Network Calculations PDF

NETWORK
CALCULATIONS

updated 11/5/13 1:02 PM
11/5/13
Network Calcula2ons (c) 2013 H. Zmuda
Introductory Comments

The typical power transmission network span a large geographic
area and involve a large number and variety of network
components.

The electrical characteris2cs of the individual components have
been developed and now we are concerned with the composite
representa2on of those components when they are interconnected
to form the network.

For large-scale system analysis the network model takes on the
form of a network matrix. with elements determined by the choice
of parameter.
11/5/13

There are two choices. The current ow through a network
component can be related to the voltage drop across it by either an
admiKance or an impedance parameter.

We now examine the admiKance representa2on in the form of a
primi2ve model which describes the electrical characteris2cs of the
network components. The primi2ve model neither requires nor
provides any informa2on about how the components are
interconnected to form the network.

The steady-state behavior of all the components ac2ng together as
a system is given by the nodal admiKance matrix based on nodal
analysis of the network equa2ons.
11/5/13

The nodal admiKance matrix of the typical power system is large
and sparse, and can be constructed in a systema2c building-block
manner.

The building-block approach provides insight for developing
algorithms to account for network changes. Because the network
matrices are very large, sparsity techniques (not covered in class)
are needed to enhance the computa2onal eciency of computer
programs employed to solve power system problems.
11/5/13
Branch and Node AdmiHances

Assume per-phase analysis, and steady-state opera2on.

Also recall source transforma2ons:
I
I

+
+

Ya = Z a1
Za
E

Network
V Network
+
V
Is = s
Za

Es ~
_

The branch impedance/admiKance is oTen called the primi2ve
impedance/admiKance.
11/5/13
RECALL NODAL ANALYSIS via example

2
Yb
Ye
Yd
3
Yf
Yc
I3
Yg
Ya
I4
0
Reference
11/5/13
RECALL NODAL ANALYSIS

KCL at node 3: I 3 = YaV3 + Yc V3 V1 + Yb V3 V2

= YcV1 YbV2 + Ya + Yb + Yc V3

KCL at node 1:
0 = Yc (V1 V3 ) + Yd (V1 V2 ) + Y f (V1 V4 )
= Yc + Yd + Y f V1 YdV2 YcV3 Y f V4
11/5/13
NODAL ADMITTANCE MATRIX YBus

Diagonal Elements:

Yii = Sum of the admiKances incident on bus (node) i

O-Diagonal Elements:

Yij = [Sum of the admiKances shared by busses i and j]

YBus is symmetric: Yij = Yji
11/5/13
BUS ADMITTANCE MATRIX Ybus

Y +Y +Y
Yd
Yc
Y f

c
d
f
Yd
Yb + Yd + Ye
Yb
Ye
Y =
Bus
Yc
Yb
Ya + Yb + Yc
0

Y f
Ye
0
Ye + Y f + Yg

BUS ADMITTANCE MATRIX Zbus
Z
Z
Z
Z

12
13
14
11
1
Bus
Z Bus = Y
11/5/13
= YBus
Z12
=
Z13
Z
14
Z 22
Z 23
Z 23
Z33
Z 24
Z34
Z 24
Z34
Z 44
AN ALTERNATIVE APPROACH FOR Ybus FORMATION

Suppose that only branch admiKance Ya is connected between
nodes m and n as part of a larger network of which only the
reference node appears is shown.
Ia
Va
Im
In
+
Vm
Ya
+
Vn
Reference Node (Bus)
11/5/13
10

Current injected into the network at any node is considered posi2ve
and current leaving the network at any node is considered nega2ve.

In the gure, current Im is that por2on of the total current injected
into node m which passes through Ya. Similarly, current In is that
por2on of the total current injected into node n which passes
through Ya. The voltages Vm and Vn are measured with respect to
the network reference.

Clearly at node a, Im = Ia, and at node b In = Ia thus
I
m
In
11/5/13
1
=
Ia
n 1
11

Similarly, Va = Vm Vn, or

m n Vm

Va = 1 1

Vn
But YaVa = Ia

V
m
YaVa = Ya 1 1
Vn
m n
Premul2ply by [1 -1]T
m

n
11/5/13
m n V
1
m
Ya 1 1
Vn
1
= Ia
1
m
=
Ia =
1
In
12

Simplifying:
V
1
m
Ya
1 1
Vn
1
1 1
=
1 1
m

n
11/5/13
m n
Ya
Ya
Ya Vm
Ya Vn
I
= m
In

I
= m
In

13

If one of the two nodes were the reference (say node n), then Vn = 0
and the result reduces to:
m

n
m n
Ya Vm I m
=
Ya Ya 0 I n
m
YV I
a m
= m m Ya Vm = I m
YaVm I n
11/5/13
Ya
14
EXAMPLE

The single-line diagram of a small power system (shown on next
slide) has a generator with voltage equal to 1.25 per unit connected
through a transformer to high-voltage node 3 while a motor with
internal voltage equal to 0.85 at an angle of 45o is connected to
node 4. The reactances for each branch (specied per unit) are
shown on the gure.

Develop the nodal admiKance matrix for each of the network

branches then write the nodal admiKance equa2ons of the system.
11/5/13
15
EXAMPLE

j0.25
j0.2
j0.125
j0.25
3
j1.15
j0.1
j0.4
j0.1
j1.15
1.250
11/5/13
0.85 45
16
REACTANCES

j0.25
j0.125
j0.25
j0.4
j0.1
j1.15
j0.1
j1.15
+
1.250
11/5/13
j0.2
+
0.85 45
17
REACTANCES

j0.25
j0.125
j0.25
j0.2
j0.4
j0.1
j0.1
Combine
j1.15
j1.15
+
1.250
11/5/13
+
0.85 45
18
REACTANCES

j0.25
j1.25
j0.125
j0.25
j0.4
j1.25
+
1.250
11/5/13
j0.2
+
0.85 45
19
SUSCEPTANCES

j4.0
1.0 90
j8.0
j4.0
j5.0
j2.5
j0.8
j0.8
0
11/5/13
+
0.68 135
20
SUSCEPTANCES

j4.0
Yb
Yc
1.0 90
j4.0
j8.0
Yf
Yd
Ye
j5.0
j2.5
Yg
j0.8
Ya
j0.8
0
11/5/13
+
0.68 135
21
Branch Matrices
3 2
3 1
2 1 1
Yb
j4.0
j4.0
3 1
1 1 1
11/5/13
4 1
1 1 1
1 Ya
j0.8
4
0
2 1 1
2
Yd
Called
PrimiSve
AdmiHance
Matrices
Yf
1 Yg
j0.8
Ye
1 1 1
4 1
2 1
j2.5
Yc
Ye
j5.0
j8.0
Yf
Yd
Yc
1.0 90
+
0.68 135
22
Nodal Matrix YBus

Column

2
3
3
1 1
Row
Y
2 1 1 b
1 Ya
11/5/13
1

2

3

4
Y +Y +Y
f
c d
Yd
Yc
Y f
1 1
Y
1 1 1 f

4
1 1
Y
2 1 1 e
Yd
1 1 1
3
1 1
Y
1 1 1 c
1
YBus =
2 1
Yd
Yc
Yb + Yd + Ye
Yb
Yb
Ya + Yb + Yc
Ye
1 Yg

Ye
Ye + Y f + Yg
Y f
23
Nodal Matrix YBus

YBus
j4.0 j8.0
j8.0
=
j4.0
j2.5
14.5 8.0
8.0 17
= j
4.0
4.0
5.0
2.5
11/5/13
j4.0 j8.0 j5.0

j4.0
j5.0
j4.0
j0.8 j4.0 j4.0
0
j5.0
0
j5.0 j2.5 j0.8
4.0 2.5
4.0 5.0
8.8
0
0
8.3
j2.5
j8.0
j4.0
j2.5
24
Nodal Matrix YBus

The numerical ordering of the rows and columns doesnt maKer so
long as they are the same.

These are ok:
2
3
3
2

3
2
1 1
1 1
Y
Yb

Row
Row
b
2 1 1
3 1 1

This is not ok:
2
3

2
1 1
Row
Y

3 1 1 b

As a conven2on, we assign node numbers in the direc2on of the
branch currents
11/5/13
25
Mutually Coupled Branches

Consider the case where two branches are mutually coupled to
each other but not induc2vely coupled to other branches.
m
Im
+
Va Z a
In
Ia
ZM
Ib
Zb
+
Vb
Ip
Iq
Va = Z a I a + Z M I b
Vb = Z M I a + Z b I b
11/5/13
26

V Z
a = a
Vb Z M
I
a
Ib
11/5/13
Z M Ia
Zb I b
Z
Z M
a
Z M
Zb
=
Z a Z b Z M2
V Y
a = a
Vb YM
YM Va
Yb Vb
27

m
Im
V
a =
Vb
11/5/13
ZM
+
Va Z a
In
Ia
Ib
Zb
+
Vb
Ip
Iq
Vm Vn 1 1 0 0
=
V p Vq 0 0 1 1
V
Vm
m
Vn
Vn
= A
Vp
Vp
Vq
Vq
28

m
Im
11/5/13
ZM
+
Va Z a
In
Ia
Ib
Zb
+
Vb
Im
In
=
Ip

Iq
1 0
I
1 0 a
0 1 Ib
0 1
Ip
Iq
I
= AT a
Ib
29

Subs2tute for V:
V
m

I a Ya YM Va Ya YM Vn
=
=
A

I b YM Yb Vb YM Yb V p

Vq
Premul2ply by AT:
V I
m
m
I
a
A
Ib
11/5/13
Y
T
= A a
YM
YM
Yb
Vn
A
Vp

Vq

In
Ip

I q
T
= A a
YM
YM
Yb
Vm
Vn
Vp
Vq
30

Im
Y
In
= AT a
Ip
YM
Iq
Im
1 0
I n 1 0
=
Ip 0 1
0 1
Iq
11/5/13
YM
Yb
Y
a
YM
Vm
Vn
Vp
Vq
YM
Yb
1
1
0
0
0 0 1 1
Vm
Vn
Vp
Vq
31

m n p q
11/5/13
I m m

I n n
=
I p p

I q q
Im
In
=
Ip
Iq
Ya
Ya
YM
Ya
Ya
YM
YM
YM
Yb
YM
YM
Yb
m

n

p
q

YM
YM
Yb
Yb
Vm
Vn
Vp
Vq
m n p q
1 1
1 1
1 1
1 1
Ya
YM
1 1
1 1
1 1
1 1
YM
Yb
Vm
Vn
Vp
Vq
32

The nodal admiKance matrix of the two coupled branches may be
formed directly by inspec2on as shown. If one of the nodes is the
reference node, that corresponding row and column is simply
eliminated.
Im
In
=
Ip
Iq
11/5/13
m

n

p
q

m n p q
1 1
1 1
1 1
1 1
Ya
YM
1 1
1 1
1 1
1 1
YM
Yb
Vm
Vn
Vp
Vq
33
Example:

Two coupled branches have the impedances shown. Find the nodal
admiKance matrix.
Z a = j0.25
3
Z M = j0.15
Z b = j0.25
11/5/13
Ya = j6.25
YM = j3.75
Yb = j6.25
34
Example:

j0.25
Z=
j0.15
I3
I1
=
I3
I 2
11/5/13
j6.25 j3.75
j0.15
=
j0.25
j3.75 j6.25
3 1 3 2
1 1
1
1

( j6.25)
( j3.75)
1 1 1
1 1

1 1
3 1 1
( j3.75)
( j6.25)

1 1
2 1 1
V3
V1
V3
V2
35
Example:

But since there are only three nodes, only a three-by-three matrix is
required.

