7-Unsymmetrical Faults PDF

UNSYMMETRICAL
FAULTS

updated 11/11/13
11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 1

Introductory Comments

Most of the faults that occur on power systems are unsymmetrical
faults, which may consist of unsymmetrical short circuits,
unsymmetrical faults through impedances, or open conductors.

Unsymmetrical faults occur as single line-‐to-‐ground faults, line-‐to-‐
line faults, or double line-‐to-‐ground faults. The path of the fault
current from line to line or line to ground may or may not contain
impedance. One or two open conductors result in unsymmetrical
faults, through either the breaking of one or two conductors or the
acHon of fuses and other devices that may not open the three
phases simultaneously.

Introductory Comments

Since any unsymmetrical fault causes unbalanced currents to flow in
the system, the method of symmetrical components is very used in
the analysis to determine the currents and voltages in all parts of
the system aLer the occurrence of the fault.

We will consider faults on a power system by applying Thévenin's
theorem, which allows us to find the current in the fault by
replacing the enHre system by a single generator and series
impedance, and we will show how the bus impedance matrix is
applied to the analysis of unsymmetrical faults.

Unsymmetrical Faults

In the derivaHon of equaHons for the symmetrical components of
currents and voltages in a general network the currents flowing out
of the original balanced system from phases a , b, and c at the fault
point will be designated as Ia, lb, and lc, respecHvely. We can
visualize these currents by as follows:
a
I fa
b
I fb
c
I fc
This shows the three lines a, b, and c of the three-‐phase system at
the part of the network where the fault occurs. The flow of current
from each line into the fault is indicated by arrows shown beside
hypotheHcal stubs connected to each line at the fault locaHon.

Appropriate connecHons of the stubs represent the various types of
fault. For instance, direct connecHon of stubs b and c produces a
line-‐to-‐line fault through zero impedance.

The current in stub a is then zero, and lb equals -‐ lc.

The line-‐to-‐ground voltages at any bus j of the system during the
fault will be designated Vja, Vjb and Vjc and we shall conHnue to use
superscripts 1, 2, and 0, respecHvely, to denote posiHve-‐, negaHve-‐,
and zero-‐sequence quanHHes.

Thus, for example, V(1)ja, V(2)jb and V(0)jc will denote, respecHvely, the
posiHve-‐, negaHve-‐, and zero-‐sequence components of the line-‐to-‐
ground voltage Vja at bus j during the fault.

The line-‐to-‐neutral voltage of phase a at the fault point before the
fault occurs will be designated simply by Vf, which is a posiHve-‐
sequence voltage since the system is balanced.

We considered the prefault voltage Vf previously when calculaHng
the currents in a power system with a symmetrical three-‐phase
fault applied.


Consider a single-‐line diagram of a power system containing two
synchronous machines. This simple system is sufficiently general
that the equaHons derived are applicable to any balanced system
regardless of the complexity. The point where a fault is assumed to
occur is marked P, and in this example it is called bus k on the
single-‐line diagram and in the sequence networks.

P

k

Single line diagram of a balanced three-‐phase system

P

k

(1)
(1) I fa
I fa P
k +
P + k (1)

Z kk
Vka( )
1
+
Vf Vf
−
−
−
Reference
Thévenin Equivalent of

PosiKve-‐Sequence Network
PosiKve-‐Sequence Network

P

k

( 2)
( 2) I fa
I fa P
k +
P k ( 2)

Z kk
Vka( )
2

−

Reference

NegaKve-‐Sequence Network
NegaKve-‐Sequence Network

P

k

(0)
(0) I fa
I fa P
k +
P k (0)

Z kk
Vka( )
0

−
Reference
Zero-‐Sequence Network
Zero-‐Sequence Network

Machines are represented by their subtransient internal voltages in
series with their subtransient reactances when subtransient fault
condiHons are being studied.

Previously we used the bus impedance matrix composed of
posiHve-‐sequence impedances to determine currents and voltages
upon the occurrence of a symmetrical three-‐phase fault. The
method can be easily extended to apply to unsymmetrical faults by
realizing that the negaHve-‐ and zero-‐sequence networks also can be
represented by bus impedance matrices. The bus impedance matrix
will now be wriben symbolically for the posiHve-‐, negaHve-‐, and
zero-‐sequence networks in the following form…


⎡ Z11( )
0,1,2
Z12( )  Z1k( )  Z1N
0,1,2 0,1,2 (0,1,2) ⎤
⎢ ⎥
⎢ (0,1,2)
Z 21 (0,1,2)
Z 22 (0,1,2)
 Z 2k  Z 2( N ) ⎥
0,1,2
⎢ ⎥
(0,1,2) ⎢       ⎥
Ζ bus = ⎢
(0,1,2) (0,1,2) (0,1,2) (0,1,2) ⎥
⎢ Z k1 Zk 2  Z kk  Z kN ⎥
⎢ ⎥
⎢       ⎥
⎢ Z N( 1 ) Z N( 2 )  Z Nk(0,1,2) (0,1,2) ⎥
0,1,2 0,1,2
 Z NN
⎣ ⎦


The Thévenin equivalent circuit between the fault point P and the
reference node in each sequence network can be used for the
analysis.

