Chapter 6

Power Flows

ANSWERS TO MULTIPLE-CHOICE TYPE QUESTIONS

6.1 Nonzero
6.2 Upper triangular
6.3 Diagonal; lower triangular
6.4 b
∂ f1 ∂ f1 ∂f
6.5  1
∂ x1 ∂ x2 ∂ xN
6.6 a
6.7 a
6.8 Independent
6.9 P1 , Q1
6.10 Vk , δ k
6.11 Pk & Vk ; Qk & δ k
6.12 − PLk ; − QLk
6.13 Bus data, transmission line data, transformer data
6.14 a
6.15 a
6.16 a
6.17 Newton-Raphson
6.18 Real power, reactive power
6.19 Negative
6.20 a
6.21 Sparse
6.22 Storage, time
6.23 a
6.24 a
6.25 a

137
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.1 − 25 5 10 10 x1  0 
 5 − 10 5 0 x2  2 

 10 5 − 10 10 x3  = 1 
    
 10 0 0 − 20 x4  − 2
There are N − 1 = 3 Gauss elimination steps.
Step 1: Use equation 1 to eliminate x1 from equations 2, 3, and 4.

 −25 5 10 10   x1   0 
 0 −9 7  
 2   x2   2 
=
 0 7 −6 14   x3   1 
    
 0 2 4 −16   x4   −2 
Step 2: Use equation 2 to eliminate x2 from equations 3 and 4.
 −25 5 10 10   0 
 0 −9 7 
2  1 x   2 
    
 −5 140   x2  =  23 
 0 0   
 9 9   x3   9 
 
 50 −140   x4   14 
 0 0
9 9   − 9 

Step 3: Use equation 3 to eliminate x3 from equation 4.

− 25 5 10 10  x   0 
 0 − 9 7 2   1  2 
  x2   
 5 140   = 23
 0 0 − x
9 9   3  9 
   
 0 0 0 140 x4  24
The matrix is now triangular. Solving via back substitution,
6
x4 =
35
1
x3 =
5
1
x2 = −
35
1
x1 =
7
6.2  6 2 1   x1   3 
 4 10 2   x  =  4 
  2  
 3 4 14   x3   2 

138
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 There are N − 1 = 2 Gauss elimination steps.
Step 1: Use equation 1 to eliminate x1 from equations 2 and 3.

6 2 1   x1  3 
 26 4  x2  = 2 
0
 3 3  x  1 
 3
 27  2
0 3
2 
Step 2: Use equation 2 to eliminate x2 from equation 3.

6 2 1  x1   3 
 26 4  x2  =  2 
0
 3 3  x   5 
 3 −
 339  26
0 0
26 
The matrix is now triangular. Solving via back substitution,
5
x3 = −
339
79
x2 =
339
48
x1 =
113

6.3 4 2 1  x1  3

4 10 2    = 4
  x2   
3 4 14 x3  2
There are N − 1 = 2 Gauss elimination steps.
Step 1: Use equation 1 to eliminate x1 from equations 2 and 3.

4 2 1   x1  3 
0 8 1  x2  = 1 

 5 53 x3  − 1
0 2 4   4 
Step 2: Use equation 2 to eliminate x2 from equation 3.

4 2 1  x1   3 
0 8 1  x2  =  1 

 207  x3   9 
0 0 16  − 16

139
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 The matrix is now triangular. Solving via back substitution,
1
x3 = −
23
3
x2 =
23
16
x1 =
23
6.4 − 10 10   x1   10
  = 
 5 − 5 x2  − 10
If we perform Gauss elimination, we arrive at:
− 10 10  x1  10
 0 0 x  = − 5
   2  
nd
The 2 row says that
0 x1 + 0 x2 = − 5
Clearly, this is not possible.
The problem in this case is that the system we started with is indeterminate. That is,
− 10 10
=0
5 −5
This system has no solution.
6.5 From Eq. (7.1.6), back substitution is given by:
yk − Ak ,k +1 X k +1 − Ak ,k + 2 X k + 2  − Ak , N X N
Xk =
Akk
which requires one division, (N − k) multiplications and (N − k) subtractions for
each k = N ,( N − 1)1 .
Summary- Back Substitution
Solving for # Divisions # Multiplications # Subtractions
XN 1 0 0
XN−1 1 1 1
XN−2 1 2 2

X1 1 (N−1) (N−1)
N −1
N ( N − 1) N −1
N ( N − 1)
Totals N i =
i =1 2
i =
i =1 2

140
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.6 6 0 0   0 −0.3333 −0.1667 
D = 0 10 0  
M = D ( D − A) =  −0.4
−1
0 −0.2 
0 0 14   −0.2143 −0.2857 0 
x (i + 1) = Mx (i) + D −1 B
x1 = 0.4243 x 2 = 0.2325 x 3 = −0.0152 after 10 iterations

6.7 6 0 0   0 −0.3333 −0.1667 

D =  4 10 0  M = D ( D − A) =  0 0.1333 −0.1333
  −1

 3 4 14   0 0.0333 0.0738 
x (i + 1) = Mx (i) + D −1 B
x1 = 0.4249 x 2 = 0.2330 x 3 = −0.0148 after 4 iterations
The Gauss-Seidel method converges over twice as fast as the Jacobi method.

6.8 10 − 2 − 4  x1  − 2
− 2 6 − 2 x  =  3
   2  
− 4 − 2 10 x3  − 1

1 
10 0 0
10 0 0  
D =  0 6 0
1
D =0
−1
0
 6 
 0 0 10  1
0 0 
 10
 1 2
0 5 5
 
1 1
M = D (D − A) = 
−1
0
3 3
2 1 
 0
5 5 
Expressing in the form of Eq. (6.2.6)
 1 2 1 
0 5  x (i ) 10
 0 0
 x1 (i + 1)  5
 1   − 2
  1 1  +0 0   3 
1
x2 (i + 1) =  3 0 x (i
3  2  
)
6 
x3 (i + 1)   x3 (i ) 
2 1 1  − 1
 0 0 0 
5 5   10 

141
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Using
1
x(0) = 1
0
we arrive at the following table:
i 0 1 2 3 4 5 6 7 8 9 10

x1 1 0 0.1667 –0.04 –0.0444 –0.0962 –0.1085 –0.1239 –0.1303 –0.1355 –0.1382

x2 1 0.8333 0.6667 0.5778 0.52 0.485 0.4633 0.45 0.4418 0.4367 0.4336

x3 0 0.5 0.0667 0.1 –0.0004 –0.0138 –0.0415 –0.0507 –0.0596 –0.0638 –0.0668

ε Inf Inf 1.24 1.0044 30 2.0095 0.2235 0.1741 0.0704 0.0484

ε for iteration 9 is less than 0.05.

