Professional Documents
Culture Documents
Power System
Power System
. Since this is a balanced network, the sum of the currents at the neutrals N (or n )
is zero. Therefore the neutral are at the same potential. Transforming the -connected load to an
equivalent y, we get the per phase equivalent circuit as shown in Fig. 1.21. In this fashion an entire
power system can be converted into its per phase equivalent. The line diagram showing a per
phase equivalent circuit is called a single-line diagram.
Section VI: Per Unit Representation
In a power system different power equipment with different voltage and power levels are connected
together through various step up or step down transformers. However the presence of various
voltage and power levels causes problem in finding out the currents (or voltages) at different points
in the network. To alleviate this problem, all the system quantities are converted into a uniform
normalized platform. This is called the per unit system . In a per unit system each system variable
or quantity is normalized with respect to its own base value. The units of these normalized values
are per unit (abbreviated as pu) and not Volt, Ampere or Ohm. The base quantities chosen are:
VA base ( P
base
): This is the three-phase apparent power (Volt-Ampere) base that is
common to the entire circuit.
Voltage Base ( V
base
): This is the line-to-line base voltage. This quantity is not uniform for
the entire circuit but gets changed by the turns ratio of the transformer.
Fig. 1.20 Three balanced sources supplying two balanced load through balanced source
impedances.
Fig. 1.21 Per phase equivalent circuit of the network of Fig. 1.20.
Based on the above two quantities the current and impedance bases can be defined as
Assume that
an impedance
Z is defined
as Z
1
per unit in a base impedance of Z
base _ old
. Then we have
The
impedance now has to be represented in a new base value denoted as Z
base_new
. Therefore
From (1.120)
Z
2
can be defined in terms of old and new values of VA base and voltage base as
(1.119)
(1.120)
(1.121)
(1.122)
(1.123)
Let us consider the circuit shown in Fig. 1.19 (a) which contains the equivalent circuit of a
transformer. Let the transformer rating be
500 MVA, 220/22 kV with a leakage reactance of 10%.
The VA base of the transformer is 500 MVA and the voltage bases in the primary and secondary side
are 200 kV and 22 kV respectively. Therefore the impedance bases of these two sides are
and W
where the subscripts 1 and 2 refer to the primary (high tension) and secondary (low tension) sides
respectively. Assume that the leakage reactance is referred to the primary side. Then for 10%, i.e.,
0.1 per unit leakage reactance we have
The above reactance when referred to the secondary side is
Hence the per unit impedance in the secondary side is 0.0968/0.968 = 0.1. Therefore we see that the
per unit leakage reactance is the same for both sides of the transformer and, as a consequence, the
transformer can be represented by only its leakage reactance. The equivalent circuit of the
transformer is then as shown in Fig. 1.22. Since this diagram only shows the reactance (or
impedance) of the circuit, this is called the reactance (or impedance ) diagram .
Fig. 1.22 Per unit equivalent circuit of a transformer.
Example 1.2:
Consider the 50 Hz power system the single-line diagram of which is shown in Fig. 1.23. The system
contains three generators, three transformers and three transmission lines. The system ratings are
Generator G
1
200 MVA, 20 kV, X
d
= 15%
Generator G
2
300 MVA, 18 kV, X
d
= 20%
Generator G
3
300 MVA, 20 kV, X
d
= 20%
Transformer T
1
300 MVA, 220Y/22 kV, X
d
= 10%
Transformer T
2
Three single-phase units each rated 100 MVA,
130Y/25 kV, X = 10%
Transformer T
3
300 MVA, 220/22 kV, X = 10%
The transmission line reactances are as indicated in the figure. We have to draw the reactance
diagram choosing the Generator 3 circuit as the base.
Fig. 1.23 Single-line diagram of the power system of Example 1.2.
As we have chosen the circuit of Generator3 as the base, the base MVA for the circuit is 300. The
high voltage side of transformer T
2
is connected wye. Therefore its ratedline to line voltage is 3 X
130 = 225 kV. Since the low voltage side is connected in D , its line to line voltage is 25 kV. The base
voltages are chosen as discussed below.
Since the base voltage of G
3
is 20 kV, the base voltage between T
3
and bus 1 will be 20 X 10 = 200
kV. Also as there is no transformer connected in bus 1, the base voltage of 200 kV must be chosen
for both the lines that are connected to either side of bus 1. Then the base voltage for the circuit of G
1
will also be 20 kV. Finally since the turns ratio of T
2
is 9 (= 225 25), the base voltage in the G 2 side
is 200 9 = 22.22 kV. The base voltages are also indicated in Fig. 1.23.
Once the base voltages for the various parts of the circuit are known, the per unit values for the
various reactances of the circuit are calculated according to (1.123) for a base MVA of 300. These
are listed below.
Generator G
1
Generator G
2
Generator G
3
Transformer T
1
Transformer T
2
Transformer T
3
The base impedance of the transmission line is
Therefore the per unit values of the line impedances are
pu and pu
The impedance diagram is shown in Fig. 1.24.
Fig. 1.24 The impedance diagram of the system of Fig. 1.23.
Closure
This completes our discussion on the modeling of power system components. In the subsequent
portion of this course we shall use these models to construct a power system and use the per unit
notation and the impedance diagram to represent the system.
Fig. 1.24 The impedance diagram of the system of Fig. 1.23.
CHAPTER 2
Overview
As we have discussed earlier in Chapter 1 that the transmission line parameters include series
resistance and inductance and shunt capacitance. In this chapter we shall discuss the various
models of the line. The line models are classified by their length. These classifications are
Short line approximation for lines that are less than 80 km long.
Medium line approximation for lines whose lengths are between 80 km to 250 km.
Long line model for lines that are longer than 250 km.
These models will be discussed in this chapter. However before that let us introduce the ABCD
parameters that are used for relating the sending end voltage and current to the receiving end
voltage and currents.
Section I: ABCD Parameters
Consider the power system shown in Fig. 2.1. In this the sending and receiving end voltages are
denoted by V
S
and V
R
respectively. Also the currents I
S
and I
R
are entering and leaving the network
respectively. The sending end voltage and current are then defined in terms of the ABCD parameters
as
From (2.1) we
see that
This implies that
A is the ratio of sending end voltage to the open circuit receiving end voltage. This quantity is
dimension less.
Similarly,
i.e., B , given in Ohm, is the ratio of sending end voltage and short circuit receiving end current. In a
similar way we
can also define
The parameter
D is dimension less.
Fig. 2.1 Two port representation of a transmission network.
Section II: Short Line Approximation
(2.1)
(2.2)
(2.3)
(2.4)
mho
(2.5)
(2.6)
The shunt capacitance for a short line is almost negligible. The series impedance is assumed to be
lumped as shown in Fig. 2.2. If the impedance per km for an l km long line is z
0
= r + jx , then the total
impedance of the line is Z = R + jX = lr + jlx . The sending end voltage and current for this
approximation
are given by
Therefore the
ABCD
parameters are
given by
Fig. 2.2 Short transmission line representation.
2.3 Medium Line Approximation
Normal - Representation
Normal- T Representation
Medium transmission lines are modeled with lumped shunt admittance. There are two different
representations - nominal- and nominal-T depending on the nature of the network. These two are
discussed here one by one.
Nominal- Representation
In this representation the lumped series impedance is placed in the middle while the shunt admittance
is divided into two equal parts and placed at the two ends. The nominal - representation is shown in
Fig. 2.3. This representation is used for load flow studies, as we shall see later. Also a long
transmission line can be modeled as an equivalent -network for load flow studies.
Fig. 2.3 Nominal- p representation.
Let us define three currents I
1
, I
2
and I
3
as indicated in Fig. 2.3. Applying KCL at nodes M and N we
get
(2.7)
(2.8)
(2.9)
(2.10)
Substituting
(2.11) in (2.10)
we get
Therefore from (2.11) and (2.12) we get the following ABCD parameters of the nominal - p
representation
Nominal- T
Representatio
n
In this
representation
the shunt
admittance is
placed in the middle and the series impedance is divided into two equal parts and these parts are
placed on either side of the shunt admittance. The nominal -T representation is shown in Fig. 2.4. Let
us denote the midpoint voltage as V
M
. Then the application of KCL at the midpoint results in
Fig. 2.4
Nominal-T
representation.
Rearranging the
above equation
can be written
as
Now the
receiving end
current is given
by
Substituting the value of V
M
from (2.16) in (2.17) and rearranging we get
(2.11)
(2.12)
(2.13)
(2.14)
(2.16)
(2.17)
Furthermore the
sending end
current is
Then
substituting the
value of V
M
from (2.16) in
(2.19) and solving
Then the ABCD
parameters of
the T-network
are
Section IV: Long Line Model
For accurate modeling of the transmission line we must not assume that the parameters are
lumped but are distributed throughout line. The single-line diagram of a long transmission line is
shown in Fig. 2.5. The length of the line is l . Let us consider a small strip x that is at a distance x
from the receiving end. The voltage and current at the end of the strip are V and I respectively and
the beginning of the strip are V + V and I + I respectively. The voltage drop across the strip is
then V . Since the length of the strip is x , the series impedance and shunt admittance are z x
and y x . It is to be noted here that the total impedance and admittance of the line are
Fig. 2.5 Long transmission line representation.
From the circuit of Fig. 2.5 we see that
Again as x
0, from (2.25) we get
(2.24)
(2.25)
(2.26)
(2.18)
(2.19)
(2.20)
(2.21)
(2.22)
Now for the current through the strip, applying KCL we get
The second
term of the above equation is the product of two small quantities and therefore can be neglected.
For x 0 we then have
Taking
derivative with respect to x of both sides of (2.26) we get
Substitution of
(2.28) in the above equation results
The roots of
the above equation are located at ( yz ). Hence the solution of (2.29) is of the form
Taking
derivative of (2.30) with respect to x we get
Combining
(2.26) with (2.31) we have
Let us define
the following two quantities
Then (2.30) and (2.32) can be written in terms of the characteristic impedance and propagation
constant as
Let us
assume that x = 0. Then V = V
R
and I = I
R
. From (2.35) and (2.36) we then get
Solving (2.37)
and (2.38) we get the following values for A
1
and A
2
.
(2.27)
(2.28)
(2.29)
(2.30)
(2.31)
(2.32)
(2.33)
(2.34)
(2.35)
(2.36)
(2.37)
(2.38)
Also note that for x = l we have V = V
s
and I = I
S
. Therefore replacing x by l and substituting the
values of A
1
and A
2
in (2.35) and (2.36) we get
Noting that
We can
rewrite (2.39) and (2.40) as
The ABCD
parameters of the long transmission line can then be written as
(2.39)
(2.40)
(2.41)
(2.42)
(2.43)
(2.44)
(2.45)
Let's do an Example
Example 2.1
Consider a 500 km long line for which the per kilometer line impedance and admittance are given
respectively by z = 0.1 + j 0.5145 W and
y = j 3.1734 x 10
-6
mho. Therefore
and
We shall now use the following two formulas for evaluating the hyperbolic forms
Application of the above two equations results in the following values
Therefore from (2.43) to (2.45) the ABCD parameters of the system can be written as
Equivalent- Representation of a Long Line
The -equivalent of a long transmission line is shown in Fig. 2.6. In this the series impedance is
denoted by Z while the shunt admittance is denoted by Y From (2.13) to (2.15) the ABCD
parameters are
defined as
Fig. 2.6 Equivalent p representation of a long transmission line.
Comparing
(2.44) with
(2.47) we can
write
where Z = zl is the total impedance of the line. Again comparing (2.43) with (2.46) we get
Rearranging (2.50) we get
where Y = yl is
the total admittance of the line. Note that for small values of l , sinh l = l and tanh ( l /2) = l /2.
Therefore from (2.49) we get Z = Z and from (2.51) we get Y = Y . This implies that when the
length of the line is small, the nominal- p representation with lumped parameters is fairly accurate.
However the lumped parameter representation becomes erroneous as the length of the line
increases.
Example 2.2
Consider the transmission line given in Example 2.1. The equivalent system parameters for both
lumped and distributed parameter representation are given in Table 2.1 for three different line
lengths. It can be seen that the error between the parameters increases as the line length
increases.
Table 2.1 Variation in equivalent parameters as the line length changes
Length of the
Line (km)
Lumped Parameters Distributed Parameters
Z() Y(mho)
100 52.41 < 79
o
3.17 X 10
-4
< 90
o
52.27< 79
o
3.17 X 10
-4
<
89.98
o
(2.46)
(2.47)
(2.48)
(2.49)
(2.50)
(2.51)
250 131.032 < 79
o
7.93 X 10
-4
< 90
o
128.81<
79.2
o
8.0 X 10
-4
< 89.9
o
500 262.064 < 79
o
1.58 X 10
-3
< 90
o
244.61<79.8
o
1.64 X 10
-3
< 89.6
o
Characterization of a Long Lossless Line
For a lossless line, the line resistance is assumed to be zero. The characteristic impedance then becomes a
pure real number and it is often referred to as the surge impedance . The propagation constant becomes a
pure imaginary number. Defining the propagation constant as = j and replacing l by x we can rewrite (2.41)
and (2.42) as
The term surge
impedance loading or SIL is often used to indicate the nominal capacity of the line. The surge impedance is
the ratio of voltage and current at any point along an infinitely long line. The term SIL or natural power is a
measure of
power delivered
by a transmission
line when
terminated by surge impedance and is given by
where V
0
is the rated voltage of the line.
At SIL Z
C
= V
R
/ I
R
and hence from equations (2.52) and (2.53) we get
This implies that as the distance x changes, the magnitudes of the voltage and current in the above equations
do not change. The voltage then has a flat profile all along the line. Also as Z
C
is real, V and I are in phase with
each other all through out the line. The phase angle difference between the sending end voltage and the
receiving end voltage is then = l. This is shown in Fig. 2.7.
Fig. 2.7 Voltage-current relationship in naturally loaded line.
Voltage and Current Characteristics of an SMIB System
For the analysis presented below we assume that the magnitudes of the voltages at the two ends
are the same. The sending and receiving voltages are given by
(2.52)
(2.53)
(2.54)
(2.55)
(2.56)
where is angle between the sources and is usually called the load angle . As the total length of
the line is l , we replace x by l to obtain the sending end voltage from (2.39) as
Solving the above equation we get
Substituting (2.59) in (2.52), the voltage equation at a point in the transmission line that is at a
distance x from the receiving end is obtained as
In a similar way the current at that point is given by
When the system is unloaded, the receiving end current is zero ( I
R
= 0). Therefore we can rewrite
(2.58) as
Substituting the above equation in (2.52) and (2.53) we get the voltage and current for the unloaded
system as
and
(2.57)
(2.58)
(2.59)
(2.60)
(2.61)
(2.62)
(2.63)
Example 2.4
Consider the system given in Example 2.3. It is assumed that the system is unloaded with V
S
= V
R
=
1< 0 per unit. The voltage and current profiles for the unloaded system is shown in Fig. 2.9. The
maximum voltage of 1.2457 per unit occurs at the receiving end while the maximum current of 0.7428
per unit is at the sending end. The current falls monotonically from the sending end and voltage rises
monotonically to the receiving end. This rise in voltage under unloaded or lightly loaded condition is
called Ferranti effect .
Fig. 2.9 Voltage and current profile over an unloaded transmission line.
Mid Point Voltage and Current of Loaded Lines
The mid point voltage of a transmission line is of significance for the reactive compensation of
transmission lines. To obtain an expression of the mid point voltage, let us assume that the line is
loaded (i.e., the load angle d is not equal to zero). At the mid point of the line we have x = l /2 such
that x = q /2. Let us denote the midpoint voltage by V
M
. Let us also assume that the line is
symmetric, i.e., V
S
= V
R
= V. We can then rewrite equation (2.60) to obtain
Again noting that
We obtain the following expression of the mid point voltage
The mid point current is similarly given by
The phase angle of the mid point voltage is half the load angle always. Also the mid point voltage and
current are in phase, i.e., the power factor at this point is unity. The variation in the magnitude of
voltage with changes in load angle is maximum at the mid point. The voltage at this point decreases
with the increase in . Also as the power through a lossless line is constant through out its length and
the mid point power factor is unity, the mid point current increases with an increase in
Example 2.5
Consider the transmission line discussed in Example 2.4. Assuming the magnitudes of both sending
and receiving end voltages to be 1.0 per unit, we can compute the magnitude of the mid point voltage
as the load angle ( ) changes. This is given in Table 2.2. The variation in voltage with is shown in
Fig. 2.10.
