POWER SYSTEM STABILITY

Introduction

Stability of a power system is its ability to return to normal or stable operating

conditions after having been subjected to some form of disturbance. Conversely,
instability means a condition denoting loss of synchronism or falling out of step.

Though stability of a power system is a single phenomenon, for the purpose of

analysis, it is classified as Steady State Analysis and Transient Stability.

Increase in load is a kind of disturbance. If increase in loading takes place

gradually and in small steps and the system withstands this change and
performs satisfactorily, then the system is said to be in STADY STATE
STABILITY. Thus the study of steady state stability is basically concerned with
ng before losing synchronism,
provided the loading is increased gradually at a slow rate.
In practice, load change may not be gradual. Further, there may be sudden
disturbances due to

i) Sudden change of load

ii) Switching operation
iii) Loss of generation
iv) Fault

Following such sudden disturbances in the power system, rotor angular

differences, rotor speeds, and power transfer undergo fast changes whose
magnitudes are dependent upon the severity of disturbances. For a large
disturbance, changes in angular differences may be so large as to cause the
machine to fall out of step. This type of instability is known as TRANSIENT
INSTABILITY. Transient stability is a fast phenomenon, usually occurring within
one second for a generator close to the cause of disturbance.
Short circuit is a severe type of disturbance. During a fault, electrical powers
from the nearby generators are reduced drastically, while powers from remote
generators are scarily affected. In some cases, the system may be stable even
with sustained fault; whereas in other cases system will be stable only if the fault
is cleared with sufficient rapidity. Whether the system is stable on the occurrence
of a fault depends not only on the system itself, but also on the type of fault,
location of fault, clearing time and the method of clearing.

Transient stability limit is almost always lower than the steady state limit and
hence it is much important. Transient stability limit depends on the type of
disturbance, location and magnitude of disturbance.

Review of mechanics

Transient stability analysis involves some mechanical properties of the machines

in the system. After every disturbance, the machines must adjust the relative
angles of their rotors to meet the condition of the power transfer involved. The
problem is mechanical as well as electrical.
The kinetic energy of an electric machine is given by

1 2
K.E. = I Mega Joules (1)
2

where I is the Moment of Inertia in Mega Joules sec.2 / elec. deg.2

Angular Momentum M = I ; Then from eqn. (1), K.E. can be written as

1
K.E. = M Mega Joules (2)
2

The angular momentum M depends on the size of the machine as well as on its
type.

The Inertia constant H is defined as the Mega Joules of stored energy of the
machine at synchronous speed per MVA of the machine. When so defined, the
relation between the Angular Momentum M and the Inertia constant H can be
derived as follows.
Relationship between M and H

Stored energy in MJ
By definition H =
Machine' s rating in MVA

Let G be the rating of the machine in MVA. Then

Stored energy = G H MJ (3)

1 1
Further, K.E. = M MJ = M MJ = M x f MJ (4)
2 2

From eqns. (3) and (4), we get

GH
M= MJ sec. / elec. rad. (5)
f

If the power is expressed in per unit, then G = 1.0 per unit and hence

H
M= (6)
f

While the angular momentum M depend on the size of the machine as well as on
its type, inertia constant H does not vary very much with the size of the machine,
The quantity H has a relatively narrow range of values for each class of machine.
Swing equation

The differential equation that relates the angular momentum M, the acceleration
power Pa and the rotor angle is known as SWING EQUATION. Solution of swing
equation will show how the rotor angle changes with respect to time following a
disturbance. The plot of Vs t is called the SWING CURVE. Once the swing curve
is known, the stability of the system can be assessed.

The flow of mechanical and electrical power in a generator and motor are shown
in Fig. 1.

Pe Pe
Te Ts
Generator Motor
Pm Pm
Ts Te
(a) Fig. 1 (b)

Consider the generator shown in Fig. 1(a). It receives mechanical power P m at the
shaft torque Ts -mover. It
delivers electrical power Pe to the power system network via. the bus bars. The
generator develops electromechanical torque Te in opposition to the shaft torque
Ts. At steady state, Ts = Te.
Assuming that the windage and the friction torque are negligible, in a generator,

accelerating torque acting on the rotor is given by

Ta = Ts Te (7)

Pa = Ps Pe (8)

In case of motor

Ta = Te Ts (9)

Pa = Pe Ps (10)

In general, the accelerating power is given by

Pa = Input Power Output Power (11)

 d2
Pa = T a I =M
dt 2

d2
Thus M = Pa (12)
dt 2
Here = angular displacement (radians)

d
= angular velocity (rad. / sec.)
dt

d d2
= = = angular acceleration
dt dt 2
Now we can see how the angular displacement can be related to rotor angle .

Consider an object moving at a linear speed of vs

displacement at any time t. For this purpose, introduce another object moving
with a constant speed of vs. Then, at any time t, the displacement of the first
object is given by

x = vs t + d

where d is the displacement of the first object wrt the second as shown in Fig. 2.
x
vs Fig. 2
d
vs
Similarly in the case of angular movement, the angular displacement , at any
time t is given by

s t + (13)

where is the angular displacement of the rotor with respect to rotating

reference axis which rotates s. The angle is also called

as LOAD ANGLE or TORQUE ANGLE. In view of eqn.(13)

d d
s + (14)
dt dt

d2 d2
2
= (15)
dt dt 2

From equations (12) and (15), we get

d2
M 2
= Pa (16)
dt

The above equation is known as SWING EQUATION

In case damping power is to be included, then eqn.(16) gets modified as

d2 d
M 2
+D = Pa (17)
dt dt

Swing curve, which is the plot of torque angle vs time t, can be obtained by
solving the swing equation. Two typical swing curves are shown in Fig. 3.

d
0
. dt

0
0

t t
(a) (b)
Fig. 3
Swing curves are used to determine the stability of the system. If the rotor angle
reaches a maximum and then decreases, then it shows that the system has
transient stability. On the other hand if the rotor angle increases indefinitely,
then it shows that the system is unstable.
We are going to study the stability of (1) a generator connected to infinite bus
and (2) a synchronous motor drawing power from infinite bus.