∴ x1 = − 0.14
x2 = 0.43
x3 = − 0.07

6.9 x2 − 3 x1 + 1.9 = 0
x2 + x12 − 3.0 = 0
Rearrange to solve for x1 and x2 :
1
x1 = x + 0.6333
3 2
x2 = 3.0 − x12
Starting with an initial guess of x1 (0) = 1 and x2 (0) = 1
1
x1 (1) = x2 (0) + 0.6333 = 0.9667
3
x2 (1) = 3 − [ x1 (0)]2 = 2
We repeat this procedure using the general equations
1
x1 (n + 1) = x2 (n) + 0.6333
3
x2 (n + 1) = 3 − [ x1 (n)]2

n 0 1 2 3 ... 47 48 49 50
x1 1 0.9667 1.3 1.3219 ... 1.1746 1.1735 1.1734 1.1743

x2 1 2.0 2.0656 1.31 ... 1.6205 1.6202 1.6229 1.6231

142
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 After 50 iterations, x1 and x2 have a precision of 3 significant digits.
x1 = 1.17
x2 = 1.62

6.10 x2 − 4 x + 1 = 0
Rearrange to solve for x:
1 2 1
x= x +
4 4
Using the general form
1 1
x(n + 1) = [ x(n)]2 + and x(0) = 1
4 4
we can obtain the following table:
n 0 1 2 3 4 5
x 1 0.5 0.3125 0.2744 0.2688 0.2681
ε 0.5 0.375 0.1219 0.0204 0.0028

ε (4) = 0.0028 < 0.01

∴ x = 0.27
Using the quadratic formula
−b ± b2 − 4 ac
x=
2a
4 ± (− 4)2 − 4
x=
2
x = 0.268, 3.732
Note: There does not exist an initial guess which will give us the second solution x = 3.732.

6.11 After 100 iterations the Jacobi method does not converge.
After 100 iterations the Gauss-Seidel method does not converge.
6.12 Rewriting the given equations,
x2
x1 = + 0.633; x2 = 1.8 − x12
3
With an initial guess of x1 (0) = 1 and x2 (0) = 1,
update x1 with the first equation above, and x2 with the second equation.
x2 (0) 1
Thus x1 = + 0.633 = + 0.633 = 0.9663
3 3
and x2 = 1.8 − x1 (0) = 1.8 − 1 = 0.8
2

143
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 In succeeding iterations, compute more generally as
x2 ( n )
x1 (n + 1) = + 0.633
3
and x2 (n + 1) = 1.8 − x12 (n)
After several iterations, x1 = 0.938 and x2 = 0.917 .
After a few more iterations, x1 = 0.93926 and x2 = 0.9178 .
However, note that an “uneducated guess” of initial values, such as x1 (0) = x2 (0) = 100 ,
would have caused the solution to diverge.

1  − 1 + 0.5 j 
6.13 x1 =  − j10 × x2 − j10
− 20 j  x1*

1 −3 + j 
x2 =  − j10 × x1 − j10
− 20 j  x2
*

Rewriting in the general form:
1  − 1 + 0.5 j 
x1 (n + 1) =  − j10 × x2 (n) − j10
− 20 j  [ x1 (n)]*

1  −3 + j 
x2 (n + 1) = − j10 × x1 (n + 1) − j10
− 20 j [ x2 (n)]* 
Solving using MATLAB with x1 (0) = 1, x2 (0) = 1, ε = 0.05 :
x1 = 0.91 − 0.16 j
x2 = 0.86 − 0.22 j
ε = 0.036 after 3 iterations

6.14 f ( x) = x3 − 6 x2 + 9x − 4 = 0
1 2 4
x = − x 3 + x 2 + = g( x )
9 3 9
Apply Gauss-Seidel algorithm with x(0) = 2.
First iteration yields
1 2 4
x (1) = g(2) = − 23 + 22 + = 2.2222
9 3 9
Second iteration: x(2) = g ( 2.2222 ) = 2.5173
Subsequent iterations will result in
2.8966,3.3376, 3.7398, 3.9568, 3.9988, and 4.0000 ←

Note: There is a repeated root at x = 1 also. With a wrong initial estimate, the solution
might diverge.

144
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.15 x 2 cos x − x + 0.5 = 0
Rearranging to solve for x,
x = x 2 cos x + 0.5
In the general form
x(n + 1) = [ x(n)]2 cos x(n) + 0.5
Solving using MATLAB with x(0) = 1, ε = 0.01:
x = 1.05
ε = 0.007 after 2 iterations
To test which initial guesses result in convergence, we use MATLAB and try out initial
guesses between − 20 and 20, with a resolution of 0.001. In the following plot the curve is
the original function we solved. The shaded area is the values of x(0) which resulted in
convergence.

145
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.16 Eq. (6.2.6) is: x (i + 1) = Mx (i ) = D −1 y
Taking the Z transform (assume zero initial conditions):
zX ( z) = MX ( z ) + D −1Y ( z )
( zU − M ) X ( z) = D −1Y ( z)
X ( z) = ( zU − M ) D −1Y ( z ) = G ( z )Y ( z )
−1

6.17 For Jacobi,

 − A12 
0
− A12   A11 
−1
 A11 0   0 
M = D −1 ( D − A) =  =
 0 A22  − A
 21 0   − A21 
 0 
 A22 
 A12 
 Z A11 
Det ( ZU − M ) = Det   = Z 2 − A12 A21 = 0
 A21  A11 A22
 Z 
 A22 
A A 
Z = ±  12 21 
 A11 A22 
For Gauss-Seidel,
 − A12 
 0
A11 
−1
 A11 0  0 − A12   
M = D ( D − A) = 
−1
  =
 A21 A22  0 0   A12 A21 
0 
 A11 A22 

 A12 
Z 
 = Z  Z − A12 A21  = 0
A11
det ( ZU − M ) = Det 
 A12 A21   
0 Z −   A11 A22 
 A11 A22 

A A
Z = 0, 12 21
A11 A22

A12 A21
When N = 2 , both Jacobi and Gauss-Seidel converge if and only if < 1.
A11 A22

146
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.18 x 3 + 8 x 2 + 2 x − 50 = 0
x(0) = 1
df
J (i) = = [3 x 2 + 16 x + 2]x = x (i )
dx x = x (i )

In the general form:

1
x (i + 1) = x (i) + × (− x (i )3 − 8 x (i)2 − 2 x (i) + 50)
3 x (i) + 16 x (i) + 2
2

Using x(0) = 1, we arrive at

i 0 1 2 3 4 5
x 1 2.857 2.243 2.129 2.126 2.126
ε 1.857 0.215 0.051 0.0018 2.0E-6

After 5 iterations, ε < 0.001.

∴ x = 2.126

6.19 Repeating Problem 6.18 with x(0) = − 2, we get:

i 0 1 2 3
x –2 –3.67 –3.610 –3.611
ε 0.833 0.015 0.00016
After 3 iterations, ε < 0.001 . Therefore,
x = − 3.611
Changing the initial guess from 1 to − 2 caused the Newton-Raphson algorithm to converge
to a different solution.

