14K.L.N.

COLLEGE OF INFORMATION TECHNOLOGY

Pottapalayam 630 611 !"#a$a%$a" &"'t(")t.
Y*a( + ,00- .,00/ !*m*'t*( + 6

0(a%)1 + III 0.E. EEE !234*)t + EE 135, Po6*( !y't*m A%aly'"'
Nam* o7 t1* 8("t*( 9Colla3o(ato(: + R. Ram*'1 P(o7*''o( 9EEE: IT &*pt.
;%"t + II &at* o7 !23m"''"o% + ,,.11.,00-
!ylla32' + II%< ;%"t
I%t(o<2)t"o% 02' Cla''"7")at"o% 02' a<m"tta%)* mat("= !ol2t"o% o7 %o%.
l"%*a( al$*3(a") *>2at"o%' Ga2''.!*"<al m*t1o< N*6to% Rap1'o% m*t1o<
Fa't <*)o2pl*< m*t1o< Flo6 )1a(t' a%< )ompa("'o% o7 t1* t1(** m*t1o<'.
......................................................................................
EE 1352 - POWER SYSTEM ANALYSIS
VI semester EEE Branch (2!-21 Batch"
#n$t % II % P&'er ()&' Ana)*s$s
1
S*))a+,s -
Introduction Bus Classification Bus admittance matrix Solution of non-linear
algebraic equations Gauss-Seidal method Newton Raphson method ast
decoupled method low charts and comparison of the three methods!
1. Intr&/,ct$&n -
The load flow studies involve the solution of the power system network under
steady-state conditions. The solution is obtained by considering certain inequality
constraints imposed on bus voltages and reactive power of generators.
The main information obtained from a load-flow study are the magnitude and
phase angle of the voltage at each bus and the real and reactive power flowing in each
line. The load flow solution also gives the initial conditions of the system when the
transient behavior of the system is to be studied.
The load flow study of a power system is essential to decide the best operation of
existing system and for planning the future expansion of the power system. The load
flow solution of power system under balanced load condition is considered for our
studies. The steps involved in load flow studies are :
1. Representation of the system by single-line diagram.
. !etermining the impedance diagram using the information in single line diagram.
". #ormulation of network equations.
$. %olution of network equations.
1.1 T*0es &1 B,ses
&n a power system the buses are meeting point of various components. The
generators will feed energy to buses and loads will draw energy from buses. &n the
network of a power system the buses become nodes and so a voltage can be specified for
each bus. The bus or bus bar is a conductor made of copper or aluminium having
negligible resistance. 'ence the bus bar will have (ero voltage drop when it conducts the
rated current. Therefore the buses are considered as points of constant voltage in a power
system.
Therefore each bus in a power system is associated with four quantities :
1. Real power) *
. Reactive power) +
". ,agnitude of voltage) -.-
$. *hase angle of voltage) /
&n a load flow problem) two quantities 0out of four1 are specified for each bus and the
remaining two quantities are obtained by solving the load flow equations.
The classification of buses are :
i1 2oad bus or *+ bus
ii1 3enerator bus or voltage controlled bus
iii1 %lack bus or swing bus or reference bus.
Types of buses can be summari(ed as a table as follows :
4us type +uantities specified +uantities to be obtained
2oad bus *) + -.-) /
3enerator bus *) -.- +) /

%lack bus -.-) / *) +

* 5 *
3
6 *
!
: -.-5 ,agnitude of bus voltage
+ 5 +
3
6 +
!
7 / 5 *hase angle of bus voltage
*
3
) +
3
6 Real and Reactive powers generated by generators connected to the bus
respectively.
*
!
) +
!
6 Real and Reactive powers drawn by the loads connected to the bus
respectively.
L&a/ B,s -
The bus called the load bus) when real and reactive components of power are
specified for the bus. The load flow equations can be solved to find the magnitude and
phase angle of the voltage. &n a load bus the voltage is allowed to vary within
permissible limits) for example 8 9 :.
2enerat&r B,s -
The bus is called generator bus) when real power and magnitude of bus voltage
are specify for the bus. The load flow equation can be solved to find the reactive power
and phase angle of bus voltage. ;sually for generator buses reactive power limits will be
specified.
S)ac3 B,s -
The bus is called slack bus) if the magnitude and phase angle of bus voltage are
specified for the bus. The slack bus is the reference bus for load flow solution and
usually one of the generator buses is selected as the slack bus.
<eed for slack bus :
4asically the power system has only two types of buses and they are load and
generator buses. &n these buses only power in=ected by generators and power drawn by
loads are specified) but the power loss in transmission lines are not accounted.
T&ta) (c&m0)e4" 0&'er )&ss $n transm$ss$&n )$ne 5 S,m &1 c&m0)e4 0&'er &1
6enerat&rs % S,m &1 c&m0)e4 0&'er &1 )&a/s.
The transmission line losses can be estimated only if the real and reactive power
of all buses are known. The powers in the buses will be known only after the solving the
load flow equations. #or these reasons) the real and reactive power of one of the
generator bus is not specified and this bus is called s)ac3 +,s. &t is assumed that the
slack bus generates the real and reactive power required for transmission line losses.
'ence for a slack bus) the magnitude and phase of bus voltage are specified and real and
reactive powers are obtained through the load flow solution.
-----------------------------------------------------------------------------------------------------------
-
172 N&/e E8,at$&ns an/ B,s a/m$ttance matr$4
>hen the power system is represented by impedance ? reactance diagram) it can
be considered as a circuit or network. The buses can be treated as nodes and the voltages
of all buses 0nodes1 can be solved by conventional node analysis technique.
2et n be the number of ma=or or principal nodes in a circuit or network. 0The
principal nodes are meeting points of more than two elements1. %ince the voltage of a
node can be measured only with respect to a reference point) one of the nodes is
considered as reference node. <ow the network will have 0<-11 independent voltages. &n
"
nodal analysis) the independent voltages are solved by writing @A2 equation for 0<-11
nodes in the circuit. #or writing @A2 equations) the voltage sources in the circuit should
be converted to equivalent current sources.
2et .
1
) .

) .
"
)BB...
n
5 <ode voltages of nodes 1))")BBn respectively.
&
11
) &

) &
""
)BBB&
nn
5 %um of current sources connected 0or in=ecting current1 to
nodes 1))")BBn respectively.
C== 5 %um of admittances connected to node-=
C=k 5 <egative of sum of admittances connected between node-= and node-k.
<ote : &f the direction of current in current source is towards the node then the
source is considered as positive and if it is away from the node then the source is
negative.
<ow the n-number of nodal equations for <-bus system will be in the form shown
below 0here n5<1.
11 1 " 1" 1 1 11
......... I V Y V Y V Y V Y
n n
+ + + +
" " 1 1
......... I V Y V Y V Y V Y
n n
+ + + +
"" " " "" " 1 "1
......... I V Y V Y V Y V Y
n n
+ + + +
. . . .
. . . .
nn n nn n n n
I V Y V Y V Y V Y + + + + .........
" " 1 1
The above n-number of equations can be arranged in the matrix form as shown in
equation 011.
1 1 ....0
.
.
.
.
. .
. . . . . .
. . . . . .
. .
. .
. .
""

11
"

1
" 1
" "" " "1
" 1
1 1" 1 11
1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
1
1
]
1

nn n nn n n n
n
n
n
I
I
I
I
V
V
V
V
Y Y Y Y
Y Y Y Y
Y Y Y Y
Y Y Y Y
&n matrix notation the equation 011 can be written as
YV 5 I . . . . 01
&n power system Y matrix is designated as C
bus
) and called the admittance matrix.
The node voltages are called bus voltages.
Therefore the equation 01 can be written as shown in equation 0"1
Y
+,s
V 5 I . . . . . 0"1
>here) Y
+,s
5 4us admittances matrix of order 0n x n1
V 5 4us voltage matrix of order 0n x 11
I 5 Aurrent sources matrix of order 0n x 11
n 5 <umber of independent buses in the system.
$
4us admittance matrix) Y
+,s
5
1 $ .......0
. .
. . . . . .
. . . . . .
. .
. .
. .
" 1
" "" " "1
" 1
1 1" 1 11
1
1
1
1
1
1
1
1
]
1

nn n n n
n
n
n
Y Y Y Y
Y Y Y Y
Y Y Y Y
Y Y Y Y
The bus admittance matrix) C
bus
is symmetrical around the principal diagonal. The
admittances C
11
) C