Table 2.2 Changes in the mid-point voltage magnitude with load angle
in degree V
M
in per unit
20 1.0373
25 1.0283
30 1.0174
(2.65)
(2.66)
Fig. 2.10 Variation in voltage profile for a loaded line
Power in a Lossless Line
The power flow through a lossless line can be given by the mid point voltage and current equations
given in (2.66) and (2.67). Since the power factor at this point is unity, real power over the line is
given by
If V = V
0
, the
rated voltage, we can rewrite the above expression in terms of the natural power as
For a short transmission line we have
where X is the total reactance of the line. Equation (2.71) then can be modified to obtain the well
known power transfer relation for the short line approximation as
In general it is not necessary for the magnitudes of the sending and receiving end voltages to be
same. The power transfer relation given in (2.72) will not be valid in that case. To derive a general
expression for power transfer, we assume
If the receiving end real and reactive powers are denoted by P
R
and Q
R
respectively, we can write
from (2.52)
(2.71)
(2.72)
(2.73)
(2.74)
and
Equating real and imaginary parts of the above equation we get
and
Rearranging (2.76) we get the power flow equation for a losslees line as
To derive expressions for the reactive powers, we rearrange (2.75) to obtain the reactive power
delivered to the receiving end as
Again from equation (2.61) we can write
The sending end apparent power is then given by
Equating the imaginary parts of the above equation we get the following expression for the reactive
generated by the source
The reactive power absorbed by the line is then
It is important to note that if the magnitude of the voltage at the two ends is equal, i.e., V
S
= V
R
= V, the reactive powers at the two ends become negative of each other, i.e., Q
S
= Q
R
. The net
reactive power absorbed by the line then becomes twice the sending end reactive power, i.e., Q
L
=
2Q
S
. Furthermore, since cos 1 for small values of , the reactive powers at the two ends for a
short transmission line are given by
(2.75)
(2.76)
(2.77)
(2.78)
(2.79)
(2.80)
(2.81)
The reactive power absorbed by the line under this condition is given by
Example 2.6
Consider a short, lossless transmission line with a line reactance of 0.5 per unit. We assume that the
magnitudes of both sending and receiving end voltages to be 1.0 per unit. The real power transfer
over the line and reactive power consumed by the line are shown in Fig. 2.11. The maximum real
power is 2.0 per unit and it occurs for = 90 . Also the maximum reactive power consumed by the
line occurs at = 180 and it has a value of 8 per unit.
Fig. 2.11 Real power flow and reactive power consumed by a transmission line.
(2.82)
Chapter 3: Network Admittance and Impedance Matrices
As we have seen in Chapter 1 that a power system network can be converted into an equivalent
impedance diagram. This diagram forms the basis of power flow (or load flow) studies and short
circuit analysis. In this chapter we shall discuss the formation of bus admittance matrix (also known
as Y
bus
matrix) and bus impedance matrix (also known as Z
bus
matrix). These two matrices are
related by
We shall discuss the formation of the Y
bus
matrix first. This will be followed by the discussion of the
formation of the Z
bus
matrix.
Section I: Formation of Bus Admittance Matrix
Node Elimination by Matrix Partitioning
Node Elimination by Kron Reduction
Inclusion of Line Charging Capacitors
Formation of Bus Admittance Matrix
Consider the voltage source V
S
with a source (series) impedance of Z
S
as shown in Fig. 3.1 (a).
Using Norton's theorem this circuit can be replaced by a current source I
S
with a parallel admittance
of Y
S
as shown in Fig. 3.1 (b). The relations between the original system and the Norton equivalent
are
We shall use this Norton's theorem for the formulation of the Y
bus
matrix.
Fig. 3.1 (a) Voltage source with a source impedance and (b) its Norton equivalent.
For the time being we shall assume the short line approximation for the formulation of the bus
admittance matrix. We shall thereafter relax this assumption and use the -representation of the
network for power flow studies. Consider the 4-bus power system shown in Fig. 3.2. This contains
two generators G
1
and G
2
that are connected through transformers T
1
and T
2
to buses 1 and 2. Let us
denote the synchronous reactances of G
1
and G
2
by X
G1
and X
G2
respectively and the leakage
reactances of T
1
and T
2
by X
T1
and X
T2
respectively. Let Z
ij
, i = 1, ..., 4 and j = 1, ... , 4 denote the line
impedance between buses i and j .
(3.1)
Fig. 3.2 Single-line diagram of a simple power network.
Then the system impedance diagram is as shown in Fig. 3.3 where Z
11
= j ( X
G1
+ X
T1
) and Z
22
= j (
X
G2
+ X
T2
).
Fig. 3.3 Impedance diagram of the power network of Fig. 3.2.
In this figure the nodes with the node voltages of V
1
to V
4
indicate the buses 1 to 4 respectively. Bus 0
indicates the reference node that is usually the neutral of the Y-connected system. The impedance
diagram is converted into an equivalent admittance diagram shown in Fig. 3.4. In this diagram Y
ij
= 1/
Z
ij
, i = 1,..., 4 and j = 1, ... , 4. The voltage sources E
G1
and E
G2
are converted into the equivalent
current sources I
1
and I
2
respectively using the Norton's theorem discussed before.
Fig. 3.4 Equivalent admittance diagram of the impedance of Fig. 3.3.
Contd from previous slide
We would like to determine the voltage-current relationships of the network shown in Fig. 3.4. It is to
be noted that this relation can be written in terms of the node (bus) voltages V
1
to V
4
and injected
currents I
1
and I
2
as follows
or,
(3.3)
(3.4)
It can be easily seen that we get (3.1) from (3.3) and (3.4).
Consider node (bus) 1 that is connected to the nodes 2 and 3. Then applying KCL at this node we get
In a similar way application of KCL at nodes 2, 3 and 4 results in the following equations
Combining (3.5) to (3.8) we get
Comparing (3.9) with (3.3) we can write
In general the format of the Y
bus
matrix for an n -bus power system is as follows
where
It is to be noted
that Y
bus
is a symmetric matrix in which the sum of all the elements of the k th column is Y
kk
.
Example 3.1
Consider the impedance diagram of Fig. 3.2 in which the system parameters are given in per unit by
Z
11
= Z
22
= j0.25, Z
12
= j0.2, Z
13
= j0.25, Z
23
= Z
34
= j0.4 and Z
24
= j0.5
(3.5)
(3.6)
(3.7)
(3.8)
(3.9)
(3.10)
(3.11)
(3.12)
The system admittance can then be written in per unit as
Y
11
= Y
22
= j4, Y
12
= j5, Y
13
= j4, Y
23
= Y
34
= j2.5 and Y
24
= j2
The Y
bus
is then given from (3.10) as
per unit
Consequently the bus impedance matrix is given by
per unit
It can be seen that like the Y
bus
matrix the Z
bus
matrix is also symmetric.
Let us now assume that the voltages E
G1
and E
G2
are given by
The current sources I
1
and I
2
are then given by
We then get the node voltages from (3.4) as
p.u.
Solving the above equation we get the node voltages as
per unit
Node Elimination by Matrix Partitioning
Sometimes it is desirable to reduce the network by eliminating the nodes in which the current do not
enter or leave.
Let (3.3) be
written as
(3.13)
In the above equation I
A
is a vector containing the currents that are injected, I
x
is a null vector and the
Y
bus
matrix is portioned with the matrices K , L and M . Note that the Y
bus
matrix contains both L and
L
T
due to its symmetric nature.
We get the following two sets of equations from (3.13)
Substituting (3.15) in (3.14) we get
Therefore we obtain the following reduced bus admittance matrix
Example 3.2
Let us consider the system of Example 3.1. Since there is no current injection in either bus 3 or bus 4,
from the Y
bus
computed we can write
We then have
per unit
Substituting I
1
= 4 60 per unit and I
2
= 4 90 per unit we shall get the same values of V
1
and V
2
as given in Example 3.1.
Inspecting the reduced Y
bus
matrix we can state that the admittance between buses 1 and 2 is - j
6.8978. Therefore the self admittance (the admittance that is connected in shunt) of the buses 1 and
2 is - j4 per unit (= - j 10.8978 + j 6.8978). The reduced admittance diagram obtained by eliminating
nodes 3 and 4 is shown in Fig. 3.5. It is to be noted that the impedance between buses 1 and 2 is the
Thevenin impedance between these two buses. The value of this impedance is 1/( - j 6.8978) = j
0.145 per unit.
Fig. 3.5 Reduced admittance diagram after the elimination of buses 3 and 4.
Node Elimination by Kron Reduction
Consider an equation of the form
(3.14)
(3.15)
(3.16)
where A is an ( n X n ) real or complex valued matrix, x and b are vectors in either R
n
or C
n
. assume
that the b vector has a zero element in the n th row such that (3.18) is given as
We can then eliminate the k
th
row and k
th
column to obtain a reduced ( n - 1) number of equations of
the form
The elimination is performed using the following elementary operations
Example 3.3
Let us consider the same system of Example 3.1. We would like to eliminate the last two rows and
columns. Let us first eliminate the last row and last column. Some of the values are given below
In a similar way we can calculate the other elements. Finally eliminating the last row and last column,
as all these elements are zero, we get the new Y
bus
matrix as
Further reducing the last row and the last column of the above matrix using (3.21), we obtain the
reduced Y
bus
matrix given in Example 3.2.
Inclusion of Line Charging Capacitors
So far we have assumed that the transmission lines are modeled with lumped series impedances
without the shunt capacitances. However in practice, the Y
bus
matrix contains the shunt admittances
for load flow analysis in which the transmission lines are represented by its -equivalent. Note that
(3.18)
(3.19)
(3.20)
(3.21)
whether the line is assumed to be of medium length or long length is irrelevant as we have seen in
Chapter 2 how both of them can be represented in a p -equivalent.
Consider now the power system of Fig. 3.2. Let us assume that all the lines are represented in an
equivalent- with the shunt admittance between the line i and j being denoted by Y
chij
. Then the
equivalent admittance at the two end of this line will be Y
chij
/2. For example the shunt capacitance at
the two ends of the line joining buses 1 and 3 will be Y
ch13
/2. We can then modify the admittance
diagram Fig. 3.4 as shown in Fig. 3.6. The Y
bus
matrix of (3.10) is then modified as
(3.22)
where
Fig. 3.6 Admittance diagram of the power system Fig. 3.2 with line charging capacitors.
Section II: Elements Of The Bus Impedance And Admittance Matrices
Equation (3.1) indicates that the bus impedance and admittance matrices are inverses of each other.
Also since Y
bus
is a symmetric matrix, Z
bus
is also a symmetric matrix. Consider a 4-bus system for
which the voltage-current relations are given in terms of the Y
bus
matrix as
We can then write
(3.23)
(3.24)
This implies that Y
11
is the admittance measured at bus-1 when buses 2, 3 and 4 are short circuited.
The admittance Y
11
is defined as the self admittance at bus-1. In a similar way the self admittances
of buses 2, 3 and 4 can also be defined that are the diagonal elements of the Y
bus
matrix. The off
diagonal elements are denoted as the mutual admittances . For example the mutual admittance
between buses 1 and 2 is defined as
The mutual admittance Y
12
is obtained as the ratio of the current injected in bus-1 to the voltage of
bus-2 when buses 1, 3 and 4 are short circuited. This is obtained by applying a voltage at bus-2 while
shorting the other three buses.
The voltage-current relation can be written in terms of the Z
bus
matrix as
The driving point impedance at bus-1 is then defined as
i.e., the driving point impedance is obtained by injecting a current at bus-1 while keeping buses 2, 3
and 4 open-circuited. Comparing (3.26) and (3.28) we can conclude that Z
11
is not the reciprocal of
Y
11
. The transfer impedance between buses 1 and 2 can be obtained by injecting a current at bus-2
while open-circuiting buses 1, 3 and 4 as
It can also be seen that Z
12
is not the reciprocal of Y
12
.
Section II: Modification of Bus Impedance Matrix
Adding a New Bus to the Reference Bus
Adding a New Bus to an Existing Bus through an Impedance
Adding an Impedance to the Reference Bus from an Existing Bus
Adding an Impedance between two Existing Buses
Direct Determination of Z
bus
Matrix
Modification of Bus Impedance Matrix
Equation (3.1) gives the relation between the bus impedance and admittance matrices. However it
may be possible that the topology of the power system changes by the inclusion of a new bus or line.
(3.25)
(3.26)
(3.27)
(3.28)
(3.29)
In that case it is not necessary to recompute the Y
bus
matrix again for the formation of Z
bus
matrix. We
shall discuss four possible cases by which an existing bus impedance matrix can be modified.
Let us assume that an n -bus power system exists in which the voltage-current relations are given in
terms of the bus impedance matrix as
The aim is to modify the matrix Z
orig
when a new bus or line is connected to the power system.
Adding a New Bus to the Reference Bus
It is assumed that a new bus p ( p > n ) is added to the reference bus through an impedance Z
p
. The
schematic diagram for this case is shown in Fig. 3.7. Since this bus is only connected to the reference
bus, the voltage-current relations the new system are
Fig. 3.7 A new bus is added to the reference bus.
Adding a New Bus to an Existing Bus through an Impedance
This is the case when a bus, which has not been a part of the original network, is added to an existing
bus through a transmission line with an impedance of Z
b
. Let us assume that p ( p > n ) is the new
bus that is connected to bus k ( k < n ) through Z
b
. Then the schematic diagram of the circuit is as
shown in Fig. 3.8. Note from this figure that the current I
p
flowing from bus p will alter the voltage of
the bus k . We shall then have
In a similar way the current I
p
will also alter the voltages of all the other buses as
Furthermore the voltage of the bus p is given by
(3.30)
(3.31)
(3.32)
(3.33)
(3.34)
Therefore the new voltage current relations are
It can be noticed that the new Z
bus
matrix is also symmetric.
Fig. 3.8 A new bus is added to an existing bus through an impedance.
Adding an Impedance to the Reference Bus from an Existing Bus
To accomplish this we first assume that an impedance Z
b
is added from a new bus p to an existing
bus k . This can be accomplished using the method discussed in Section 3.3.2. Then to add this bus
k to the reference bus through Z
b
, we set the voltage V
p
of the new bus to zero. However now we
have an ( n + 1) X ( n + 1) Z bus matrix instead of an n X n matrix. We can then remove the last row
and last column of the new Z
bus
matrix using the Kron's reduction given in (3.21).
Adding an Impedance between two Existing Buses
Let us assume that we add an impedance Z
b
between two existing buses k and j as shown in Fig. 3.9.
Therefore the current injected into the network from the bus k side will be I
k
- I
b
instead of I
k
. Similarly
the current injected into the network from the bus j side will be I
j
+ I
b
instead of I
j
. Consequently the
voltage of the i
th
bus will be
Similarly we have
and
Fig. 3.9 An impedance is added between two existing buses.
We shall now have to eliminate I
b
from the above equations. To do that we note from Fig. 3.9 that
(3.35)
(3.36)
(3.37)
(3.38)
Substituting (3.37) and (3.38) in (3.39) we get
We can then write the voltage current relations as
where
We can now eliminate the last row and last column using the Kron's reduction given in (3.21).
Direct Determination of Z
bus
Matrix
We shall now use the methods given in Sections 3.3.1 to 3.3.4 for the direct determination of the Z
bus
matrix without forming the Y
bus
matrix first. To accomplish this we shall consider the system of Fig. 3.2
and shall use the system data given in Example 3.1. Note that for the construction of the Z bus matrix
we first eliminate all the voltage sources from the system.
Step-1 : Start with bus-1. Assume that no other buses or lines exist in the system. We add this bus to
the reference bus with the impedance of j 0.25 per unit. Then the Z
bus
matrix is
Step-2 : We now add bus-2 to the reference bus using (3.31). The system impedance diagram is
shown in Fig. 3.10. We then can modify (3.43) as
Fig. 3.10 Network of step-2.
Step-3 : We now add an impedance of j 0.2 per unit between buses 1 and 2 as shown in Fig. 3.11.
The interim Z bus matrix is then obtained by applying (3.41) on (3.44) as
Eliminating the last row and last column using the Kron's reduction of (3.31) we get
(3.39)
(3.40)
(3.41)
(3.42)
(3.43)
(3.44)
Step-4: We now
add bus-3 to bus-1 through an impedance of j 0.25 per unit as shown in Fig. 3.12. The application of
(3.35) on (3.45) will then result in the following matrix
Fig. 3.11 Network of step-3.
Fig. 3.12 Network of step-4.
Step-5 : Connect buses 2 and 3 through an impedance of j 0.4 per unit as shown in Fig. 3.13. The
interim Z
bus
matrix is then formed from (3.41) and (3.46) as
Fig. 3.13 Network of step-5.
Using the Kron's reduction we get the following matrix
Step-6 : We now add a new bus-4 to bus-2 through an impedance of j 0.5 as shown in Fig. 3.14.
Then the application of (3.35) on (3.47) results in the following matrix
(3.45)
(3.46)
(3.47)
Fig. 3.14 Network of step-6.
Step-7 : Finally we add buses 3 and 4 through an impedance of j 0.4 to obtain the network of Fig. 3.3
minus the voltage sources. The application of (3.41) on (3.48) results in the interim Z
bus
matrix of
Eliminating the
5
th
row and column through Kron's reduction we get the final Z
bus
as
The Z
bus
matrix
given in (3.49) is the as that given in Example 3.1 which is obtained by inverting the Y
bus
matrix.
Section IV: Thevenin Impedance And Z
bus
Matrix
To establish relationships between the elements of the Z
bus
matrix and Thevenin equivalent, let us
consider the following example
Example 3.4
Consider the two bus power system shown in Fig. 3.15. It can be seen that the open-circuit voltages
of buses a and b are V
a
and V
b
respectively. From (3.11) we can write the Y
bus
matrix of the system
as
Fig. 3.15 Two-bus power system of Example 3.4.
The determinant of the above matrix is
(3.48)
(3.49)
Therefore the Z
bus
matrix is
Solving the last two equations we get
Now consider the system of Fig. 3.15. The Thevenin impedance of looking into the system at bus- a
is the parallel combination of Z
aa
and Z
ab
+ Z
bb
, i.e.,
Similarly the Thevenin impedance obtained by looking into the system at bus- b is the parallel
combination of Z
bb
and Z
aa
+ Z
ab
, i.e.,
Hence the driving point impedances of the two buses are their Thevenin impedances.