As before, the voltage source in the posiHve-‐sequence network and
its Thévenin equivalent circuit is Vf, the prefault voltage to neutral
at the fault point P, which happens to be bus k in this illustraHon.
The Thévenin impedance measured between point P and the
reference node of the posiHve-‐sequence network is Z(1)kk, and its
value depends on the values of the reactances used in the network.

Recall that subtransient reactances of generators and 1.5 Hmes the
subtransient reactances (or else the transient reactances) of
synchronous motors are the values used in calculaHng the
symmetrical current to be interrupted.

There are no negaHve-‐ or zero-‐sequence currents flowing before
the fault occurs, and the prefault voltages are zero at all buses of
the negaHve-‐ and zero-‐sequence networks. Therefore, the prefault
voltage between point P and the reference node is zero in the
negaHve-‐ and zero-‐sequence networks and no electromoHve forces
(emfs) appear in their Thévenin equivalents.

The negaHve-‐ and zero-‐sequence impedances between point P at
bus k and the reference node in the respecHve networks are
represented by the the impedances Z(2)kk and Z(0)kk, the Diagonal
elements of Z(2)bus and Z(0)bus, respecHvely.


Since Ifa is the current flowing from the system into the fault, its
symmetrical components flow out of the respecHve sequence
networks and their equivalent circuits at point P, as shown.

Thus, the currents –I(1)fa, –I(2)g and –I(0)fc represent injected currents
into the faulted bus k of the posiHve-‐, negaHve-‐, and zero-‐sequence
networks due to the fault.

These current injecHons cause voltage changes at the buses of the
posiHve-‐, negaHve-‐, and zero-‐sequence networks, which can be
calculated from the bus impedance matrices in the manner similare
to what we have done before.


For instance, due to the injecHon –I(1)fa into bus k, the voltage
changes in the posiHve-‐sequence network of the N-‐bus system are
given in general terms by:
⎡ ⎡ −Z1k( ) I (fa) ⎤⎥
ΔV1a ⎥ ⎡⎢
(1) ⎤ (1) (1) (1) (1) ⎤ 1 1
Z11 Z12  Z1k  Z1N ⎥ ⎡
⎢ 0 ⎤ ⎢
⎢ ⎢ ⎥ ⎢ (1) (1) ⎥
ΔV2a( ) ⎥ ⎢ () () ()
Z 2( N) ⎥ ⎢ 0 ⎥ ⎢ −Z 2k
1 1 1 1 1
Z 21 Z 22  Z 2k  I fa ⎥
⎢ ⎥ ⎢ ⎥
⎢  ⎥ ⎢       ⎥⎢  ⎥ ⎢  ⎥
⎢ ⎥ =⎢ ⎥ ⎢ (1) ⎥ = ⎢ ⎥
ΔVka( ) ⎥ ⎢ ()
Z k( 2)  Z kk( )  ( ) ⎢ −I ⎢ (1) (1) ⎥
1 1 1 1 1
⎢ Z k1 Z kN ⎥ fa ⎥ −Z kk I fa
⎢ ⎥ ⎢ ⎥⎢  ⎥ ⎢ ⎥
⎢  ⎥ ⎢       ⎥⎢ ⎥ ⎢  ⎥
(1) ⎥ ⎢ (1) ⎥ ⎢⎣ 0 ⎥⎦ ⎢ (1) (1) ⎥
⎢ Z N( 1) Z N( )2  Z Nk
()
1 1 1
ΔVNa  Z NN −Z Nk I fa ⎥
⎣ ⎦ ⎣ ⎦ ⎢⎣ ⎦


Once again, it is industry pracHce to regard all prefault currents as
being zero and to designate the voltage Vf as the posiHve-‐sequence
voltage at all buses of the system before the fault occurs. Using
superposiHon, the total posiHve-‐sequence voltage of phase a at
each bus during the fault is:
⎡ (1) ⎤ ⎡ V f − Z1k( ) I (fa) ⎤⎥
V1a ⎥ ⎡ Vf ⎤ ⎡ (1) ⎤ 1 1
⎢ ΔV1a ⎥ ⎢
⎢ ⎥ ⎢
⎢ ⎢ (1) (1) ⎥
V2a( ) ⎥ ⎢ Vf ⎥ ⎢ ΔV2a( ) ⎥ ⎢ V f − Z 2k
1 1
I fa ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢  ⎥ ⎢  ⎥ ⎢  ⎥ ⎢  ⎥
⎢ ⎥ =⎢ ⎥ +⎢ ⎥ =⎢ ⎥
Vka( ) ⎥ ⎢ Vf ΔVka( ) ⎥ ⎢ (1) (1) ⎥
1 1
⎢ ⎥ ⎢ V f − Z kk I fa
⎢ ⎥  ⎥ ⎢ ⎥ ⎢ ⎥
⎢  ⎥ ⎢  ⎥ ⎢  ⎥
⎢ ⎥ ⎢ (1) ⎥ ⎢
⎢ (1) ⎥
VNa ⎢ Vf ⎥ ⎢ ΔVNa (1) (1) ⎥
⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎢⎣ V f − Z Nk I fa ⎥
⎦