Since currents m and n add to form a common injected current (I3 in
this case), the common columns and rows of the common node
(node 3) are added. Do it once to see it.
11/5/13
Im
In
=
Ip
Iq
Ya
Ya
YM
Ya
Ya
YM
YM
YM
Yb
YM
YM
Yb
YM
YM
Yb
Yb
Vm
Vn
Vm
Vq
36
Example:

I
Y
V
Y
Y
Y
m
a
M
M
a
m
In
Ya
Ya YM YM Vn
Yb
Yb Vm
Ip
YM YM
Yb
Yb q
I q
YM YM
I m = YaVm YaVn + YMVm YMVq = (Ya + YM )Vm YaVn YMVq
I n = YaVm + YaVn YMVm + YMVq = ( Ya YM )Vm + YaVn + YMVq

I p = YMVm YMVn + YbVm YbVq = (YM + Yb )Vm YMVn YbVq
I q = YMVm + YMVn YbVm + YbVq = ( YM Yb )Vm + YMVn + YbVq

11/5/13
37
Example:

But
I3 = I m + I p
Im
In
ZM
+
Va Z a
Ya = j6.25
Zb
+
Vb
Ip
YM = j3.75
Yb = j6.25
Iq
I m = (Ya + YM )Vm YaVn YMVq
I n = ( Ya YM )Vm + YaVn + YMVq

I p = (YM + Yb )Vm YMVn YbVq
I q = ( YM Yb )Vm + YMVn + YbVq

11/5/13
38
Example:

I 3 = I m + I p = Ya + 2YM + Yb Vm + YM Ya Vn + Yb YM Vq

I n = Ya YM Vm + YaVn + YMVq

I q = YM Yb Vm + YMVn + YbVq

All one need do is add the corresponding common equa2ons and
gather like terms.

(
(
(
)
)
I 3 = I m + I p = (Ya + 2YM + Yb )V3 + ( YM Ya )V1 + ( Yb YM )V2

I1 = ( Ya YM )V3 + YaV1 + YMV2
I 2 = ( YM Yb )V3 + YMV1 + YbV2
11/5/13
39
Example:

I 1 = YaV1 + YMV2 + ( Ya YM )V3
I 2 = YMV1 + YbV2 + ( YM Yb )V3
I 3 = ( YM Ya )V1 + ( Yb YM )V2 + (Ya + 2YM + Yb )V3

I 1 = j6.25V1 + j3.75V2 + ( j6.25 j3.75)V3
I 2 = j3.75V1 j6.25V2 + ( j3.75 + j6.25)V3
I 3 = ( j3.75 + j6.25)V1 + ( j6.25 j3.75)V2 + ( j6.25 + 2 j3.75 j6.25)V3

j6.25 j3.75 j2.5
j6.25
j3.75
j6.25 j3.75

Y =
j3.75
j6.25
j3.75 + j6.25
= j3.75 j6.25 j2.5
j3.75 + j6.25 j6.25 j3.75 j6.25 + 2 j3.75 j6.25 j2.5
j2.5
j5

Remember, for passive networks, the Ybus or Zbus matrices MUST
ALWAYS be reciprocal.
11/5/13
40
An Equivalent Network

We can now eliminate the mutual coupling; oTen a useful feature
for large networks.

Recall our result from Slide 33:

Im
In
=
Ip
Iq
11/5/13
m

n

p
q

m n p q
1 1
1 1
1 1
1 1
Ya
YM
1 1
1 1
1 1
1 1
YM
Yb
Vm
Vn
Vp
Vq
41

Consider the equa2on for Im, I m = YaVm YaVn + YMV p YMVq

Add and subtract YMVm on the right,

I m = YaVm YaVn + YMV p YMVq

= Ya Vm Vn + YMV p YMVm YMVq + YMVm

= Ya Vm Vn + YM Vm V p + YM Vm Vq

But this is just a KCL at node m, i.e.,
(
(
11/5/13
)
) (
)(
42

KCL at node m: I m = Ya Vm Vn + YM Vm V p + YM Vm Vq
(
)
(
)
(
)

I mn
Im
+ I mn
Va Z a
In
11/5/13
Ia
I mp
I mp
Ib
ZM
I mq
Zb
(
)
I mq
+
Vb
Ip
Iq
43

I m = Ya (Vm Vn ) + YM Vm V p + YM Vm Vq

I mn
I mp
I mq
Im
YM
m
I mp
Ya
n
11/5/13
YM
I mn
I mq
q
44

Similarly for the second and third equa2on:
I n = YaVm + YaVn YMV p + YMVq
= Ya (Vn Vm ) + ( YM ) Vn Vq + YM Vn V p

I nm
I nq
I np
I p = YMVm YMVn + YbV p YbVq
= Yb V p Vq + ( YM ) V p Vm + YM V p Vn

I pq
I pm
I pn
11/5/13
45

I n = Ya (Vn Vm ) + ( YM ) Vn Vq + YM Vn V p

I nm
I nq
I np
Ya
In
I nm
YM
I np
I nq
YM
11/5/13
46

I p = Yb V p Vq + ( YM ) V p Vm + YM V p Vn

I pq
I pm
I pn
YM
m
I pm
I pn
11/5/13
YM
Ip
Yb
I pq
47

I
I
m
Im
+
Va Z a
In
ZM
Zb
+
Vb
Ip
Iq
Im
YM
m
YM
Ya
In
11/5/13
YM
YM
Ip
Yb
q
48
Iq
Example: Consider our earlier example

2
j4.0
1.0 90
j8.0
j4.0
j5.0
j2.5
j0.8
j0.8
0
11/5/13
+
0.68 135
49
Example: Replace these

j4.0
1.0 90
j8.0
j4.0
j5.0
j2.5
j0.8
j0.8
0
11/5/13
+
0.68 135
50
Example: With these

2
Ya = j6.25

3
YM = j3.75

Yb = j6.25

1

and nd the new Ybus.
11/5/13
1 2 3
1 j6.25 j3.75 j2.5

Y = 2 j3.75 j6.25 j2.5
3 j2.5
j2.5
j5
51
Example:

j6.25
j8.0
j3.75
3
1.0 90
j6.25
j5.0
j2.5
j0.8
j0.8
0
11/5/13
+
0.68 135
52
Example: Focus just on the part that changes

2
j6.25
j8.0
j3.75
j6.25
Ip
j8.0
I3
YM
j3.75
j2.5
I3
Im
j2.5
Im
Ip
YM
Ya
In
1
11/5/13
YM
YM
Ya + YM = j2.5
Yb
2
Iq
In
Yb + YM = j2.5
YM
1
Iq
53
Example: Replace the part that changed:

2
j2.5
j5.0
j11.75
1.0 90
j2.5
j2.5
j0.8
j0.8
0.68 135
0
11/5/13
54
Example: Now write the nodal equa2ons by inspec2on:
1.0 90
0.68 135
11/5/13
j11.75
j2.5
j2.5
2 ( j2.5) j11.75
j11.75
j2.5 j5.0 j11.75
j2.5
j5.0
=

j2.5
j2.5
2 ( j2.5) j0.8
0

j2.5
j5.0
0
j2.5
j5.0
j0.8
j16.75 j11.75
j2.5
j2.5 V1
j11.75
j19.25
j2.5
j5.0
V2
=
V
j2.5
j2.5
j5.8
0
j2.5
j5.0
0
j8.30 V4
55
V1
V2
V3
V4
ModicaSon of Ybus

The building-block and the equivalent circuits approach provide

insights into how each branch self- and mutual admiKance
contributes to the entries of Ybus and the corresponding equivalent
network of the overall system.

Clearly Ybus is merely a systema2c means of combining the nodal
admiKance matrices of the various branches.

Consequently, we can easily see how to modify Ybus to account for
branch addi2ons or other changes to the system network.

For instance, consider the modica2on of an exis2ng Ybus to reect
the addi2on of the branch admiKance Ya between the nodes m and n.
11/5/13
56
ModicaSon of Ybus

We simply add Ya to the elements Ymm and Ynn of Ybus and subtract Ya
from elements Ymn = Ynm.

Moreover, to remove a branch admiKance Ya already connected
between nodes m and n of the network, we simply add the branch
admiKance Ya be tween the same nodes.
11/5/13
57
Example

Determine the bus admiKance matrix for previous example when

the coupling eects are removed.

We will do this by
a. removing the two coupled branches completely
b. restoring each of these branches without mutual coupling.
For the original network with coupling present we found:
Ybus
11/5/13
j16.75 j11.75
j2.5
j2.5
j5.0
j11.75 j19.25 j2.5
=
j2.5
j2.5
j5.8
0
j2.5
j5.0
0
j8.30
58
Example

First remove the coupled branches:
a
Ybus
j16.75 j11.75
j2.5
j2.5
j5.0
j11.75 j19.25 j2.5
=
j2.5
j2.5
j5.8
0
j2.5
j5.0
0
j8.30
j10.5
j8.0
0
j2.5
0
j5.0
j8.0 j13.0
=
0
0
j0.8
0
j2.5
j5.0
0
j8.30
11/5/13
j6.25 j3.75 j2.5

j3.75 j6.25 j2.5
j2.5
j5
j2.5

0
0
0
0
0
0
59
Example

Connect j4.0 between branches 1 and 3:
b
Ybus
j10.5
j8.0
0
j2.5

0
j5.0
j8.0 j13.0
==
+
0
0
j0.8
0

j2.5
j5.0
0
j8.30
j14.5
j8.0
j4
j2.5
0
j5.0
j8.0 j13.0
=
j4
0
j4.8
0
j2.5
j5.0
0
j8.30
11/5/13
1
0
1
0
0 1 0
0 0 0
j4 )
(
0 1 0
0 0 0
60
Example

Connect j4.0 between branches 2 and 3:

j14.5

j8.0
j4
j2.5 0 0 0

c j8.0 j13.0
0
j5.0
0 1 1
Y
=
+
0 1 1
bus j4
0
j4.8
0

j2.5
0 0 0
j5.0
0
j8.30

j14.5
j8.0
j4
j2.5

j5.0
j8.0 j17.0 j4.0
= j4
j4.0
j8.8
0

j2.5
j5.0
0
j8.30

Which agrees with the original result.
11/5/13
0
0
0
0
( j4 )
61
The Bus Impedance Matrix Zbus

The bus admiKance matrix of a large-scale interconnected power
system is typically a very sparse matrix with many zero elements.

We have seen how Ybus is constructed branch by branch from
primi2ve elements.