Let us now consider the Thevenin impedance while looking at the system between the buses a and b
. From Fig. 3.15 it is evident that this Thevenin impedance is the parall el combination of Z
ab
and Z
aa
+
Z
bb
, i.e.,
With the values given in (3.50) we can write
Comparing the last two equations we can write
As we have seen in the above example in the relation V = Z
bus
I , the node or bus voltages V
i
, i = 1, ...
, n are the open circuit voltages. Let us assume that the currents injected in buses 1, ... , k - 1 and k +
1, ... , n are zero when a short circuit occurs at bus k . Then Thevenin impedance at bus k is
(3.50)
(3.51)
(3.52)
(3.53
From (3.51), (3.52) and (3.54) we can surmise that the driving point impedance at each bus is the
Thevenin impedance.
Let us now find the Thevenin impedance between two buses j and k of a power system. Let the open
circuit voltages be defined by the voltage vector V and corresponding currents be defined by I such
that
Now suppose the currents are changed by I such that the voltages are changed by V . Then
Comparing (3.55) and (3.56) we can write
Let us now assume that additional currents I
k
and I
k
are injected at the buses k and j respectively
while the currents injected at the other buses remain the same. Then from (3.57) we can write
We can therefore write the following two equations form (3.58)
The above two equations can be rewritten as
Since Z
jk
= Z
kj
the network can be drawn as shown in Fig. 3.16. By inspection we can see that the
open circuit voltage between the buses k and j is
and the short circuit current through these two buses is
Also during the short circuit V
k
- V
j
= 0. Therefore combining (3.59) and (3.60) we get
(3.54)
(3.55)
(3.56)
(3.57)
(3.58)
(3.59)
(3.60)
(3.61)
(3.62)
Combining (3.61) to (3.63) we find the Thevenin impedance between the buses k and j as
The above equation agrees with our earlier derivation of the two bus network given in (3.53).
Fig 3.16 Thevenin equivalent between buses k and j
Chapter 4: Load Flow Studies
Overview
Load flow studies are one of the most important aspects of power system planning and operation.
The load flow gives us the sinusoidal steady state of the entire system - voltages, real and reactive
power generated and absorbed and line losses. Since the load is a static quantity and it is the
power that flows through transmission lines, the purists prefer to call this Power Flow studies
rather than load flow studies. We shall however stick to the original nomenclature of load flow.
Through the load flow studies we can obtain the voltage magnitudes and angles at each bus in the
steady state. This is rather important as the magnitudes of the bus voltages are required to be held
within a specified limit. Once the bus voltage magnitudes and their angles are computed using the
load flow, the real and reactive power flow through each line can be computed. Also based on the
difference between power flow in the sending and receiving ends, the losses in a particular line can
also be computed. Furthermore, from the line flow we can also determine the over and under load
conditions.
The steady state power and reactive power supplied by a bus in a power network are expressed in
terms of nonlinear algebraic equations. We therefore would require iterative methods for solving
these equations. In this chapter we shall discuss two of the load flow methods. We shall also
delineate how to interpret the load flow results.
Section I: Real And Reactive Power Injected in a Bus
For the formulation of the real and reactive power entering a b us, we need to define the following
quantities. Let the voltage at the i
th
bus be denoted by
Also let us define the self admittance at bus- i as
(3.63)
(3.64)
(4.1)
(4.2)
Similarly the mutual admittance between the buses i and j can be written as
Let the power system contains a total number of n buses. The current injected at bus- i is given as
It is to be noted we shall assume the current entering a bus to be positive and that leaving the bus to
be negative. As a consequence the power and reactive power entering a bus will also be assumed to
be positive. The complex power at bus- i is then given by
Note that
Therefore substituting in (4.5) we get the real and reactive power as
Section II: Classification Of Buses
For load flow studies it is assumed that the loads are constant and they are defined by their real and
reactive power consumption. It is further assumed that the generator terminal voltages are tightly
regulated and therefore are constant. The main objective of the load flow is to find the voltage
magnitude of each bus and its angle when the powers generated and loads are pre-specified. To
facilitate this we classify the different buses of the power system shown in the chart below.
(4.3)
(4.4)
(4.5)
(4.6)
Classification Of Buses
Load Buses : In these buses no generators are connected and hence the generated real power P
Gi
and reactive power Q
Gi
are taken as zero. The load drawn by these buses are defined by real power -
P
Li
and reactive power -Q
Li
in which the negative sign accommodates for the power flowing out of the
bus. This is why these buses are sometimes referred to as P-Q bus. The objective of the load flow is
to find the bus voltage magnitude |V
i
| and its angle
i
.
Voltage Controlled Buses : These are the buses where generators are connected. Therefore the
power generation in such buses is controlled through a prime mover while the terminal voltage is
controlled through the generator excitation. Keeping the input power constant through turbine-
governor control and keeping the bus voltage constant using automatic voltage regulator, we can
specify constant P
Gi
and | V
i
| for these buses. This is why such buses are also referred to as P-V
buses. It is to be noted that the reactive power supplied by the generator Q
Gi
depends on the system
configuration and cannot be specified in advance. Furthermore we have to find the unknown angle
i
of the bus voltage.
Slack or Swing Bus : Usually this bus is numbered 1 for the load flow studies. This bus sets the
angular reference for all the other buses. Since it is the angle difference between two voltage sources
that dictates the real and reactive power flow between them, the particular angle of the slack bus is
not important. However it sets the reference against which angles of all the other bus voltages are
measured. For this reason the angle of this bus is usually chosen as 0 . Furthermore it is assumed
that the magnitude of the voltage of this bus is known.
Now consider a typical load flow problem in which all the load demands are known. Even if the
generation matches the sum total of these demands exactly, the mismatch between generation and
load will persist because of the line I
2
R losses. Since the I
2
R loss of a line depends on the line
current which, in turn, depends on the magnitudes and angles of vol tages of the two buses connected
to the line, it is rather difficult to estimate the loss without calculating the voltages and angles. For this
reason a generator bus is usually chosen as the slack bus without specifying its real power. It is
assumed that the generator connected to this bus will supply the balance of the real power required
and the line losses.
Section III: Preparation Of Data For Load Flow
Let real and reactive power generated at bus- i be denoted by P
Gi
and Q
Gi
respectively. Also let us
denote the real and reactive power consumed at the i
th
th bus by P
Li
and Q
Li
respectively. Then the
net real power injected in bus- i is
Let the injected power calculated by the load flow program be P
i, calc
. Then the mismatch between
the actual injected and calculated values is given by
In a similar way the mismatch between the reactive power injected and calculated values is given
by
The purpose
of the load flow is to minimize the above two mismatches. It is to be noted that (4.6) and (4.7) are
used for the calculation of real and reactive power in (4.9) and (4.10). However since the
magnitudes of all the voltages and their angles are not known a priori, an iterative procedure must
be used to estimate the bus voltages and their angles in order to calculate the mismatches. It is
expected that mismatches P
i
and Q
i
reduce with each iteration and the load flow is said to have
converged when the mismatches of all the buses become less than a very small number.
For the load flow studies we shall consider the system of Fig. 4.1, which has 2 generator and 3 load
buses. We define bus-1 as the slack bus while taking bus-5 as the P-V bus. Buses 2, 3 and 4 are
(4.8)
(4.9)
(4.10)
P-Q buses. The line impedances and the line charging admittances are given in Table 4.1. Based
on this data the Y bus matrix is given in Table 4.2. This matrix is formed using the same procedure
as given in Section 3.1.3. It is to be noted here that the sources and their internal impedances are
not considered while forming the Y
bus
matrix for load flow studies which deal only with the bus
voltages.
Fig. 4.1 The simple power system used for load flow studies.
Table 4.1 Line impedance and line charging data of the system of Fig. 4.1.
Line(bus to bus) Impedance Line charging ( Y /2)
1-2 0.02 + j 0.10 j 0.030
1-5 0.05 + j 0.25 j 0.020
2-3 0.04 + j 0.20 j 0.025
2-5 0.05 + j 0.25 j 0.020
3-4 0.05 + j 0.25 j 0.020
3-5 0.08 + j 0.40 j 0.010
4-5 0.10 + j 0.50 j 0.075
Table 4.2 Y
bus
matrix of the system of Fig. 4.1.
1 2 3 4 5
1 2.6923 - j 13.4115 - 1.9231 + j 9.6154 0 0 - 0.7692 + j 3.8462
2 - 1.9231 + j 9.6154 3.6538 - j 18.1942 - 0.9615 + j 4.8077 0 - 0.7692 + j 3.8462
3 0 - 0.9615 + j 4.8077 2.2115 - j 11.0027 - 0.7692 + j 3.8462 - 0.4808 + j 2.4038
4 0 0 - 0.7692 + j 3.8462 1.1538 - j 5.6742 - 0.3846 + j 1.9231
5 - 0.7692 + j 3.8462 - 0.7692 + j 3.8462 - 0.4808 + j 2.4038 - 0.3846 + j 1.9231 2.4038 - j 11.8942
The bus voltage magnitudes, their angles, the power generated and consumed at each bus are
given in Table 4.3. In this table some of the voltages and their angles are given in boldface letters.
This indicates that these are initial data used for starting the load flow program. The power and
reactive power generated at the slack bus and the reactive power generated at the P-V bus are
unknown. Therefore each of these quantities are indicated by a dash ( - ). Since we do not need
these quantities for our load flow calculations, their initial estimates are not required. Al so note from
Fig. 4.1 that the slack bus does not contain any load while the P-V bus 5 has a local load and this is
indicated in the load column.
Table 4.3 Bus voltages, power generated and load - initial data.
Bus no. Bus voltage Power generated Load
Magnitude (pu) Angle (deg) P (MW) Q (MVAr) P (MW) P (MVAr)
1 1.05 0 - - 0 0
2 1 0 0 0 96 62
3 1 0 0 0 35 14
4 1 0 0 0 16 8
5 1.02 0 48 - 24 11
Section IV: Load Flow by Gauss-Seidel Method
Updating Load Bus Voltages
Updating P-V Bus Voltages
Convergence of the Algorithm
The basic power flow equations (4.6) and (4.7) are nonlinear. In an n -bus power system, let the
number of P-Q buses be n
p
and the number of P-V (generator) buses be n
g
such that n = n
p
+ n
g
+ 1.
Both voltage magnitudes and angles of the P-Q buses and voltage angles of the P-V buses are
unknown making a total number of 2n
p
+ n
g
quantities to be determined. Amongst the known
quantities are 2n
p
numbers of real and reactive powers of the P-Q buses, 2n
g
numbers of real powers
and voltage magnitudes of the P-V buses and voltage magnitude and angle of the slack bus.
Therefore there are sufficient numbers of known quantities to obtain a solution of the load flow
problem. However, it is rather difficult to obtain a set of closed form equations from (4.6) and (4.7).
We therefore have to resort to obtain iterative solutions of the load flow problem.
At the beginning of an iterative method, a set of values for the unknown quantities are chosen. These
are then updated at each iteration. The process continues till errors between all the known and actual
quantities reduce below a pre-specified value. In the Gauss-Seidel load flow we denote the initial
voltage of the i
th
bus by V
i
(0)
, i = 2, ... , n . This should read as the voltage of the i
th
bus at the 0
th
iteration, or initial guess. Similarly this voltage after the first iteration will be denoted by V
i
(1)
. In this
Gauss-Seidel load flow the load buses and voltage controlled buses are treated differently. However
in both these type of buses we use the complex power equation given in (4.5) for updating the
voltages. Knowing the real and reactive power injected at any bus we can expand (4.5) as
We can rewrite (4.11) as
In this fashion the voltages of all the buses are updated. We shall outline this procedure with the help
of the system of Fig. 4.1, with the system data given in Tables 4.1 to 4.3. It is to be noted that the real
and reactive powers are given respectively in MW and MVAr. However they are converted into per
unit quantities where a base of 100 MVA is chosen.
Updating Load Bus Voltages
Let us start the procedure with bus-2 of the 5 bus 7 line system given in fig: 4.1. Since this is load
bus, both the real and reactive power into this bus is known. We can therefore write from (4.12)
(4.11)
(4.12)
(4.13)
From the data given in Table 4.3 we can write
It is to be noted that since the real and reactive power is drawn from this bus, both these quantities
appear in the above equation with a negative sign. With the values of the Y bus elements given in
Table 4.2 we get V
2
1
= 0.9927 < 2.5959 .
The first iteration voltage of bus-3 is given by
Note that in the above equation since the update for the bus-2 voltage is already available, we used
the 1
st
iteration value of this rather than the initial value. Substituting the numerical data we get V
3
(1)
=
0.9883 < 2. 8258 . Finally the bus-4 voltage is given by
Solving we get
V
4
(1)
= 0. 9968 < 3.4849 .
Updating P-V Bus Voltages
It can be seen from Table 4.3 that even though the real power i s specified for the P-V bus-5, its
reactive power is unknown. Therefore to update the voltage of this bus, we must first estimate the
reactive power of this bus. Note from Fig. 4.11 that
And hence we can write the k
th
iteration values as
For the system of Fig. 4.1 we have
This is computed as 0.0899 per unit. Once the reactive power is estimated, the bus-5 voltage is
updated as
It is to be noted that even though the power generation in bus-5 is 48 MW, there is a local load that is
consuming half that amount. Therefore the net power injected by this bus is 24 MW and consequently
the injected power P
5, inj
in this case is taken as 0.24 per unit. The voltage is calculated as V
5
(1)
=
1.0169 < 0.8894 . Unfortunately however the magnitude of the voltage obtained above is not equal
to the magnitude given in Table 4.3. We must therefore force this voltage magnitude to be equal to
that specified. This is accomplished by
(4.14)
(4.15)
(4.16)
(4.17)
(4.18)
(4.19)
This will fix the voltage magnitude to be 1.02 per unit while retaining the phase of 0.8894 . The
corrected voltage is used in the next iteration.
Convergence of the Algorithm
As can be seen from Table 4.3 that a total number of 4 real and 3 reactive powers are known to us.
We must then calculate each of these from (4.6) and (4.7) using the values of the voltage magnitudes
and their angle obtained after each iteration. The power mismatches are then calculated from (4.9)
and (4.10). The process is assumed to have converged when each of P
2
, P
3
, P
4
, P
5
, Q
2
,
Q
3
and Q
4
is below a small pre-specified value. At this point the process is terminated.
Sometimes to accelerate computation in the P-Q buses the voltages obtained from (4.12) is multiplied
by a constant. The voltage update of bus- i is then given by
where is a constant that is known as the acceleration factor . The value of has to be below 2.0
for the convergence to occur. Table 4.4 lists the values of the bus voltages after the 1
st
iteration and
number of iterations required for the algorithm to converge for different values of . It can be seen that
the algorithm converges in the least number of iterations when is 1.4 and the maximum number of
iterations are required when is 2. In fact the algorithm will start to diverge if larger values of
acceleration factor are chosen. The system data after the convergence of the algorithm will be
discussed later.
Table 4.4 Gauss-Seidel method: bus voltages after 1 st iteration and number of iterations required for
convergence for different values of l .
k Bus voltages (per unit) after 1
st
iteration No of iterations
for convergence
V
2
V
3
V
4
V
5
1 0.9927 2.6 0.9883 2.83 0.9968 3.48 1.02 0.89 28
2 0.9874 5.22 0.9766 8.04 0.9918 14.02 1.02 4.39 860
1.8 0.9883 4.7 0.9785 6.8 0.9903 11.12 1.02 3.52 54
1.6 0.9893 4.17 0.9807 5.67 0.9909 8.65 1.02 2.74 24
1.4 0.9903 3.64 0.9831 4.62 0.9926 6.57 1.02 2.05 14
1.2 0.9915 3.11 0.9857 3.68 0.9947 4.87 1.02 1.43 19
Section V: Solution of a Set of Nonlinear Equations by Newton-Raphson Method
In this section we shall discuss the solution of a set of nonlinear equations through Newton-Raphson
method. Let us consider that we have a set of n nonlinear equations of a total number of n variables
x
1
, x
2
, ... , x
n
. Let these equations be given by
where f
1
, ... , f
n
are functions of the variables x
1
, x
2
, ... , x
n
. We can then define another set of functions g
1
, ... , g
n
as
given below
(4.20)
(4.21)
(4.22)
Let us assume
that the initial estimates of the n variables are x
1
(0)
, x
2
(0)
, ... , x
n
(0)
. Let us add corrections x
1
(0)
,
x
2
(0)
, ... , x
n
(0)
to these variables such that we get the correct solution of these variables defined by
The functions in (4.23) then can be written in terms of the variables given in (4.24) as
We can then expand the above equation in Taylor 's series around the nominal values of x
1
(0)
, x
2
(0)
,
... , x
n
(0)
. Neglecting the second and higher order terms of the series, the expansion of g
k
, k = 1, ... ,
n is given as
where is the partial derivative of g
k
evaluated at x
2
(1)
, ... , x
n
(1)
.
Equation (4.26) can be written in vector-matrix form as
The square matrix of partial derivatives is called the Jacobian matrix J with J
(1)
indicating that the
matrix is evaluated for the initial values of x
2
(0)
, ... , x
n
(0)
. We can then write the solution of (4.27) as
Since the Taylor 's series is truncated by neglecting the 2
nd
and higher order terms, we cannot expect
to find the correct solution at the end of first iteration. We shall then have
(4.23)
(4.24)
(4.25)
(4.26)
(4.27)
(4.28)
, k= 1, ....n
These are then used to find J
(1)
and g
k
(1)
, k = 1, ... , n . We can then find x
2
(1)
, ... , x
n
(1)
from an
equation like (4.28) and subsequently calculate x
2
(1)
, ... , x
n
(1)
. The process continues till g
k
, k = 1,
... , n becomes less than a small quantity.