This is the same equaKon as found for symmetrical faults, the only
difference being the added superscripts and subscripts denoKng
the posiKve-‐sequence components of the phase a quanKKes.


The negaHve-‐ and zero-‐sequence voltage changes due to the
fault at bus k of the N-‐bus system are similarly wriben with the
superscripts changed accordingly. Because the prefault voltages are
zero in the negaHve-‐ and zero-‐sequence networks, the voltage
changes express the total negaHve-‐ and zero-‐sequence voltages
during the fault, namely,
⎡ ( 2) ⎤ ⎡ −Z1k( ) I (fa) ⎤⎥
2 2
⎡ ⎡
(0) ⎤ −Z1k( ) I (fa) ⎤⎥
0 0
⎢ V1a ⎥ ⎢ ⎢ V1a ⎥ ⎢
⎢ ( 2) ⎥ ⎢ ( 2) ( 2) ⎥
−Z 2k I fa ⎥ ⎢ (0) ⎥ ⎢ (0) (0) ⎥
−Z 2k I fa ⎥
V2a V2a
⎢ ⎥ ⎢ ⎢ ⎥ ⎢
⎢  ⎥ ⎢  ⎥ ⎢  ⎥ ⎢  ⎥
⎢ =⎢ ⎥, =⎢ ⎥
( 2) ⎥ ⎢ ( 2) ( 2) ⎥ ⎢ (0) ⎥ ⎢ (0) (0) ⎥
⎢ Vka ⎥ −Z kk I fa ⎢ Vka ⎥ −Z kk I fa
⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢  ⎥ ⎢  ⎥ ⎢  ⎥ ⎢  ⎥
⎢ ( 2) ⎥ ⎢ ( 2) ( 2) ⎥ ⎢ (0) ⎥ ⎢ (0) (0) ⎥
VNa −Z Nk I fa ⎥ VNa −Z Nk I fa ⎥
⎣ ⎦ ⎢⎣ ⎦ ⎣ ⎦ ⎢⎣ ⎦

When the fault is at bus k, note that only the entries in columns k of
Z(2)bus and Z(0)bus are involved in the calculaHons of negaHve-‐ and
zero-‐sequence voltages.

Thus, knowing the symmetrical components I(0)fa, I(1)fa and I(2)fa of
the fault currents at bus k, we can determine the sequence voltages
at any bus j of the system from the jth rows of
⎡ ⎡ V f − Z1k( ) I (fa) ⎤⎥ ⎡ −Z1k( ) I (fa) ⎤⎥ ⎡ −Z1k( ) I (fa) ⎤⎥
V1a( ) ⎤⎥ ⎢ ⎡ V1a( ) ⎤⎥ ⎢ ⎡ V1a( ) ⎤⎥ ⎢
1 1 1 2 2 2 0 0 0
⎢ ⎢ ⎢
⎢ ⎢ (1) (1) ⎥ ⎢ ( 2) ( 2) ⎥ ⎢ (0) (0) ⎥
V2a( ) ⎥ ⎢ V f − Z 2k ⎢ V2a( ) ⎥ ⎢ −Z 2k ⎢ V2a( ) ⎥ ⎢ −Z 2k
1 2 0
I fa ⎥ I fa ⎥ I fa ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢  ⎥ ⎢  ⎥ ⎢  ⎥ ⎢  ⎥ ⎢  ⎥ ⎢  ⎥
⎢ =⎢ ⎥, =⎢ ⎥, =⎢ ⎥
(1) ⎥ ⎢ (1) (1) ⎥ ⎢ ( 2) ⎥ ⎢ ( 2) ( 2) ⎥ ⎢ (0) ⎥ ⎢ (0) (0) ⎥
⎢ Vka ⎥ V f − Z kk I fa ⎢ Vka ⎥ −Z kk I fa ⎢ Vka ⎥ −Z kk I fa
⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢  ⎥ ⎢  ⎥ ⎢  ⎥ ⎢  ⎥ ⎢  ⎥ ⎢  ⎥
⎢ (1) ⎥ ⎢ (1) (1) ⎥ ⎢ ( 2) ⎥ ⎢ ( 2) ( 2) ⎥ ⎢ (0) ⎥ ⎢ (0) (0) ⎥
VNa V f − Z Nk I fa ⎥ VNa −Z Nk I fa ⎥ VNa −Z Nk I fa ⎥
⎣ ⎦ ⎢⎣ ⎦ ⎣ ⎦ ⎢⎣ ⎦ ⎣ ⎦ ⎢⎣ ⎦