Conceptually we can invert Ybus to obtain Zbus, but direct inversion is
rarely performed because of the large scale.

The bus impedance matrix can be directly constructed element by
element as was done for Ybus though the work involved is greater.

Zbus however has some very useful proper2es.
11/5/13
62

For example, we will see that the diagonal elements of Zbus is the
Thvenin impedance at the corresponding bus.

Unlike Ybus, Zbus is never sparse and contains zeros only when the
system is regarded as being subdivided into independent parts by
open circuits.

Although Ybus is widely used in power-ow analysis, Zbus is favored
for fault analysis, hence both have important roles.
11/5/13
63

1
Recall the rela2onship: bus = Ybus

For simplicity, consider a network with three independent busses:

Z
Z
Z
12
13
11

bus = Z12 Z 22 Z 23

Z13 Z 23 Z33

The diagonal elements are called driving point impedances and the
o-diagonal terms are called the transfer impedances of the buses.
11/5/13
64

Lets gain some insights. For bus 2 we can write:

= Ybus V

I 2 = Y21V1 + Y22V2 + Y23V3

If V1 and V3 are reduced to zero by shor2ng buses 1 and 3 to the
reference node and V2 is applied to bus 2 so that current I2 enters
bus 2, the self-admiKance at bus 2 is:
I2
Y22 =
V2 V =V =0
1
11/5/13
65

For a 3-bus reac2ve network:
Ya
Yd
V1
Yb
I1
I3
Yc
+
V3
I2
V2
Ye
Yf
0 (Reference)
11/5/13
66

Disable V1 and V3:
Ya
Yb
I1
Yd
11/5/13
Yc
+
V2
I3
Ye
Yf
I2
67

Disable V1 and V3:

Ya
1

Yb 3

I1
I3

Yd

11/5/13
Yc
Ye
+
V2
Yf
I2
68

Disable V1 and V3:

Ya
1

Yb 3

I1
I3

Yd

11/5/13
Yc
Ye
+
V2
Yf
I2
69

Disable V1 and V3:

Ya
1
2

Yb 3
Yc
+

I1
I3

V2

Yd

Yf
Ye

Clearly I 2 = Y21V1 + Y22V2 + Y23V3 = Y22V2 = Ya + Yc + Y f V2

Y22 = Sum of conductances incident on bus 2
11/5/13
I2
70

For bus 2 we can write:

= Ybus V

I1 = Y11V1 + Y12V2 + Y13V3

If V1 and V3 are reduced to zero by shor2ng buses 1 and 3 to the
reference node and V2 is applied to bus 2 and causes current I1 to
leave bus 1, hence the 2-1 transfer-admiKance:
I1
Y12 =
V2 V =V =0
1
11/5/13
71

Disable V1 and V3:

Ya
1

Yb 3

I3

Yd

I1

See how V2 is directly across Ya.

11/5/13
Yc
Ye
+
V2
Yf
I2
72

See how V2 is directly across Ya:

V2 +
1

Ya

Yb 3
Yc
I =0

I =0
I3

Yd

I1
Ye

11/5/13
+
V2
Yf
I2
73

Disable V1 and V3:

V2 +
1
2

Ya

Yb 3
Yc
I =0
I2
+

I =0
I3

V2

Yd

Yf
I1
Ye

I1
See how V2 is directly across Ya. Y12 =
V2 V =V =0

1
3
Y12 = Y21 is the nega2ve of the admiKance connec2ng buses 1 and 2.

11/5/13
74

How about the impedances representa2on?

Again for a network of three independent buses:
V = bus
V1 = Z11 I1 + Z12 I 2 + Z13 I 3
V2 = Z 21 I1 + Z 22 I 2 + Z 23 I 3
Z ij = Z ji
V3 = Z31 I1 + Z31 I 2 + Z33 I 3
V2
Z 22 =
I2
11/5/13
I1 = I3 =0
V1
Z12 =
I2
I1 = I3 =0
75

1
Zb
Zd
V1
Zc
Ze
V2
Z 22 =
I2
11/5/13
Za
I1 = I3 =0
+
V3
I2
Zf
V1
Z12 =
I2
+
V2
I1 = I3 =0
76

1
Zb
Zd
V1
Zc
Ze
V2
Z 22 =
I2
11/5/13
Za
I1 = I3 =0
+
V3
I2
Zf
V1
Z12 =
I2
+
V2
I1 = I3 =0
77


1
Za

Z b 3 Z c
+

+
Z
d
V1

Z e V3

Not as simple.
11/5/13
V2
Z 22 =
I2
I1 = I3 =0
I2
Zf
+
V2
78
Thvenins Theorem and Zbus

The bus impedance matrix provides important informa2on
regarding the Thvenin impedance presented by the network to
each of its buses.

To establish this, denote ini2al values of bus current as I0 with
corresponding bus voltages of V0. V01 to V0N are the open circuit bus
voltages measure with respect to the reference node.

0
0
V
=
bus

When the bus currents change from their ini2al value I0 to a new
value I0 + I, the new bus voltages are given by:
V = Z bus 0 + = bus 0 + bus = V 0 + V

11/5/13
79
Thevenins Theorem and Zbus

Consider a large-scale network where bus k has been extracted
along with the reference bus.

3
n

Original
2
Network

Zbus
k

1

0
Reference

Ini2ally consider the network not to be energized so that as I0 and
V0 are zero.
11/5/13
80

Now inject a current Ik into bus k from a current source connected
to the reference node. This will result in a voltage change at all the
buses of the network as shown. This can be expressed as
V3
V2
V1
n
Original
Network
Zbus
11/5/13
Vn
Reference
Vk
I k
81

This can be expressed as:
11/5/13
V1

V2

=
Vk

VN

Z11
Z12
Z1k
Z 21
Z 22
Z 2k
Z k1
Zk 2

Z kk

Z N 1 Z N 2 Z Nk
Z1N 0
Z2 N 0
Z kN
k
Z NN 0
82

V
Mul2plying:
1

V2

=

Vk

V
N

Since
Z1k
Z 2k

I k
Z kk

Z Nk
V = V 0 + V = bus 0 + bus
Vk = Vk0 + Z kk I k
11/5/13
83

This suggests the equivalent circuit:
Vk = Vk0 + Z kk I k
ZTh = Z kk
+
Vk0
11/5/13
Original
Network
Zbus
Reference
+
Vk
I k
84

In a similar way we can determine the Thvenin impedance
between any two buses j and k of the network.

Again consider an otherwise unenergized (dead) network with
current injec2ons at buses j and k:
V3
V2
V1
k
Original
Network
Zbus
0
11/5/13
Vk I k
V j
I j
Reference
85

Now we write:
11/5/13
V1

V j
=
Vk

VN

Z11 Z1 j
Z1k

Z j1 Z jj
Z jk
Z k1 Z kj
Z kk

Z N 1 Z Nj
Z Nk
Z1N
0
Z jN I j
Z kN I k
Z NN 0
86

Mul2plying:
11/5/13
V1

V j
=
Vk

VN

Z1 j I j + Z1k I k
Z jj I j + Z jk I k
Z kj I j + Z kk I k
Z Nj I j + Z Nk I k
87

Thus: V = V 0 + Z I + Z I
Recall:
j
j
jj
j
jk
k

Z ij = Z ji
0
Vk = Vk + Z kj I j + Z kk I k

Algebra: add zero to each equa2on as shown:

0
V
=
V
+ Z jj I j + Z jk I k + Z jk I j Z jk I j

j
j

Vk = Vk0 + Z kj I j + Z kk I k + Z kj I k Z kj I k

Rearrange:
(
(
)
)
V j = V j0 + Z jj Z jk I j + Z jk I k + I j
) (
Vk = Vk0 + Z kj I j + I k + Z kk Z kj I k
11/5/13
88

Thvenin Equivalent Circuit of the system between buses j and k:
V j = V j0 + Z jj Z jk I j + Z jk I k + I j
) (
Vk = Vk0 + Z kj I j + I k + Z kk Z kj I k
Original
Network
Vk0 +
Zbus
Z kk Z kj
I k
+
V j0+
Vopen = Vk0 V j0
Z jj Z jk
circuit
I j
Z jk = Z kj
0
11/5/13
Reference
89

0
0
Open Circuit Voltage: Vopen = Vk V j
circuit
Z kk Z kj
Original
Network
Vk0 +
Zbus
I k
+
0
Vj
Vopen = Vk0 V j0
Z jj Z jk
circuit
I j
Z jk = Z kj
0
11/5/13
Reference
90

Short Circuit Current:
Original
Network
Vk0 +
Zbus
Z kk Z kj
I k
V j0 +
I sc
Z jj Z jk
I j
Z jk = Z kj
0
11/5/13
Reference
ZTh,i, j
I sc =
Vk0 V j0
Z jj + Z kk 2Z jk
Voc
=
= Z jj + Z kk 2Z jk
I sc
91

Thvenin equivalent circuit between buses j and k:
ZTh,i, j = Z jj + Z kk 2Z jk
VTh,i, j = Vk0 V j0
11/5/13
92

Note how this also captures the Thvenin equivalent from bus to
reference:
At bus k, Vth,k = V0k and Zth,k = Zkk Zkj + Zkj= Zkk
At bus j, Vth,j = V0j and Zth,j = Zjj Zkj + Zkj= Zjj
Original
Network
Vk0 +
Zbus
Z kk Z kj
I k
Z jj Z jk
I j
V j0+
Z jk = Z kj
0
11/5/13
Reference
93

For termina2on in an arbitrary branch impedance Zb:

Original
Z kk Z kj
k
I k
Network

0+
V
Z
k
bus

Zb I b

j
Z jj Z jk
I j

0
Vj +

Vk V j
Vk0 V j0
Z jk = Z kj

Ib =
=
Zb
ZTh, j,k + Z b

0
Reference

We will use this result momentarily to see how to modify Zbus when
a branch is added to the network.
11/5/13
94
Example Consider again our earlier example:

2
j6.25
j8.0
j3.75
3
1.0 90
j6.25
j5.0
j2.5
j0.8
j0.8
0
11/5/13
+
0.68 135
95
Example There we found:

V
j16.75
j11.75
j2.5
j2.5
0
1

j5.0 V2
0
= j11.75 j19.25 j2.5

1.0 90
j2.5
j2.5
j5.8
0
V3

0.68
135
j2.5
j5.0
0
j8.30 V4

A capacitor having reactance of 5.0 per unit is connected between
bus 4 and the reference node.

Find the current drawn by the capacitor.
11/5/13
96
Example We can nd Zbus by inver2ng Ybus using Mathcad:

This is BEFORE the capacitor is inserted.
11/5/13
97
Example
11/5/13
98
Example Solving:
11/5/13
99
Example If an addi2on current of I 4 = 0.2205669.2534 pu

is injected into the network at bus 4 (without the capacitor) nd the
resul2ng bus voltages.
11/5/13
100
Example
11/5/13
101
Example These examples illustrate the importance of the bus

impedance matrix and incidentally show how adding a capacitor at
a bus causes a rise in bus voltages (in an induc2ve network).