6.20 x 4 + 3 x 3 − 15 x 2 − 19 x + 30 = 7
df
J (i) = = [4 x 3 + 9 x 2 − 30 x − 19]x = x ( i )
dx x = x (1)

In general form
1
x (i + 1) = x (i) + × (7 − x (i)4 − 3x (i )3 + 15 x (i )2 + 19 x (i) − 30)
4 x (i)3 + 9 x (i)2 − 30 x (i) − 19
Using x(0) = 0, we arrive at:
i 0 1 2 3 4
x 0 1.211 0.796 0.804 0.804
ε ∞ 0.342 0.010 8.7E-6
After 4 iterations, ε < 0.001. Therefore,
x = 0.804

147
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.21 Repeating Problem 6.20 with x(0) = 4, we arrive at x = 3.082 after 4 iterations.
6.22

For points near x = 2.2, the function is flat, meaning the Jacobian is close to zero.
In Newton-Raphson, we iterate using the equation
x(i + 1) = x(i) + J −1 (i)[ y − f ( x(i))]
If J (i) is close to zero, J −1 (i) will be a very large positive or negative number.
If we use an initial guess of x(0) which is near 2.2, the first iteration will cause x(i) to jump
to a point far away from the solution, as illustrated in the following plot:

148
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 df
6.23 J= = 4 x 3 + 36 x 2 + 108 x + 108
dx
From Eq. (6.3.9):
x(i + 1) = x(i) +  4 x 3 (i ) + 36 x 2 (i ) + 108 x(i ) + 108 
{0 −  x 4
(i ) + 12 x 3 (i ) + 54 x 2 (i ) + 108 x(i ) + 81}

i 0 1 2 3 4 … 17 18 19
x(i ) −1 − 1.5 − 1.875 − 2.15625 − 2.3671875 … − 2.9848614 − 2.989901 − 2.9923223

x(19) − x(18) −2.9923223 + 2.989901

= = 0.0008
x(18) −2.989901
Stop after 19 iterations. x(19) = −2.9923223 . Note that x = −3 is one of four solutions to
this 4 degree polynomial. The other three solutions are x = −3 , x = −3 , and x = −3 .
th

6.24 2 x12 + x22 − 10 = 0

x12 − x22 + x1 x2 − 4 = 0
df 4 x 2 x2 
J (i) = = 1 
dx x = x (i ) 2 x1 + x2 − 2 x2 + x1 x = x (i )
In general form
x(i + 1) = x(i) − J −1 (i ) f ( x )
Solving using MATLAB with x1 (0) = 1, x2 (0) = 1, ε = 0.001:
i 0 1 2 3 4
x1 1 2.3 1.934 1.901 1.900

x2 1 1.9 1.683 1.667 1.667

ε 1.30 0.157 0.0175 2E-4
After 4 iterations ε < 0.001. Therefore,
x1 = 1.900
x2 = 1.667

6.25 10 x1 sin x2 + 2 = 0
10 x12 − 10 x1 cos x2 + 1 = 0
df 10sin x2 10 x1 sin x2 
J (i) = = 
dx x = x (i ) 20 x1 − 10cos x2 10 x1 sin x2  x = x (i )
In the general form
x(i + 1) = x(i) − J −1 (i ) f ( x )

149
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Solving using MATLAB with x1 (0) = 1, x2 (0) = 0, ε = 1E − 4 :
i 0 1 2 3 4
x1 1 0.90 0.8587 0.8554 0.8554

x2 0 – 0.20 – 0.2333 – 0.2360 – 0.2360

ε ∞ 0.1667 0.01131 7.0E-5
After 4 iterations, Newton-Raphson converges to
x1 = 0.8554 rad
x2 = − 0.2360 rad

6.26 Repeating Problem 6.25 with x1 (0) = 0.25 and x2 (0) = 0, Newton-Raphson converges to
x1 = 0.2614 rad
x2 = − 0.8711 rad
after 6 iterations.
6.27 To determine which values of x1 (0) result in the answer obtained in Problem 6.25, we test
out initial guesses for x1 between 0 and 1, with a resolution of 0.0001. The conclusion of
this search is that if 0.5 < x1 (0) ≤ 1 and x2 (0) = 0, Newton-Raphson will converge to
x1 = 0.8554 rad and x2 = − 0.2360 rad.
Note: If x1 (0) = 0.5 and x2 (0) = 0. Newton-Raphson does not converge. This is because
4.79 5
J (0) =  
 0 0
Thus, J −1 does not exist.
6.28 (a) Y11 = 2 − j 4 + 3 − j 6 = 5 − j10 = 11.18 ∠ − 63.43 °
Y22 = 2 − j 4 = 4.47 ∠ − 63.43 °
Y33 = 3 − j6 = 6.71 ∠ − 63.43 °
Y12 = Y21 = − 2 + j 4 = 4.47∠116.57 °
Y13 = Y31 = − 3 + j6 = 6.71 ∠ 116.57 °
Y23 = Y32 = 0
N
(b) Pk = Vk  YknVn cos (δ k − δ n − θkn )
n =1

At bus 2,
P2 = V2 Y21V1 cos (δ 2 − δ1 + θ21 ) + Y22V2 cos (δ 2 − δ 2 − θ22 )
+ Y23V3 cos (δ 2 − δ 3 − θ23 )
1.5 = 1.1[4.47 cos (δ 2 − 116.57 °) + 4.47 × 1.1 cos (63.43 °)]

150
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Solving for δ 2
δ 2 = 15.795 °
(c) At bus3
P3 = V3 Y31V1 cos (δ 3 − δ1 − θ31) + Y33V3 cos (−θ33 )
− 1.5 = V3 6.71 cos (δ 3 − 116.57 °) + 6.71V3 cos (63.43 °) Eq. 1
Also
Q3 = V3 Y31V1 sin (δ 3 − δ1 − θ31) + Y33V3 sin (−θ33 )
0.8 = V3 6.71 sin (δ 3 − 116.57 °) + 6.71V3 sin (63.43) Eq. 2
Solving for V3 and δ 3 using Eq. 1 and Eq. 2:
First, rearrange Eq. 1 and Eq. 2 to separate V3 and δ 3
1.5
− − V3 cos (63.43 °) = cos (δ 3 − 116.57 °)
6.71V3
0.8
− V3 sin (63.43 °) = sin (δ 3 − 116.57 °)
6.71V3
Next, square both sides of each equation, then add the two equations. The trigonometric
identity sin 2 (θ ) + cos2 (θ ) = 1 can be used to simplify the right hand side.
2 2
 1.5   0.8 
6.71V − V3 cos (63.43) + 6.71V − V3 sin (63.43) = 1
 3   3 
Solving, we get V3 = 0.9723 p.u.
Substitute V3 back into Eq. 1 or Eq. 2 to get δ 3 = − 15.10 ° .
(d) At bus 1
P1 = V1 Y11V1 cos (−θ11) + Y12V2 cos (−δ 2 − θ12 ) + Y13V3 cos (−δ 3 − θ13 )
= 11.18 cos (63.43 °) + 4.47 × 1.1 cos (− 15.795 ° − 116.57 °)
+ 6.71 × 0.9723 cos (15.10 ° − 116.57 °)
= 0.39 p.u.
(e) Ploss = Pgen − Pload
= P1 + P2 + P3
= 0.39 + 1.5 − 1.5
= 0.39 p.u.