) C
""
)BB..C
nn
are called self admittances at the buses and the other
admittances are called mutual admittances.
&n general)
C
==
5 %um of all admittances connected to bus-=
C
=k
5 <egative of sum of all admittances between bus-= and bus-k.
Dlso C
=k
5 C
k=
<ote : &f the power system is represented by reactance diagram) then all the elements of
the network are inductive susceptances 0which are negative1. &n this case C
==
will be
negative and :C
=k
will be positive.
1. 3 S&),t$&n &1 +,s 9&)ta6es
Aonsider the node basis matrix equation of <-bus system
Y
+,s
V 5 I . . . . . 091
En premultiplying the equation 091 by
1
bus
Y ) we get
I Y V
bus
1
. . . . 0F1
>e know that)
bus
bus
bus
antofY Deter
ofY Adjo
Y
min
int
1

2et G 5 !eterminant of C
bus
and G
==
5 Aofactor of C
=k

<ow Dd=oint of C
bus
5
1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
1
1
]
1

nn n n n
n
n
n
T
nn n n n
n
n
n
. .
. . . . . .
. . . . . .
. .
. .
. .
. .
. . . . . .
. . . . . .
. .
. .
. .
" 1
" "" " 1"
" 1
1 "1 1 11
" 1
" "" " "1
" 1
1 1" 1 11
. . . . 0H1
9
Dlso) . 5
1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
1
1
]
1

nn n
I
I
I
I
andI
V
V
V
V
.
.
.
.
""

11
"

1
. . . . 0I1
;sing equations 0H1 and 0I1) the equation 0F1 can be written as shown in equation 0J1
1
1
1
1
1
1
1
1
]
1

n
V
V
V
V
.
.
"

1

1
1
1
1
1
1
1
1
]
1

nn n n n
n
n
n
. .
. . . . . .
. . . . . .
. .
. .
. .
1
" 1
" "" " 1"
" 1
1 "1 1 11

1
1
1
1
1
1
1
1
]
1

nn
I
I
I
I
.
.
""

11
. . . 0J1
4y matrix multiplication the equation 0J1 can be expressed as n-number of linear
independent equations shown below.
[ ]
nn n
I I I I V
1 "" "1 1 11 11 1
.........
1
+ + + +

[ ]
nn n
I I I I V
"" " 11 1
.........
1
+ + + +

[ ]
nn n
I I I I V
" "" "" " 11 1" "
.........
1
+ + + +

. . . . .
. . . . .
. . . . .
[ ]
nn nn n n n n
I I I I V + + + +

.........
1
"" " 11 1
&n general the k
th
bus voltage is given by
[ ]
nn nk k k k k
I I I I V + + + +

.........
1
"" " 11 1

n
j
jj jk k
I V
1
1
. . . . 01K1
>here) G 5 !eterminant of C
bus
matrix.
&
==
5 %um of current sources in=ecting current to node-=
:
=k
5 Aofactor o the element C
=k
of bus admittance matrix.
<ote : The equation 01K1 is AramerLs rule and this equation can be expressed in another
simpler form as shown below.
2et G 5 !eterminant of C
bus
matrix
G
k
5 !eterminant of C
bus
matrix after replacing k
th
column by current source
vector &.
F
<ow) k
th
bus voltage

k
k
V
1.; B,s &r N&/e e)$m$nat$&n +* matr$4 a)6e+ra (<r&n re/,ct$&n techn$8,e"
The buses or nodes which does not have any current sources can be eliminated by
matrix manipulation of the standard node equation.
Aonsider the general form of node basis matrix equation.
Y
+,s
V 5 I . . . . . 0111
<ow the matices in the equation can be partitioned using the guidelines given
below.
1. The column matrices V and I are rearranged such that the element associated with
buses to be eliminated are in the lower rows of the matrices.
. The square matrix C
bus
is rearranged such that the elements associated with buses to be
eliminated are in the last rows and columns of the matrices.
". The bus admittance matrix is partitioned so that elements identified only with nodes to
be eliminated are separated from the other elements by hori(ontal and vertical lines.
Aonsider a bus admittance matrix of order 09 x 91. <ow the matrix equation 0111
for n 5 9 can be written as shown in equation 01. 2et us assume that the buses $ and 9
does not have any current source and so they can be eliminated.
1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
1
1
]
1

99
$$
""

11
9
$
"

1
99 9$ 9" 9 91
$9 $$ $" $ $1
"9 "$ "" " "1
9 $ " 1
19 1$ M 1" 1 11
N N
M
M
N N M N N N
M
M
M
I
I
I
I
I
V
V
V
V
V
Y Y Y Y Y
Y Y Y Y Y
Y Y Y Y Y
Y Y Y Y Y
Y Y Y Y Y
. . . . . 011
2et us partition the matrix equation 011 and define the following submatrices.
@ 5
1
1
1
]
1

"" " "1

" 1
1" 1 11
Y Y Y
Y Y Y
Y Y Y
25
1
1
1
]
1

"9 "$
9 $
19 1$
Y Y
Y Y
Y Y

1
]
1

9" 9 91
$" $ $1
Y Y Y
Y Y Y
L
T
1
]
1

99 9$
$9 $$
Y Y
Y Y
M

1
1
1
]
1

"

1
V
V
V
V
A

1
1
1
]
1

""

11
I
I
I
I
A

1
]
1

99
$$
I
I
I
X
>here &
x
5 %ubmatrix composed of the currents entering the buses to be eliminated.
.
x
5 %ubmatrix composed of the voltages of the buses to be eliminated.
@ 5 %ubmatrix composed of self and mutual admittances identified only with
buses to be retained.
, 5 %ubmatrix composed of self and mutual admittances identified only with
buses to be eliminated.
2 5 %ubmatrix composed of only mutual admittances between buses to be
retained and eliminated.
2
T
5 Transpose of 2
H
;sing the submatrices as defined above) the matrix equation 011 can be written
as shown in equation 01"1.
1
]
1

1
]
1

1
]
1

X
A
B
A
T
I
I
V
V
M L
L K
. . . . 01"1
4y matrix multiplication the equation 01"1 can be written as)
<V
A
= LV
>
5 I
A
. . . . 01$1
L
T
V
A
= MV
>
5 I
>
. . . .0191
#rom equation 0191 we get)
MV
>
5 I
>
- L
T
V
A
. . . . .01F1
'ere all the elements of the submatrix &O are (ero) because the buses to be
eliminated does not have any current source.
MV4 5 -L
T
V
A
. . . . . 01H1
En premultiplying equation 01H1 by M
-1
we get)
V4 5 -M
-1
L
T
V
A
. . . . . . 01I1
En substituting for V
>
from equation 01I1 in equation 01$1 we get)
<V
A
% LM
-1
L
T
V
A
5 I
A
[ ]
A A
T
I V L LM K
1
C
bus)new
.
D
5 &
D
. . . . . .01J1
>here Y
+,s7ne'
5
T
L LM K
1

. . . . 0K1
The new bus admittance matrix C
bus)new
used to reconstruct the circuit with the
unwanted buses eliminated.
The matrix partitioning method discussed above is a general procedure and it is
suitable for computer solutions. >hen large number of buses are to be eliminated) then
the si(e of matrix , will be large and finding ,
-1
will be tedious.
The process of bus elimination can be simplified if one bus is eliminated at a time.
&n this case the matrix , will have only one element and ,
-1
is the reciprocal of this
element. 'ere) the bus to be eliminated must be the highest numbered bus and
renumbering may be required. This can be achieved by row interchange and column
interchange. >hen bus-p has to be eliminated in a system with n-independent buses) the
p
th
row is interchanged with n
th
row and p
th
column is interchanged with n
th
column in the
bus admittance matrix.
Aonsider a bus admittance matrix of order 0n x n1 in which the n
th
bus has to be
eliminated.
1
1
1
1
1
1
1
1
1
]
1

.
M . .
M
M . .
. M . . . . .
. M . . . . .
M . .
M . .
1 1 0 1
1 1 0 1 1 10 1 0 1 1 0 1 1 1 0
1 1 0 1
1 1 1 0 1 1 11
nn n n n n
n n n n n n
n n
n n
bus
Y Y Y Y
Y Y Y Y
Y Y Y Y
Y Y Y Y
Y
. . . . . 011
I
2et us partition the bus admittance matrix as shown in equation 011. The
submatrices are given below.
1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
]
1

n n
n
n
n n n n
n
n
n
Y
Y
Y
L
Y Y Y
Y
Y Y Y
Y Y Y
K
1 1 0

1
1 1 10 1 0 1
1 1 0 "
1 1 0 1
1 1 0 1 1 11
.
.
.
7
. .
. . . . .
. . . .
. .
. .
[ ]
nn
nn n n n n
T
Y
M Y M Y Y Y L
1
7 7 . . .
1
1 1 0 1

#rom equation 0K1 we get)

Y
+,s7ne'
5 < % L M
-1
L
T
. . . . . 01
En substituting the submatrices in equation 01 we get)
Y
+,s7ne'
5
[ ]
1 1 0 1
1 1 0

1
1 1 10 1 0 1
1 1 0 "
1 1 0 1
1 1 0 1 1 11
.
1
.
.
. .
. . . . .
. . . .
. .
. .