The assump2on that the angles of voltage and current sources
remain constant aTer connec2ng capacitors at a bus is not en2rely
valid if we are considering opera2on of a power system.
11/5/13
102
ModicaSon of an ExisSng Zbus

We now see how to use the Thvenin equivalent circuit and the
exis2ng Zbus to solve for new bus voltages in a network following the
addi2on of a branch (without having to recompute a new Zbus).

Since Zbus is such an important quan2ty in power system analysis, let
us now see how to add new lines or addi2onal buses to an exis2ng
network without star2ng from scratch.

When we see how to modify Zbus we will also see how to construct it
directly without matrix inversion.

There are four dis2nct cases. (Of increasing complexity.)
11/5/13
103

Case 1: Adding Zb from a new bus p to the reference node.

Adding a new bus p connected to the reference node through Zb
without a connec2on to any of the other buses of the original
network cannot alter the original bus voltages when a current Ip is
injected at the new bus.
Ip
p
Original
Network
Zbus
0
11/5/13
Zb
Reference
+
Vp
104

Case 1: Adding Zb from a new bus p to the reference node.

11/5/13
V10
V20
=
VN0
Vp
original
p 0 0 0 Z
b
I1
I2

IN
Ip
105

Case 2: Adding Zb from a new bus p to an exis2ng bus k.

With current Ip injected at bus p, the current entering the original
network at bus k will become the sum of Ik injected at bus k plus the
current Ip entering through Zb.
k
Original
Network
Zbus
0
11/5/13
Ip
Zb
Reference
106


Current Ip entering bus k will cause the original voltage to increase:
Vk = Vk0 + I p Z kk
k
Original
Network
Zbus
0
11/5/13
Ip
Z kk
Zb
Reference
107


Vp will be larger than the new Vk by: V p = Vk0 + I p Z kk + I p Z b
= Vk0 + I p ( Z kk + Z b )
k
Original
Network
Zbus
0
11/5/13
+
Vk
Vb +
Zb
Ip
+
Vp
Reference
108


V p = Z k1 I1 + Z k 2 I 2 ++ Z kN I N + I p Z kk + Z b

Vk0
A new row must be added to Zorig:
Z bus( new)
Z1k
1
V2
Z 2k
Z
original
Z
Nk
N
V p p Z k1 Z k 2 Z kN Z kk + Z b
11/5/13
I1
I2

IN
Ip
109

Case 3: Adding Zb from an exis2ng bus k to the reference node.

k
Original
Network
Zbus
11/5/13
Ik
Zb
Reference
110


Add a new bus p connected to bus k through Zb then short-circuit
bus p to the reference node by lewng Vp = 0.
k
Original
Network
Zbus
0
11/5/13
Ip
Zb
Reference
111


k
Original
Network
Zbus
0
11/5/13
Ip
Zb
Reference
112


V1
V2
=
VN
11/5/13
Z k1
Z 2k
Z Nk
Z kk + Z b
Z1k
original
Z k 2 Z kN
I1
I2

IN
Ip
113


This can be simplied. Pulling out the ith equa2on:

Vi = Z i1 I1 + Z i2 I 2 ++ Z iN I N + Z ik I p

Write out the last equa2on:

0 = Z k1 I1 + Z k 2 I 2 ++ Z kN I N + Z kk + Z b I p

Solve for Ip:
Z k1
Zk 2
Z kN
Ip =
I1
I 2
IN
Z kk + Z b
Z kk + Z b
Z kk + Z b
11/5/13
114


Subs2tute for Ip back into the ith equa2on:
Vi = Z i1 I1 + Z i2 I 2 ++ Z iN I N + Z ik I p
= Z i1 I1 + Z i2 I 2 ++ Z iN I N
Z k1
Zk 2
Z kN
+ Z ik
I1
I 2
IN
Z kk + Z b
Z kk + Z b
Z kk + Z b
Z ik Z k1
Z ik Z k 2
Z ik Z kN
= Z i1
I1 + Z i2
I 2 ++ Z iN
IN
Z kk + Z b
Z kk + Z b
Z kk + Z b
11/5/13
115


We conclude:

Z ik Z kj
new

Z ij = Z ij
Z kk + Z b

FYI The procedure just performed is known as Kron Reduc2on.
11/5/13
116


We conclude:

Z ik Z kj
new

Z ij = Z ij
Z kk + Z b

FYI The procedure just performed is known as Kron Reduc2on.
11/5/13
117

Case 4: Adding Zb between two exis2ng buses j and k.

I j + Ib
Original
Network
Zbus
Zb
Ib
Ik Ib
0
11/5/13
Ij
Ik
Reference
118


This case is very similar to our discussion of Thvenin equivalents.
Recall we found that if we have an otherwise unenergized (dead)
network with current injec2ons at buses j and k then the resul2ng
change in bus voltages was given by:
V1

V j
=
Vk

VN

11/5/13
Z11 Z1 j
Z1k

Z j1 Z jj
Z jk
Z k1 Z kj
Z kk

Z N 1 Z Nj
Z Nk
Z1N
0
Z jN I j
Z kN I k
Z NN 0
Z1 j I j + Z1k I k
Z jj I j + Z jk I k
Z kj I j + Z kk I k
Z Nj I j + Z Nk I k
119


For the present case we inject Ib at bus j and Ib at bus k
V1

V j
=
Vk

VN

11/5/13
Z11 Z1 j
Z1k

Z j1 Z jj
Z jk
Z k1 Z kj
Z kk

Z N 1 Z Nj
Z Nk
Z1N
0
Z jN I b
Z kN I b
Z NN 0
Z1 j I b Z1k I b
Z jj I b Z jk I b
Z kj I b Z kk I b
Z Nj I b Z Nk I b
120


Examine a few equa2ons:
V1 = Z11 I1 ++ Z1 j I j + Z1k I k ++ Z1N I N + I b Z1 j Z1k

V10
V1
V j = Z j1 I1 ++ Z jj I j + Z jk I k ++ Z kN I N + I b Z jj Z jk

V j0
V j
Vk = Z k1 I1 ++ Z kj I j + Z kk I k ++ Z kN I N + I b Z kj Z kk

Vk0
Vk
11/5/13
121


Thus, for the ith equa2on:

Vi = Z ij I b Z ik I b

= Z ij Z ik I b

This means that the vector V is found by subtrac2ng column k
from column j of Zorig and mul2plying by Ib.
11/5/13
122


We have N equa2ons for V1 ,,VN

but N + 1 unknowns: I1 ,, I N , I b

We need one more equa2on. It is always true from Thvenin's
theorem that (from an earlier slide)
Ib =
11/5/13
Vk0 V j0
ZTh, j,k + Z b
0 = V j0 Vk0 + I b ZTh, j,k + Z b
)
123


No2ce from V j = Z j1 I1 ++ Z jj I j + Z jk I k ++ Z kN I N + I b Z jj Z jk

0
V
V j

j
that Vj0 is the inner product of row j of Zorig and the vector of bus
currents I.

I ++ Z kj I j + Z kk I k ++ Z kN I N + I b Z kj Z kk
k1 1
Similarly, Vk = Z

Vk0
Vk

Vk0 is the inner product of row k of Zorig and the vector of bus
currents I. Subs2tu2ng this fact in the last equa2on on the previous
slide (the one obtained from the Thvenin circuit):
11/5/13
124


0
0
0
=
V
V
+ I b ZTh, j,k + Z b

j
k

j
k
0
=
+ I b ZTh, j,k + Z b
orig
orig

where orig
i denotes the ith row of the original bus impedance
matrix.
11/5/13
125


Puwng it all together:
11/5/13
V1

Vj
Vk =

VN
orig
j k
orig
orig
j
orig
k
orig
Z bb
I1

Ij
Ik

IN
Ib
126


j
k
orig
orig orig
=

0
I b

j k
Z

orig
bb
orig

Where: Z bb = ZTh, j,k + Z b

But recall: ZTh,i, j = Z jj + Z kk 2Z jk

Thus: Z bb = Z jj + Z kk 2Z jk + Z b
11/5/13
127


The matrix itself can be reduced using Krons method again (the
details are leT as an exercise):
V
1
V
j
V =
V
N
11/5/13
Z11
Z1 j
Z1k
Z1N
Z j1 Z k1
Z j1
Z jj
Z jk
Z jN
Z jj Z kj
Z k1
Z kj
Z kk
Z kN
Z jk Z kk
Z1N
Z Nj
Z Nk
Z NN
Z j1 Z k1
Z jj Z kj
Z jk Z kk
Z jN Z kN
Z jN
Z kN
Z bb
I1

Ij
Ik

IN
Ib
128


Write out the ith and the last equa2on and do the math:
Vi = Z i1 I1 + + Z ij I j + Z ik I k + + Z iN I N + Z ji Z ki I b
Ib =
Z j1 Z k1
Z bb
I1
Z jj Z kj
)(
Z bb
Z ji Z ki Z j1 Z k1
Vi = Z i1
Z bb
)(
Ij
Z bb
I k
) I ++ Z ( Z
Z ji Z ki Z jk Z kk
+ Z ik
Z bb
11/5/13
Z jk Z kk
) I
ij
ji
Z jN Z kN
Z bb
)(
IN
Z ki Z jj Z kj
Ij
Z bb
)(
Z ji Z ki Z jN Z kN
++
Z
Z bb
k
iN
) I
129


Thus:

Z jm Z km Z jn Z kn
new
Z mn = Z mn

Z bb

Z jm Z km Z jn Z kn
= Z mn

Z jj + Z kk 2Z jk + Z b

Check the math!
11/5/13
)(
)(
130
Example

From our earlier example where:

Modify Zbus to account for a capacitor having reactance 5.0 per unit
between bus 4 and the reference node, then nd V4 using the
impedance of the new matrix and the current sources used and
compare with our previous result.
11/5/13
131
Example
j8.0
3
j5.0
j2.5
ZC = j5
1.0 90
YC = j0.2
j0.8
j0.8
0
11/5/13
+
0.68 135
132
Example

here k = 4
V1
V2
=
VN
original
Z k1 Z k 2 Z kN
Z 2k
Z Nk
Z kk + Z b
j0.7321
j0.6914
j0.6132
j0.6368
j0.6914
j0.7197
j0.6082
j0.6418
j0.6132
j0.6082
j0.6989
j0.5511
j0.6368
j0.6418
j0.5511
j0.6989
j0.6368
j0.6418
j0.5511
j0.6989
11/5/13
Z1k
I1
I2

IN
Ip
j0.6418
j0.5511
j0.6989
j0.6989 j5
= j 4.30110
j0.6368
133
Example

V j0.7321 j0.6914
1
V2 j0.6914 j0.7197

V3 = j0.6132 j0.6082
V j0.6368 j0.6418
4
0 j0.6368 j0.6418

Via Kron reduc2on (k = 4):
j0.6132
j0.6368
j0.6082
j0.6418
j0.6989
j0.5511
j0.5511
j0.6989
j0.5511
j0.6989
j0.6418
j0.5511
j0.6989
j4.30110
j0.6368
Z ik Z kj
new
ij
= Z ij
new
11
Z14 Z 41
j0.6368 j0.6368
= Z11
= j0.7321
= j0.82555
Z 44 + Z b
j4.30110
new
24
Z 24 Z 44
j0.6418 j0.6989
= Z 24
= j0.6418
= j0.74606
Z 44 + Z b
j4.30110
Z
Z
Z
11/5/13
I1
I2
I3
I 4
Ib
Z kk + Z b
134
Example

Similarly for the other terms which gives:
new
Z bus
=
11/5/13
j0.82555
j0.78641
j0.69482
j0.78641
j0.81542
j0.69045
j0.69482
j0.69045
j0.76951
j0.74024
j0.74606
j0.64065
j0.74024
j0.74606
j0.64065
j0.81247
135
Example

This is the same result obtained previously.
11/5/13
136
Example

For this example we could have also found V4 from:

0
0
V
V
V
V
j
j

Ib = k
= k
Zb
ZTh, j,k + Z b

with V 0 = 0, Z
=Z
j
Th, j,k
44
Vk
Vk0
Zb
=
Vk =
Vk0
Z b Z 44 + Z b
Z 44 + Z b
11/5/13
137
Direct DeterminaSon of Zbus

We could always determine Zbus by inver2ng Ybus, but this is
problema2c for large networks.