151
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.29 Y21 = 0
1 1 1.72 0.88
Y22 = + +j +j
0.009 + j 0.100 0.009 + j 0.100 2 2
= 18.625 ∠ − 84.50 °
Y23 = 0
1
Y24 = − = 9.960 ∠ 95.14 °
0.009 + j 0.100
1
Y25 = − = 9.960 ∠ 95.14 °
0.009 + j 0.100

6.30 6.25 − j18.695 − 5.00 + j15.00 − 1.25 + j 3.75 0 0

− 5.00 + j15.00 12.9167 − j 38.665 − 1.6667 + j 5.00 − 1.25 + j 3.75 − 5.00 + j15.00

− 1.25 + j 3.75 − 1.6667 + j 5.00 8.7990 − j 32.2294 − 5.8824 + j 23.5294 0

0 − 1.25 + j 3.75 − 5.8824 + j 23.5294 9.8846 − j 36.4037 − 2.7523 + j 9.1743
0 − 5.00 + j15.00 0 − 2.7523 + j 9.1743 7.7523 − j24.1443

6.31 First, we need to find the per-unit shunt admittance of the added capacitor.
75 Mvar = 0.75 p.u.
To find Y, we use the relation
S = V 2Y
S 0.75
Y= = 2 = 0.75 p.u.
V2 1
Now, we can find Y44
1 1 1
Y44 = + +
0.08 + j 0.24 0.01 + j 0.04 0.03 + j 0.10
0.05 + 0.01 + 0.04
+j + j 0.75
2
= 9.8846 − j35.6537

6.32

152
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Assume flat start V1 (0) = 1.0 ∠ 0, V2 (0) = 1.0 ∠ 0
1  P2 − JQ2 
V2 (i + 1) =  *
− Y21V1 (i)
Y22  V2 (i) 
1
Y21 = − = − 4 + j8
0.05 + j 0.1
1
Y22 = = 4 − j8
0.05 + j 0.1
Using MATLAB, with V1 (0) = 1.0 ∠ 0 and V2 (0) = 1.0 ∠ 0 :
i 0 1 2 3
V2 1.0 ∠ 0 0.9220 ∠ − 3.7314 ° 0.9111∠ − 3.7314 0.9101 ∠ − 3.7802 °

After 3 iterations,
V2 = 0.9101 ∠ − 3.7802 °

6.33 If we repeat Problem 6.32 with V2 (0) = 1.0 ∠ 30 °, we get V2 = 0.9107 ∠ − 3.8037 ° after 3
iterations.

6.34 V1 = 1.0 ∠ 0
1  P2 − jQ2 
V2 (i + 1) = − Y21V1 − Y23V3 (i)
Y22  V2* (i) 
1 1 − j 0.5 
=  − j 5V1 − j 5V3 (i)
− j10  V2 (i )
*

1  P3 − jQ3 
V3 (i + 1) =  V * (i) − Y31V1 − Y32V2 (i + 1)
Y33 3 
1 1.5 − j 0.75 
=  − j 5V1 − j 5V2 (i + 1)
− j10  V3 (i)
*

Solving using MATLAB with V2 (0) = 1.0 ∠ 0 and V3 (0) = 1.0 ∠ 0 :
i 0 1 2
V2 1.0 ∠ 0 0.9552 ∠ − 6.0090 ° 0.9090 ∠ − 12.6225 °

V3 1.0 ∠ 0 0.9220 ∠ − 12.5288 ° 0.8630 ∠ − 16.1813 °

After 2 iterations,
V2 = 0.9090 ∠ − 12.6225 °
V3 = 0.8630 ∠ − 16.1813 °

153
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.35 Repeating Problem 6.34 with V1 = 1.05 ∠ 0
i 0 1 2
V2 1.0 ∠ 0 0.9801 ∠ − 5.8560 ° 0.9530 ∠ − 11.8861 °

V3 1.0 ∠ 0 0.9586 ∠ − 120426 ° 0.9129 ∠ − 14.9085 °

After 2 iterations,
V2 = 0.9530 ∠ − 11.8861 °
V3 = 0.9129 ∠ − 14.9085 °

1  P2 − jQ2 
6.36 V21 =  0 *
− Y21V1 − Y23V30 − Y24V40 
Y22  (V2 ) 
1  0.5 + j 0.2 
=  − 1.04(−2 + j 6) − (−0.666 + j 2) − (−1 + j 3) 
Y22  1 − j 0 
4.246 − j11.04
= = 1.019 + j 0.046 = 1.02∠2.58° p.u.
3.666 − j11
1  P − jQ 
V22 =  2 2
− Y21V1 − Y23V31 − Y24V41 
 (V21 ) 
*
Y22

Determine V31 and V41 in the same way as V21 . Then

1  0.5 + j 0.2
V22 = − 1.04(−2 + j 6) − (−0.666 + j 2.0)(1.028 − j 0.087)
Y22 1.019 + j 0.046
− (−1 + j 3)(1.025 − j 0.0093)]
4.0862 − j11.6119
= = 1.061 + j 0.0179 = 1.0616∠0.97° p.u.
3.666 − j11.0
6.37 Gauss-Seidel iterative scheme:
With δ 2 = 0;V3 = 1 p.u .; δ 3 = 0 (starting values)
 7 −2 −5 
Given YBus = − j  −2 6 −4  ;
 −5 −4 9 

Also given V1 = 1.0∠0° p.u.

V2 = 1.0 p.u.; P2 = 60 MW
P3 = −80 MW; Q3 = −60 M var (lag)

154
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Iteration 1:
Q2 = − Im(Y22V2 + Y21V1 + Y23V3 )V2*
= − Im {[ − j 6(1∠0) + j 2(1∠0) + j 4(1∠0)]1∠0} = 0
1  P2 − jQ2 
V2 =  − Y21V1 − Y23V3 
Y22  V2* 
1  0.6 + j 0 
=  − j 2(1∠0) − j 4(1∠0)
− j 6  1∠0 
= 1 + j 0.1  1∠5.71°
0.1∠90°
Repeat: V2 = 1 + = 0.99005 + j 0.0995
1∠ − 5.71°
= 0.995∠5.74°
V2 = 1∠5.74° = 0.995 + j 0.1
1  P3 − jQ3 
V3 =  *
− Y31V1 − Y32V2 
Y33  V3 
1  −0.8 + j 0.6 
= − j 5 (1∠0 ) − j 4 ( 0.995 + j 0.1) 
− j 9  1∠0 
0.1111∠53.13°
= 0.9978 + j 0.0444 −
1∠0
= 0.9311 − j 0.0444 = 0.9322∠ − 2.74°

0.1111∠53.13°
Repeat: V3 = 0.9978 + j 0.0444 −
0.9322∠2.74°
= 0.9218 − j 0.0474 = 0.923∠ − 2.94°

Check: ΔV2 = (0.995 + j 0.100) − (1 − j 0) = −0.005 + j 0.1

ΔV3 = (0.9218 + j 0.0474) − (1 − j 0) = −0.0782 − j 0.0474
Δxmax = 0.1

Iteration 2:
Q2 = − Im {[ − j 6(0.995 + j 0.1) + j 2(1∠0)
+ j 4(0.9218 − j 0.0474)] (0.995 − j 0.1)}
= 0.36
1  0.6 − j 0.36 
V2 =  − j 2(1∠0) − j 4(0.9218 − j 0.0474) 
− j 6  1∠ − 5.74° 
0.1166∠59.04
= 0.9479 + j 0.316 + = 1∠4.24 = 0.9973 + j 0.0739
1∠ − 5.74