1
]
1

1
1
1
1
1
1
]
1

1
1
1
1
1
1
]
1

n n n n
nn
n n
n
n
n n n n
n
n
n
Y Y Y
Y
Y
Y
Y
Y Y Y
Y
Y Y Y
Y Y Y
Y
+,s7ne'
5
1
1
1
1
1
]
1

1
1
1
1
1
1
]
1

1 1 0 1 1 0 1 1 0 1 1 1 0
1 1 0 1
1 1 0 1 1 1 1
1 1 10 1 0 1
1 1 0 "
1 1 0 1
1 1 0 1 1 11
.
. . . .
.
.
1
. .
. . . . .
. . . .
. .
. .
n n n n n n n n n n
n n n n n n n
n n n n n n n
nn
n n n n
n
n
n
Y Y Y Y Y Y
Y Y Y Y Y Y
Y Y Y Y Y Y
Y
Y Y Y
Y
Y Y Y
Y Y Y

. . . . 0"1
#rom equation 0"1 the element C
=k
0i.e. the element in row-k and column-=1 of the
resulting 0n-11 x 0n-11 new bus admittance matrix is given by
nn
nk jn
jk new jk
Y
Y Y
Y Y
)
. . . . 0$1
#or = 5 1))")B.0n-11 and k 5 1))")BB0n-11
>here C
=k
) C
=n
) C
nk
and C
nn
are elements of original or given bus admittance matrix.
The equation 0$1 can be used to compute the elements of new bus admittance matrix
directly from the elements of original bus admittance matrix.
The following procedure can be used to compute an element of bus admittance matrix
using equation 0$1.
Ste0 1 - Aonsider an element C
=k
of original bus admittance matrix.
Ste0 2 - 3et the product of last element of row-= and last element of column-k.
J
Ste0 3 - !ivide the product obtained in step- by C
nn
of original bus admittance matrix.
Ste0 ; - En subtracting the resultant value obtained in step-" from C
=k
of original bus
admittance matrix we get) the new value of C
=k
.
<ote : %ince bus admittance matrix is symmetrical we can take C
=k)new
5 C
k=)new
and can
avoid m0m-11? calculations for a new bus admittance matrix of order 0m x m1 where m
5 0n-11.
---------------------------------------------------------------------------------------------------------
1.! B#S IMPE?AN@E MATRI>
The impedance matrix is given by A
+,s
5 Y
+,s
-1
%ince the bus admittance matrix is symmetrical) the bus impedance matrix is also
symmetrical around the principal diagonal. &n bus impedance matrix) the elements on the
main diagonal are called driving point impedance of the buses or nodes and the off-
diagonal elements are called the transfer impedances of the buses or nodes. The bus
impedance matrix is very useful in fault analysis or calculations.
The bus impedance matrix can be determined by
11 Taking the inverse of the bus admittance matrix formed.
1 4us building algorithm.
B,s +,$)/$n6 a)6&r$thm
2et us denote the original P
bus
of a system with n-number of independent buses as
P
orig
. >hen a branch of impedance P
b
is added to the system the P
orig
. 3ets modified. The
branch impedance P
b
can be added to the original system in the following four different
ways.
@ase 1 - Ddding a branch of impedance P
b
from a new-p to the reference bus.
@ase 2- Ddding a branch of impedance P
b
from a new-p to an existing bus-q.
@ase 3 - Ddding a branch of impedance P
b
from an existing bus-1 to the reference bus.
@ase ; - Ddding a branch of impedance P
b
between two existing buses h and q.
Aase 1 : Ddding P
b
from a new bus-p to the reference bus.
Aonsider a n-bus system as shown in fig.a. 2et us add a bus-p through an
impedance P
b
to the reference bus. The addition of a bus will increase the order of the
bus impedance matrix by one.
&n this case the elements of 0nQ11
th
column and row all (eros except the diagonal.
The diagonal element is the added branch impedance P
b
. The elements of original P
bus
matrix are not altered. The new bus impedance matrix will be as shown in equation 091.
1
1
1
1
1
1
]
1

b
orig
new bus
Z
Z
Z
M K K K
M
K M
K M
K M
)
. . . . 091
Aase : Ddding P
b
from a new bus6p to existing
bus-q
1K
Aonsider a n-bus system as shown in fig.b in which a new bus-p is added through
an impedance P
b
to an existing bus-q. The addition of a bus will increase the order of the
bus impedance matrix by one.
&n this the element of 0nQ11
th
column are the elements of q
th
column and elements
of 0nQ11
th
row are the elements of q
th
row. The diagonal element is given by sum of P
qq
and P
b
. The elements of original P
bus
matrix are not altered. The new bus impedance
matrix will be as shown in equation 0F1.
1
1
1
1
1
1
1
1
]
1

b qq qq q q
q
q
new bus
Z Z Z Z Z
Zbus
Z
Z
Z
M .
M
. M
. M
M
M
1

1
)

. . . . 0F1

Aase " : Ddding P
b
from an existing bus-q to the reference bus
Aonsider a n-bus system as shown in fig.c in which an impedance P
b
is added
from an existing bus-q to the reference bus. 2et us consider as if he impedance P
b
is
connected from a new bus-p and existing bus-q. <ow it will be an addition as that of
case-. Then we can short-circuit the bus-q to
reference bus. This is equivalent to eliminating
0nQ11
th
bus 0i.e. bus-p in this case1 and so the
bus impedance matrix has to be modified by
eliminating 0nQ11
th
row and 0nQ11
th
column. The
reduced bus impedance matrix can be formed by
a procedure similar to that of bus elimination in
bus admittance matrix. This reduced bus
impedance matrix is the actual new bus
11
impedance matrix. Rvery element of actual new bus impedance matrix can be
determined using the equation 0H1.
1 1 10 1 0
1 1 0 1 1 0
)
+ +
+ +

n n
k n n j
jk act jk
Z
Z Z
Z Z
. . . . 0H1
<ote : 1. P
=k)act
is the impedance corresponding to row-= and column-k of actual new bus
impedance matrix.
. P
=k
) P
0nQ11k
) P
=0nQ11
) P
0nQ110nQ11
are impedance of new bus impedance matrix of order 0nQ11.
". %ince bus impedance matrix is symmetrical P
=k) act
5 P
k=) act
Aase $ : Ddding P
b
between two existing buses h and q
Aonsider a n-bus system shown in fig. d) in which
an impedance P
b
is added between two existing buses h and
q.
&n this case the bus impedance matrix is formed as
shown in equation 0I1. 'ere the elements of 0nQ11
th
column is the difference between the elements of column-h
and column-q. The elements of 0nQ11
th
row is the
difference between the elements of row-h and row-q. The
diagonal element is given by equation 0I1.
1
1
1
1
1
1
1
1
]
1

b qq qn hn q he q h
nq nh
q h
q h
new bus
Z Z Z Z Z Z Z
Z Z
Zbus
Z Z
Z Z
Z
M .
M
M
. M
M
M
1 1

1 1
)
. . . . 0I1
hq qq hh hh b n n
Z Z Z Z Z Z
1 1 10 1 0
+ + +
+ +
. . . . . 0J1
%ince the modification does not involve addition of new bus) the order of new bus
impedance matrix has to be reduced to n x n by eliminating the 0nQ11
th
column and
0nQ11
th
row.
This reduced bus impedance matrix is the actual new bus impedance matrix.
Rvery element of the actual new bus impedance matrix can be determined using equation
0H1 which is also given below for reference.
1 1 10 1 0
1 1 0 1 1 0
)
+ +
+ +

n n
k n n j
jk act jk
Z
Z Z
Z Z
!irect determination of a bus impedance matrix
The bus impedance matrix can be directly obtained from
impedance or reactance diagram instead of forming C
bus
and then
1
inverting it. &n direct determination of P
bus
first consider an impedance P
b
connected
between bus-1 to reference bus impedance matrix will have a single element a shown
below.
Then add each element of impedance or reactance diagram one by one and
modify the P
bus
in each step. Rach modification of P
bus
involve any one of the four cases
discussed above.
-----------------------------------------------------------------------------------------------------------
-
1.B (&rm,)at$&n &1 L&a/ ()&' e8,at$&ns ,s$n6 Y
+,s
matr$4
The load flow equations can be formed using either the mesh or node base
equations of a power system. 'owever) from the view point of computer time and
memory) the nodal admittance formulation using the nodal voltages as the independent
variables is the most economic.
The node basis matrix equation of a n-bus system is given by
Y
+,s
V 5 I . . . . C
>here Y
+,s
5 4us admittance matrix of order 0n x n1
V 5 4us 0node1 voltage matrix of order 0n x 11
I 5 %ource current matrix of order 0n x 11
The equation C can be written in the form shown below :
1
1
1
1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
1
1
1
1
1
]
1

nn n n n
n
n
n
n

V
V
V
V
Y Y Y Y
Y Y Y Y
Y Y Y Y
Y Y Y Y
I
I
I
I
.
.
.
.
. . . .
. . . . . . . .
. . . . . . . .
. . . .
. . . . . . . .
. . . . . . . .
. . . .
. . . .
.
.
.
.