Here we develop an algorithm to build Zbus directly, one that is
readily programmable.

When we begin we have a list of branch impedances and knowledge
of the busses to which they are connected.

For one bus and one impedance Za connected to the reference node
the maKer is simple.
V1 = Z a I1
11/5/13
138

We can add a new bus connected to the reference node through Zb:

V Z
I

0
1
a
1
=

V2 0 Z b I 2

We then use the cases just developed to add (modify) the exis2ng
Zbus by adding elements.

Lets do an example.
11/5/13
139

4
j0.2
j0.125
j0.25
(5)
(6)
j0.4
(2)
1
(3)
3
(4)
j1.25
(1)
11/5/13
j1.25
140

To establish bus 1:
j0.2
j0.125
V1 = j1.25 I1
j0.25
(5)
(6)
j0.4
(2)
1
(3)
3
(4)
j1.25
(1)
11/5/13
j1.25
141

To establish bus 2 use:


Here k = 1, p = 2

11/5/13
V1
V2
=
VN
Vp
Z k1
Z 2k
Z Nk
Z kk + Z b
Z1k
original
Z k 2 Z kN
I1
I1

IN
Ip
142

To establish bus 2 use:


Here k = 1, p = 2

V j1.25
Z11
1
=
V2 Z11 Z11 + Z b

j1.25
=
j1.25
11/5/13
I
1
I2
I1 j1.25
j1.25
=
j1.25 + j0.25 I 2 j1.25
j1.25 I1
j1.5 I 2
143

Now establish bus 3 in a similar way adding impedance element (3):

p = 3, k = 2
V j1.25
1
V2 = j1.25

V3 j1.25
j1.25
= j1.25
j1.25
11/5/13
j1.25
j1.5
j1.5
j1.25
j1.5
j1.5
I1
j1.5
I2
j1.5 + j0.4 I 3
j1.25 I1
j1.5 I 2
j1.9 I 3
j1.25
144

Now add j1.25 from bus 3 (k = 3) to the reference bus using Case 3:
new
ij
Z13Z31
j1.25
Z33 + Z b
Z 23Z31
new
Z = j1.25
Z33 + Z b
j1.25 Z33Z31
Z33 + Z b
11/5/13
= Z ij
Z ik Z kj
Z kk + Z b
Z13Z32
j1.25
Z33 + Z b
Z 23Z32
j1.5
Z33 + Z b
Z33Z32
j1.5
Z33 + Z b
Z13Z33
j1.25
Z33 + Z b
Z 23Z33
j1.5
Z33 + Z b
Z33Z33
j1.9
Z33 + Z b
145

Now add j1.25 from bus 3 to the reference bus using Case 3 (Krons
reduc2on):
1.25 1.25
1.25 1.5
1.25 1.9
1.25
1.25
1.25
1.9 + 1.25
1.9 + 1.25
1.9 + 1.25
1.5 1.25
1.5 1.5
1.5 1.9
new
Z = j 1.25
1.5
1.5
1.9 + 1.25
1.9 + 1.25
1.9 + 1.25
1.9 1.25
1.5 1.5
1.9 1.9
1.25 1.9 + 1.25 1.5 1.9 + 1.25 1.9 1.9 + 1.25
11/5/13
146

Now add j1.25 from bus 3 to the reference bus using Case 3 (Krons
reduc2on):
new
Z = j
= j
11/5/13
1.56
1.88
2.38
1.25
1.25
1.25
3.15
3.15
3.15
1.88
2.25
2.85
1.25
1.5
1.5
3.15
3.15
3.15
2.38
2.25
2.85
1.25
1.5
1.9
3.15
3.15
3.15
0.75397 0.65476 0.49603
0.65476 0.78571 0.59524

0.49603 0.59524 0.75397
147

Now add j0.20 from bus 3 to establish bus 4:
new
Z =
11/5/13
j0.75397
j0.65476
j0.49603
j0.65476
j0.78571
j0.59524
j0.49603
j0.59524
j0.75397
j0.49603
j0.59524
j0.75397
j0.75397
j0.65476
j0.49603
j0.65476
j0.78571
j0.59524
j0.49603
j0.59524
j0.75397
j0.49603
j0.59524
j0.75397
j0.59524
j0.75397
j0.75397 + j0.2
j0.49603
j0.59524
j0.75397
j0.95397
j0.49603
148

Finally add Zb = j0.125 between bus 2 and bus 4. Use:

Z jm Z km Z jn Z kn
new

Z mn = Z mn

with j = 2 and k = 4 and using:
11/5/13
old
)(
j0.75397
j0.65476
j0.49603
j0.65476
j0.78571
j0.59524
j0.49603
j0.59524
j0.75397
j0.49603
j0.59524
j0.75397
j0.49603
j0.59524
j0.75397
j0.95397
149

11/5/13
150

Final answer:
11/5/13
151

Sanity check:
j0.2
j0.125
j0.25
(5)
(6)
j0.4
(2)
1
Z11
(3)
3
(4)
j1.25
(1)
j1.25
Z11 = j1.25 j0.25 + ( j0.125 + j0.2 ) j0.4 + j1.25

11/5/13
152

Sanity check: Z11 = j1.25 j0.25 + j0.125 + j0.2 j0.4 + j1.25
j0.4 j0.125 + j0.2

= j1.25 j0.25 +
+ j1.25
j0.4 + j0.125 + j0.2
j0.4 j0.125 + j0.2

j1.25 j0.25 +
+ j1.25
j0.4 + j0.125 + j0.2
=
j0.4 j0.125 + j0.2
j1.25 + j0.25 +
+ j1.25
j0.4 + j0.125 + j0.2
11/5/13
153
Summary Direct DeterminaSon of Zbus

We shall again refer to these results in a future example.

The Zbus building procedures are simple for a computer which rst
must determine the types of modica2on involved as each branch
impedance is added.
11/5/13
154
Return to Fault Analysis
11/5/13
155
Fault CalculaSons Using Zbus

Again consider the
system considered
j0.25
earlier:
j0.25
j0.125
j0.4
j0.1
j1.15
j0.1
j1.15
+
1.250
11/5/13
j0.2
+
0.85 45
156

Suppose there is a three-phase fault from bus 2 to ground.

We change the reactances in series with the generated voltages
from synchronous values to subtransient values.

This again is regarded as the per-phase equivalent of a balanced
three-phase system.

We designate the voltage from bus 2 to neutral as Vf the actual
voltage at bus 2 before the fault occurs, as shown:
11/5/13
157

j0.25
j0.125
j0.25
j0.4
j0.1
j1.15 j0.2
+
Ea
11/5/13
j0.2
Vf
j0.1
j1.15 j0.2
+
Eb
158

j0.25
j0.125
j0.25
j0.1
j0.2
11/5/13
j0.1
j0.2
+
Ea
No change made
j0.4
I =0
j0.2
+
Vf
+
Eb
159

j0.25
Simulated
fault
j0.125
I f
+
j0.2
Vf
j0.25
j0.4
j0.1
j0.2
+
Ea
11/5/13
j0.1
+
V f
j0.2
+
Eb
160

If, caused by the addi2on of Vf, distributes itself throughout the
system before owing out of bus 2 through the Vf source.

In doing so, it produces whatever changes that occur in the system
due to the fault.

If all other sources are disabled leaving Vf to act alone, then If
into bus 2 is the only current entering the network from external
sources. With Vf as the only source, the network has Zbus of the
form:
11/5/13
161

j0.25
j0.125
j0.25
Simulated
fault
j0.4
j0.1
j0.2
+
Ea = 0
11/5/13
j0.2
I f
j0.1
+
V f
j0.2
+
Eb = 0
162

V V Z
Z12 Z13 Z14 0
1
1
11

V f
Z12 Z 22 Z 23 Z 24 I f
V2
=
=

V3 V3 Z13 Z 23 Z33 Z334 0

0
V
Z
Z
Z
Z
V
4
24
34
44

4

14

Recall that the matrix elements would be dierent from the
previous example since they now would contain the subtransient
values.
11/5/13
163

Simplifying:

Z
V1 V1 Z11 Z12 Z13 Z14 0
12

Z 22
V
Z
Z
Z
Z
I f

f
2
12
22
23
24
=
= I f
=
V V Z
Z 23 Z33 Z34
Z 23
0
3
13
3
Z
0
V
Z
Z
Z
Z
V
4
24
34
44
4
24

14

V = I Z
f
f 22

Vf

I f =
Z 22

Recall that Z22 is the Thvenin impedance of the network seen at
bus 2.
11/5/13
164

Subs2tu2ng:
11/5/13
V1
V2
= I f
V3
V4
Z12
Z 22
Vf
=
Z 22
Z 23
Z 24
Z12
Z 22
Z 23
Z 24
Z12
Vf
Z 22
Z 22
Vf
Z 22
Z 23
V f
Z 22
Z 24
Vf
Z 22
165

Thus:
11/5/13
V1
V2
V3
V4
Z
12 V f
Z 22
V
f

= Z 23
Vf

Z 22
Z
24 V
Z 22 f
166

When the generator voltage Vf is disabled and sources Ea, Eb, and
Vf are restored into the network, the currents and voltages
everywhere in the network will be the same as those exis2ng before
the fault.

Using superposi2on, these pre-fault voltages add to the changes
just derived to yield the total voltages exis2ng aTer the fault occurs.

A major simplifying assump2on: The faulted network is usually (but
not not always) assumed to be without load before the fault occurs.
In the absence of loads, no pre-fault currents ow and there are no
voltage dierences across branch impedances; all bus voltages
throughout the network are then the same as Vf, the pre-fault
voltage at the fault point. This simplies things considerably.
11/5/13
167

V1
V2
=
V3
V4

11/5/13
V f + V1
V f + V2
V f + V3
V f + V4
Z12
Vf
Vf
Z 22

V V
f
f

=
Z 23
Vf
Vf
Z 22

Z 24

Vf
Vf
Z 22
=V
f
Z12
1
Z 22
Z 23
1 Z
22
Z 24
1
Z 22
168

Thus the voltage at all busses of the network can be calculated
using the prefault voltage Vf of the fault bus and the elements of
the column of Zbus corresponding to the fault bus.

