to see that they are equally
valid for any number of interconnection points.
The process of network analysis by the use of a common voltage base can now
be seen to consist of three stages:
(a) the conversion of the given system network to an equivalent network having
a single common nominal rated voltage, namely the chosen base voltage, for
all its sections
(b) the solution of this equivalent network in terms of its common-base values of
voltage, current and impedance
(c) the reconversion of the equivalent network, with its now determined
common-base values of current and voltage, to its original actual form con-
taining actual values of current, voltage and impedance.
Eqns. [Link], used to determine the referred values of driving voltage and
impedance for the equivalent system network, are, of course, equally valid for the
final stage of reconversion from referred values to actual values. These equations,
rewritten in the form appropriate to this reconversion, are

E= Et/K
v = V'/K
I =KI' [Link]
z = z'/x~
86 Fault calculation

where the primed symbols represent common-base values and the unprimed
symbols actual values. The symbol K is the ratio of common base voltage to actual
nominal rated voltage, as already defined.
The use of the common voltage base is illustrated by the simple example
shown in Fig. 3.3.2F.
It is important to note that the use of the common voltage base, as so far des-
cribed, has assumed ideal-transformer interconnections between the different
network sections, the voltage-transformation ratios of these transformers being
11/33 k V 3316.6 k V
3 + j6S i'

I -i j i l l

4 ~ jS.~L /.-..~..~ [
12 kV
phase-to-phase
I l / 6 . 6 kV

0.12 + 0,24Sg
4160 +.i0v
o ~ o
I~ I + jlS~

1.44 + j l , 8 0 , ~
! ; o

4 1 6 0 + jOY
0.11 + .iO.21~ + jl.~

j0.72 S.~

4 160 + jOY
[Link] , j1.'43S~

~31 j1620A

Fig. 3 . 3 . 2 F Application o f the common voltage base for a system with ideal-transformer
interconnections (common base voltage 6.6 k V)
 Fault calculation 87

identical to the ratios of the nominal rated voltages of their associated network
sections. As already shown, such ideal-transformer intercormections are represented
in the common-voltage equivalent circuit by direct intereonnections of zero
impedance. The representation necessary ha the case of practical transformers is
discussed in the following section.

3.3.3 Representation of nominal-ratio transformer circuits

Practical transformer circuits differ from the ideal transformers of the previous
section in that they have non-hat'mite values of exciting impedance and non-zero
values of leakage impedance. From the point of view of their representation in a
common voltage equivalent circuit they may ha addition, have transformation ratios
which differ from the nominal ratios so far considered.
Two-winding transformers: The common-voltage equivalent circuit of two-
winding transformer can be derived by considering the single-phase transformer
shown ha Fig. 3.3.3A, where H and L denote the high-voltage and low-voltage
windings, respectively. The transformer is assumed to have a high-voltage-to-low-
voltage turns-ratio N (= NH/NL) equal to the ratio of the nominal rated voltages of
the associated network sections. As shown in the figure, the given transformer can
be represented completely by an ideal transformer of this same turns-ratio N
together with the lumped impedances ZH, ZL and Ze. The impedances Zt4 and ZL
are the leakage impedances of the windings H and L, respectively, and Ze the
exciting impedance, Ze + Ztt being the impedance which would be measured by an
open-circuit test on the high-voltage winding.

N:I

I"-
"" i"
I,,,,.

H ~ L

.< ,.

IH ZH N : I ZL ]L
c "; " :- t ;'; o

Fig. 3.3.3A
VH

T I' IIITr
C .....
Ze EH

Equivalent circuit o f a two-winding transformer

,
EL

O
VL
88 Fault calculation

Now, the voltage E L across the low-voltage winding of the ideal transformer is
given by
EL = Vz, + I L Z L [Link]
where VL is the voltage and/i, the outflowing current at the terminal of winding L.
The voltage across the high-voltage winding of the ideal transformer is therefore
EH = NEL = IV(VL + IL Z L ) [Link]
which can be rearranged to read
EH = NVL + (IL/N)N2 ZL [Link]
It will be noted that NVL and I L / N are the equivalent values of V'L and IL when
referred to the nominal rated voltage of the [Link] winding, so that
E H = V~L + fLZ~ [Link]
where Z~ (= N:ZL)is the referred value ofZL.
From Eqn. [Link] it will be evident that the equivalent T circuit of Fig. 3.3.3B
is the required common-voltage equivalent circuit of the transformer referred to
the nominal rated voltage of the high-voltage winding H.
The exciting impedance Ze can be ignored in the great majority of calculations
because the exciting current le can normally be considered to be negligibly small.
iH ZH z~ (=N2Z L ) =[(=~L/N)
c .~ ; ; -

V'L(=NV L)
VHI Ze
EH(=E L)

!H ZHL(=ZH+N2ZL) I~
o > r-'-I > O

Fig. 3.3,3B Common-base equivalent circuit~ o f a two-winding transformer

In such cases, therefore the common-voltage equivalent circuit reduces to the

single series impedance
ZH + NaZL = ZHL , say [Link]
shown in Fig. 3.3.3B. It will be noted that ZHL is the short-circuit impedance of
the transformer as measured from the high-voltage side with the low-voltage
winding short-circuited. The equivalent impedance of the transformer referred to
the nominal rated voltage of the low-voltage winding, namely ZLH is readily
 Fault calculation 89

obtained from the fact that

ZHL = ~ ZLH
~Ving
ZLH = Z L + (I/N)~.ZH [Link]
The impedance ZLH is,of course, the short-circuitimpedance of the transformer as
measured from the low-voltage side with the high-voltage winding short-circuited.
Auto-transformers:The auto-transformer differs from the [Link] trans-
former in that a single tapped winding is used in place of two separate windings, the
arrangement being as shown in Fig. 3.3.3C. The fullwinding, which constitutes the
[Link] winding, is assumed to have 2VH turns and the common portion, which
constitutes the low-voltage winding, is assumed to have NL turns. As before, the
transformer is assumed to have a high-voltage-to-low-voltage turns-ratio
N (= N H / N L ) equal to the ratio of the nominal rated voltages of the associated
network sections. As shown in the figure, the given transformer can be represented
completely by an ideal auto-transformer of this same turns-ratio N together with
N:I

IH ZS

VH
zelItL 1L
f

VL

ZC .,I
0 0
Fig. 3.3.3C Equivalent circuit o f an auto-transformer

lumped impedances Zs, Z e and Ze. The impedance Z s is the leakage impedance of
that portion of the winding which lies between the high-voltage terminal and the
tapping point, the series portion, this portion having NH - Nt. turns. Similarly,
Zc is the leakage impedance of the remaining portion of the winding, namely that
between the tapping point and the common terminal. The impedance Z¢ is the
90 Fault calculation

exciting impedance, Zs + Ze + Ze being the impedance which would be measured

by an open-circuit test on the high-voltage winding. The tapping point, it will be
noted, constitutes the low-voltage terminal of the transformer.
Now, the voltage E L across the low-voltage (common) winding of the ideal
transformer is given by
Ez, = VL + (1L - 1H)Zc [Link]
where VL is the voltage and IL the outflowing current at the low-voltage terminal
of the winding L and 1H the inflowing current at the high-voltage terminal of
winding H. The terminal voltage VH of the high-voltage winding H is given by
VH = VL + ( N - 1)EL +IHZs [Link]
Substituting for EL in this latter equation usin~ eqn. [Link] gives
VH =NVL +IH[Zs- ( N - 1)Zc] +IL ( N - 1)Ze [Link]
which, rearranged, gives
VH =NVL +IH[Zs- ( N - 1)Zc] + (IL/N)N(N- 1)Ze [Link]
It will be noted, as before, that NVL and IL/N are the equivalent values of VL
and I£ when referred to the nominal rated voltage of the high-voltage winding,
so that
VH = IfL +IH[Z s - ( N - 1)Zc] + ~ N ( N - 1)Zc [Link]
Now the open-circut impedance of the auto-transformer as measured from the
high-voltage side is
Zs + Ze + Zc [Link]
and that measured from the low-voltage side, but referred to the high-voltage side,
is
N 2 Ze + Ze [Link]
IH ZS-(N-Z)Z c N(N-I)Z c l~
q ; i ; ; >

Ze + NZ c

IH ZHL [= Zs+(N- I)2Zc] I~.

c > I--"l >

VH

Fig. 3.3.3D Common-base equivalent circuits o f an auto-transformer

 Fault calculation 91

The required common-voltage equivalent-circuit for the auto-transformer can now

be obtained using eqns. [Link], [Link] and [Link] and is as shown in
Fig. 3.3.3D, this equivalent circuit being that referred to the nominal rated voltage
of the high-voltage winding.
If the magnetising impedance Ze can be ignored, as is usually the case, the
common-voltage equivalent circuit reduces to the single series impedance

Zs + (N- 1)2Ze = ZHL [Link]

shown in Fig. 3.3.3D. It will be noted that ZHL is the short-circuit impedance of
the transformer as measured from the high-voltage side with the low-voltago wind-
ing short-circuited. The equivalent impedance of the transformer referred to the
low-voltage side is obtained by dividing the above value by Na , as before, the result
being

ZLH = (1]N)2 Zs + (N- 1)2Zc/N 2 3.3.3,15

As before, the impedance ZLH is the short-circuit impedance of the transformer as

measured from the low-voltage side with the [Link] winding short-circuited.
Three-winding transformers: The common-voltage equivalent-circuit of a single.
phase three-winding transformer can be obtained by similar reasoning to that
already employed, the result being the equivalent-star circuit shown in Fig. 3,3.3E,
It is assumed, as before, that the turns ratio of each pair of windings is equal to the
ratio of the nominal rated voltages of the associated network sections. The letters
H, L and T denote the high-voltage, low-voltage and tertiary-voltage windings,
respectively, and the component branch impedances Z H, Z'L and Z'T are all assumed
to be values referred to the high voltage side. (Note that primed symbols denote
referred values.) These component branch impedances are obtained by similar

o
T

z,L
I ! | ~ U
ZH
H -~ ; j I
]
z'T T'
Fig. 3.3.3E Common-base equivalent circuit o f a Chree-winding ransforrnet
92 Fault calculation

short-circuit impedance measurements to those used for two-winding transformers

and autotransformers. Thus, let

ZH/~ = effective impedance of winding H with winding L short-circuited and

winding T open-circuited
Z ~ , ~ = effective impedance of winding L with winding T short-circuited and
winding H open-circuited
Z~H = effective impedance of winding T with winding H short-circuited and
winding L open-circuited

all three values being referred to the high-voltage side. Then, from the equivalent
circuit it is seen that

ZHL = ZH + Z'L
Z'Lr = Z'L + Zr [Link]
t
grH = Zr + ZH
from which

zH = I(ZHL + Z~H Z~T)

-

Z'L = I ( z . L + Z'Lr - Z'rH) [Link]

t 1 t
ZT = 2(Z~r + Z ~ H - Z.L)

z,,.
~ LI

ZH
HC ."'1-----

| "1 0 T'

Ze

Fig, 3.3.3F [Link] equivalent circuit o f a three-winding transformer with exciting

impedance included.

The equivalent-star representation of Fig. 3.3.3E is equally valid whether the three
windings are electrically separate or not and is therefore also applicable to the quite
common case in which two of the windings form an auto-transformer, the third
winding only being electrically separate.
If it is required to represent exciting impedance in the equivalent circuit this can
be done by adding an impedance of appropriate value between the star-point of the
 F a u l t calculation 93

I1/33 kV 33/6.6 kV

ZHL=J1.512, ZHL=J2,~

3 + j6,$I l ~ i+jl~,

12 kV
I I
phase-to-phase 4 + jS [~.
ZHL=J I,~I

1I/6.6 kV
jO.06~l jO.O8~l
4 1 6 0 + j0V
.-[Link]-
0 . 1 2 + jO.241Z

1.44 +jI.8012

4160 +jOV

O.12 + j 0 . 3 3 ~ 1 +jill
I I E::::3-
j0.7212

4 1 6 0 + jOV
1.12 + [Link]
r'-"l
855 - j l S 6 S A

Fig. 3.3.3G Example of Fig. 3.3.2F but with practical transformer Interconnectlona.
(common base voltage 6.6 k W
impedances ZH, Z'L and Z~- and neutral, as shown in Fig. 3.3.3F. This exciting
impedance must, of course, be to the same common voltage base as th¢ other
impedances.
The representation of transformer impedance is illustrated by the simplo
example shown in Fig. 3.3.3G, this being the same as that given in Fig. 3,3,2F
except that the transformers, previously assumed ideal, now have the given values
of winding impedance.
 94 Fault calculation

It might be expected that transformers having more than three windings would
be represented by an equivalent-star circuit similar to that shown in Fig. 33.3E
but with the number of branches in the star (exciting impedance ignored) equal to
the number of windings. This form of representation is not normally valid,
however, for transformers having more than three windings, and such transformers
normally require to be represented by an equivalent-mesh circuit. This mesh circuit

o BI

c 0

(, 1) I (,e
c o

IJl

0
B
A 0

c A (,i
0

E
¢

EI DI

Fig. 3.3.3H Common-base equivalent circuits of four-winding and five-winding transformers

with exciting impedance ignored

has one terminal per winding with every pair of terminals interconnected by an
impedance. Thus, a four-winding transformer would be represented by four
terminals interconnected by six impedances and a five-winding transformer by five
terminals interconnected by ten impedances. This form of representation is shown
in Fig. 3.3.3H.

3.3,4 Representation of off-nominal-ratio transformer circuits

It has so far been assumed that the transformers are all nominal-ratio transformers,
that is that the turns ratio of each transformer is equal to the ratio of the nominal
rated voltages of its associated network sections. Put in another way, it has been
 Fault calculation 95

assumed that the nominal rated voltage of each transformer winding is identical to
that of the network section to which it is connected. This assumption is commonly
made in fault calculations but it is nevertheless sometimes necessary to consider
cases in which the transformer ratio is different from this nominal value, the most
common cause of such a difference being the ratio change produced by transformer
tap-change equipment. A treatment of off-nominal-ratio transformers is given in
Appendix 3.8.1.

3.3.5 Transformer phase-shifts

The single-phase (reference-phase) representation of three-phase transformers, as

required for the system positive-sequence network, has so far proceeded on the
tacit assumption that all the transformer windings are star connected. It is necessary,

R (a) r (a)

v (b)
r"y"y'"y~ y(h)

a(c) r*'T*'T-Y~ b (c)

Fig. 3.3.5A

therefore, to examine the validity of this single-phase representation for those cases
where other winding connections (for example delta and interconnected-star
connections) are employed. The point at issue arises from the fact that, for a power
system without transformers, the three system phases have an identity which
extends throughout the network. With transformers present, however, the identity
of a given phase conductor is limited to the particular network section concerned,
this identity terminating at each and every transformer interconnection between
that network section and the rest of the system.
The single-phase representation is valid if all the transformers are star connected
since the phase references a, b and c can then apply to the corresponding phases of
all the network sections, the reference phase (phase a) then having an identity
which extends throughout the network. These conditions are shown in Fig. 3.3.5A.
Consider, however, the star-delta transformer shown in Fig. 3.3.5B in which the
star-side and delta-side phase-terminals are denoted by the phase letters R, Y, B and
r, y, b, respectively, allocated as shown. The relationship between the star-side and
the delta-side currents and voltages can be obtained by considering each limb of
the transformer core and its associated two windings to constitute a single-phase
transformer, and combining the delta-side phase-winding voltages and currents to
96 Faultcalculation
obtain the phase-terminal values. The vector diagrams show the resulting relation-
ship between the star-side and delta-side positive-sequence voltages under no-load
conditions and that between the star-side and delta-side positive-sequence currents,
exciting current ignored. The delta-side currents and no-load voltages, it wig be
noted, all lag their corresponding star-side values by 30 ° , this phase shift being
caused by the particular star-delta connection of the windings and the given alloca-
tion of the phase references. It readily follows, therefore, that with the transformer
providing the interconnection between two network sections, the effect of the
star-delta connection as compared with the star-star connection is that the currents
and voltages at any point in the delta-side network-section are all retarded in phase
by 30 ° with respect to any reference vector in the star-side network. Now, all the

R
Y
B

rE1 Vb

No-LoadVoltages
1B Ib

illl
l\x [

Iy ~0
Currents
Fig.3.3.6B Phase-shiftsintroducedby a star/delta transformer
 Fault calculation 97

transformers in an interconnected power system must have compatible phase-shifts.

In other words, the phase-shift introduced between any two network sections as a
result of transformer [Link] must be the same for all the inter-
connecting paths between the two sections. This means, in effect, that the phase-
shifts introduced by the transformer winding-connections have no effect on the
magnitudes of the current and voltage values, these magnitudes being identical to
those which would be obtained if all the transformer windings were star-connected.
The only effect of the transformer winding-connections, therefore, is to shift the
vector [Link] of the currents and voltages in any given network section
with respect to those in any other network section or with respect, for example, to
the chosen common reference vector for the whole power system. The amount of
this phase-shift is known if the winding connections of the interconnecting trans-
formers are known, the total phase-shift introduced by a number of transformers
in series being the algebraic sum of their individual phase-shifts.
The effect of transformer winding-connections can now be seen to be readily
obtained by simply analysing the system network with all the transformer windings
treated as if they were star-connected (that is by using the single-phase equivalent
circuits already described) and using the phase reference a, b and c to denote the
phases of this hypothetical system, these phase references denoting actual phase
conductors in one, or possibly more, of the network sections. The phase currents
and voltages so obtained can then be corrected in phase, as necessary, to allow for
the transformer phase-shifts, the resulting values then being correct in both magni-
tude and phase.
Thus, considering the particular case of a two-section system interconnected by the
star-delta transformer of Fig. 3.3.5B, let it be assumed that the reference phase
(phase a) has been chosen to correspond to phase R on the star side of the trans.
former. Then, at any given point in the star-side network-section, the phase voltages
and currents are given by

v s = Vo z s = I°
V y = a 2 Va I r = a2 Ia
VB =aV° 18 =a/o

where Va and Ia are the values of the reference-phase positive-sequence voltage and
current at the point in question. The symbol a is the 120 ° operator already defined.
Now the reference phase on the delta side of the transformer is a hypothetical con-
tinuation of that on the star side based on the assumed star-star transformer-inter-
connection and hence, for any point in the delta-side network-section, the phase
voltages and currents are given by

v, = 1/- 30°1I, i , = l/- 30%

Vy = 1[- 300a 2 Va ly = 1/- 30°aZ la
Vb = 1/- 30*aVa /b = 1/-3O°a/.
98 Fault calculation

where V a and Ia are now the values of the reference-phase positive-sequence voltage
and current at the point in question in the delta-side network section. The symbol a
is the 120 ° operator, as before, the vector multiplier l/s_.~_* representing the
necessary phase-shift correction.
The phase correction required in any given network-section is determined from a
knowledge of the winding connections of the transformers present in any inter-
connecting path between that particular network-section and the network-section
containing the chosen system reference vector. The phase-shifts are normally
known from the system phasing diagram but can otherwise be obtained from a
knowledge of the vector-symbol reference (for example, Ydl, Dyl 1, Yy0, etc.)of
the transformers concerned and their phase markings. The vector-symbol references
are defined in British Standard BS 171:1978, the first (capital)letter denoting the
connection of the high-voltage winding (Y for star, D for delta and Z for inter-
connected-star), the second (small) letter the connection of the low-voltage winding
using similar small-letter symbols (y, d and z) and the remaining figure (0, I, 6 or
11) being the clock-dial reference representing the hour position of the low-voltage
phase-vector in relation to that of the corresponding (similarly lettered) high-voltage
phase-vector, the latter being assumed to occupy the 0 (that is 12 o'clock) position.
In determining the network-section phase-shifts due account must, of course, be
taken of the actual phase references employed, usually red, yellow and blue, and
their allocation with respect to the B.S. transformer phase-terminal references. The
latter are A, B and C for the high-voltage winding and a, b and c for the low-voltage
winding and should not be confused with the general phase references a, b and c
employed in the network analysis.
In so far as balanced-load and three-phase short-circuit conditions are concerned,
it is of interest to note that correction for transformer phase-shifts is not normally
of any practical interest or value, since such correction merely ensures the correct
relative phase displacements of the currents and voltages in different network-
sections. The currents and voltages in any given network-section have correct phase
angles with respect to each other whether phase correction is applied or not, since
the effect of phase correction is merely to shift all the voltage and current vectors
in any network-section by the same fixed amount. In analysing balanced-load or
three-phase short-circuit conditions, therefore, phase correction is usually ignored
and the network analysed by assuming star-connected transformer windings.
As will be seen in later sections, however, the effect of transformer phase-shifts
must be considered in obtaining the correct analysis of unbalanced-fault conditions.

3.3.6 Representation of synchronous machines

The representation of a synchronous machine in the positive-sequence network is

by means of the single-phase (reference-phase) circuit shown in Fig. 3.3.6A, the
 Fault calculation 99

Fig. 3.3.6A Positive-sequence equivalent circuit o f a synchronous machine

ruJVVVVVUVUVVI .... ~q.%%.

~ ~-~ ~-~ ~A~ -~- Steady-stateva|ue

o • L___ ~JV'J",

~VVVUvvuuvv
Ih Steady-statevalue
o -----'u u v v

Fig. 3.3.6B Instantaneous phase-currents o f a synchronous machine during a three-phase

short-circuit at its terminals
100 Fault calculation

circuit comprising a positive-sequence [Link] in series with the machine

positive-sequence impedance. The values of voltage and impedance to be employed
in the representation, however, are dependent on the machine operating conditions
at the instant in question and on whether the condition to be analysed is a steady-
state condition or a condition following a sudden change in electrical conditions,
for example a system short-circuit.
The factors which determine the required representation can be understood by
considering the effect of a three-phase terminal short-circuit on an unloaded
machine as shown by the osciUograms of Fig. 3.3.6B, the field excitation voltage
being assumed constant at the value required to produce rated voltage at the
machine terminals prior to the application of the fault. The variation with time of

S t e a d y - s t a t e value

0:2 oY3 0:,

T i m e in s e c o n d s

Fig. 3.3.6C Variation with time o f the r.m.s, a.c. component o f short=circuit current o f
Fig. 3.3.6B
the r.m.s.a.c, component of the phase current is shown in Fig. 3.3.6C, this value
being the same in each of the three phases. It will be noted that the reduction in
the short-circuit current magnitude with time is exponential in form, and this
reduction is caused by the effects of mutual coupling between the stator and rotor
windings, as represented by armature reaction, and of similar coupling between
each of these windings and another effective winding representing induced-current
paths in the body and poles of the rotor and in any damping windings which might
be present.
The r.m.s, current magnitude at any time t after the instant of short-circuit is
given by t t
Est Et Tst Et Es Tt E s
I= e + e +-- [Link]
Zst Zt Zt Zs Zs

and is seen to consist of a constant component and two exponentially-decaying

components. The first exponentially-decaying component is a rapidly-decaying
component of time-constant Tst and the second a more slowly decaying component
of time-constant T t. The machine impedances Zs~, Zt and Z s are termed the sub-
transient, transient and synchronous impedances, respectively, and can be regarded
 Fault calculation 101

for most practical purposes as purely reactive, that is as/Xst,/Xt and/Xs. The
voltages Est, Et and E s are the voltages behind these impedances, their values being
the vector sum of the machine terminal voltage and the appropriate impedance
voltage-drop in the machine immediately prior to the short-circuit. The time-
constants Tst and Tt are termed the sub-transient and transient short-circuit time-
constant, respectively, the former usually being of the order of twenty or thirty
milli-seconds and the latter of the order of one or two seconds. For the case of a
short-circuit on an unloaded machine, the voltages Est, Et and E s would, of course,
all be equal to the pre-fault value of the open-circuit phase-to-neutral terminal
voltage of the machine.

Es

I"'.., p /
"~'.7

Fig. 3.3.6D Phasor diagram f o r a round-rotor synchronous machine

It will be noted from eqn. [Link] that the current at the instant of short-circuit
(t = 0) is equal to Est/Zst and is thus determined by the sub-transient impedance
only. Similarly, the current under steady-state conditions (t' much greater than Tst
and Tt) is given by Es/Z s and is determined by the synchronous impedance only. If
the initial, rapidly-decaying, transient component is ignored, the initial current is
effectively given by Et/Z t and is thus determined by the transient impedance only.
The vector diagram for a synchronous machine (cylindrical-rotor type assumed) is
shown in Fig. 3.3.6D where V represents the phase-to-neutral terminal-voltage of
the machine and I the current flowing out of the machine.
In addition to the a.c. components of short-circuit current, two at least of the
three phases will have an exponentially-decaying d.c. component of current as
shown by the off-set of the current waves with respect to the zero axis in Fig.
3.3.6B. The initial value of this d.c. component in any given phase depends on the
instantaneous value of the phase voltage at the instant of short-circuit and is a maxi-
mum (equal to the peak value of the initial a.c. component of the phase-current)
when the instant of short-circuit (t -- O) coincides with an instant at which the
voltage behind the sub-transient impedance is instantaneously zero for the phase in
102 Fault calculation

question. Conversely, the d.c. component of current will be completely absent

when the instant of short-circuit coincides with an instant at which the voltage
behind the subtransient impedance is at its maximum (peak) value. The time-
constant of the d.c. component of current is practically the same as the sub-transient
short-circuit time-constant Tst. Because the d.c. currents in the rotor field-winding
and damping-circuits are associated with a.c. machine-frequency currents in the
stator windings, it follows that the transient d.c. component of current in the stator
phases will be associated with a corresponding a.c. machine-frequency component
of current in the rotor field winding and damping-circuits. The variation in the
field-current of a short-circuited machine operating with constant field voltage is
shown in Fig. 3.3.6E.

If

S t e a d y-sigHt, v a l u e

0 0 I 0 2 0.3 0.4
I'Jlllt" ill ~eCcbllds

Fig. 3.3.6E Field current o f a synchronous machine during a three-phase short-circuit at its
terminals

Effect of automatic voltage regulators: The consideration of synchronous-

machine performance presented so far has assumed constant field voltage. In
practice, however, machines are commonly equipped with automatic voltage
regulators whose function is to control the excitation and thereby attempt to main-
tain the machine terminal-voltage at a constant pre-set value. Under short-circuit
conditions, therefore, the regulator will increase the field voltage in an attempt to
restore the terminal voltage to its normal pre-fault value and will thus increase the
short-circuit current to a higher value corresponding to the higher field current. The
effect of a typical regulator for a terminal three-phase short-circuit on an unloaded
generator is shown in Fig. 3.3.6F, the variation in the short-circuit current with
time being a function of the response characteristics of the regulator, the field-
circuit time-constant and the maximum available field-supply voltage.
Saturation effects." The synchronous reactance of a synchronous machine is
normally defined, in [Link] value (see Section 3.3.7), as the ratio of the field
current required to produce rated armature current on sustained three-phase
terminal [Link] to that which would be required to produce rated armature
voltage on no-load if no saturation were present, that is with the open-circuit-
voltage/field-current characteristic assumed to be an extrapolation of its initial
straight-line form. (See Fig. 3.3.6G.) The synchronous reactance is thus, by defini-
tion, an unsaturated value and is therefore applicable to those steady-state condi-
tions in which the main flux paths of the machine can be assumed to be unsaturated.
 Fault calculation 103

C Without A.V.R.
. . . . With A . V . R .
cJ 6
G e n e r a t o r e x c i t e d to gnve
rated v o l t a g e on N o - L o a d

.m

! 2 3 4
T i m e in s e c o n d s

Fig. 3.3.6F

In the case of the sub-transient and transient reactance, the flux paths concerned
are leakage-flux paths which are partly in iron and partly in air. These reactances
are therefore not strictly constant but are dependent on the degree of saturation of
the iron paths, it being necessary, for precision, to know the conditions for which
the stated values apply. The values obtained from a terminal short-circuit test with
rated pre-fault terminal voltage are known as the rated-voltage values of sub-
transient and transient reactance, the values being saturated values because of the
high value of armature current for this condition. Unsaturated values are obtained
by performing the short-circuit test at reduced pre-fault terminal voltage, for
example, at 50% of rated voltage.
Quoted values of sub-transient and transient reactance can usually be assumed to
be unsaturated values, unless otherwise stated, these values usually being some 10
or 15% higher than the saturated values.
Salient-pole machines: The equivalent circuit of Fig. 3.3.6A is, strictly speaking,
only valid for the case of round-rotor machines, there being no equivalent represen-
tation for salient-pole machines, the analysis of which requires the use of the two-
104 Fault calculation

Open-circuit test Short-circuit test

1.5
m

• .~(~o~..
:=

¢-,
~.o -y ~ i.o
._~.