We know that the complex power is given by

P + j Q = V I* i.e. P j Q = V* I Thus real power P = Re {V * I}

Consider a generator connected to infinite bus.

V is the voltage at infinite bus.
Xd XT
E is internal voltage of generator.

X is the total reactance

E I V
Taking this as ref. V = V 00

V+jXI=E phasor dia. can be obtained as

E Internal voltage E leads V

jX I Thus E = E
V
I 1
Current I = [ E cos j E sin V ]
jX

E V
Electric output power Pe = Re [ V I ] = sin = Pmax
X
Consider a synchronous motor drawing power from infinite bus.

Xd XT
V jXI = E
V
E I V
-jXI
I
E

1
Thus E = E Current I = [ V ( E cos j E sin ) ]
jX

E V
Electric input power Pe = Re [ V I ] = sin = Pmax
X
Thus Swing equation for alternator is

d2
M = Pm Pmax
dt 2
Swing equation for motor is

d2
M = Pmax - Pm
dt 2
Notice that the swing equation is second order nonlinear differential equation
Equal area criterion

The accelerating power in swing equation will have sine term. Therefore the
swing equation is non-linear differential equation and obtaining its solution is not
simple. For two machine system and one machine connected to infinite bus bar,
it is possible to say whether a system has transient stability or not, without
solving the swing equation. Such criteria which decides the stability, makes use
of equal area in power angle diagram and hence it is known as EQUAL AREA
CRITERION. Thus the principle by which stability under transient conditions is
determined without solving the swing equation, but makes use of areas in power
angle diagram, is called the EQUAL AREA CRITERION.

From the Fig. 3, it is clear that if the rotor angle oscillates, then the system is
stable. For to oscillate, it should reach a maximum value and then should
d
decrease. At that point = 0. Because of damping inherently present in the
dt
system, subsequence oscillations will be smaller and smaller. Thus while
d
changes, if at one instant of time, = 0, then the stability is ensured.
dt
 d
Let us find the condition for to become zero.
dt

The swing equation for the alternator connected to the infinite bus bars is

d2
M = Ps Pe (18)
dt 2

d
Multiplying both sides by , we get
dt

d2 d d 1 d d 2 d
M = (Ps Pe) i.e. M ( ) (Ps Pe) (19)
dt 2 dt dt 2 dt dt dt

Thus

d d 2 dt 2 (Ps Pe ) d d 2 2 (Ps Pe )
( ) ; i.e. ( ) On integration
dt dt d M d dt M

d 2 2 (Ps Pe ) d d 2 (Ps Pe ) d (20)

( ) = i.e. =
dt 0
M dt 0
M

d
Before the disturbance occurs, 0 was the torque angle. At that time = 0. As
dt
d
soon as the disturbance occurs, is no longer zero and starts changing.
dt
Torque angle will cease to change and the machine will again be operating at
d
synchronous speed after a disturbance, when = 0 or when
dt

2 (Ps Pe )
d = 0 i.e.
0
M

(Ps Pe ) d = 0 (21)
0

If there exist a torque angle for which the above is satisfied, then the machine
will attain a new operating point and hence it has transient stability.

The machine will not remain at rest with respect to infinite bus at the first time
d
when = 0. But due to damping present in the system, during subsequent
dt
oscillation, maximum value of keeps on decreasing. Therefore, the fact that
has momentarily stopped changing may be taken to indicate stability.
Sudden load increase on Synchronous motor

Let us consider a synchronous motor connected to an infinite bus bars.

- E jX V
+
E=V I ( jX )
I

E V
E Pe sin Pmax sin
X
I

P
Input power

Pe= Pmax sin

c

Ps d b A2 e Changed output

A1

a Initial output
P0

0 s m
Fig. 4
 P Input power

c Pe= Pmax sin

V
Ps d b A2 e Changed output

A1 E

Initial output I
P0 a
Fig. 4

0 s m

The following changes occur when the load is increased suddenly.

Point a Initial condition; Input = output = P0 s 0

Due to sudden loading, output = Ps; output > Input;

s 0.

Between a-b Output > Input; Rotating mass starts loosing energy resulting
d

Point b min s s

s
 P Input power

Pe= Pmax sin

c

V
Ps d b A2 e Changed output

A1 E

Initial output I
P0 a

0 s m

Between b-c Input > output; Rotating masses start gaining energy;

Point c Input s m

s m.

Between c-b Input > output;

Point b max s s,

Between b-a max ; but

Point a s 0; Output > Input; The cycle repeats.

Because of damping present in the system, subsequent oscillations become
smaller and smaller and finally b will be the steady state operating point.

Interpretation of equal area

As discussed earlier (eqn. 21), the condition for stability is

P
(Ps Pe ) d = 0 i.e. Pe d = Ps d Input power
0 0 0

c Pe= Pmax sin

m

From Fig. 4, Pe d 0 m
0
Ps d b A2 e Changed output

A1
m

and Ps d 0 m
a Initial output
0 P0

Thus for stability,

0 m 0 m 0 s m
Fig. 4
Subtracting 0 m from both sides of above equation, we get A2 = A1.
Thus for stability,

A2 = A1 (22)
Fig. 5 shows three different cases: The one shown in case a is STABLE. Case b
indicates CRITICALLY STABLE while case c falls under UNSTABLE.

P P P
A2
A2 Ps
Ps
A2 A1 A1
Ps
A1

P0 P0 P0

0 m 0 s m 0
s

Case a Case b Case c

Fig. 5

Note that the areas A1 and A2 are obtained by finding the difference

between INPUT and OUTPUT.

Example 1

A synchronous motor having a steady state stability limit of 200 MW is receiving

50 MW from the infinite bus bars. Find the maximum additional load that can be
applied suddenly without causing instability.

Solution
P
Referring to Fig. 6,
200
A2
for critical stability PS C D
A1 Electric Input
A2 = A1 power

50 B
200 0 = 50 i.e.
A E
50 -
0 = sin-1 0.25268 rad. 0 s S
200 Fig. 6
S = PS

Adding area ABCDEA to both A1 and A2 and equating the resulting areas
 S

S S 0) = 200 sin d i.e.