That is, during the fault at bus k the voltages at any bus j are:

(0) (0) (0) (1) (1) (1) ( 2) ( 2) ( 2)
V ja
= −Z jk fa
I , V ja
= V f
− Z I
jk fa
, V ja
= −Z jk fa
I

If the prefault voltage at bus CD is not Vf, then simply replace Vf in
by the actual value of the prefault (posiHve-‐sequence) voltage at
that bus. Since Vf is by definiHon the actual prefault voltage at the
faulted bus k, we always have at that bus:

(0) (0) (0) (1) (1) (1) ( 2) ( 2) ( 2)
Vka = −Z kk I fa , Vka = V f − Z kk I fa , Vka = −Z kk I fa

and these are the terminal voltage equaHons for the Thévenin
equivalents of the sequence networks previously shown.


It is important to remember that the currents I(0)fa, I(1)fa and I(2)fa are
symmetrical-‐component currents in the stubs hypotheHcally
abached to the system at the fault point.

These currents take on values determined by the parHcular type of
fault being studied, and once they have been calculated, they
can be regarded as negaHve injecHons into the corresponding
sequence networks.


If the system has Δ-‐Y transformers, some of the sequence voltages
may have to be shiLed in phase angle before being combined with
other components to form the new bus voltages of the faulted
system. There are no phase shiLs involved in

Vka( ) = −Z kk( ) I (fa) , Vka( ) = V f − Z kk( ) I (fa) , Vka( ) = −Z kk( ) I (fa)
0 0 0 1 1 1 2 2 2

when the voltage Vf at the fault point is chosen as reference, which
is customary.


In a system with Δ-‐Y transformers the open circuits encountered in
the zero-‐sequence network requires some care in the Zbus building
algorithm.

Consider, for instance, the solidly grounded Δ-‐Y transformer
connected between buses m and n as shown along with their
posiHve and zero-‐sequence circuits:
Z

m m n m Z n
n Reference Reference

Δ-‐Y transformer with PosiKve-‐sequence Zero-‐sequence
leakage impedance Z circuit circuit

The negaHve-‐sequence circuit is the same as the posiHve-‐sequence
circuit.
11/11/13
Unsymmetrical Faults (c) 2013 H. Zmuda 24

It is straighlorward using the circuit representaHons shown to
generate the bus impedance matrices Zbus (0,1,2). This will be done
subsequently.

Suppose, however, that we wish to represent removal of the
transformer connecHons from bus n in a computer algorithm which
cannot make use of circuit (schemaHc) representaHons.

We can easily undo the connecHons to bus n in the posiHve-‐ and
negaHve-‐sequence networks by applying the building algorithm the
Zbus (1,2) matrices in the usual manner, i.e., by adding the negaHve of
the leakage impedance Z between buses m and m in the posiHve-‐
and negaHve-‐sequence networks. (Next slide)


PosiKve-‐sequence circuit:

-‐ Z
Z
m n è m n
Reference Reference


This strategy does not apply to the zero-‐sequence matrix Zbus (0) if it
has been formed directly from the schemaHc representaHon shown.
Adding -‐ Z between buses m and m does not remove the zero-‐
sequence connecHon from bus n.
-‐ Z
m Z n
Reference


To permit uniform procedures for all sequence networks, one
strategy is to include an internal node p, as shown below. Note that
the leakage impedance is now subdivided into two parts between
node p and the other nodes as shown. ConnecHng –Z/2 between
buses n and p in each of the sequence circuits will open the
transformer connecHons to bus n.
Z
m n m Z n
Reference Reference
è
è
Z/2 p Z/2 p Z/2
m n m Z/2 n
Reference Reference

ConnecHng –Z/2 between buses n and p in each of the sequence
circuits will open the transformer connecHons to bus n.
-‐ Z/2 -‐ Z/2
Z/2 Z/2 Z/2

p
m p n m n
Z/2
Reference Reference
è
è
Z/2
p
m p n m n
Z/2
Reference Reference


The faults to be discussed in succeeding secHons may involve
impedance Zf between lines and from one or two lines to ground.

When Zf = 0, we have a direct short circuit, which is called a bolted
fault .

Although such direct short circuits result in the highest value of fault
current and are therefore the most conservaHve values to use when
determining the effects of anHcipated faults, the fault impedance is
seldom zero.


Most faults are the result of insulator flashovers, where the
impedance between the line and ground depends on the resistance
of the arc, of the tower itself, and of the tower fooHng if ground
wires are not used.

Tower-‐fooHng resistances form the major part of the resistance
between line and ground and depend on the soil condiHons. The
resistance of dry earth is 10 to 100 Hmes the resistance of swampy
ground.