The calculated values of the bus voltages will yield the subtransient
currents in the branches of the network if the system Zbus has been
formed with subtransient values for the machine reactances.

More generally, when a three-phase fault occurs on bus k of a large
scale network we have:
I f =
11/5/13
Vf
Z kk
169

Neglec2ng prefault load currents, the voltage at any bus j during the
fault is:

Z jk
V j = V f Z jk I f = V f
Vf

Z kk

If the prefault voltage of bus j is not the same as the prefault
voltage of bus k, then we simply replace Vf on the leT in the above
equa2on by the actual prefault voltage of bus j.
11/5/13
170

Knowing the bus voltages during the fault, we can calculate the
subtransient current Iij from bus i to bus j in the line of impedance
Zb connec2ng those two busses as:
I ij =
Vi V j
Zb
Z jk
Z ik
V f Z V f V f Z V f
kk
kk
Zb
Z jk
Z ik
Vf +
Vf
Z kk
Z kk
=
Zb
= Z kk
Z ik Z jk V f
Z ik Z jk
=
=
I f
Z kk Z b
Z kk
11/5/13
171

This gives Iij as a frac2on of If appearing as a line ow from bus i to
bus j in the faulted network.

If bus j is directly connected to the faulted bus k by a line of series
impedance Zb, then the current contributed from bus j to the
current in the fault at bus k is simply

Vj

I jk =
Zb

Z jk
where, as found previously, V = V Z I = V
V
j
11/5/13
jk
Z kk
172

k
We see that only column k of Zbus, denoted from now on as Z bus
is required to evaluate the impact on the system of a symmetrical
three-phase fault at bus k.
11/5/13
173
Example

A three-phase fault occurs at bus 2 of
2
j0.25
j0.125
j0.25
j0.4
j0.1
j1.15
j0.1
j1.15
+
1.250
11/5/13
j0.2
+
0.85 45
174
Example

Determine:

the subtransient current in the fault
the voltage at busses 1, 3, and 4 during the fault
the current ow in the line from bus 3 to bus 1
the current contribu2ons to the fault from lines 3-2, 1-2, and 4-2.
Take the prefault voltage Vf at bus 2 as 1.0 per unit and neglect all
prefault currents.
11/5/13
175
Example

We rst nd Zbus via the Zbus building algorithm:
Z bus
11/5/13
j0.2436
j0.1938
j0.1544
j0.1938
j0.2295
j0.1954
j0.1544
j0.1954
j0.1954
j0.1456
j0.1506
j0.1046
j0.1456
j0.1506
j0.1046
j0.1954
176
Example

Since the load currents are neglected, the prefault voltage at each
bus is 1.0 per unit, the same as Vf at bus 2.

When the fault occurs,
Vf
1.0
I f =
=
= j4.3573 pu
Z 22 j0.2295
11/5/13
177
Example

The voltages during the fault are:
V1
V2
V3
V4
11/5/13
Z12
Vf
Vf
Z 22

0

Z 23
=
Vf Z Vf
22

Z 24
Vf
Vf
Z 22
j0.1938
1

j0.2295
0.1556

=
j0.1494 =
1 j0.2295 0.3490

0.3438

j0.1506
1
j0.2295

pu
178
Example

Fault currents contributed to bus 2 by the adjacent unfaulted
busses are:

V1 0.1556
From bus 1:
=
= j1.2448 pu
Z b1 j0.125

V3 0.3490
From bus 3:
=
= j1.3960 pu
Z b3
j0.25

V4 0.3438
From bus 4:
=
= j1.7190 pu
Z b4
j0.20

Note that these sum to If as required.
11/5/13
179
Fault CalculaSons Using Zbus Equivalent Circuits

Recall our Thvenin equivalent circuit model between any pair of
buses in a network that may be of interest.

This model is very helpful in developing and understanding
symmetrical fault equa2ons.
V j0
+
Z jj Z jk
Z kk Z kj
Vk0 +
Z jk = Z kj
0
11/5/13
Reference
180

Suppose bus k is the fault bus while bus j is unfaulted. Prefault bus
voltages are Vf if load currents are neglected.
Unfaulted bus
Vf
+
Z jj Z jk
Bus to be faulted
Vf +
Z kk Z kj
Z jk = Z kj
0
11/5/13
181

Since points x are at the same poten2al they can be shorted without
causing any change in the circuit.

Unfaulted bus
This can be redrawn:
+
Vf
Z jj Z jk
Bus to be faulted
Vf
Z jk = Z kj
11/5/13
Z kk Z kj
+
x
182

When the fault occurs:
Unfaulted bus
j
I f =
Vf
Vf
Z kk
Z jj Z jk
Faulted bus
Z kk Z kj
Z jk = Z kj
11/5/13
I f
I f
183

Consistent with our earlier result.

This suggests another approach for determining transient
quan22es.
j
Unfaulted bus
I f =
Vf
Vf
Z kk
Z jj Z jk
Z kk Z kj
Faulted bus
k
S
Z jk I f =
Z jk
Z kk
Vf
I f
Vj = Vf
Z jk
Z kk
Vf
I f
+
11/5/13
184
Example

A ve-bus network has generators at buses 1 and 3 rated 270 and
225 MVA, respec2vely. The generator subtransient reactances plus
the reactances of the transformers connec2ng them to the buses
are each 0.30 per unit on the generator ra2ng as a base.
1
2 j0.126
j0.168
5
j0.126
j0.21
j0.252 j0.336
4
0
11/5/13
185
Example

The turns ra2o of the transformers are such that the voltage base in
each generator circuit is equal to the voltage ra2ng of the
generator. Line impedances in per unit on a 100 MVA system base
are given in the gure. All resistances are neglected.

Using the bus impedance matrix for the network which includes the
generator and transformer reactances, nd the subtransient current
in a three-phase fault at bus 4 and the current coming to the faulted
bus over each line. Prefault current is to be neglected and all
voltages are assumed to be 1.0 per unit before the fault occurs.
11/5/13
186
Example

1
1.0
2 j0.126
j0.1111
j0.168
1.0
j0.1111
5
j0.126
j0.21
j0.252 j0.336
4
11/5/13
187
Example

Converted to the 100-MVA base, the combined generator and
transformer reactances are:

100
Generator at bus 1: X = 0.30
= 0.1111 pu
270

100
Generator at bus 2: X = 0.30
= 0.1333 pu
225

Use the Zbus building algorithm to obtain (leT as an exercise)
11/5/13
188
Example

Z bus
11/5/13
j0.0793
j0.0558
j0.0382
j0.0511
j0.0558
j0.1338
j0.0664
j0.0630
j0.0382
j0.0664
j0.0875
j0.0720
j0.0511
j0.0630
j0.0720
j0.2321
j0.0608
j0.0605
j0.0603
j0.1002
j0.0608
j0.0605
j0.0603
j0.1002
j0.1301
189
Example

Faulted bus k = 4

Unfaulted buses j = 3
Z33 Z34
= j0.0875 j0.0720
3
= j0.0155
Unfaulted bus
Z 44 Z 43
I f
+
1.0
Z34 = Z 43 = j0.0720
Faulted bus
= j0.2321 j0.0720
4
S
= j0.1601
I f
I f
I f =
11/5/13
1
1
=
= j4.308
Z 44 j0.2321
190
Example

Faulted bus k = 4

Unfaulted buses j = 5
Z55 Z54
= j0.1301 j0.1002
= j0.0299
Z 44 Z 45
+
1.0
Z54 = Z 45 = j0.1002
I f
5
Unfaulted bus
= j0.2321 j0.1002
Faulted bus
= j0.1319
4
S
I f
I f
1
1
I f =
=
= j4.308
Z 44 j0.2321
11/5/13
191
Example

The voltages (per unit) at bus 3 and 5 during the fault is:
Vj = Vf
Z jk
Z kk
Vf
Z34
j0.0720
V3 = 1
= 1
= 0.68979
Z 44
j0.2321
V5 = 1
11/5/13
Z54
j0.1002
= 1
= 0.56829
Z 44
j0.2321
192
Example

Currents (per unit) into the fault at bus 4 are j 2.255 from bus 5
and j 2.053 from bus 3. Note that these sum to j 4.308 which is
If.
1
j0.168
5
j0.126
j0.21
j0.252 4 j0.336
V3 = 0.68979
V5 = 0.56829
I3 =
2 j0.126
0.56829
= j2.25512
j0.252
I3 =
0.68979
= j2.0529
j0.336
11/5/13
193
Circuit Breakers

Three-phase faults occur more oTen on transmission lines than on
substa2on buses because of the greater exposure of the lines to
storms and accidental disturbances.

To analyze a line fault, the point of fault on the line can be assigned
a new bus number and Zbus for the normal congura2on of the
network can then be modied to accommodate the new bus.

Some2mes the circuit breakers at the two ends of the line do not
open simultaneously when a line fault is being cleared. If only one
circuit breaker has opened and the fault is not fully cleared, and
short-circuit current persists.
11/5/13
194
Circuit Breakers

hKp://www.abb.us/product/us/9AAC710003.aspx?country=US

hKp://www.energy.siemens.com/hq/en/power-transmission/high-
voltage-products/circuit-breaker/

11/5/13
195
Circuit Breakers

The so-called line-end fault represents the par2cular situa2on
where the three-phase fault occurs very close to one of the
termina2ng buses of the line, on the line side of the rst breaker
(near the fault) to open. The line breaker near the fault is called the
near-end breaker and that at the end away from the fault is called
the remote-end breaker.

The single-line diagram on the next slide shows a four-bus network
with a line-end fault at point P on the line connec2ng buses 1 and 2.
The line has series impedance Zb.

The near-end breaker at bus 2 is open and the remote-end breaker
is closed, leaving the fault s2ll on at point P, which we call bus k.
11/5/13
196
Circuit Breakers

1
Closed breaker
5
4
Open breaker 2
k

Fault
Point P
4
11/5/13
197
Circuit Breakers

To study this fault condi2on, we need to modify the exis2ng bus
impedance matrix Zorig to reect the near-end breaker opera2on.

This is accomplished in two steps:

1. Establish the new bus k by adding a line of series impedance
Zb between bus 1 and bus k using the procedure earlier
designated as Case 2.