Section VI: Load Flow By Newton-Raphson Method
Load Flow Algorithm
Formation of the Jacobian Matrix
Solution of Newton-Raphson Load Flow
Let us assume that an n -bus power system contains a total n
p
number of P-Q buses while the
number of P-V (generator) buses be n
g
such that n = n
p
+ n
g
+ 1. Bus-1 is assumed to be the slack
bus. We shall further use the mismatch equations of P
i
and Q
i
given in (4.9) and (4.10)
respectively. The approach to Newton-Raphson load flow is similar to that of solving a system of
nonlinear equations using the Newton-Raphson method: At each iteration we have to form a
Jacobian matrix and solve for the corrections from an equation of the type given in (4.27). For the
load flow problem, this equation is of the form
where the Jacobian matrix is divided into submatrices as
It can be seen that the size of the Jacobian matrix is ( n + n
p
1) x ( n + n
p
1). For example for the 5-
bus problem of Fig. 4.1 this matrix will be of the size (7 x 7). The dimensions of the submatrices are
as follows:
J
11
: (n 1) (n 1), J
12
: (n 1) n
p
, J
21
: n
p
(n 1) and J
22
: n
p
n
p
The sub-matrices are
(4.29)
(4.30)
(4.31)
Load Flow
Algorithm
The Newton-
Raphson
procedure is as
follows:
Step-1: Choose
the initial values
of the voltage
magnitudes |V|
(0)
of all n
p
load
buses and n 1
angles
(0)
of the voltages of all the buses except the slack bus.
Step-2: Use the estimated |V|
(0)
and
(0)
to calculate a total n 1 number of injected real power
P
calc
(0)
and equal number of real power mismatch P
(0)
.
Step-3: Use the estimated |V|
(0)
and
(0)
to calculate a total n
p
number of injected reactive power
Q
calc
(0)
and equal number of reactive power mismatch Q
(0)
.
Step-3: Use the estimated |V|
(0)
and
(0)
to formulate the Jacobian matrix J
(0)
.
Step-4: Solve (4.30) for
(0)
and |V|
(0)
|V|
(0)
.
Step-5 : Obtain the updates from
Step-6: Check if all the mismatches are below a small number. Terminate the process if yes.
Otherwise go back to step-1 to start the next iteration with the updates given by (4.36) and (4.37).
Formation of the Jacobian Matrix
We shall now discuss the formation of the submatrices of the Jacobian matrix. To do that we shall
use the real and reactive power equations of (4.6) and (4.7). Let us rewrite them with the help of
(4.2) as
(4.32)
(4.33)
(4.34)
(4.36)
(4.37)
(4.38)
A. Formation of J
11
Let us define J
11
as
It can be seen from (4.32) that L
ik
's are the partial derivatives of P
i
with respect to
k
. The derivative P
i
(4.38) with respect to k for i k is given by
Similarly the derivative P
i
with respect to k for i = k is given by
Comparing the above equation with (4.39) we can write
B. Formation of J
21
Let us define J
21
as
From (4.34) it is evident that the elements of J
21
are the partial derivative of Q with respect to . From
(4.39) we can write
Similarly for i = k we have
(4.39)
(4.40)
(4.41)
(4.42)
(4.43)
(4.44)
(4.45)
The last equality of (4.45) is evident from (4.38).
C. Formation of J
12
Let us define J
12
as
As evident from (4.33), the elements of J
21
involve the derivatives of real power P with respect to
magnitude of bus voltage |V| . For i k , we can write from (4.38)
For i = k we have
Formation of J
22
For the formation of J
22
let us define
For i k we can
write from (4.39)
Finally for i = k we have
We therefore see that once the submatrices J
11
and J
21
are computed, the formation of the
submatrices J
12
and J
22
is fairly straightforward. For large system this will result in considerable
saving in the computation time.
(4.46)
(4.47)
(4.48)
(4.49)
(4.50)
(4.51)
Solution of Newton-Raphson Load Flow
The Newton-Raphson load flow program is tested on the system of Fig. 4.1 with the system data and
initial conditions given in Tables 4.1 to 4.3. From (4.41) we can write
Similarly from (4.39) we have
Hence from (4.42) we get
In a similar way the rest of the components of the matrix J
11
(0)
are calculated. This matrix is given by
For forming the off diagonal elements of J
21
we note from (4.44) that
Also from (4.38) the real power injected at bus-2 is calculated as
Hence from (4.45) we have
Similarly the rest of the elements of the matrix J
21
are calculated. This matrix is then given as
For calculating the off diagonal elements of the matrix J
12
we note from (4.47) that they are negative
of the off diagonal elements of J
21
. However the size of J
21
is (3 X 4) while the size of J
12
is (4 X 3).
Therefore to avoid this discrepancy we first compute a matrix M that is given by
The elements of the above matrix are computed in accordance with (4.44) and (4.45). We can then
define
Furthermore the diagonal elements of J
12
are overwritten in accordance with (4.48). This matrix is
then given by
Finally it can be noticed from (4.50) that J
22
= J
11
(1:3, 1:3). However the diagonal elements of J
22
are
then overwritten in accordance with (4.51). This gives the following matrix
From the initial conditions the power and reactive power are computed as
Consequently the mismatches are found to be
Then the updates at the end of the first iteration are given as
The load flow converges in 7 iterations when all the power and reactive power mismatches are below
10
6
.
Section VII: Load Flow Results
In this section we shall discuss the results of the load flow. It is to be noted here that both Gauss-
Seidel and Newton-Raphson methods yielded the same result. However the Newton-Raphson
method converged faster than the Gauss-Seidel method. The bus voltage magnitudes, angles of
each bus along with power generated and consumed at each bus are given in Table 4.4. It can be
seen from this table that the total power generated is 174.6 MW whereas the total load is 171 MW.
This indicates that there is a line loss of about 3.6 MW for all the lines put together. It is to be noted
that the real and reactive power of the slack bus and the reactive power of the P-V bus are computed
from (4.6) and (4.7) after the convergence of the load flow.
Table 4.4 Bus voltages, power generated and load after load flow convergence.
Bus
no.
Bus voltage Power generated Load
Magnitude (pu) Angle (deg) P (MW) Q (MVAr) P (MW) P (MVAr)
1 1.05 0 126.60 57.11 0 0
2 0.9826 5.0124 0 0 96 62
3 0.9777 7.1322 0 0 35 14
4 0.9876 7.3705 0 0 16 8
5 1.02 3.2014 48 15.59 24 11
The current flowing between the buses i and k can be written as
Therefore the complex power leaving bus- i is given by
Similarly the complex power entering bus- k is
Therefore the I
2
R loss in the line segment i-k is
The real power flow over different lines is listed in Table 4.5. This table also gives the I
2
R loss along
various segments. It can be seen that all the losses add up to 3.6 MW, which is the net difference
between power generation and load. Finally we can compute the line I
2
X drops in a similar fashion.
This drop is given by
However we have to consider the effect of line charging separately.
Table 4.5 Real power flow over different lines.
Power dispatched Power received
Line loss (MW)
from (bus) amount (MW) in (bus) amount (MW)
1 101.0395 2 98.6494 2.3901
1 25.5561 5 25.2297 0.3264
2 17.6170 3 17.4882 0.1288
3 0.7976 4 0.7888 0.0089
5 15.1520 2 14.9676 0.1844
5 18.6212 3 18.3095 0.3117
5 15.4566 4 15.2112 0.2454
Total = 3.5956
Consider the line segment 1-2. The voltage of bus-1 is V
1
= 1.05 < 0 per unit while that of bus-2 is V
2
= 0.9826 < 5.0124 per unit. From (4.52) we then have
per unit
Therefore the complex power dispatched from bus-1 is
where the negative signal indicates the power is leaving bus-1. The complex power received at bus-2
is
MW
(4.52)
(4.53)
(4.54)
(4.55)
(4.56)
Therefore out of a total amount of 101.0395 MW of real power is dispatched from bus-1 over the line
segment 1-2, 98.6494 MW reaches bus-2. This indicates that the drop in the line segment is 2.3901
MW. Note that
MW
where R
12
is resistance of the line segment 1-2. Therefore we can also use this method to calculate
the line loss.
Now the reactive drop in the line segment 1-2 is
MV Ar
We also get this quantity by subtracting the reactive power absorbed by bus-2 from that supplied by
bus-1. The above calculation however does not include the line charging. Note that since the line is
modeled by an equivalent- p , the voltage across the shunt capacitor is the bus voltage to which the
shunt capacitor is connected. Therefore the current I 12 flowing through line segment is not the
current leaving bus-1 or entering bus-2 - it is the current flowing in between the two charging
capacitors. Since the shunt branches are purely reactive, the real power flow does not get affected by
the charging capacitors. Each charging capacitor is assumed to inject a reactive power that is the
product of the half line charging admittance and square of the magnitude of the voltage of that at bus.
The half line charging admittance of this line is 0.03. Therefore line charging capacitor will inject
MV Ar
at bus-1. Similarly the reactive injected at bus-2 will be
MV Ar
The power flow through the line segments 1-2 and 1-5 are shown in Fig. 4.2.
(a)
(b)
Fig. 4.2 Real and reactive power flow through (a) line segment 1-2 and (b) line segment 1-5. The thin
lines indicate reactive power flow while the thick lines indicate real power flow.
Section VIII: Load Flow Programs In Matlab
The load flow programs are developed in MATLAB. Altogether there are 4 mfiles that are attached
with this chapter. The program listings and descriptions of these mfiles are given below. It must
however be emphasized that these are not general purpose programs and are written only for the
examples of this chapter.
Forming Y
bus
Matrix
Gauss- Seidel Load Flow
Solving Nonlinear Equations using Newton- Raphson
Newton - Raphson Load Flow
Forming Y
bus
Matrix
This is a function that can be called by various programs. The function can be invoked by the
statement
[yb,ych]=ybus;
where 'yb' and 'ych' are respectively the Y
bus
matrix and a matrix containing the line charging
admittances. It is assumed that the system data of Table 4.1 are given in matrix form and the matrix
that contains line impedances is 'zz', while 'ych' contains the line charging information. This program
is stored in the file ybus.m. The program listing is given below.
Chapter 5: Economic Operation of Power Systems
Overview
A good business practice is the one in which the production cost is minimized without sacrificing the
quality. This is not any different in the power sector as well. The main aim here is to reduce the
production cost while maintaining the voltage magnitudes at each bus. In this chapter we shall
discuss the economic operation strategy along with the turbine-governor control that are required to
maintain the power dispatch economically.
A power plant has to cater to load conditions all throughout the day, come summer or winter. It is
therefore illogical to assume that the same level of power must be generated at all time. The power
generation must vary according to the load pattern, which may in turn vary with season. Therefore the
economic operation must take into account the load condition at all times. Moreover once the
economic generation condition has been calculated, the turbine-governor must be controlled in such
a way that this generation condition is maintained. In this chapter we shall discuss these two aspects.
Section I: Economic Operation Of Power System
Economic Distribution of Loads between the Units of a Plant
Generating Limits
Economic Sharing of Loads between Different Plants
In an early attempt at economic operation it was decided to supply power from the most efficient
plant at light load conditions. As the load increased, the power was supplied by this most efficient
plant till the point of maximum efficiency of this plant was reached. With further increase in load, the
next most efficient plant would supply power till its maximum efficiency is reached. In this way the
power would be supplied by the most efficient to the least efficient plant to reach the peak demand.
Unfortunately however, this method failed to minimize the total cost of electricity generation. We
must therefore search for alternative method which takes into account the total cost generation of
all the units of a plant that is supplying a load.
Economic Distribution of Loads between the Units of a Plant
To determine the economic distribution of a load amongst the different units of a plant, the variable
operating costs of each unit must be expressed in terms of its power output. The fuel cost is the main
cost in a thermal or nuclear unit. Then the fuel cost must be expressed in terms of the power output.
Other costs, such as the operation and maintenance costs, can also be expressed in terms of the
power output. Fixed costs, such as the capital cost, depreciation etc., are not included in the fuel cost.
The fuel requirement of each generator is given in terms of the Rupees/hour. Let us define the input
cost of an unit- i , f
i
in Rs./h and the power output of the unit as P
i
. Then the input cost can be
expressed in terms of the power output as
Rs./h
(5.1)
The operating cost given by the above quadratic equation is obtained by approximating the power in
MW versus the cost in Rupees curve. The incremental operating cost of each unit is then computed
as
Let us now assume that only two units having different incremental costs supply a load. There will be
a reduction in cost if some amount of load is transferred from the unit with higher incremental cost to
the unit with lower incremental cost. In this fashion, the load is transferred from the less efficient unit
to the more efficient unit thereby reducing the total operation cost. The load transfer will continue till
the incremental costs of both the units are same. This will be optimum point of operation for both the
units.
The above principle can be extended to plants with a total of N number of units. The total fuel cost will
then be the summation of the individual fuel cost f
i
, i = 1, ... , N of each unit, i.e.,
Let us denote
that the total power that the plant is required to supply by P
T
, such that
where P
1
, ... , P
N
are the power supplied by the N different units.
The objective is minimize f
T
for a given P
T
. This can be achieved when the total difference df
T
becomes zero,
i.e.,
Now since the power supplied is assumed to be constant we have
Multiplying (5.6)
by and subtracting from (5.5) we get
The equality in (5.7) is satisfied when each individual term given in brackets is zero. This gives us
Also the partial derivative becomes a full derivative since only the term f
i
of f
T
varies with P
i
, i = 1, ...,
N . We then have
Rs./MWh
(5.2)
(5.3)
(5.4)
(5.5)
(5.6)
(5.7)
(5.8)
Generating Limits
It is not always necessary that all the units of a plant are available to share a load. Some of the units
may be taken off due to scheduled maintenance. Also it is not necessary that the less efficient units
are switched off during off peak hours. There is a certain amount of shut down and start up costs
associated with shutting down a unit during the off peak hours and servicing it back on-line during the
peak hours. To complicate the problem further, it may take about eight hours or more to restore the
boiler of a unit and synchronizing the unit with the bus. To meet the sudden change in the power
demand, it may therefore be necessary to keep more units than it necessary to meet the load
demand during that time. This safety margin in generation is called spinning reserve . The optimal
load dispatch problem must then incorporate this startup and shut down cost for without endangering
the system security.
The power
generation limit
of each unit is
then given by
the inequality constraints
The maximum limit P
max
is the upper limit of power generation capacity of each unit. On the other
hand, the lower limit P
min
pertains to the thermal consideration of operating a boiler in a thermal or
nuclear generating station. An operational unit must produce a minimum amount of power such that
the boiler thermal components are stabilized at the minimum design operating temperature.
Example 5.2
let us consider a generating station that contains a total number of three generating units. The fuel
costs of these units are given by
Rs./h
Rs./h
Rs./h
The generation limits of the units are
The total load that these units supply varies between 90 MW and 1250 MW. Assuming that all the
three units are operational all the time, we have to compute the economic operating settings as the
load changes.
The incremental costs of these units are
(5.10)
Rs./MWh
Rs./MWh
Rs./MWh
At the minimum load the incremental cost of the units are
Rs./MWh
Rs./MWh
Rs./MWh
Since units 1 and 3 have higher incremental cost, they must therefore operate at 30 MW each. The
incremental cost during this time will be due to unit-2 and will be equal to 26 Rs./MWh. With the
generation of units 1 and 3 remaining constant, the generation of unit-2 is increased till its incremental
cost is equal to that of unit-1, i.e., 34 Rs./MWh. This is achieved when P
2
is equal to 41.4286 MW, at
a total power of 101.4286 MW.
An increase in the total load beyond 101.4286 MW is shared between units 1 and 2, till their
incremental costs are equal to that of unit-3, i.e., 43.5 Rs./MWh. This point is reached when P
1
=
41.875 MW and P
2
= 55 MW. The total load that can be supplied at that point is equal to 126.875.
From this point onwards the load is shared between the three units in such a way that the incremental
costs of all the units are same. For example for a total load of 200 MW, from (5.4) and (5.9) we have
Solving the above three equations we get P
1
= 66.37 MW, P
2
= 80 MW and P
3
= 50.63 MW and an
incremental cost ( ) of 63.1 Rs./MWh. In a similar way the economic dispatch for various other load
settings are computed. The load distribution and the incremental costs are listed in Table 5.1 for
various total power conditions.
Table 5.1 Load distribution and incremental cost for the units of Example 5.1
P
T
(MW) P
1
(MW) P
2
(MW) P
3
(MW) (Rs./MWh)
90 30 30 30 26
101.4286 30 41.4286 30 34
120 38.67 51.33 30 40.93
126.875 41.875 55 30 43.5
150 49.62 63.85 36.53 49.7
200 66.37 83 50.63 63.1
300 99.87 121.28 78.85 89.9
400 133.38 159.57 107.05 116.7
500 166.88 197.86 135.26 143.5
600 200.38 236.15 163.47 170.3
700 233.88 274.43 191.69 197.1
800 267.38 312.72 219.9 223.9
906.6964 303.125 353.5714 250 252.5
1000 346.67 403.33 250 287.33
1100 393.33 456.67 250 324.67
1181.25 431.25 500 250 355
1200 450 500 250 370
1250 500 500 250 410
At a total load of 906.6964, unit-3 reaches its maximum load of 250 MW. From this point onwards
then, the generation of this unit is kept fixed and the economic dispatch problem involves the other
two units. For example for a total load of 1000 MW, we get the following two equations from (5.4) and
(5.9)
Solving which we get P
1
= 346.67 MW and P
2
= 403.33 MW and an incremental cost of 287.33
Rs./MWh. Furthermore, unit-2 reaches its peak output at a total load of 1181.25. Therefore any
further increase in the total load must be supplied by unit-1 and the incremental cost will only be
borne by this unit. The power distribution curve is shown in Fig. 5.1.