0

S 0 S 0 S 0 S i.e.

S S - S = 0.9682458

The above equation can be solved by trial and error method.

S 0.85 0.9 0.95

RHS 0.8718 0.9363 0.9954

Using linear interpolation S = 0.927 rad.

0.927 rad. = 53.11 deg.

Thus PS = 200 sin 53.110 = 159.96 MW

Maximum additional load possible = 159.96 50 = 109.96 MW

Opening of one of the parallel lines

When a generator is supplying power to an infinite bus over two parallel

transmission lines, the opening of one of the lines will result in increase in the
equivalent reactance and hence decrease in the maximum power transferred.
Because of this, depending upon the initial operating power, the generator may
loose synchronism even though the load could be supplied over the remaining
line under steady state condition.

Consider the system shown in Fig. 7. The power angle diagrams corresponding
to stable and unstable conditions are shown in Fig. 8.

G Fig. 7

Electric Output
power (2 lines) A2

Electric Output
A2
power (1 line) PS A1
PS
A1

0 s m

A2 = A1 ; Stable A2 < A1 ; Unstable

Fig. 8
 Short circuit occurring in the system

Short circuit occurring in the system often causes loss of stability even though
the fault may be removed by isolating it from the rest of the system in a relatively
short time. A three phase fault at one end of a double circuit line is shown in Fig.
9(a) which can be reduced as shown in Fig. 9(b).

+ + +
+ Ei Eg Ei
- - -
Eg Ig
-

(a) (b)
Fig. 9
It is to be noted that all the current from the generator flows through the fault and
this current Ig lags the generator voltage by 900. Thus the real power output of the
generator is zero. Normally the input power to the generator remains unaltered.
ncrease indefinitely
because entire the input power will be used for acceleration. This may result in
unstable condition.
When the three phase fault occurring at one end of a double circuit line is
disconnected by opening the circuit breakers at both ends of the faulted line,
power is again transmitted. If the fault is cleared before the rotor angle reaches a
particular value, the system will remain stable; otherwise it will loose stability as
shown in Fig. 10.

Pre-fault output
power

Post-fault output
power
A2 A2 A2
PS Input power
A1 A1 A1

0 C m 0 CC m 0 C
During fault
output power
C< CC C> CC

A2 = A1 ; Stable A2 = A1 ; Critically stable A2 < A1 ; Unstable

Fig. 10

Note that the areas A1 and A2 are obtained by finding difference between INPUT
and OUTPUT.
When a three phase fault occurs at some point on a double circuit line, other than
on the extreme ends, as shown in Fig. 11(a), there is some finite impedance
between the paralleling buses and the fault. Therefore, some power is transmitted
during the fault and it may be calculated after reducing the network to a delta
connected circuit between the internal voltage of the generator and the infinite
bus as shown in Fig. 11(b).

Xb

+ + +
Xa Xc
+ Em Eg Em
Eg - - -
-

(a) (b)
Fig. 11

Eg Em
Power transmitted during the fault = sin (23)
Xb
Stable, critically stable and unstable conditions of such systems are shown:

Pre-fault output power

C < CC;
Post-fault output power

A2 = A1; A2
Input power
STABLE
A1
During-fault
output power

0 C m
 Pre-fault output power

Post-fault output power

C = CC; A2
Input power
A2 = A1;
A1
CRITICALLY During-fault
STABLE output power

0 CC m

Pre-fault output power

Post-fault output power

C > CC;

A2 < A1;
A2 Input power
UNSTABLE
A1
During-fault
output power

0 C
Example 2

In the power system shown in Fig. 12, three phase fault occurs at P and the faulty
line was opened a little later. Find the power output equations for the pre-fault,
during fault and post-fault conditions.
0.16 0.16
Xg = 0.28 0.24 0.16

G
P
Eg
'
1.25 p.u. x
V 1.0 p.u.
0.24
0.16 0.16
Values marked are p.u. reactances
Solution Fig. 12

Pre-fault condition
0.16
0.56 0.72

0.28 0.56 + + +
1.0 1.25 1.0
+
1.25 - - -
-

1.25 x 1.0
Power output Pe = sin 1.736 sin
0.72
During fault condition:

0.16 0.56
0.56
0.16 0.4

0.28 + 0.28 0.16

+ 1.0 + 0.16 0.4 +
1.25 -
- - -

2.99
0.08 0.2 0.16
0.28
+ 1.25 1.0
+
1.25 0.057 1.0
- -

(0.36x0.36 + 0.36x0.057 + 0.057x0.36) / 0.057 = 2.99

1.25 x 1.0
Power output Pe = sin 0.418 sin
2.99
Post-fault condition:

0.16
0.56
1.0
0.28 +
1.0 1.25 1.0
+
1.25 -
-

1.25 x 1.0
Power output Pe = sin 1.25 sin
1.0

Thus power output equations are:

Pre-fault Pe = Pm 1 sin = 1.736 sin

During fault Pe = Pm 2 sin = 0.418 sin

Post fault Pe = Pm 3 sin = 1.25 sin

Here

Pm 1 = 1.736; Pm 2 = 0.418; Pm 3 = 1.25;

Expression for critical clearing angle CC

Pm1
Consider the power Pre-fault
angle diagrams output power
shown in Fig. 13
Pm3
Post-fault
A2 output power
Pm 1 = Ps PS
0 Input power
Pm2
Pm 3 sin = Ps A1
s
During-fault
output power

CC 0 s CC m - s

Area A1 = PS( CC 0 ) Pm 2 sin d Fig. 13

0

= PS CC PS 0 Pm 2 cos CC Pm 2 cos 0 (24)

Area A2 = Pm3 sin d - PS ( m CC )

CC

= Pm3 cos CC Pm3 cos m PS m PS CC (25)

A1 = PS CC PS 0 Pm 2 cos CC Pm 2 cos 0 (24)

