PS 0523 POWER SYSTEM

MODELLING and ANALYSIS I

SYMMETRICAL FAULT ANALYSIS USING

BUS IMPEDANCE MATRIX

1
1 INTRODUCTION

Short circuit study is one of the basic power system analysis problems. It is
also known as fault analysis. Whenever a fault occurs, the bus voltages and
flow of current in the network elements get affected. Faults occur in power
system i) due to insulation failure in the equipments
ii) due to flashover of lines initiated by lightning stroke
iii) due to mechanical damage to conductors and towers
iv) due to accidental faulty operation.

Faults can be classified as Shunt faults and Series faults. Series faults are not
so severe as compared to shunt faults. Shunt faults are further classified as
1. Three phase fault
2. Single line to ground fault
3. Line to line fault
4. Double line to ground.

2
Three phase fault is a symmetrical fault and hence it can be analyzed on per
phase basis. All other faults mentioned above are unsymmetrical faults and
they can be analyzed using symmetrical components.

The relative frequency of occurrence of various faults in the power systems, in

the order of severity is as follows;
Three phase fault 5%

Double line to ground fault 10 %

Line to line fault 15 %

Single line to ground fault 70 %

3
When a fault occurs, large fault current flows through the fault causing large
current flows in the lines and reduced bus voltages. Heavy current flows may
cause damage to the equipments. Hence faulty section should be isolated from
the rest of the network quickly on the occurrence of a fault. This can be
achieved by providing relays and circuit breakers. The protective relays sense
the occurrence of the fault and send signals to circuit breakers to open the
circuit under faulty condition. The circuit breakers should able to interrupt
the current flowing after a few cycles.

The calculation of currents in network elements for different types of faults

occurring at different locations is called SHORT CIRCUIT STUDY. The
results obtained from the short circuit study are used to find the relay
settings and the circuit breaker ratings which are essential for power system
protection.

Short circuit study is carried out using Bus Impedance Matrix. Symmetrical
short circuit study using Bus Impedance Matrix is presented first. Followed
by this unsymmetrical fault analysis using symmetrical components is
discussed.

4
Bus impedance building algorithm

5
Consider the power system shown in Fig. 1. The values marked are p.u.
impedances. The p.u. reactances of the generator 1 and 2 are 0.15 and
0.075 respectively. Compute the bus impedance matrix of the generator –
transmission network.

1 j 0.1 2

G1 G2

j 0.1 j 0.1
G1 3
G1

Fig. 1

6
The ground bus is numbered as 0 and it is taken as reference bus. The
p.u. impedance diagram is shown in Fig. 2.

1 j 0.1 2

j0.15 j 0.1 j 0.1 j0.075

3
0 0
Fig. 2

Its bus admittance matrix can be obtained as

26.6667 - 10 - 10
YBus =  j  - 10 33.3333 - 10 
 
 - 10 - 10 20 

7
One way of finding its bus impedance matrix, Z Bus, is to invert the above bus
admittance matrix. If the number of buses is more, it is difficult to get direct
inverse. Alternatively ZBus matrix can be obtained by constructing it adding the
element one by one.

The network graph of this power system is shown in Fig. 3.

3
1 2

4 5
3 2
1

Fig. 3

8
The network consisting of elements 1, 2 and 3 is a partial network with
buses 0, 1 and 2. To this if element 4 is added, the network will be as
shown in Fig. 4.

3
1 2

3
1 2

Fig. 4

Now a new bus 3 is created. The added element is a BRANCH. For the
next step, network with elements 1,2,3 and 4 will be taken as partial
network. This contains buses 0,1,2 and 3
9
When element 5 is added to this, the network will be as shown in Fig. 4.

3
1 2

4 5
3
1 2

0
Fig. 4

In this case, no new bus is created and the added element links buses 2
and 3 and hence it is called a LINK.