Fig.5.1 Power distribution between the units of Example 5.2.
Example 5.3
Consider two generating plant with same fuel cost and generation limits. These are given by
For a particular time of a year, the total load in a day varies as shown in Fig. 5.2. Also an additional
cost of Rs. 5,000 is incurred by switching of a unit during the off peak hours and switching it back on
during the during the peak hours. We have to determine whether it is economical to have both units
operational all the time.
Fig. 5.2 Hourly distribution of load for the units of Example 5.2.
Since both the units have identical fuel costs, we can switch of any one of the two units during the off
peak hour. Therefore the cost of running one unit from midnight to 9 in the morning while delivering
200 MW is
Rs.
Adding the cost of Rs. 5,000 for decommissioning and commissioning the other unit after nine hours,
the total cost becomes Rs. 167,225.
On the other hand, if both the units operate all through the off peak hours sharing power equally, then
we get a total cost of
Rs.
which is significantly less that the cost of running one unit alone.
Economic Sharing of Loads between Different Plants
So far we have considered the economic operation of a single plant in which we have discussed how
a particular amount of load is shared between the different units of a plant. In this problem we did not
have to consider the transmission line losses and assumed that the losses were a part of the load
supplied. However if now consider how a load is distributed between the different plants that are
joined by transmission lines, then the line losses have to be explicitly i ncluded in the economic
dispatch problem. In this section we shall discuss this problem.
When the transmission losses are included in the economic dispatch problem, we can modify (5.4) as
where P
LOSS
is
the total line loss. Since P
T
is assumed to be constant, we have
In the above equation dP
LOSS
includes the power loss due to every generator, i.e.
Also minimum generation cost implies df
T
= 0 as given in (5.5). Multiplying both (5.12) and (5.13) by
and combining we get
(5.11)
(5.12)
(5.13)
Adding (5.14) with (5.5) we obtain
The above equation satisfies when
Again since
from (5.16) we get
where L
i
is called the penalty factor of load- i and is given by
Example 5.4
Consider an area with N number of units. The power generated are defined by the vector
Then the transmission losses are expressed in general as
where B is a symmetric matrix given by
The elements B
ij
of the matrix B are called the loss coefficients . These coefficients are not constant
but vary with plant loading. However for the simplified calculation of the penalty factor L
i
these
coefficients are often assumed to be constant.
When the incremental cost equations are linear, we can use analytical equations to find out the
economic settings. However in practice, the incremental costs are given by nonlinear equations that
may even contain nonlinearities. In that case iterative solutions are required to find the optimal
generator settings.
(5.14)
(5.15)
(5.16)
(5.17)
(5.18)
(5.19)
Section II: Automatic Generation Control
Load Frequency Control
Automatic Generation Control
Electric power is generated by converting mechanical energy into electrical energy. The rotor mass,
which contains turbine and generator units, stores kinetic energy due to its rotation. This stored
kinetic energy accounts for sudden increase in the load. Let us denote the mechanical torque input by
T
m
and the output electrical torque by T
e
. Neglecting the rotational losses, a generator unit is said to
be operating in the steady state at a constant speed when the difference between these two elements
of torque is zero. In this case we say that the accelerating torque
(5.20)
is zero.
When the electric power demand increases suddenly, the electric torque increases. However, without
any feedback mechanism to alter the mechanical torque, T
m
remains constant. Therefore the
accelerating torque given by (5.20) becomes negative causing a deceleration of the rotor mass. As
the rotor decelerates, kinetic energy is released to supply the increase in the load. Also note that
during this time, the system frequency, which is proportional to the rotor speed, also decreases. We
can thus infer that any deviation in the frequency for its nominal value of 50 or 60 Hz is indicative of
the imbalance between T
m
and T
e
. The frequency drops when T
m
< T
e
and rises when T
m
> T
e
.
The steady state power-frequency relation is shown in Fig. 5.3. In this figure the slope of the P
ref
line
is negative and
is given by
where R is called the regulating constant . From this figure we can write the steady state power
frequency relation as
Fig. 5.3 A typical steady-state power-frequency curve.
Suppose an interconnected power system contains N turbine-generator units. Then the steady-state
power-frequency relation is given by the summation of (5.22) for each of these units as
(5.21)
(5.22)
In the above equation, P
m
is the total change in turbine-generator mechanical power and P
ref
is the
total change in the reference power settings in the power system. Also note that since all the
generators are supposed to work in synchronism, the change is frequency of each of the units is the
same and is denoted by f. Then the frequency response characteristics is defined as
We can therefore modify (5.23) as
Example 5.5
Consider an interconnected 50-Hz power system that contains four turbine-generator units rated 750
MW, 500 MW, 220 MW and 110 MW. The regulating constant of each unit is 0.05 per unit based on
its own rating. Each unit is operating on 75% of its own rating when the load is suddenly dropped by
250 MW. We shall choose a common base of 500 MW and calculate the rise in frequency and drop in
the mechanical power output of each unit.
The first step in the process is to convert the regulating constant, which is given in per unit in the base
of each generator, to a common base. This is given as
(5.26)
We can therefore write
Therefore
p.u.
We can therefore calculate the total change in the frequency from (5.25) while assuming P
ref
= 0,
i.e., for no change in the reference setting. Since the per unit change in load - 250/500 = - 0.5 with the
negative sign accounting for load reduction, the change in frequency is given by
(5.23)
(5.24)
Then the change in the mechanical power of each unit is calculated from (5.22) as
It is to be noted that once P
m2
is calculated to be - 79.11 MW, we can also calculate the changes in
the mechanical power of the other turbine-generators units as
This implies that each turbine-generator unit shares the load change in accordance with its own
rating.
Load Frequency Control
Modern day power systems are divided into various areas. For example in India , there are five
regional grids, e.g., Eastern Region, Western Region etc. Each of these areas is generally
interconnected to its neighboring areas. The transmission lines that connect an area to its
neighboring area are called tie-lines . Power sharing between two areas occurs through these tie-
lines. Load frequency control, as the name signifies, regulates the power flow between different areas
while holding the frequency constant.
As we have in Example 5.5 that the system frequency rises when the load decreases if P
ref
is kept
at zero. Similarly the frequency may drop if the load increases. However it is desirable to maintain the
frequency constant such that f=0 . The power flow through different tie-lines are scheduled - for
example, area- i may export a pre-specified amount of power to area- j while importing another pre-
specified amount of power from area- k . However it is expected that to fulfill this obligation, area- i
absorbs its own load change, i.e., increase generation to supply extra load in the area or decrease
generation when the load demand in the area has reduced. While doing this area- i must however
maintain its obligation to areas j and k as far as importing and exporting power is concerned. A
conceptual diagram of the interconnected areas is shown in Fig. 5.4.
Fig. 5.4 Interconnected areas in a power system.
We can therefore state that the load frequency control (LFC) has the following two objectives:
Hold the frequency constant ( f = 0) against any load change. Each area must contribute to
absorb any load change such that frequency does not deviate.
Each area must maintain the tie-line power flow to its pre-specified value.
The first step in
the LFC is to
form the area
control error (ACE) that is defined as
where P
tie
and P
sch
are tie-line power and scheduled power through tie-line respectively and the
constant B
f
is called the frequency bias constant .
The change in the reference of the power setting P
ref, i
, of the area- i is then obtained by the
feedback of the ACE through an integral controller of the form
where K
i
is the integral gain. The ACE is negative if the net power flow out of an area is low or if the
frequency has dropped or both. In this case the generation must be increased. This can be achieved
by increasing P
ref, i
. This negative sign accounts for this inverse relation between P
ref, i
and ACE.
The tie-line power flow and frequency of each area are monitored in its control center. Once the ACE
is computed and P
ref, i
is obtained from (5.28), commands are given to various turbine-generator
controls to adjust their reference power settings.
Example 5.6
Consider a two-area power system in which area-1 generates a total of 2500 MW, while area-2
generates 2000 MW. Area-1 supplies 200 MW to area-2 through the inter-tie lines connected
between the two areas. The bias constant of area-1 (
1
) is 875 MW/Hz and that of area-2 (
2
) is
700 MW/Hz. With the two areas operating in the steady state, the load of area-2 suddenly increases
by 100 MW. It is desirable that area-2 absorbs its own load change while not allowing the frequency
to drift.
The area control errors of the two areas are given by
and
Since the net change in the power flow through tie-lines connecting these two areas must be zero, we
have
Also as the transients die out, the drift in the frequency of both these areas is assumed to be
constant, i.e.,
If the load frequency controller (5.28) is able to set the power reference of area-2 properly, the ACE of
the two areas will be zero, i.e., ACE
1
= ACE
2
= 0. Then we have
This will imply that f will be equal to zero while maintaining P
tie1
=P
tie2
= 0. This signifies that area-
2 picks up the additional load in the steady state.
Coordination Between LFC And Economic Dispatch
Both the load frequency control and the economic dispatch issue commands to change the power
setting of each turbine-governor unit. At a first glance it may seem that these two commands can be
conflicting. This however is not true. A typical automatic generation control strategy is shown in Fig.
(5.27)
(5.28)
5.5 in which both the objective are coordinated. First we compute the area control error. A share of
this ACE, proportional to
i
, is allocated to each of the turbine-generator unit of an area. Also the
share of unit- i ,
i
X ( P
DK
- P
k
), for the deviation of total generation from actual generation is
computed. Also the error between the economic power setting and actual power setting of unit- i is
computed. All these signals are then combined and passed through a proportional gain K
i
to obtain
the turbine-governor control signal.
Fig. 5.5 Automatic generation control of unit-i.
Chapter 6: Short Circuit Studies - Symmetrical Faults
Introduction
Short circuits occur in power system due to various reasons like, equipment failure, lightning strikes,
falling of branches or trees on the transmission lines, switching surges, insulation failures and other
electrical or mechanical causes. All these are collectively called faults in power systems.
A fault usually results in high current flowing through the lines and if adequate protection is not taken,
may result in damages in the power apparatus.
In this chapter we shall discuss the effects of symmetrical faults on the system. Here the term
symmetrical fault refers to those conditions in which all three phases of a power system are grounded
at the same point. For this reason the symmetrical faults sometimes are also called three-line-to-
ground (3LG) faults.
Section I: Transients in R-L Circuits
DC Sourse
AC Sourse
Fault in an AC Circuit
Transients in R-L Circuits
In this section we shall consider transients in a circuit that contains a resistor and inductor (R - L
circuit). Consider the circuit shown in Fig. 6.1 that contains an ideal source (
s
), a resistor ( R ), an
inductor ( L ) and a switch ( S ). It is assumed that the switch is open and is closed at an instant of
time t = 0. This implies that the current i is zero before the closing of the switch. We shall first discuss
the effect of closing the switch on the line current (i ) when the source is dc. Following this we shall
study the effect when the source is ac and will show that the shape of the transient current changes
with the changes in the phase of the source voltage waveform at the instant of closing the switch.
Fig 6.1 A Simple R - L Circuit
DC Source
Let us assume that the source voltage is dc and is given by
s
= V
dc
. Then the line current is given by
the differential equation
The solution of the above equation is written in the form
Since the initial current i (0) = 0 and since
s
( ) = V
dc
for 0 t < , we can rewrite the above
equation as
where T = L / R is the time constant of the circuit.
Let us assume R = 1 , L = 10 mH and V
dc
= 100 V. Then the time response of the current is as
shown in Fig. 6.2. It can be seen that the current reaches at steady state value of 100 A. The time
constant of the circuit is 0.01 s. This is defined by the time in which the current i ( t ) reaches 63.2% of
its final value and is obtained by substituting t = T. Note that the slope of the curve is given by
Fig 6.2 Current in the R-L circuit when the source is dc
AC Source
The current response remains unchanged when the voltage source is dc. This however is not the
case when the circuit is excited by an ac source. Let us assume that the source voltage is now given
by
where is the
(6.1)
(6.2)
(6.3)
(6.4)
(6.5)
phase angle of the applied voltage. We shall show that the system response changes with a change
in .
The solution of (6.2) for the source voltage given in (6.5) is
The system response for V
m
= 100 V and = 45 is shown in Fig. 6.3. In this figure both i
ac
and i
dc
are also shown. It can be seen that i
ac
is the steady state waveform of the circuit, while i
dc
dies out
once the initial transient phase is over. Fig. 6.4 shows the response of the current for different values
of a . Since the current is almost inductive, it can be seen that the transient is minimum when = 90
, i.e., the circuit is switched on almost at the zero-crossing of the current. On the other hand, the
transient is maximum when = 0 , i.e., almost at the peak of the current.
Fig 6.3 Transient in current and its ac and dc components at the instant of switch closing
Fig 6.4 Transient in current for different values of
(6.6)
A
(6.7)
A
(6.8)
,
Fault in an AC Circuit
Now consider the single-phase circuit of Fig. 6.5 where V
s
= 240 V (rms), the system frequency is 50
Hz, R = 0.864 , L = 11 mH ( L = 3.46 ) and the load is R-L comprising of an 8.64 W resistor and
a 49.5 mH inductor ( L = 15.55 ). With the system operating in the steady state, the switch S is
suddenly closed creating a short circuit. The current (i ) waveform is shown in Fig. 6.6. The current
phasor before the short circuit occurs is
A
This means that the pre-fault current has a peak value of 15.97 A.
Fig. 6.5 A single-phase circuit in which a source supplies a load through a source impedance.
Fig 6.6 The current waveform of the circuit of Fig 6.5 before and after the closing of the switch S
Once the fault occurs and the system is allowed to reach the steady state, the current phasor is given
by
This current has a peak value of 95.28 A. However it can be seen that the current rises suddenly and
the first peak following the fault is 124 A which is about 30% higher than the post-fault steady-state
value. Also note that the peak value of the current will vary with the instant of the occurrence of the
fault. However the peak value of the current is nearly 8 times the pre-fault current value in this case.
In general, depending on the ratio of source and load impedances, the faulted current may shoot up
anywhere between 10 and 20 times the pre-fault current.
Short Circuit in an Unloaded Synchronous Generator
Fig. 6.7 shows a typical response of the armature current when a three-phase symmetrical short
circuit occurs at the terminals of an unloaded synchronous generator.
Fig. 6.7 Armature current of a synchronous generator as a short circuit occurs at its terminals.
It is assumed that there is no dc offset in the armature current. The magnitude of the current
decreases exponentially from a high initial value. The instantaneous expression for the fault current is
given by
where V
t
is the
magnitude of the terminal voltage, is its phase angle and
is the direct axis subtransient reactance
is the direct axis transient reactance
is the direct axis synchronous reactance
with . The time constants are
is the direct axis subtransient time constant
is the direct axis transient time constant
In the expression of (6.9) we have neglected the effect of the armature resistance hence = /2. Let
us assume that the fault occurs at time t = 0. From (6.9) we get the rms value of the current as
which is called the subtransient fault current. The duration of the subtransient current is dictated by
the time constant T
d
. As the time progresses and T
d
< t < T
d
, the first exponential term of (6.9)
will start decaying and will eventually vanish. However since t is still nearly equal to zero, we have the
following rms value of the current
This is called the transient fault current. Now as the time progress further and the second
exponential term also decays, we get the following rms value of the current for the sinusoidal steady
state
(6.9)
(6.10)
(6.11)
(6.12)
In addition to the ac, the fault currents will also contain the dc offset. Note that a symmetrical fault
occurs when three different phases are in three different locations in the ac cycle. Therefore the dc
offsets in the three phases are different. The maximum value of the dc offset is given by
where T
A
is the armature time constant.
Section III: Symmetrical Fault in a Power System
Calculation of Fault Current Using Impedance Diagram
Calculation of Fault Current Using Z
bus
Matrix
Calculation of Fault Current Using Impedance Diagram
Let us first illustrate the calculation of the fault current using the impedance diagram with the help of
the following examples
Example 6.1
Consider the power system of Fig. 6.8 in which a synchronous generator supplies a synchronous
motor. The motor is operating at rated voltage and rated MVA while drawing a load current at a power
factor of 0.9 (lagging) when a three phase symmetrical short circuit occurs at its terminals. We shall
calculate the fault current that flow from both the generator and the motor.
We shall choose a base of 50 MVA, 20 kV in the circuit of the generator. Then the motor synchronous
reactance is given by
per unit
Also the base impedance in the circuit of the transmission line is
Fig. 6.8 A generator supplying a motor load though a transmission line.
Therefore the impedance of the transmission line is
per unit
The impedance diagram for the circuit is shown in Fig. 6.9 in which the switch S indicates the fault.
Fig. 6.9 Impedance diagram of the circuit of Fig. 6.8.
The motor draws a load current at rated voltage and rated MVA with 0.9 lagging power factor.
Therefore
(6.13)
per unit
Then the subtransient voltages of the motor and the generator are
per unit
per unit
Hence the subtransient fault currents fed by the motor and the generator are
per unit
per unit
and the total current flowing to the fault is
per unit
Note that the base current in the circuit of the motor is
A
Therefore while the load current was 1603.8 A, the fault current is 7124.7 A.