1
1
1
1
1 1 1 11

1
. . . C
2et &
p
5 Aurrent in=ected to bus p
.
p
5 .oltage of bus p
#rom equation C the current &p can be expressed as shown in equation C
&
p
5 C
p1
.
1
QC
p
.

QBB.QC
pp
.
p
QB.QC
pn
.
n

+
+ +
1
1 1

q
n
q
q q q q
V Y V Y V Y I
B.. C
2et %
p
5 Aomplex power of bus-p
*
p
5 Real power of bus-p
+
p
5 Reactive power of bus-p
<ow %
p
5 *
p
Q = +
p
Dlso %
p
5 .
p
&
p
S
1"
Therefore) .
p
&
p
S
5 *
p
Q = +
p
. . . . . . C
The load flow problem can be handled more conveniently by use of &
p
rather than &
p
S
Therefore take con=ugate of equation C
0.
p
&
p
S
1
S
5 0*
p
Q = +
p
1
S
.
p
S
&
p
5 *
p
- = +
p
S

V
j! "
I


. . .
En equating the equationC and we get)
S
1
1
1

n
q
q q

q
q q
V
j! "
V Y V Y V Y

+ +

+

. . . . C
Meth&/-1 - 2a,ss &r 2a,ss-Se$/a) meth&/
The equation C can be written as shown below

1
1 1
S

q
n
q
q q q q

V Y V Y
V
j! "
V Y
1
1
]
1

+
1
1 1
S
1

q
n
q
q q q q

V Y V Y
V
j! "
Y
V
. . . . .C
1
1
1
]
1

n
q
q
q q

V Y
V
j! "
Y
V
1
S
1
. . . .
The equation C is called the load flow equation for 3auss or 3auss-%eidal
method.
Meth&/ % 2 - Ne't&n - Ra0hs&n meth&/
The equation C can be written as shown below

n
q
q q
V Y V j! "
1
S
. . . .
2et) .
p
5 e
p
Q = f
p
. . . 01K1
.
q
5 e
q
Q = f
q
. . . 0111
C
pq
5 3
pq
Q = 4
pq
. . . 011
>here) e
p
) f
p
5 Real and imaginary part of .
p
respectively
e
q
)

f
q
5 Real and imaginary part of .
q
respectively
3
pq
) 4
pq
5 Aonductance T susceptance of the admittance C
pq
respectively.
En substituting for .
p
) .
q
and C
pq
from equation 01K1) 0111 and 011 we get)

+ +
n
q
q q q q
jf e jB # jf e j! "
1
S
1 10 0 1 0
1$

+
n
q
q q q q q q q q
B f j # jf B je # e jf e
1

1 0 1 0
[ ]

+ +
n
q
q q q q q q q q
B e # f j B f # e jf e
1
1
0 1 0 1 0
[ ]

+ +
n
q
q q q q q q q q
B e # f j B f # e jf e
1
1
0 1 0 1 0
[ ]

+ + +
n
q
q q q q q q q q q q q q q q q q
B e # f f j B f # e jf B e # f je B f # e e
1

1 0 1 0 1 0 1 0
[ ] [ ]


+ + +
n
q
n
q
q q q q q q q q q q q q q q q q
B e # f e B f # e f j B e # f f B f # e e
1 1
1
0 1 0 1 0 1 0
. . .01"1

En separating the real and imaginary parts of equation 01"1 we get)
[ ]

+ +
n
q
q q q q q q q q
B e # f f B f # e e "
1
1 0 1 0
. . . . . 01$1
[ ]

+
n
q
q q q q q q q q
B e # f e B f # e f !
1
1 0 1 0
. . . . . 0191
#rom equation we get)

f e V +
. . . . . . . 01F1
The equations 01$1 and 0191 are called load flow equations of <ewton - Raphson
method.
-----------------------------------------------------------------------------------------
1.3 LOA? (LOW SOL#TION BY 2A#SS % SEI?AL METCO?
The 3auss 6 %eidal method is an iterative algorithm for solving a set of non-linear
load flow equations. The non-linear load flow equations are given by equation C) when
p 5 1)). . . . n and this equation is
1
1
]
1

+
1
1 1
S
1

q
n
q
q q q q

V Y V Y
V
j! "
Y
V
. . . . . 01H1
where p 5 1))")BB..n
The variables in these equations are the node voltages .1) .) .")BB.n. &n
3auss-%eidal method an initial value of voltages are assumed and they are denoted as .
1
K
)
.

K
) .
"
K
)BB.
n
K
. En substituting these initial values in equation 01H1 and by taking p51)
the revised value of bus-1 voltage .
1
1
is computed. The revised value of bus voltage is
.
1
1
is replaced for initial value .
1
K
and the revised bus- voltage .

1
is computed. <ow
replace the .
1
1
for .
1
K
and .

1
for .

K
and perform the calculation for bus-" and so on.
The process of computing all the bus voltages as explained above is called one
iteration. The iterative process is then repeated till the bus voltage converges within
prescribed accuracy. The convergence of bus voltage is quite sensitive to the initial
19
values assumed. 4ased on practical experience it is easier to get a set of initial voltages
very close to final solution.
( )
1
1
]
1

+
+ +
1
1 1
1
S
1
1

q
n
q
k
q
k
q q
k

V Y V Y
V
j! "
Y
V
. . . . 01I1
>here) .
i
k
5 k
th
iteration value of bus voltage .
i
.
i
kQ1
5 0kQ11
th
iteration value of bus voltage .
i
&n equation 01I1) to compute the 0kQ11
th
iteration values of bus-p voltage ) the
0kQ11
th
iteration values of voltages are used for all buses less than p and k
th
iteration
values are used for all buses greater than or equal to p. The equation 01I1 is applicable
for load bus) since in load bus changes in both magnitude and phase of voltages are
allowed. 4ut in generator bus the magnitude of voltage remains constant and so the
equation 01I1 is used to calculate the phase of the voltage.
&t is important to note that the slack bus is a reference bus and so its voltage will
not change. Therefore in each iteration the slack bus voltage is not modified.
#or a generator bus) the reactive power is not specified. Therefore in order to
calculate the phase of bus voltage of a generator bus using equation 01I1 we have to
estimate the reactive power from the bus voltages and admittances as shown below :
#rom equation 0F1 we get)
S
1
1
1

n
q
q q

q
q q
V
j! "
V Y V Y V Y

+ +

+

1
]
1

1
1
S

q
q
n
q
q q q
V Y V Y V j! "
. . . . 01J1
#rom equation 01J1 the equation for complex power in bus-p during 0kQ11
th
iteration can be obtained as shown in equation 0K1
1
]
1

+ + +
1
1
1 S 1 1
1 0

q
k

n
q
q
k
q
k

V Y V Y V j! " . . . . 0K1
The reactive power of bus-p during 0kQ11
th
iteration is given by imaginary part of
equation 0K1.
Reactive power of bus-p during 0kQ11
th
iteration is given by
1
1
]
1

1
]
1

+ +
1
1
1 S 1
1 0 &m 1 1 0

q
k

n
q
q
k
q
k

V Y V Y V aginar$ % !
. . . 011
Dlso) for generator buses a lower and upper limit for reactive powers will be
specified. &n each iteration) the reactive power of generator bus is calculated using
equation 011 and then checked with specified limits. &f it violates the specified limit)
then the reactive power of the bus is equated to the limits violated and it is treated as load
bus. &f it does not violate the limits then the bus is treated as generator bus.
-----------------------------------------------------------------------------------------
1.; @&m0,tat$&n &1 s)ac3 +,s 0&'er an/ )$ne 1)&'s
1F
The slack bus power can be calculated after computing the bus voltage upto the
specified accuracy. The equation 01J1 can be used the slack bus power. 'ere) bus-p is a
slack bus.
1
]
1

n
q
k
q q
V Y V j! "
1
S
. . . . 01
The line flows are the power fed by the buses into various lines and they are
calculated as shown below :
Aonsider a line connecting bus-p and bus-q as shown in fig. 1 below. ;sually the
transmission line is connected to buses using transformers at its ends. The U equivalent
of the transmission line with transformers at its end consist of a series admittance C
pq
and
shunt admittances y
L
pq
as shown in fig. below .
($6.1 D e8,$9a)ent &1 a transm$ss$&n )$ne 1&r e9a),at$n6 )$ne 1)&'s
&n fig.1)

1 0
V
1
q
q q q q q
Y
V Y V V I I I + +
. . . 0"1
Aomplex power in=ected by bus-p in line pq - %
pq
is given by)
S S
q q q q
V I V j! " &
1
1
]
1

1 0
V
q
q q
Y
V Y V V
. . . 0$1
&n fig.)