Consider elements 1 and 2 alone with impedances j0.15 and j0.075. Its YBus is
given by

6.6667 0 
YBus =  j  and hence its ZBus is given by
 0 13.333 

1 2 1 2
j 0.15 j 0.075
1 0.15 0 
ZBus = j
2  0 0.075  - j 6.6667 - j 13.3333
10
Note that these element values are the values of the shunt impedances at buses 1
and 2. This result can be extended to more number of buses also. The above ZBus
corresponds to the partial network shown in Fig. 5.

1 2

1 2
j 0.15 j 0.075

0
Fig. 5

Now add element 4. This it from bus 1 to 3 with an impedance j 0.1. The added
element is a branch. It is from the existing bus 1 and it creates a new bus 3 as
shown in Fig. 6. The modified ZBus matrix is

1 4 2
j 0.1
3
1 2

0
Fig. 6 11
 1 2 3
1 0.15 0 0.15 
ZBus = j  0 0.075 0 
2  
3 0.15 0 0.25 

Now add element 3. It is from bus 1 to 2 with an impedance of j 0.1. Since it is

linking two of the existing buses 1 and 2 as shown in Fig. 7, it is a Link between
buses 1 and 2.
3

j 0.1
1 2

3
1 2

Fig. 7

12
The modified ZBus matrix with ℓ th bus is

1 2 3 

1 0.15 0 0.15 0.15 

 0 0.075 0  0.075 
ZBus = j 2  
0.15 0 0.25 0.15 
3  
 0.15  0.075 0.15 0.325 

Eliminating bus ℓ

1 2 3
1 0.0808 0.0346 0.0808 
ZBus = j 2 0.0346 0.0577 0.0346 
 
3 0.0808 0.0346 0.1808 

13
Now add element 5. It is from bus 2 to 3 with an impedance of j 0.1. Since it is
linking two of the existing buses 2 and 3 as shown in Fig. 8, it is a Link.

3
1 2
5
4
j 0.1
3
1 2

Fig. 8

The modified ZBus matrix with ℓ th bus is

1 2 3 

1  0.0808 0.0346 0.0808 - 0.0462 

 0.0346 0.0577 0.0346 0.0231 
ZBus = j 2 
 0.0808 0.0346 0.1808 - 0.1462 
3  
 - 0.0462 0.0231 - 0.1462 0.2692 
14
 1 2 3 
1  0.0808 0.0346 0.0808 - 0.0462 
 0.0346 0.0577 0.0346 0.0231 
ZBus = j 2 
 0.0808 0.0346 0.1808 - 0.1462 
3  
 - 0.0462 0.0231 - 0.1462 0.2692 

Eliminating bus ℓ the final bus impedance matrix is obtained as

1 2 3
1 0.0729 0.0386 0.0557 
ZBus = j 2 0.0386 0.0557 0.0471
 
3 0.0557 0.0471 0.1014 

The correctness of the result can be verified by multiplying the values of the two
matrices ZBus and YBus.

15
2 BUS IMPEDANCE MATRIX - BUILDING ALGORITHM

Bus impedance matrix Z bus of a power network can be obtained by inverting

the bus admittance matrix Ybus , which is easy to construct. However, when
the order of matrix is large, direct inversion requires more core storage and
enormous computer time. Therefore inversion of Ybus is prohibited for large
size network.

Bus impedance matrix can be constructed by adding the network elements

one after the other. Using impedance parameters, performance equations in
bus frame of reference can be written as

E bus  Z bus I bus (1)

16
E bus  Z bus I bus (1)

In expanded form the above becomes

 E1   Z 11 Z 12  Z 1N   I1 
E  Z Z 22  Z 2N  I 
  =  21
2  2 (2)
         
     
E
 N Z N1 Z N2  Z NN  I N 

From this we can write

E p  Z p1 I 1  Z p2 I 2    Z p q I q    Z pN I N (3)

17
E p  Z p1 I 1  Z p2 I 2    Z p q I q    Z pN I N (3)

With I q  1 p.u., other bus currents set to zero, E p  Z pq . Thus Z pq can be

obtained by measuring E p when 1 p.u. current is injected at bus q and
leaving the other bus currents as zero. p and q can be varied from 1 to N to
cover all the elements of Z bus .