E E

.~o.5 -r - 0.5
P

If I If 2 If3
Field c u r r e n t Field current

Per unit s y n c h r o n o u s impedance = lf~3 Short-circuit ratio = lf.-.~

2
ifl l f3

Fig. 3.3.6G Open-circuit and short-circuit characteristics o f a synchronous machine

axis theory. The given representation is, however, sufficiently valid for most
practical cases of fault calculation provided the impedances used are the direct-axis
values.

3.3.7 Use of per-unit and per-cent values

The network analysis described so far has all been in terms of actual or referred
values of voltage, current and impedance. In the great majority of cases, however,
the plant impedance values are given in per-unit or per-cent value, and it is accord-
ingly normally preferable to employ such values directly in the analysis rather than
convert them to their actual or referred ohmic equivalents.
Per-unit values: The basis of the per-unit method can be understood by consider-
ing the application of Ohm's Law to a single impedance of Z ohms, the voltage-drop
V in volts produced by a current of Iamps flowing through the impedance being
given by
V=I Z [Link]
where V and I are vector values and Z is the complex impedance, all values being
expressed by complex numbers. Now, letting Vbase and lt,ase be real numbers
representing stated values of voltage in volts and current in amps, the above equa-
 Fault calculation 105

tion may be written

"I '1I'-1
= 7=

which, in per-unit notation is written

=
[Link]

V p.u. = (I [Link] p.u.) [Link]

The per-unit values of voltage, current and impedance are thus defined by
V
V p.u. = [Link]
V~se
I
I p.u. = ~ [Link]

Zp.u. = m -- Z [Link]
Zbase

where the base impedance is given by

Vt,ase
Zt,a~ = [Link]

It should be noted that the base values of voltage, current and impedance are all
real numbers and it therefore follows that the per-unit values V [Link], I p.u. and
Z p.u. are complex numbers with the same arguments as V, I and Z. The per-unit
values, being ratios of similar quantities (that is volts, amps or ohms), are dimen-
sionless.
The base values of voltage and current, which must of course be stated, can be
any convenient values. In the analysis of three-phase systems the base voltage is
always chosen to be equal to the rated phase-to-neutral voltage of the plant item
or network section concerned and the base current, the current corresponding to a
stated three-phase power-level (usually expressed in MVA), at this same rated
voltage. In referring to individual plant items it is usual for the stated powerqevel to
be chosen to be equal to the load rating in MVA of the plant item concerned. Thus
considering a 132 kV transmission line with a positive-sequence impedance of
0-076 +/0.379 I2 per km and a load rating of 175 MVA, it follows that
132000
Vt,ase = - - = 76 300 V
43
so that
175 x lO s
It~se = = 765.4 A
~ / 3 x 132000
106 Fault calculation

from which
76 300
Zt~e = = 99.7 I2
765.4

Hence, the per-unit impedance of the transmission line is given by

0-076 +]0-379
Zp.u. =' 99.7 per km

from which Zt,.u ' = 0.000762 +]0~0380 per km to the stated rating of 175 MVA.
The base current for a given base voltage is directly proportional to the chosen
MVA base and hence the per-unit impedance of any given item of plant is also
directly proportional to the MVA base. Thus the per-unit impedance of the above-
mentioned transmission line to a base of 100 MVA is given by
100(0-000762 +/0.00380)
Zp.u. = per km
175

giving Zp.u. = 0.000435 +/0.00217 per km to the new base of 100 MVA.
Per-cent values: It is common practice in power system analysis to express plant
impedances in per-cent value rather than per-unit value, the former being simply the
latter value multiplied by 100. Expressed in per-cent values, Ohm's Law becomes
I%Z~
V% = ~ [Link]
100
where the per-cent values of voltage and current are similarly the corresponding per-
unit values multiplied by 100. Thus, the transmission line referred to in the previous
example has an impedance of 0.0435 + ]0.217 per-cent per km to a base of I00
/VIVA.
Use of common-base per-unit and per-cent values: The use of per-unit or per-
cent values of voltage, current and impedance in the analysis of a complete network
requires that all such values should be referred to a common MVA base,just as the
use of values expressed in volts, amps and ohms requires that these latter-mentioned
values, should be referred to a common voltage base. Thus, consider any item of
plant, say a transmission line, forming a part of a complete three-phase power
system and assume it to have a rated line-to-line voltage of V, and VA rating of S
and ohmic impedance of Z ~2 per phase. Then, as already shown in Section 3.3.2,
this item of plant would be represented in the system positive-sequence network by
the referred impedance value

Z' = Z [Link]

where Vt~zseis the line-to-line value of the chosen common base voltage. Expressing
this referred impedance in per-unit value on its own three-phase VA load-rating S,
 Fault calculation 107

and remembering that this referred impedance value now represents an equivalent
plant item with a rated line-to-line voltage of l/b~, it is seen from eqn. 3,3,7,6 that

Z p.u. = [Link]
v'b / 43
where
S
43Vba =/base

and
Vbase[ x/3 = phase-to-neutral value of the base voltage
Eqn. [Link] reduces to

Z p.u. = Z 3,3,?.11

Eqn. [Link], it will be noted, is the per-unit value of the actual plant Impedance
Z, thus showing that reference to a common voltage base is unnecessary when using
per-unit or per-cent values.
Consider, now, the representation of the given plant item in the system positive,
sequence network. The positive-sequence network expressed in terms of per-unit or
per-cent values of impedance must be identical except for a common multiplier,
to the same network expressed in terms of referred ohmic impedances to a common
voltage base Vbase. In other words, the impedances of corresponding branches in
the two networks will be in the same constant ratio for all the branches. Now, from
eqns. [Link] and [Link] it is seen that

Zp.u _ S 3.3.7,12
Z' V~

which is seen to be a constant ratio if all the plant items are assumed to have the
same VA load rating S. Conversion of all the per-unit or per-cent impedance values
to a common MVA base is readily achieved bearing in mind the direct proportion,
ality between MVA rating and per-unit or per-cent impedance. Thus, the required
conversion of all the impedance values to a stated common/VIVA base, which carl
be arbitrarily chosen, is given by
[ Z p.u. (or Z%) to ] = [ C o m m o n M V A b a s e ] [ Z p . u . ( o r . Z ~ ) t o 13,3.7.13
common MVA base | Plant MVA rating 1 [ plant MVA rating

A frequently used value for the common MVA base is the MVA load rating of the
108 Fault ca/cu/at/on

largest individual plant item concerned or some standard value such as 100 MVA,
500 MVA or 1000 MVA appropriate to the system under examinaffon.

3,3,8 Fault-calculation procedure

The analysis of a three-phase balanced-fault condition consists, in general, of three

parts, the first being the representation of the given power system with its fault
condition by the equivalent positive-sequence network, the second the solution of
this network in terms of its common-base values of voltage, current and impedance,
and the third the conversion of the resulting common-base values to actual values.
The system [Link] network as already stated, is the equivalent single-
phase (reference-phase) representation of the complete power system. In this
network each component item of plant is represented by its equivalent positive-
sequence circuit using common-base values of voltage, current and impedance,
namely values to a stated common voltage base or, as is more usual, per-unit or
pot-cent values to a stated common MVA base.
[Link] and cable circuits: Overhead-line and cable circuits are represented
by their equivalent-lt circuits, it being usually sufficient to employ the nominal-It
circuit in which the series arm represents the total series impedance of the circuit

( k + j~,.)L)I
A

c .......... t ; i

-2j I
wC!
-2J
~(.'1

. . . . ii

(a) N o m i n a l - ~ representation

(R + j ~ L ) I
¢ ....... I I 0

III

(h) Simple series-circult representation

Fig, 3,a,SA Equivalent circuits o f an overhead-line or cable circuit o f length I and

parsrnete~ R, L and C per unit length
 Fault calculation 109

concerned and each of the two shunt arms the impedance corresponding to one half
of the total phase-to-neutral capacitance. The shunt-arm impedances are always
large in comparison with the series-arm impedance, and representation by the series
arm alone is usually sufficiently exact for most practical purposes, particularly for
[Link] circuits.
Transformers and synchronous machines: Transformer and synchronous
machine impedances are predominantly reactive with X/R ratios typically between
ten and twenty, or sometimes more. It is, therefore, usually sufficiently exact to
ignore the resistive component of the impedances and to assume all the impedances
to be purely reactive.
Loads: Load impedances are always large in value in compariaon with the series
impedances of the power system plant and they therefore have only a small effect
on the value of the total fault current under short-circuit conditions. Load can
therefore be ignored in the majority of short-circuit calculations unless the analysis
is particularly concerned with the combined effect of load conditions and short,
circuit conditions.
Transformer tap-position: For the great majority of fault calculations it is
usually sufficient to ignore actual tap positions and to assume all the transformers
to be operating on the nominal-ratio tap-position, the error introduced by this
assumption being normally quite small insofar as values of total short-circuit
current are concerned. The use of the exact off-nominal-ratio representation,
referred to in Section 3.3.4, has many applications in practice, however, where it
is required to determine the precise division of load or fault current between
transformers operating on different taps.
Equivalent sources: The representation of a complex power,system network
can often be simplified considerably by the use of an equivalent generator to
represent the whole or certain parts of a given network. Thus, a complete network,
as seen from any given point, may be represented, using Thevenin's Theorem, as a
single driving voltage in series with a single impedance. This equivalent-generator
circuit can often be obtained with sufficient accuracy from an estimated knowledge
of the system three-phase fault-level at the point in question, the pre-fault value of
the voltage at this point being assumed equal to the nominal rated value.
Treatment of complex impedances: The impedances in the positive,sequence
network are all complex impedances and must, therefore, be represented in R +iX
form. In many cases, however, the resistance components of the impedances are
small compared with the reactance components and in such cases it is often
sufficient to treat the impedances as pure reactances, thus ensuring a considerable
simplification in computation. The use of such a pure-reactance form of representa,
tion, it should be noted, results in a short-circuit current slightly greater than the
true value.
Plant-impedance values: The impedance values employed in any particular fault
calculation should, wherever possible, be the known values appropriate to the
particular items of plant concerned. Where precise actual values are not known,
however, it may be permissible to use typical values appropriate to similar plant of
 1I0 Fault calculation

similar load and voltage rating. Tables of typical plant impedance values are given
in Section 3.4.3.
Neutral earthlng: Neutral earthing arrangements have no effect on balanced
three-phase conditions, whether load conditions or three-phase short-circuit condi-
tions, and are therefore disregarded in the derivation of the system positive-sequence
network.
3.3.9 Example 1

The power system shown in Fig. 3.3.9A develops a three-phase short-circuit on the
132 kV busbars at Station C. Determine the resulting value of the three-phase fault
current and its distribution in the 132 kV system, given that the pre-fault value of
Zg = j ! 8 . 0 % Zt = j i 2 . S % Zt = j l O . 0 % Zs = j 1 7 . 5 %

I I kV I00 M V A AJ 132 kV |B 60 M V A I I kV
I00 M V A J[ l 60 M V A

4.6 + j26.0.~ 2.3 + jl 3.0~

[Link]
the [Link] wltage at the point of fault is 140 kV. The system can be assumed
to be operating at no4oad prior to the fault.
Solution: Using a common voltage base of 132 kV, the resulting positive-sequence
network is as shown in Fig. 3.3.9B, the generator and transformer impedances
having been converted to equivalent 132 kV ~2. Thus, considering the 100 MVA
generator, its rated current is given by
100 x 106
I= = 5250 A
× 11000

and its rated phase-to-neutral voltage is given by

11 000
V= ~ = 6350 V
45

from which the corresponding base impedance is seen to be

6350
Z~ = = 1.21
5250
 Fault calculation 111

80.8 kV 80,8 kV
@ j3! .41Z j21.8~ 6.9 +j39.0~ j29.1l~ j50,912
r---1 l " A~ I/B ~ I I '~ Illl' '~'~

4.6 +j26,012~ C " 2.3 +jl3.0~2

Fig.3.3.9B
The ohmic impedance of the generator is therefore given by

Z =/'0.180 x 1.21 =/0.218

and the value referred to the common base voltage of 132 kV is given by

= ~'~-/ j0"218 =J31"4~2

Because the pre-fault conditions have been assumed to be no-load eondhions, the
generator driving-voltages, referred to the common base voltage of 132 kV, are both
given by
140 000
E = = 80 800 V

j53.2~2 ~ ~ 4 . 6 +j26,01~

Fig.3.3.9C
112 Fault calculation

this voltage being used as the vector base. Because the two driving voltages are equal
they can be connected in parallel, thus giving the equivalent circuit shown in
Fig. 3.3.9C.
B

j80.O~l~l.I 5 + [Link]~l

j53.2~1~ , ~ 2.30 + jl 3.O~

Fig, 3.3.9D

The delta-connected line-impedances 2.3 +/13.0, 4.6 +/26.0 and 6.9 +/39.0 can
now be replaced by the equivalent star-connected impedances shown in Fig. 3.3.9D.
Thus, the impedance connecting Station C to the star point is given by
,(2.3 +/13.0X4.6 +/26.0)
= 0.77 +/4.33 12
(2,3 +/13.0) + (4.6 +j26.0) + (6.9 +/39.0)

udng eqn, 3,2,5.8, the two remaining impedances connecting Stations A and B to
the star point being similarly obtained.
The remaining steps in the reduction process are shown in Fig. 3.3.9E from
which the resultant three-phase fault current is seen to be given by
80 800
If" = 79.7 -/1930 A
1.72 +/41.8

~ving
left 1932/-- 87° 40' A
The distribution of the fault current in the power system can now be obtained by
the process of back-substitution. This consists of retracing the steps in the reduc-
tion process starting with the total fault current and dividing this fault current
among the different branches by straightforward application of the network laws.
Thus, the current fed into the 132 kV system by the generator/transformer unit at
 Fault calculation 113

Station A is given by
(1.15 +]86.5X79.7 - / 1 9 3 0 )
= 55.3 - 11094 A
(1.15 +/86.5) + (2.30 +/66.2)

= 1095 [ - 87 ° 7' A

and hence the phase-to-neutral voltage of the reference phase at Station A Is given
by

80 800 -/53.2(55.3 - / 1 0 9 4 ) = 22 550 - / 2 9 4 4 V

= 22 720 [ - 7 ° 26' V
1.15 + j86.5~
80.8~kV
0.77 + j4,3311

2.30 + j66.21Z

80.8 kV

0.954 + j37.512 0.77 + j 4 , 3 3 ~

f----]

80.8 kV

1.72 + j41.8~2
f-"l

C Jplf

Fig. 3.3.9E

Similarly, the current fed into the 132 kV system from Station B is given by

(79.7 - / 1 9 3 0 ) - (55.3 - / 1 0 9 4 ) = 24.4 - 1836 A

= 836 [ - 88* 19' A
114 Fault calculation

giving a voltage at Station B of

80 800 - ]80.0(24-4 - / 8 3 6 ) = 13 920 - ]1952 V

= 14050 [ - 7 ° 5 8 ' V

Now, since the voltage at Station C is zero because of the fault, the current
flowing from Station A to Station C is

22 550 - / 2 9 4 4
- 38.9 -/.860 A
4.6 +/26.0
= 861 [ - 8 7 ° 25' A

Also, the current flowing from Station B to Station C, by subtraction, is

(79"7 - / 1 9 3 0 ) - (38.9 - j860) -- 40.8 - / 1 0 7 0 A

= 1071 [ - 87 ° 4 9 ' A

109~ [ 7-87071A 235 [ - 8 5 ° S g P A 836 [ - 8 8 ° 1 9 / A

"--~'-!A •I ]---a
(
861 [ - 8 7 ° 2 5 t A t 1071/ -870491A
• C
Fig. 3.3,9F
I f = 1932 [ - 8 7 ° 4 0 t A

Finally, the current flowing from A to B is

(55.3 - ]1094) - (38.9 - / ' 8 6 0 = 16.4 -/.234 A

=235[-85"59'A
The required fault-current distribution is thus as shown in Fig. 3.3.9F.
The voltage at the generator terminal at Station A, to the c o m m o n base voltage of
132 kV, is

80 800 - ]31-4(55.3 - ]1094) = 46 430 - j1737 V

the actual value, transformer phase-shift ignored, being

11(460430 - ]1737)
- 3 8 7 0 - j145 V
132
= 3870 [ - 2 ° 8 ' V
 Fault calculation 115

Similarly, the actual generator current at Station A, transformer phase-shift ignored,

is

132(55.3 - / 1 0 9 4 )
= 663.5 - ]13 130 A
11
= 13 1 5 0 / - 8 7 ° 7' A

The corresponding values of voltage and current for the generator at Station B can
be similarly shown to be 3190 -/103-6V (=3190 [ - 1° 51' V) and 293 - / 1 0 0 3 0 A
(= 10 030 L- 88 ° 19' A), respectively.

3.3.10 Example 2

The power system shown in Fig. 3.3.10A develops a three-phase short-circuit at the
point F on the h.v. terminals of one of the two auto-transformers at Station B, the
pre-fault voltage at the 132 kV busbars of Station B being 138 kV. The load
supplied by the 132 kV busbars can be assumed to be passive (that is, to contain no
120 MVA
275/132 kV
j i s.o%
210 MVA
200 MW 16.5/275 kV~ 5.72 + j24.M~ ~ 150 MVA

Zst -- j18.2% l 5.72 + j 2 4 . 8 ~

Zs = j143% , ~ ~ 138 kV

120 MVA
Fig. 3.3.10A 275/132 kV
j i s .0%

driving voltages) and to be representable by a constant impedance. Determine the

value of the short-circuit current and the network branch currents at the instant of
fault. Determine also the final steady-state value of the fault current and the net-
work branch currents assuming the fault to remain on the system and there to be
no voltage-regulator action.
Solution: Using the per-unit method and a common base of 200 /VIVA, the
system positive-sequence network for the pre-fault load condition is as shown in
Fig. 3,3.10B, the branch impedances having all been converted to common-base
per-unit values. The generator driving-voltage and impedance depend on the fault
condition being investigated, as already mentioned, and the values are therefore
denoted by per-unit values Eg and Zg as shown, the symbol Vg denoting the
generator terminal voltage.
116 Fault calculation

The conversion to common-base per-unit values has already been described in

Section 3.3.7. Thus, the impedance of the generator step-up transformer, namely
/14.0 per-cent on its own rating of 210 MVA, becomes

200 x/0.140
=/0.133 p.u. on 200 MVA
210
Similarly, the impedance of each auto-transformer, namely/15.0 per-cent on their
rating of 120 MVA, becomes

200 x/0.150
=/0.250 [Link] 200 M V A
120

The 275 kV transmission-line impedances, each of which corresponds to 76.5 km of

a twin-conductor-bundle line design, are given in ohmic value in Fig. 3.3.10A and
these have been converted by the use of eqn. [Link], the resulting per-unit
impedance of each line being given by

200 x l0 s x (5-72 +i24.8)

= 0-015 +i0.0655 p.u.
(275000)
to the common base of 200 MVA.
Now, the current taken by the load of 150 MVA at the stated pre-fault voltage
of 138 kV is
150 x 106
= 628 A
1.732 x 138000

and the current corresponding to the base MVA of 200 at a rated voltage of 132 kV is
200 x 106
= 875 A
1.732 x 132000

Hence, the scalar value of the load current in per-unit value is

628
= 0.718 p.u. on 200 MVA
875

The scalar value of the voltage across the load, namely 138 kV, is
138
= 1.045 p.u.
132
and hence the scalar value of the load impedance is given by
1.045
= 1.455 p.u. on 200 MVA
0-718
 Fault calculation 117

Now the power factor of the load is 0.850, lagging, corresponding to a load-
impedance phase-angle of 31.8", the sine and cosine of this phase angle being 0.527
and 0.850 respectively. The complex value of the load impedance is therefore given
by

1.455(0.850 +]0.527) = 1.236 +]0.767 p.u.

Adopting the pre-fault reference-phase voltage at the 132 kV busbar as the vector
base, its complex value thus being 1.045 + ]0 p.u., the current supplied to the load
is given by
1045
= 0"718/- 31"8 °
1.455 [ 31.8 °

= 0.610 -/0.379 p.u.

This is also the current supplied by the generator through its step-up transformer,
the current dividing equally between the two transformer-feeder circuits.
[Link] + jO.O6SS jO.2SO

~A Z j0.133
g B ~ : ~ ! ' 2+3j0.767
6
0.015 +j0.06SS jo.2so

Vg= 1.160+j0.175 0.610 -- j0.379

Fig. 3.3.10B

The generator terminal voltage is given by

Vg = (1.045 + / 0 ) + [~(0-015 + /0.3155) +/0.133] (0.610 - /0.379)

giving
Vg = 1.160 +/0.175 p.u.
The pre-fault conditions are thus as shown in Fig. 3.3.10B, and the fault analysis
can now proceed by applying the fault connection between the point of fault F and
the neutral bar with appropriate values of Eg and Zg for the condition under
investigation.
Considering, first, the conditions at the instant of short-circuit, the required
generator impedance is the sub-transient value, namely/18.2 per-cent on the
machine rating of 222 MVA, the required common-base per-unit value being

200 xj0.182
=]0.164 p.u. on 200 MVA
222
118 Fault calculation

The generator driving-voltage is the pre-fault value of the generator terminal-voltage

plus the pre-fault impedance drop in the machine sub-transient impedance, giving

Eg = Est = (1.160 +/0.175) +/0.164(0.610 -/0.379)

from which
Eg = 1.222 +]0.275 p.u.
The positive-sequence network for the conditions at the instant of fault is shown in
Fig. 3.3.10C and this can be solved for the network branch currents by the process
of network reduction and back-substitution, the reduction procedure being shown
in Fig. 3.3.10D.
0.015 + j0.0655 j0.250
1.222 + ~o.275 !"-"-! o /
jO.164 jO.133 ~ 1.236 + j0.767
~ C"-I-o-f~ A 8 ,""'}
0.015 + j0.0655 j0.250
I ; ; I

F"
i If

Fig. 3.3.10C

From Fig. 3.3.10D, the given network is seen to reduce to the generator driving-
voltage Est in series with the single impedance 0.0124 +]0.356 p.u. and, hence, the
generator current is given by
1.222 +/'0.275
= 0.892 -/3.405 p.u
0.0124 +/0.356

Applying the process of back-substitution to obtain the remaining branch currents,

the voltage at the point A is obtained from

VA = (0.892 - ]3.405) (0.0124 +/0.0590)

giving
V.4 = 0.2105 +/0.00983 p.u.
The current flowing from A to F through the 275 kV line associated with the faulted
transformer can now be obtained and is

0.2105 +/0.00983
= 0.842 - 3.020 p.u.
0.015 +/0.0655

The current flowing from A to B through the other 275 kV line is the generator
current minus the current just obtained and is given by
 Fault calculation 119

1.222 + j0.275
0.015 + j 0 . 3 1 5 5 B i.236 + j0.767
Ij0.297
; i ["-'3

0.015 + j0.0655 j0.250

1.222 + jO.275
0.OI5 + j0.3155 B
i""'1

0.015 + jO.0655 0.0302 + j0.2252

1.222 + j0.275
0.0452 + j0.5407
t jO.297
; "i | I

0.015 + jO.0655

1.222 + j0.275 1.222 + j0.275

0.0124 + jO.356
__~ j0.297
f ; i - ~ ; 1
lg
0.0124 + j0.0590

Fig. 3.3.10D

(0.892 - j3.405) - (0.842 - j3.020)

= 0.050 - j0.385 p.u.

The voltage at B can now be obtained and is given by

(0.2105 +/0.00983)- (0.050- ]0.385) (0.015 +]0.3155)

= 0.0882 - ]0-00027 p.u.
The current flowing to the fault through the transformer on the faulted circuit is
 120 Fault calculation

given by
0.0882 -/0.00027
-- - 0.001 - 0-353 p.u.;
i0-250
and the total fault current at F is thus:

/ f = (0.842 - ./3.020) + (- 0.001 - / 0 . 3 5 3 )

from which
I / = 0.841 - i3.373 p.u.
The base current corresponding to 200 MVA at 275 kV is given b y
2 0 0 xlO 6
- = 420 A
1.732 x 275 000
and hence the fault current in amps is given b y
I! = (0.841 -/3.373)420
giving
I / = 353 - / 1 4 1 6 A = 1460 / - 76.0°A
The fault current and the branch currents are shown in per-unit value in
Fig. 3.3.10F.
Considering now the determination o f the final steady-state values o f the fault

1.222 + j0.275 [----U'--1 ~ ;.-..J I

•~ ] [Link] - j0.385 /
-@ - / o.o , - jo.o,

'> ; 2t--rg--f-'--~. ~-
0.842 -j3.0 0 ' [ -0.001 -j0.353

,~ If = 0.841 - j3.373

Fig. 3.3.10E

163 / -82.6 °
ll.9~/J2.7°kV • >

A
148F/ -9~B.
s2"s / ~32'1°
"~
[-~

If = 1 4 6 0 [ - - 7 6 . 0 °

Fig. 3.3.10F
 Fault calculation 121

current and the network branch-currents, the generator impedance applicable to

this condition is the synchronous value, namely/143 per-cent on the machine rating
of 222 MVA. The required common-base per-unit value is thus

200 x/1.43
=/1.288 p.u. on 200 MVA
222

The required generator driving-voltage is the pre-fault voltage behind the

synchronous impedance (the excitation being assumed to remain unaltered) and
this is given by

Eg = E s = (1.160 +/0.175) +/1.288(0.610 -/0.379)

from which Eg = 1.648 +/0.961 p.u.

The positive-sequence network for the final steady-state condition, therefore, is

as shown in Fig. 3.3.10G and this can be solved for the network branch currents, as
before, by the process of network reduction and back-substitution. Because the

0.015 + j0.0655 j0.250

1.648 + j0.961 ~ ~ } l
~ jO.13 1.236 ÷ j0.767
~' A B, f--'l
l O . O J s jo.o6ss jo.2so

Fig. 3.3.10G

positive-sequence network external to the generator terminal is the same as for the
previous case, its equivalent impedance as seen from the generator terminal is the
same as before, namely

/0.133 + (0.0124 +/0.0590) -- 0.0124 +/0.192 p.u.

The equivalent positive-sequence network, as seen from the generator, is therefore

as shown in Fig. 3.3.10H and comprises a driving voltage of 1-648 +/0.961 p.u. in
series with an impedance of

/1.288 + (0-0124 +/0.192) = 0.0124 + / I 4 8 0 p.u.

122 Fault calculation

1.648 + j0.961

0.0124 + j 1 . 4 8 0

lg

Fig. 3.3.10H

The current supplied by the generator, therefore, is given by

1.648 +/0.951
= 0.659 -/1-108 p.u
0.0124 +/1.480.

and, using the process of back-substitution, the network branch currents are found
as shown in Fig. 3.3.101, the fault current at F now being given by

I f = 0.637 -/1.101 p.u.