1 0
V
1
q
q q q q q q
Y
V Y V V I I I + +
. . . . 091
Aomplex power in=ected by bus-q in line pq - %
pq
is given by)
S S
q q q q q q
V I V j! " &
1
1
]
1

1 0
V
q
q q q
Y
V Y V V
. . . 0F1
*ower loss in the transmission line pq is %
pq) loss
5 %
pq
6 %
qp
. . . .0H1
---------------------------------------------------------------------------------------------
1.$ Pr&ce/,re 1&r )&a/ 1)&' s&),t$&n +* 2a,ss % Se$/a) meth&/
Ste0 1 - Dssume a flat voltage profile 1 Q =K for all buses 0nodes1 except in the slack bus.
The voltage of slack bus is the specified voltage and it is not modified in any iteration.
Ste0 2 - Dssume a suitable value of W called convergence criterion. 'ere W is a specified
change in bus voltage that is used to compare the actual change in bus voltage between k
th
and 0kQ11
th
iteration.
1H
Ste0 3 - %et iteration count k 5 K and assumed profile of the buses are denoted as .
1
K
)
.

K
) .
"
K
)BB.
n
K
except slack bus.
Ste0 ; - %et bus count p 5 1.
Ste0 5 - Aheck for slack bus. &f it is a slack bus then go to step-1) or otherwise go to
next step.
Ste0 ! - Aheck for generator bus. &f it is a generator bus go to next step or otherwise 0i.e.
if it is load bus1 go to step I.
Ste0 B - Temporarily set
sec

V V
and phase of .
p
k
as the k
th
iteration value if the
bus-p is a generator bus where
sec

V
is the specified magnitude of voltage for bus-p.
Then calculate the reactive power of the generator bus using the following equation :
1
1
]
1

1
]
1

+ +
1
1
1 S 1
1 0 &m 1 1 0

q
k

n
q
q
k
q
k

V Y V Y V aginar$ % !
The calculated reactive power may be within specified limits or it may violate the
limits. &f the calculated reactive power is within the specified limits then consider the bus
as generator bus and set
1
)
+

k
ca'
! !
for this iteration and go to step-I.
&f the calculated reactive power violates the specified limit for reactive power then
treat this bus as a load bus. The magnitude of the reactive power at this bus will
correspond to the limit it has violated.
i.e. if
min ) min )
1
)
k
ca'
! then! ! !
+
0or1
max ) max )
1
)
k
ca'
! then! ! !
+
%ince the bus is treated as load bus) take actual value of .
p
k
for 0kQ11
th
iteration.
i.e. -.
p
k
-need not be replaced by -.
p
-
spec
when the generator bus is treated as load bus.
Then go to step-J.
Ste0 E - #or generator bus the magnitude of voltage does not change and so for all
iterations the magnitude of bus voltage is the specified value. The phase of the bus
voltage can be calculated as shown below.
( )
1
1
]
1

+
+ +
1
1 1
1
S
1
)
1

q
n
q
k
q
k
q q
k

k
te(
V Y V Y
V
j! "
Y
V
1
1
]
1

+
+
+
1
)
1
) 1 1
Re
. &m
tan
k
te(
k
te( k

a'artofV
artofV ag

<ow) the 0kQ11th iteration voltage of the generator bus is given by

1 1 + +

k

sec

V V
where -.
p
-
spec
5 magnitude of specified bus voltage.
Dfter calculating .
p
kQ1
for generator bus go to step-11.
Ste0 F - #or the load bus the 0kQ11
th
iteration value of load bus-p voltage) .
p
kQ1
can be
calculated using the following equation.
( )
1
1
]
1

+
+ +
1
1 1
1
S
1
1

q
n
q
k
q
k
q q
k

V Y V Y
V
j! "
Y
V
1I
Ste0 1 - Dn acceleration factor) X can be used for faster convergence. &f acceleration
factor is specified then modify the 0kQ11
th
iteration value of bus-p voltage using the
following equation.
1 0
1 1
)
k

k
acc
V V V V +
+ +

Then set) .
p
kQ1
5 .
p)acc
kQ1
.
<ote : D suitable value of X for a particular system can be obtained by running trial load
flows. #or most of the systems the recommended value of X is 1.F.
Ste0 11 - Aalculate the change in the bus-p voltage) using the relation)
k

V V V
+ + 1 1
Ste0 12 - Repeat the steps $ to 11 until all the bus voltages have been calculated. #or
this increment the bus count by 1 and go to step-9) until the bus count is n.
Ste0 13 - #ind out the largest of the absolute value of the change in voltage. i.e. find the
largest among -G.
1
kQ1
-)-G.

kQ1
-)-G.
"
kQ1
-)BBB..-G.
n
kQ1
-. 2et this largest change
be -G.
max
-. Aheck whether this largest change -G.
max
- is less than the prescribed
tolerance W. &f -G.
max
- is less than W then move to next step. Etherwise increment the
iteration count and go to step-$.
Ste0 1; - Aalculate the line flows and slack bus power using the bus 0node1 voltages.
D computer program can be developed using the procedure given above. D flowchart for
the computational procedure have been presented in fig..
<ote : #or pure load bus the real and reactive power are considered as negative. #or pure
generator buses the real and reactive power are always considered as positive) even if the
bus is treated as load bus. &n buses having generators and loads connected to it) either the
net power will be specified or the generated and load powers will be individually
specified.
The specified net power is the difference between generator T load power and so it can
be used as such in calculation without changing the sign. >hen the generator and load
power are individually specified) then the power used for calculation is the value obtained
by subtracting load power from generator power 0i.e. %
p
5 Y%
3
6 Y%
2
and %
p
5 *
p
Q =+
p
1.
1J
($6. 2 ()&'chart 1&r )&a/ 1)&' s&),t$&n +* 2a,ss - Se$/a) meth&/
K
. NEWTON - RAPCSON METCO? O( LOA? (LOW ANALYSIS -
1
The <ewton 6 Raphson method of load flow analysis is an iterative method which
approximates the set of non-linear simultaneous equations to a set of linear
simultaneous 0load flow1 equations using TaylorLs series expansion and the terms are
limited to first order approximation.
The load flow equations of <ewton-Raphson method Zequation 011 to 0"1[ are
non-linear equations in terms of real and imaginary part of bus voltages.
[ ]

+ +
n
q
q q q q q q q q
B e # f f B f # e e "
1
1 0 1 0
. . . . 011
[ ]

+
n
q
q q q q q q q q
B e # f e B f # e f !
1
1 0 1 0
. . . . 01

f e V + . . . . 0"1
where) e
p
5 Real part of .
p
0.oltage of bus-p1
f
q
5 &maginary part of .
p
0.oltage of bus-p1
3
pq
) 4
pq
5 Aonductance T susceptance of admittance C
pq
respectively.
The voltage of a slack bus will be known quantity in a power system and so it
need not be solved. #or load buses *
p
and +
p
will be specified and we have to solve .
p
.
#or a generator bus pp and -.
p
- will be specified and we have to solve +
p
and phase of
.
p
.
&n order to calculate .
p
of load bus and +
p
T phase of .
p
for generator bus we
have to first calculate the real and imaginary part of bus voltages. Therefore the
unknown quantities are real part 0e
p
1 of bus voltages and imaginary part 0f
p
1 of bus
voltages.
#or a n-bus system) the bus-1 is considered as slack bus and the real and
imaginary part of remaining 0n-11 bus voltages have to be solved. 'ence we have 0n-11
variables to be solved and they are e

) e
"
) BB.e
n
) f

) f
"
)BB.f
n
. Thus to solve the
problem for 0n-11 variables we need to solve 0n-11 set of equations.
-----------------------------------------------------------------------------------------------------------
2.1 Mathemat$ca) +ac36r&,n/ 1&r Ne't&n - Ra0hs&n meth&/ &1 ana)*s$s
2et x
1
) x

) BBB.x
n
be a set of unknown variables and y
1
) y

)BB..y
n
be a set of
specified quantities. <ow the specified quantities can be expressed as a non-linear
function of unknown variables as shown below.
y
1
5 f
1
0x
1
) x

)B..)x
n
1M
y

5 f

0x
1
) x

)B..)x
n
1M
. . -
. . -. . . . 011
. . -
y
n
5 f
n
0x
1
) x

)B..)x
n
1\
2et us assume an approximate initial solution x
K
1
)x
K

)B..)x
K
n
for the above equations.
The prefix (ero refers to (eroth iteration in the process of solving the above non-linear
equations.