155
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 0.1166∠59.04
Repeat: V2 = 0.9479 − j 0.0316 + = 1.0003 + j 0.0725
1∠ − 4.24
 1∠4.15 = 0.9974 + j 0.0723
1  −0.8 + j 0.6 
V3 =  − j 5(1∠0) − j 4(0.9974 + j 0.0723)
j 9  0.9230∠2.94 
0.1111∠53.13°
= 0.9988 + j 0.0321 − = 0.9217 − j 0.0604
0.923∠2.94
= 0.9237∠ − 3.75

0.1111∠53.13°
Repeat: V3 = 0.9988 + j 0.0321 − = 0.9205 − j 0.0592
0.9237∠3.75°
= 0.9224∠ − 3.68

Check: ΔV2 = (0.9974 + j 0.0732) − (0.995 + j 0.1) = 0.0024 − j 0.0277

ΔV3 = (0.9205 − j 0.0592) − (0.9218 − j 0.0474) = −0.0013 − j 0.0118
Δxmax = 0.028
Third iteration yields the following results within the desired tolerance:
V2 = 0.9981 + j 0.0610 = 1∠3.5° 
 ←
V3 = 0.9208 − j 0.0644 = 0.923∠ − 4°

6.38

First, convert all values to per-unit.

P − 150 MW
Ppu = = = − 1.5 p.u.
Sbase 100 MVA
Q − 50 MW
Qpu = = = − 0.5 p.u.
Sbase 100 MVA
Δ ymax 0.1 MVA
Δ ymax,pu = = = 1E-4 p.u.
Sbase 100 MVA
Since there are 2 bases, we need to solve 2(n − 1) = 2 equations.
Therefore, J has dimension 2 × 2.

156
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Using Table 6.5
J122 = −V2V21V1 sin (δ 2 − δ1 − θ21 )
J 222 = V2Y22 cos(θ22 ) + Y21V1 cos (δ 2 − δ1 − θ21 ) + Y22V2 cos (−θ22 )
J 322 = V2Y21V1 cos (δ 2 − δ1 − θ21 )
J 422 = −V2Y22 sin θ22 + Y21V1 sin (δ 2 − δ1 − θ21 ) + Y22 + V2 sin (−θ22 )
Also,
− j10 j10  10 ∠ − 90 10 ∠90 
Ybus =  = 
 j10 − j10 10 ∠90 10 ∠ − 90
Finally, using Eqs. (6.6.2) and (6.6.3),
P2 (δ 2 , V2 ) = V2 Y21V1 cos (δ 2 − δ1 − θ21 ) + Y22V2 cos (−θ22 )
Q2 (δ 2 , V2 ) = V2 Y21V1 sin (δ 2 − δ1 − θ21 ) + Y22V2 sin (−θ22 )
Solving using MATLAB with V2 (0) = 1.0 ∠ 0,
i 0 1 2 3
V2 1.0 ∠ 0 0.9500 ∠ − 8,5944 ° 0.9338 ∠ − 9.2319 ° 0.9334 ∠ − 9.2473 °

After 3 iterations
V2 = 0.9334 ∠ − 9.2473 °

6.39 If we repeat Problem 6.38 with V2 (0) = 1.0 ∠ 30 °, Newton-Raphson converges to

0.9334 ∠ − 9.2473 ° after 5 iterations.

6.40 If we repeat Problem 6.38 with V2 (0) = 0.25 ∠0 °, Newton-Raphson converges to

V2 = 0.1694 ∠ − 27422.32 ° = 0.1694 ∠ − 62.32 ° after 14 iterations.

6.41 The number of equations is 2(n − 1) = 4. Therefore, J will have dimension 4 × 4.

Using Table 6.5
10 − 5 0 0
 
− 5 10 0 0
J (0) = 
 0 0 10 − 5
 
 0 0 − 5 10

6.42 First, we convert the mismatch into per-unit 0.1 MVA = 1E-4 p.u. with Sbase = 100 MVA.
Solving using MATLAB with V2 (0) = 1.0 ∠ 0 and V3 (0) = 1.0 ∠ 0,

i 0 1 2 3 4

V2 1.0 ∠ 0 0.8833 ∠ − 13.3690 ° 0.8145 ∠ − 16.4110 ° 0.8019 ∠ − 16.9610 ° 0.8015∠ − 16.9811 °

V3 1.0 ∠ 0 0.8667 ∠ − 15.2789 ° 0.7882 ∠ − 19.2756 ° 0.7731 ∠ − 20.1021 ° 0.7725 ∠ − 20.1361 °

157
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 After 4 iterations, Newton-Raphson converges to
V2 = 0.8015 ∠ − 16.9811 °
V3 = 0.7725 ∠ − 20.1361 °

6.43 (a) Step 1

By inspection:
− j12.5 + j10.0 + j 2.5
 
YBus = + j10.0 − j15.0 + j 5.0
+ j 2.5 + j 5.0 − j 7.5
 
δ 2 (0) = δ 3 (0) = 0° V2 (0) = 1.0
Compute Δ y(0)

P2 ( X ) = 1.0 [10 cos (−90°) + 5 cos ( −90°)] = 0

P3 ( X ) = 2.5 cos (−90°) + 5cos (−90°) = 0
Q2 ( X ) = 1.0 [10 sin (−90°) + 15 + 5 sin (−90°) ] = 0
 P2 − P2 ( X )   −2.0 − 0   −2.0 
Δy(0) =  P3 − P3 ( X )  =  1.0 − 0  =  1.0 
Q2 − Q2 ( X )   −0.5 − 0   −0.5

(b) Step 2 Compute J(0) (see Table 6.5 text)

∂P2
J 122 = = −V2 [Y21V1 sin (δ 2 − δ1 − θ 21 ) + Y23V3 sin (δ 2 − δ 3 − θ 23 )]
∂δ 2
= −1.0 [10(1) sin (−90°) + 5(1) sin ( −90°) ] = 15.
∂P2
J 123 = = V2Y23V3 sin (δ 2 − δ 3 − θ 23 ) = (1.0)(5) sin (−90°) = −5.
∂δ 3
∂P3
J 132 = = V3Y32V2 sin (δ 3 − δ 2 − θ32 ) = (1)(5)(1) sin (−90°) = −5
∂δ 2
∂P3
J 133 = = −V3 [Y31V1 sin (δ 3 − δ1 − θ31 ) + Y32V2 sin (δ 3 − δ 2 − θ32 ) ]
∂δ 3
= −1.0 ( 2.5 ) (1) sin ( −90°) + (5)(1) sin ( −90°)  = 7.5
∂P2
J 222 = = V2Y22 cos (θ 22 ) + [Y21V1 cos (δ 2 − δ1 − θ 21 ) + Y22V2 cos (−θ 22 )
∂V2
+ Y23V3 cos (δ 2 − δ 3 − θ 23 )] = 0