Example 6.2
We shall now solve the above problem differently. The Thevenin impedance at the circuit between the
terminals A and B of the circuit of Fig. 6.9 is the parallel combination of the impedances j 0.4 and j
0.5148. This is then given as
per unit
Since voltage at the motor terminals before the fault is 1.0 per unit, the fault current is
per unit
If we neglect the pre-fault current flowing through the circuit, then fault current fed by the motor and
the generator can be determined using the current divider principle, i.e.,
per unit
per unit
If, on the other hand, the pre-fault current is not neglected, then the fault current supplied by the
motor and the generator are
per unit
per unit
Calculation of Fault Current Using Z
bus
Matrix
Consider the circuit of Fig. 3.3 which is redrawn as shown in Fig. 6.10.
Fig. 6.10 Network depicting a symmetrical fault at bus-4.
We assume that a symmetrical fault has occurred in bus-4 such that it is now connected to the
reference bus. Let us assume that the pre-fault voltage at this bus is V
f
. To denote that bus-4 is short
circuit, we add two voltage sources V
f
and - V
f
together in series between bus-4 and the reference
bus. Also note that the subtransient fault current I
f
flows from bus-4 to the reference bus. This
implies that a current that is equal to - I
f
is injected into bus-4. This current, which is due to the
source - V
f
will flow through the various branches of the network and will cause a change in the bus
voltages. Assuming that the two sources and V
f
are short circuited. Then - V
f
is the only source left in
the network that injects a current - I
f
into bus-4. The voltages of the different nodes that are caused
by the voltage - V
f
and the current - I
f
are then given by
where the prefix indicates the changes in the bus voltages due to the current - I
f
.
From the fourth row of (6.14) we can write
Combining (6.14) and (6.15) we get
We further assume that the system is unloaded before the fault occurs and that the magnitude and
phase angles of all the generator internal emfs are the same. Then there will be no current circulating
anywhere in the network and the bus voltages of all the nodes before the fault will be same and equal
to V
f
. Then the new altered bus voltages due to the fault will be given from (6.16) by
Example 6.3
(6.14)
(6.15)
(6.16)
Let us consider the same system as discussed in Example 3.1 except that we assume that the
internal voltages of both the generators are equal to 1.0 < 0 . Then the current injected in both bus-1
and 2 will be given by 1.0/ j 0.25 = - j 4.0 per unit. We therefore get the pre-fault bus voltages using
the Z
bus
matrix given in Example 3.1 as
Now the altered bus voltages for a symmetrical fault in bus-4 are given from (6.17) as
per unit
per unit
per unit
per unit
Also since the Thevenin impedance looking into the network at bus-4 is Z
44
(see Section 3.4), the
subtransient fault current flowing from bus-4 is
pu
Section IV: Circuit Breaker Selection
A typical circuit breaker operating time is given in Fig. 6.11. Once the fault occurs, the protective
devices get activated. A certain amount of time elapses before the protective relays determine that
there is overcurrent in the circuit and initiate trip command. This time is called the detection time.
The contacts of the circuit breakers are held together by spring mechanism and, with the trip
command, the spring mechanism releases the contacts. When two current carrying contacts part, a
voltage instantly appears at the contacts and a large voltage gradient appears in the medium
between the two contacts. This voltage gradient ionizes the medium thereby maintaining the flow of
current. This current generates extreme heat and light that is called electric arc. Different
mechanisms are used for elongating the arc such that it can be cooled and extinguished. Therefore
the circuit breaker has to withstand fault current from the instant of initiation of the fault to the time the
arc is extinguished.
Fig. 6.11 Typical circuit breaker operating time.
Two factors are of utmost importance for the selection of circuit breakers. These are:
The maximum instantaneous current that a breaker must withstand and
The total current when the breaker contacts part.
In this chapter we have discussed the calculation of symmetrical subtransient fault current in a
network. However the instantaneous current following a fault will also contain the dc component. In a
high power circuit breaker selection, the subtransient current is multiplied by a factor of 1.6 to
determine the rms value of the current the circuit breaker must withstand. This current is called the
momentary current . The interrupting current of a circuit breaker is lower than the momentary
current and will depend upon the speed of the circuit breaker. The interrupting current may be
asymmetrical since some dc component may still continue to decay.
Breakers are usually classified by their nominal voltage, continuous current rating, rated maximum
voltage, K -factor which is the voltage range factor, rated short circuit current at maximum voltage and
operating time. The K -factor is the ratio of rated maximum voltage to the lower limit of the range of
the operating voltage. The maximum symmetrical interrupting current of a circuit breaker is given by K
times the rated short circuit current.
Overview
An unbalanced three-phase system can be resolved into three balanced systems in the sinusoidal
steady state. This method of resolving an unbalanced system into three balanced phasor system has
been proposed by C. L. Fortescue. This method is called resolving symmetrical components of the
original phasors or simply symmetrical components.
In this chapter we shall discuss symmetrical components transformation and then will present how
unbalanced components like Y- or -connected loads, transformers, generators and transmission
lines can be resolved into symmetrical components. We can then combine all these components
together to form what are called sequence networks .
Section I: Symmetrical Components
Symmetrical Component Transformation
Real and Reactive Power
Orthogonal Transformation
Symmetrical Components
A system of three unbalanced phasors can be resolved in the following three symmetrical
components:
Positive Sequence: A balanced three-phase system with the same phase sequence as the
original sequence.
Negative sequence: A balanced three-phase system with the opposite phase sequence as the
original sequence.
Zero Sequence: Three phasors that are equal in magnitude and phase.
Fig. 7.1 depicts a set of three unbalanced phasors that are resolved into the three sequence
components mentioned above. In this the original set of three phasors are denoted by V
a
, V
b
and V
c
,
while their positive, negative and zero sequence components are denoted by the subscripts 1, 2 and
0 respectively. This implies that the positive, negative and zero sequence components of phase-a are
denoted by V
a1
, V
a2
and V
a0
respectively. Note that just like the voltage phasors given in Fig. 7.1 we
can also resolve three unbalanced current phasors into three symmetrical components.
Fig. 7.1 Representation of (a) an unbalanced network, its (b) positive sequence, (c) negative
sequence and (d) zero sequence.
Symmetrical Component Transformation
Before we discuss the symmetrical component transformation, let us first define the -operator. This
has been given in (1.34) and is reproduced below.
Note that for the above operator the following relations hold
Also note that we have
Using the a -operator we can write from Fig. 7.1 (b)
Similarly from Fig. 7.1 (c) we get
Finally from Fig. 7.1 (d) we get
Therefore,
The symmetrical component transformation matrix is then given by
(7.1)
(7.2)
(7.3)
(7.4)
(7.5)
(7.6)
(7.7)
(7.8)
(7.9)
(7.10)
(7.11)
Defining the vectors V
a012
and V
abc
as
we can write (7.4) as
where C is the symmetrical component transformation matrix and is given by
The original phasor components can be obtained from the inverse symmetrical component
transformation, i.e.,
Finally, if we define a set of unbalanced current phasors as I
abc
and their symmetrical components as
I
a012
, we can then define
Example 7.1
Let us consider a set of balanced voltages given in per unit by
These imply
Then from (7.7) we get
(7.12)
(7.13)
(7.14)
(7.15)
We then see that for a balanced system the zero and negative sequence voltages are zero. Also the
positive sequence voltage is the same as the original system, i.e.,
Example 7.2
All the quantities given in this example are in per unit. Let us now consider the following set of three
unbalanced voltages
If we resolve them using (7.4) we then have
Therefore we have
Furthermore note that
Real and Reactive Power
The three-phase power in the original unbalanced system is given by
where I* is the complex conjugate of the vector I . Now from (7.10) and (7.15) we get
From (7.11) we get
Therefore from
(7.17) we get
We then find that the complex power is three times the summation of the complex power of the three
phase sequences.
Example 7.3
Let us consider the voltages given in Example 7.2. Let us further assume that these voltages are line-
to-neutral voltages and they supply a balanced Y-connected load whose per phase impedance is Z
Y
= 0.2 + j 0.8 per unit. Then the per unit currents in the three phases are
(7.16)
(7.17)
(7.18)
pu
pu
pu
Then the real and reactive power consumed by the load is given by
Now using the transformation (7.15) we get
pu
From the results given in Example 7.2 and from the above values we can compute the zero sequence
complex power as
pu
The positive sequence complex power is
pu
Finally the negative sequence complex power is
pu
Adding the three complex powers together we get the total complex power consumed by the load as
pu
Orthogonal Transformation
Instead of the transformation matrix given in (7.13), let us instead use the transformation matrix
We then have
Note from (7.19) and (7.20) that C
-1
= ( C
T
)* . We can therefore state C( C
T
)* = I
3
, where I
3
is (3 3)
identity matrix. Therefore the transformation matrices given in (7.19) and (7.20) are orthogonal. Now
since
(7.19)
(7.20)
we can write from (7.17)
We shall now discuss how different elements of a power system are represented in terms of their
sequence components. In fact we shall show that each element is represented by three equivalent
circuits, one for each symmetrical component sequence.
Section II: Sequence Circuits for Loads
In this section we shall construct sequence circuits for both Y and -connected loads separately.
Sequence Circuit for a Y-Connected Load
Sequence Circuit for a -Connected Load
Sequence Circuit for a Y-Connected Load
Consider the balanced Y-connected load that is shown in Fig. 7.2. The neutral point (n) of the
windings are grounded through an impedance Z
n
. The load in each phase is denoted by Z
Y
. Let us
consider phase-a of the load. The voltage between line and ground is denoted by V
a
, the line-to-
neutral voltage is denoted by V
an
and voltage between the neutral and ground is denoted by V
n
. The
neutral current
is then
Therefore there will not be any positive or negative sequence current flowing out of the neutral point.
Fig. 7.2 Schematic diagram of a balanced Y-connected load.
The voltage drop between the neutral and ground is
Now
We can write similar expression for the other two phases. We can therefore write
(7.21)
(7.22)
(7.23)
(7.24)
(7.25)
Pre-multiplying both sides of the above equation by the matrix C and using (7.8) we get
Now since
We get from (7.26)
We then find that the zero, positive and negative sequence voltages only depend on their respective
sequence component currents. The sequence component equivalent circuits are shown in Fig. 7.3.
While the positive and negative sequence impedances are both equal to Z
Y
, the zero sequence
impedance is equal to
If the neutral is grounded directly (i.e., Z
n
= 0), then Z
0
= Z
Y
. On the other hand, if the neutral is kept
floating (i.e., Z
n
= ), then there will not be any zero sequence current flowing in the circuit at all.
Fig. 7.3 Sequence circuits of Y-connected load: (a) positive, (b) negative and (c) zero sequence.
Sequence Circuit for a -Connected Load
Consider the balanced -connected load shown in Fig. 7.4 in which the load in each phase is
denoted by Z
.
Example 7.4
Consider the circuit shown in Fig. 7.5 in which a -connected load is connected in parallel with a Y-
connected load. The neutral point of the Y-connected load is grounded through an impedance.
Applying Kirchoff's current law at the point P in the circuit we get
The above expression can be written in terms of the vector V
abc
as
Since the load is balanced we can write
Fig. 7.5 Parallel connection of balanced D and Y-connected loads.
Pre-multiplying both sides of the above expression by the transformation matrix C we get
Now since
(7.30)
(7.31)
we get
Separating the three components, we can write from the above equation
Suppose now if we convert the -connected load into an equivalent Y, then the composite load will
be a parallel combination of two Y-connected circuits - one with an impedance of Z
Y
and the other
with an impedance of Z
/3. Therefore the positive and the negative sequence impedances are given
by the parallel combination of these two impedances. The positive and negative sequence impedance
is then given by
Now refer to Fig. 7.5. The voltage V
n
is given by
From Fig. 7.5 we can also write I
a
= I
a
+ I
aY
. Therefore
This implies that I
a0
= I
ay0
and hence V
n
= 3Z
n
I
a0
. We can then rewrite the zero sequence current
expression as
It can be seen that the Z
0
= 23.96 = 0.4182 rad and the maximum power transferred as
(9.15)
(9.16)
(9.17)
(9.18)
per unit
We have to find the critical clearing angle.
From (9.15) the accelerating area is computed as by note that P
e
= 0 during this time. This is then
given by
To calculate the decelerating area we note that
m
= - 0.4182 = 2.7234 rad. This area is computed
by noting that P
e
= 2.2164 sin( ) during this time. Therefore
Equating A
1
= A
2
and rearranging we get
Now a frequently asked question is what does the critical clearing angle mean?
Since we are interested in finding out the maximum time that the circuit breakers may take for
opening, we should be more concerned about the critical clearing time rather than clearing angle.
Furthermore, notice that the clearing angle is independent of the generalized inertia constant H .
Hence we can comment that the critical clearing angle in this case is true for any generator that has a
d-axis transient reactance of 0.20 per unit. The critical clearing time, however, is dependent on H and
will vary as this parameter varies.
To obtain a description for the critical clearing time, let us consider the period during whi ch the fault
occurs. We then have P
e
= 0. We can therefore write from
Integrating the above equation with the initial acceleration being zero we get
Further integration will lead to
Replacing by
cr
and t by t
cr
in the above equation, we get the critical clearing time as
Example 9.4:
In Example 9.2, let us choose the system frequency as 50 Hz such that
s
is 100. Also let us
choose H as 4 MJ/MVA. Then with
cr
being 1.5486 rad,
0
being 0.4182 rad and P
m
being 0.9 per
unit, we get the following critical clearing time from (9.20)
s
To illustrate the response of the load angle , the swing equation is simulated in MATLAB. The swing
equation of (9.14) is then expressed as
where
r
is the deviation for the rotor speed from the synchronous speed
s
. It is to be noted that
the swing equation of (9.21) does not contain any damping. Usually a damping term, that is
proportional to the machine speed
r
, is added with the accelerating power. Without the damping
the load angle will exhibit a sustained oscillation even when the system remains stable when the fault
cleared within the critical clearing time.
Fig. 9.4 Stable and unstable system response as a function of clearing time.
Fig. 9.4 depicts the response of the load angle for two different values of load angle. It is assumed
that the fault occurs at 0.5 s when the system is operating in the steady state delivering 0.9 per unit
power. The load angle during this time is constant at 23.96 . The load angle remains stable, albeit
the sustained oscillation when the clearing time t
cl
is 0.253 s. The clearing angle during this time is
88.72 . The system however becomes unstable when the clearing time 0.2531s and the load angle
increases asymptotically. The clearing time in this case is 88.77 . This is called the Loss of
Synchronism. It is to be noted that such increase in the load angle is not permissible and the
protection device will isolate the generator from the system.
The clearing time of (8.20) is derived based on the assumption that the electrical power P
e
becomes
zero during the fault as in (8.19). This need not be the case always. In that even we have to resort to
finding the clearing time using the numerical integration of the swing equation. See example 9.5 to
illustrates the point.
Example 9.5
Consider the system in which a generator is connected to an infinite bus through a double circuit
transmission line as shown in Fig. 9.5. The per unit system reactances that are converted in a
common base, are also shown in this figure. Let us assume that the infinite bus voltage is 1< 0 . The
generator is delivering 1.0 per unit real power at a lagging power factor of 0.9839 to the infinite bus.
While the generator is operating in steady state, a three-phase bolted short circuit occurs in the
transmission line connecting buses 2 and 4 - very near to bus 4. The fault is cleared by opening the
circuit breakers at the two ends of this line. We have to find the critical clearing time for various values
of H .
(9.21)
Fig. 9.5 Schematic diagram of the power system of Example 9.4.
Let the current flowing through the line be denoted by I . Then the power delivered to the infinite bus
is
From the above equation we get
The total impedance during the time when both the lines are operational, the impedance between the
generator and the infinite bus is j(0.3+0.1+0.1) = j 0.5 per unit. Then the generator internal voltage is
Therefore the machine internal voltage is E' = 1.2 per unit its angle is 24.625 or 0.4298 rad.
The pre-fault equivalent circuit is shown in Fig. 9.6 (a). From this figure we can write the power
transfer equation as
Once the fault is cleared by opening of the breakers connected near buses 2 and 4, only of one the
two lines will be operational. Therefore during the post fault period, the impedance between the
generator and the infinite bus is j (0.3+0.2+0.6) = j 0.5 per unit as shown in Fig. 9.6 (b). Then the
post-fault power transfer equation is given by
Fig. 9.6 Equivalent circuit: (a) pre-fault and (b) post-fault.
It is to be noted that since one of the two lines is functional during the fault, the power transfer during
the fault will not be zero. The equivalent circuit during the fault is shown in Fig. 9.7. Since the fault
has occurred very near to bus-4, we can assume that this bus has been short circuited. We shall find
the Thevenin equivalent of the portion of the circuit to the right of the points A and B .
Fig. 9.7 Faulted equivalent circuit.
The circuit between buses 2 and 3 has been converted into an equivalent Y using -Y transformation.
This is shown in Fig. 9.8 (a). From this figure we find the Thevenin impedance as
per unit
Also the Thevenin voltage is then given by
per unit
The reduced circuit is shown in Fig. 9.8 (b). From this circuit we can write the following power transfer
equation during the fault
Fig. 9.8 Equivalent circuit during the fault: (a) D -Y transformed and (b) Thevenin equivalent.
Three power-angle curves are shown in Fig. 9.9. From this figure we find that
rad
Fig. 9.9 Power-angles curves for the three modes of operation of the system of Example 9.4.