While making measurements all the buses except one, are open circuited.
Hence, the bus impedance parameters are called open circuit impedances. The
diagonal elements in Z bus are known as driving point impedances, while the

off-diagonal elements are called transfer impedances.

While constructing Z bus using building algorithm, elements are added one by

one. At any stage, the added element may be a branch or a link as

explained below.
18
The performance equation of the network shown in Fig. 4 is

2
Partial
network 


m

0
Fig. 4
E bus  Z bus I bus (4)

where E bus = an m x 1 vector of bus voltages measured with reference to

reference bus 0
I bus = an m x 1 vector of bus current

When an element p-q is added to the partial network, it may be a branch or a

link.
19
Addition of a branch
An element having an impedance of z α is added from bus p, creating a new
bus q as shown in Fig. 5.
1

2

Partial
network q
 p zα

m

0
Fig. 5
The performance equation of the new network will be

 E1   Z 11 Z 12  Z 1p  Z 1m Z 1q   I1 
E  Z Z 22  Z 2p  Z 2m Z 2q  I 
 2  21  2
          
     
  =  Z p1
E p Z p2  Z pp  Z pm Z pq  Ip  (5)
          
     
E
 m Z m1 Z m2  Z mp  Z mm Z mq  I m 
     
      20
 E q   Z q1 Z q2  Z qp  Z qm Z qq   I q 
To obtain the elements in k th column of Z bus , k  1, ...., m ; k  q , 1 p.u.
current is injected at bus k , other bus currents are kept zero, the voltages are
measured at buses 1,…., m. These values remain same as those in partial
network.
Hence it can be concluded that while finding the new bus impedance matrix, it
is required to compute only the elements in the new row and column alone,
other elements remain unaltered.

It is assumed that the network consists of bilateral passive elements. Then

Z i q  Z q i for i = 1,2,….,m.
To compute Z q i :
The element Z q i can be determined by calculating the voltage at the q th bus
w.r.t. reference bus, by injecting 1 p.u. current at the i th bus and keeping all
other buses open circuited as shown in Fig. 6.

21
 1

2


Partial p q

network

i
Ep
1 p.u.
Eq

Fig. 6 0
Now the bus currents are

0
 
 
0
 
I bus  1 i  (6)
0
 
  22
0
 
Corresponding to this bus current, from eqn. (5) we get

E1  Z 1i
E2  Z 2i

Ep  Zpi (7)

Em  Zm i
Eq  Zqi

There is no current flow in the added element. It is also assumed that there is
no mutual coupling between the added element and the elements in the partial
network. Then

Eq = Ep (8)

23
Using eqn. (8) in eqn. (7) we get

Zq i  Zpi

Thus Zq i  Zpi i  1,2,........., m (9)

iq
If p happens to be reference bus, then E p = 0 and hence Z p i  0 . Thus
Z qi  0 (10)

To compute Z q q

The element Z q q can be calculated by measuring the voltage at bus q due to

current injection of 1 p.u. at bus q and keeping the other buses open circuited,
as shown in Fig. 7.

24
 1

2


Partial  p q
network zα


 1 p.u.



Fig. 7 0

Now the bus currents are

0
 
 
I bus    (11)
 
0
1q 
  25
Using the above in the performance eqn. (5), we get
E1  Z 1q
E2  Z 2q

Ep  Zp q (12)

Em  Zmq
Eq  Zq q
The added element p-q has an impedance of z α and there is no mutual coupling
between the added element and the elements in the partial network. Then from
Fig. 7 it is clear that

Eq  z α  Ep i.e. Eq  Ep  z α

Using eqn. (12) in the above equation, we get

Z qq  Z pq  z α (13)

If p happens to be the reference bus, then E p  0 and hence Z pq = 0. Thus

Z qq  z α (14)
26
Addition of a link

If the added element p-q is a link, the procedure for recalculating the elements
of Z bus is to connect in series with the added element a voltage source e  as
shown in Fig. 8.
1

 i

 p
Partial
network

e
q

m
0
Fig. 8

27
This creates a fictitious bus ℓ which will be eliminated later. The voltage source
e  is selected such that the current through the added link is zero. Now the
performance equation is