A2 = Pm3 cos CC Pm3 cos m PS m PS CC (25)

Area A2 = Area A1

Pm3 cos CC Pm3 cos m PS m PS CC = PS CC PS 0 Pm 2 cos CC Pm 2 cos 0

( Pm 3 Pm 2 ) cos CC PS ( m 0 ) Pm 3 cos m Pm 2 cos 0

PS ( m 0 ) Pm 3 cos m Pm 2 cos 0
cos CC
Pm 3 Pm 2

Thus CRITICAL CLEARING ANGLE is given by

1
PS ( m 0 ) Pm3 cos m Pm2 cos 0
CC cos [ ] (26)
Pm3 Pm2

Here the angles are in radian. Further, since

Pm1 0 = Ps , Pm3 sin s = Ps m - s 0 m are given by

1 PS 1 PS
0 sin ( ) m sin ( ) (27)
Pm1 Pm 3
Example 3

In the power system described in the previous example, if the generator was
delivering 1.0 p.u. just before the fault occurs, calculate CC .

Solution

Pm 1 = 1.736; Pm 2 = 0.418; Pm 3 = 1.25; PS = 1.0

1.736 sin 0 = 1.0; sin 0 = 0.576; 0 = 0.6139 rad.

1.25 sin s = 1.0; sin s = 0.8; s = 0.9273 rad.; m = s = 2.2143 rad.

PS ( m 0 ) Pm 3 cos m Pm 2 cos 0
cos CC
Pm 3 Pm 2

1.0 (2.2143 0.6139) 1.25 cos 2.2143 - 0.418 cos 0.6139

= 0.6114
1.25 0.418

Critical clearing angle CC = 52.310

STEP BY STEP SOLUTION OF OBTAINING SWING CURVE

The equal area criterion of stability is useful in determining whether or not a

system will remain stable and in determining the angle through which the
machine may be permitted to swing before a fault is cleared. It does not
determine directly the length of time permitted before clearing a fault if stability is
to be maintained.

In order to specify a circuit breaker of proper speed, the engineer must know the
CRITICAL CLEARING TIME, which is the time taken by the machine to swing from
its initial position to its critical clearing angle. If the Critical Clearing Angle (CCA)
is determined by the equal area criterion, then to determine corresponding
Critical Clearing Time (CCT), the swing curve for the sustained faulted condition
is required.

The step by step method of obtaining swing curve, using hand calculation is
necessarily simpler than some of the methods recommended for digital
computer. In the method suitable for hand calculation,the period of interest is
divided into several short intervals. The change in the angular position of the
rotor during a short interval of time is computed by making the following
assumptions.
 1. The accelerating power Pa computed at the beginning of an interval is
constant from the middle of the proceeding interval to the middle of the
interval considered.

d
2. is constant throughout any interval at the value computed at the
dt
middle of the interval.

d
Above assumptions are made to approximate continuously varying P a and as
dt
stepped curve. Fig. 14 will help in visualizing the assumptions. The accelerating
power is computed for the points enclosed in circles, at the beginning of n-1, n
and n+1 th intervals. The step of Pa in the figure results from assumption 1.

d d
( = - s), the excess of angular velocity over the synchronous
dt dt
angular velocity is shown as a step curve that is constant throughout the interval,
at the value computed at the midpoint.

3 1
Between the ordinates n - and n -
2 2
caused by constant angular acceleration (caused by constant Pa).
 Pa
Pa (n 2)

Pa (n 1)

Pa (n)

n-2 n-1 n t
n th interval
n - 1 th interval

(n 1/2)

(n 3/2)

n 3/2 n 1/2 t
n

n-1

Fig. 14 n-1
n-2

n-2 n-1 n t
The is

1 3
(n - )- (n - ) = Constant angular acceleration x time duration
2 2

Pa ( n 1) d2 1
= ( Because = Pa ) (28)
M dt 2 M

x time
duration. Thus

1
(n) (n - (29)
2

3
(n 1) (n - (30)
2

1 3 Pa ( n 1) 2
Therefore (n) - (n 1) (n - )- (n - =
2 2 M

Pa ( n 1) 2
Thus (n) (n 1) + (31)
M
 Pa ( n 1) 2
(n) (n 1) + (31)
M

Equation (31) shows that the change in torque angle during a given interval is
equal to the change in torque angle during the proceeding interval plus the
( t)2
accelerating power at the beginning of the interval X .
M

at the end of nth interval can be computed as

(n) (n 1) (n) (32)

Pa ( n 1) 2
where (n) (n 1) + (33)
M
The above two equations are known as Recursive equations using which
approximate swing curve can be obtained.

The process of computation is now repeated to obtain P a (n) (n+1) (n+1). The
solution in discrete form is thus carried out over the desired length of time
normally 0.05 sec. Greater accuracy of solution can be achieved by reducing the
time duration of interval.
Any switching event such as occurrence of a fault or clearing of the fault causes
discontinuity in the accelerating power P a. If such a discontinuity occurs at the
beginning of an interval then the average of the values of P a just before and just
after the discontinuity must be used.

Thus in computing the increment of angle occurring during the first interval after
a fault is applied at time t = 0, becomes

1 - + ( t) 2
1 = 0 + (Pa 0 + Pa 0 )
2 M

2
1 + ( t)
= Pa 0 (Because Pa 0- = 0)
2 M
If the discontinuity occurs at the middle of an interval, no special procedure is
needed. The correctness of this can be seen from Fig. 15.

n th interval n th interval
n - 1 th interval n - 1 th interval

Discontinuity at the Discontinuity at the

beginning of an interval middle of an interval

Fig. 15
Example 4

A 20 MVA, 50 Hz generator delivers 18 MW over a double circuit line to an infinite

bus. The generator has KE of 2.52 MJ / MVA at rated speed and its transient
reactance is Xd = 0.35 p.u. Each transmission line has a reactance of 0.2 p.u. on a
20 MVA base. E = 1.1 p.u. and infinite bus voltage V = 1.0 p.u. A three phase

fault occurs at the mid point of one of the transmission lines. Obtain the swing
curve over a period of 0.5 sec. if the fault is sustained.