,6.+,090, : > : , ,
0,057 - - j 0 . 1 3 2 /

~ ~t ~~ ~ 1 I 1
"~" 0.659 -- j1.108 1 0 . 0 2 2 -- jO.O07

•
o.~o~ - j O ~ ;
35 - j o<. l": 5
ll = 0.637 - j l . l O l

Fig. 3.3.101
I i

The fault current value in amps, therefore, is given by

If = (0.637 - i1.101)420
giving
I f = 268 - ]463 A = 535 [ - 60.0°A
The values of the network branch currents in amps are shown in polar form, in
Fig. 3.3.10J.
It should be noted that the results obtained by the method of solution employed
above could have been obtained equally well by the use of Thevenin's Theorem and
the Superposition Theorem. Thus, the impedance of the network of Fig. 3.3.10C,
as measured between point F and the neutral bar with the fault connection
removed and the driving voltage Est short-circuited, is 0.0549 + ]0-310 p.u., as can
be seen from the network reduction shown in Fig. 3.3.10K. Also the pre-fault
 Fault calculation 123

() I ' ~ Z ! -,6 6~'

18.2 [ 30.2__ kV

~ol s I ...s~.=~'_' , ~ ' ~ " ~ I " - k"A 1' ~ ' ' - ' 6.'" I - " - - " 1
Q >
LLJ-t" ..... ;

I
If = 535 [ - - 6 0 . 0 °

Fig. 3.3.10J

voltage Ef at the point of fault F is given by

Ef = (1.045 +i0) + 0.500(0.610 -/'0.379)(/0.250)

giving
Ef = 1.092 +/0.076 p.u.
Hence, applying Thevenin's Theorem, the fault current for a short-circuit at F is
given by
1-092 +/0.076
0.0549 +/'0.310
giving
If = 0.841 -/3.373 p.u.
This is seen to be identical to the value obtained by direct solution for the network
branch currents.
Knowing the total fault current and the pre-fault branch currents, the network
branch currents with the fault condition present are readily obtained by use of the
Superposition Theorem as shown in Fig. 3.3.10L. Thus considering the pre-fault
load condition represented by diagram (a) of this figure let E¢ denote the pre-fault
voltage at the point of fault, the value of Ef being 1.091 + j0.076 p.u. as already
computed. It will be evident that the network branch currents in diagram (a) will
be undisturbed if a fictitious driving voltage of value Ef is assumed to be connected
between the neutral bar and the point F, the direction of El being from the neutral
bar to the point F. The pre-fault load condition indicated by diagram (a) can there.
fore be regarded as caused by the joint action of the driving voltages Est and El. By
Thevenin's Theorem, however, the fault current at F is obtained by applying a
driving voltage Ef between the neutral bar and the point of fault F, a~ shown in
diagram (b), with the driving voltage Est removed (that is short-circuited), It should
be noted in this diagram that the voltage Ef acts from the point F to the neutral
bar, as is necessary to obtain the correct direction for the fault current, this being the
same as an applied driving voltage of - E l acting from the neutral bar to the point
of fault. Diagram (c) represents the required fault-plus-load condition for wtdeh it is
required to determine the values of the network branch currents and it will be
noted that this corresponds to diagram (a) with the assumed fictitious driving
124 Fault calculation

0.01S + j0.0655 j0.2S0

j0,164 jo.~3131 ~ c"-! 1.236 + j0.767

:~ /- -:' ~ . ~ A B ! !

1 0.015
I
+ j0.0655
I o--1
F
j0.250
"'1

j0,297 0.0075 + j0.0345 j0.125 1.236 + j0.767

A S~ B

0.0047 + j0.02617

F
iii II ]]I IIII

|,236 + j0.892

0.0047 + j0.02617 0.0549 +jO.310

[Link],,
. 98s
• t t I o
F'
f'7"'l

I II III III I I

Fig. 3,3,10K

voltage E/short-circuited.
Let the current in any given branch in diagrams (a), (b) and (c) be denoted by
Ik~d, licit and I/oad~a~rault, respectively. Then by the Superposition Theorem it
readily follows that:

lJoad = IIoad-p~fault - [fault

 Fault calculation 125

1.222 + jO.27S
0.305 -- jO., 89 B ~ 3 , 9
~.610-~j0.379 0.305 j0.189
(a)
FIE
fA = 1,092+jO,O'/6
/

[--'1 > ; I
--0.255 --j0,196 I
[ : > , ) ^ B < :'I
0.282 - j3.026
0.S37 - j2.831 F" ,. ~. -/ ° ' $ 5 9 - j ° ' 3 4 7

(b)

1,222 +j0.275 ] ~ ~ ; J [
- - " / [Link] - jO.38S /
89~ - j3.4os| ~ /[Link],- jo,o32 I
-->' ~ ;<" I
(c)
°"'-'"°'° I-°°°' -'°"" I

Fig.3.3.10L
from which the required value of branch current for the [Link]-fault condition
is given by:

l~oad.o~fauu = Ik,ad + Ifau#

This relationship is true for each branch of the network as is easily soon from
Fig. 3.3.10L. The use of the Superposition Theorem in this manner is particularly
useful in cases where fault analysis is required to take due account of a known pre.
fault load condition.
 126 Fault calculation

3,4 Calculation of unbalanced fault conditions

3.4,1 Symmetricaleomponents

A full and proper analysis of unbalanced conditions in a three-phase network is

made possible by the fact that any given set of unbalanced three-phase vectors,
which may be voltages or currents, can be represented by the sum of three sets of
balanced or symmetrical vectors, namely: the positive-sequence set, consisting of
three vectors all equal in magnitude and symmetrically spaced, at 120 ° intervals, in
[Link] with a stated phase order (termed the positive-sequence phase-order)
equal to the phase order of the system generated voltages; the negative-sequence
set, consisting of three vectors all equal in magnitude and symmetrically spaced, at
120 ° intervals, in time-phase, their phase-order being the reverse of the positive-
sequence phase-order; and finally, the zero-sequence set, consisting of three vectors,
all of which are equal in both magnitude and phase.
These three sets of component vectors are called the positive, negative and zero
phase-sequence components of the given set of vectors, and are represented
vectoriaUy in the manner shown in Fig. 3.4.1A. The assumed direction of rotation
for all voltage and current vectors is the anti-clockwise direction, and it thus follows

f
a

Positive sequencu Nugativ*: sequence Zero suquence

Fig, $,4.1A Phase-diQgramrepresentation o f phase-sequence components

that the positive-sequence vectors rotate in the positive-sequence phase-order a, b,

c and the negative-sequence vectors in the phase-order a, c, b.
Relations between phase vectors and their sequence components: Let Is, I b and
Io be any set of unbalanced three-phase vectors, where the subscripts a, b and c
denote the three phases in positive-sequence phase-order. Then denoting the posi-
tive, negative and zero phase-sequences by the second subscripts l, 2 and 0,
respectively, the three sets of component vectors are the positive-sequence set
 Fault calculation 127

lax ,lt,l (= a21at) and Ict (= alal) 3,4A.1

the negative-sequence set
Ia2 , It, 2 (= ala2 ) and le2 (= a2 Ia2 ) 3,4.1.2
and finally, the zero sequence set
Iao, lt,o (=Iao)andleo (= Ia0) 3.4.1,3
where the symbol a (not to be confused with the phase reference ) is the 120'
operator already referred to in Section 3.3.1.
It will be noted that the nine sequence components can all be expressed in
terms of the three components of any one chosen reference phase, here denoted
by phase a.
From what has already been said, therefore, we have that

It, = lb~ +It,2 +1be 3,4,1.4

and rewriting these equations in terms of the reference-phase (phase-a) sequence

components only, we get

la =11 +12 +1o

It, = a2 It +a/2 +Io [Link]
l~=aI~ +a212 +1o

the phase a subscript being omitted from the sequence components because phase-a
is understood.
From these equations, the sequence components of the reference phase (phase a)
can be obtained in terms of the three phase currents, the resulting equations being

I, =~- (Io +atb +a~Ic)

,q =~- (Io +a2~rt, + a,rc) [Link]
Io =~-6ra +It, +L)
as can readily be checked by substituting for Is, Ib and le in this latter set of
equations using the values (expressed in terms of the phase-a sequence components)
given by Eqns. [Link].
The validity of these relationships between phase vectors and their sequence
components can be demonstrated by the three vectors

I a = 0 +j14
.rb = 5 +/5
Ic =4-]8

shown in Fig. 3.4.1B. Applying eqns. [Link], [Link] and [Link] to these phase
t28 Fault calculation

values we obtain

Jal ~
(0 +/14) * + .(-
........
0.5 +/0.866X5 +/5) + (- 0-5 =/0.866)(4 - / 8 )
........

(0 +/14) + (- 0.5 -/0.866X5 +/5) + (- 0.5 +/0.866X4 - / 8 )

I , ' ) '~l m; . . . . . . . . . . . . . . . . . .
3

(0 +/14) + (5 +/5) + (4 - / 8 )
IIIO ga . . . . . . . . . . . . . . . . . . . . .
3

from wkleh ial =" 5,253 +/5.455

la2 ffi2.253 +/4.878
lao ffi3.000 +/'3.667

Using these values to obtain those for the other two phases gives
lbt = (- 0.5 - i0,866) (- 5.253 +/5-455) = 7.350 +/1.822
Ib2 " (- 0,5 +/0.866)(2.253 +/'4.878) = - 5.350-/0.488
lao ffi3.000 ÷/3.667
and
lcl = (- 0.5 +/0.866) (- 5.253 + i5.455) = - 2.101 -/6.976
lea = (- 0.5 -/0.866) (2.253 +/4.878) = 3.098 -/4.390
/co - 3.000 +i3.667

The== sequence components, rewritten in polar form, are

L,t ffi 7,5731133:9', 182 = 5.372[65"2 ° ) Iao = 4.737/50.7 °
/bt = 7'573113"9*, Ib: = 5.372/-178-8", It, o = 4.737[50.7"
Icl = 7'573/" 106'1", Ic2 = 5.372[-54.8 °, lco = 4.737[50.7 °

The sequence components are shown in vector-diagram form in Fig. 3.4.1B.

The relationships between phase values and sequence-component values,
although applied to current vectors in eqns. [Link] to [Link], are, of course,
equally applicable to voltage vectors, the resulting equation being

Va = V~ + V~ + v° 1
Vl, = ai V~ + aV2 + Vo [Link]
Vc "-aVl +a2V2 + Vo

for the phase values in terms of the reference-phase (phase-a) sequence components
and
Vt + aVb + a 2 Vc) I
I/2 ='~(Va + a2 Vb +aVe) I [Link]
Vo + vb +
 Fault calculation 129

for the converse relationships giving the sequence components of the reference
phase (phase a ) in terms of the phase values.
Because the symmetrical components of current and voltage in each of the three
phases can be expressed in terms of those of the chosen reference phase (phase a), it

a2l~ al h

\
\ la
\
h \

0
7 Ibl

Posltive
~1 c
a21h /al c

r// \\ \
Ia \ Ic
I / /
la2 /
//
/f
Ib2 Negative /luo, Iho, leo

Ic2
0 ~¢ro

I I
0 .~ I0 Amps
Fig. 3.4.1B
is usual to perform symmetrical-component analysis in terms of the sequence
components of this reference phase. This permits the phase subscripts to be deleted
from the sequence quantities, it being understood that, in the absence of the phase
subscript, the sequence quantities referred to are the phase-a values.

3.4.2 Phase-sequence networks and impedances

The balanced nature of power-system plant impedances has already been referred to
130 Fault calculation

in Section 3.3,1, and it follows directly from this characteristic phase symmetry
that, in any such balanced-impedance circuit, the voltage-drops produced in the
three phases by phase currents of any given phase-sequence will themselves be of
that same phase-sequence. Thus, the flow of positive sequence currents through
such a circuit will produce positive-sequence voltage-drops in the three phases and
no others. Similarly, negative-sequence currents will produce only negative-sequence
voltage-drops and zero-sequence currents only zero-sequence voltage-drops.
The vector ratio of the phase-sequence voltage-drop to the phase-sequence
current producing it ts the same in all three phases and is termed the appropriate
phase-sequence impedance of the circuit concerned. There are thus, in general,
three phase-sequence impedances for any given three-phase circuit, namely the
positive-sequence impedance denoted by Zt, the negative sequence impedance
denoted by Z~, arid the zero-sequence impedance denoted by Zo. These impedances
and their associated phase-sequence currents and voltage-drops are related by
Ohm's Law, giving:

Val = Z l l a l Va2 = Z2Ia2 Vao = Zol~o ]

Vbl =Zllbl Vb2 =Z2Ib2 VbO =Zo[bO [Link]
V¢1 = ZIIcl go2 = Z2Ic2 Vco = Zolco t
In general, the phase-sequence impedances Z~, Z 2 and Zo of any given three-phase
circuit have different values, although for static (that is, non-rotating) plant the
positive and negative sequence impedances Z~ and Z2 are always equal.
Expressed in terms of phase a, the voltage-drop equations are
Vl = Zsll
V2 = Z212 [Link]
Vo = Zolo

the phase subscript being deleted since phase a is understood.

Considering, now, any balanced-impedance three-phase circuit carrying un-
balanced phase currents la, lb and lc, the corresponding reference-phase (phase a)
sequence currents are given by eqns. [Link] and the resulting phase-sequence
voltage-drops (phase-a values) by eqns. [Link]. The resulting total voltage-drops in
the three phases are therefore

Ira ~ IiZl +I2Z2 + IoZo }

Vb =a2IiZs +a12Z2 + IoZo [Link]
Ve =aliZi +a212Z2 +IoZo

those equations being obtained by substituting from equations [Link]. into eqns.
[Link].
[Link] networks: The analysis of balancedqoad and three-phase short-
circuit conditions has already been discussed in Section 3.3 where use was made of
the equivalent single-phase (reference-phase) representation of the power system
 Fault calculation 131

appropriate to such balanced conditions. This equivalent network, termed the

system positive-sequence network, is valid for the balanced conditions in question
since, under such conditions, the power system voltages and currents, and the
appropriate power system impedances, are all positive-sequence quantities. In order
to be able to analyse unbalanced fault conditions, therefore, it is necessary first to
consider the equivalent system networks appropriate to such unbalanced conditions,
bearing in mind that the system voltages, currents and impedances to be taken into
account now include, in general, both negative and zero sequence values in addition
to positive-sequence values.
In balanced-impedance circuits, as already stated, currents of any given phase-
sequence produce voltage-drops of that same sequence only, the value of the
voltage drop for a given value of phase-sequence current being fixed by the appro-
priate phase-sequence impedance of the circuit concerned and being quite inde-
pendent of the presence of currents of any other phase sequence in the circuit in
question. This independence of the phase-sequence quantities means that in any
power system consisting of balanced circuits:
(a) the positive-sequence currents are determined solely by the positive-sequence
driving voltages produced by the power sources, the [Link] voltage
at the point of fault and the system positive-sequence impedances;
(b) the negative-sequence currents are determined solely by the negative-sequence
voltage at the point of fault and the system negative-sequence impedances;
and,
(c) the zero-sequence currents are determined solely by the zero-sequence voltage
at the point of fault and the system zero-sequence impedances.
It is important to note that the e.m.f.s produced by the generators and other
synchronous machines are positive-sequence voltages, there being no generated
negative- or zero-sequence voltages in the power system. Hence, any negative- or
zero-sequence currents present in the power system exist solely as a result of the
negative-sequence and zero-sequence voltages, respectively, at the point of fault.
It will now be evident from the above-mentioned considerations that the analysis
of unbalanced fault conditions requires, in general, the use of three single-phase
(reference-phase) networks, one for each of the three phase-sequences. These
networks are termed the system positive, negative and zero sequence networks, the
several items of plant which constitute the given power system being represented in
each network by their appropriate phase-sequence equivalent circuits.
Positive-sequence network: The system positive-sequence network, already
referred to in Section 3.3.1, is a single-phase network representing the reference
phase (phase a ) of the given power system in so far as positive-sequence quantities
are concerned. Each three-phase circuit is therefore represented by its positive-
sequence impedance or impedances and, in the case of a power source, by a driving
voltage representing the generated e.m.f, behind the source positive-sequence
impedance.
Negative-sequence network:The system negative-sequence network is a single-
phase network representing the reference phase ([Link] a ) of the given power
132 Fault calculation

system in so far as negative-sequence quantities are concerned. Each three-phase

circuit is therefore represented by its negative-sequence impedance or impedances.
Because there are no generated negative-sequence e.m.f.s in the power system, every
power source is represented simply by its negative-sequence impedance, there being
no driving voltages corresponding to those in the positive-sequence network. As
already pointed out, the positive and negative sequence impedances of any item of
static plant are always equal, and it therefore follows that the positive and negative
sequence networks differ only in so far as the representation of rotating machines is
concerned.
Zero-sequence network: The system zero-sequence network is a single-phase
network representing the reference phase of the given power system in so far as
zero-sequence quantities are concerned. Each three-phase circuit is therefore
represented by its zero-sequence impedance or impedances. Because there are no
generated zero-sequence e.m.f.s in the power system, the zero-sequence network,
like the [Link] network, contains no driving voltages. The zero-sequence
network is the only network in which neutral-to-earth connections are shown, the
currents in such earthing connections consisting of zero-sequence currents only.
Sequence network representation: It is convenient to depict the three phase-
sequence networks, in general terms, by the simple block-diagram form of repre-
sentation shown in Fig. 3.4.2A, it being appreciated that the rectangle representing
any given sequence network is assumed to contain the whole of that sequence
network for the particular power system under consideration. Any particular
point in the power system (for example, a point of fault F) can be represented by a
suitably lettered corresponding point in each of the three sequence networks.

Positive-seq u e n c e Negative-seq uent'e Zero-seq uence

network network network

Fig. 3.4.2A Block diagram representation of sequence networks

The thick base line of each rectangle in the block-diagram representation is
termed the [Link] bar and represents the zero-potential reference for the
power system, namely earth potential. It is important to note, however, that under
fault conditions the earth potential will, in general, be different at different points
in the power system, these potential differences being caused by the flow of current
through the earth. However, because it is never necessary to determine the values of
the phase-sequence voltages at any given point on the system with respect to the
earth potential at some other point, any ambiguity is avoided by always interpret-
ing the term 'earth potential' to mean local earth potential. Thus, in the case of the
zero-sequence network, the voltage at any point in this network, measured with
respect to the zero-potential bar, represents the zero-sequence voltage at the corres-
ponding point in the power system, measured with respect to the local earth at that
point. Similar reasoning is applicable in the case of the positive and negative
 Fault calculation 133

sequence networks, although for these two networks it is equally valid to regard the
zero-potential bar as representing the system neutral.
Considering any short-circuit condition at a point F in the power system, this
may be represented by the sequence networks in the manner shown in Fig.
3.4.2B(a), where the phase-a sequence currents flowing in to the fault are denoted by
If t, If2 and I/~ and the phase-a sequence voltages at the point of fault by Vft, Vf2
and Vfo. In the case of the open-circuit fault condition, the sequence-network
representation is as shown in Fig. 3.4.2B(b), the two sides of the open-circuit in the
faulted circuit being denoted by P and Q. In this latter case, the voltages Vft, Vf2
and Vfo are the phase-a sequence voltages produced between the points P and Q by
the open-circuit fault condition, I~,/2 and Io being the resultant phase-a sequence
currents in the faulted circuit.
The phase-sequence voltages and currents produced by the different types of
fault are determined in Sections 3.4.4 and 3.4.5 together with the sequence-
network interconnections necessary to represent these fault conditions.

Ifl

FI Pl QI
Positive-seq uence posit ire -seq ue nce
network network
v

If2

~i'
F2
Negative-sequence
P2 Q2
Negative-seq uence
network network

lfo
>

I
F"0
1 i
Zero-seq uence
network
Zern°e;ewq:/knce
v

(a) Short-circuit condition (~) Opon-circuit condition

Fig. 3.4.2B Representation of fault conditions by sequence networks

134 Fault calculation

3.4.3 Phase-sequence equivalent circuits

The positive, negative and zero phase-sequence impedances of any given three-phase
circuit can be determined by test by applying three-phase voltages of the appropriate
phase-sequence and measuring the resultant phase currents, short-circuiting connec-
tions being applied to the circuit terminals, as necessary, in order to provide a path
1 P Q
--~ ) e --- o •
a21 1
---~ ) o --- c

__@ ]
.=.-

Z I = Z 2 = E/I
z~ (=z2)
PO ! I OQ
Positive (and negative)
sequence circuit

Fig. 3.4.3A Measurement o f the positive (and negative) sequence impedance o f a trans-
mission line
for the flow of the phase-sequence currents. Thus, the positive (and negative)
sequence impedances o f an overhead-line circuit (shunt-admittance ignored) and a
star/delta transformer could be measured in the manner shown in Figs. 3.4.3A and

__Q •
-@ a21
,
a,
)

....-

Z I = Z 2 = E/I = Z H L

z i (=z2)
It O | | 0 L'
Positive (and negative)
sequence circuit

Fig. 3.4.3B Measurement o f positive (and negative) sequence impedance o f a transformer

 Fault calculation 135

I P Q
o

I o c

I
I

z o = Ell

Zo
v o I I OQ

Zero-seq uence
circuit

Fig. 3.4.3C Measurement o f the zero-sequence impedance o f • transmission line

I H
>

rQ
1
r

I
I
31 I
_L
Z 0 = E/I = Z H L

Zo

Zero-sequence H 0 [ I I n 0 L'
circuit
/
Fig. 3.4.3D Measurement o f the zero-sequence impedance o f a transformer
3.4.3B, respectively, and the zero-sequence impedance in the manner shown in
Figs. 3.4.3C and 3.4.3D. Because these are both items of static (that is non-rotating)
plant, the negative-sequence impedance is in each case equal to the positive-sequence
value.
Table 3.4.3A gives the phase-sequence circuits of a number of arrangements of
lumped impedances, the arrangements in question being representative of a number
of typical items of power system plant. Thus, the first arrangement shown, namely
a simple three-phase series circuit could represent a three-phase series reactor, and
the second, namely the star-connected impedances, a three-phase shunt-reactor,
136 Fault calculation

Circuit Pos've and neg've Zero

arrangement sequence c i r c u i t sequence circuit

Z
0...1 ; 0
Z Z Z
C ~ I ¢ ; ; 0
o ~

;/111////////////

Z
0"I r
z Z
o C ~ on

.////////////////

o..4---=i
Z
Z Z n
0 I Z]_~,.~l ,n 0 l"=-=l i 0 i :
L I
o ,
3Z n

i ii
.///i////////, ./
n n

Z.//////~//////
// Not
applicable

t 3Z n

:o
z/3
C ; I

"///////////////

shunt-capacitor or load. The phase-sequence circuits appropriate to the arrangement

of delta-connected impedances are obtained by applying the delta-to-star trans-
formation, the resulting positive (and negative) sequence impedance being Z/3 as
shown, where Z is the value of each impedance in the delta. The zero-sequence
impedance of the delta-connected impedances, as seen from the circuit terminals,
is obviously infinite since there is no path to earth for the flow of zero-sequence
current. The zero-sequence circuit in this case, therefore, is simply an uncon-
nected terminal, as shown.
In obtaining the positive and negative-sequence circuits, it is important to note
 Fault calculation 137

that the system electrical neutral points are all at zero potential with respect to
earth in so far as positive and negative-sequence voltages are concerned. The star
point n of the set of three star-connected impedances (all equal) is therefore shown
connected to the zero-potential bar in the positive and negative-sequence networks.
In the case of the delta-connected impedances, the positive and negative-sequence
circuits are the same in form as that for the star-connected impedances, the impe-
dance value being Z/3 instead of Z as a result of the delta-to-star transformation.
In this case it is the star point of the equivalent-star arrangement which can be
regarded as connected to the zero-potential bar.
Neutral earthing impedances: It will be noted from Table 3.4.3A that the zero-
sequence circuit for a neutral earthing impedance, of value Zn is 3Zn and the
reason for this can be readily understood from Fig. 3.4.3E. At the neutral point, as

El . . . .

b .... ~ ~ n

c .... Zn 3Z n

Zero-sequence circuit

Fig. 3.4,3E Zero-sequencecircuit of a neutral earthing impedance

shown, the zero-sequence currents Io in the three phases combine to give a current
of 3/o in the neutral earthing impedance, the three zero-sequence currents being all
equal in both magnitude and phase. Now the potential Vo of the neutral point with
respect to earth is the same as the potential of the three phase conductors at the
neutral point and hence the zero-sequence voltage at the neutral point, from eqn.
[Link], is
Vo =~z(v. + v. + v.)= v.
But,
Vn = 3IoZn
and because the zero-sequence impedance is the zero-sequence voltage-drop divided
by the per-phase zero-sequence current, it follows that the zero-sequence
impedance of the earthing impedance is given by

3IoZn
Z o =' = 3Z n
Io
and is thus seen to be three times the value of the neutral earthing impedance as
shown in the figure.
Synchronous machines:The positive, negative and zero-sequence circuits for a
synchronous machine (that is a generator, synchronous motor or synchronous
138 Fault calculation

condenser) are shown in Fig. 3.4.3F and require little further comment. The
negative-sequence impedance is usually of roughly the same value as the positive-
sequence sub-transient impedance, the zero-sequence impedance tending to vary
considerably from one design of machine to another but being usually less than the
negative-sequence impedance. All the machine impedances can generally be
regarded, for most practical purposes as pure inductive reactances. It should be
noted that the negative and zero sequence impedances have single fixed values for

ZI

Positive s e q u e n c e

Z2

! | -I

Negative sequence

Zo
nO f-'--I o

Zero sequence

Fig. 3.4.3F Phase-sequence circuits o f a synchronous machine

any given machine (subject to slight variation due to saturation effects) whereas
there are three values of positive-sequence impedance, as already discussed in
Section 3.3.6, namely the subtransient, transient and synchronous values.
Overhead-line and cable circuits: The phase-sequence circuits for an overhead-
line or underground cable circuit can generally be represented by the simple series
circuits shown in Fig. 3.4.3G, these circuits being valid if the [Link]-phase and
phase-to-earth admittances of the circuit are negligibly small. If these shunt ad-
mittances are not negligibly small, the phase-sequence circuits may be represented
by the nominal-rr circuits shown in Fig. 3.4.3H, where Y~ and Yo are, respect-
ively, the positive-sequence and zero-sequence admittances for the total length of
the given circuit.
The value of the zero-sequence impedance Zo depends on the fact that the
 Fault calculation 139

0
0
0

IIIIlllllllllll )llllllllllllll

z I ( = z 2)
P o (" ] oo

Positive and negative

seq u¢ i| ('e

Zo
po |" 1 o Q

Zdr(i $¢q uer~ce

Fig. 3.4.3G Phase-sequence circuits o f an overhead-line or cable circult

Z1

Y1 Y1 Positive and negative

2 ~ sequeflc¢

Zo

Zero $¢quen¢¢

Fig, 3.4.3H Phase-sequence circuits o f an overhead-line; or cable circuit including shunt.

admittances
140 Fault calculation

0
PA ¢ -++ o QA
4~ . . . . . ---- 0

C . . . . . . m ~ _ ~ O

PB ¢ ++ -------- 0 QB
¢ ++ ---- 0

•% , , , , . , , , , , , , "-----,.,,,,,,,,,,,,

Z]A (=Z2A)
PA C .... : : :) QA
Positive and negative
ZIB (=Z2B) sequence
PB C - .... | ; ~ QB

Z0A - Z0M Z0M

PA C ...... "'- i ,, r ~ ~ QA

Z0B - Z0M
PB C+++ ; ..... : - f / ~ i' L " O QB Zero sequence

Transformer

Fig. 3.4,31 [Link] circuits o f a pair o f [Link] or cable circuits with zero.
sequence mutual coupling

return path for the zero-sequence currents is through earth, the term 'earth' here
being meant to refer to the body of the earth together with any circuit earth-wires
or cable-sheaths connected to it. The zero-sequence impedance is generally greater
than the positive (and negative) sequence impedance being usually in the order of
two to three times the positive-sequence value in the case of overhead-line circuits.
The zero-sequence circuits of Figs. 3.4.3G and 3.4.3H assume the overhead-line or
cable circuit to be sufficiently remote from other circuits as to have no electro-
magnetic or electrostatic coupling with such circuits. Where this is not the case, as
with the two circuits of a double-circuit overhead line, the required zero-sequence
circuits (self and mutual shunt-admittances ignored) are as shown in Fig. 3.4.3I, the
 Fault calculation I41

Z0A
PA C ~ ~ - - ~1 QA

-ZoM ~ - Z O M

PB C ~ ; - :) QB
ZOB

Fig. 3.4.3J Alternative form o f zero-sequence circuit for the two c/rcu/tl o f Fig. & & 3 1

circuits shown taking due account of the zero-sequence mutual impedance ZOM,
between the two circuits concerned. The use of the perfect one-to-one ratio
coupling-transformer in the zero-sequence circuit can be avoided if the given circuit
is replaced by the equivalent circuit shown in Fig. 3.4.3J. If the two circuits are
connected to a common busbar at one end, the [Link] circuit is
that shown in Fig. 3.4.3K, Fig. 3.4.3L shows the equivalent zero-sequence circuit
Z0A - Z0M
PA O ~ [
Z0M
1 ; j..--,---,--o QA' QB
ZOB - ZOM [
PB C ; I I

Fig. 3.4.3K Alternative form o f zero-sequence circuit for the two circuits o f Fig. 3,4,31e the
two circuits being connected to a common busbar at end Q

for the case where the two circuits are connected to common bushars at both ends.
Allowance for shunt-admittance effects can be made in a similar way to that
already referred to in connection with single-circuit lines, that is by means of

ZoAZoB - Z20M
ZOA + ZOB - 2ZoM
PA'PB 0 ~ ~, QA'QB

Fig. 3.4.3L Alternative form o f zero-sequence circuit for the two clrcutt~ o f Fig, 3,4,31, the
two circuits being connected to common busban at both ends
 142 Fault caloulation

shunt-connected impedances at the circuit ends, these impedances including circuit-

to-circuit impedances in the zero-sequence circuit to represent the effects of zero-
sequence mutual admittance between the two circuits. Positive and negative
sequence mutual admittance can usually be assumed to be negligibly small.
Transformer circuits: The positive-sequence circuits of transformers have been
considered in some detail in Section 3.3.3 and require tittle further comment other
than to point out that they are equally valid for the negative-sequence network,
there being no difference between the positive and negative-sequence circuits in so
far as static plant is concerned. As pointed out in Section 3.3.5, the positive-
sequence circuit of a transformer is based on the assumption that all the windings
are star-connected, and this assumption applies similarly to the negative-sequence
circuit. The effect of the actual transformer-winding connections (for example star/
delta, star/interconnected-star, etc.) being simply to produce known phase-shifts
of the network currents and voltages. This question of transformer phase-shifts is
considered further in Section 3.4.5.
The zero-sequence circuit of a transformer is dependent on the paths which the
transformer provides for the flow of zero-sequence current, the flow of zero-
sequence currents in one winding requiring a balancing flow of zero-sequence
currents in one or more other windings. Thus, considering the star/delta trans-
former of Fig. 3,4.3D, it will be noted that zero-sequence currents can flow in the
star winding provided the neutral is earthed, the ampere-turns produced by these
currents being balanced by a circulating zero-sequence current in the delta-winding
of the transformer, It is important to note that the delta winding of the transformer
provides no connection to earth for zero-sequence currents flowing in from the
delta-winding terminals. The zero-sequence impedance of the transformer as seen
from the delta-winding terminals is therefore int~mite.
The zero-sequence circuit for the star/delta transformer is thus as shown in
Fig. 3,4,3D, the transformer impedance ZHL being the sum of the star-winding
leakage-impedance Z//per phase, and the referred delta-winding leakage-impedance,
Z~ per phase. For most practical purposes, the impedance ZHL can be assumed to
be the same for zero-sequence conditions as for positive (and negative) sequence
conditions. The impedance ZnL is the phase-sequence impedance in ohms referred
to the nominal rated voltage of winding H and can be referred to any other rated
voltage or converted to per-unit or per-cent value by the methods already described.
It is of interest to note, in Fig. 3.4.3D, that short-circuiting of the delta-winding
terminals is unnecessary in the zero-sequence test, since the delta winding already
provides a closed path for the flow of zero-sequence current. If the delta-winding
were replaced by a second star-winding, however, the zero-sequence test would, of
course, require the neutral and all three phase terminals of this star winding to be
connected together, thus providing the necessary path for the flow of the zero-
sequence currents,
The phase-sequence circuits for a star/delta transformer with an impedance-
earthed neutral are shown in Fig. 3.4.3M, the zero-sequence circuit being readily
derived from Figs. 3.4.3D and 3.4.3E. It will be noted that the earthing impedance
 Fault calculation 143

affects only the zero-sequence circuit, the positive (and negative) sequence circuit
consisting of the transformer impedance only.