2et Gx
K
1
) Gx
K

)B..) Gx
K
n
are the corrections required for x
K
1
) x
K

)B..)x
K
n
respectively
for the next better solution. <ow the non-linear equations can be expressed as shown
below) i.e. they can be expressed as functions of modified variables) x
K
1
Q Gx
K
1
) x
K

Q
Gx
K

)B..)x
K
n
Q

Gx
K
n
.
y
1
5 f
1
0x
K
1
Q Gx
K
1
) x
K

Q Gx
K

)B..) x
K
n
Q

Gx
K
n
1 M
y

5 f

0x
K
1
Q Gx
K
1
) x
K

Q Gx
K

)B..) x
K
n
Q

Gx
K
n
1 M
. . -
. . -. . . . 01
. . -
y
n
5 f
n
0x
K
1
Q Gx
K
1
) x
K

Q Gx
K

)B..) x
K
n
Q

Gx
K
n
1\
The above equations are linearised about the initial guess using TaylorLs series
expansion. The linearised equations with second order and higher derivatives neglected
are given below.

n
n
n n
%
f
%
%
f
%
%
f
% % % % f $

+ +

+
K

1
1 K
1
K K

K
1 1 1
........ 1 ) )......... ) 0

n
n
n n
%
f
%
%
f
%
%
f
% % % % f $

+ +

+
K

1
1 K
1
K K

K
1
........ 1 ) )......... ) 0

. .
. . . . . . . 0"1
. .
n
n
n n n n
%
f
%
%
f
%
%
f
% % % % f $

+ +

+
K

1
1 K
1
K K

K
1
........ 1 ) )......... ) 0
The equation 0"1 can be rearranged as shown in equation 0$1
n
n
n n
%
f
%
%
f
%
%
f
% % % % f $

+ +

1
1 K
1
K K

K
1 1 1
........ 1 ) )......... ) 0
n
n
n n
%
f
%
%
f
%
%
f
% % % % f $

+ +

1
1 K
1
K K

K
1
........ 1 ) )......... ) 0
. .
. . . . . . . 0$1
. .
n
n
n n n n
%
f
%
%
f
%
%
f
% % % % f $

+ +

1
1 K
1
K K

K
1
........ 1 ) )......... ) 0
2et
1
K K

K
1 1 1
1 ) )......... ) 0 $ % % % f $
n

K K

K
1
1 ) )......... ) 0 $ % % % f $
n

. .
. .
. .

n n n n
$ % % % f $ 1 ) )......... ) 0
K K

K
1
<ow the equation 0$1 can be written as
"
n
n
n
%
f
%
%
f
%
%
f
% $

+ +

1
1 K
1 1
........
n
n
n
%
f
%
%
f
%
%
f
% $

+ +

1
1 K
1
........
. .
. . . . . . . 091
. .
n
n
n n
%
f
%
%
f
%
%
f
% $

+ +

1
1 K
1
........
The equation 091 can be arranged in the matrix form as shown in equation 0F1
1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
]
1

1
1
1
1
1
1
1
1
]
1

K
J

K
1
1

1
1
1

1
.
.
.
..
. . .
. . . . . .
. . . . . .
. . . . . .
. . .
. . .
.
.
.
n
n
n n n
n
n
n
%
%
%
%
f
%
f
%
f
%
f
%
f
%
f
%
f
%
f
%
f
$
$
$
. . . . . . 0F1
&n matrix 0vector1 notation) the equation 0F1 can be expressed as
B 5 G@ . . . . 0H1
>here B 5
1
1
1
1
1
1
1
1
]
1

n
$
$
$
.
.
.

1
7 G 5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
]
1

..
. . .
. . . . . .
. . . . . .
. . . . . .
. . .
. . .
1

1
1
1
n
n n n
n
n
%
f
%
f
%
f
%
f
%
f
%
f
%
f
%
f
%
f
7 @ 5
1
1
1
1
1
1
1
1
]
1

K
J

K
1
.
.
.
n
%
%
%
'ere ] is the first derivative matrix and it is called ]acobian matrix. The elements of
]acobian matrix are obtained by evaluating the first derivative at the assumed solution.
The 4 matrix is called residual column vector. The elements of matrix 4 are the
difference between the specified quantities and calculated quantities at the assumed
$
solution. >ith the elements of ] and 4 are known) the elements of matrix A are obtained
by solving the matrix equations 0H1. The A matrix is called increment solution vector.
The solution of the matrix equation 0H1 gives
K K

K
1
)........ )
n
% % % . The next better
solution is obtained as follows.
K
1
K
1
1
1
% % % +
K

% % % +
. .
. . . . . . 0I1
. .
K K 1
n n n
% % % +
>ith the new solution equation 0I1 the process is repeated to find next new solution.
The iterative process is terminated if any one of the following condition is satisfied.
i1 The largest 0magnitude of the1 element in the 4 matrix is less than a
prespecified value.
ii1 The largest 0magnitude of the1 element in the A matrix is less than a
prespecified value.
&n this method the initial values assumed should be close to actual solution.
Etherwise) there are chances of the solution diverging rather than converging and it may
not be possible to get a solution. &n power system problems) the choice of initial values
may not pose a problem because) a flat voltage profile) i.e. .
p
5 1 Q =K for p51))B.)n
except the slack bus has been found to be satisfactory for almost all practical systems.
------------------------------------------------------------------------------------------------------
2.2 A00)*$n6 Ne't&n % Ra0hs&n techn$8,e t& )&a/ 1)&' 0r&+)em
Aonsider a system with n-buses. The bus-1 is usually selected as slack bus. The
other buses 0i.e. bus- to bus-n1 can be either generator bus or load bus. #or load buses
the specified quantities are *
p
and +
p
. #or generator buses the specified quantities are *
p
and M.
p
M.
2.2.1 (&rmat$&n &1 2(n-1" e8,at$&n 1&r a s*stem '$th n +,ses -
@ase ($" When a)) the (n-1" +,ses are )&a/ +,ses -
&n this case) bus-1 is slack bus and bus- to bus-n are load buses.
2et *

) *
"
).. . . . .*
n
be the specified real powers and +

) +
"
). . . . . +
n
be the
specified reactive powers of 0n-11 load buses. The unknown variables are ral part of bus
voltages e

) e
"
)BB.e
n
and reactive part of bus voltages f

)f
"
)B..f
n
. <ow the matrix
equation B 5 G@ for this power system problem will be in the form shown in equation 0J1
9
. . . 0J1
@ase ($$" When the s*stem has +&th )&a/ an/ 6enerat&r +,ses
&n this case also the bus-1 is slack bus. 2et buses )")$)BBm) be load buses and
the remaining buses the generator buses. 'ere buses are numbered such that the first bus
is slack bus) bus- to bus-m are load buses and bus-0mQ11 to bus-n are generator buses.
2et *

) *
"
)BBB.*
n
be the specified real powers of 0n-11 buses. 2et +

) +")B..
+
m
be the specified reactive powers of load buses. 2et M.
mQ1
M)M.
mQ
M)BBM.
n
M be the
specified magnitude of voltages of generator buses. The unknown variables are real part
of bus voltages e

) e
"
)BBe
n
and imaginary part of bus voltages f) f")BB.fn. <ow the
matrix B 5 G@ for this case will be in the form shown in equation 01K1
F
. . . 01K1
2.3 @&m0,t$n6 the e)ements &1 matr$ces -
The elements of ]acobian matrix (G" can be derived from the three load flow
equations 011 to 0"1. The elements of the matrices are obtained by partially differentiating
the load flow equations with respect to a unknown variable and then evaluating the first
derivaties using the solution of previous iteration. #or first iteration the initial assumed
values are used as the solution of previous iteration. The initial values are e
p
K
and f
p
K
for
p51))")BBn.
The element of the residual column (B" is the difference between the specified
value of the quantity and the calculated value of the quantity using the solution of
previous iteration.
2et *
p)spec
) +
p)spec
and M.
p
M
spec
be the specified quantities at the bus-p. #or the initial
solution the values of *
p
K
) +
p
K
and M.
p
K
M can be calculated using the load flow equations.
<ow for the first iteration)
G*
p
5 *
p)spec
6 *
p
K
G+
p
5 +
p)spec
6 +
p
K
Dfter calculating the elements of ]acobian matrix (G" and residual column vector
(B") the elements of increment voltage vector (@" can be calculated by using any standard
technique.
<ow the next better solution will be
K K 1

e e e +
K K 1

f f f +
These values of voltages will be used in the next iteration. This process will be
repeated and in general the new better estimates for bus voltages will be
k

e e e +
+1
H
k

f f f +
+1
The process is repeated till the magnitude of the largest element in the residual
column vector is less than the prescribed value.
2.3.1 Pr&ce/,re 1&r )&a/ 1)&' s&),t$&n +* Ne't&n % Ra0hs&n meth&/ -
Ste0 1 - Dssume a flat voltage profile 1 Q =K for all buses 0nodes1 except the slack bus.
The voltage of the slack bus is the specified voltage and it is not modified in any
iteration.
Ste0 2 - Dssume a suitable value of W called convergence criterion. 'ence W is a
specified change in the residue that is used to compare the actual residues 0G* and G+ or
G.1 at the end of each iteration.
Ste0 3 - %et iteration count k 5 K) and assumed voltage profile of the buses are denoted
as .
1
K
) .