E = 1.1 p.u. 0.2 p.u.

0.2 p.u.

20 MVA 50 Hz Infinite bus

Delivers 18 MW V = 1.0 p.u.
Xd = 0.35 p.u.
H = 2.52 MJ/MVA

G = 20 MVA = 1.0 p.u.

GH 1.0 x 2.52 4
Angular momentum M = 2.8 x 10 sec2 / elec. deg.
180 f 180 x 50

( t) 2 (0.05)2 x 10 4
Let us choose Then = 8.929
M 2.8
Recursive equations are

(n) (n 1) (n)

(n) (n 1) + 8.929 Pa (n-1)

1.1 x 1.0
Pre fault: X = 0.45 p.u.; Pe = sin
0.45

During fault:
0.2
0.2
0.1 0.1

0.35 + 0.35 +
+ 1.0 + 0.1 0.1
1.1 -
- - -

1.25
Converting the star 0.35, 0.1 and 0.2 as delta
1.1 1.0
1.1 x 1.0
Pe = sin
1.25
Initial calculations:

Before the occurrence of fault, there will not be acceleration i.e.

Input power is equal to output power. Therefore

Input power Ps = 18 MW = 0.9 p.u.

Initial power angle is given by

2.44 sin 0 = 0.9; Thus 0 = 21.64

Pa 0- = 0; Pa 0+ = 0.9 0.88 sin 21.640 = 0.576 p.u.

Pa average = ( 0 + 0.576 ) / 2 = 0.288 p.u.

( t) 2
1 = 0 + Pa average x = 0.288 x 8.929 = 2.570
M

= 24.210

Subsequent calculations are shown below.

t sec. Pmax Pe Pa = 0.9-Pe 8.929 Pa
-
0 21.64 2.44 0.9 0
+
0 21.64 0.88 0.324 0.576

0 average 21.64 0.288 2.57 2.57

0.05 24.21 0.88 0.361 0.539 4.81 7.38

0.10 31.59 0.88 0.461 0.439 3.92 11.30

0.15 42.89 0.88 0.598 0.301 2.68 13.98

0.20 56.87 0.88 0.736 0.163 1.45 15.43

0.25 72.30 0.88 0.838 0.062 0.55 15.98

0.30 88.28 0.88 0.879 0.021 0.18 16.16

0.35 104.44 0.88 0.852 0.048 0.426 16.58

0.40 121.02 0.88 0.754 0.145 1.30 17.88

0.45 138.90 0.88 0.578 0.321 2.87 20.75

0.50 159.65
Swing
shown in Fig. 16.

Fig. 16
Example 5

In the power system considered in the previous example, fault is cleared by

opening the circuit breakers at both ends of the faulty line. Calculate the CCA and
hence find CCT.

Solution

From the previous example: Ps = 0.9; Pm1 = 2.44 and Pm2 = 0.88

For the Post fault condition:

1.1 x 1.0
X = 0.55 p.u; Pe = sin 2.0 sin
0.55

Thus Ps = 0.9; Pm1 = 2.44; Pm2 = 0.88; Pm3 = 2.0

PS ( m 0 ) Pm 3 cos m Pm 2 cos 0
cos CC
Pm 3 Pm 2

0 0 = 0.3778 rad.

s s = 0.4668 Therefore m = - s = 2.6748 rad

0.9 ( 2.6748 0.3778 ) 2 cos (2.6748) 0.88 cos (0.3778)

cos CC 0.47915
2 0.88

CC = 118.630
 Fig. 17

Referring to the swing curve obtained for sustained fault condition,

corresponding to CCA of 118.630, CCT can be obtained as 0.38 sec. as shown in
Fig. 17.
equations (DE)

Let us first consider solution of first order differential equation. Later we shall
extend it for solving a set of first order DE. The swing equation is a second order
DE which can be written as two first order DE and solution can be obtained using

Let the given first order DE be

dx
f ( t, x ) (34)
dt

where t is the independent variable and x is the dependent variable. Let (t 0, x 0)

t1=t0 2 =t1 n =tn 1

First estimate of x1 (value of x at time t1) is denoted as x 1(0). Then

dx
x 1(0) = x 0 + |0
dt
 (0)
Thus (t 1, x 1 ) is the first estimated point of (t 1, x 1). Second and the final
estimate of x 1 is calculated as

1 dx dx (0)
x1 = x0 + ( |0 + |1 ) (36)
2 dt dt

dx (0) dx
where |1 is the value of computed at (t 1, x 1(0)). Thus the next point
dt dt

(t 1, x 1) is now known. Same procedure can be followed to get (t 2, x 2) and it can

be repeated to obtain points (t 3, x 3), (t 4, x 4

Knowing (t n - 1, x n - 1 ), next point (t n, x n) can be computed as follows:

tn=tn 1 (37)

dx
x n(0) = x n 1 + |n 1 (38)
dt

dx (0) dx
Compute |n which is computed at (t n, x n(0)). (39)
dt dt

1 dx dx (0)
Then xn = xn-1 + ( |n - 1 + |n (40)
2 dt dt
Same procedure can be extended to solve a set of two first order DE given by

dx dy
= f1 (t, x, y) and = f2 (t, x, y)
dt dt

Knowing (t n - 1, x n - 1 , y n - 1), next point (t n, x n, yn) can be computed as follows:

tn=tn 1

dx
x n(0) = x n 1 + |n 1
dt

dy
y n(0) = y n 1 + |n 1
dt

dx (0) dx
Compute |n which is computed at (t n , x n(0), y n(0)) and
dt dt

dy (0) dy
|n which is computed at (t n , x n(0), y n(0))
dt dt

1 dx dx (0)
Then x n = x n - 1 + ( |n - 1 + |n
2 dt dt

1 dy dy (0)
yn = yn-1 + ( |n - 1 + |n
2 dt dt
We know that the swing equation is

d2
M = Pa
dt 2

H
M=
f

Therefore for a generator

d2 f f f
2
= Pa ( Ps Pe ) = K ( Ps Pe ) where K =
dt H H H

d2
The second order DE 2
= K ( Ps Pe )
dt

d
s
dt

d
= K ( Ps Pe )
dt
 d d
Note that generally of the form = f1
dt dt
d d
generally of the form = f2
dt dt

Just prior to the occurrence of the disturbance, P s Pe s. The rotor

Thus the initial point is

As soon as disturbance occurs, electric network changes and the expression for
electric power Pe
Pe shall be computed by the said expression.