Ho

Zn

ZHL
H 0 l ; c, L~

Positive (and negative)

seq uence

ZHL

3Zn Zero seq uonce

Fig. 3.4.3M Phase-sequence circuits of a star/delta transformer with impedance-earthed

neutral

Fig. 3.4.3N shows the [Link] circuits of a star/star transformer, each of

the two neutral points being individually earthed through an impedance. Here
again, the given circuits are referred to the nominal rated voltage of winding H, The
zero-sequence impedance of the circuit is seen to be

g i l l "1"3Znh + 3Zn'1

where Z~r is the neutral impedance Znt referred to winding H and is therefore given
by
7-.'i -- N~Z.l
As before, N is the transformation-ratio of the transformer.
The zero-sequence circuit of Fig. 3.4.3N is, strictly speaking, only valid for the
case of a five-limb transformer or its equivalent (for example, a three,phase unit
144 Fault calculation

~ L

Znh Znl

ZHL
H O , , - . . . . . o u
Positive (and negative)
sequence

ZHL + 3Znh + 3Z~I

HC . . . . . . . . :_ ~ o L'

Zero sequence

Fie. 3.4,3N [Link] circuits of a star~star transformer with impedance-earthed

neutrals

composed of three single-phase transformers). In the case of a three-limb star/star

transformer, the zero-sequence fluxes cause currents to be induced in the sides of
the transformer tank (tel Edith Clarke Vol. II) with the result that the latter acts
as an effective high-impedance delta winding.
The phase-sequence circuits for a star/star auto-transformer are shown in Fig.
3A.3P, the transformer neutral being impedance earthed. Here again, the zero-
sequence circuit is only valid for the case of a five-limb transformer or its equivalent.
Star/star transformers are commonly provided with a delta-connected tertiary
winding, and the phase-sequence circuits for such a transformer are shown in
Fig, 3,4.3Q, the two [Link] neutrals being individually earthed through impe-
dances. The zero-sequence circuit applies equally to the case where the delta-
winding does not exist physically but is an effective winding formed by the trans-
former tank. As before, the given phase-sequence circuits are referred to the nominal
rated voltage of winding H, referred impedance values being indicated by dashed
symbols,
Fig, 3A.3R shows the phase-sequence circuits for a star/star auto-transformer
equipped with a [Link] tertiary winding, the neutral being again impe-
 Fault calculation 145

V o l t a g e r a t i o H tO L -" N

Zn

ZHL
H ~ ~' I ~, L~

Positive (and negative)

sequence

ZHL + 3(N-I)2Z n
H O t : ~ L~

Zero sequence

Fig. 3.4.3P Phase-sequencecircuits of a star/delta auto-transformer with [Link]

neutral

dance earthed. The phase-sequence circuits, as previously, are referred to winding H,

In all the zero-sequence circuits which have been given, the case of solid
earthing of any given neutral is obtained by simply assuming a z0ro value for the
earthing impedance in question. Similarly, in all cases except that given in Fig,
3.4.3R, the representation of an unearthed neutral is obtained by simply assuming
an infinite value (that is an open-circuit) for the earthing impedance concerned, In
the case of the star/star auto-transformer of Fig. 3.4.3R, putting Z n equal to
infinity in the given zero-sequence circuit gives an indeterminate result. For this
case, therefore, the required zero-sequence circuit with the neutral unearthed is as
shown in circuit (a) of Fig. 3.4.3 S. This latter circuit is obtained by transforming
the star circuit of Fig. 3.4.3R into its equivalent delta and then putting Z, equal to
infinity. An alternative form of the zero-sequence circuit is shown in circuit (b) of
Fig. 3.4.3S.
As already stated, it is common practice to assume the [Link]
leakage-impedances to be the same for zero and positive (and negative) soquenoo
conditions. This equality of the sequence-impedances is exact for the special case of
three-phase banks of singie-phase transformers but for the more usual case of throe.
146 Fault calculation

phase transformers it should be noted that differences may exist between the
positive (and negative) and zero-sequence values. These differences are generally
small, however, and are usually ignored in the great majority of fault calculations.

H o T O L

0 O

Znh Znl

zL
0 LI

ZH
HC : :
a
ZT
I ; 0 T#

Positive ( a n d n e g a t i v e ) s e q u e n c e

zl. + 3zl, i
! l 0 L~

Z H + 3Znh

O T)

Zero sequence

Fig, 3.4.3Q Phase-sequencecircuits o f a star/star

In the case of three-winding transformers in which one winding is a permanently

connected delta winding, the measurement of the zero-sequence equivalent circuit
is complicated slightly by the presence of the delta winding. Thus, consider a
transformer with star-connected high-voltage and low-voltage windings and a delta-
 Fault calculation 147

¢
o L

H 0 T

O
0

Zn Voltage ratio H to L = N

zt.
I ; O LI

zH
HO I
z~
I ; O T~

Positive (and negative) s e q u e n c e

Z~. + 3 N ( N - I ) Z n
! .... ," O L~

ZH - 3(N-I)Z n |
H O

0 T#
Z~ + 3NZ n

Zero s e q u e n c e

Fig. 3 , 4 . 3 R Phase-sequence circuits o f a star/star auto-transformer with delta-tertiary

winding

connected tertiary winding and let the common-base zero-sequence equivalent

circuit be the star circuit composed of impedances ZH, ZL and ZT.
Then, let Z A = zero-sequence impedance measured at the terminals of winding
H with winding L an open-circuit
Z B = zero-sequence impedance measured at the terminals of winding
H with winding L short-circuited to neutral
Z C = zero-sequence impedance measured at the terminals of winding
L with winding H on open-circuit
148 Fault calculation

NZ
U"'! ~) LI

H ¢

N2Z 0 T' (a)

I-N

{
0 Lf

Z I:N
H O U"'!

N2 Z~

Fig. 3.4.3S Alternative forms of zero-sequence circuit for the transformer of Fig. 3.4.3R
when the neutral is unearthed

p o..r,"v'Y"..

Zn

Star/delta Interconnected-star

Z0
P 0
Po---f"-:, 1" 3Z n

III I

Positive (and negative) sequence Zero sequence

Fig. 3.4.3T Phase-sequencecircuits of an aarthing transformer

 Fault calculation 149

Z D = zero-sequence impedance measured at the terminals of winding

L with winding H short-circuited to neutral
Assuming all impedances to be to a common base, then

Z~ = ZH +Z r

Z B = Z H + -ZTZL
-
Z T +Z L

ZC = ZL +Z r

ZTZH
ZD=ZL+-
ZT +ZH
from which

ZT =~/ (ZA(Zc- ZD))=~/ (Zc(ZA - ZB))

ZH =ZA - ZT
ZL = Zc - Z r
The same test method is applicable to the case of the star/star auto-transformer
with delta-tertiary winding.
Earthing transformers: The transformers considered so far have all been of the
type required to provide interconnection facilities between network sections and,
as already seen, such transformers commonly provide neutral earthing facilities for
one or more of the network sections concerned. In certain cases, however, a trans-
former may be used solely for the provision of a neutral earthing point where
such a neutral point would not otherwise exist, for example, on the delta side of a
star/delta transformer.
The requirements of a neutral earthing transformer are that it should present a
low impedance to the flow of zero-sequence currents but a high impedance (ideally
infinitely high) to the flow of positive (and negative) sequence currents. Two
commonly used earthing-transformer arrangements are shown in Fig.3.4.3T, namely
the star/delta transformer and the inter-connected-star earthing-reactor. In each
case, -the positive (and negative) sequence impedance can be regarded as infinitely
high for most practical purposes, this impedance being the positive (and negative)
sequence magnetising-impedance. The positive-sequence circuit thus consists of the
unconnected terminal as shown. In the case of the star/delta transformer, the zero-
sequence impedance Zo is the sum of the [Link] and delta-winding lea~ge
impedances as already explained, while in the case of the interconnected-star
earthing reactor it is the sum of the leakage impedances of the two equal half-
windings which constitute each leg of the reactor. It will be noted that the flow of
zero-sequence current through the earthing reactor produces zero magnetising
ampere-turns on each leg of the reactor since the ampere-turns produced by one
150 Fault calculation

Table 3.4.3B Typical per¢ent reactances of synchronous machines at 5O Hz

Positive sequence Negative Zero Short-

Type and rating of machine sequence sequence circuit
Xst Xt X, X2 Xo ratio .d

I 1 kV salient-pole
alternator without dampers 22-0 33.0 110 22.0 6.0 -

11-8 kV 60 MW 75 MVA
turbo-alternator 12-5 17-5 201 13-5 6.7 0.55

11.8 kV 56 MW 70 MVA
gas-turbine turbo-alternator 10.0 14-0 175 13.0 5.0 0-68

11-8 kV 70 MW 87.5 MVA

gas-turbine turbo-alternator 14.0 19.0 195 16.0 7.5 0.5.5

13.8 kV 100 MW 125 MVA

turbo-alternator 20.0 28-0 206 22-4 9.4 0.58

16.0 kV 275 MW 324 MVA

turbo-alternator 16.0 21-5 260 18-0 6.0 0.40

18.5 kV 300 MW 353 MVA

turbo-alternator 19.0 25.5 265 19.0 11.0 0.40

22 kV 500 MW 588 MVA

turbo-alternator 20.5 28-0 255 20.0 6.0- 12.0 0-40

23 kV 660 MW 776 MVA

turbo-alternator 23.0 28.0 207 26.0 15.0 0.50

Table 3.4.3C Typical impedance values for overhead lines at 50 Hz

Sefiesimpedance Shunt suseeptance

Conductor fZ/km /af2-l/kin Charging
Voltage size in current
kV mm 2 al. Zl Zo Zorn ° Zl Be Born* amps/
(and Z2) (and B2) km

33 75 0.36 + 0.56 + - -
/0.41 /1.02
33 100 0.27 + 0.48 + - - _ _ m

/0-39 /'0.98
33 175 0.15 + 0.35 + - - D _ - -

/0.37 /0-97
132 175 0.177 + 0.354 + 0-178 + 2.85 1-66 -0-463 0-22
/0.402 jl.022 /0.509
132 400 0.076 + 0.191 + 0-115 + 3.04 1.73 -0-494 0.24
/'0.379 /0.963 /0-474
274 2 x 175 0.089 + 0.198 + 0.110 + 3.52 2.04 -0.681 0.58
i0.324 /0-854 /'0.462
275 2 x 400 0.038 + 0-147 + 0-109 + 3.59 2-06 -0.682 0.58
/0.320 /0.839 /0-450
400 2 x 400 0.039 + 0.146 + 0.107 + 3-59 2..09 -0.663 0.85
/0.328 /'0.851 /'0.445
400 4 x 400 0.019 + 0.105 + 0.086 + 4.13 2-33 -0-684 0.98
/0-277 i0.790 /0-425
 Fault calculation 151
e~ o. ~ 6 ~ o.
o I I ~. I "'. "', + "'. ""
+ 4- 4" ~0 4- 4-
t,j
",.0 ~.0 ("4 ',.0 ~ "~
o. ~ o. ~ o.
6 o 6 o 6 o
"0
~ =..=. =. ~. ~ . . . + +
-f. 4- 4- 4- 4- + ',0 cO 0 t'~
N •~ o. o. o. o. o. o. o. o.
6 o o o o o o o o
.,= E E E E E ~
~o'~ o o ~ ~ ~ o ~ o ,,, ~
.g
014
g E
~m
0
0 0 0 0 0 0 0 O 0 0
~.~ .,-,, ~ ¢'~ ',,0 hr~ ',,0 ,'~ 0 0 0
&
~4
4
.11.
152 Fault calculation
;> @@
',,0 ',.0
s,--+ ~,-,<
0 I I I I I I I I I I i
0
,,¢ ,,.¢
1 , - ~ 1-,-4
=> @
0
I"'.,.
("4
=>
0,13
.+
= >.
,,o o. ,,?.
:~ ~ ,.o ,,or-. @ @ @ @ @
~ O O 0 0
~ .~ :> :,.o. O..:,0 ,;,
'T " ""
<~. ....+
I I
0 0 0 v'+
Z O0
0
O0
<'9. t":. I I I I
<:
¢z.1
+:
 Fault calculation 153

Table 3.4.3F Typical percent reactances of auto-transformers at 50 Hz

All transformers equipped with delta-tertiary windings

Per-cent reactance
Transformer
rating XHL XHT XLT XH XL XT

120 MVA
275/132 kV 15-0 38.0 18.0 17.5 - 2.5 20.5

180 MVA
275/132 kV 15-0 40.0 20-0 17.5 - 2.5 22.5

240 MVA
275/132 kV 20.0 52.0 25.0 23.5 - 3.5 28-5
240 MVA
400/132 kV 20.0 52.0 30.0 21-0 - 1.0 31.0
500 M V A
400/132 k V 12.0 80.0 52.0 20.0 - 8.0 60-0

750 MVA
400/275 kV 12.0 85.0 60.0 18.5 -6.5 66.5
1000 MVA
400/275 kV 16.0 110.0 84.0 21.0 - 5.0 89.0

half-winding are equal and opposite to those produced by the other.

Plant impedance values:Typical impedance values for synchronous machines,
overhead lines, cable circuits and transformers are given in Tables 3.4.3B, 3.4.3C,
3.4.3D, 3.4.3E and 3.4.3F.

3.4.4 Analysis of short-circuit conditions

The analysis of unbalanced fault conditions has already been shown to involve, in
general, the three phase-sequence networks of the given power system, namely the
positive, negative and zero-sequence networks. It is convenient to represent each
sequence network in its simplest form as viewed from the point of short-circuit, F,
the positive-sequence network being thus represented by a driving voltage E in
series with a positive-sequence impedance ZI, the negative-sequence network by the
negative-sequence impedance Z2 and the zero-sequence network by the zero-
sequence impedance Z0. In other words, the power system as seen from the point
of fault, F, is considered as a single equivalent three-phase generator of terminal
voltage E(reference-phase-to-neutral) and phase-sequence impedancesZl, Z2 andZ0.
These conditions'are shown in Fig. 3.4.4A.
It will be noted that the voltage E is the pre-fault reference-phase voltage at the
point of fault and that Z~, Z2 and Zo are the impedances of the positive, negative
154 Fault calculation

and zero-sequence networks, respectively, as measured from the point of fault.

The symmetrical-component equations representing any given fault condition
are obtained from a knowledge of the voltage and current constraints at the point
of fault, these constraints being expressed in symmetrical component terms. The
standard formulae for a number of short-circuit conditions are derived below, phase
'a' being used as the reference phase. The symbols VI, I/2 and Vo denote the
reference-phase sequence voltages at the point of short-circuit and the symbols 11,
12 and Io the reference-phase sequence currents flowing into the fault.
a

Power system -- O F
-- O

_I_
'w

] Z2 [ ZO
I- l r--"-).--.--.-o ~2 ! ; O 1"0

v
Positive Negative Zero
Fig. 3.4.4A Equivalent phase-sequence circuits o f a power system as seen from the p o i n t o f
fault
(a) Three-phase fault: Considering, first, the three-phase fault clear of earth,
shown in Fig. 3.4.4B, the conditions at the point of fault are

Va = Vb = Vc = V, say [Link]
and
I~ +Ib +1c=o [Link]
Expressing eqn. [Link] in terms of symmetrical components gives

(I~ +/2 +lo)+(a2I~ +a12 +Io)+(aI~ +a212 + I o ) = 0

__[a b
0
Power system
-- [ c ¢

_L
Fig. 3.4.4.B Three-phase fault
 Fault calculation 155

from which
Io=0
because 1 + a + a 2 is equal to zero.
Expressing eqns. [Link] in terms of symmetrical components (with I0 = 0) gives

Va = E - I i Z l - 12Z2 = V ]
Vb = a a E - a ~ I i Z l - aI2Z2 = V [Link]
V~ = a E - a l l Z l - a212Z2 = V
and adding the three equations together we obtain
I
Va+Vb+Vc=O=3V
Hence
Va=Vb= Vc=O

Now, from eqn. [Link] it is seen that

v , - aVb =(a 2 - 1)/2z2 =(I - a)V=O

Hence
/2=O
The required symmetrical-component equations for the [Link] fault are
therefore
E

~q=o 3,4.4.4
io=o

Turning, now, to the three-phase-to-earth fault, the only difference from the
conditions just considered is that the three phase conductors, in addition to being
connected to each other, are also connected to earth. The conditions at the point of
fault are, therefore, now given by

Va = E- IIZ1 - IaZ2 - loZo = 0

Vt, *~a2 E - a2 I i Z l - aI2Z2 - l o Z o = 0 3,4,4.5
V c r. a E - a / l Z 1 - a212Z2 - l o Z o = 0

Adding these three equations together gives

Va + lib + Ve = - 31oZo = 0
from which
Io=0

because 1 + a + a 2 is equal to zero.

156 Fault calculatlo~

Also from these same equations

Va " aYb = (a 2 - 1)I2Z2 = 0

from which
15=0
The required symmetrical-component equations are therefore precisely the same as
those already given by eqns. [Link] for the three-phase fault clear of earth. The
electrical conditions are thus the same for both types of fault, there being zero
current in the earth connection for the three-phase-to-earth fault.
Since the only currents present are positive-sequence currents, the phase currents
flowing into the fault are
E
la =11 = ~
ZI
a=E
Ib =a21 = [Link]

aE
tc =a/i ='~-'-

the sum of the three currents being zero as already indicated.

(b) [Link]-phase fault: It is convenient to assume the phase-to-phase fault to be
between phases b and c, as shown in Fig. 3.4.4C, the conditions at the point of
fault being
la = 0 [Link]

Ib+~=0 [Link]

Vb = Vc = V, say [Link]

a
m II

b
Power O
system
m Co 1
F

Fig. 3,4,4C Phas~to-phasafault

Writing eqns. 3,4,4,7, and [Link] in terms of symmetrical components gives

Ia =Iz +/a +Io = 0

Io +I~ = (a2Ii +a/2 +Io) + (a/l +aaI2 +Io) = 0
 Fault calculation 157

from which
Ib + Ic =- I 1 - 1 2 + 2/o = 0
Hence, it is seen that
~r~+Ib +,r =3Io =0
so that
lo=0
Also, from the equations for Ia and Ib + Ic it is seen that

~r,- (lb +t~) = 2/, +2/2 =0

from which
12 =-11

Expressing eqns. [Link] in symmetrical-component form gives

Vt, = aa E - aaliZl - a12Z2 = V

Ve = a E - aliZl - a212Z2 = V
from which
let, - Vc = (a s - a)E- (a" - a ) l ~ Z l - (a - a2)12Z2 = 0
giving
E - IIZ1 +I2Z2 = 0
But 15 = -11 and hence it is seen that
E-11(ZI + Z2) =0
The required symmetrical-component equations for the [Link] fault are
therefore
E
/1 = ~
ZI + Z2

-E
3,4,4.10
Za +Z2

Zo=O
The phase currents flowing into the fault can now be obtained and are given by

Ia =Ii +I2 =0

-/43e
It, =a2ll +a12 = - - 3,4.4,11
ZI + Z2

/x/3E
Ic=all +a212 =
Z~ +Z2

because a 2 - a is equal to - ix/3.

 158 Fault calculation

a
- a o

b 0
Power
system c
o
F

±
Fig. 3.4.40 [Link]-earth f a u l t

(c) Single-phase-to earth fault: For this fault condition, it is convenient to

assume the short-circuit to be between phase a and earth, as shown in Fig. 3.4.4.D,
the conditions at the point of fault being

Va = 0 [Link]

Ib = Ic = 0 [Link]

Writing eqns [Link] in symmetrical-component form, we obtain

lb =a2lz +a12 +Io = 0

I c = a l z +aal2 +1o = 0

and from these equations it is seen that

Ib - to = (a s - a ) t , + ( a - a ~ ) I 2 = 0
from which
11 =12

Substituting for 12 therefore

lt, + Ic = - Ii -I1 +2/0=0

from which
Io =Ii
Hence
11 =12 =Io

Wdting eqn. 3,4.4.12 in symmetrical-component form gives

ga = E = I z Z l = I2Z2 - IoZo = 0

but because Is and Io are both equal to 11, this can be written

E - I i ( g l + Z= + Z o ) = 0
 ~ult~u~n i~

Hence, the required symmetrical-component equations for the single-phase-to.

earth fault are
E
I1 =/2 = Io = [Link]
Zz + Z2 + Zo
The phase currents flowing into the fault can now be obtained and are given by

3E
Ia =11 +12 +-to =
Zz +Z2 +Zo
3.4,4,15
Ib = a 2 I i +a/2 +Io =0
Ic =aI1 +a212 +Io = 0

(d) Two-phase-to-earth fault: It is convenient to assume the [Link]-to-earth

fault to be between phases b and c and earth, as shown in Fig. 3.4.4E, the eondi.
tions at the point of fault being

=0 3.4,4.16
= ~=0 3.4.4,17

a
u .
0

Power
system

i_
Fig. 3.4.4E Two-phase-to-earth fault

Writing eqn. [Link] in symmetrical-component form gives

Vb = a 2 E - a 2 I z Z l - a/2Z2 - IoZo = 0
V~ = a E - a l l Z t - a212Z2 - IoZo = 0

from which
Vt, - a v e = (1 - a)I2Z2 - (1 - a ) l o Z o = 0
giving
I2Z2
/o =
Zo
Writing eqn. [Link] in terms of symmetrical components gives
Ia =11 +12 +1o = 0
160 Faultcalculation

which, on substitutingfor Io, becomes

(Z2 +Zo)I2
Io=I1 + =0
Zo
Hence
-ZoI1
Za + Zo

-Z 211
Io ~ ~
Z~ +Zo
Now
Vb + Ve = - E + I,Z, +I2Z2- 2IoZo =O
which on substituting for 12 and Io, becomes

i .. +- :ol
E=II ZI Z2+Zo Z2+Zo

from which
Z2Zo ]
E=II ZI +Z2 +Zo
The symmetricM-component equations for the two-phase-to-earth fault are therefore

(Z2 + Zo)E
Z l Z 2 +Z2Z 0 + Z o Z 1
-ZoE
[Link]
I~ =ZIZ2 +Z2Zo +ZoZ~

-Z2E
Io=
ZtZ2 +Z2Zo +ZoZt
The phase currents flowing into the fault can now be obtained and are given by

Ia =I~ +/2 +Io =0

-/V3~'(Zo - a z 2 )
Ib = a211 + al2 +Io =
Z]Z2 +Z2Zo +ZoZI
[Link]
/x/3E(Zo - a 2Z2 )
I¢=aIl +aaI2 +I'o =
ZIZ2 +Z2Zo +ZoZt
 Fault calculation !61

(e) The phase-to-phase plus single-phase-to-earth fault: It is convenient to assume

the [Link]-phase fault to be between phases b and c and the single-pha,w-to-earth
fault between phase a and earth, as shown in Fig. 3.4.4F, the conditions at the
point of fault being

~=0 3,4A.20
= ~=V, say 3.4,4.21
Ib+~=O 3.4.4,22

Fig. 3.4.4F
Power
system

3_
[Link] single-phase-to-earth fault
!
Writing equations [Link] and [Link] in terms of symmetrical components
gives

Va = E - I i Z l " 12Z2 - IoZo = 0

Vb = a a E - a 2 1 1 Z l - oluZa - l o Z o = V
V c = a E - a l i Z 1 - a212Z2 - IoZo = V

Adding these equations together gives

Va + Vb + Vc = -3IoZo = 2V
from which
- 3IoZo
V=
2

Now, from the equations for Vt, and Vc

Vb - a v e = (1 - a)IaZ2 - (1 - a)IoZo = (1 - a)V

giving
I~Z2 - IoZo = V

and substituting for V gives

- 31o Zo
12Za - IoZo =
2
162 Fault calculat/on

Hence
-2/2Z2
Zo
Writing eqn. 3,4,4.22 in terms of symmetrical components gives

Ib +I¢ =(aZll +aIz +lo)+(al, +a2Iz +Io) =0

from which
-11 - 12 + 2/o = 0

Substituting for Io gives

412Z2
-I1 - 1 2 - '=0
Zo
from which
- Ii Zo
4Z2 +Zo
so that
2Zzll
4Za +Zo

Substituting for 12 and Io in the expression for Va gives

Z2Zo 2Z2Zo
I', =E-11 ZI + =0
4Z2+Zo 4Z2+Zo
from which
[ 4ZIZ~ +Z2Zo +ZoZI ]
E - 11 4Z2 +Zo =0

The symmetrical-componentequations for this fault condition are thus

(4Z2 +Zo)E
Il =
4ZIZ2 +Z2Zo +ZoZI

-ZoE
12 [Link]
=4ZlZ2" + Z2Zo + ZoZ~

2Z2E
4Z~Z2 +Z2Zo +ZoZI
 Fault calculation 163

from which the phase currents flowing into the fault are
6Z2E
la =11 +12 +1o -
4ZlZ2 +Z2Zo +ZoZl