 E1   Z 11 Z 12  Z 1p  Z 1m Z 1   I1 
E  Z Z 22  Z 2p  Z 2m Z 2  I 
 2  21  2
          
     
E
 p =  Z p1 Z p2  Z pp  Z pm Z p  Ip  (15)
          
     
E m  Z m1 Z m2  Z mp  Z mm Z m  I m 
e  Z Z 2  Z p  Z m Z   I 
   1  

To determine Z  i
The element Z  i can be determined by calculating the voltage at the  th bus
w.r.t. to bus q when 1 p.u. current is injected into bus i and other buses are
open circuited as shown in Fig. 9. For this condition, the bus current vector is

28
 1


i

p 0
Partial  
network  
 0
1 p.u.  
I bus  1 i  (16)
e
0
q  
E  
0
m Eq  

0
Fig. 9

Substituting eqn. (16) in eqn. (15), performance equation reduces to

29
E1  Z1i
E2  Z2i

Ep  Zpi (17)

Em  Zmi
e  Z i

As there is no current in p-  , E q  e   E p i.e.

e  Ep  Eq (18)

Using eqn. (17) in the above equation

Z i  Zpi  Zqi i  1,2,......, m (19)

If p happens to be reference bus E p  0 and hence Z p i  0 . Thus

Z i   Zqi i  1,2,......, m (20)

30
To determine Z  
The element Z   can be calculated by measuring the voltage at the  th bus
w.r.t. bus q after injecting 1 p.u. current at the  th bus and keeping other buses
open circuited as shown in Fig. 10. Now the bus current is


i

p 0
Partial  
zα  
network
 0
 
I bus    (21)
e 1 p.u.
0
q  
 
m 1 
 

0
Fig. 10 31
Substituting eqn. (21) in eqn. (15), performance equation reduces to

E1  Z1
E2  Z 2

Ep  Zp
 (22)
Eq  Zq

Em  Zm
e  Z 

From the Fig. 10 it is clear that E q  e   z α  E p i.e.

e  Ep  Eq  z α (23)
Substituting eqn. (22) in the above equation we get

Z   Zp   Zq   zα (24)

If p happens to be reference bus, E p  0 and hence Z p   0 . Thus

32
Z    Zq   zα (25)
Elimination of bus 

Once we have calculated Z  i i = 1,2,……..,m and Z   we have an

( m+1) x ( m+1) impedance matrix. However, we know that eventually we must
connect  and q together. This will make the resultant bus impedance matrix of
dimension m x m . Equation (15) can be written as

 _

E bus  Z bus Z i   I bus  i = 1,2,….,m
 e  = (26)
    _   I   j = 1,2,….,m
 Z  j Z  
_
Therefore E bus  Z bus I bus  Z i  I  (27)
_
e   Z  j I bus  Z   I  (28)
If we join  and q, e  = 0. Then we can solve for I  from eqn. (28). Thus
_
Z  j I bus
I   . Substituting this in equation (27) we get
Z 

33
 _ _
Zi  Zj
E bus  Z bus I bus  I bus
Z 
_ _
Zi  Zj
= [ Z bus  ] I bus
Z 
_ _
Zi  Zj
Thus Z bus ( modified )  Z bus ( before modification )  (29)
Z 
_ _
Note that Z i  Z  j will be a m x m matrix. Elements of modified bus impedance
matrix can be obtained as

Z i  Z j i = 1,2,….,m
Zi j ( modified )  Z i j ( before modification )  (30)
Z  j = 1,2,….,m

34
Summary of formulas

p is not reference bus p is the reference bus

Added
element

p–q Zq i  Zp i Zq i  0
i  1,2,......., m i  1,2,......., m
is a
iq iq

Branch Zqq  Zpq  zα Zqq  zα

Added Z i  Zp i  Zq i Z i   Zq i
i  1,2,......., m i  1,2,......., m
element
i i

p–q Z   Zp   Zq   z α Z    Zq   zα

is a
Zi  Zj
Link Zi j ( modified )  Z i j ( before modification ) 
Z 
i  1,2,.........., m and j  1,2,.........., m
35
Example 1

Consider the power system shown in Fig. 11. The values marked are p.u.
impedances. The p.u. reactances of the generator 1 and 2 are 0.15 and 0.075
respectively. Compute the bus impedance matrix of the generator – transmission
network.