158
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 ∂P3
J 232 = = V3Y32 cos (δ 3 − δ 2 − θ32 ) = 0
∂V2
∂Q2
J 322 = = V2 [Y21V1 cos (δ 2 − δ1 − θ 21 )] + Y23V3 cos (δ 2 − δ 3 − θ 23 ) = 0
∂δ 2
∂Q2
J 323 = = −V2Y23 V3 cos (δ 2 − δ 3 − θ 23 ) = 0
∂δ 3

∂Q2 Y21V1 sin (δ 2 − δ1 − θ 21 ) + Y22V2 sin (−θ 22 ) 

J 422 = = −V2Y22 sin θ 22 +  
∂V2  +Y23V3 sin (δ 2 − δ 3 − θ 23 ) 
J 422 = (−1)(15) sin (−90°) + [(10)(1) sin (−90°) + 15(1) sin (90°) + 5(1) sin ( −90°) ]
= 15
15 −5 0 
 J 1 J 2  
J (0) =   =  −5 7.5 0  per unit
 J 3 J 4   0 0 15

Step 3 Solve J Δx = Δy
15 −5 0   Δδ 2   −2.0 
 −5 7.5 0   Δδ  =  1.0 
  3  
 0 0 15  ΔV2   −0.5 

Using Gauss elimination, multiply the first equation by (−5/15) and subtract from the second
equation:
15 −5 0   Δδ 2   −2.0 
 0 5.833333 0   Δδ  =  0.33333
  3  
 0 0 15  ΔV2   −0.5 

Back substitution:
ΔV2 = −0.5/15 = −0.033333
Δδ 3 = 0.33333/5.833333 = 0.05714285
Δδ 2 = [ −2.0 + 5(0.05714285) ] 15 = −0.1142857
 Δδ 2   −0.1142857 
Δx =  Δδ 3  =  0.05714285 
 ΔV2   −0.033333 

Step 4 Compute x(1)

δ 2 (1)   0   −.1142857   −0.1142857  radians
x (1) = δ 3 (1)  = x (0) + Δx =  0  +  .05714285  =  0.05714285  radians
 
V2 (1)  1   −.0333333  0.96666667  per unit

Check QG3 using Eq. (6.5.3)

159
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Q3 = V3 [Y31V1 sin (δ 3 − δ1 − θ 31 ) + Y32V2 sin (δ 3 − δ 2 − θ32 ) + Y33V3 sin (−θ33 ) ]
  π  π
= 1  (2.5)(1) sin  0.05714
  − 2  + 5(0.966666) sin  0.05714 + 0.11429 − 2 
  radians   
 π 
+7.5 (1) sin  
 2 
Q3 = 1[ −2.4959 − 4.7625 + 7.5] = 0.2416 per unit
Q G3 = Q3 + QL 3 = 0.2416 + 0 = 0.2416 per unit

Since QG3 = 0.2416 is within the limits [−5.0, +5.0], bus 3 remains a voltage-controlled bus.
This completes the first Newton-Raphson iteration.

6.44 Repeating Problem 6.43 with PG 2 = 1.0 p.u.

(a) Step 1
Ybus stays the same.
P2 ( x(0)) , P3 ( x(0)) , Q2 ( x(0)) also stay the same.
P2 − P2 ( x(0))  − 2.0 + 1.0 − 0 − 1.0
     
Δ y(0) = P3 − P3 ( x(0))  =  1.0 − 0  =  1.0
    − 0.5
Q2 − Q2 ( x(0)) − 0.5 − 0
(b) Step 2
The Jacobian stays the same.
15 − 5 0 
 
J (0) = − 5 7.5 0 
 0 0 15
(c) Step 3
Solve J Δ x = Δ y
15 − 5 0 Δ δ 2  − 1.0
    
− 5 7.5 0 Δ δ 3  =  1.0
 0 0 15 Δ V2  − 0.5
−5
Multiply the first equation by and substract from the second equation.
15
15 − 5 0 Δ δ 2  − 1.0
    
 0 35 6 0 Δ δ 3  =  2 3
 0 0 15 Δ V2  − 0.5

160
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Via back substitution
− 0.5
Δ V2 = = − 0.033333
15
23
Δδ3 = = 0.11429
35 6
Δ δ 2 = (− 1.0 + 5 × 0.11429) 15 = − 0.028571
(d) Step 4
0 − 0.028571 − 0.028571
    
x(1) = x(0) + Δ x = 0 +  0.11429  =  0.11429 
1 − 0.033333  0.96667 
Check QG3 using Eq. (6.5.3)
Q3 = V3 Y31V1 sin (δ 3 − δ1 − θ31) + Y32V2 sin (δ 3 − δ 2 − θ32 ) + Y33V3 sin (−θ33 )
  π  π
= 1.0 2.5 × 1.0 sin 0.11429 −  + 5 × 0.96667 sin  0.11429 + 0.028571 − 
  2  2
 π 
+ 7.5 × 1.0 sin  
 2 
= 0.2322 p.u.
QG3 = Q3 + QL3 = 0.2322 + 0 = 0.2322 p.u.
Since QG3 is within the limits [− 5.0, 5.0] , bus 3 remains a voltage controlled bus.
∴ After the first iteration,
δ 2  − 0.028571
   
x = δ 3  =  0.11429 
V2   0.96667 

6.45 After the first three iterations J22 = 104.41, 108.07, 107.24; and with the next iteration it
converges to 106.66.

6.46 First, convert all values to per-unit.

PG 2 = 80 MW = 0.8 p.u.
PL3 = 180 MW = 1.8 p.u.
For convergence, Δ ymax = 0.1 MVA = 1E-4 p.u.
Second, find Ybus .

8 − j16 − 4 + j8 − 4 + j8
 
Ybus = − 4 + j8 8 − j16 − 4 + j8
− 4 + j8 − 4 + j8 8 − j16 

161
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 To solve the case, we follow the steps outlined in section 6.6.
Iteration 0
Step 1
δ 2 (0) 0
 
Start with x(0) = δ 3 (0) = 0
V3 (0) 1

P ( x(0)) 0
P ( x(0)) =  2 = 
P3 ( x(0)) 0
Q ( x(0)) = Q3 ( x(0)) = 0
 P2 − P2 ( x(0))   0.8 
   
Δ y(0) =  P3 − P3 ( x(0))  = − 1.8
Q3 − Q3 ( x(0))  0 
 
Step 2
16 − 8 − 4
J1 =   J2 =  
− 8 16 8
J 3 = [4 − 8] J 4 = [16]
16 − 8 − 4
 
J (0) = − 8 16 8 
 4 − 8 16
 
Step 3
J (0)Δ x(0) = Δ y(0)
16 − 8 − 4 Δ δ 2 (0)  0.8
    
− 8 16 8 Δ δ 3 (0) = − 1.8
 4 − 8 16 Δ V3 (0)  0 
− 0.0083333
 
Δ x(0) = − 0.094167 
− 0.045 
Step 4
− 0.0083333
 
x(1) = x(0) + Δ x(0) = − 0.094167 
 0.955 

162
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Iteration 1
Step 1
 0.7825
P ( x(1)) =  
− 1.6861
Q ( x(1)) = 0.0610
 0.017502
 