-/x/3e(2z2 + Zo)
Ib =a211 +a12 +Io = [Link]
4ZIZ2 +Z2Zo +ZoZI
/\/3E(2Z2 + Zo)
I,=all +a212 +Io =
4ZlZ2 +Z2Zo + ZoZI

(f) Summary of symmetrical-component formulas: The symmetrical components

of fault current for the five short-circuit conditions are summarised in Table
3.4.4B.
The symmetrical components of voltage at the point of fault are obtained from
the equations:
VI = E - I1Z 1
V2 = -I2Z2 [Link]
Vo = -ioZo
the resulting values being given in Table 3.4.4C. The phase-to-earth voltages at the
point of fault are given in Table 3.4.4D.
From the symmetrical components of fault current and fault voltage, given in
Tables 3.4.4A and 3.4.4C, it will be seen that the given short-circuit conditions can
be represented by interconnection of the sequence networks in the manner shown
in Table 3.4.4E.
The analysis of each fault condition is thus seen to reduce to the solution of an
equivalent single-phase network, the latter being composed, in general, of the three
sequence networks. It is important to note that each sequence network represents
the chosen reference phase of the power system, namely phase a, and that the
sequence network currents and voltages are, in consequence, all phase a values.
In the case of the three-phase short-circuit, the negative and zero sequence net-
works are not interconnected with the positive-sequence network and because the
latter network is the only one containing driving voltages, the currents and voltages
in the negative and zero-sequence networks are all zero. In analysing the three-phase
short-circuit condition, therefore, the only network which requires to be considered
is the positive-sequence network. Similarly, in the case of the phase-to-phase short-
circuit condition, the zero-sequence network has no interconnection with the
positive-sequence network and, hence, for this fault condition the only networks
which require to be considered are the positive and negative-sequence networks. It
will be noted that each of the remaining three fault conditions involves a short-
circuit or short-circuits to earth and that in all three cases the analysis requires
consideration of all three sequence networks.
Table 3.4.4F shows the phase-sequence components of fault current, in vector-
164 Fault calculation
o 0
4" 4"
0 0
0
¢4
4"
~q
o 0
~q
4" 4"
..9 o , + 0
~q
4" 4"
0
0
+
~a
~I~ +. o
~q
¢4
4"
~q
~q
ot
f-%
~.
e~
~o
c~
 Fault calculation 165
0
% +
4" 0
~qo
~÷
o
÷
o
o t~
÷
+
o
0 0
+
÷
to
y
~ ~ ~.~
166 F a u l t calculation
o o o
~q ~q
4- -p +
o
+ e~
I ~q
-F
~q
0
..,,,. o o
~q ~q
-F -F
~q
+
0
o eJ~
¢) ~q ~q
~q -p +
'.h. -F
"F -p
~q
o~
?
?
J3
o ~ ~ o
o
o b J=
6
6
p, E-
 Fault calculation 167
o
~q
.p
+ +
o
÷
÷
I
o
! o
~q ~q
+
+
÷ ÷ o
+
I
!
¢J
°~
o
~q
+
o ÷ 0 0
~q ~q
+
N
9
o=
9
9
o~ 9
9
9 J=
o, .0
o, c~
9 0
0
m
t- o=
O == o ¢0
o
~J
6
168 Fault calculation

Table 3.4.4E Intarconnection o f the sequence networks for the representation o f system
short¢ircuit conditions

Fault condition Positive Negative Zero

Three-phase
(a-h-c or a-h-c-~)

Phase-to-phase
(b-c)

Single-phase-to-earth
(a~)

Two-phase-to-eart h
(b-c-e)

1:2
Phase-to-phase plus
Single-phase-to-earth
(b-c plus a-e)

*Closed for three-phase-to-earth fault.

diagram form, for the five short-circuit conditions considered, the impedances Zs,
Z2 and Zo of the power system, as seen from the point of fault, being assumed to
be all equal. The three impedances are all assumed to have an X/R ratio of 5, corres-
ponding to a phase angle of 78* 42'. The table also shows the phase values of the
fault currents, and it can be readily checked from the vector diagrams that each
phase-current vector is the sum of its associated sequence-current vectors. In
addition to the current vectors, the table also gives the vector diagrams of the
phase-to-earth voltages at the point of fault, the pre-fault phase voltages at the
point of fault, Ea(=E),Ez, and Ec, being shown by broken lines.
It is important to note that the vector diagrams of Table 3.4.4F refer specifically
to the stated case, namely that in which the three phase-sequence impedances are
all equal in both magnitude and phase. It will be noted that, for this case, the
voltage to earth of any unfaulted phase or phases at the point of fault is equal to
the pre-fault voltage of the phase or phases in question. This is not the case, in
general, when the sequence impedances are not equal.

3.4.5 Effect of fault impedance

The analysis of short-circuit conditions, as presented in the previous Section, has

assumed the short-circuiting connections to be of zero impedance. This assump-
tion is sufficiently accurate for the great majority of fault calculations but it is,
nevertheless, sometimes necessary to take due account of fault-path impedance,
 Fault calculation 169
0
m I o
1
I °
,,7
H
~ o.~o ~ o
I Ol
0
2 ~
c~
J
N
Ii
J~
N N 3
o!11 ~ ~o o~ m
~d
~ z
u
,p-
",3
J=
N .c
0
0
Y
~ Y
&
170 Fault calculation

~-- a P

Power b

.~- 'T

3Z e
F'I0 zp
FIQ I
Positive Zero I J,
T l
Fig. 3.4.5A Three-phase fault with fault impedance

a particular instance being the consideration of fault-arc resistance in its effect on

distance protection.
The effect of fault impedance is most readily understood by considering the
phase-sequence impedances which must be added, externally, to the system
sequence networks when the latter are interconnected to represent the short-circuit
condition in question. Thus, considering the three-phase fault shown in Fig. 3.4.5A,
let the fault-path impedances be Zp in each phase, together with Ze between the
star-point of these three impedances and earth. Considering the position F' as the
true short-circuit position, rather than F, it will be evident that the impedances Zp
simply appear as external impedances Zp connected to the fault points FI, F2 and
Fo of the three sequence networks. Similarly, the impedance Ze, as seen from F',
appears as an external impedance 3Ze, connected to the zero-potential (earth) bar
of the zero-sequence network, as shown in the figure. Now, it will be noted that at
F', the three phases are connected together through zero impedance and that this
star-point is connected through zero-impedance to the terminal of the impedance
Z e. It follows, therefore, that the sequence-network connections appropriate to
this fault condition are as shown in the figure, the points FI' and F2' being connected
to the zero-potential bar of the positive and negative-sequence networks, respect-
ively, and the point F0' being connected to the terminal of the impedance 3Ze
associated with the zero-sequence network. As already pointed out in the previous
Section, the inclusion of the negative and zero-sequence networks in the representa-
tion of the three-phase short-circuit is only of academic interest, since the currents
and voltages present in the system are all positive-sequence values.
Considering, now, the phase-to-phase fault shown in Fig. 3.4.5B, let the fault-
path impedance between the faulted phases b and c be 2Zt, and let this be repre-
sented by an impedance Zt, in each of the two phases. Now, there is no flow of
 Fault calculat/on t71

Zp
a
Zp
b
Power ---
system

I_

~ Zp Zp Zp

"1
Positive
I "°
Negative
l F°'
. Zero

T r
Fig. 3.4.5B Phase-to-phase fault with fault impedance

current from phase a into the fault, and hence the electrical conditions will be quite
unaltered by the insertion of a fictitious impedance Zp in series with phase a, as
shown. Each phase now contains an impedance Zp between the fault position F
and the position F', and at this latter position, it will be noted, the phase b and c
are connected together through zero impedance. The fault condition, now
considered as being at F', can therefore be analysed by applying the results of
Section 3.4.4 with the three sequence networks augmented by the added
impedance Zp connected to the fault points F~, F 2 and Fo, as shown in the figure.
The remaining fault conditions, namely the single-phase-to-earth fault, the two-
Zp

Power ---
system Zp

Zp Zp

[ ':'°1
Positive
FT°I
Negative
I ~o6
Zero
~ I
Fig. 3.4.5C Single-phase-to-earth fault with fault impedance
172 Fault calculation

Zp
I -

Z
Power --
system Zp

I I I t Ii " ~

'T: ..... !
Fig. 3,4.8D [Link]-earth fault with fault impedance

[Link]-earth fault and the phase-to-phase plus single-phase-to-earth fault, are

treated in a similar manner, the resulting sequence-network interconnections being
given in Figs, 3.4.5C, 3.4.5D and 3.4.5E.
The treatment of fault impedance described above is based on the assumption
that the fault-path impedances are known fixed values. This assumption is accept-
able for many purposes but it is important to note that in a particularly important
case, namely that of a fault arc, the fault-path impedance (normally regarded as a
pure resistance) is a function of the fault current and the arc length. Warrington

Zp Ze

Power
system ¢

F" FI
. . . . . . A

i
1:2

v26

Fig, 3,4,5E Phase,to-phaseplus slngle-phase-to-earth fault with fault impedance

 Fault ~alculation 173

o\
0
0

.=_ ~

P
<
o
o
o

0
I00 I000 I0000 IOOOO0
Arc current in r.m.s, amps

Fig. 3.4.5F A . C arc voltage as a function o f arc current

expresses the characteristic of such a.c. arcs, in air, by the formula

8750L
V = ~
/0.4

where V is the arc voltage-drop in r.m.s, volts, L is the length of the arc in foot and
I is the arc current in r.m.s, amps. This arc characteristic is represented graphically
in Fig. 3.4.5F in terms of voltage-drop per metre of arc length as a function of arc
current in amps.

3.4.6 Analysis of open-circuit conditions

The open-circuit conditions normally of interest in power-system analysis are the

single-phase open-circuit and the two-phase open-circuit, the remaining possibility,
namely the three-phase open-circuit, being of little practical sigrdficanoe, The
analysis of these conditions is achieved by the use of symmetrical components In a
similar manner to that employed for short-circuit conditions, that rtS by eonsid0rtng
 174 Fault calculation

the voltage and current constraints produced by the fault condition at the point of
fault,
(a) Single-phase open-circuit: It is convenient to assume the single-phase open-
circuit to be in phase a of the given circuit, as shown in Fig. 3.4.6A, the positions P
and Q denoting the points in the circuit between which the open-circuit is assumed
to have occurred, The positions P and Q are assumed to be so close together that
the impedances of the healthy phases between P and Q can be assumed to be zero.
P Q
Ia
>

b Ib
>

¢ Ic
>

T A 1

Positive ! [Link] i
Fig. 3 , 4 . 6 A [Link] open-circuit condition

Writing down the voltage and current constraints at the point of fault gives the
equations

(v~)eQ = (v,)p- (v~)Q ~ o

(Vb)pQ -- (Vb)e - (V~)Q -- 0 [Link]
(Vc)pQ = (Vc)e - (V~)Q -- 0

'a=0 /
Ib:~o [Link]
lc~O

The symbols (Va)eQ , (Vb)eQ and (Ve)eQ, it will be noted, represent the series
[Link] along the phase conductors from P to Q.

The phase sequence components of the voltage-drops are given by

(VI)pQ = ~ [(Va)pQ + a(Vb)pQ + a2(Vc)l,Q] = ](Va)~, Q

(V2)eO = ~ [(Zo)eQ + a2 (Vb)ee + a(Vc)eO] = A(V,)eO [Link]
(VO)eQ = '3 [(Vo)ee + (Vb)eO + (Vc)eO] = ~(V,)eO
 ~ult~uD~n 1~

because (Vb)pQ and (Vc)pQ are both zero. The positive, negative and zero-sequence
voltage-drops between P and Q are thus seen to be all equal.
Considering, now, the phase-sequence components of the currents, because
la = 0 it follows that

It = ~(Ia +aIo +a2Ie) = ~(alb +a2Ic)

I2 = ~(Io +a2Ib +ale) = ~02Ib +aZc) 3,4.6.4
Io = ~(Ia + Ib + I~) = ~(I~ + Ic)
It is thus seen that

It +12 +Io = 0 3.4,6.5

because 1 + a + a 2 is equal to zero. and hence
12 +Io = -It [Link]

The constraints imposed by the fault condition, as expressed by eqns. 3,4.6.3 and
[Link], are seen to be satisfied by the interconnection of the sequence networks
shown in Fig. 3.4.6A. It should be noted that, in each sequence network, the
branch representing the given crcuit is open-circuited between the points corres-
ponding to the positions P and Q in the given circuit, that is the points Pt and Qt,
P2 and Q2 and Po and Qo.
Now let Zt, Z~ and Zo be the impedances of the positive, negative and zero-
sequence networks as measured between Pt and Qt, P~ and Q2 and Po and Q0,
respectively, in Fig. 3.4.6A. Then, the positive-sequence network, as seen from the
points Pt and Qt, reduces to a positive-sequence voltage in series with the impe-
dance Z t , the positive-sequence voltage being given (by Thevenin's Theorem) by
Ztltp1,, where It p.r denotes the pre-fault value of the positive,sequence current in
phase a of the given circuit.
From Fig. 3.4.6A, it follows that the required phase-sequence currents in the
faulted circuit are given by

Zllxp/
I 1 =
Z2Zo
Z~ +
Z2 + Zo

-Zol~
3.4.6,7
Z2 + Zo

-ZII !
/o =
Z2 + Zo
176 Fault calculation

from which the phase currents in the faulted circuit can be shown to be given by

~=o

[, (a2 - I)Z 2 + (a 2 - a)Z o

It, 1 Ill
Z of [Link]

Ie = ZIZ2 +Z2Zo ÷ZoZ~

It will be noted that the sequence currents and phase currents are zero if the pre-
fault condition is one n which the circuit in question is carrying zero current.
(b) [Link] open-circuit:For this case it is convenient to assume the open-
circuits to be in phases b and e of the given circuit, as shown in Fig. 3.4.6B.

0
Ia
>

Ib
)

C |C
)

__1L_
, P,~",I v~, ,
[2,L
Fig, 3.4,6B Two~ohaseopen-circuit condition

Proceeding as in the previous case, the voltage and current constraints at the
point of fault are

(re)co = (re)e- ( v , ) e = o I
(Vb)pQ = (lib) P - ( V b ) Q ~ 0 [Link]
(v,)eQ = (v~)e- (v~) o ~ o

,,+° /
Ib = 0 [Link]
I~=0
 Fault calculation 177

As before, the symbols (Va)pQ, (Vv)pQ and (Vc)~ represent the series voltage.
drops along the phase conductors from P to Q.
The phase-sequence components of the voltage-drops are given by

(Vt)efl = ~ [(Va)pfl + a(Vb ) efl + a2 ( V c ) ~ ] = ~ [a(Vb)pfl + a2 (Vc)~ ]

(V2)pQ = ~ [(Va)pQ +a2(Vt,)pQ + a ( V c ) ~ ] = ~ [a2(Vb)efl +a(Vc)eQ]. 3,4.6.11
(Vo)~,Q = i [(vo)~ + (v~)pQ + ( v ~ ) ~ ] = ~ [ ( v ~ ) ~ + (vc)m]

because (Va)eQ= 0. From these equations it is seen that:

(VI)p~ + (V2~Q + (Vo)eQ = 0 3,4.6.12

because 1 + a + a 2 in equal to zero.

Considering, now, the phase-sequence components of the currents it follows that

Iz = ~(la + all, + a2Ic) = ~ Ia

•q = ~(I. +a2Ib +ale) = gIa [Link]
.Io = [(S. + I~ + Zc)= [Zo

because It, and Ic are both zero. The positive, negative and [Link] currents
in the faulted circuit are thus seen to be equal.
The constraints imposed by the fault condition as expressed by eqns, 3,4.6.12
and [Link], are seen to be satisfied by the interconnection of the sequence
networks shown in Fig. 3.4.6B. As before, the branch representing the given circuit
in each sequence network is open-circuited between the points corresponding to the
positions P and Q in the given circuit, that is the points Pt and Qi, P2 and Q2 and
Po and Qo.
Proceeding as in the previous case, it follows from Fig. 3.4.6B that the required
phase-sequence currents in the faulted circuit are given by
Zlllpl"
It =/2 =Io = [Link]
= Zt +Z2 +Zo
from which the phase currents in the faulted circuit are seen to be given by

3Zlltpl"
Io=
Zt +Z2 +Zo

I b =0 [Link]
I,=0

It will again be noted that the sequence currents and phase currents are zero if
the pre-fault condition is one in which the circuit in question is carrying zero
current.
178 Fault caloulation

3.4.7 Transformer phase-shifts

The representation of a given power system by its positive, negative and zero-
sequence networks, all of which represent the chosen reference phase, namely phase
a, implies that the phase references a, b and c can be regarded as applying to the
whole of the power system, even though the latter may consist of a number of
network sections interconnected through transformers. This continuity of the phase
references through the transformer interconnections results from the assumption,
for the positive and negative-sequence networks, that all the transformer windings
are star-connected, this assumption having been the basis for the derivation of the
transformer positive and negative-sequence circuits obtained in Section 3.3.3. Since
not all transformer windings are star-connected, the phase references a, b and c,
which denote actual phase conductors in a given network section (for example
that containing th0 point of fault), must be regarded as denoting hypothetical phase
conductors in the remaining network sections. It is therefore necessary, for
complete analysis, to be able to obtain the sequence currents and voltages appli-
cable to the actual phase conductors in any given network section, given the corres-
ponding reference-phase (phase-a) values obtained from the sequence networks and
taking due account of the actual transformer-winding connections employed in the
power system,
The effect of transformer-winding connections has already been discussed in
Section 3.3.5 for the particular case of the system positive-sequence network, the
assumption of star,connected transformer windings having been shown to give valid
results provided that the positive-sequence currents and voltages obtained from the
sequence network (that is the phase-a values) are suitably corrected in phase, as
necessary, to take duo account of the actual transformer-winding connections. This
reasoning is similarly applicable in the case of the negative-sequence network, the
negative-sequence currents and voltages obtained from the negative sequence net-
work (that is the phase-a values) being also required to be suitably corrected in
phase, as necessary, to take due account of the actual transformer-winding connec-
tions.
Insofar as negative-sequence quantities are concerned, the phase-shift introduced
by a particular star/delta transformer is illustrated in Fig. 3.4.7A, the transformer in
question being identical to that given in Fig. 3.3.5B. It will be noted that the delta-
side negative-sequence currents and no-load voltages all lead their corresponding
star-side values by 30 °. It has already been shown in Fig. 3.3.5B, however, that the
delta-side positive-sequence currents and no-load voltages all lag their corresponding
star-side values by 30*. The phase-shift produced is thus seen to be the same in
mal~rtltude for both positive and negative sequence quantities, the direction of the
phase-shift in the ease of the negative-sequence vectors being the reverse of that
applicable In the case of the positive-sequence vectors. Although illustrated for a
particular ease, this relationship between the positive and negative-sequence phase-
shifts Is true in general, the two phase-shifts being always equal in magnitude and
opposite in direction. The magnitude of the phase-shifts and their respective direc-
 Fault calculation 179

tions are, of course, dependent on the particular transformer-winding connections

and the allocation of the phase-references.

------ > Y ~ y

vy

v y ~'... ..f~c~3oo
--~v R ~ ._L
v/ vb
No-load voltages

Iy

|._

,/ Currents
/il °

(~\~
Ib
Fig. 3.4.7A Phase-shifting effect o f a star/delta transformer on negative-sequence currents
and voltages

Considering, again, the star/delta transformer of Fig. 3.4.7A, let the phase
references, a, b and e correspond to the given star-side phase-references R, Y and B,
respectively, and let 11 and 12 denote the phase-a positive and negative-sequence
~urrents at any point in the star-side network section. Then, the corresponding
180 Fault calculation

sequence currents in the actual phases R, Y and B at the point in question are given
by:

IRI =Ii I.~s =12

I y I = a211 IY2 = als [Link]
IBI =all IB, =a212
with a similar set of equations relating the sequence voltages.
At any point in the delta-side network section, however, the positive and
negative-sequence currents in the actual phases r, y and b must be obtained by
applying the appropriate phase correction to the phase-a sequence currents 11 and
/2 at the point in question, giving

Iri = 1[- 30 ° I1 I n = 1 [+30 °/2

Iyt = 1[- 30 ° a21t Iy2 = l/+ 3o ° at2 3A.7.2
Ib i = 1 [- 30 ° a/l G2 = 1/+ 30 ° a2/2

with, again, a similar set of equations relating the sequence voltages. The vector
multipliers 1[- 30 ° and 1[+ 30 ° represent the positive and negative-sequence
phase-shift correction which must be applied to take due account of the
transformer-winding connections.
The effect of transformer phase-shift is illustrated in Fig. 3.4.7B for the case of
the given star/delta transformer, the transformer being assumed to be supplying
current to a red-phase earth-fault on the star-winding side, the yellow and blue
phase currents on that side being both zero. The positive, negative and zero-sequence
currents on the star-winding side are as shown (per-unit values being assumed) and,
since the transformer is identical to that already considered in Figs. 3.3.5B and
3.4.7A, it follows that the delta-side positive-sequence currents are identical (in per-
unit value) to the corresponding star-side values, but retarded in phase by 30 ° ,
while the delta-side negative-sequence currents are identical to the corresponding
star-side values but advanced in phase by 30 ° . The star-side and delta-side sequence
currents are thus as shown in Fig. 3.4.7B, there being no zero-sequence currents on
the delta-side of the transformer.
The phase currents on the two sides of the transformer are obtained by vector
summation of the sequence currents, the results being as shown with current in the
red and yellow phases only on the delta side of the transformer. These delta-side
phase currents are seen to be equal in magnitude and 180 ° apart in phase, the
magnitude (in per-unit value) being 1/~/3 times that of the red-phase current on the
star-side of the transformer. The validity of this phase-current distribution is readily
verified by considering the flow of current in the transformer phase-windings, as
shown by the arrows.
The question of transformer-winding connections has so far been considered
solely in relation to the phase-shift corrections which must be applied to the
quantities obtained from the positive and negative sequence networks. In fact, the
 Fault calculation 181

positive and negative-sequence networks are the only ones in which phase-shift
correction is required, the zero-sequence quantities requiring no such correction.
This is due to the fact that the flow of zero-sequence current from one network
section to another can only occur if each of the associated transformer windings is
connected in star or inter-connected-star with both the neutral points earthed, such
a winding arrangement producing no [Link] of the zero-sequence currents on
one side of the transformer with respect to those on ~ e other. In the case of the
star/delta transformer already referred to, the transformer-winding connections
limit the zero-sequence currents to the network-section connected to the star
winding and hence, since there is no flow of zero-sequence current through the
transformer, the question of phase-shift correction does not arise.

IR

L.:'c"c-'c'x

?
I
L __..J
IBI iyl
I y I ~ ~ l b !

0 POS. g/ POS.
IIRI Irl
ZERO

IROIyoIBo
IB2~ NEG.
Ib2
ly2

NEG.
klr2
IR2
0 ly

Iy =IB=O Ib=O

|R Ir
Fig. 3.4.7B Effect o f transformer phase-shifts on fault-current distribution
182 Fault calculation

3.4.8 Fault-calculation procedure

The analysis of a power-system fault condition consists, in general, of the following

steps:
(a) The representation of the given power system by its positive, negative and
zero-sequence networks, as required for the particular fault condition under
consideration, the zero-sequence network being omitted for faults which do
not involve earth, and both the negative and zero-sequence networks being
omitted for the balanced three-phase fault condition. This representation is
achieved by the adoption of common-base values, that is, amps, volts and
ohms to a common base-voltage or [Link] or per-cent values to a common
MVA base, the latter being the more usual.
(b) The reduction of each of the sequence networks concerned to its simplest
form, namely a single branch, and the use of the branch values so obtained in
the appropriate symmetrical-component equations to give the phase-a
sequence currents at the point of fault.
(c) The use of the process of back-substitution to obtain the distribution of the
phase-a sequence currents in their respective sequence networks, the resulting
positive and negative-sequence voltages and currents in the sequence networks
being phase-corrected, as necessary, to give the values appropriate to the
actual phase conductors in each of the network sections concerned.
(d) The summation of the phase-sequence currents and voltages to give the
resultant values of the phase currents and voltages at any desired points in the
power system.
The procedure outlined above will provide a complete analysis of the given
power system for the specified fault condition, but in many cases the information
required is limited to the conditions at the point of fault or in the network section
containing the point of fault. In the latter-mentioned cases, the network analysis
is obviously very much less than would be required for a complete analysis.

3.4.9 Example 3

The generator/transformer unit shown in Fig. 3.4.9A develops a short-circuit on the

h.v. terminals of the step-up transformer, the generator and transformer impedances
having the values indicated. Determine the short-circuit currents at the point of
fault, assuming the fault condition to be
(a) a three-phase fault,
(b) a phase-to-phase fault,
(c) a single-phase-to-earth fault.
It can be assumed that the pre-fault voltage at the point of fault is 70 kV phase-
to-phase and that the generator/transformer unit is not connected to the rest of the
power system.

Solution: Adopting a common base MVA of 100, the common-base values of the
 Fault calculation 183

1 1.8/66 kV
7S M V A
jlo.o%
11.8 k V
~F
60 MW 75 MVA

= 58 + jO~
z 2 = j] 3.s% ~J-"

..L
Fig. 3.4.9A

generator impedances are given by

]0-175 x 100
Zgl = 75 =]0.234 p.u.

]0.135 x 100
Zg2 = =i0.180 p.u.
75
and the transformer impedances are given by

]0.i00 x I00
Z n = Zt2 = Zto = = ]0.133 p.u.
75

The zero-sequence impedance of the neutral earthing resistor is given by

Zn o = 3Zn = 174 + j0 ~2

which must be converted to per-unit value to the common base MVA of 100. Now,
the rated current corresponding to 100 MVA at 66 kV is given by
100 x l0 s
= 874A
~,/3"x 66 000
Also, the phase to neutral voltage corresponding to 66 kV is

66 000
= 38 100V

and hence, the base impedance corresponding to the base MVA of 100 at 66 kV is

38 100
= 43-6 I2
874

The required per-unit zero-sequence impedance of the earthing resistor is therefore

184 Fault calculation

given by
174 +/0
Zno = - - = 3.99 +]0 p.u
43.6.

The pre-fauit voltage at the point of fault, F, is stated to be 70 kV and this

expressed in per-unit value is

7O
= 1.060 p.u.
66
The positive, negative and zero-sequence networks for the system are thus as shown
in Fig. 3.4.9B from which it is seen that

E = 1.060 +10 p.u. (the vector reference)

Z1 =/"0-367 p.u.
Z2 =]0.313 p.u.
Zo = 3.99 +]0.133 p.u.
where E is the pre-fauit voltage of phase a at the point of fault and ZI, Z2 and Zo
are the positive, negative and zero-sequence impedances of the system as would be
measured from the point of fault.
These values can now be used in the appropriate symmetrical-component
formulae of Table 3.4.4A as follows:

Case(a) Three-phasefault: For this fault condition the sequence currents at the
point of fault are
E
ZI

I2--o
io=o
from which
1.060 + ]0
I1 = = -/2.89 p.u.
]'o.367

The required phase-current values at the point of fault are thus seen to be given by

Ia = I1 = 0 - j2-89 p.u.
I~ =a~Iz = -2.50 +]1445 p.u.
Ic = a/z = 2.50 +/1-445 p.u.

Multiplying these values by the base current corresponding to 100 MVA at 66 kV

(874 A) gives the required phase currents in amps, the resultant values, expressed in
 Fault calculation 185

1.06 + j0 p.u.
j0.234 p.u. j0.133 p.u.
I

Positive-seq uence
ne t w ork

j0.180 p.u. j0.133 p.u.

f---.-] f--.-.} 0 I"2

Negative-sequence
ne t w ork

j0.133 p.u.
O F0

nI I ;
3.99 + j0 p.u.
gero-seq uence
ne t w ork

Fig. 3.4.9B
polar form, being

la =2526/-90 ° A
Ib= 2526/+150" A
Ic= 2526/+30"A

Case (b} Phase-to-phasefault:For this fault condition the sequence currents at

the point of fault are given by
E
Z~ + Z2

-E
Za +Z2

~o=o
from which
1.060 +/0
11 = =-/1.560 p.u.
/O.68O

-/o
- 1 .o6o
12 = =/1.560 p.u
/0.680.
186 Fault calculation

The required phase-current values at the point of fault are therefore

Ia=It +12 =0p.u.

lb = a2Ii +a/2 = ' 2 . 7 0 +/0 p.u.
lc=all +a212 = 2.70 +10 p.u.