1 j 0.1 2

G1 G2

G1 j 0.1 j 0.1 G1
3

Fig. 11

36
Solution
The ground bus is numbered as 0 and it is taken as reference bus. The p.u.
impedance diagram is shown in Fig. 12.

1 j 0.1 2

j0.15 j 0.1 j 0.1 j0.075

3
0 0
Fig. 12

37
When element 0 – 1 is included

1 1 2
1 0.15 0 
Z bus  j 1  0.15 ; When element 0 – 2 is included Z bus  j 
2  0 0.075

Element 1 – 2 is added ; it is a link between buses 1 and 2. With bus 

Eliminating the  th bus

1 2 
1 2
1 0.15 0 0.15 
 0  ; 1  0.08077 0.034615
Z bus  j 2  0.075  0.075  Z bus j  
2 0.034615 0.05769 
 0.15  0.075 0.325 

38
Add element 1 – 3. It is a branch from bus 1 and it creates bus 3.

1 2 3
1  0.08077 0.034615 0.08077 
Z bus  j 2 0.034615 0.05769 0.034615
 
3  0.08077 0.034615 0.18077 

Finally add element 2 – 3. It is a link between buses 2 and 3. With bus 

1 2 3 
1  0.08077 0.034615 0.08077  0.046155
 0.034615 0.05769 0.034615 0.023075 
2  
Z bus  j
3  0.08077 0.034615 0.18077  0.146155
 
   0.046155 0.023075  0.146155 0.26923 

Eliminating the  th bus, final bus impedance matrix is obtained as

1 2 3
1 0.07286 0.03857 0.05571
Z bus  j 0.03857 0.05571 0.04714
2  
3  0.05571 0.04714 0.10143 39
Consider the circuit shown.
1 2 3
J 0.15 J 0.12

J 0.1

0.1 0.1 0.1 

Its bus impedance is Z bus  j 0.1 0.25 0.25 
0.1 0.25 0.37 
For the circuit
1 2 3
J 0.15 J 0.12

J 0.8
J 0.1

0.1 0.1 0.1   0   0.08 

0.1 0.25 0.25  j 0.8    0.2 
Vbus  j      
0.1 0.25 0.37   0    0.2 

40
 1 2 3
For this circuit J 0.15 J 0.12

J 0.1
 V 1  1.0 
+  V   1.0 
1.0  2  
-  V3  1.0 

For this combined circuit

1 2 3
J 0.15 J 0.12

J 0.8
J 0.1
+
1.0  V 1  1.0  0.08  0.92 
-  V   1.0    0.2    
 2      0.8 
 V3  1.0   0.2   0.8 

41
3 SYMMETRICAL FAULT ANALYSIS USING Z bus MATRIX

In classical method of symmetrical fault analysis, Thevenin’s impedance is to be

calculated. This impedance is the equivalent impedance between the fault point
and the reference bus. This will call for network reduction if the network is
large. Practical systems are very large and the network reduction will be very
tedious. In such cases, the bus impedance matrix Z bus can be readily used for
the fault analysis.

In this method, once the bus impedance matrix is constructed, with a very few
calculations, bus voltages and currents in various elements can be computed
quickly. When faults are to be simulated at different buses, this method proved
to be good.

Symmetrical short circuit analysis essentially consists of determining the steady

state solution of linear network with balanced sources.

42
Normally, in the short circuit study the following assumptions are made.
1. all the shunt parameters like loads, line charging admittances etc. are
neglected.
2. all the transformer taps are at nominal position.
3. prior to the fault, all the generators are assumed to operate at rated
voltage of 1.0 p.u. with their emf’s in phase.