1 1 can be computed. Thus we get the next

solution point as ( t1 1 1). The procedure can be repeated to get subsequent
solution points until next change in electric network, such as removal of faulted
line occurs. As soon as electric network changes, corresponding expression for
electric power need to be obtained and used in subsequent calculation.

The whole procedure can be carried out until t reaches the time upto which
transient stability analysis is required.
Example 6

An alternator rated for 100 MVA supplies 100 MW to an infinite bus through a line
of reactance 0.08 p.u. on 100 MVA base. The machine has a transient reactance of
0.2 p.u. and its inertia constant is 4.0 p.u. on 100 MVA base. Taking the infinite
bus voltage as reference, current supplied by the alternator is ( 1.0 j 0.6375 )
p.u.

Calculate the torque angle and speed of the alternator for a period of 0.14 sec.
when there is a three phase fault at the machine terminals and the fault is cleared

Solution
100 MVA
Ps = 100 MW
H=4 I 00

E = ( 1.0 + j0) + j 0.28 ( 1.0 j 0.6375) = 1.1785 + j 0.28 = 1.2113 13.36510 p.u.
0
= 0.2333 rad.
Shaft power Ps = 100 MW = 1.0 p.u. This remains same

throughout the calculations.

Just before the fault, Pe = Ps = 1.0 p.u.; Swing equation is:

d2 f 50
2
= ( Ps Pe ) = (1 Pe ) 39.2699 ( 1 - Pe )
dt H 4

The two first order DEs are:

d
314.1593 Initial point is:
dt

d
= 39.2699 ( 1 Pe )
dt

Since the fault is at the generator terminals, during fault P e = 0

d d
314.1593 39.2699 (1 Pe )
dt dt

Initial point is:

First estimate:
rad. / sec.
d
= 314.1593 314.1593 = 0
dt

d
= 39.2699 ( 1 0 ) = 39.2699
dt
x 0.02 ) = 0.2333 rad.

Second estimate: First estimated point is:

d
= 314.9447 314.1593 = 0.7854
dt

d
= 39.2699 ( 1 0 ) = 39.2699; Thus
dt

1
( 0 + 0.7854 ) x 0.02 = 0.24115 rad.
2

1
( 39.2699 + 39.2699 ) x 0.02 = 314.9447 rad. / sec.
2
d d
314.1593 39.2699 (1 Pe )
dt dt

(0.04 (0.04) Latest point is:

First estimate

d
= 314.9447 314.1593 = 0.7854
dt

d
= 39.2699 ( 1 0 ) = 39.2699
dt
= 0.24115 + ( 0.7854 x 0.02 ) = 0.2569 rad.
= 314.9447 + ( 39.2699 x 0.02 ) = 315.7301 rad. / sec.
Second estimate: First estimated point is:
d
= 315.7301 314.1593 = 1.5708
dt

d
= 39.2699 ( 1 0 ) = 39.2699; Thus
dt

1
(0.04) = 0.24115 + ( 0.7854 + 1.5708 ) x 0.02 = 0.2647 rad.
2

1
(0.04) = 314.9447 + ( 39.2699 + 39.2699 ) x 0.02 = 315.7301 rad. / sec.
2
Calculations can be repeated until the fault is cleared i.e. t = 0.1. The results are
tabulated. Thus

Once the fault is cleared, reactance between internal voltage and the infinite bus
is 0.28 and thus generator out put is;

1.2113 x 1.0
Pe = sin 4.3261 sin
0.28

In the subsequent calculation Pe must be obtained from the above equation.

(0.12 (0.12) Latest point is:

d d
314.1593 39.2699 (1 Pe )
dt dt

(0.12 (0.12) Latest point is:

First estimate
. / sec.
d
= 318.0869 314.1593 = 3.9276
dt

d
= 39.2699 ( 1 4.3261 sin 0.4297 rad. ) = - 31.5041
dt

- 31.5041 x 0.02 ) = 317.4568 rad. / sec.

Second estimate: First estimated point is:
d
= 317.4568 314.1593 = 3.2975
dt

d
= 39.2699 ( 1 4.3261 sin 0.50825 rad. ) = - 43.4047; Thus
dt
1
(0.12) = 0.4297 + ( 3.9276 + 3.2975 ) x 0.02 = 0.50195 rad.
2

1
(0.12) = 318.0869 + ( - 31.5041 - 43.4047 ) x 0.02 = 317.3378 rad. / sec.
2
Complete calculations are shown in the Table:

d d
314.1593; 39.2699 (1 Pe ); Pe = 0 for t < 0.1 and Pe = 4.3261 sin
dt dt

First Estimate Second Estimate

t
sec. rad/sec
rad/sec rad/sec
-
0 0.2333 314.1593
+
0 0.2333 314.1593 0 39.2699 0.2333 314.9447 0.7854 39.2699 0.2412 314.9447

0.02 0.2412 314.9447 0.7854 39.2699 0.2569 315.7301 1.5708 39.2699 0.2647 315.7301

0.04 0.2647 315.7301 1.5708 39.2699 0.2961 316.5155 2.3562 39.2699 0.304 316.5155

0.06 0.304 316.5155 2.3562 39.2699 0.3511 317.3009 3.1416 39.2699 0.359 317.3009

0.08 0.359 317.3009 3.1416 39.2699 0.4218 318.0863 3.927 39.2699 0.4297 318.0869

-
0.10 0.4297 318.0869

+
0.10 0.4297 318.0869 3.9276 -31.504 0.5083 317.4568 3.2975 -43.405 0.502 317.3378

0.12 0.502 317.3378 3.1785 -42.468 0.5655 316.4884 2.3291 -51.761 0.557 316.3955

0.14 0.557 316.3955

 t sec 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14

0.2333 0.24115 0.2647 0.304 0.359 0.4297 0.50195 0.5570

13.37 13.82 15.17 17.42 20.57 24.62 28.76 31.91

314.1593 314.9447 315.7301 316.5155 317.3009 318.0869 317.3378 316.3955

rad/sec
SOLUTION OF SWING EQUATION BY RUNGE KUTTA METHOD

Fourth order Runge Kutta (RK) method is one of the most commonly used
methods of solving differential equation.