Δ y(1) = − 0.11385 
− 0.06104 
Step 2
15.9057 − 7.93935 − 3.29945
 
J (1) = − 7.28439 14.5314 5.8744 
 4.4609 − 8.98235 15.3439 
Step 3
− 0.0038007
 
Δ x(1) = − 0.0069372
− 0.0069342
Step 4
− 0.012134
 
x(2) = − 0.1011 
 0.94807 
Iteration 2
Step 1
 0.7999
P ( x(2)) =  
− 1.7990
Q ( x(2)) = 6.484 E-4
 1.3602 E-4
 
Δ y(2) = − 1.0458 E-4
− 6.484 E-4 
Step 2
15.8424 − 7.89148 − 3.27336
J (2) = − 7.21758 14.3806 5.68703
 4.45117 − 8.98958 15.1697 
Step 3
− 3.7748 E-5
Δ x (2) = − 6.412E-5 
− 6.9664 E-5

163
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Step 4
− 0.012172
 
x(3) = − 0.10117 
 0.948 
Iteration 3
Step 1
 0.8000
P ( x(3)) =  
− 1.8000
Q ( x(3)) = 6.4875 E-8
 1.1294 E-8 
 
Δ y(3) = − 9.7437 E-8
− 6.4875 E-8
Δ y(3) < 1E-4
∴ Newton-Raphson has converged in 3 iterations to
V1 = 1.0 ∠ 0
V2 = 1.0 ∠ − 0.012172 rad = 1.0 ∠ − 0.6974°
V3 = 0.948 ∠ − 0.10117 rad = 0.948 ∠ − 5.797°
Using the voltages, we can calculate the powers at buses 1 and 2 with Eqs. (6.6.2) and
(6.6.3).
P1 = 1.0910 p.u. = 109.10 MW
Q1 = 0.0237 p.u. = 2.37 Mvar
Q2 = 0.1583 p.u. = 15.83 Mvar
Using PowerWorld simulator, we get
P1 = 109.1 MW
Q1 = 2.4 Mvar
Q2 = 15.8 Mvar
V3 = 0.948 p.u.
which agrees with our solution.

6.47 First, convert values to per-unit

Q2 = 50 Mvar = 0.5 p.u.
Ybus stays the same as in Problem 6.46.

164
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Since bus 2 is now a PQ bus, we introduce a fourth equation in order to solve for V2 .
Iteration 0
Step 1
δ 2 (0) 0

 δ 3 (0) 0
Start with x(0) = =
V2 (0) 1
  
V3 (0) 1
P ( x(0)) 0
P ( x(0)) =  2 = 
P3 ( x(0)) 0
Q ( x(0)) 0
Q ( x(0)) =  2 = 
Q3 ( x(0)) 0
 P2 − P2 ( x(0))   0.8
 
 P3 − P3 ( x(0))  − 1.8
Δ y(0) = =
Q2 − Q2 ( x(0)   0.5
   
Q3 − Q3 ( x(0))  0
Step 2
16 −8 8 − 4
 
−8 16 − 4 8 
J (0) = 
− 8 4 16 − 8
 
4 − 8 − 8 16 
Step 3
J (0)Δ x(0) = Δ y(0)
− 0.023333
 
− 0.10167 
Δ x(0) = 
 1.03 
 
 0.97 
Step 4
− 0.023333
 
− 0.10167 
x(1) = x(0) + Δ x(0) = 
 1.03 
 
 0.97 

165
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Iteration 1
Step 1
 0.8174
P ( x(1)) =  
− 1.7299
0.5517
Q ( x(1)) =  
0.0727
− 0.01739 
 
 − 0.070052
Δ y(1) =
− 0.051746
 
− 0.072698
Step 2
16.4227 − 8.28102 9.03358 − 3.46256
 
− 7.65556 14.9817 − 4.47535 5.97655
J (1) = 
− 7.66981 3.35868 17.0157 − 8.53714
 
 4.60961 − 9.25715 − 7.43258 15.5949 
Step 3
− 0.00057587
 
− 0.0032484 
Δ x(1) = 
− 0.0077284 
 
− 0.010103 
Step 4
− 0.023904
 
− 0.10492 
x(2) = 
 1.0223 
 
 0.9599 
Iteration 2
Step 1
 0.7999
P ( x(2)) =  
− 1.7989
0.5005
Q ( x(2)) =  
0.0007
 0.00013185
 
− 0.0011097 
Δ y(2) = 
− 0.00048034
 
− 0.00072284

166
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Step 2
 16.2201 − 8.14207 8.96061 − 3.41392
 
 − 7.50685 14.7417 − 4.44838 5.80513 
J (2) =
− 7.56045 3.27701 16.8459 − 8.48223
 
 4.54745 − 9.17011 − 7.34331 15.3591 
Step 3
 − 6.3502 E-7 
 
− 5.3289E-5 
Δ x(2) = 
 − 7.6468E-5 
 
− 0.00011525
Step 4
− 0.02391
 
− 0.10497
x(3) = 
 1.0222 
 
 0.95978 
Iteration 3
Step 1
 0.8000 
P ( x(3)) =  
− 1.8000
0.5000
Q ( x(3)) =  
0.0000
 6.1267 E-8 
 
− 2.0161E-7 
Δ y(3) = 
− 6.5136 E-8
 
 − 8.9553E-8 
Δ y(3) < 1 E-4
∴ Newton-Raphson has converged in 3 iterations to
V1 = 1.0 ∠ 0
V2 = 1.0222 ∠ − 0.02391 rad =1.0222 ∠ − 1.370 °
V3 = 0.95978 ∠ − 0.10497 rad = 0.95978 ∠ − 6.014°
Using the voltages, we can calculate the powers at bus 1 with Eqs. (6.6.2) and (6.6.3).
P1 = 1.0944 p.u. = 109.44 MW
Q1 = − 0.3112 p.u. = − 31.12 Mvar

167
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
 Using PowerWorld Simulator, we get
P1 = 109.4 MW
Q1 = − 31.1 Mvar
V2 = 1.025 p.u.
V3 = 0.960 p.u.
Which agrees with our solution.

6.48

6.49
Tap setting Mvar @ G1 V5 (p.u.) V2 (p.u.) P losses
0.975 94 0.954 0.806 37.64
0.98125 90 0.961 0.817 36.63
0.9875 98 0.965 0.823 36.00
0.99375 106 0.97 0.828 35.4
1.0 114 0.974 0.834 34.84
1.00625 123 0.979 0.839 34.31
1.0125 131 0.983 0.845 33.81
1.01875 140 0.987 0.850 33.33
1.025 149 0.992 0.855 32.89
1.03125 158 0.996 0.86 32.47
1.0375 167 1.0 0.865 32.08
1.04375 176 1.004 0.87 31.72
1.05 185 1.008 0.874 31.38
1.05625 195 1.012 0.879 31.06
1.0625 204 1.016 0.884 30.76
1.06875 214 1.02 0.888 30.49
1.075 224 1.024 0.893 30.23
1.08125 233 1.028 0.897 30.00
1.0875 243 1.031 0.901 29.79
1.09375 253 1.035 0.906 29.59
1.1 263 1.039 0.910 29.42

168
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.50 Shunt Capacitor Shunt Capacitor Shunt Capacitor V2 Line Line Line Total Power
State Rating (Mvar) Output (Mvar) 2-4 2-5 4-5 Losses (MW)
(p.u.)
Disconnected 0 0 0.834 27% 49% 19% 34.37
Connected 200 184.0 0.959 25% 44% 16% 25.37
Connected 261 261.0 1.000 25% 43% 15% 23.95
When the capacitor rating is set at 261 Mvar, the voltage at bus 2 becomes 1.0 per unit. As a result of this, the
line loadings decrease on all lines, and thus the power loss decrease as well.