Multiplying by 874 to convert to amps, the required values, expressed in polar

form, are

/a=O
I b = 2360/+180" A
1e = 2360/0°A

Case (c) Single-phase-to-earth fault: The sequence currents at the point of fault
for this condition are
E
z~ =~q =Xo =
Z 1 "l-Z2 "{-Z0

1.060 +/0
= 3.99 +/0.81.3 p.u.

= 0-255 - / 0 . 0 5 2 p.u.

The required phase currents at the point of fault are therefore given by

Ia =11 +12 +Io = 0-765 -/0.156 p.u.

Ib =a2Ii +aI2 +Io =0
Ie=all +aaI2 +Io = 0

Multiplying by 874 to convert to amps, the resulting values, expressed in polar

form, are
Ia = 683/-11 ° 31' A
ib=o
ic=o

3.4.10 Example 4

The system shown in Fig. 3.4.10A develops a [Link]-earth fault at the point
F on the 275 kV terminals of the auto-transformer, the plant impedance values
being as indicated. Determine the value of the fault current at F assuming the pre-
fault phase-to-phase voltage at F to be the rated value, namely 275 kV.
 Fault calculation 187

13.8 kV 1 3 . 8 / 1 3 2kV 132/275 kV

100MW 120 M V A 180MVA
125 MVA j12.0% Z L = - j 2 . 9 % , Z H =jlT.S%

JjF

T = j22.7% ~

Fig. 3.4.10A

S o l u t i o n : Adopting a common base MVA of 100, the required common-base

values of the plant impedances are

]0.280 x I00
Zg I = =/0-244 p.u.
125

/0-224 x 100
Zg2 = =/0-179 p.u.
125

/0-120 x 100
Zt - -/0-100 p.u.
120

/0.175 x I00
ZH - =/0.0972 p.u.
180

-j0.029 x I00
ZL = =-]0.0161 p.u.
180

]0.227 x I00
Zr = =/0.126 p.u.
180

Because the pre-fault voltage at F is equal to the rated voltage its per-unit value is
unity. The system positive, negative and zero-sequence networks are thus as shown
in Fig. 3.4.10B from which it will be seen that the resultant sequence impedances,
as seen from the point of fault are

Zl =i0.405 p.u.
Z2 =/0-360 p.u.
Zo =]0-148 p.u.
188 Fault calculation

I + jO p.u.
jO.l O0 -jO.O161 j0.0972
• , o

T ~ Z 1 = j0.405 p.u.

jO. 179 j0.10o -jO.O161 j0.0972

.... •.- I I I ~ F2
T Z 2 = j0.360 p.u.

I............
jO.lO0 -jO.O161 j0.0972
--/-"7 ] 0 F0

Z 0 : j0.148 p.u.

All impedances in [Link] on 100 M V A

Fig, 3,4,10B

The sequence currents flowing into the fault at F are therefore

E
It =/2 =/o =
Z] + Z2 + Zo

I +i0
- p.u.
/0.913

=-/1.095 p.u.
from wh/ch
la "lt +1: +Io = -/3.29 p.u.
the rema/n/ng phase currents lb and Ic being both zero.
 Fault calculation !89

Now the current corresponding to 100 MVA at 275 kV is

100 x 106
= 210 A
~/3 x 275 000

and hence, multiplying the given per-unit value by 210, the resulting value of the
fault current at F is
I a = -/3.29 x 210 = -/689 A
or, expressed in polar form
Ia = 689/-90 ° A

3.4.11 Example 5

The power system shown in Fig. 3.4.11A develops a two-phase-to-earth fault at the
point F on the 132 kV overhead-line transmission system, the pre-fault phase-to.
phase voltage at F being equal to the rated value, namely 132 kV, Determine the
current and voltage conditions at the point of fault and the values of the phase

(; I G2 G3
G3
LOA D LOAD

A 24.2 km

8 132 kV Transmission system

/
16.1 km D 8.05 km
A

LOAD

G4

Fig. 3 . 4 . 1 1 A
190 Fault calculation

currents which flow to the point of fault from the directions of stations B and D.
Determine also the phase-to-earth voltages at station B. It can be assumed that the
system loads are all zero. What is the three-phase fault level at F?
The plant impedance values are given in Table 3.4.11A, the generator positive-
sequence impedances being the transient values

Table 3,4.11A Plant impedance data for Example 5

Per-cent impedance
Plant Rating
unit to MVA rating to 100 MVA base

(31, G2 60 MW 21 =/17.0 ZI =/22.7

75 MVA Z 2 =1"12.0 Z2 =/16.0
O3, G4 30MW ZI =./15.0 Z t =/40.0
37.5 MVA Z2 =]11.5 Z2 =j30.7
TI, T~ 75 MVA /12.0 ./16.0
Ts, 'I'4 37.5 MVA /10.0 /'26.7
Ts, T6 30 MVA /10.0 /'33.3
TT, Te, T9 60 MVA /12.0 /'20-0
O.H. Line - - Zz = 22 = 0.089 +/0.236
per km - - Zo = 0.197 +/'0.589

1,44+j3.79 ~0.72 +jI.89 2.88+j7.58

2.88 + iT.S8 2.16 + j5.68

i ! r-'-'l
c A B

i
°1°
,4 ,4

All impedances are in per-cent on IOO M V A

Fig. 3.4,11B Positive-sequencenetwork

 Fault calculation 191

Solution: Using a c o m m o n MVA base of 100, the required common-base per-

cent impedance values are as given in Table 3.4.11A, the positivej negative and zero-
sequence networks being as shown in Figs. 3.4.11B, 3.4.11C and 3.4.11D. Ignoring
the load impedances (that is assuming them to be infinite) the steps in the network
reduction process are as shown in Figs. 3.4.11E, 3.4.11F and 3.4.11G, for the
positive, negative and zero-sequence networks, respectively.
From the network reduction, the sequence impedances as seen from the point of
fault are
ZI = 1.69 +/16.7 per-cent = 0.0169 +/0.167 p.u.
Z2 = 1.71 +/14.9 per-cent = 0.0171 +/0.149 p.u.
Zo = 1.90 +/'9.25 per-cent = 0.0190 +/0.0925 p.u.

1.44 + j3.79
I| 0.72 + j1.89 2.88 + jT,S8
t '~ ".3 ! I 0 I . . . .

D F2 "

2.88 + j 7 . S 8
I I "~
2.16+j5.68
:
1
C A . IB

o o Q r- , ,.,~[
O
f~

,,[
v

All i m p e d a n c e s are in per-cent on 100 MVA

Fig. 3.4.1 lC Negative-sequencenetwork

3.16 + j9.48
1| 1.58 + j 4 . 7 4 6.32 + j 1 9 . 0
! . e't I I C : i:
D F0

6.32 + j 1 9 . 0 4.74 + j14.2

• iA ; I " ~B
i , I

o.

All i m p e d a n c e s are in per-cent on I 0 0 MVA

Fig. 3.4.11D Zero-sequence network

192 Fault calculation

The [Link] voltage at F is equal to the rated voltage. Hence,

E = 1 +,/0 p.u. (the vector reference)
The sequence currents at the point of fault can now be obtained from the equations
for the two-phase-to-earth fault in Table 3.4.4A, namely

(Z2 + Z o ) E
11 = . . . . . . . . . . . 0.521 - ./4.40p.u.
Z~Z2 +Z2Zo +ZoZ,

-ZoE
= •= - 0.109 +,/1.71 p.u.
Z, Z2 + Z 2 Z o + Z o Z ,

-ZaE
I o ffi = - 0.412 +,/2.69 p.u.
ZIZ~ +Z2Zo +ZoZ,

The phase currents flowing into the fault can now be obtained and are
I a " I i +12 +Io =O
I~ "- a211 +a12 +Io = - 5 . 9 1 +,/3.49 p.u.
Ic =at, +a~I2 +Io =4.68 +./4.58 p.u.
The current corresponding to 100 MVA at 132 kV is
100 x 106
=437 A
X/3 x 132000
Hence, multiplying the per-unit values by 437 to convert to amperes, the resultant
fault currents, in polar form, are
/o=0
Ib = 3000/149 ° 26' A
~/= 2865/44 ° 2 3 ' A
The voltage to earth of phase a at the point of fault is given by
Va = E " l,Z, - 12Z2 - loZo
which on substitutingthe appropriate values gives
Va = 0.780 -./0.0386 p.u.
The [Link] voltage corresponding to the rated phase-to-phase value of
132 kV is

132000
- - - 76300 V
V3
Hence, multiplying the per-unit value of Va by 76 300 the required voltage to earth
of phase a at the point of fault is given by
Ya = 59 500 - ./2940 V
 Fault calculation 193

or, in polar form, by

V, = 59.6/-2 ° 501 k V
Now, by the process of back-substitution', the sequence currents which flow to the

1.44+j3.79 0.72 + j 1 . 8 9 2.88+j'7.S8 2,16+jS,68 2.88+j'/.$8

2.16 + j5.68 2.88 + j7.58 2,16 + j 5 , 6 8 2.88 +J"/.$8

B C B

2.16 +jfl.68 2.88 + j 7 . S 8 FI

°01
I
2.23 ÷ j 6 . $ 8 1.9o+j5.$4
C'
C~

~
1.6q + Jl6,7
I ] -o.,, +jlO., 0.598 + j10.4

)
All impedances are in per-cent on 100 M V A

Fig, 3,4,11E Reduction of the positive-sequence network

194 Fault calculation

fault from the direction of station B are, from Figs. 3.4.11 E, 3.4.11F and 3.4.11G

(4.39 +j12.3X0.521 -/4.40)

11 = p.u.
(4.39 4/12.3) + (4.78 +/,13.1)

(4-34 +/12.2X- 0.109 +/1.71)

I2 = p.u.
(4.34 +/12"2) + (4.74 +/13-1)

(2.80 +/,13.8X-0.412 +/.2.69)

Io -- p.u.
(240 +/'13.8) + (7-05 +]254)

i~44+j3.79 0,72 + j l . 8 9 2.88 +j7.58 2.16 + jS.68 2.88 + j 7 . 5 8

2.88 + j7,58 2.16 + j5.68 2.88 + j 7 . 5 8 2.16 + j 5 . 6 8

, . B C,
• ; I A

2.16 + j $ . 6 8 2.88 + j 7 . 5 8 2.16 + j 5 . 6 8 2.88 + j 7 . 5 8

7~09 +j15.6 c/ 7.09 + ~15.6 /

C ¸

t~

2,16 +j5,68 2.88 +j7.58 F2 F2

co--f,-3 ~ ,r~k--OB
i0 0
,/ - 4
+

1.71 + j 1 4 . 9

All i m p e d a n c e s are m p e r - c e n t o n 100 MVA

Fig. 3.4.11 F Reduction o f the negative-sequence network

 Fault calculation 195

i.58 + j4.74 b'0 6.32 + j i g . 0

o I I

3.16 + j9.48 6.32 + j i g . 0 4.74 + j14.2

P ~B

o.
P4

1.58 +j4.74 F0 6.32 + j 1 9 . 0 1.58 + j4.74 F0 6.32 + j19.0

I I o I ;
3.16 +j9.48 6.32 +j19.0 4.74 +j14.2 20.0 + j42.4 4.74 + j14.2
D B ~B

",.,z~- ~;,~" rh

1.58 +j4.T4 6.32 + j l g . 0 1.58 + j 4 . 7 4 6.32 + j i g . 0

20.0 +j42.4 4.74 +j14.2 ) 117.6 + j 1 5 6 . 3

B

1.58 + j4.74 6.32 + j19.0 1.$8 + j4.74 6.32 + j i 9 . 0

I 11~.6 +jlS6.3 /
I I.22 + j9.04 0.73 + j6.47~B

-0.13g+j0.214

oo
F0 F,0

"0
~
,4 1.90 + j 9 . 2 5

--0.139 + j0.294

A l l impedances are in per-cent on 100 M V A

Fig. 3.4.11G Reduction of the zero-sequence network
196 Fault calculation

from which
I, = 0.259 -]2.13 p.u.
12 = -0.0546 +]0.823 p.u.
Io = -0.187 +]0.930 p.u.
The phase currents flowing from the direction of station B can now be obtained
and are
la =It +12 +Io = 0.0174 - ]0.377 p.u.
Ib = a21z + a12 + Io =-2.84 +]1.31 p.u.
Ic =all +a212 +Io = 2.27 +/1.86 p.u.
Multipyling by 437 to convert to amps, the required phase currents flowing from
the direction of station B are,in polar form
la = 1651-87 ° 22'A
Ib = 1370[155 ° 14' A
Ic = 1280/39 ° 21' A
The phase currents flowing from the direction of station D can now be obtained by
subtraction of the values just obtained from the total phase-current values at the
point of fault. The required current and voltage conditions at the point of fault are
therefore as shown in Fig. 3.4.11H. The figure also gives the vector diagram showing
the total fault currents Ib and Ic, and the phase-a voltage Va in relation to the pre-
fault voltages Ea, E b and E c.
To obtain the phase voltages at station B, the sequence voltage-drops in each
phase between F and B must be added to the corresponding phase voltages at
F, giving the following per-unit voltages at B

Va = (0.780 - ]0.0386) + (0.259 - ]2.13X0.0288 +/0.0758)

+ (-0.0546 +/0.823X0.0288 +]0.0758)
+ (-0.187 +]0.930)(0.0632 +/0.190)

v b = 0 + a2(0.259 -/2.13X0.0288 +]0.0758)

+ a(-0.0546 +/0.823X0.0288 +]0-0758)
+ (-0.187 +/0.930X0.0632 +/0.190)

Vc = 0 + a(0.259 - ]2.13X0.0288 +]0.0758)

+ a 2 (-0.0546 +/0.823X0.0288 +]0-0758)
+ (-0.187 +]0.930X0~)632 +]0.190)
from which
Va = 0.697 -/0.0374 p.u. = 53.3[-3 ° 4' kV
Vb = -0.293 -/0.168 p.u. = 25.81-.150° I0' k V
V c = -0.189 +]0.236 p.u. = 23.1[128 ° 40' k V
Finally, to determine the three-pham faultlevelat F, the scalarvalue of the system
positive-sequence impedance as seen from F is
x/0.01693 + 0.1672 = 0.168 p.u. on I00 M V A
 Fault calculation 197

and hence the scalar value of the fault current from a three-phase fault at F is given
by
E
I= = 5.96 p.u.
0.168
since E = 1.00 p.u.
Multiplying by 437 to convert to amps
I = 2604 A
and hence the required three-phase short-circuit level is
x/r~ 132 x 2.604 MVA = 596 MVA
It should be noted that this result could have been obtained directly from the fact
that the fault MVA for a three-phase short-circuit at rated voltage is equal to the
ratio of the base MVA (that is 100) to the per-unit scalar value of the system
positive-sequence impedance (that is 0.168).

V a = 59.6 / - 2 ° S 0 1 k V

F
165 / 9 2 ° 3 8 ' A a 165 / -8"/°22_°A
> o <

164S / 144°38'A b 1370/lSS°i4.~A

B

,
1588 / 4 8 ° 2 8 a A
>

lfb = 3000 / 1 4 9 ° 2 6 ' A

Ij c 1280 [ 39°2 I ' A
<

lfc = 2865 / 4 4 ° 2 3 ' A

. . . .

3570 [ 9 8 ° 4 0 ° A
,-j,,

~ . - p - ca
Vb = Vc = 0 /"0 Va

I
I
I

Fig. 3.4.11H
198 Fault calculation

3.4.12 Example 6

The power system shown in Fig. 3.4.12A develops a single-phase open-circuit at

the point F on one of the two 80.5 km 132 kV transmission lines connecting
stations A and B, each line having positive and zero sequence impedances, per km,
of 0-089 + j0.236 per-cent and 0.197 +i0-589 per-cent, respectively, on 100/vIVA.
Determine the resulting phase currents in the two 132 kV circuits immediately after
the open-circuiting, given that the load voltage on the 33 kV side of the transformer
at station B is 32 kV. The load can be assumed to be wholly passive and to contain
no rotating plant, its positive and negative-sequence impedances thus being equal.

Solution: Using the load voltage as the vector reference, its value in per-unit is
given by
32
- - + j 0 = 0.970 +/0
33
Adopting 100 MVA as the common base, the scalar value of the load power is
85
---0.85 p.u.
100
and the scalar value of the load current is therefore

0.85
- 0.876 p.u.
0.97
Now, the phase-angle of the load impedance, that is the angle whose cosine is
0.950, is 18" 12', the sine of this angle being 0.312. The complex value of the load
current is therefore
0.876(0.950 - j0.312) -- 0.832 - j0.274 p.u.
and, hence, the complex value of the load impedance is
0.970 +/(3
= 1.051 +j0.346 p.u.
0.832 -/'0.274

13.8 kV 13.8/132kV 132/33 kV 8S MVA

100 MW 125 MVA 90 MVA 0.95 P.b'.lag
125 MVA j12.$% • 80.5 km • j12.0%
~ A 4 8 - 3 k m B ~

Fig. 3.4.12~
 Fault calculation 199

The positive, negative and zero-sequence networks of the power system, open-
circuited at the point of fault, are shown in Fig. 3.4.12B, the plant impedance
values having been all converted to the common base of 100 MVA. Considering the
positive-sequence network, the effective impedance of the network, as would be

7.20 ÷ jl8.95
! 1

0 4.32 +j11.37
o i !
Pl QI

Positive-sequence n e t w o r k

7.20 + jl8.95

jl7.9 jl0.0 ~ . ~ 13.3I IOS.l + j34.6

[" I r--'l-~ Ar------1 B ~
2.88+j7.58 4.32 +j11.37 /
,, O 0 I J
P2 Q2

Negative-seq uence n e t w o r k

I 5.80 + j47.40

A B
6.32 +j18.96 9.48+j28.44
t-'-1 oo I-"'3
P0 Q0

Zero-seq uence n e t w o r k

All i m p e d a n c e s in per-cent on 100 MVA

Fig. 3 . 4 . 1 2 B
200 Faultcalculation
measured between the points PI and Qt, is seen to be

Z~ = (2.88 +/.7.58) + (4.32 +/,11.37)

(7.20 +/,18.95)(105.1 +j80.3)
+ per-cent
(7.20 +/,18.95) + (105.1 +/.80.3)
from which
Z1 = 0.147 +/'0.352 p.u.
Similarly, the impedance of the negative-sequence netwok, as would be measured
between the points 1)2 and Q2, is seen to be

Z2 = (2.88 +/.7.58) + (4.32 +/,11.37)

-t
(7.20 +/,18.95X105.1 +/.75.8)
per-cent
(7.20 +/,,18-95) + (105.1 +/.75.8)
from which
Z2 = 0-148 +/,0.351 p.u.
Finally the impedance of the zero-sequence network, as would be measured
between the points Po and Qo, is seen to be

Zo = (6.32 +/,18.96) + (948 +j28-44)

(15.80 +/.474)(0 +/.23.3)
+ per-cent
(15.80 +j474) + (0 +/.23.3)

from which
Zo = 0.174 +/.0.634 p.u.
Now, the pre-fault value of the positive-sequence current in the circuit containing

Ic
c Ic c

h/lib 'h b

Faulted circuit Healthy circuit

Fig. 3.4.12C
 Fault calculation 201

the point of fault, F, is equal to one half the current taken by the load, giving
Izp! = 0-416 -/0.137 p.u.
Hence, using the values obtained above in eqns. [Link], the required positive,
negative and zero-sequence currents in the faulted circuit are

Zlllp/
11 ffi = 0.2535 - ]0.0884 p.u.
Z2Zo
Z~ +
Z2 + Zo

-Zoll
=
= - 0.1633 +/0.0484 p.u.
Z2 + Zo

-Z2Ia
Io= = - 0.0902 +]0.0400 p.u.
Z2 +Zo
The phase currents in the faulted circuit can now be obtained and are

la =11 +12 +Io = 0 p . u .

Ib = a2Iz +o22 +Io =-0.254-]0.301 p.u.
lefalz + a212 +Io =-0.0167+]0.421 p.u.

the direction of flow being from A to B. The base current corresponding to the
base MVA of 100 at 132 kV is 437 A and, hence, multiplying the above values by
437 to convert to amps, the resulting phase currents, in polar form, are

/a=0
It, = 172/-130010 ' A
I e = 185/92°16 ' A
Now the effective impedance of the faulted circuit to the flow of positive-sequence
current is seen, from Fig. 3.4.6A, to be its normal healthy positive-sequence impe-
dance (7.20 + ]18.95 per-cent) plus the impedance formed by the parallel combina-
tion ofZ~ andZo, that is
(0.148 +]0.351X0.174 +/0.634)
(0.072 +]0.1895) + p.u.
(0.148 +]0.351) + (0.174 +]0.634)

= 0.155 +/0.416 p.u.

With the open-circuit fault present, therefore, the positive-sequence voltage
(reference-pha@ value) of point A with respect to point B is
(0.2535 - ]0.0884X0.155 +]0.416) = 0.0761 +]0.0917 p.u.
202 Fault calculation

The positive-sequence current flowing from A to B in the other, healthy, 132 kV

circuit is therefore
0.0761 +/0-0917
11 = = 0.556 -/0.190 p.u.
0.072 +/0.1895

Also, knowing the negative sequence current flowing in the faulted circuit, the
negative-sequence current flowing from A to B in the healthy circuit is, from
Fig. 3.4.12B,
- (-0.1633 +/0.0484X105.1 +/75.8)
p.u.
(105.1 +/75.8) + (7.20 +/18.95)
from which
I2 = 0.140 -/0.0535 p.u.
Similarly, the zero-sequence current flowing from A to B in the healthy circuit is
-(-0-0902 +/0.0400)(/23.3)
Io =
(15-80 +/47.40) + (/23.3)

from which
Io = 0.0311 -/0.00621 p.u.
The phase currents flowing from A to B in the healthy circuit are therefore
Ia = I~ + 12 + Io = 0.727 -/0-250 p.u.
I b =a211 +a/2 +1 o = - 0 4 3 5 -/'0.245 p.u.
Ic =aI1 +a212 +1o =-0.199 +j0.476 p.u.
Multiplying these values by 437 to convert to amps, the required phase currents in
the healthy circuit ate, in polar form;
la = 336/-18" 59' A
I b = 218].-.150 ° 36' A
I c = 226/112" 41' A
The phase currents in the two circuits are shown in vector-diagram form in Fig.
3.4.12C, the pre-fault values being shown by the broken lines. It will be noted that
the missing phase-a current in the faulted circuit produces an increased phase-a
current in the parallel healthy circuit.
It is important to note that the generator impedance used in the given calcula-
tions is the transient value and that, in consequence, the phase currents obtained
are those appropriate to the conditions obtaining immediately after the instant of
fault.

3.5 Calculation of simultaneous fault conditions

3.5.1 Sequence networks
The fault conditions so far considered have all been single fault conditions occurring
at a stated single point on the power system, the analysis of these conditions
 Fault calculation 203

having been achieved by considering the positive, negative and zero-sequence net-
works of the given power system and reducing these networks to their simplest
form as seen from the point of fault. A similar procedure is used in dealing with
two faults which occur simultaneously at, say, two different points on the power
system, the requirement now being that the three sequence networks must be
considered in relation to the two points of fault, namely A and B, instead of the
single point F considered previously.

a a
0
b b ,0
0 Power ---
C
system ¢
¢ ,0

B!

, ~ ZCI ....

Positive

ZA2 ZB2
A2 ~ l r ' - ' - ' l " ° e2

Zc2

Negative

ZAO ZBO
A0 ~ T r " T " l " ° BO

ZCO

i iii

Zero

Fig. 3.5.1A Equivalent phase-sequence networks for a system subjected to two simultaneous
faults at points A and B
204 Fault calculation

A B
¢ a a
o

.~. b I I b

, $

Fig, 3,5,2A Block diagram representation o f the cross~ountry [Link] condition

The three-phase power system is shown in block-diagram form in Fig. 3.5.1A,

the phase references a, b and c being the same at the two points of fault, A and
B, The corresponding phase-sequence networks, reduced to their simplest form, are
the equivalent-T circuits shown in the figure, each of these networks representing
the chosen reference phase, namely phase a. The use of the two driving voltages
EA and E~ in the positive-sequence network permits the treatment of cases in which
the pre-fault voltages at A and B are different due to the network pro-fault condi-
tions (for example load flows). Where such pro-fault voltage difference is zero or
can be ignored, however, the two driving voltages EA and EB can be removed and
replaced by a single driving voltage in series with Z c t.
The analysis of a simultaneous fault condition is achieved in a similar manner to
that already described for the single fault condition, namely by considering the
current and voltage constraints imposed at the points of fault, these constraints
being used to determine the values of the sequence currents flowing from the power
system into the two points of fault. It is not possible, in the space available, to
consider all the possible combinations of fault conditions and the treatment given is
therefore limited to particular fault conditions with special reference to the cross-
country earth-fault condition.

3,$,2 Cross-country earth-fault

This fault condition is represented, in block-diagram form, in Fig. 3.5.2A, the fault
condition at A being a single-phase-to-earth fault on phase b and that at B a similar
fault, but on phase c.
Let I a 1,/,4 2 and I a o be the phase-a sequence currents flowing into the fault at
A and let IB t, IB2 and IBo be, similarly, the phase-a sequence currents flowing into
the fault at B. Let EA and EB be the pro-fault phase-a voltages at the points A and
B, respectively.
The constraints imposed by the fault condition at A are
(io)A = (zc)A =o [
(Vb)A = 0 / 3.5.2.]
 Fault calculation 205

and writing the current constraints in symmetrical-component form gives

(la)A =IAt +IAa +IAo = 0 [ 3,5,2,2
(Ich =a1A, +a~IA2 +I,4o = 0
from which it follows that
IA 2 -- a/A, [Link]
1.4o =a2Ia]
The fault current at A is thus given by
(Ib)A =aal•, +a/A2 +IAo
from which
(Ib )a = 3a2I A , [Link]
Similarly, the constraints imposed by the fault condition at B are
/

(Ia)B = (Ib)s = 0 [ [Link]

(vc)B = o [
J
and, again, writing the current constraints in symmetrical-component form gives
(Ia)s =IBI +Is2 +Iso = 0 / 3.5,2,6
(Ib)s =a21st +a/B2 +Iso =0
from which it follows that
]
/sols~=a2Ist=a/B, ] 3.5,2,7

The fault current at B is thus given by

(l~)s =alst +a21s2 + Iso
from which
(Ic)s = 3a/s t 3,5,2.8
Considering the voltage constraint at A, it follows from Fig. 3.5,2A that
(Vb)A =a2Ea- a2IAa(gA, + Zca)-alA2(ZAz +Zc2)
-IAo(ZAo + Z¢o)- a2ls,Z~,
-als2Zc2 - lsoZco = 0 3,5,2,9
which reduces to
e~4 =L4,[(ZA, +Za~ +Zao) +(Zc, +Zc~ +Zco)]
+Is1 [Zc, + ~ C 2 +a2Zco] 3.5,2.10
Similarly, considering the voltage constraint at B, it is seen that
(g~)s = ~ s - a/s~(Zs, + Z c ~ ) - a~ls2(Zs2 + Z c 2 )
- lso + (Zso + Z c o ) - aL4~Zct
-aa l A 2Zc2 - 1,4oZco = 0 3.5,2.11
which reduces to
~s =Is, [(zs~ +zs2 +Zso)+(Zc~ +Zc2 +Zco)l
+1.4, [Zc~ +a~Zc2 +aZco] [Link]
206 Fault calculation

Solving the simultaneous equations [Link] and [Link] gives

GEA HEB-

Ia,- [Link]
GJ - HK
JEB - KEA
IBI = [Link]
GJ- HK
where
G= (Zaa +zn2 +ZBo)+(Zc~ +Zc2 +Zco)
H=Zcl +aZc2 +a2Zco
]=(ZA~ +ZA~ +ZAo)+(Zc~ +Zc~ +Zco)
K =Zct +a2Zc2 +aZco
The required fault currents at A and B can now be obtained from eqns. [Link] and
[Link], giving:

[ GE.4.. HEal [Link]

(Ib)a = 3a2 GJ- HK

[ JEB-KEA ]
(I¢)B = 3a [Link]
GJ- HK
The resulting current distribution in the power system is obtained by the process of
back-substitution in the manner already described.
The treatment of other simultaneous fault conditions can be achieved in a
similar manner to that described above, assuming, as in the case treated, that the
two faults both occur on the same network section. The treatment of simultaneous
faults in which the two faults occur on different network sections (that is, are
separated by transformer interconnections) is given by Edith Clarke in the book
referred to in the Bibliography. This book gives a full treatment of a wide range of
simultaneous fault conditions.