With these assumptions, in the prefault condition, there will not be any
current flow in the network and all the bus voltages will be equal to 1.0 p.u.

The linear network that has to be solved comprises of

i) Transmission network
ii) Generation system and
iii) Fault
By properly combining the representations of the above three components, we
can solve the short circuit problem

43
Consider the transmission network shown in Fig. 13. Taking the ground as the
reference bus, the bus admittance matrix is obtained as
1 2 3
1 3
y2 y 1  y 2  y1  y2 
1
Ybus =  y y1  y 3  y 3 
2  1

3   y 2  y3 y 2  y 3 
y1 y3
2
Fig. 13
If we add all the columns ( or rows ) we get a column ( or row ) of all zero
elements. Hence this Ybus matrix is singular and hence corresponding Z bus matrix
of this transmission network does not exist. Thus, when all the shunt parameters
are neglected, Z bus matrix will not exist for the transmission network.

44
However, connection to ground is established at the generator buses, representing
the generator as a constant voltage source behind appropriate reactance as shown
in Fig. 14.
1 3

Fig. 14

45
As stated earlier, there is no current flow in the network in the pre-fault
condition. During short circuit, injection of bus current arises due to fault.
When such injection is there, then there will be current flow in different lines
and bus voltages will no longer remain at the pre-fault values.

Consider the network shown in Fig. 14. Symmetrical fault occurring at bus 2 can
be simulated by closing the switch shown in Fig. 15. Here Z F is the fault
impedance.

1 3

ZF

Fig. 15
46
Any general power system with a number of generators and N number of buses
subjected to symmetrical fault at p th bus will be represented as shown in Fig. 16.
 Vf 


Transmission
 Vf   Network p


 Vf 
ZF
Fig. 16

In the faulted system there are two types of sources:

1. Current injection at the faulted bus

2. Generated voltage sources.

The bus voltages in the faulted system can be obtained using Superposition
theorem. 47
Bus voltages due to current injection:

Make all the generator voltages to zero. Then we have Generator-Transmission

system without voltage sources. Such network has transmission parameters and
generator reactances between generator buses and the ground. Let Z bus be the
bus impedance matrix of such Generator-Transmission network. Then the bus
voltages due to the current injection will be given by

Vbus  Z bus I bus ( F ) (31)

where I bus ( F ) is the bus current vector having only one non-zero element. Thus
when the fault is at the p th bus

 0 
  
 
 0 
 
I bus ( F )  I p ( F )  Here I p ( F ) is the faulted bus current. (32)
 0 
 
  
 0 
  48
Thus, bus voltages due to current injection will be

 Z 11 Z 12  Z 1p  Z 1N   0   Z 1p 
Z Z 22  Z 2p  Z 2N   0  Z 
 21    2p 
           
Vbus      =   Ip (F) (33)
 Z p1 Z p2  Zpp  Zp N  I p ( F )  Z
 pp 
           
     
Z N1 Z N2  ZNp  Z N N   0  Z
 N p 

Bus voltages due to generator voltages

Make the fault current to be zero. Since there is no shunt element, there will be
no current flow and all the bus voltages are equal to V0 , the pre-fault voltage
which will be normally equal to 1.0 p.u. Thus, bus voltages due to generator
voltages will be

49
 1
1
 
 
Vbus    V0 (34)
1
 
 
1
Thus for the faulted system, wherein both the current injection and generator
sources are simultaneously present, the bus voltages can be obtained by adding
the voltages given by eqns. (33) and (34). Therefore, for the faulted system the
bus voltages are

 V1 ( F )   Z 1p  1
V  Z  1
 2( F )   2p   
       
Vbus ( F )    =   Ip (F) +   V0 (35)
V
 p(F)  Z
 pp  1
       
     
V
 N ( F )  Z
 N p  1

To calculate Vbus ( F ) we need the faulted bus current I p ( F ) which can be

50
determined as discussed below.
The fault can be described as shown in Fig. 17.

p
Ip (F)

VF
IF
Vp ( F )

ZF

Fig. 17
Here Vp ( F ) and I p ( F ) are the faulted bus voltage and current respectively.
Further VF , I F and Z F are the fault voltage, current and impedance respectively.