K
) .
"
K
)BB..
n
K
except slack bus. The real part of the voltage are denoted as e
p
K
and the reactive part as f
p
K
for p 5 1))")BB.n.
Ste0 ; - %et bus count p 5 1.
Ste0 5 - Aheck for slack bus. &f it is a slack bus then go to step-1") otherwise go to next
step.
Ste0 ! - Aalculate the real and reactive power of bus-p using the following equation.
[ ]

+ +
n
q
q q q q q q q q
B e # f f B f # e e "
1
1 0 1 0

[ ]

+
n
q
q q q q q q q q
B e # f e B f # e f !
1
1 0 1 0

Ste0 B - Aalculate the change in real power.
Ahange in real power)
k
sec
k

" "
)
where

sec
"
)
%pecified real power of bus-p
Ste0 E - Aheck for generator bus. &f it is a generator bus then go to next step 0step-J1 or
otherwise 0i.e. if it is a load bus1 go to step-1.
Ste0 F - Aheck for reactive power limit violation of generator buses. #or this compare
the calculated reactive power +
p
k
with specified limits. &f the limit is violated go to
step-11) otherwise go to next step.
Ste0 1 - &f the calculated reactive power is within the specified limits then consider this
bus as generator bus. <ow calculate the voltage residue 0change in voltage1 using the
following equation.

k

sec

V V V
>here

sec

V
specified voltage magnitude for generator bus. Then go to step-1".
Ste0 11- &f the reactive power limit is violated then treat this bus as a load bus. <ow the
specified reactive power for this bus will correspond to the limit violated.
i.e. if
min ) ) min ) sec
k

! then! ! !
i.e. if
max ) ) max ) sec
k

! then! ! !
Ste0 12 - Aalculate the change in reactive power for load bus 0or for the generator bus
treated as load bus1.
I
Ahange in reactive power)
k
sec
k

! ! !
)
Ste0 13 - Repeat steps 9 to 1 until all resides 0change in * and + or .1 are calculated.
#or this increment the bus count by 1 and go to step F) until the bus count is n.
Ste0 1; - !etermine the largest of the absolute value of the residue 0i.e. find the largest
among . )

V or ! " 2et this largest change by GR.

Ste0 15 - Aompare GR and W then go to step K. &f GR ^ W then go to step K. &f GR _ W
then go to next step 0step 1F1
Ste0 1! - !etermine the elements of ]acobian matrix (G" by partially differentiating the
load flow equations and evaluating the equations using the k
th
iteration values.
Ste0 1B - Aalculate the increments in real and reactive part of voltages)
k

e
and
k

f
by
solving the matrix equation B 5 G@ ZThe elements of B matrix are

)
k

V or ! "
calculated in the previous steps The elements of @ matrix are
k

e
and
k

f
which are
voltage increments to be solved from the matrix equation B 5 G@[.
Ste0 1E - Aalculate the new bus voltages as shown below

k

e e e +
+1
7 p 5 1))") BBBn except slack bus.

k

f f f +
+1
7 p 5 1))")BBB.n except slack bus.

,
_

+
+
+
+ + + +
1
1
1 1 1 1 1
tan 1 0 1 0
k

k
k

f
e
and f e V
1 1 1 + + +

k

V V
Ste0 1F - Ddvance the iteration count) i.e. k 5 kQ1 and to step-$.
Ste0 2 - Aalculate the line flows.
J
($6. 2.3 ()&'chaart 1&r )&a/ 1)&' s&),t$&n +* Ne't&n % Ra0hs&n meth&/
"K
2.3.1 @OMPARISION O( 2.S. AN? N.R. METCO? O( LOA? (LOW ST#?IES
1. #or 3.%. method the variables are expressed in rectangular coordinates whereas
in <.R. method) they are expressed in polar coordinates. &f rectangular
coordinates are used for <.R. method then memory requirement will be more.
. The number of mathematical operations per iteration will be lesser in 3.%.
method.
". The 3.%. method has linear convergence characteristics whereas the <.R. method
has quadratic convergence characteristics. 'ence <.R. method converges faster
than 3.%. method.
$. &n 3.%. method number of iterations increases with number of buses but in <.R.
method the number of iteration remains constant and it does not depend on the
si(e of the system.
9. &n 3.%. method convergence is affected by the choice of slack bus and the
presence of series capacitors but the <.R. method is less sensitive to these factors.
F. The <.R. method needs only " to 9 iterations to reach an acceptable solution for a
large system. 4ut 3.%. method requires large number of iterations 0"K or more1
for same level of accuracy.
A/9anta6es &1 2.S. meth&/
1. Aalculations are simple and so the programming task is faster.
. The memory requirement is less.
". ;seful for small si(e system.
?$sa/9anta6es &1 2.S. meth&/
1. Require large number of iteration to reach convergence.
. <ot suitable for larger systems.
". Aonvergence time increases with si(e of the system.
A/9anta6es &1 N.R. meth&/
1. The <.R. method is faster) more reliable and the results are accurate.
. Requires less number of iterations for convergence.
". The number of iteration are independent of the si(e of the system 0number of
buses1.
$. %uitable for large si(e systems.
?$sa/9anta6es &1 N.R. meth&/
1. The programming logic is more complex than 3.% method
. The memory requirement is more.
". <umber of calculations per iterations are higher than 3.%. method.
-------------------------------------------------------------------------------------------------------
(AST ?E@O#PLE? METCO?

"1
3.1 Re0resentat$&n &1 Trans1&rmers -
1.1 ($4e/ ta0 sett$n6 trans1&rmer
&n a system having different voltage levels) two buses are separated by a
transformer operating with either a nominal or an off nominal turns ratio. &n the case
of a nominal turns ratio) the transformer is represented by its leakage admittance.
1.2 O11 n&m$na) t,rns rat$& trans1&rmer
&f an off nominal turns ratio is present) the transformer can be represented by
its leakage admittance in series with an ideal autotransformer as shown.
Representation of off nominal turns ratio transformer
'ere the leakage admittance y of the transformer has been referred to the base
value associated with bus q) the non-tap side of the transformer and the tap side bus is
connected directly with the ideal transformer whose turns ratio X 5 01 8 t1 where t is the
per unit tap setting.
The current flow in the transformer in the direction from `xL to `qL is given by
q
q % % $ ) ) I 1 0
%ince the autotransformer is an ideal one) there is no power loss in the transformer.
% % I ) I )
S S

% %

%
I
a
I
)
)
I
S
S
S
1

,
_

%ince X is real) XS 5 X) 5 % I
a
1
q
q

$ )
a
)
a
I
1
]
1

1
q
q

q
)
a
$
)
a
$

. . . . . 011
#rom which
a
$
andY
a
$
Y
q
q
q

%imilarly the current flowing from q to x is

"
q
% q q $ ) ) I 1 0
5
q

q $
a
)
)

,
_

5
q
q
q
)
a
$
) $ . . . . . . 01
&n general
q q qq q
) Y ) Y I +
#rom which
a
$
andY $ Y
q
q q qq

%ince C
pq
5 C
qp
) we can suggest a U equivalent of the above transformer which is shown
in fig. below.
($6. D e8,$9a)ent c$rc,$t &1 &11 n&m$na) t,rns rat$& trans1&rmer
#rom the U equivalent circuit) the self-admittance at any bus is the sum of the
admittances connected at that particular bus and the shunt admittance to ground.
To find the shunt admittance to ground at bus p.
C
pp
5 C
pq
Q y
qo
o
q q
$
a
$
a
$
+

a
$
a
$
$
q q
o

1
]
1

1
1
a a
$
$
q
o
. . . . . . 0"1
Dt bus q)
qo q qq
$ Y Y +
qo
q
q
$
a
$
$ +
a
$
$ $
q
q qo

1
]
1

a
$ $
q qo
1
1
Thus the elements are marked in the U equivalent circuit.
>hen the off nominal turns ratio is represented at bus p for a transformer
connecting buses `pL and `qL) the self admittance at bus p is
""
n q
q

$ $
a a a
$
$ $ Y + +
,
_

+ + + + + .......... 1
1 1
...... ........
1
n
q

$
a
$
$ $ + + + + + ..... .....