3,5.3 Sequence networkinterconnections

The treatment of single fault conditions in Sections 3.4.4, 3.4.5 and 3.4.6 has
shown that any given fault condition may be represented by an appropriate inter-
connection of the sequence networks. Such interconnection of the sequence net-
works earl also be used for the representation of simultaneous fault conditions, the
required lnterconneetion for any given fault condition being that which applies the
appropriate constraints to the sequence currents and voltages at the two points of
fault,
In Interconnecting the sequence networks it is important to note that
(a) the three sequence networks represent the chosen reference phase, namely
phase a of the power system
 Fault calculation 207

(b) the interconnections applied must be such as will satisfy the fault constraints
at the two points of fault, those applied to achieve the necessary constraints
at one point of fault being applied in such a way as not to introduce additional
constraints at the other point of fault.
Fig. 3.5.3A shows the required interconnection of the sequence networks to
represent
(i) a three-phase fault at A plus a three-phase fault at B
(ii) a single-phase-to-earth fault (a--e) at A plus a three-phase fault at B
('di) a phase-to-phase fault ( b - c ) at A plus a three-phase fault at B.

I~
ositiv

Tll t
egati

I I
(

T h r e e - p h a s e f a u l t s at A a n d B
t lA'z.,0

I I "

Phase-a e a r t h - f a u l t at A w i t h t h r e e - p h a s e f a u l t at B

P h a s e - t o - p h a s e f a u l t (b to c) at A w i t h t h r e e - p h a s e f a u l t at B

Fig. 3.5.3A Representation of simultaneous fault conditions by interconnection o f the

sequence networks
208 Fault calculation

,i , ,

I ..... I - - -

1"6Positive 6B,II , Zero

. T

rq
M

Phase-a earth-faults at A and B

,, , , , , i

L I i". I
] ---6 6---6B,1
II I
I
|
P' ~'
Positi ve -
I
I I°:0o It
°.

'fcl -~
Phase-a o p e n - c i r c u i t at A ( P , Q ) w i t h a p h a s e - a e a r t h - f a u l t at B

Fig, 3,6,3B Representation o f simultaneous fault conditions by interconnec6on o f the

sequence networks and the use o f real-retio in terposing transformers

For the reasons already given, the negative and zero-sequence networks are super-
fluous in case (i) since the currents and voltages in both these networks are zero.
Similarly, the [Link] network is superfluous in ease Ctii). It is important to
note in case (ii) that the fault condition depends significantly on whether the
three-phase fault at B is clear of earth (switch open) or to earth (switch closed),
because this will affect the zero-sequence impedance of the system as seen from the
other point of fault, A.
Fig. 3.5.3B shows the required interconnection of the sequence networks to
represent
(iv) a single-phase-to-earth fault (a-e) at A plus a [Link]-earth fault
(a-e) at B
 Fault calculation 209

(v) a single-phase open-circuit in phase a at A (P, Q) plus a single-phase-to-earth

fault (a-e) at B.
It will be noted that in both the latter-mentioned cases the interconneotions
have required the use of perfect 1 : 1 interposing transformers, these being necessary
to avoid the introduction of additional constraints at the points of fault over and
above those necessary to represent the stated fault conditions.
The sequence-network interconnections shown in Figs. 3.5.3A and 3,5,3B result
from applying the known relations between the phase-a sequence currents and
voltages at the points of fault, the interconnections necessary being a combination
of those required to satisfy the two fault conditions considered ind/vlduaUy,
Other simultaneous fault conditions can be similarly represented by appropriate
interconnection of the sequence networks. It is important to note, however, that
the sequence currents and voltages represented in the sequence networks are the
phase-a values, and the sequence network interconnections which must be applied
are those required to satisfy the known relations between the phase-a sequence
quantities at the two points of fault.
In certain cases the required interconnection of the sequence networks will
involve the use of ideal phase-shifting transformers of ratio l:a and 1 :a 2 , where
a = 1/120 °, and this can be illustrated for a particular case, namely that ofa phase-a
earth fault at A plus a phase-b earth-fault at B. At the point A the relationship be-
tween the phase-a sequence currents is
(Z,)A = (Z~).,I = (Zo)A
and at point B the relationship is
a3(I1)8 = a(I~)B = (Io)B
At the point B, therefore, it is required that
(z2)B ~(zl)B
=

(Io)B = a~(,r,)B
It will be noted that these constraints are satisfied by the interconnection of the
sequence networks shown in Fig. 3.53C.

A! c
~o,i,iv~ i
I

rT'L G"]
ee

Fig. 3.5.3C Representation o f a simultaneous fault condition (phase-a earth-fault at A with

phase.b earth-fault at B) by interconnection o f the sequence networks and the use
of complex-ratio interposing transformers
 210 Fault calcu/ation

3.$.4 Example 7

The 33 kV system shown in Fig. 3.5.4A develops a phase-b earth-fault at the point
A and a phase-c earth-fault at the point B, the positive and zero sequence impe-
dances of the 33 kV overhead lines being 0.155 +/'0-367 I2 per km and 0.9354 +
/0.969 ~2 per krn, respectively. Determine the values of the fault currents at A and
B and the value of the current in the neutral-earthing resistor. It can be assumed
that the load currents are zero and that the pre-fault phase-to-phase voltages at two
points of fault are equal to the rated value of 33 kV. It can also be assumed that the
negative-sequence impedance of the 132 kV system is the same as the positive-
sequence value.

Solution; Using a common MVA base of 100, the 132 kV system impedance is
seen to be
100
= 0.0667 p.u. = 6.67 per-cent
1500
the stated X/R ration of 5 corresponding to an impedance phase-angle of 78 ° 41'.
The complex value of the 132 kV system impedance is therefore given by
6.67 (cos 78 ° 41' +/sin 78* 41') = 1.31 +j6.53 per-cent
The impedance of the 60 MVA transformer to the common 100 MVA base is
100 x112.5
=/'20.8 per-cent
60
Using the nomenclature of Fig. 3.5.1A, therefore, it follows that
Zcl =Zc2 = (1.31 +/6.53)+(/20.8)= 1.31 +j27.3 per-cent

132/33 kV ,ta.~ _--L

60 MVA y -
. ~ • Phase-b earth-fault

132 kV
System

Zo = j36.0~ " ~

Fault level = 1500 M V A - - R n = i 5.01~

X / R ratio = 5
Z I =Z 2

%-

Phase-c earth-fault

Fig. $ , 5 , 4 A
 Fault calculation 211

The impedance Zco is the zero-sequence impedance of the earthing transformer plus.
the zero-sequence impedance of the earthing resistor, giving
Zco = 45.0 +j36.0 ~2
Now the base current corresponding to 100 MVA at 33 kV is
100 x 106
= 1750 A
X/3 x 33 000
the rated phase-to-neutral voltage being
33 000
= 19050 V
,/3
The base impedance for the 33 kV system is therefore
19 050
~ = 10.88 ~2
1750
and the common-base value of Zc.o is therefore
45.0 + 136.0
=4-135 +j3-31 p.u.
10.88
giving
Zco = 413.5 +]331.0 per-cent

The common-base values of the line impedances are

0.155 + 0.367
Positive-sequence impedance = p.u. per km
10.88

= 1.42 +]3.37 per-cent per km

0.354 +j0.969
Zero-sequence impedance = p.u. per km
10.88

= 3.25 +/'8.91 per-cent per km

From the given line lengths, therefore, it is seen that
ZA 1 = Z42 = 23.0 +j54.2 per-cent
ZAo = 52.4 +]143.4 per-cent
ZBI = ZB2 = 34.5 +i81-3 per-cent
ZBo = 78-6 +j215.1 per-cent
The positive, negative and zero-sequence networks for the power system are thus as
shown in Fig. 3.5.4B. Combining the impedance values for use in the fault-current
equations gives
G =(ZB~ +Z82 +ZBo)+(Zc~ +Zc2 +Zco)
212 Fault calculation

~vmg
G = 5.64 +j7.63 p.u.
H=Zcl +aZc2 +a2Zco 0.570- j5.09 p.u.=

Jffi(ZAI +ZA2 +ZAo)+(ZcI +Zc2 +Zco)

#~ng
J = 5.15 +j6.37 p.u.
K =Zca +a2Zc2 + aZco = - 4 . 6 9 +j2.05 p.u.

A, o230+j54 34 13 0.,
v fJ'1 v Zb I
Za i Zc I
~ i.31 + j27.3

Positive-seq uence network

23.0 + j54.2 34.5 + j81.3

A20 [ ; ~ [ | o B2
Za2 ~ Zb2
Zc2 1.31 + j27.3

Negative-seq uence network

52.4 + jl43.4 78.6 + j21$.1

AO0 {' ; ~ | | 0 B0

Za0 Z c 0 ~ 413.5 +j331.0 zb0

Zero-seq uence network

EA = E B= 1 +jOp.u.
All impedance values are in per-cent on I00 MVA

Fig. 3.5.48
 Fault calculation 213

Substituting these values into eqns. [Link] and [Link], with E A = E B = I +/(3,
giVeS
(Ib)A = -0.710 -/0.122 p.u.
(Ic)B = 0.511 +/0.?.,40 p.u.
Multiplying by the base current of 1750 A to convert to amps, the required fault
currents, expressed in polar form, are
(Ib)A = 12621-170 ° 15'A
(/c)B = 987/25° 10' g
The current flowing from earth into the neutral earthing resistor is
In = (Ib ).4 + (Ie)B = -0.199 +/0.118 p.u.
which multiplied by 1750 and converted to polar form gives
In = 405/149 ° 20' A
The fault currents are shown in vector-diagram form in Fig. 3.5.4C.

Ec

\
\
\\\ 7 (Ic)B
In ~ , /

Ob)A /
/
/
/
/

Fig. 3.5.4C

3.5.5 Example 8

The power system shown in Fig. 3.5.5A develops a single-phase (phase a) earth-
fault at the point F on the h.v. terminals of the 132/33 kV transformer, together,
with a phase-a open-circuit immediately on the line side of the earth fault. Deter-
mine the value of the fault current at F and compare this with the value obtained
with the open-circuit absent. The line can be assumed to have a positive-sequence
impedance of 0.089 +/0.236 per-cent per km and a zero-sequence impedance of
0.197 +/0.589 per-cent per km on a 100 MVA base. It can be assumed that the
system load is zero.
 214 Fault calculation

132/33 kV
90 MVA
j16.0%
32.2 km
132 kV System -----

F a u l t level = 2 5 0 0 M V A
X/R ratio = 6
Z I --Z 2
Z 0 = 0.8Z 1

Fig. 3.5.5A

Solution: Adopting a common base MVA of 100, the sequence impedances of

the 132 kV system are
100
ZI =Z2 = 2 ~ ( c o s 80 ° 32' + l sin 80 ° 32') p.u.

= 0.00658 +/'0.0395 p.u.

Zo = 0.8Zl -- 0.00526 +i0.0316 p.u.
and those of the 132 kV line are

Z1 = Z2 = 0.0288 +i0.0758 p.u.

Z0 = 0.0632 +/0.190 p.u.
The common-base impedance of the transformer is
100 x j16.0
Zt = = l l 7.8 per-cent
90
=/'0.178 per-unit
The positive, negative and zero-sequence networks thus reduce to the circuits
shown in Fig. 3.5.5B, it being assumed that the pre-fault phase-to-phase voltage at
F is equal to the rated value of 132 kV.
Considering, now, the conditions at the point of fault, these are shown in Fig.
3.5.5C, the currents flowing into the transformer consisting of zero-sequence
currents only, I per phase, as shown.
The phase currents which flow in the transmission line towards the point of
fault are thus

Ia=0
lb =I
lc=I
the corresponding sequence currents in the line thus being
Is, = ~(la +a/b +a2Ie) = - ~I
Is2 = "~'(fa + a2 lb + a I c ) = - ½[
 Fault calculation 215

Let the sequence impedances of the power system to the left of the point of fault,
as seen from P, be Zsl, Zs2 and Zso. Then, the phase-to-earth voltages at F are
given by
(V~)F = 0
( V b ) v = (Vb)I" = a a E - a 2 I s l Z , 1 - a l , 2 Z , 2 - I , o Z , o
(Ve)F = (Vc)p = aE- alslZsl - aals2Zs2 - lsoZso

] +jO
0.0354 + j0.115
o o--o
Zsl PI Q! FI

Positive-sequence network

0.0354 + j O . I 15
[ I 0 0---0
Zs2 P2 Q2 F'2

Negative-seq uence network

0.0685 + j0.222 jO.178

I 1 o o--.o. {,--']
Zs0 P0 Q0 F0 Zt

Zero-sequence network

All impedances are in per-unit on 100 MVA

Fig. 3 . 5 . 5 B
216 Fault calculation

and, hence, the zero-sequence voltage at the point F is

(Vo)~ --4A(vo)p + (Vb)~ + (Vc)H
= ~ [ - E + I s ~ Z , l + 1 , 2 Z , 2 - 2/,oZ, o]
But
1,1 = 1,5 = - ~I
and
I,o =3I
and hence
(Vo.)F =- ~[3E + l(z,1 + z , + 4Z, o)]
Now, the zero-sequence voltage at F must be equal to the zero-sequence voltage-
drop in the transformer, so that
(Vo)F --Izt
and, hence, substituting for (Vo)F in the preceding equation gives
9IZt = - 3 E - I(Zsl + Z,2 + 4Z, o)
from which - 3E
I=
9Zt + Z,l + Z,2 + 4Zso
The fault current I f which flows from phase a to earth at F is equal to minus l,
hence the required equation for the fault current is
3E
9Z t + Z s l + Zs2 + 4 Z , o
Substituting the values of voltage and impedance for the given example, shown in
Fig. 3.5.5B, gives
3(1 +/0)
I/= p.u.
0.345 +/2.721
The base current corresponding to the base MVA of 100 at 132 kV is 437 A and,

! I I
c--- > I >
I
i I
b - - - > I '.
r
I
o /I !
a--- > ~o-- >
p Q
31
i = -If

21 31

Fig. 3.5.5C
 Fault calculation 217

hence, multiplying the per-unit value of the fault current by 437 to convert to
amps, the required fault current, expressed in polar form, is
I / = 479/-82 ° 43'A
With the open-circuit absent, the value of the fault current is
3//
Zx + Z2 +Zo
where Zz, Z2 and Zo are the sequence impedances of the complete system as seen
from the point of fault F. With the open-circuit removed in the sequence networks
of Fig. 3.5.5B, therefore, it is seen that the above-mentioned impedances values are
Zl =Zal =0~)354+/D.115 p.u.
Z2 =Zs2 = 0.0354 +10.115 p.u.
Z,oZt
Zo = ~ = 0.0132 +/0.101 p.u.
Zso +Zt
The value of the fault current with the open-circuit absent is therefore
3 0 +/0)
If ffi p.u.
0.084 +/0.331
giving
I f = 2.16 -/8.52 p.u.
Multiplying by 437 and converting to polar form, the required current is
I/= 3840/-75 ° 47' A
It will be noted that the open-circuit produces a considerable reduction in the value
of the earth-fault current.

3.6 Practical network analysis

It will be evident from the examples already given that network analysis by manual
computation may involve considerable expenditure of time and effort if the
network under examination has many nodes and branches, the method becoming
virtually impracticable if the network is particularly large as may often be the case
in some of the studies required to be carried out for the larger power systems.
Accordingly, recourse may sometimes necessarily have to be made to alternative
methods of analysis to overcome this difficulty, the first, based on the analogue
principle, being the use of a network analyser and the second, and now much more
common method, the use of a digital computer.

3.6.1 Network analysers

The network analyser provides, through analogue simulation, a means of network

218 Fault calculation

analysis by which the required power-system currents and voltages can be obtained
by direct measurement carried out on an appropriate electrical model of the
network under consideration.
The simulation principle can be understood by considering a given network and
assuming it to be represented by a second similar network in which the voltages,
currents and impedances are fixed multiples of the corresponding voltages, currents
and impedances in the given network. The second network is said to be a simulated
equivalent of the given network in that the electrical behaviour of the model
network is identical in form to that of the given network. The response of the given
network can therefore be obtained by direct measurement of the currents and
voltages in the model network, the values so obtained being then multiplied by
known constants to obtain the actual values appropriate to the given network.
Thus, let I/, I and Z denote actual voltages, currents and impedances in the given
network and let I'm, lm and Zm be corresponding values in the model network.
Then the relationships between actual and model values can be written
v=K~v~
I=Kilm
z = K . Zm = ( K e / K 3 Z . ,
where Ke, Ki and Kz are constants. (It should be noted that the constants Ke, Ki
and Kz are not independent in that a given choice of values for any two of these
constants automatically determines the value of the third).
The simulation principle is illustrated by the simple example shown in Fig.
3.6.1A in which the per-unit quantities in the given network correspond to actual
I + jO p.u.
0.2 + j0.S p.u.
o l I t

T 0,825 - j0.056 p.u.

v
0.4*/2 --j! .460 p.u.

Given network

IO+jOV
40 + j100,~

0 8 . 2 5 _ jO.56r"'--'JV t 0.0236 - j0.0730 A

Model network
Fig. 3.6.1A Iflustration o f network simulation
Fault calculation 219
U.
 220 Fault calculation

voltages, currents and impedances in the model network, the relationships between
the two sets of quantities being:
1 per-unit voltage -- 10V
1 per-unit current = 0.050 A
1 per-unit impedance -- 200 F2
It will be noted that the values of voltage, current and impedance used in the model
are such as can be easily provided in the required physical model of the given
network,

3.6,1,1 A.C. analyser

Using the simulation principle described in the previous Section, a given single-
phase a.c, network (for example, the positive-sequence network shown in Fig.
3,4,11B) may be represented by an equivalent singie-phase model network, the
latter consisting of
(a) an appropriate number of single-phase voltage sources, adjustable in magni-
tude and phase, for the simulation of the given network driving-voltages
(b) an appropriate number of impedance units built up from resistors, inductors,
and capacitors, as required, for the simulation of the network branch impe-
dances
(c) suitable interconnection facilities
(d) suitable metering facilities to permit the measurement of currents and
voltages in complex-number form with respect to some common reference
phase
An equipment providing the above-mentioned facilities is termed an a.c. network
analyser, and such equipment has been widely used for fault calculation and network
analysis generally, A typical installation is shown in Fig. 3.6.1.1A, the racks in the
background being used to accommodate the source units and impedance units.
Interconnection facilities are provided by a patch panel shown in the middle of
the racks, the metering facilities being housed in the two consoles shown in the
foreground. The equipment shown is termed a universal network analyser in that it
may be used to represent any given network by suitable interconnection of the
individual source units and impedance units. The impedance units are arranged so
that they can be set at will to any desired impedance value, this being achieved by
the use of [Link] components or decade switches. The source units, similarly, are
provided with means of setting the driving-voltage values in both magnitude and
phase, The impedances are equipped with jack-in points to which the metering
(that is current and voltage measuring facilities) can be connected when required.
In simulating a given power system fault condition it is generally necessary to
represent all three sequence networks and to interconnect these in the network
analyser to represent the particular fault condition being investigated. The sequence
currents and voltages at any point in the network can then be measured at the
corresponding points in the sequence networks, the values obtained being then used
 Fault calculation 221

to obtain the required power-system currents and voltages, Using a network

analyser, therefore, the steps in the fault calculation process are
(a) the representation of the given power system by its positive, negative and
zero-sequence networks to a chosen common base
(b) the physical representation of these sequence networks by means of the
network analyser
(c) the interconnection of the sequence networks in the network analyser to
represent the power system fault conditions under consideration
(d) the measurement of the required sequence currents and voltages in the
network analyser
(e) the calculation of the required power system currents and voltages from the
measured sequence quantities.
The network analyser thus replaces computation of the [Link]
currents and voltages by direct measurement of these quantities in the model
network, the resultant saving in time and labour being usually very considerable,
The accuracy obtained from a network analyser depends essentiaUy on the accuracy
of calibration of the analyser components and that of the associated measuring
equipment, overall error limits of +2% being typical for modem instaLlations,

[Link] D.C. analyser

In many cases, fault calculations can be achieved with sufficient accuracy by

treating all the impedances as scalar quantifies rather than complex quantities, this
simplification being usually restricted to those cases in which the network impe,
dances can all be regarded as having the same phase angle. This assumption makes it
possible to represent all the network impedances by resistors in the network
analyser and also enables the network driving-voltages to be simulated in the
analyser by d.c. voltage sources. In this simplified method of representation,
therefore, the given a.c. network is simulated by an equivalent, or approximately
equivalent, d.c. model, the a.c. currents and voltages in the given network being
represented by d.c. values in the model network. An analyser of this type is termed
a d.c. network analyser or d.c. fault calculator.

3.6.2 Digital-computer analysis

Network analysis by solution of the relevant network equations is readily achieved

by the use of digital-computer methods of analysis, the availability of suitable
large-capacity computers and appropriate analysis programs making it possible to
analyse even the largest networks with required speed and accuracy. Computer-
based methods of analysis are accordingly now very widely used for load-flow
studies, fault calculations and system stability investigations to the extent that they
have now largely superseded the previous use of network analysers for such work.
222 Fault calculation

The use of a computer requires the adoption of an orderly and logical method of
description of the network to be studied, in terms of its nodes and branches, the
appropriate data defining the network being passed to the computer by means of
punched cards or directly from a keyboard. The information so provided is handled
by a program management system, the function of which is to store, retrieve
and generally manipulate the network data for input and transfer to any of the
several analysis programs which may require to be used in the solution of any
particular problem. Such a program management system, developed and used by
the former CEGB and named TRAM, is capable of handling problems involving
up to 1000 nodes and 15000 branches. The network equations involved in any
particular problem are solved by appropriate direct or iterative computational
procedures within the computer using the necessary analysis programs, the latter
covering such standard needs as load-flow determination, three-phase short-circuit
studies, unbalanced-fault studies, network reduction and system stability studies.
Analysts programs are available which also have provision for the inclusion of
power-system protection data, thus making it possible, in fault studies, to
investigate also the response of the protection to the fault conditions under study
and thereby assist, for example, in the determination of required protection settings
and the assessment of expected fault clearance times.

3.6.3 Transient analysis

The methods of analysis so far considered have all been concerned with determina-
tion of the power-frequency components of current and voltage on the assumption
of steady-state conditions, it being sufficient for most studies to assume that only
power-frequency components are present or that such transient components as may
be present, for example in fault conditions, are of only negligible significance in the
particular study in question. Such an assumption may not always be admissible,
however, particularly in studies concerned with the performance of high-speed
protection, recognlsing that such protection is required to perform its function
within a very short time after the instant of fault when significant transient non-
power-frequency components of current and voltage can also be expected to be
present.
The transient components in question which may need to be considered in such
studies include both d.c. and a.c. components, with the latter possibly ranging in
frequency from below fundamental frequency (particularly if large underground-
cable systems are involved) up to several kHz and with decay time-constants, for
both d,¢, and a.c. components, of up to a few hundreds of milliseconds.
The accurate analysis of power-system fault conditions to include such transient
phenomena is recognised to present particular difficulties, the first requirement
being a much more detailed specification of the system under study, including both
lumped and distributed parameters and, where relevant, an appropriate modelling
of non-linear parameters. The second requirement is a suitable method of transient
 Fault calculation 223

analysis of the resulting network, the methods employed to date including both
analogue simulation through use of a transient network analyser and digital analysis
by means of a computer. Details of such methods will be found in the relevant
references contained in the Bibliography.

3.7 Winding faults

3.7.1 General considerations

The analysis of faults in machine and transformer windings presents particular

difficulty in that the effective impedances of the affected windings under internal
fault conditions are not normally known. The methods of analysis which may be
employed are, therefore, severely limited and are usually concerned with obtaining
a reasonable estimate based on certain simplifying assumptions. A particular case in
which simplifying assumptions can yield quite accurate results is that concerned
with the analysis of earth faults on impedance-earthed windings, where the neutral-
earthing impedance can be regarded as large compared with the winding impedance.

1.o =!
I
•.,,,-- x ---~ F I

= I "
I

Zn
I

Fig. 3.7.2A Synchronous machine with fault position F, a per-unlt d/stance x from the
neutral point

3.7.2 Generator-winding faults

Consider the generator shown in Fig. 3.7.2A and let it be assumed that a short-
circuit condition occurs at the point F, a per-unit distance x from the generator
neutral point. Let it be further assumed that the generated e.m.f, (phase-a value)
in that portion of the winding between the neutral point and the point of fault is
xE, where E is the value appropriate to the full winding, the e.m.f, in the remaining
224 Fault calculation

xl~: (1 -x)E

; I ~ :LI
ZIq

I,' I

Positive-sequence circuit

ZN2 ZL2
"-7 ~, I ; O L2

ZF2

F2

Negatlve-sequence circuit

ZN0 ZL0
| I O L0

3Z n ZFO

FO

Zero-sequence circuit

Fig. 3,7,2B Phase-aequencecircuits o f the synchronous machine o f Fig. 3. 7.2,4

portion of the winding thus being (1 - x)E. The generator winding can now be
regarded as a three-ended circuit, these three ends being the neutral end N, the line
or pha~.terminal end L and the point of fault F. It follows, therefore, that the
positive, negative and zero-sequence circuits of the generator will have the form
shown in Fig. 3.7.2B. It will be noted that there are now three generator impedance
values to be considered for each of the three phase-sequence circuits, namely ZN,
Z L and ZF, respectively, with appropriate phase-sequence suffixes, the values of
these impedances depending on the fault position x and the normal full-winding
values of the generator phase-sequence impedances ZI, Z2 and Zo.
In order to apply the generator phase-sequence circuits of Fig. 3.7.2B it is, of
course, necessary to determine the required impedance values, given the values of
x, Zt, Z2 and Zo, Some supposition must, therefore, be made to obtain these
values and the simplest of these is to assume that there is no mutual coupling
between the neutral-to-fault and fault-to-line sections of the winding, and that the
impedances appropriate to each of these two winding sections are proportional to
lengths of these sections. The phase-sequence circuits which result from these
 Fault calculation 225

xE O-x)E
xZ l ~ (I -x)Z l
L1

F1

Positive-sequence circuit

xZ 2 (I-x)Z 2
o L2

!-"2

Negative-sequence circuit

xZ o (l-x)Z 0
I ; ....... O L o

3Z n FO

Zero-seq u e n c e c i r c u i t

Fig. 3.7.2C [Link] circuits of the synchronous machine of Fig. 3.7.24 assuming a
linear winding-impedance distribution

assumptions are shown in Fig. 3.7.2C, from which it will be seon that
ZN1 = xZI ZL l = (1 - x)ZI [
ZN2 = xZz ZL 2 ffi(1 - x)Z2 I 3,7,2,1
ZNo = xZo ZL o = ( 1 - x ) Z 0
the remaining impedances ZFI , ZF2 and ZFO all being zero.
Considering the simple case in which there is assumed to be no infood to the
fault from the generator terminals, the phase-sequence impedance values as soon
from the point of fault are xZl, xZ2 and xZo, the pro-fault voltage at th0 point
of fault (phaso-a value) being xE.
Hence, considering a three-phase fault at F, the resultant fault current value is
given by xE E
Ij, = = [Link]
xZl Zl
the value of the fault current thus being seen to be independent of fault position.
Similarly, for a phase-to-phase fault at F, the value of the phase-fault current is
226 Fault calculation

given by
43e
I, = ~ [Link]
-~ =xZ~ + xZ2 ZI + Z2
which is also seen to be independent of the fault position.
Considering now, a single-phase-to-earth fault at F, the value of the fault current
is given by
3xE
I/= [Link]
xZ1 + xZ2 + xZo + 3Zn
and if Z n is very much greater than ZI, Z2 and Zo, the expression for the fault
current reduces to the approximate form
xE
If = ~ [Link]
z..
The ease of the short-circuited turns condition can be treated by assuming the
short-circuited portion of the winding to constitute a fraction x of the full winding,
as shown, in Fig. 3.7.2D, it being further assumed that the short-circuit current I¢ is
quite independent of the position of the short-circuited turns within the winding.
With these assumptions, the value of the fault current will be precisely the same
as for a [Link] fault at a per-unit distance x from the neutral point, the
fault current being therefore given by
3xE 3E
11.= = [Link]
xZ 1 + xZ 2 + xZ 0 Z 1 + Z 2 +Z 0
The value of the fault current is thus seen to be independent of the fraction of the
winding spanned by the short-circuit.
In all the above-mentioned fault conditions, the effect of fault infeed from the
generator terminals can be taken into account by connecting the generator phase-
sequence circuits of Fig. 3.7.2C to those applicable to the rest of the system.