It is clear that VF  Z F I F , Vp ( F ) = VF and Ip (F) = - IF (36)

Therefore
Vp ( F )   Z F I p ( F ) (37)
51
The p th equation extracted from eqn. (35) gives

Vp ( F )  Z p p I p ( F )  V0 (38)

Substituting eqn. (37) in the above, we get

 Z F I p ( F )  Z p p I p ( F )  V0

Thus the faulted bus current I p ( F ) is given by

V0
Ip (F) =  (39)
Zpp  ZF

Substituting the above in eqn. (37), the faulted bus voltage Vp ( F ) is

ZF
Vp ( F ) = V0 (40)
Zpp  ZF

Finally voltages at other buses at faulted condition are to be obtained. The i th

equation extracted from eqn. (35) gives
52
Vi ( F )  Z i p I p ( F )  V0

Substituting eqn. (39) in the above, we get

Zip
Vi ( F )  V0  V0 i  1,2,.........., N (41)
Zpp  ZF
ip
Knowing all the bus voltages, current flowing through the various network
elements can be computed as

i k m ( F )  ( Vk ( F )  Vm ( F ) ) y k m where y k m is the admittance of element k  m . (42)

When the fault is direct, Z F  0 and hence

V0 Zip i  1,2,.........., N
Ip (F) =  , Vp ( F ) = 0 and Vi ( F )  V0  V0 (43)
Zpp Zpp ip

It is to be noted that when the fault occurs at the p th bus, only the p th column
of Z bus matrix ( and not the entire Z bus matrix ) is required for further
calculations.
53
The following are the various steps for conducting symmetrical short circuit
analysis.

Step 1 Read i) Transmission line data

ii) Generator reactances data
iii) Faulted bus number p and
iv) Fault impedance Z F .

Step 2 Construct the bus impedance matrix of the transmission network

including the generator reactances.

V0
Step 3 Compute I p ( F ) from I p ( F ) = 
Zpp  ZF

ZF
Step 4 Compute Vp ( F ) from Vp ( F ) = V0
Zpp  ZF

Zip i  1,2,.........., N
Step 5 Compute Vi ( F ) from Vi ( F )  V0  V0
Zpp  ZF ip

Step 6 Calculate the element currents from i k m ( F )  ( Vk ( F )  Vm ( F ) ) y k m 54

Example 2
Consider the power system discussed in Example 1. The p.u. impedances are on
a base of 50 MVA and 12 kV. Symmetrical short circuit occurs at bus 3 with
zero fault impedance. Using Z bus matrix determine the fault current, bus voltages
and also the currents contributed by the generators.

Solution
As seen in example 1, Z bus matrix of the transmission-generator network is
1 2 3
1 0.07286 0.03857 0.05571
Z bus  j 0.03857 0.05571 0.04714 Faulted system is shown in Fig. 18.
2  
3 0.05571 0.04714 0.10143

1 2

j 0.15 3 j 0.075
+ +
1.0 p.u. 1.0 p.u.
_ Fig. 18 _ _

55
 1
Faulted bus current I 3 ( F )    j9.8590 p.u.
j0.10143
Fault current I F   I 3 ( F )   j9.8590 p.u.
50 x 1000
Base current = = 2405.6 amp.
3 x 12
Fault current I F = - j 9.8590 x 2405.6 = - j 23717 amp.
Since fault impedance is zero, V3 ( F )  0
Taking the pre-fault voltage V0  1.0 p.u.
Z13 0.05571 12
V1 ( F )  1.0   1.0   0.45075 p.u. = x 0.45075  3.1229 kV
Z33 0.10143 3
Z23 0.04714 12
V2 ( F )  1.0   1.0   0.53525 p.u. = x 0.53525  3.7083 kV
Z33 0.10143 3
1.0  0.45075
Current supplied by gen. 1, I G 1    j3.6617 p.u. = - j8808.6 amp.
j0.15
1.0  0.53525
Current supplied by gen. 2, I G 2    j6.1967 p.u. = -j14906.8 amp.
j0.075
Note that in this example as the fault occurs at bus 3 we used only the column
3 of Z bus matrix and not the entire Z bus matrix.
56
Example 3