2. Remove the line between bus 1 and bus 2 by adding line
impedance Zb between those two buses using the
procedure earlier designated as Case 4.
11/5/13
198
Circuit Breakers

1. Add a line of series impedance Zb between bus 1 and bus k.
Case 2: Adding Zb from a new bus p to an exis2ng bus k:
Z bus( new) =
k: Z k1
11/5/13
k:
Z 2k
Z Nk
Z kk + Z b
Z1k
Z original
Z k 2 Z kN
199
Circuit Breakers

1. Add a line of series impedance Zb between bus 1 and bus k.
Case 2: Adding Zb from a new bus k to bus 1:
Z bus( new) =
k: Z11
11/5/13
k:
Z 21
Z31
Z 41
Z11 + Z b
Z11
Z original
Z12
Z13
Z14
200
Circuit Breakers

2. Remove the line between bus i and bus j (1 and 2 in this case)
by adding line impedance Zb between those two buses.

Recall Case 4:
11/5/13
new
mn
= Z mn
Z
(
jm
)(
Z im Z jn Z in
Z jj + Z ii 2Z ji Z b
201
Circuit Breakers

2. Remove the line between bus 1 and bus 2 adding line
impedance Zb between those two buses:

For i = 1 and j = 2:
11/5/13
new
mn
= Z mn
Z
(
2m
Z1m ) ( Z 2n Z1n )
Z 22 + Z11 2Z 21 Z b
202
Circuit Breakers

But Zkk is the only element needed to nd the current in the
fault, hence:
new
kk
= Z kk
Z
(
2k
Z1k ) ( Z 2k Z1k )
Z 22 + Z11 2Z 21 Z b
Z bus( new) =
k: Z11
11/5/13
= Z kk
(Z
1k
Z 22 + Z11 2Z 21 Z b
k:
Z original
Z12
Z13
Z 2k )
Z14
Z 21
Z31
Z 41
Z11 + Z b
Z11
203
Circuit Breakers

But Zkk is the only element needed to nd the current in the
fault, hence:

2

Z 21 Z11
new

Z kk = Z11 + Z b
Z 22 + Z11 2Z 21 Z b

Neglec2ng prefault currents:
I f =
11/5/13
Vf
Z
new
kk
1.0
=
Z11 + Z b
(Z
21
Z11 )
Z 22 + Z11 2Z 21 Z b
204
Circuit Breakers

Does this make sense?
Z kk = Z b + ( Z 22 Z 21 Z b ) ( Z11 Z12 ) + Z 21
1
Zb
Z11 Z12
+
Vf
Zth,12
Z 22 Z 21
Z b
Z12 = Z 21
11/5/13
Z kk
Zth,12 = Z11 + Z 22 2Z12
205
Circuit Breakers

Does this make sense?

Z kk = Zb + ( Z22 Z21 Zb ) ( Z11 Z12 ) + Z21

Z 22 Z 21 Z b ) ( Z11 Z12 )
(
= Zb + Z21 + Z + Z 2Z Z
22
11
21
b

Z11 + Z 22 2Z 21 + Z 21 Z11 Z b ) ( Z11 Z12 )
(
= Zb + Z21 +
Z 22 + Z11 2Z 21 Z b

= Z + Z + ( Zth1,2 Zb + Z21 Z11 )( Z11 Z12 ) = Z + Z + 1+ Z21 Z11 ( Z Z )
11 12
b
21
b
21
Z
Z
Z

th1,2
b
th1,2
b
2
2

Z
Z
Z
Z
( 11 12 ) = Z + Z ( 21 11 )
=
Z
+
Z
+
Z

b
21
11
12
11
b
Zth1,2 Z b
Z 22 + Z11 2Z 21 Z b

Yes.
11/5/13
206
Circuit Breakers

The faulted circuit reduces to the following model (for this
example):
Z kk = Z11 + Z b
( Z21 Z11 )
Z 22 + Z11 2Z 21 Z b
+
Vf
I f
11/5/13
207
Circuit Breakers Example

For the previous example, a line-end short-circuit fault occurs on
line 1 2, on the line side of the breaker at bus 2. Neglec2ng
prefault currents and assuming rated voltage at the fault,
calculate the subtransient current into the fault when only the
near end breaker at bus 2 opens.
1
1.0
2 j0.126
j0.1111
j0.168 Fault
1.0
j0.1111
j0.126
j0.21
j0.252 j0.336
4
0
11/5/13
208
Z kk = Z11 + Z b
( Z21 Z11 )
Z 22 + Z11 2Z 21 Z b
+
Vf
I f
11/5/13
209

new
Z mn
= Z mn
11/5/13
)(
Z jm Z km Z jn Z kn
)Z
kk
= Z11 + Z b
(Z
21
Z11 )
Z 22 + Z11 2Z 21 Z b
210
Circuit Breaker SelecSon

The electric u2lity company furnishes data to a customer who
must determine the fault current in order to specify circuit
breakers properly for an industrial plant or industrial power
distribu2on system connected to the u2lity system at a certain
point. Instead of providing the Thvenin impedance of the
system at the point of connec2on, usually the power company
informs the customer of the short-circuit MVA which can be
expected at nominal voltage, namely

Short-Circuit MVA = 3 nominal kV I SC 103

where |Isc| in amperes is the rms magnitude of the short-circuit
current in a three-phase fault at the connec2on point.
11/5/13
211

Base megavoltamperes are related to base kilovolts and base
amperes |Ibase| by

Base MVA = 3 base kV I base 103

If base kilovolts equal nominal kilovolts, then dividing converts
the former to per unit, and we obtain

Short-Circuit MVA in per unit = I base in per unit

At nominal voltage the Thvenin equivalent circuit looking back
into the system from the point of connec2on is an emf of 1.0 per
unit in series with the per-unit impedance Zth.
11/5/13
212

Therefore, under short-circuit condi2ons,

1
1

Zth =
per unit =
per unit
Short-Circuit MVA
I SC

OTen resistance and shunt capacitance are neglected, in which
case Zth = Xth.

Thus, by specifying short-circuit megavoltamperes at the
customer's bus, the electric u2lity is eec2vely describing the
short-circuit current at nominal voltage and the reciprocal of the
Thvenin impedance of the system at the point of connec2on.

11/5/13
213

Much study has been given to circuit-breaker ra2ngs and
applica2ons, and this discussion gives only a brief introduc2on.

The inten2on is NOT to study study of breakers or their
applica2ons but rather to underscore the importance of
understanding fault calcula2ons.

From the current viewpoint two factors to be considered in
selec2ng circuit breakers are:
The maximum instantaneous current which the breaker must
carry (withstand) and
The total current when the breaker contacts part to interrupt
the circuit.

11/5/13
214

Up to this point we have devoted most of our aKen2on to the
subtransient current called the ini2al symmetrical current, which
does not include the dc component. Inclusion of the dc
component results in a rms value of current immediately aTer
the fault, which is higher than the subtransient current. For oil
circuit breakers above 5 kV the subtransient current mul2plied
by 1.6 is considered to be the rms value of the current whose
disrup2ve forces the breaker must with stand during the rst
half cycle aTer the fault occurs.

This current is called the momentary current, and circuit
breakers are oTen rated in terms of their momentary current as
well as other criteria.

11/5/13
215

The interrup2ng ra2ng of a circuit breaker was specied in KVA
or MVA. The interrup2ng KVA equals 31/2 (the kV of the bus to
which the breaker is connected) (the current which the
breaker must be capable of interrup2ng when its contacts part).
This interrup2ng current is of course lower than the momentary
current and depends on the speed of the breaker, such as 8, 5, 3,
or 2 cycles, which is a measure of the 2me from the occurrence
of the fault to the ex2nc2on of the arc. Breakers of dierent
speeds are classied by their rated interrup2ng 2mes.

The rated interrup2ng 2me of a circuit breaker is the period
between the instant of energizing the trip circuit and the arc
ex2nc2on on an opening opera2on, as shown (next slide).
Preceding this period is the tripping delay 2me, which is usually
assumed to be cycle for relays to pick up.
11/5/13
216

Ini2a2on of
Short Circuit
Ex2nc2on of Arc on
Primary Contacts
Energiza2on of
of Trip Circuit
Tripping
Delay
Par2ng of
Primary Arcing
Contacts
Interrup2ng Time
Opening
Time
Time
Arcing
Time
Contact Par2ng Time

DeniSon of InterrupSng Time, ANSI/IEEE Standard C.37.010-1979, Applica8on
Guide for AC High Voltage Circuit Breakers Rated on Symmetrical Current Basis.
11/5/13
217

The current which a breaker must interrupt is usually
asymmetrical since it s2ll contains some of the decaying dc
component.

Ra2ngs for ac high-voltage oil circuit breakers species the
interrup2ng current ra2ngs of breakers in terms of the
component of the asymmetrical current which is symmetrical
about the zero axis.

This current is properly called the required symmetrical
interrup2ng capability or the rated symmetrical short-circuit
current. OTen the adjec2ve symmetrical is omiKed. Selec2on of
circuit breakers may also be made on the basis of total current
(dc component included). Here we consider a brief treatment of
the symmetrical basis of breaker selec2on.
11/5/13
218

Breakers are iden2ed by nominal-voltage class, such as 69 kV.
Other factors specied are rated con2nuous current, rated
maximum voltage, voltage range factor K, and rated short-circuit
current at rated maximum kilovolts.

The rated maximum voltage of a circuit breaker is the highest
rms voltage for which the circuit breaker is designed.

The rated voltage range factor K is the ra2o (rated maximum
voltage the lower limit of the range of opera2ng voltage).

K determines the range of voltage over which the product of the
rated short-circuit current 2mes the opera2ng voltage is
constant.
11/5/13
219

In the applica2on of circuit breakers it is important not to exceed
the short-circuit capabili2es of the breakers.

A breaker is required to have a maximum symmetrical
interrup2ng capability equal to K 2mes the rated short-circuit
current.

Between the rated maximum voltage and 1/K 2mes the rated
maximum voltage the symmetrical interrup2ng capability is
dened as the product of the rated short-circuit current 2mes
rated maximum voltage/opera2ng voltage.

An example will clarify.
11/5/13
220
Example

A 69-kV circuit breaker having a voltage range factor of K = 1.21
and a con2nuous current ra2ng of 1200 A has a rated short-
circuit current of 19,000 A at the maximum rated voltage of 72.5
kV. Determine the maximum symmetrical interrup2ng capability
of the breaker and explain its signicance at lower opera2ng
voltages.
11/5/13
221
Example

The maximum symmetrical interrup2ng capability is given by

K rated short-circuit current = 1.21 19,000 = 22,900 A

This value of symmetrical interrup2ng current must not be
exceeded. From the deni2on for K:

Lower limit of opera2ng voltage = (rated maximum voltage)/K

=72.5/1.21 ~60 kV

This means that in the opera2ng range 72.5 60 kV, the
symmetrical interrup2ng current may exceed the rated short-
circuit current of 19,000 A, but it is limited to 22,900 A.
11/5/13
222
Example

For example, at 66 kV, the interrup2ng current can be

72.5/66 19,000 = 2,871 A

Breakers of the 115-kV class and higher have a K equal to 1.0
11/5/13
223

A simplied procedure for calcula2ng the symmetrical short-
circuit current, called the E/X method.

It disregards all resistance, all sta2c load, and all prefault current.

Subtransient reactance is used for generators in the E/X method,
and for synchronous motors the recommended reactance is the
Xd of the motor 2mes 1.5, which is the approximate value of the
transient reactance Xd of the motor.