1.0
I

I I I
! i

Fig. 3.7,2D
 Fault calculation 227

It is important to note that the generator phase-sequence circuits of Fig. 3.7.2C

are based on a particular and somewhat arbitrary supposition regarding the sub-
division of the generator impedance values. The results obtained must not,
therefore, be regarded as anything more than a very rough approximation to the
actual conditions.
It should be noted also that this simple treatment is applicable only to the case
of a generator with a single-conductor stator winding. A more rigorous treatment,
applicable to generators with stator windings having a number of paths in parallel,
is given by V.A. Kinitsky.

L. 1.0 LI
I
I
I
C o

I
C 0

7
. . . . J

/
1.0

/
.~ 0.8
E

/
E
0.6

/
~. 0.4

/
0.2

0.2 0.4 0.6 0.8 1.0

X

Fig. 3.7.3A Variations o f fault current with fault position for an impedance-earthed star
winding
228 Fault calculation

3.7.3 Transformer-winding faults

In dealing with earth faults on transformer windings, the special case of impedance-
earthed windings is amenable to the simplified treatment already described in the
previous Section, the value of the fault current being the pre-fault phase-to-earth
voltage at the point of fault divided by the effective value of the system earthing
impedance. The two cases of most common interest are shown in Figs. 3.7.3A and
3.7.3B, namely the star-connected winding and the delta-connected winding.

I I
L..___ _> . . . . J

1.0

0.8

E
E
'2 0.6
\ /
E \ /
•.& 0.4
.=
E

~ 0.2

0.2 0.4 0.6 0.8 1.0

Fig, 3 , 7 , 3 B Variation o f fault current with fault position for an impedance-earthed delta
winding
 Fault calculation 229

It is of interest to note, in the case of the delta-connected winding, that the pre-
fault phase-to-earth voltage has a minimum value when the fault is at the centre of
the affected winding, this minimum value being one half the phase-to-neutral
voltage. As will be seen from Fig. 3.7.3C, the magnitude of the pre-fault phase-to-
neutral voltage, E/, at the point of fault is
E/= V1 +3x(x- 1)
where E is the phase-to-neutral voltage of the winding and x is the per unit distance
of the fault from one terminal of the winding.

From the Cosine Rule:-

E~= F.2 + (x~/~E) 2 - 2xur3E 2 cos 30 °

giving Ef = E~/I + 3 x ( x - I )

',l ,erhe=
¢

I x45~: I

J~E I
I

Fig. 3.7.3C Prefault phase-to-neutral voltage In a delta winding and its dependence on fault
position

The case of an earth fault on a solidly-earthed star-connected transformer-

winding is of particular interest, and has been investigated by tests on a 132133 kV
45 MVA star/delta transformer, the results being as shown in Fig. 3.7.3D. It will be
noted that, within the limited range of fault positions investigated (accessible by
virtue of the tap-change connections), the earth-fault current is a maximum for a
fault close to the neutral point, the current flowing into the winding from the phase
terminal having a minimum value of zero for a fault in this position. Extending
these results to the case of short-circuited turns it is not difficult to see that the
fault current in the short-circuited turns can be expected to be high if the fraction
of the winding which is short-circuited is small, the high value of current in the
short-circuited turns being accompanied by a relatively low value of current in the
remainder of the winding.
 230 Fault calculation

IA
1.0 --I

I x

I I I

f
Ip

t. . . .

12
\
\
\
10 \

t. f
J
/
/ I s '~"
1 1

0.1 0.2 0.3 0.4

x

Fig. 3.7.3D Variation o f fault current with fault position for an earth fault in the solidly-
earthed star-connected winding o f a star/delta transformer

3.8 Appendixes

3.8.1 Representation of [Link] transformers

The representation o f transformer circuits given in Section 3.3.3 was on the basis
that the transformer windings could be assumed to have the same rated voltage as
 Fault calculation 231

their associated network sections, such transformers being termed nominal-ratio

transformers. Thus, a transformer interconnecting a 132 kV system and a 33 kV
system would, on this basis, be assumed to have a transformation ratio of exactly
132 kV to 33 kV with a corresponding turns ratio (star/star windings assumed) of
exactly 4 to 1. It is, however, sometimes ne6essary to consider cases in which the
transformer voltage ratio is not the same as the ratio of the rated voltages of the
associated network sections, the most common instance being that in which the
effect of transformer'tap-change equipment must be taken into account.

Z H + k2N2ZL
O ~ ". 3 ,.~ O
iH IL/kN

Iv

Fig. 3.8.1A Representation o f an off-nominal-ratio transformer

In considering the representation of such an off-nominal-ratio transformer, let it

be assumed that the two-winding transformer of Fig. 3.3.3A has a turns ratio (star•
star windings assumed) of k N instead of N, so that
rated voltage of winding H
kN =
rated voltage of winding L
and N, as before, is the ratio of the nominal rated voltages of the network sections
to which the windings H and L are connected. Thus, considering the case of a
118.8/33 kV transformer interconnecting a 132 kV system and a 33 kV system it
is seen that
118.8
kN = ~ = 3.60
33
132
N = = 4.00
33
so that
3-60
k- - 0.90
4.00
The symbol k thus denotes a positive real number whose value may be less than,
equal to, or greater than unity, its value being unity when the transformer ratio is
exactly equal to the ratio of the nominal rated voltages of the associated network
sections.
232 Fault calculation.

ZH + k2N2Z L IL/kN IL
HO C-'1 >

VL (=NV L)

Fig. 3.8.1B Equivalent circuit of the off-nominal-raffo transfoffner

Proceeding as in Section 3.3.3, the equivalent circuit may be obtained directly by

replacing N by kN in the equivalent-T circuit shown in Fig. 3.3.3B, the resulting
circuit with magnetising impedance ignored being as shown in Fig. 3.1.8A. It will
be noted, however, that the low-voltage-side values, previously NVL and IL/N,
are now kNVL and IL/kN and are therefore no longer compatible with the referred
values of the other voltages and currents in the low-voltage network section. These
latter values, it will be remembered, are referred to the nominal rated voltage of the
high-voltage network section by multiplication by N and division by N respectively,
not by kN. To overcome this difficulty, therefore, the referred values N V L and
IL/N must be obtained by applying the values kdVVL and IL/kN in Fig. 3.8.1A to
an ideal transformer of turns-ratio k as shown in Fig. 3.8,1B, the low-voltage-side
values obtained from the transformer being the true referred values V~(ffiNVz.)
and I~ (= IL/N). Fig. 3.8.1B is therefore the reqttired off-nominal-ratio equivalent
circuit of the transformer.
It is important to note in Fig. 3.8.1B that the impedance ZH + k2AraZL is the
short-circuit impedance ZHL of the transformer, as measured from the high-voltage
side H, with the given transformation ratio of kN. If the non-unity value of k is due
to the effect of tap-changer equipment, it is of interest to know the relationship
between this off-nominal-ratio impedance and the value applicable for the nominal-
ratio (k = 1) tap-position, the latter-mentioned value being the one normally known.
To obtain this relationship let
k = kH/k L
where

kH-- rated voltage of winding H for given tap-position

rated voltage of winding H for [Link] tap-position
and
rated voltage of winding L for given tap-position
kL-
rated voltage of winding L for nominal-ratio tap-position

the values of kz/and kL being each identical to the ratio of the numbers of turns on
the winding referred to for the given and nominal-ratio tap-positions.
Now the leakage impedances Z H and Z L are predominantly reactive and can be
assumed to be approximately proportional, within limits, to the square of the
 Fault calculation 233

jlO.125% j10.125%
HO r"-I OL

0.9 "~ • 1.0

Positive (k = 0 . 9 ) Z e r o (k = 0 . 9 )

j12.5% j l 2.$%
HO I I OL H O I I = oL

Positive (k = !.0) Zero (k = 1.0)

j15.125% jlS.125%
HC ; I ~ OL

1.1 1.0

Positive ( k = 1.1) Z e r o (k = I . I )

Fig. 3.8.1C Effect o f off-nominal-ratio operation o f a 132/33 k V 60 MVA star/delta trans-

former on the transformer phaze-sequence circuits

number of turns on the windings concerned. It therefore follows that, to a close

approximation
zM = kb(Z.).om J
= eL(z, ).ore [' 3.8.1 1
where (ZH)nom and (ZL)nom are the horrdnal-rafio values of the two impdeances.
The value of the transformer impedance ZHL, referred to the high-voltage
winding is therefore
ZHL = ZH + k~N2Z~.
= k~l(ZH)nom + (kH/kL)2N~k~(ZL)nom
= k~[(ZH)no m + ~ ( Z L ) n o m ]

ZH L = k~l(ZHL)nom
2 [Link]
where (ZHL)nom is the nominal-ratio impedance of the transformer referred to the
high-voltage winding.
234 Fault calculation

The application of these results to the representation of an [Link]-ratio

transformer is illustrated in Fig. 3.8.1C for the case of a 60 MVA star/delta trans-
former with a nominal ratio of 132[33 kV and a nominal-ratio impedance of
j12.5%, the star connected 132 kV winding being solidly earthed. The tap changer
is assumed to be on the 132 kV winding so that k n = k and kL = 1, the equivalent
circuits shown in Fig. 3.8.1C representing k values of 0.9, 1.0 and 1.1, correspond-
ing to transformation ratios of 118.8/33 kV, 132/33 kV and 145.2/33 kV,
respectively.
The representation of off-nominal-ratio transformers by means of circuits
containing ideal transformers, as shown in Fig. 3.8.1C, has the disadvantage that
such circuits, although suitable for network analyser use, are not satisfactory for
normal network analysis by computation. To overcome this difficulty, therefore,
the ideal transformer of turns-ratio k must be replaced by a suitable equivalent
circuit. The derivation of this equivalent circuit can be understood by considering
the ideal transformer of Fig. 3.8.1D with a primary-to-secondary turns-ratio k and
an impedance Z in series with the primary winding. Let it be assumed that this
circuit can be replaced by an equivalent-zt circuit consisting of a primary-side
shunt-impedance A, a secondary-side shunt-impedance B and a series impedance C.
For the two circuits to be electrically equivalent the no-load transformation-ratios,
the short-circuit impedances and the open-circuit impedances must each be the
same for both circuits. Considering, first, the no-load voltage-transformation ratios,
it is seen that the primary-to-secondary ratio is
A
k=~ [Link]
A+C
and the secondary-to-primary ratio

1 B
- - - [Link]
k B+C
Similarly, the impedance measured at the primary terminals with the secondary
terminals short-circuited is
AC
Z- - - [Link]
A+C

k:l
Z Z/k
os

I I)

0 0

Fig. 3.8.1D
 Fault calculation 235

and that measured at the secondary terminals with the primary terminals short-
circuited is
Z BC
- - - 3.8,1,6

k2 B+C
From eqns. [Link] and [Link] it is seen that
Z
C=~ 3,8,1,7
k
and substituting this value of C in eqns. [Link] and [Link] gives
Z
A - 3,8,1.8
1-k

jl 1.25% jlO.125%
H~ - [ I - oL HC [ ~| OL

~j101.25% t-j112.5%

v v

Positive (k : 0.9) Zero (k = 0.9)

j12.5% j12.5%
H C ~ oL HO r--"7 OL

Positive (k = 1.0) Zero (k = 1.0)

jl 3.75% jl S.12S%

-jl s ~ .2 s% j 137.5%

, f, i

Positive (k := 1.1) Zero (k = 1,1)

Fig. 3.8.1E Application of the equivalent-n representation to the phase-sequence el¢cuits of

Fig. 3.8.1C
236 Fault calculation

and
Z
B = [Link]
k(k- 1)
These values of A, B and C are the required equivalent-It values and it can easily
be checked that they satisfy the remaining condition, namely that the open-circuit
impedances as measured from the two sides should both be equal to infinity.
Applying these results to the phase-sequence circuits shown in Fig. 3.8.1C, the
resulting circuits are as shown in Fig. 3.8.1E.

1 3 2 / 3 3 kV
45 M V A
j12.5%

I S 0 0 M V A fault-level - - - - A B
X/R ratio -" S

1 3 2 / 3 3 kV
45 MVA
jl2.S%
Fig. 3.8.1F

The representation of an [Link]-ratio auto-transformer is achieved by

precisely the same reasoning as that used for the two-winding transformer, the
relationship in this case between the transformer impedance ZHL and the nominal-
ratio value (ZHL)nom being

[kHN-kL] '
ZHL = N - 1 (ZHL)nom [Link]

where kH and k L are as already defined for the case of the two-winding transformer.
If tap-changing is effected by variation of the [Link]-winding tapping-point,
as is commonly the case, then kH = 1 and k L = 1/k so that Equation [Link]
becomes

ZHL = k(N- 1) (ZHL)n°m [Link]

Example: The bulk-supply point feeding a 33kV distribution network consists of

two 45 MVA 132/33 kV transformers in parallel, as shown in Fig. 3.8.1F, each
transformer having a nominal-ratio impedance of ]12.5 per-cent. The 132 kV
system, as seen from the 132 kV busbars of the bulk-supply point can be assumed
to have an impedance corresponding to a three-phase short-circuit level of 1500
MVA and an X/R ratio of 5, the generated voltage behind the system impedance
 Fault calculation 237

j0.125
I +jO {'---I
_ . @ 0 . 0 0 5 8 8 + j0.02942
l "~ A
j0.12S

I +jO

__~ 0.00588 + j0.02942 j0.062S

r--1 O I I 0 -
A B

I +jo

-Q 0.00S88 + j0.09192
I I

Fig. 3.8.1G

being the nominal phase-to-phase value of 132 kV. Determin0 the value of the
three-phase fault current for a fault on the 33 kV busbars, assuming:
(a) that both transformers are on the nominal-ratio tap-position
(b) that one transformer, T1, is on the nominal-ratio tap-position and the oth0r
7"2, is on the tap-position corresponding to a voltage-ratio of 145.2/33 kV,
the tap-changer being on the h.v. winding.
238 Fault calculation

Determine also the distribution of the fault current in these two cases. What are
the values of the transformer currents in case (b) when the fault is absent, assuming
the load on the 33kV busbars to be zero?

Solution: using the per unit method and adopting 45 MVA as the common base,
the common-base per-unit impedance (nominal-ratio" value) of each transformer is
/,0,125 p.u.
The scalar value of the 132 kV system impedance is given by
45
= 0.0300 p.u.
1500
the phase-angle of this impedance being the angle whose tangent is 5, that is 78.7 °.
The complex value of the 132 kV system impedance is therefore given by
0.0300(cos 78.7* +/.sin 78.7*) = 0~2300(0-1959 +/.0.9806)
= 0.00588 +/'0.02942 p.u.
The positive-sequence network for case (a) is therefore as shown in Fig. 3.8.1G,
the network reducing to the driving voltage of 1 +/,0p.u., used as the vector base,
in series with an impedance of 0-00588 +/.0.09192 p.u.
The current flowing into the fault at F is therefore given by
1 +/'0
If = = 0.693 -/,10.83 p.u.
0-00588 +/.0.09192
The current corresponding to the common base of 45 MVA at 33kV is
45 x 106
= 788 A
1.732 × 33 000
The total fault current in amps is therefore given by
If = (0.693 -/,10.83)788 = 546 -/.8537 A
giving
l f = 8552/.-86.33* A
Turning, now, to case (b), the positive-sequence network for the system is shown in
Fig. 3.8.1H, transformer T2 being now represented by its equivalent-Tr circuit. The
[Link] impedance value of/,0.125 p.u. becomes the off-nominal-ratio value
(1-1)2 (j0.125) =/.0.15125 p.u.,
since
145.2
k---- 1.1
132
and henoe the series arm of the equivalent-~r circuit has the value
/.0.15125
=/,0.1375 p.u.
1-1
 Fault calculation 239

The left-hand shunt arm of the equivalent-fr circuit has the value
]0.15125
=-]1.5125 p.u.
1-1.1
and the right-hand shunt arm the value
]0.15125
=]1.375 p.u.
1.1(1.1- 1)

j0.125
I +jO

_ ~ 0.00588 + j0.02942
I I A S •

- t jO.1375

-jl.5125

l j1.3,75

I +j0
. ~ 0.00588 + j0.02942 j0.06548
I I A1 | }

--ji.5125

] +j0

0.00588 + j0.02942 j0.06844

I I O ! I
A

Fig. 3.8.1H
240 Fault calculation

The reduction of the positive-sequence network is shown in Fig. 3.8.1H from which
it will be seen that the equivalent circuit, as seen from the 132 kV system, is a
driving voltage of 1 +/0 p.u. in series with an impedance of
(0.00588 +]0-02942) +/'04)6844 = 0.00588 +/0439786 p.u.
The current flowing into the station, from the system, at the 132 kV busbars is
therefore 1 +]0
= 0.612-/10.18 p.u.
0.00588 +/0.09786
and hence the voltage at A, namely the 132 kV busbars, is given by
(0.612 - / 10.18) (/0.06844) = 0.6967 +/0.04189 p.u.
The currents in the remaining network branches can now be readily obtained and
are shown in [Link] value in Fig. 3.8.1I, the currents in amps being shown in
Fig, 3.8,1J along with those for case (a).
it will be noted that the total fault current at F is given by
I f = 0.640 -/10.641 p.u.

the value in amps being

It'--(0.640-/I 0-641)788 = 504-/8385 A
giving
I/= 84001-86.56 ° A
This value, it will be noted, differs only slightly from the value obtained for case
(a), the current magnitude being approximately 1.8% below that obtained in case
(a).
Considering, now, the final part of the question in which the conditions are as
for case (b) but with the fault absent, the required positive-sequence network is as
shown in Fig. 3.8.1K. The steps in the network reduction process are as shown and
these lead to an equivalent circuit, as seen from the 132 kV system, consisting of
the driving voltage of 1 +/0 p.u. in series with an impedance of
(0.00588 +/0.02942) +/30.26 = 0.00588 +/30.29 p.u.

0.335 - jS.574

+j__.o
, I ; ' 1
0 . 6 1 2 -- j l O . 1 8 J 1 0 . 6 4 0 -- j 1 0 . 6 4 1

. . t

Fig, 3,8,11
 Fault calculation 241

Tl
1069/-86.33° 4276/-86.3.30
76"3L0----'~~V
'° 1A ' ~ ' B"
8552[~8~~6"33° ,. F
!1069/286~: : ~°3 3T2 4276/=8633~
o]

a) Transforme~on sametap position

Ti
1100[-86.$6° 4400/T86.$6
~

---_ oi, 1 S400/--~6.$6 °

b) Transformers on dissimilar tap positions

Fig. 3.8.1J

The current flowing from the 132 kV system into the 132 kV busbars is therefore
1 +/0
- 0.0000064 -/0.03301 p.u. on 45 MVA,
0.00588 +/30.29
and the voltage at A, namely the 132 kV busbars, is therefore

(0-0000064 -/0.03301) (/'30.26) = 0-9989 +/0.000194 p.u.

The current flowing to neutral through the shunt impedance of-/1.5125 p,u, is
thus
0.9989 +/0.000194
= -0.000128 +/0.6604 p.u.
-/1.5125
and hence the current flowing from A to B through the series impedance of
/'(3. 06548 p.u. is
The voltage at
(0.0000064 -/0.0331)- (-0.000128 +/0.6604)
= 0-000134 -/0.6935 p.u.
The voltage at B, namely the 33 kV busbars is therefore
(0-9989 +/0.000194) - (0.000134 -/0.6935) (/0.06548)
= 0.9535 +/0.000185 p.u.
242 Faultcalculation

j0.125
1 +jo I !
~ 0,00588+j0.02942
L ! A B

jO.1375

-jl.5125 i jl.375

I +jo

_.~ 0.005~8+ j0.02942 j0.06548

l 2 ['--'I B

I +jo
-ji.5125

t
jl.375

~ 0.00588+j0.02942 j30.26
I I o ! I
A

Fig. 3,8.1K
The remaining branch currents can now be readily obtained and are shown in per-
unit value in Fig. 3,8.1L. The currents in amperes are given in polar form in Fig.
3.8.1M.

3,8,2 Effects of overhead-line a s y m m e t r y

In the great majority of network calculations by symmetrical components, it is

sufficient to assume that all the items of plant concerned have balanced impedances,
 Fault calculation 243

0.0000702 - j0.3633

- - Q
'+'--°°°°°°°"-'°°33°'/
, , . .
" ' ' 1
-0.0000~3s +,0.3303 N ~] -0.0o007o2 ÷ j0.3633

-[Link]~2s ÷j0.6604 ~, ~, 0.000134 -,0.693s

.Ik

Fig. 3.8.1L

that i§ that the impedance presented to current of any given phase sequence is the
same for all three phases. Although this assumption is valid for all practical
purposes in so far as generators, transformers, reactors and cable-circuits are con-
cerned, it is not strictly true for overhead-line circuits and occasions may arise in
which the effect of this phase unbalance may have to be assessed.
For the general unbalanced three-phase circuit, the following equations relate
the sequence voltage-drops to the sequence currents producing them
VI =IiZll +I2Zs2 +loZlo
112 =IIZ21 +I2Z22 +loZ2o [Link]
Vo =IiZo, +12Z02 +IoZoo
The impedances Z t l , Z22 and Zoo are known as the positive, negative and zero-
sequence self-impedances, respectively, of the given circuit, and are the values
normally quoted as Zz, Z2 and Zo when unbalance effects are not being con-
sidered. The impedances Zm, Zlo, Z2s, Z20,Zo, and Z02 are known as the
sequence mutual-impedances. Thus, the impedance Zo, represents the zero-
sequence voltage-drop produced in the circuit when unit value of positive-sequence
current flows through it.
For static plant
Zll =Z22 1
Zol =Z2o ~ [Link]
Z s o = Zo2 /
T1
.71"5[-90°~ ~ 286/--90°
76.3/0° kV
Q 6.50/__900
A
T2
65"0/~90° ~ ) 286/-90°.~,

Fig. 3.8.1M
244 Fault calculation

It is important to note that the sequence mutual-impedance values are not

reciprocal; in other words
Zlo 4:Zol
Z12 ~Z21
Z2o 4=Zo~
Values of the sequence self and mutual impedances for 132 kV overhead lines have
been given by Butterworth, the following values referring to one circuit of a 132 kV
0.175 in2 e.c. (175 mm 2 al.) double-circuit line with a single earth-wire.
Z11 = Z22 =0.158 +]0.406 ~2 per km
Zoo = 0.359 +/,1.10 I2 per km
Zlo = Zo2 = -0.00143 -/0.0206 I2 per km
Z12 = -0.0214 +/0.0172 12 per km
Z21 = 0-0220 +/'0.0168 ~2 per km
Zol = Z2o = 0.0111 -/.0.0240 I2 per km
It will be noted that the sequence mutual-impedances are small compared with the
sequence self-impedances, the largest sequence mutual-impedance having a
magnitude equal to about 6% of the positive-sequence self-impedance. In a
perfectly balanced circuit, all the sequence-mutual impedances are zero and the
phase-sequence voltage-drop equations given in eqns. [Link] then reduce to the
form shown in eqns. [Link].
The following example illustrates the use of the sequence self and mutual
impedances.
Example : Two stations A and B are interconnected by a 32.2 km 132 kV over-
head line, the latter having a three-phase fault clear of earth at end B. Determine
the zero-sequence current in the line, assuming station A to have a positive-
sequence source-impedance of/.8.70 I2, stations A and B both having zero-sequence
impedances of/'6.96 ~2 at 132 kV. The overhead line can be assumed to have the
sequence self and mutual impedance values already quoted.
Solution: An exact solution of the problem will not be attempted because of the
cosiderable arithmetic involved. The following approximate solution is acceptable,
however, because of the relatively small magnitudes of the sequence mutual-
impedances, compared with the sequence self-impedances.
From the given data, the positive-sequence self-impedance of the overhead line is
ZlI ---32.2 (0.158 +/.0.406) 12 = 5.10 +/13.06 ~2
The positive-sequence current flow through the line to the three-phase fault is
therefore
76300 +/,0
11 --
/.8.70 + (5.10 +/13.06)
= 780 +/3330 A
giving
11 = 3420/-76 ° 45' A
 Fault calculation 245

The appropriate sequence [Link] of the overhead Line, from the given
data, is
Zot = 32.2 (0.0111 -/.0.0240) [2 = 0.356 -/'0.774
The zero-sequence voltage-drop produced in the Line by the positive-sequence
current is therefore given by
Vo = I l Z o t
= (780 -/3330) (0.356 -/0.774)
giving
Vo = -2302 -/1789 V
Now, the zero-sequence self-hnpedance of the overhead Line,from the given data, is
Zoo = 32.2 (0.359 +/,I.10)I2 -- II.54 +/35.4 [2
and hence, the total impedance presented to the flow of the zero-sequence current
in the line is
Zo = 2(/6.96) + (11.54 +/.35.4) = 11.54 +/.49.3 ~2
The zero-sequence current in the line is therefore
Vo
Io = ~
Zo

- 2302 -/,I 789

A
11.54 +/..49.3
---44.8 +/36.2 A
giving
Io = 57"6/!41 ° 4' A
This zero-sequence current, which flows in the opposite direction to the positive-
sequence current in the line, is seen to have a magnitude equal to about 1.7% of
the positive-sequence current.

3.9 Bibliography

British Standards

BS.171 : 1978,'Power transformers'

Books

Symmetrical components by C F Wagner and R D Evans (McGraw Hill, 1933)

Circuit analysis ofa.c, power systems by Edith Clarke (John Wiley & Sons, 1950)
Power system analysis by J R Mortloek and M W Humphrey Davies (Chapman &
Hail, 1952)
246 Fault calculation

The general theory of electrical machines by B Adkins (Chapman & Hall, 1957)
Advanced studies in electrical power system design by R A Hore (Chapman & Hall,
1966)
Electric energy systems theory by O Elgard (McGraw Hill, 1971)
Protective relays application guide (GEC Measurements, 1975)
Power system planning by R L Sullivan (McGraw Hill, 1976)
Computation of power-system transients by J P Bickford, N Mullineux and J R Reed
(Peter Peregrinus, 1976)

Articles

'Digital computers in power system analysis' by P P Gupta and M W Humphrey

Davies (Proc. IEE, 1961, 108, pp. 383-398)
'Some improved methods for digital network analysis' by A Brameller and J K
Denmead (Proc. lEE, 1962, 109, pp. 109-116)
'A conjugate-impedance network analyser operating at 50 c/s' by W Casson and
A W Hales (Proc. IEE, 1958, 105, pp. 295-303)
'Electrical characteristics of overhead lines' by S Butterworth (ERA Report 0T4,
Pt. I, 1950 and Pt. 1I, 1951)
'Calculations of internal fault currents in synchronous machines' by V A Kinitsky
(IEEE Int. Conf. Record, Pt. III - Power, 23rd-26th March 1964, pp. 13-31)
'Digital simulation of faulted e.h.v, transmission lines with particular reference to
very high speed protection' by A T Johns and R K Aggarwal (Proc. IEE, 1976, 123,
pp. 353-359)
'Developments in techniques for simulating faults in eJ~.v, transmission systems' by
A T Johns and M M T E1-Kateb (Proc. lEE, 1978, 125, pp. 221-229)