6.51 Before new line After new line

Bus voltage V2 (p.u.) 0.834 0.953
Total real power losses (MW) 34.8 18.3
Branch b/w bus 1–5 (% loading) 68.5 63.1
Branch b/w bus 2–4 (% loading) 27.3 17.5
49.0 25.4
Branch b/w bus 2–5 (% loading) (both lines)
Branch b/w bus 3–4 (% loading) 53.1 45.7
Branch b/w bus 4–5 (% loading) 18.8 22.1

6.52 With the line connecting HOMER 69 and LAUF 69 removed, the voltage at HANNAH 69
drops to 0.966 p.u. To raise that voltage to 1.0 p.u., we require 32.0 Mvar from the capacitor
bank of HANNAH 69.

6.53

When the generation at BLT138 is 300 MW, the losses on the system are minimized to
10.24 MW.

169
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.54

When the generation at BLT138 is 220 MW or 240 MW, the losses on the system are
minimized to 11.21 MW.

6.55 DIAG = [17 25 9 2 14 15]

OFFDIAG = [−9.1 −2.1 −7.1 −9.1 −8.1 −1.1 −6.1 −8.1 −1.1 −2.1 −6.1 −5.1 −7.1 −5.1]
C0L =[ 2 5 6 1 3 4 5 2 2 1 2 6 1 5]
ROW =[ 3 4 1 1 3 2]
6.56 By the process of node elimination and active branch designation, in Fig. 6.9:

The fill-in (dashed) branch after Step 6 is shown below:

Note that two fill-ins are unavoidable. When the bus numbers are assigned to Fig. 6.9 in
accordance with the step numbers above, the rows and columns of YBUS will be optimally
ordered for Gaussian elimination, and as a result, the triangular factors L and U will
require minimum storage and computing time for solving the nodal equations.

170
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.57 Table 6.6 w/DC Approximation
Generation Load
Bus # Voltage Phase Angle PG QG PL QL
Magnitude (degrees) (per unit) (per unit) (per unit) (per unit)
(per unit)
1 1.000 0.000 3.600 0.000 0.000 0.000
2 1.000 −18.695 0.000 0.000 8.000 0.000
3 1.000 0.524 5.200 0.000 0.800 0.000
4 1.000 −1.997 0.000 0.000 0.000 0.000
5 1.000 −4.125 0.000 0.000 0.000 0.000
TOTAL 8.800 0.000 8.800 0.000

When comparing these results to Table 6.6 in the book, voltage magnitudes are all constant.
Most phase angles are close to the NR algorithm except bus 3 has a positive angle in DC and
a negative value in NR. Total generation is less since losses are not taken into account and
reactive power is completely ignored in DC power flow.
Table 6.7 w/ DC Approximation

Line # Bus to Bus P Q S

1 2 4 −2.914 0.000 2.914
4 2 2.914 0.000 2.914
2 2 5 −5.086 0.000 5.086
5 2 5.086 0.000 5.086
3 4 5 1.486 0.000 1.486
5 4 −1.486 0.000 1.486

With the DC power flow, all reactive power flows are ignored. Real power flows are close
to the NR algorithm except losses are not taken into account, so each end of the line has the
same flow.

Table 6.8 w/DC Approximation

Tran. # Bus to Bus P Q S

1 1 5 3.600 0.000 3.600
5 1 −3.600 0.000 3.600
2 3 4 4.400 0.000 4.400
4 3 −4.400 0.000 4.400

With DC approximation the reactive power flows in transformers are also ignored and losses
are also assumed to be zero.

171
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.58 With the reactance on the line from bus 2 to bus 5 changed, the B matrix becomes:
− 43.333 0 10 33.333
 
0 − 100 100 0
B= 
 10 100 − 150 40 
 
 33.333 0 40 − 123.333
P remains the same:
− 8.0
 
4.4 
P=
 0 
 
 0 
− 0.2443 − 13.995° 
   
 0.0255  1.4638°
δ = −B P =
−1
rad =
− 0.0185  − 1.0572°
   
− 0.0720  − 4.1253°

6.59 (a) Since bus 7 is the slack bus, we neglect row 7 and column 7 in our B matrix and P
vector.
− 20.833 16.667 4.167 0 0 0 
 
 16.667 − 52.778 5.556 5.556 8.333 16.667
 4.167 5.556 − 43.056 33.333 0 0 
B= 
 0 5.556 33.333 − 43.056 4.167 0 
 0 8.333 0 4.167 − 29.167 0 
 
 0 16.667 0 0 0 − 25 
 160  1.6
   
 110  1.1
− 110 − 1.1
P=  MW=   p.u.
− 30  − 0.3
− 130 − 1.3
   
 0   0 
(b) δ = − B−1 P
 7.6758 °
 
 4.2020 °
− 0.4315 °
= 
− 0.3265 °
− 1.3999 °
 
 2.8014 °
This agrees with the angles shown in PowerWorld.

172
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
6.60 First, we find the angle of bus 1 required to give a 65 MW line loading to the line between
buses 1 and 3.
δ1 − δ 3
P13 =
x13
δ1 − δ 3
0.65 =
0.24
δ1 = δ 3 + 0.156
Now, we solve for the angles:
P = − Bδ
1.6 + Δ P1   δ 3 + 0.156
 1.1    δ2 
    
 − 1.1   B  δ3 
  = −  
 − 0.3    δ4 
 − 1.3    δ5 
    
 0     δ6 
This is a system of 6 unknowns (δ 2 , δ 3 δ 6 , Δ P1 ) and 6 equations. Solving yields
Δ P1 = 0.3194 p.u.
=31.94 MW
Therefore, if the generator at bus 1 increases output by 31.94 MW, the flow on the line
between buses 1 and 3 will reach 100%.

6.61 A type 1 or 2 wind turbine is one which maintains a constant real and reactive power output.
Hence, the bus it is connected to will be a PQ bus. A type 3 or 4 wind turbine is one which
maintains a constant voltage and real power output; the bus it is connected to will be a PV
bus. In a contingency such as a line outage, the voltage in the PQ bus will change due to the
redistribution of power flows on the remaining lines. However, the voltage on a PV bus
remains constant in a contingency, due to the fact that the type 3 or 4 wind turbine changes
its reactive power output to maintain that voltage. In this regard, a type 3 or 4 wind turbine
is more favorable.
From Eq. (6.6.3), we can see that the reactive power injection at a bus is directly
proportional to the voltage magnitude at that bus.

173
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.