1
The mutual admittance from p to q is
a
$
Y
q
q

The self admittance at bus q is

qn q
q
q q qq
$
a
$
a
$
$ $ Y +
,
_

+ + + + +
1
1 ...... .......
1
qn q q q
$ $ $ $ + + + + + ..... ......
1
and is left unchanged.
Enly on the tapped side) the admittance value is to be divided by a

and the
untapped side is left as such.
1.3 P&'er 1)&' thr&,6h 0hase sh$1t$n6 trans1&rmer
D phase shifter regulates the flow of active power by varying its phase angle a. &n load
flow studies) it can be represented by an admittance in series with an ideal auto
transformer having a complex turns ratio as shown in figure below.
($6. Re0resentat$&n &1 a 0hase sh$1t$n6 trans1&rmer
#rom the figure) the terminal voltage at x and p are related by
s s
%

jb a
)
)
+
%ince there is no power loss in an ideal transformer)
% % I ) I )
S S

I
I
S
1

,
_

s s
%

jb a )
)
s s
jb a

1
I
s s
jb a

1
% I
#rom the figure) [ ]
q
q % $ ) ) I
q
q
s s

s s
$ )
jb a
)
jb a
I
1
]
1

+

1
"$
q
s s
q

s s
q
)
jb a
$
)
b a
$
I

&n general q
q

) Y ) Y I +
#rom which
s s
q
q
s s
q

jb a
$
andY
b a
$
Y

%imilarly) the current from q to s is
( )
q
s q q $ ) ) I
q
s s

q $
jb a
)
)
1
]
1

s s
q
q
q
)
jb a
$
) $
+

&n general)
q
q
qq
q ) Y ) Y I +
#rom which
s s
q
q qq qq
jb a
$
andY $ Y
+

%ince
q q
Y Y
) an equivalent U circuit cannot be formed.
>hen the phase shifting transformer is connected between buses `pL and `qL) the
self admittance at bus-p is
n
s s
q

$
b a
$
$ $ Y + +
+
+ + + ...... ......

1
The mutual admittance between buses `pL and `qL is given by
s s
q
q
jb a
$
Y

The self admittance at bus `qL is

qn q q q qq
$ $ $ $ Y + + + + + ...... .......
1
The mutual admittance between buses `qL and `pL is given by
s s
q
q
jb a
$
$
+

&n matrix form

0p1 0q1
0p1
0q1
%ince C is unsymmetrical) the equivalent circuit representation is not possible.
"9
1
1
1
1
1
]
1

+
q
s s
q
s s
q
s s
q
$
jb a
$
jb a
$
b a
$

Re9$e' H,est$&ns (2 mar3s"
1. <ame any two iterative methods need for the solution of load flow
problems.
. 'ow is the swing bus selected in a load flow study b
". >hat are the quantities specified at a generator bus b
$. >hat is meant by acceleration factor in load flow solution b >hat is its
best value b
F. >hat are the uses of phase shifting transformer b
H. >hat is the important factor on which the real power transfer over line
dependb
I. %tate 3auss-%eidal load flow formula.
J. %tate <ewton-Raphson load flow formula.
1K. >hat are the applications of load flow analysisb
11. ,ention the items specified and not specified in a reference bus.
1. !efine voltage-controlled bus.
1". %tate the advantages of <ewton-Raphson method over 3auss-%eidal
method.
1$. Rxplain the convergence criterion for 3auss-%eidal load flow studies b
19. >hat is a ]acobian matrixb
1F. >hat are the advantages of 3auss-%eidal load flow methodb
1H. &n a load flow problem) how various buses are classified b
1I. %tate the elements of a ]acobian matrix.
1J. >hat are power balance equations in power flow analysisb
K. >hat is the necessity for slack busb
1. >hat is a load busb
. 3iven the simultaneous equations: x
1
Qx

5 $7 x
1
Qx

597 using initial

values x
1
K
5 and x

K
5. >rite down the values for x
1
1
and x

1
0first
iteration values1 using 3auss-%eidal method.
". >rite down the general expression for complex power in=ected by the
source into the ith bus of a power system.
$. >hat is half line charging admittanceb
9. >hat is meant by a transformer with off-nominal turns ratiob
F. !raw the equivalent circuit of tapped transformer for nodal admittance
matrix.
H. !raw the circuit representation of a line with voltage regulating
transformer.
Re9$e' H,est$&ns (1! mar3s"
1.i1 %tate the load flow problem.
ii1 !iscuss the classification of buses.
. !raw the flow chart for load flow solution by 3auss-%eidal iterative
method and explain.
"F
". !erive the static load flow equations of a power system. Rxplain the
<ewton-Raphson method of solving the load flow equations.
$. !raw the flow chart for load flow solution by <ewton-Raphson method.
9. Aompare 3auss-%eidal method and <ewton-Raphson method in detail.
F. Rxplain clearly how the nodal admittance matrix of a system is changed
when an off nominal turns ratio transformer is introduced in a line connected
between two buses.
H. !iscuss the various types of buses and their significance in detail.
I. !erive the static load flow equations using C
bus
.
J. !erive <ewton-Raphson power flow analysis algorithm and give steps for
implementation of this algorithm.
1K. !erive the basic equations for the load flow study using 3auss-%eidal
method. >ith respect to this method) explain the following :
a. Dcceleration factor. b. Aonvergence criteria c. 'andling of *. buses.
11. !raw the representation schemes for
a. *hase shifting transformer b. Tap changing transformer.
1. !erive mathematical model of phase shifting transformer to be used in
power flow analysis.
1". Aarry out one iteration of load flow analysis of the system given below by
3auss-%eidal method.
4us
<o.
4us
Type
-.-
p.u.
p.u. generation p.u. load
* + * +
1 %lack 1.K - - - -
*. 1.K K.I - - -
" *+ - - - 1.K K.$
2ine reactance in p.u. are given below :
4us code &mpedance
1- ]K.9
-" ]K.9
"-1 ]K.9
1$. Aonsider the power system with the following data :
1 "
NNNNNNNNNNNNNNN
1 I -J12 JE J;
Cbus 5 M =I -=1 =$
" M =$ =$ -=I

4us
<o.
4us
Type
3eneration 2oad .oltage
* + * + -.- /
1 %lack - - - - 1.K Kc
*. F.K - K - 1.K$ -
" *+ K K $.K K.9 - -
Dssume that bus can supply any amount of reactive power. Dssuming
flat start) perform the first iteration of power flow analysis using <ewton-
Raphson method.
19. The bus admittance matrix of a " bus power system is
"H
1
1
1
]
1

19 9 1K
9 19 1K
1K 1K K
j j j
j j j
j j j
Y
bus
4us
<o.
4us
Type
*3 +3 *2 +2
1 *.
.JK"
$
- - -
*+ - -
$.KKI
J
1.HJ19
" %lack - - - -
The latest solution is
.
1
5 1.K9 F.JFc
.

5 K.J""I-I.Ic
.
"
5 1.KKc
!etermine the bus voltages at the end of next iteration using 0i1 3% method
0ii1 <R method
1F. The following data are given for a three bus power system.
4us 1 slack bus . specified 5 1.K9Kc
4us *. bus -.- specified 5 1.K p.u. *
3
5 " p.u.
4us " *+ bus *
2
5 $ p.u. +
2
5 p.u.
2ine reactances in p.u. are given below:
Aarry out one
iteration of load
flow solution by
3% method. Take
+ limits of generator is K d + d $.
1H. Repeat problem <o. 1F when the limits of reactive power is specified as K d
+ d .9
1I. The load flow data of a four bus system is given in tables. Dssume the bus
voltage of bus <o." as 1.K$ p.u. and the maximum and minimum reactive
power constraints for bus " are -K.1 and 1.K p.u. respectively. Taking bus 1 as
slack determine the voltages of all the buses t the end of first iteration starting
with flat voltage profile for all buses expect the slack bus using 0i1 3auss-
%eidal method and 0ii1 <ewton-Raphson method.
4us
<o.
Dssumed
voltages
3eneration 2oad
,> ,.DR ,> ,.DR
1 1.KF Q =K.K K.K K.K K.K K.K
1.K Q =K.K K.K K.K K. K.1
" 1.KF Q =K.K K.F K." K.$ K.
$ 1.KF Q =K.K K.K K.K K.$ K.K
4us code &mpedance
1- =K.9
-" =K.9
"-1 =K.9
"I
4us code &mpedance 2ine charging
admittance yL
pq
?
1- K.K Q =K.KI K.K Q =K.K$
1-" K.KF Q =K.$ K.K Q =K.K"
-" K.K$ Q =K.1F K.K Q =K.K9
-$ K.K$ Q =K.1F K.K Q =K.K9
"-$ K.K1 Q =K.K$ K.K Q =K.K19

1J Rxplain clearly with flow chart the computational procedure for load flow
solution using 3auss-%eidal method when the system contains all types of
buses.
K Rxplain clearly with flow chart the computational procedure for load flow
solution using 3auss-%eidal method when the system contains all types of
buses.
----------------------------------------------------------------------------------------------
"J