Consider the first order DE

dx
f ( t, x )
dt

Let (t m, x m) be the initial point and h be the increment in time. Then

tm+1 = tm + h

Fourth order RK method can be defined by the following five equations.

1
xm+1 = xm + ( k1 + 2 k2 + 2 k3 + k4) where
6

h k
k 1 = f ( t m, x m ) h k2 = f ( t m + ,xm+ 1)h
2 2

h k
k3 = f ( t m + ,xm+ 2 )h k4 = f ( t m + h, x m + k3 )
2 2
Note that in this method, the function has to be evaluated four times in each step.
Same procedure can be extended to solve a set of first order DE such as

dx dy
f1 ( t , x , y ) and f2 ( t , x , y )
dt dt

Initial solution point is ( t m, xm, ym ). Then

1 1
xm+1 = xm + ( k1 + 2 k2 + 2 k3 + k4) ym+1 = ym + 1 2 3 4)
6 6

where

k1 = f1 ( t m, xm, ym 1 = f2 ( t m, x m, ym ) h

h k h k
k2 = f1 ( t m + , xm + 1 , ym + 1 2 = f2 ( t m + , xm + 1 , ym + 1 ) h
2 2 2 2 2 2

h k h k
k3 = f1 ( t m + , xm + 2 , ym + 2 3 = f2 ( t m + , xm + 2 , ym + 2 ) h
2 2 2 2 2 2

k4 = f1 ( t m + h, xm + k3, ym 3 4 = f2 ( t m + h, xm + k3, ym 3 )h
We know that the swing equation can be written as

d
s
dt

d f
= K ( Ps Pe ) where K =
dt H
th
The initial solution point is ( order RK method is used, k1,
1, k2 2, k3 3, k4 4 are computed and then the next solution point is obtained as

( t1 1 1). This procedure can be repeated to get subsequent solution points.

Example 7

Consider the problem given in previous example and solve it using 4 th order RK
method.

Solution

As seen in the previous example, two first order DEs are

d
314.1593 Initial point is:
dt

d
= 39.2699 ( 1 Pe )
dt

During the first switching interval t = 0+ to 0.1 sec. electric output power Pe = 0.
d d
314.1593 39.2699 (1 Pe ) Initial point is:
dt dt

k1 = (314.1593 314.1593) x 0.02 = 0 1 = 39.2699 ( 1 0 ) x 0.02 = 0.7854

) + k1 1 / 2 = 314.1593 + 0.3927 = 314.552

k2 = (314.552 314.1593) x 0.02 = 0.007854 2 = 39.2699 ( 1 0 ) x 0.02 = 0.7854

) + k2 / 2 = 0.2372 2 / 2 = 314.1593 + 0.3927 = 314.552

k3 = (314.552 314.1593) x 0.02 = 0.007854 3 = 39.2699 ( 1 0 ) x 0.02 = 0.7854

) + k3 = 0.2412; 3 = 314.1593 + 0.7854 = 314.9447

k4 = (314.9447 314.1593) x 0.02 = 0.0157 4 = 39.2699 ( 1 0 ) x 0.02 = 0.7854

1
[ 0 + 2 (0.007854) + 2 (0.007854) + 0.0157 ] = 0.24115 rad.
6

1
[ 0.7854 + 2 (0.7854) + 2 (0.7854) + 0.7854 ]
6

= 314.9447 rad / sec.

It is to be noted that up to 0.1 sec., since Pe remains at zero,

1 = 2 3 4 = 0.7854

d d
314.1593 39.2699 (1 Pe )
dt dt
Latest point is:
)

k1 = (314.9447 314.1593) x 0.02 = 0.01571

1 / 2 = 314.9447 + 0.3927 = 315.3374

k2 = (315.3374 314.1593) x 0.02 = 0.02356

.02 2 / 2 = 314.9447 + 0.3927 = 315.3374

k3 = (315.3374 314.1593) x 0.02 = 0.02356

.02 3 = 314.9947 + 0.7854 = 315.7301

k4 = (315.7301 314.1593) x 0.02 = 0.03142

1
[ 0.01571 + 2 (0.02356) + 2 (0.02356) + 0.03142 ] = 0.2647 rad.
6

) = 314.9447 + 0.7854 = 315.7301 rad / sec.

d d
314.1593 39.2699 (1 Pe )
dt dt

) Latest point is:

rad.
k1 = (315.7301 314.1593) x 0.02 = 0.03142 ) = 315.7301 rad. / sec.

1 / 2 = 315.7301 + 0.3927 = 316.1228

k2 = (316.1228 314.1593) x 0.02 = 0.03927

.04 2 / 2 = 315.7301 + 0.3927 = 316.1228

k3 = (316.1228 314.1593) x 0.02 = 0.03917

.04 3 = 315.7301 + 0.7854 = 316.5155

k4 = (316.5155 314.1593) x 0.02 = 0.04712

1
+ [ 0.03142 + 2 (0.03927) + 2 (0.03927) + 0.04712 ] = 0.304 rad.
6

.7301 + 0.7854 = 316.5155 rad / sec.

d d
314.1593 39.2699 (1 Pe )
dt dt
Latest point is:
)

k1 = (316.5155 314.1593) x 0.02 = 0.04712

1 / 2 = 316.5155 + 0.3927 = 316.9082

k2 = (316.9082 314.1593) x 0.02 = 0.05498

2 / 2 = 316.5155 + 0.3927 = 316.9082

k3 = (316.9082 314.1593) x 0.02 = 0.05498

3 = 316.5155 + 0.7854 = 317.3009

k4 = (317.3009 314.1593) x 0.02 = 0.06283

1
[ 0.04712 + 2 (0.05498) + 2 (0.05498) + 0.06283 ] = 0.359 rad.
6
d d
314.1593 39.2699 (1 Pe )
dt dt

Latest point is:

)
) = 0.359 rad.
k1 = (317.3009 314.1593) x 0.02 = 0.06283 rad. / sec.