The accelerating area is given by
and the decelerating area is
Equating the two areas we get
As mentioned earlier, the critical clearing angle depends on the system network configuration. The
critical clearing time, however, is dependent on H and will vary as this parameter varies. Usually
numerical methods are employed for finding out the clearing time. We shall however demonstrate the
effect of H through a MATLAB program. The program uses the built-in ordinary differential equation
solver though which the swing equations are solved. The results obtained are listed in Table 9.1. It
can be seen that as the value of H increases, the clearing time also increases, even though the
clearing angle remains the same. This is obvious as the value of H increases, the response of the
system becomes more sluggish due to larger inertia. Hence, the rotor takes more time to accelerate.
Table 9.1 Effect of H on critical clearing time
H ( MJ/MVA) Approximate Critical Clearing Time (s)
2 0.2783
4 0.3936
6 0.4821
8 0.5566
10 0.6223
Section IV: Multimachine Stability
Oscillations in a Two Area System
Consider Fig. 9.10, which depicts a number of weights that are suspended by elastic strings. The
weights represent generators and the electric transmission lines being represented by the strings.
Note that in a transmission system, each transmission line is loaded below its static stability limit.
Similarly, when the mechanical system is in static steady state, each string is loaded below its break
point. At this point one of the strings is suddenly cut. This will result in transient oscillations in the
coupled strings and all the weights will wobble. In the best possible case, this may result in the
coupled system settling down to a new steady state. On the other hand, in the worst possible
scenario this may result in the breaking of one more additional string, resulting in a chain reaction in
which more strings may break forcing a system collapse. In a similar way, in an interconnected
electric power network, the tripping of a transmission line may cause a catastrophic failure in which a
large number of generators are lost forcing a blackout in a large area.
Modern power systems are interconnected and operate close to their transient and steady state
stability limits. In large interconnected systems, it is common to find a natural response of a group of
closely coupled machines oscillating against other groups of machines. These oscillations have a
frequency range of 0.1 Hz to 0.8 Hz. The lowest frequency mode involves all generators of the
system. This oscillation groups the system into two parts - with generators in one part oscillating
against those of the the other part. The higher frequency modes are usually localized with small
groups oscillating against each other. Unfortunately, the inter-area oscillation can be initiated by a
small disturbance in any part of the system. These small frequency oscillations fall under the category
of dynamic stability and are analysed in linear domain through the liberalisation of the entire
interconnected systems model.
Inter-area oscillations manifest wherever the power system is heavily interconnected. The
oscillations, unless damped, can lead to grid failure and total system collapse. Low frequency
oscillations in the range of 0.04 Hz to 0.06 Hz were observed in the Pacific North West region as
early as 1950. Improper speed governor control of hydro units created these oscillations. The
Northern and Southern regions of WSCC were interconnected by a 230 kV line in 1964. Immediately
the system experienced a 0.1 Hz oscillation resulting in over 100 instances of opening of the tie line in
the first nine months of operation. Some system damping was provided through the modification in
the hydro turbine governors.
A 500 kV pacific intertie and another 400 kV HVDC system was commissioned in 1968. This raised
the frequency of oscillation from 0.1 Hz to 0.33 Hz and these oscillations could no longer be
controlled through governor action alone. In late 1980's a new intertie joined the WSCC system to
Alberta and British Columbia in Canada . As a result of this interconnection, the two different
oscillation frequencies manifested - one at 0.29 Hz and the other at 0.45 Hz.
Ontario Hydro is one of the largest utilities in North America . Due to the vast and sparsely populated
topology of Canada , the operating span of Ontario hydro is over 1000 km from East to West and
from North to South. The Ontario Hydro system is connected to the neighbouring Canadian provinces
and the North Western region of the United States . In 1959 Ontario Hydro was connected to
Michigan in the South and Quebec Hydro in the East. As a result of this connection, a 0.25 Hz
oscillation was observed and a result of this it was decided to remove the tie with Quebec and retain
the tie to Michigan . The Western portion of Ontario was connected to neighbouring Manitoba in 1956
and then Manitoba was connected to its neighbour Saskatchewan in 1960. This resulted in oscillation
in the frequency range 0.35 Hz to 0.45 Hz often tripping the tie. As a result of this, Ontario Hydro
decided to commission power system stabilizers for all their generating units since early 1960's. It has
also sponsored extensive research in this area.
Through research it was established that the action of automatic voltage regulators caused these
oscillations. An automatic voltage regulator (AVR) regulates the generator terminal voltage and also
helps in the enhancement of transient stability by reducing the peak of the first swing following any
disturbance. However, its high gain contributed to negative damping to the system. The knowledge of
this relation resulted in the commissioning of power system stabilizers. It was observed that these
oscillations were results of the periodic interchange of kinetic energy between the generator rotors. A
power system stabilizer (PSS) provides a negative feedback of the changes in rotor kinetic energy
when it is connected to the excitation system thereby providing damping to these small oscillations.
The PSS has been a subject of extensive research. The team of Dr. P. Kundur, then with Ontario
Hydro, and his co-workers has done extensive research in the area of PSS tuning and its
characteristics. Through their vast experience and extensive research, they reported the
enhancement of inter-area and local modes through PSS reported in. Since a power system is piece-
wise linear, its system characteristics changes with operating point. Therefore an adaptive controller
that can tune with the changes in the system has been developed and reported in. It was shown that
the adaptive PSS is effective in damping large as well as small disturbances.
The power flow between generators, as evident from (9.4), is dependent on the angle between those
generators. The stable operating point of the power system is where the generated power at each
station is matched by the electrical power sent out from that station. When there is a mismatch
between electrical power out and the generated mechanical shaft power, the generator will accelerate
at a rate determined by the power mismatch and the machine inertia as given in (9.14).
Oscillations in s Two Area System
Consider the simple power system shown in Fig. 9.11 in which two machines are operating. Let us
assume that starting with the initial angles
1
and
2
with respect to some reference at nominal
frequency, machine 1 accelerates while machine 2 decelerates from this nominal frequency. We then
have
where the subscripts 1 and 2 refer to machines 1 and 2 respectively. Let us assume that the
transmission line is loss less. Then in the simple case where the power from machine 1 flows to
machine 2, we
get
where
12
=
1
-
2
.
Fig. 9.11 Single-line diagram of a two-machine power system.
Now since the system is lossless, (9.26) will also imply that P
m1
= - P
m2
. This means that in the
steady state, the power generated at machine 1 is absorbed through machine 2. Combining (9.25)
and (9.26) we get
(9.25)
(9.26)
Let us now
assume that H
1
= H
2
= H , V
1
= V
2
= 1.0 per unit and P
m1
= 0. We then get from (9.27)
where the oscillation frequency is given by
Thus the weighted difference of angles will approximate simple harmonic motion for small changes in
12
and the frequency will decrease for an increase in inertia H or impedance X . Another aspect can
be seen by adding the system to give
Thus the overall acceleration of the machine group will depend on the overall balance between power
generated and consumed. Usually there are governors on the generators to reduce generated power
if the system frequency increases.
Introduction
Ideal Shunt Compensator
Improving Volatage Profile
Improving Power-Angle Characteristics
Improving Stability Margin
Improving Damping to Power Oscillations
Ideal Series Compensator
Impact of Series Compensator on Voltage Profile
Improving Power-Angle Characteristics
An Alternate Method of Voltage Injection
Improving Stability Margin
Comparisons of the Two Modes of Operation
Power Flow Control and Power Swing Damping
Introduction
The two major problems that the modern power systems are facing are voltage and angle stabilities.
There are various approaches to overcome the problem of stability arising due to small signal
oscillations in an interconnected power system. As mentioned in the previous chapter, installing
power system stabilizers with generator excitation control system provides damping to these
oscillations. However, with the advancement in the power electronic technology, various reactive
power control equipment are increasingly used in power transmission systems.
A power network is mostly reactive. A synchronous generator usually generates active power that is
specified by the mechanical power input. The reactive power supplied by the generator is dictated by
the network and load requirements. A generator usually does not have any control over it. However
the lack of reactive power can cause voltage collapse in a system. It is therefore important to
supply/absorb excess reactive power to/from the network. Shunt compensation is one possible
approach of providing reactive power support.
A device that is connected in parallel with a transmission line is called a shunt compensator , while a
device that is connected in series with the transmission line is called a series compensator . These
(9.27)
(9.28)
are referred to as compensators since they compensate for the reactive power in the ac system. We
shall assume that the shunt compensator is always connected at the midpoint of transmission
system, while the series compensator can be connected at any point in the line. We shall
demonstrate that such connections in an SMIB power system improves
voltage profile
power-angle characteristics
stability margin
damping to power oscillations
A static var compensator ( SVC ) is the first generation shunt compensator. It has been around since
1960s. In the beginning it was used for load compensation such as to provide var support for large
industrial loads, for flicker mitigation etc. However with the advancement of semiconductor
technology, the SVC started appearing in the transmission systems in 1970s. Today a large number
of SVCs are connected to many transmission systems all over the world. An SVC is constructed
using the thyristor technology and therefore does not have gate turn off capability.
With the advancement in the power electronic technology, the application of a gate turn off thyristor
(GTO) to high power application became commercially feasible. With this the second generation
shunt compensator device was conceptualized and constructed. These devices use synchronous
voltage sources for generating or absorbing reactive power. A synchronous voltage source (SVS) is
constructed using a voltage source converter (VSC). Such a shunt compensating devi ce is called
static compensator or STATCOM . A STATCOM usually contains an SVS that is driven from a dc
storage capacitor and the SVS is connected to the ac system bus through an interface transformer.
The transformer steps the ac system voltage down such that the voltage rating of the SVS switches
are within specified limit. Furthermore, the leakage reactance of the transformer plays a very
significant role in the operation of the STATCOM.
Like the SVC, a thyristor controlled series compensator ( TCSC ) is a thyristor based series
compensator that connects a thyristor controlled reactor ( TCR ) in parallel with a fixed capacitor.
By varying the firing angle of the anti-parallel thyristors that are connected in series with a reactor in
the TCR, the fundamental frequency inductive reactance of the TCR can be changed. This effects a
change in the reactance of the TCSC and it can be controlled to produce either inductive or capacitive
reactance.
Alternatively a static synchronous series compensator or SSSC can be used for series
compensation. An SSSC is an SVS based all GTO based device which contains a VSC. The VSC is
driven by a dc capacitor. The output of the VSC is connected to a three-phase transformer. The other
end of the transformer is connected in series with the transmission line. Unlike the TCSC, which
changes the impedance of the line, an SSSC injects a voltage in the line in quadrature with the line
current. By making the SSSC voltage to lead or lag the line current by 90 , the SSSC can emulate
the behavior of an inductance or capacitance.
In this chapter, we shall discuss the ideal behavior of these compensating devices. For simplicity we
shall consider the ideal models and broadly discuss the advantages of series and shunt
compensation.
Section I: Ideal Shunt Compensator
The ideal shunt compensator is an ideal current source. We call this an ideal shunt compensator
because we assume that it only supplies reactive power and no real power to the system. It is
needless to say that this assumption is not valid for practical systems. However, for an introduction,
the assumption is more than adequate. We shall investigate the behavior of the compensator when
connected in the middle of a transmission line. This is shown in Fig. 10.1, where the shunt
compensator, represented by an ideal current source, is placed in the middle of a lossless
transmission line. We shall demonstrate that such a configuration improves the four points that are
mentioned above.
Fig 10.1 Schematic diagram of an ideal, midpoint shunt compensation
Improving Voltage Profile
Let the sending and receiving voltages be given by and respectively. The ideal shunt
compensator is expected to regulate the midpoint voltage to
against any variation in the compensator current. The voltage current characteristic of the
compensator is shown in Fig. 10.2. This ideal behavior however is not feasible in practical systems
where we get a slight droop in the voltage characteristic. This will be discussed later.
Fig. 10.2 Voltage-current characteristic of an ideal shunt compensator.
Under the assumption that the shunt compensator regulates the midpoint voltage tightly as given by
(10.1), we can write the following expressions for the sending and receiving end currents
Again from Fig. 10.1 we write
Combing (10.2)-(10.4) and solving we get
We thus have to generate a current that is in phase with the midpoint voltage and has a magnitude of
(4V / X
L
){1 - cos( /2)}. The apparent power injected by the shunt compensator to the ac bus is then
Since the real part of the injected power is zero, we conclude that the ideal shunt compensator injects
only reactive power to the ac system and no real power.
Improving Power-Angle Characteristics
The apparent power supplied by the source is given by
(10.5)
(10.6)
Similarly the apparent power delivered at the receiving end is
Hence the real power transmitted over the line is given by
Combining (10.6)-(10.8), we find the reactive power consumed by the line as
The power-angle characteristics of the shunt compensated line are shown in Fig. 10.3. In this figure
P
max
= V
2
/X is chosen as the power base.
Fig. 10.3 Power-angle characteristics of ideal shunt compensated line.
Fig. 10.3 depicts P
e
- and Q
Q
- characteristics. It can be seen from fig 10.4 that for a real power
transfer of 1 per unit, a reactive power injection of roughly 0.5359 per unit will be required from the
shunt compensator if the midpoint voltage is regulated as per (10.1). Similarly for increasing the real
power transmitted to 2 per unit, the shunt compensator has to inject 4 per unit of reactive power. This
will obviously increase the device rating and may not be practical. Therefore power transfer
enhancement using midpoint shunt compensation may not be feasible from the device rating point of
view.
Fig. 10.4 Variations in transmitted real power and reactive power injection by the shunt compensator
with load angle for perfect midpoint voltage regulation.
Let us now relax the condition that the midpoint voltage is regulated to 1.0 per unit. We then obtain
some very interesting plots as shown in Fig. 10.5. In this figure, the x-axis shows the reactive power
available from the shunt device, while the y-axis shows the maximum power that can be transferred
over the line without violating the voltage constraint. There are three different P-Q relationships given
for three midpoint voltage constraints. For a reactive power injection of 0.5 per unit, the power
transfer can be increased from about 0.97 per unit to 1.17 per unit by lowering the midpoint voltage to
0.9 per unit. For a reactive power injection greater than 2.0 per unit, the best power transfer capability
is obtained for V
M
= 1.0 per unit. Thus there will be no benefit in reducing the voltage constraint
when the shunt device is capable of injecting a large amount of reactive power. In practice, the level
to which the midpoint voltage can be regulated depends on the rating of the installed shunt device as
well the power being transferred.
Fig. 10.5 Power transfer versus shunt reactive injection under midpoint voltage constraint.
Improving Stability Margin
This is a consequence of the improvement in the power angle characteri stics and is one of the major
benefits of using midpoint shunt compensation. As mentioned before, the stability margin of the
system pertains to the regions of acceleration and deceleration in the power-angle curve. We shall
use this concept to delineate the advantage of mid point shunt compensation.
Consider the power angle curves shown in Fig. 10.6.
The curve of Fig. 10.6 (a) is for an uncompensated system, while that of Fig. 10.6 (b) for the
compensated system. Both these curves are drawn assuming that the base power is V
2
/X . Let us
assume that the uncompensated system is operating on steady state delivering an electrical power
equal to P
m
with a load angle of
0
when a three-phase fault occurs that forces the real power to zero.
To obtain the critical clearing angle for the uncompensated system is
cr
, we equate the accelerating
area A
1
with the decelerating area A
2
, where
with
max
= -
0
. Equating the areas we obtain the value of
cr
as
Let us now consider that the midpoint shunt compensated system is working with the same
mechanical power input P
m.
The operating angle in this case is
1
and the maximum power that can
be transferred in this case is 2 per unit. Let the fault be cleared at the same clearing angle
cr
as
before. Then equating areas A
3
and A
4
in Fig. 10.6 (b) we get
2
, where
Example 10.1
Let an uncompensated SMIB power system is operating in steady state with a mechanical power
input P
m
equal to 0.5 per unit. Then
0
= 30 = 0.5236 rad and
max
= 150 = 2.6180 rad.
Consequently, the critical clearing angle is calculated as (see Chapter 9)
cr
= 79.56 = 1.3886 rad.
Let us now put an ideal shunt compensator at the midpoint. The pre-fault steady state operating angle
of the compensated system can be obtained by solving 2 sin(/2) = 0.5, which gives
1
= 28.96 =
0.5054 rad. Let us assume that we use the same critical clearing angle as obtained above for clearing
a fault in the compensated system as well.
The accelerating area is then given by A
3
= 0.4416. Equating with area A
4
we get a nonlinear
equation of the form
Fig 10.6 Power-angel curve showing clearing angles: (a) for uncompensated sytem and (b) for
compensated system
Solving the above equation we get
2
= 104.34 = 1.856 rad. It is needless to say that the stability
margin has increased significantly in the compensated system.
Improving Damping to Power Oscillations
The swing equation of a synchronous machine is given by (9.14). For any variation in the electrical
quantities, the mechanical power input remains constant. Assuming that the magnitude of the
midpoint voltage of the system is controllable by the shunt compensating device, the accelerating
power in (9.14) becomes a function of two independent variables, | V
M
| and . Again since the
mechanical power is constant, its perturbation with the independent variables is zero. We then get the
following small perturbation expression of the swing equation
where indicates a perturbation around the nominal values.