ConnecHons of the hypotheHcal stubs for faults through impedance
Zf are as follows:

a

I fa
Zf

b
I fb Zf

c

I fc Zf

Three-‐Phase Fault

Zf are as follows:

a

I fa
Zf

b
I fb

c

I fc

Single Line-‐to-‐Ground Fault

Zf are as follows:

a

I fa

b
I fb Zf

c

I fc

Line-‐to-‐Line Fault

Zf are as follows:

a

I fa

b
I fb Zf

c

I fc

Double Line-‐to-‐Ground Fault

Other types of faults:

a
Ia

b
Ib

c

Ic
Open-‐Conductor Faults


Other types of faults:

a
Ia

b
Ib

c

Ic


A balanced system remains symmetrical aLer the occurrence of a
three-‐phase fault having the same impedance between each line
and a common point. Only posiHve-‐sequence currents flow. With
the fault impedance Zf equal in all phases, as in a three-‐phase fault,
we simply add impedance Zf to the usual (posiHve-‐sequence)
Thévenin equivalent circuit of the system at the fault bus k and
calculate the fault current from the equaHon:

(1) Vf
I fa = (1)
Z kk + Z f

For each of the other types of faults, formal derivaHons of the
equaHons for the symmetrical-‐component currents follow. In each
case the fault point P is designated as bus k.
Example

Two synchronous machines are connected through three-‐phase
transformers to the transmission line as shown.
Machine 1
1 T1 2 3 T2 4
Machine 2

The raHngs and reactances of the machines and transformers are:
Machine 1 & 2: 100MVA, 20 kV, Xd” = X1 = X2 = 20%
X0 = 4%, Xn = 5%

Transformers T1 & T2: 100 MVA, 20Δ/345Y kV, X = 8%

On a chosen base of 100 MVA, 345 kV in the transmission line
circuit the line reactances are X1 = X2 = 15% and X0 = 50%.
Determine Zbus for the sequence networks.
Example

PosiHve-‐ and negaHve-‐sequence circuit:
1
(2)
2
(3)
3
(4)
4
(5)
j0.04 j0.04 j0.15 j0.04 j0.04
j0.20 (1)

j0.20
+
V +
f Vf
− Reference −

X0 = 4%, Xn = 5%
Transformers T1, T2: 100 MVA, 20Δ/345Y kV, X = 8% (split as 4% + 4%)
Transmission line reactances are X1 = X2 = 15% and X0 = 50%.
Example

1 (2) 2 (3) 3 (4) 4
(5)
j0.20 j0.04 j0.04 j0.15 j0.04 j0.04
(1) j0.20

+ +
V
f − Vf
−

Reference
Add branch (1): Z1( ) = ⎡⎣ j0.20 ⎤⎦
1,2

Add branch (2):
⎡ j0.20 j0.20 ⎤ ⎡ j0.20 j0.20 ⎤
(1,2)
Z2 =⎢ ⎥=⎢ ⎥
⎢⎣ j0.20 j0.20 + j0.08 ⎥⎦ ⎢⎣ j0.20 j0.28 ⎥⎦

Example

1 (2) 2 (3) 3 (4) 4
(5)
j0.20 j0.04 j0.04 j0.15 j0.04 j0.04
(1) j0.20

+ +
V
f − Vf
−

Reference
Add branch (3):
⎡ j0.20 j0.20 ⎤ ⎡ j0.20
j0.20 j0.20 j0.20 ⎤
(1,2) ⎢ ⎥ ⎢ ⎥
Z3 = ⎢ j0.20 j0.28 j0.28 ⎥ = ⎢ j0.20 j0.28 j0.28 ⎥
⎢ j0.20 j0.28 j0.28 + j0.15 ⎥⎦ ⎢⎣ j0.20 j0.28 j0.43 ⎥⎦
⎣

Example

1 (2) 2 (3) 3 (4) 4
(5)
j0.20 j0.04 j0.04 j0.15 j0.04 j0.04
(1) j0.20

+ +
V
f − Vf
−

Reference
Add branch (4):
⎡ j0.20 j0.20 j0.20 ⎤ ⎡
j0.20 j0.20 j0.20 j0.20 j0.20 ⎤
⎢ ⎥ ⎢ ⎥
⎢ j0.20 j0.28 j0.28 j0.28 ⎥ ⎢ j0.20 j0.28 j0.28 j0.28 ⎥
Z 4( ) = ⎢
1,2
⎥ =⎢
⎢ j0.20 j0.28 j0.43 j0.43 ⎥ ⎢ j0.20 j0.28 j0.43 j0.43 ⎥⎥
⎢ j0.20 j0.28 j0.43 j0.43+ j0.08 ⎥⎦ ⎢⎣ j0.20 j0.28 j0.43 j0.51 ⎥⎦
⎣