1 / 2 = 317.3009 + 0.3927 = 317.6936

k2 = (317.6936 314.1593) x 0.02 = 0.07069

.08 2 / 2 = 317.3009 + 0.3927 = 317.6936

k3 = (317.6936 314.1593) x 0.02 = 0.07069

.08 3 = 317.3009 + 0.7854 = 318.0863

k4 = (318.0863 314.1593) x 0.02 = 0.07854

1
1) = 0.359 + [ 0.06283 + 2 (0.07069) + 2 (0.07069) + 0.07854 ] = 0.4297 rad.
6

.1593 + 0.7854 = 318.0863 rad. / sec.

At t = 0.1 sec., the fault is cleared. As seen in the previous example, for t 0.1
sec., electric power output of the alternator is given by Pe
d d
314.1593 39.2699 (1 Pe )
dt dt

Latest point is:

2)

k1 = (318.0863 314.1593) x 0.02 = 0.07854 ) = 318.0863 rad. / sec.

Pe = 4.3261 sin (0.4297 rad.) = 1.8022

1 = 39.2699 ( 1 1.8022 ) x 0.02 = - 0.63

) + k1 1 / 2 = 318.0863 - 0.315 = 317.7713

k2 = (317.7713 314.1593) x 0.02 = 0.07224

Pe = 4.3261 sin (0.469 rad.) = 1.9554

2 = 39.2699 ( 1 1.9554 ) x 0.02 = - 0.7504

.1) + k2 / 2 = 0.4658 .1 2 / 2 = 318.0863 + 0.3752 = 317.7111

k3 = (317.7111 314.1593) x 0.02 = 0.07104

Pe = 4.3261 sin (0.4658 rad.) = 1.9430

3 = 39.2699 ( 1 1.9430 ) x 0.02 = - 0.7406

.1) + k3 = 0.5007 .1 3 = 318.0863 - 0.7406 = 317.3457

k4 = (317.3457 314.1593) x 0.02 = 0.06373

Pe = 4.3261 sin (0.5007 rad.) = 2.0767

4 = 39.2699 ( 1 2.0767 ) x 0.02 = - 0.8456

1
+ [ 0.07854 + 2 (0.07224) + 2 (0.07104) + 0.06373 ] = 0.5012 rad.
6

1
+ [ - 0.63 - 2 (0.7504) - 2 (0.7406) - 0.8456 ]= 317.3434 rad. / sec.
6
d d
314.1593 39.2699 (1 Pe )
dt dt

)
Latest point is:
k1 = (317.3434 314.1593) x 0.02 = 0.06368
d. / sec.
Pe = 4.3261 sin (0.5012 rad.) = 2.0786

1 = 39.2699 ( 1 2.0786 ) x 0.02 = - 0.8471

) + k1 1 / 2 = 317.3434 - 0.42355 = 316.91985

k2 = (316.91985 314.1593) x 0.02 = 0.05521

Pe = 4.3261 sin (0.533 rad.) = 2.1982

2 = 39.2699 ( 1 2.1982 ) x 0.02 = - 0.9411

2) + k2 / 2 = 0.5288 2 2 / 2 = 317.3434 - 0.47055 = 316.87285

k3 = (316.87285 314.1593) x 0.02 = 0.05427

Pe = 4.3261 sin (0.5288 rad.) = 2.1825

 3 = 39.2699 ( 1 2.1825 ) x 0.02 = - 0.9287

2) + k3 = 0.5555 2 3 = 317.3434 - 0.9287 = 316.4147

k4 = (316.4147 314.1593) x 0.02 = 0.04511

Pe = 4.3261 sin (0.5555 rad.) = 2.2814

4 = 39.2699 ( 1 2.2814 ) x 0.02 = - 1.0064

1
+ [ 0.06368 + 2 (0.05521) + 2 (0.05427) + 0.04511 ] = 0.5558 rad.
6

1
+ [- 0.8471- 2 (0.9411)- 2 (0.9287)- 1.0064 ]= 316.4112 rad. / sec.
6

The results are tabulated.

t sec 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14

0.2333 0.24115 0.2647 0.304 0.359 0.4297 0.5012 0.5558

13.37 13.82 15.17 17.42 20.57 24.62 28.72 31.84

314.1593 314.9447 315.7301 316.5155 317.3009 318.0863 317.3434 316.4112

rad/sec
Factors Affecting Transient Stability

The two factors mainly affecting the stability of a generator are

INERTIA CONSTANT H and TRANSIENT REACTANCE Xd .

Smaller value of H:

smaller. As seen in the step by step method

Pa ( n 1) 2
(n) (n 1) +
M

the angular swing of the machine in any interval is larger. This will result in
lesser CCT and hence instability may result.
Larger value of Xd :

As the transient reactance of the machine increases, P max decreases. This

is so because the transient reactance forms part of over all series
reactance of the system. All the three power output curves are lowered
when Pmax is decreased. Accordingly, for a given shaft power P s, the initial
0 m is decreased.
0 m as seen in Fig. 17.

Pre-fault output
power

Post-fault
output power

Input power

During-fault
output power

0 m m
0

Fig. 17
The net result is that increased value of constrains
a machine to swing though a smaller angle from its original position before it
reaches the critical clearing angle and the possibility of instability is more.

Thus any developments which lower the H constant and increase the
transient reactance of the machine cause the CCT to decrease and lessen
the possibility of maintaining the stability under transient conditions.