Induc2on motors below 50 hp are neglected, and various
mul2plying factors are applied to the Xd of larger induc2on
motors depending on their size. If no motors are present,
symmetrical short-circuit current equals subtransient current.
11/5/13
224

The impedance by which the voltage Vf at the fault is divided to
nd short-circuit current must be examined when the E/X
method is used.

In specifying a breaker for bus k, this impedance is Zkk of the bus
impedance matrix with the proper machine reactances since the
short-circuit current is expressed by:

Vf

I f =

Z kk

If the ra2o of X/R of this impedance is 15 or less, a breaker of the
correct voltage and KVA may be used if its interrup2ng current
ra2ng is equal to or exceeds the calculated current.
11/5/13
225

If the X/R ra2o is unknown, the calculated current should be no
more than 80% of the allowed value for the breaker at the
exis2ng bus voltage.

The ANSI applica2on guide species a corrected method to
account for ac and dc 2me constants for the decay of the current
amplitude if the X/R ra2o exceeds 15. The corrected method also
considers breaker speed.

Once again an example will help.
11/5/13
226
Motors
Example A 25,000-kVA 13.8-kV generator with Xd = 15% is

connected through a transformer to a bus which supplies four
iden2cal motors. The subtransient reactance Xd of each motor
is 20% on a base of 5000 kVA, 6.9 kV. The three-phase ra2ng of
the transformer is 25,000 kVA, 18.8/6.9 kV, with a leakage
reactance of 10%. The bus voltage at the motors is 6.9 kV when
a three-phase fault occurs at point P. For the fault specied,
determine:
a. the subtransient current in the fault
b. the subtransient current in breaker A
c. the symmetrical short-circuit interrup2ng current in the fault
and a breaker A.
Generator
P
A
11/5/13
227
Example

a) For base quan22es of 25,000 kVA, 13.8 kV in the generator
circuit, the base for the motors is 25,000 kVA, 6.9 kV.

The subtransient reactance of each motor is

25,000

X d = 0.20
=1.0 pu
5,000

(see handout on Base Quan22es, specically Change of Base)

We can draw the reactance diagram:
11/5/13
228
Example

For a fault at point P:
1
j0.15
j1.0
j1.0
j0.10
j1.0
j1.0
11/5/13
229
Example

What is Zth?
1
j0.15
j1.0
j1.0
j0.10
j1.0
j1.0
I f
11/5/13
Fault
230
Example

What is Zth?
1
j0.15
j1.0
j1.0
j0.10
j1.0
j1.0
Zth
j1.0
Zth =
( j0.15 + j0.10 ) = j0.125
4
Vf
1
I f =
=
= j8.0
Zth j0.125
11/5/13
pu
231
Motors
Example

25,000
= 2090 A
The base current in the 6.9 kV circuit is I base =
3 6.9

So I f = 8 2090 = 16,720 A

b) The current in breaker A comes from the generator and three
of the four motors.
Generator
A
11/5/13
232
Example

The generator contributes a current of (by current divider):

0.25
j8.0
= j4.0 pu

0.25 + 0.25

Each motor contributes 25% of the remaining fault current, or
j 1.0 per unit each.

The current through breaker A is: I A = j4.0 + 3 j1.0 = j7.0 pu

or
7 2090 = 14,630 A
11/5/13
233
Example

c) To determine the current to be interrupted by breaker A,
replace the subtransient reactance of j1.0 by the transient
reactance of j1.5 in the motor circuits, which is the factor called
for in the E/X method.

j1.5
This changes Zth: Zth =
j0.15 + j0.10
4
0.375 0.25
= j
0.375 + 0.25
= j0.15
11/5/13
234
Example

The generator contributes a current (again by current divider):

1.0
1.5 4
1.0
0.375
I gen =
= j4.0

j0.15 1.5 4 + 0.25 j0.15 0.625

and each motor contributes a current of:

1 1.0
0.25

I motor =
= j0.67
4 j0.15 0.625

The symmetrical short-circuit current to be interrupted is
I interrupt = j4.0 + 4 ( j0.67 ) 2090 = 12,560 A
11/5/13
235
Generator
Motors
Example

Suppose that all the breakers connected to the bus are rated on
the basis of the current into a fault on the bus. The the short-
circuit current interrup2ng ra2ng of the breakers connected to
the 6.9 kV bus must be at least:
4 + 4 0.67 = 6.67 pu or 6.67 2090 = 13,940 A
11/5/13
236
Example

A 14.4-kV circuit breaker has a rated maximum voltage of 15.5kV
and a K of 2.67. At 15.5 kV its rated short-circuit interrup2ng
current is 8900 A. This breaker is rated for 2.67 8900 = 23,760 A at
a voltage of 15.5/2.67 = 5.8 kV.

This current is the maximum that can be interrupted even though
the breaker may be in a circuit or lower voltage.

15.5
The short-circuit ra2ng at 6.9 kV is
8900 = 20,000 A
6.9

The required capability of 13,940 A is well below 80% of 20,000 A,
and the breaker is suitable with respect to short-circuit current.
11/5/13
237
Example

Note that the short-circuit current could also have been found using
the bus impedance matrix.
1
j1.5
j0.15
j1.5
j0.10
j1.5
j1.5
P
0
Ybus
1
j1.5
+
j0.1 4
=
1
j0.1
11/5/13
j12.67
j10
j10
j16.67
1
1

+
j0.1 j0.15
j0.1
Z bus
j16.67
j10
j12.67
j10
=
111.11
238
j16.67
Example
j10

j12.67 j0.15
j10
j0.90
Z bus =
=

111.11
j0.090 j0.114

(try this using the Zbus building algorithm)
Z11 Z12 = j0.15 j0.09
+
1.0
Z12 = Z 21 = j0.09
11/5/13
Z 22 Z 21 = j0.114 j0.09
S1
S2
I sc
239
Example

Closing S1 with S2 open represents a fault on bus 1.

The symmetrical short-circuit interrup2ng current in a three-phase
fault at bus 1 is:
1.0

I sc =
= j6.67 pu
j0.15

which agrees with our earlier calcula2on.

Zbus also gives us us V2 with the fault on bus 1:

V2 = V f Z 21 I sc = 1.0 ( j6.67 ) j0.09 = 0.4 pu

and since the admiKance between buses 1 and 2 is j10, the current
into the fault from the transformer is
( 0.4
0 )( j10 ) = j4.0 pu again agrees.
11/5/13
240
Example

We also immediately know the short-circuit current in a three-
phase fault at bus 2. With S1 open and S2 closed,

1.0

I sc =
= j8.77 pu
j0.114

This simple example shows the value of Zbus where the eects of a
fault at a number of buses are to be studied.
11/5/13
241
Summary

The current owing immediately aTer a fault occurs in a power
network is determined by the impedances of the network
components and the synchronous machines. The ini2al
symmetrical rms fault current can be determined by
represen2ng each machine by its subtransient reactance in
series with its subtransient internal voltage . Subtransient
currents are larger than the transient and steady-state currents.

Circuit breakers have ra2ngs determined by the maximum
instantaneous current which the breaker must withstand and
later interrupt. Interrup2ng currents depend on the speed of
breaker opera2on. Proper selec2on and applica2on of circuit
breakers should follow the recommenda2ons of ANSI standards.
11/5/13
242
Summary

Simplifying assump2ons usually made in industry-based fault
studies are:

All shunt connec2ons from system buses to the reference
node (neutral) can be neglected in the equivalent circuits
represen2ng transmission lines and transformers.
Load impedances are much larger than those of network
components, and so they can be neglected in system
modeling.
All buses of the system have rated/nominal voltage of 1.0 per
unit so that no prefault currents ow in the network.
11/5/13
243
Summary

Simplifying assump2ons usually made in industry-based fault
studies are:

Synchronous machines can be represented by voltage of 1.0
per unit behind subtransient or transient reactance,
depending on the speed of the circuit breakers and whether
the momentary or interrup2ng fault current is being
calculated (ANSI standards should be consulted).
The voltage-source-plus-series-impedance equivalent circuit
of each synchronous machine can be transformed to an
equivalent current-source-plus shunt-impedance model.
Then, the shunt impedances of the machine models represent
the only shunt connec2ons to the reference node.
11/5/13
244
Summary

The bus impedance matrix is most oTen used for fault current
calcula2ons. The elements of Zbus can be made available
explicitly using the Zbus building algorithm or they can be
generated from the inversion of Ybus.

Equivalent circuits based on the elements of Zbus can simplify
fault-current calcula2ons as demonstrated in the example for
the line-end fault.

Next Symmetrical Components and Sequence Networks.

11/5/13
245

Network Calculations PDF

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Network Calculations PDF

Încărcat de

Drepturi de autor:

Formate disponibile

NETWORK

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Branch and Node AdmiHances

Network Calcula2ons (c) 2013 H. Zmuda

RECALL NODAL ANALYSIS via example

Network Calcula2ons (c) 2013 H. Zmuda

RECALL NODAL ANALYSIS

0 = Yc (V1 V3 ) + Yd (V1 V2 ) + Y f (V1 V4 )

Network Calcula2ons (c) 2013 H. Zmuda

NODAL ADMITTANCE MATRIX YBus

Network Calcula2ons (c) 2013 H. Zmuda

BUS ADMITTANCE MATRIX Ybus

Network Calcula2ons (c) 2013 H. Zmuda

AN ALTERNATIVE APPROACH FOR Ybus FORMATION

Reference Node (Bus)

Network Calcula2ons (c) 2013 H. Zmuda

AN ALTERNATIVE APPROACH FOR Ybus FORMATION

Network Calcula2ons (c) 2013 H. Zmuda

AN ALTERNATIVE APPROACH FOR Ybus FORMATION

Network Calcula2ons (c) 2013 H. Zmuda

AN ALTERNATIVE APPROACH FOR Ybus FORMATION

Network Calcula2ons (c) 2013 H. Zmuda

AN ALTERNATIVE APPROACH FOR Ybus FORMATION

Network Calcula2ons (c) 2013 H. Zmuda

Develop the nodal admiKance matrix for each of the network

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Nodal Matrix YBus

Network Calcula2ons (c) 2013 H. Zmuda

Nodal Matrix YBus

j4.0 j8.0 j5.0

Network Calcula2ons (c) 2013 H. Zmuda

Nodal Matrix YBus

Network Calcula2ons (c) 2013 H. Zmuda

Mutually Coupled Branches

Network Calcula2ons (c) 2013 H. Zmuda

Mutually Coupled Branches

Network Calcula2ons (c) 2013 H. Zmuda

Mutually Coupled Branches

Network Calcula2ons (c) 2013 H. Zmuda

Mutually Coupled Branches

Network Calcula2ons (c) 2013 H. Zmuda

Mutually Coupled Branches

Network Calcula2ons (c) 2013 H. Zmuda

Mutually Coupled Branches

Network Calcula2ons (c) 2013 H. Zmuda

Mutually Coupled Branches

Network Calcula2ons (c) 2013 H. Zmuda

Mutually Coupled Branches

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

Network Calcula2ons (c) 2013 H. Zmuda

I m = YaVm YaVn + YMVm YMVq = (Ya + YM )Vm YaVn YMVq