For the transmission-generator system shown in Fig.19, the bus impedance

matrix is obtained as

1 j 0.1 2

j0.15 j 0.1 j 0.1 j0.075

3
0 0
Fig. 19

1 2 3
1 0.072857 0.038571 0.055714 
Z bus  j  0.038571 0.055714 0.047143 
2  
3 0.055714 0.047143 0.101429 

Symmetrical three phase fault with fault impedance j 0.052143 p.u. occurs
at bus 1. Find the p.u. currents in all the elements and mark them on the
single line diagram. 57
Solution
Fault occurs at bus 1 and we need the first column of Z bus , which is

1
1 0.072857 
j  0.038571 and ZF = j 0.052143
2  
3 0.055714 

1 1
Faulted bus current I1(F) =    j 8 p.u.
j 0.072857  j 0.052143 j 0.125
Fault current IF = - I1(F) = - j 8 p.u.

VF = ZF IF = ( j 0.052143 ) ( - j 8) = 0.41714 p.u.

V1 (F) = VF = 0.41714 p.u.
0.038571
V2(F) = 1   0.69143 p..u.
0.125
0.055714
V3(F) = 1   0.55429 p..u.
0.125 58
i2-1 = (0.69143 – 0.41714) / (j 0.1) = - j 2.7429 p.u.
V1(F) = 0.41714 p.u.
i3-1 = (0.55429 – 0.41714) / (j 0.1) = - j 1.3715 p.u.
V2(F) = 0.69143 p.u.
i2-3 = (0.69143 – 0.55429) / (j 0.1) = - j 1.3714 p.u.
V3(F) = 0.55429 p.u.
iG1 = ( 1 – 0.41714 ) / ( j 0.15 ) = - j 3.8857 p.u.

iG2 = ( 1 – 0.69143 ) / ( j 0.075 ) = - j 4.1143 p.u.

Currents are marked in Fig. 20.

1 - j 2.7429 2

- j 4.1143
- j 3.715 - j 1.3714
- j3.8857
3
+ ZF -j8 + 1.0
1.0
- - j 3.715

- - j 3.715

Fig. 20
59
SHORT CIRCUIT MVA

The Short Circuit MVA is also known as FAULT LEVEL or FAULT MVA. The
circuit breaker breaking capacity must be equal to or greater than the three
phase fault MVA. The estimation of circuit breaker capacity is made on the
basis that it must clear a three phase fault as that is generally the worst
case. By simulating three phase fault at a point, short circuit level at that
point can be computed as

Short Circuit MVA =prefault voltage in p.u.x fault current in p.u. x Base MVA

Unless it is given otherwise prefault voltage shall be taken as 1.0 p.u.

60
Example 4

For the transmission-generator system shown in Fig.21, the bus impedance

matrix is obtained as

1 j 0.1 2

j0.15 j 0.1 j 0.1 j0.075

3
0 0
Fig. 19

1 2 3
1 0.072857 0.038571 0.055714 
Z bus  j  0.038571 0.055714 0.047143 
2  
3 0.055714 0.047143 0.101429 

The p.u. impedances are on a base of 50 MVA and 12 kV. Obtain the
fault level at bus 1 and 3. 61
Solution

1 2 3
1 0.072857 0.038571 0.055714 
Z bus  j  0.038571 0.055714 0.047143 
2  
3 0.055714 0.047143 0.101429 

When three phase fault occurs at bus 1:

1
Fault current =  13.7255 p.u.
0.072857

Fault level = 1 x 13.7255 x 50 = 686 MVA

When three phase fault occurs at bus 3:

1
Fault current =  9.8591 p.u.
0.101429
Fault level = 1 x 9.8591 x 50 = 493 MVA

62