Example

1 (2) 2 (3) 3 (4) 4
(5)
j0.20 j0.04 j0.04 j0.15 j0.04 j0.04
(1) j0.20

+ +
V
f − Vf
−

Reference
Add branch (5):
⎡ j0.20 j0.20 j0.20 j0.20 ⎤
⎢ ⎥ Z ik Z kj
⎢ j0.20 j0.28 j0.28 j0.28 ⎥ new
= Z ij −
Z 4( ) = ⎢ Z
1,2
j0.43 ⎥⎥ Z kk + Z b
ij
⎢ j0.20 j0.28 j0.43
⎢ j0.20 j0.28 j0.43 j0.51 ⎥⎦
⎣

Example


Example

( )
1,2
Z bus :

Example Zero-‐sequence circuit:
See Slide 196 See Slide 199
of Symmetrical of Symmetrical
Components Components
j0.04 1 j0.04 2 j0.5 3 j0.08 4 j0.04

(1) Transformer
Node Bus (3) (4) (5)
Transformer
Node Bus (6)
(2)

3X n = j0.15 j0.04 3X n = j0.15

Reference

X0 = 4%, Xn = 5%
Transformers T1 & T2: 100 MVA, 20Δ/345Y kV, X = 8%
Transmission line reactances are X1 = X2 = 15% and X0 = 50%.
Example

1 2 j0.5 3 j0.08 4
(3) (4)
(1) j0.19 j0.08 (2) j0.19 (5)
Reference

Example

1 2 j0.5 3 j0.08 4

(3) (4)

j0.19 j0.08 (2)
(1) j0.19 (5)

Reference
Add branch (1): Z1( ) = ⎡⎣ j0.19

0
⎤⎦ Add branch (3):

⎡ j0.19 0 0 ⎤
Add branch (2): ⎢ ⎥
Z3( ) = ⎢
0
0 j0.08 j0.08 ⎥
⎡ j0.19 ⎤ ⎢
Z2 (0)
=⎢
0
⎥ ⎣ 0 j0.08 j0.08 + j0.5 ⎥⎦
⎢⎣ 0 j0.08 ⎥⎦

Example

1 2 j0.5 3 j0.08 4

(3) (4)

j0.19 j0.08
(1) (2) j0.19 (5)

Reference
Add branch (5):

⎡ j0.19 0 0 0 ⎤
⎢ ⎥
(0) ⎢ 0 j0.08 j0.08 0 ⎥
Z bus = ⎢
⎢ 0 j0.08 j0.58 0 ⎥⎥
⎢ 0 0 0 j0.19 ⎥⎦
⎣
Example

We’ll use these in subsequent examples.
⎡ j0.19 0 0 0 ⎤
⎢ ⎥
(0) ⎢ 0 j0.08 j0.08 0 ⎥
Z bus = ⎢
⎢ 0 j0.08 j0.58 0 ⎥⎥
⎢ 0 0 0 j0.19 ⎥⎦
⎣
⎡ j0.1437 j0.1211 j0.0789 j0.0563 ⎤

⎢ ⎥
(1,2) ⎢ j0.1211 j0.1696 j0.1104 j0.0789 ⎥
Z bus = ⎢
⎢ j0.0789 j0.1104 j0.1696 j0.1211 ⎥⎥
⎢ j0.0563 j0.0789 j0.1211 j0.1437 ⎥⎦
⎣

Single Line-‐to-‐Ground Faults

The single line-‐to-‐ground fault, the most common type, is caused by
lightning or by conductors making contact with grounded
structures. A single line-‐to-‐ground fault on phase a through
impedance Zf is shown below:
a
k
I fa Z f
b
I fb
c
I fc
The relaHons to be developed will apply only when the fault is on
phase a, but any phase can be designated as phase a. The
condiHons at the fault bus k are expressed by the following
equaHons:
I fb = I fc = 0, Vka = Z f I fa

The symmetrical components of the current are, with I fb =
I fc = 0 :

⎡ I (fa0) ⎤
⎢ ⎥ ⎡ 1 1 1 ⎤ ⎡ I fa ⎤
⎢ 1 ⎥ 1 ⎢ ⎥ ⎢ ⎥ ( ) ( ) ( ) I fa
() = 0 1 2
⎢ I fa ⎥ ⎢ 0 ⎥ ⇒ I fa = I fa = I fa =
2
⎢ 1 a a ⎥
3 ⎢ ⎥ 3
⎢ ( 2) ⎥ ⎢⎣ 1 a 2
a ⎥⎦ 0
⎣ ⎦
⎢⎣ I fa ⎥⎦

But recall from Slide 21:
(0) (0) (0) (1) (1) (1) ( 2) ( 2) ( 2)
V ka
= −Z I
kk fa
, V ka
= V f
− Z I
kk fa
, V ka
= −Z I
kk fa
Thus:
(0) (0) (0) (1) (1) (0) ( 2) ( 2) (0)
Vka = −Z kk I fa , Vka = V f − Z kk I fa , Vka = −Z kk I fa


Summing, and noHng that Vka = 3Z f I (fa0)

(0) (1) ( 2)
Vka = Vka + Vka + Vka

( 2 0
)
= V f − Z kk( ) + Z kk( ) + Z kk( ) I (fa) = 3Z f I (fa)
0 1 0
Solve for I (fa)

0

(0) (1) ( 2) Vf
I fa = I fa = I fa = (0) (1) ( 2)
Z kk + Z kk + Z kk + 3Z f

This has the circuit representaHon shown on the next slide.