If the mid point voltage is regulated at a constant magnitude, | V
M
| will be equal to zero. Hence the
above equation will reduce to
The 2
nd
order differential equation given in (10.13) can be written in the Laplace domain by neglecting
the initial conditions as
The roots of the above equation are located on the imaginary axis of the s-plane at locations j
m
where
This implies that the load angle will oscillate with a constant frequency of
m
. Obviously, this solution
is not acceptable. Thus in order to provide damping, the mid point voltage must be varied according
to in sympathy with the rate of change in . We can then write
where K
M
is a proportional gain. Substituting (10.15) in (10.12) we get
Provided that
K
M
is positive definite, the introduction of the control action (10.15) ensures that the roots of the
second order equation will have negative real parts. Therefore through the feedback, damping to
power swings can be provided by placing the poles of the above equation to provide the necessary
damping ratio and undamped natural frequency of oscillations.
Example 10.2
Consider the SMIB power system shown in Fig. 10.7. The generator is connected to the infinite bus
through a double circuit transmission line. At the midpoint bus of the lines, a shunt compensator is
connected. The shunt compensator is realized by the voltage source V
F
that is connected to the
midpoint bus through a pure inductor X
F
, also known as an interface inductor . The voltage source
V
F
is driven such that it is always in phase with the midpoint voltage V
M
. The current I
Q
is then purely
inductive, its direction being dependent on the relative magnitudes of the two voltages. If the
magnitude of the midpoint voltage is higher than the voltage source V
F
, inductive current will flow
from the ac system to the voltage source. This implies that the source is absorbing var in this
configuration. On the other hand, the source will generate var if its magnitude is higher than that of
the midpoint voltage.
The system is simulated in MATLAB. The three-phase transmission line equations are simulated
using their differential equations, while the generator is represented by a pure voltage source. The
second order swing equation is simulated in which the mechanical power input is chosen such that
the initial operating angle of the generator voltage is (0.6981 rad). The instantaneous electrical power
is computed from the dot product of the three-phase source current vector and source voltage vector.
The system parameters chosen for simulation are:
(10.12)
(10.13)
(10.14
(10.15)
(10.16)
Fig. 10.7 SMIB system used in the numerical example.
Sending end voltage, V
S
= 1 < 40 per unit,
Receiving end voltage, V
R
= 1 < 0 per unit,
System Frequency
s
= 100 rad/s,
Line reactance, X=0.5 per unit,
Interface reactance, X
F
per unit,
Generator inertia constant, H=4.0 MJ/MVA.
Two different tests are performed. In the first one, the midpoint voltage is regulated to 1 per unit using
a proportional-plus-integral (PI) controller. The magnitude of the midpoint voltage is first calculated
using the d-q transformation of the three phase quantities. The magnitude is then compared with the
set reference (1.0) and the error is passed through the PI controller to determine the magnitude of the
source voltage,
ie,
The source voltage is then generated by phase locking it with the midpoint voltage using
Fig. 10.8 depicts the system quantities when the system is perturbed for its nominal operating
condition. The proportional gain ( K
P
) is chosen as 2.0, while the integral gain ( K
I
) is chosen as 10.
In Fig. 10.8 (a) the a-phase of the midpoint voltage, source voltage and the injected current are
shown once the system transients die out. It can be seen that the source and midpoint voltages are
phase aligned, while the injected current is lagging these two voltages by 90 . Furthermore, the
midpoint voltage magnitude is tightly regulated. Fig. 10.8 (b) depicts the perturbation in the load angle
and the injected reactive power. It can be seen that the load angle undergoes sustained oscillation
and this oscillation is in phase with the injected reactive power. This implies that, by tightly regulating
the midpoint voltage though a high gain integral controller, the injected reactive power oscillates in
sympathy with the rotor angle. Therefore to damp out the rotor oscillation, a controller must be
designed such that the injected reactive power is in phase opposition with the load angle. It is to be
noted that the source voltage also modulates in sympathy with the injected reactive power. This
however is not evident from Fig. 10.8 (a) as the time axis has been shortened here.
(10.17)
(a) (b)
Fig. 10.8 Sustained oscillation in rotor angle due to strong regulation of midpoint voltage.
To improve damping, we now introduce a term that is proportional to the deviation of machine speed
in the feedback loop such that the control law is given by
The values of proportional gain K
P
and integral gain K
I
chosen are same as before, while the value of
C
P
chosen is 50. With the system operating on steady state, delivering power at a load angle of 40
for 50 ms, breaker B (see Fig. 10.7) opens inadvertently. The magnitude of the midpoint voltage is
shown in Fig. 10.9 (a). It can be seen that the magnitude settles to the desired value of 1.0 per unit
once the initial transients die down. Fig. 10.9 (b) depicts perturbations in load angle and reactive
power injected from their prefault steady state values. It can be seen that these two quantities have a
phase difference of about 90 and this is essential for damping of power oscillations.
(a) (b)
Fig 10.9 System response with the damping controller
Ideal Series Compensator
Let us assume that the series compensator is represented by an ideal voltage source. This is shown
in Fig. 10.10. Let us further assume that the series compensator is ideal, i.e., it only supplies reactive
power and no real power to the system. It is needless to say that this assumption is not valid for
practical systems. However, for an introduction, the assumption is more than adequate. It is to be
noted that, unlike the shunt compensator, the location of the series compensator is not crucial, and it
can be placed anywhere along the transmission line.
Fig. 10.10 Schematic diagram of an ideal series compensated system.
Impact of Series Compensator on Voltage Profile
In the equivalent schematic diagram of a series compensated power system is shown in Fig. 10.10,
the receiving end current is equal to the sending end current, i.e., I
S
= I
R
. The series voltage V
Q
is
injected in such a way that the magnitude of the injected voltage is made proportional to that of the
line current. Furthermore, the phase of the voltage is forced to be in quadrature with the line current.
We then have
The ratio /X is
called the compensation level and is often expressed in percentage. This compensation level is
usually measured with respect to the transmission line reactance. For example, we shall refer the
compensation level as 50% when = X /2. In the analysis presented below, we assume that the
injected voltage lags the line current. The implication of the voltage leading the current will be
discussed later.
Applying KVL we get
Assuming V
S
= V < and V
R
= V <0 , we get the following expression for the line current
When we choose V
Q
= I
S
e
- j90
, the line current equation becomes
Thus we see that is subtracted from X . This choice of the sign corresponds to the voltage source
acting as a pure capacitor. Hence we call this as the capacitive mode of operation . In contrast, if
we choose V
Q
= I
S
e
+j90
, is added to X , and this mode is referred to as the inductive mode of
operation . Since this voltage injection using (10.20) add to or subtract from the line reactance,
we shall refer it as voltage injection in constant reactance mode. We shall consider the implication
of series voltage injection on the transmission line voltage through the following example.
(10.20)
Example 10.3
Consider a lossless transmission line that has a 0.5 per unit line reactance (X). The sending end and
receiving end voltages are given by 1< and 1< 0 per unit respectively where is chosen as 30 .
Let us choose = 0.5 and operation in the capacitive mode. For this line, this implies a 30% level of
line impedance compensation. The line current is then given from (10.21) as I
S
= 1.479 7 < 15 per
unit and the injected voltage calculated from (10.20) is V
Q
= 0.2218 < - 75 per unit. The phasor
diagrams of the two end voltages, line current and injected voltage are shown in Fig. 10.11 (a). We
shall now consider a few different cases.
Let us assume that the series compensator is placed in the middle of the transmission line. We then
define two voltages, one at either side of the series compensator. These are:
Voltage on the left: V
QL
= V
S
- jXI
S
/ 2 = 0.9723 < 8.45 per unit
Voltage on the right: V
QR
= V
R
+ jXI
S
/ 2 = 0.9723 < 21.55 per unit
The difference of these two voltages is the injected voltage. This is shown in Fig. 10.11 (b), where
the angle = 8.45 . The worst case voltage along the line will then be at the two points on either side
of the series compensator where the voltage phasors are aligned with the line current phasor. These
two points are equidistant from the series compensator. However, their particular locations will be
dependent on the system parameters.
As a second case, let us consider that the series compensator is placed at the end of the
transmission line, just before the infinite bus. We then have the following voltage
Voltage on the left of the compensator: V
QL
= V
R
+ V
Q
= 1.0789 < - 11.46 per unit
This is shown in Fig. 10.11 (c). The maximum voltage rise in the line is then to the immediate left of
the compensator, i.e., at V
QL
. The maximum voltage drop however still occurs at the point where the
voltage phasor is aligned with the line current phasor.
As a third case, let us increase the level of compensation from 30% to 70% (i.e., change from 0.15
to 0.35). We however, do not want to change the level of steady state power transfer. The relation
between power transfer and compensation level will be discussed in the next subsection. It will
however suffice to say that this is accomplished by lowering the value of the angle of the sending
end voltage to 12.37 . Let us further assume that the series compensator is placed in the middle of
the transmission line. We then have V
QL
= 1.0255 < - 8.01 per unit and V
QR
= 1.0255 < 20.38 per
unit. This is shown in Fig. 10.11 (d). It is obvious that the voltage along the line rises to a maximum
level at either side of the series compensator.
Improving Power-Angle Characteristics
Noting that the sending end apparent power is V
S
I
S
* , we can write
Similarly the receiving end apparent power is given by
Hence the real power transmitted over the line is given by
The power-angle characteristics of a series compensated power system are given in Fig. 10.12. In
this figure the base power is chosen as V
2
/ X . Three curves are shown, of which the curve P
0
is the
power-angle curve when the line is not compensated. Curves which have maximum powers greater
than the base power pertain to capacitive mode of operation. On the other hand, all curves the
inductive mode of operation will have maximum values less than 1. For example, in Fig. 10.12, the
curve P
1
is for capacitive mode and the curve P
2
is for inductive mode of operation.
Fig. 10.12 Power-angle characteristics in constant reactance mode.
Let us now have a look at the reactive power. For simplicity let us restrict our attention to capacitive
mode of operation only as this represents the normal mode of operation in which the power transfer
over the line is enhanced. From (10.20) and (10.21) we get the reactive power supplied by the
compensator as
Solving the
above equation
we get
(10.25)
In Fig. 10.13, the reactive power injected by the series compensator is plotted against the maximum
power transfer as the compensation level changes from 10% to 60%. As the compensation level
increases, the maximum power transfer also increases. However, at the same time, the reactive
injection requirement from the series compensator also increases. It is interesting to note that at 50%
compensation level, the reactive power injection requirement from a series compensator is same that
from shunt compensator that is regulating the midpoint voltage to 1.0 per unit.
Fig. 10.13 Reactive power injection by a series compensator versus maximum power transfer as the
level of compensation changes in constant reactance mode.
An Alternate Method of Voltage Injection
So far we have assumed that the series compensator injects a voltage that is in quadrature with the
line current and its magnitude is proportional to the magnitude of the line current. A set of very
interesting equations can be obtained if the last assumption about the magnitude is relaxed. The
injected voltage
is then given by
We can then write the above equation as
i.e., the voltage source in quadrature with the current is represented as a pure reactance that is either
inductive or capacitive. Since in this form we injected a constant voltage in quadrature with the line
current, we shall refer this as constant voltage injection mode. The total equivalent inductance of
the line is then
Defining V
S
= V < and V
R
< 0 , we can then write the power transfer equation as
Since | V
Q
| / | I
S
| = X
Q
, we can modify the above equation as
(10.26)
Consider the phasor diagram of Fig. 10.14 (a), which is for capacitive operation of the series
compensator. From this diagram we get
Similarly from the inductive operation phasor diagram shown in Fig. 10.14 (b), we get
Substituting the above two equations in (10.28) and rearranging we get
where the positive sign is for capacitive operation.
Fig. 10.14 Phasor diagram of series compensated system: (a) capacitive operation and (b) inductive
operation.
The power-angle characteristics of this particular series connection are given in Fig. 10.15. In this
figure the base power is chosen as V
2
/X . Three curves are shown, of which the curve P
0
is the
power-angle curve when the line is not compensated. Curves which have maximum powers greater
than the base power pertain to capacitive mode of operation. On the other hand, all curves the
inductive mode of operation will have maximum values less than 1. For example, in Fig. 10.15, the
curve P
1
is for capacitive mode and the curve P
2
is for inductive mode of operation.
Fig. 10.15 Power-angle characteristics for constant voltage mode.
The reactive power supplied by the compensator in this case will be
(10.30)
Improving Stability Margin
From the power-angle curves of Figs. 10.13 and 10.15 it can be seen that the same amount of power
can be transmitted over a capacitive compensated line at a lower load angle than an uncompensated
system. Furthermore, an increase in the height in the power-angle curve means that a larger amount
of decelerating area is available for a compensated system. Thus improvement in stability margin for
a capacitive series compensated system over an uncompensated system is obvious.
Comparisons of the Two Modes of Operation
As a comparison between the two different modes of voltage injection, let us first consider the
constant reactance mode of voltage injection with a compensation level of 50%. Choosing V
2
/ X as
the base power, the power-angle characteristic reaches a maximum of 2.0 per unit at a load angle /
2. Now | V
Q
| in constant voltage mode is chosen such that the real power is 2.0 per unit at a load
angle of / 2. This is accomplished using (10.29) where we get
per unit
The power-angle characteristics of the two different modes are now drawn in Fig. 10.16 (a). It can be
seen that the two curves match at / 2. However, the maximum power for constant voltage case is
about 2.1 per unit and occurs at an angle of 67 .
Fig. 10.16 (b) depicts the line current for the two cases. It can be seen that the increase in line current
in either case is monotonic. This is not surprising for the case of constant reactance mode since as
the load angle increases, both real power and line currents increase. Now consider the case of
constant voltage control. When the load angle moves backwards from /2 to 67 , the power moves
from 2.0 per unit to its peak value of 2.1 per unit. The line current during this stage decreases from
about 2.83 to 2.50 per unit. Thus, even though the power through the line increases, the line current
decreases.
Power Flow Control and Power Swing Damping
One of the major advantages of series compensation is that through its use real power flow over
transmission corridors can be effectively controlled. Consider, for example, the SMIB system shown
in Fig. 10.17 in which the generator and infinite bus are connected through a double circuit
transmission line, labeled line-1 and line-2. Of the two transmission lines, line-2 is compensated by a
series compensator. The compensator then can be utilized to regulate power flow over the entire
system.
(a) (b)
Fig. 10.16 Power-angle and line current-angle characteristics of the two different methods of voltage
injection: solid line showing constant reactance mode and dashed line showing constant voltage
mode.
For example, let us consider that the system is operating in the steady state delivering a power of
P
m0
at a load angle of
0
. Lines 1 and 2 are then sending power P
e1
and P
e2
respectively, such that
P
m0
= P
e1
+ P
e2
. The mechanical power input suddenly goes up to P
m1
. There are two ways of
controlling the power in this situation:
Regulating Control: Channeling the increase in power through line-1. In this case the series
compensator maintains the power flow over line-2 at P
e2
. The load angle in this case goes up
in sympathy with the increase in P
e1
.
Tracking Control: Channeling the increase in power through line-2. In this case the series
compensator helps in maintaining the power flow over line-1 at P
e1
while holding the load angle
to
0
.
Let us illustrate these two aspects with the help of a numerical example.
Example 10.4
Let us consider the system of Fig. 7.8 where the system parameters are given by
System Frequency = 50 Hz, | V S | = | V R | = 1.0 per unit, X = 0.5 per unit and d 0 = 30 /
It is assumed that the series compensator operates in constant reactance mode with a compensation
level of 30%. We then have
P
e1
= 1.0 per unit, P
e2
= 1.43 per unit, P
m
= 2.43 per unit
The objective of the control scheme here is to maintain the power through line-2 to a pre-specified
value, P
ref
. To accomplish this a proportional-plus-integral (PI) controller is placed in the feedback
loop of P
e2
. In addition, to improve damping a term that is proportional to the deviation of machine
speed is introduced in the feedback loop. The control law is then given by
where C
L
= / X is the compensation level. For the simulation studies performed, the following
controller parameters are chosen
K
P
= 0.1, K
I
= 1.0 and C
P
= 75
Regulating Control: With the system operating in the nominal steady state, the mechanical power
input is suddenly raised by 10%. It is expected that the series compensator will hold the power
through line-2 constant at line-2 at P
e2
such that entire power increase is channeled through line-1.
We then expect that the power P
e1
will increase to 1.243 per unit and the load angle to go up to 0.67
rad. The compensation level will then change to 13%.
The time responses for various quantities for this test are given in Fig. 10.18. In Fig. 10.18 (a), the
power through the two line is plotted. It can be seen that while the power through line-2 comes back
to its nominal value following the transient, the power through the other line is raised to expected
level. Similarly, the load angle and the compensation level reach their expected values, as shown in
Figs. 10.18 (b) and (c), respectively. Finally, Fig. 10.18 (d) depicts the last two cycles of phase-a of
the line current and injected voltage. It can be clearly seen that these two quantities are in
quadrature, with the line current leading the injected voltage.
(10.31)
Fig. 10.18 System response with regulating power flow controller.
Tracking Control: With the system operating in the nominal steady state, the mechanical power
input is suddenly raised by 25%. It is expected that the series compensator will make the entire power
increase to flow through line-2 such that both P
e1
and load angle are maintained constant at their
nominal values. The power P
e2
through line-2 will then increase to about 2.04 per unit and the
compensation level will change to 51%.
The time responses for various quantities for this test are given in Fig. 10.19. It can be seen that while
the power through line-1 comes back to its nominal value following the transient, the power through
the other line is raised to level expected. Similarly, the load angle comes back to its nominal value
and the compensation level is raised 51%, as shown in Figs. 10.19 (b) and (c), respectively. Finally,
Fig. 7.19 (d) depicts the last two cycles of phase-a of the line current and injected voltage.
Fig. 10.19 System response with regulating power flow controller.