(1)
I fa
k
+ (1) + (1)
Vf Z kk Vka
− −
I (fa)
2
k
I (fa) = I (fa) = I (fa)
0 1 2
Z kk( ) + ( 2)
2
Vka
−
(0)
I fa k
(0) + (0)
Z kk Vka
− 3Z f


The last result are the fault current equaHons parHcular to the
single line-‐to-‐ground fault through impedance Zf and they are used
with the symmetrical-‐component relaHons to determine all the
voltages and currents at the fault point P.

If the Thévenin equivalent circuits of the three sequence networks
of the system are connected in series, as shown on the previous
slide, we see that the resulHng currents and voltages saHsfy the
above equaHons – for the Thévenin impedances looking into the
three sequence networks at fault bus k are then in series with the
fault impedance 3Zf and the prefault voltage source Vf.


With the equivalent circuits so connected, the voltage across each
sequence network is the corresponding symmetrical component of
the voltage Vka at the fault bus k, and the current injected into each
sequence network at bus k is the negaHve of the corresponding
sequence current in the fault.

The series connecHon of the Thévenin equivalents of the sequence
networks, as shown on Slide 55 is a convenient means of
remembering the equaHons for the soluHon of the single line-‐to-‐
ground fault, for all the necessary equaHons for the fault point can
be determined from the sequence-‐network connecHon.


Once the sequence components of the fault currents are known,
the components of voltages at all other buses of the system can be
determined from the bus impedance matrices of the sequence
networks according to Slide 21, namely:
V ja( ) = −Z (jk ) I (fa)

0 0 0
(1) (1) (1)

V ja = V f − Z jk I fa
V ja( ) = −Z (jk ) I (fa)
2 2 2

EXAMPLE – Single Line-‐to-‐Ground Faults

Two synchronous machines are connected through three-‐phase
transformers to the transmission line as shown.

Machine 1
1 T1 2 3 T2 4
Machine 2

X0 = 4%, Xn = 5%

Transformers T1 & T2: 100 MVA, 20Y/345Y kV, X = 8%


Both transformers are solidly grounded on on two sides. On a
chosen base of 100 MVA, 345 kV in the transmission line circuit the
line reactances are X1 = X2 = 15 % and X0 = 50 %.

The system is operaHng at nominal voltage without prefault
currents when a bolted (Zf = 0) single line-‐to -‐ground fault occurs on
phase A at bus 3.

Using the bus impedance matrix for each of the three sequence
networks, determine the subtransient current to ground at the
fault, the line-‐to-‐ground voltages at the terminals of machine 2, and
the subtransient current out of phase c of machine 2.


The system is the same as on Slide 39, except that the transformers
are now Y-‐Y connected. Therefore, we can conHnue to use Zbus (1,2)
from Slide 51, however, because the transformers are solidly
grounded on both sides, the zero-‐sequence network is fully
connected, as shown below, and has the bus impedance matrix
1 2 3 4
j0.04 j0.08 j0.15 j0.08 j0.04
j0.15 j0.15
Reference


The bus impedance matrix is:
1 2 3 4
j0.04 j0.08 j0.15 j0.08 j0.04
j0.15 j0.15
Reference
⎡ j0.1553 j0.1407 j0.0493 j0.0347 ⎤

⎢ ⎥
(0) ⎢ j0.1407 j0.1999 j0.0701 j0.0493 ⎥
Z bus = ⎢
⎢ j0.0493 j0.0701 j0.1999 j0.1407 ⎥⎥
⎢ j0.0347 j0.0493 j0.1407 j0.1553 ⎥⎦
⎣
EXAMPLE – Single Line-‐to-‐Ground Faults ⎡
j0.1553 j0.1407 j0.0493 j0.0347 ⎤
Since the line-‐to-‐ground fault is at bus 3: ( ) ⎢⎢ j0.1407 j0.1999 j0.0701 j0.0493 ⎥⎥
0
Z =⎢
()
1 ⎢
bus
j0.0493 j0.0701 j0.1999 j0.1407 ⎥
⎥
I fA
k ⎢ j0.0347 j0.0493 j0.1407 j0.1553 ⎥
⎣ ⎦
⎡ j0.1437 j0.1211 j0.0789 j0.0563 ⎤
() 1
+ (1) ⎢ ⎥
V f = 1.0 ±
Z 33 V3a Z( ) = ⎢
⎢ j0.1211
1,2 j0.1696 j0.1104 j0.0789 ⎥
j0.0789 j0.1104 j0.1696 j0.1211 ⎥⎥
−
bus
⎢
⎢ j0.0563 j0.0789 j0.1211 j0.1437 ⎥
⎣ ⎦
( 2)
I fA k
I fA Vf
(0) (1) ( 2)
I fA = I fA = I fA = (0) (1) ( 2)
( 2) + ( 2) 3 Z + Z + Z + 3Z f
Z33 V3a
kk kk kk
1
− = = − j1.8549
j0.1999 + j1.696 + j1.696 + 0
I (fA)
0
k

Z (0) + V ( 0) The total current in the fault is:
3a
33
− 3Z f = 0
⇒ I fA = 3I (fA) = − j5.5648
0


Since the base current in the high-‐voltage transmission line is:

100,000
I base = = 167.35 amps
3 × 345

Then: I fA = − j5.5648
× 167.35
= 931∠270°
amps

The phase-‐a sequence voltages at bus 4, the terminals of machine 2,
are (from Slide 21 with k = 3 and j = 4):

V
4a
(0)
= −Z (0) (0)
( )(
I = − j0.1407 − j1.8549 = −0.261
43 fA )
(
V (1) = V − Z (1) I (1) = 1− j0.1211 − j1.8549 = −0.7754
)( )
4a f 43 fA per unit
V4a( ) = −Z 43
2 ( 2) ( 2)
( )(
I fA = − j0.1211 − j1.8549 = −0.2246 )

Note that the subscripts A and a denote voltages in the high-‐voltage
and low-‐voltage circuits, respecHvely, of the Y-‐Y connected
transformer. No phase shiL is involved.

From the symmetrical components we can calculate the a-‐b-‐c line-‐
to-‐ground voltages at bus 4 as:
⎡ V (0) ⎤
⎡ V ⎤
⎢ 4a ⎥ 1 ⎡⎢ 1 1 1 ⎤⎥ ⎢ 4a ⎥ 1 ⎡⎢ 1 1 1 ⎤⎥ ⎡ −0.261 ⎤
⎢ V (1) ⎥ = ⎢ ⎥
⎢ V4b
⎥ = ⎢ 1 a 2
a ⎥ ⎢ 4a ⎥ 3 ⎢ 1 a 2
a ⎥⎢ −0.7754 ⎥
⎢ ⎥ 3⎢
⎢ V4c ⎥ ⎣ 1 a a ⎦ ⎢ V (2) ⎥ ⎢⎣ 1 a a ⎥⎦ ⎢⎣ −0.2246 ⎥⎦
2 ⎥ 2
⎣ ⎦ ⎢⎣ 4a ⎥⎦

⎡ 0.2898 ⎤
⎢ ⎥
= ⎢ −0.5346
− j0.8660
⎥ per unit
⎢ −0.5346 − j0.8660 ⎥
⎣ ⎦

To express the line-‐to-‐ground voltages of machine 2 (in kV) mulHply
by 20/31/2:

⎡ V ⎤ ⎡ ⎤ ⎡ ⎤
⎢ 4a
⎥ 20 ⎢ 0.2898 3.346
⎥ ⎢ ⎥
⎢ ⎥ = −0.5346 − j0.8660 =
4b ⎢ ⎥ ⎢ ⎥ kV
V 1.0187∠ − 121.8°
⎢ ⎥ 3⎢ ⎥ ⎢⎣ 1.0187∠ − 121.8° ⎥⎦
⎢⎣ V4c ⎥⎦ ⎣ −0.5346 − j0.8660 ⎦


To determine phase-‐c current out of machine 2 we must first
calculate the symmetrical components of the phase-‐a current in the
branches represenHng the machine in the sequence networks.

From the zero-‐sequence circuit, the zero-‐sequence current out of
the machine is: 1 I a( )
0
2 3 4

j0.04 j0.08 j0.15 j0.08 j0.04
+
j0.15 V4a( ) j0.15
0
−
Reference

V (0)
0.261
(0)
I a = − = = − j6.525
4a
per unit
jX 0 j0.04

Similarly for the posiHve-‐ and negaHve-‐sequence currents:

1 2 3 4

(1,2)
j0.04 j0.04 j0.15 j0.04 j0.04 I a
j0.20 + j0.20

V4a( )
1,2
+
V f +
Vf
− − −
Reference

V − V (1)
1− 0.7754
I a( ) = f
1 4a
= = − j1.123
j X ′′ j0.2
per unit
() 1
( 2) V 0.2246
Ia = − 4a
= = − j1.123
jX 2 j0.2

The phase-‐c current in machine 2 is

I c = I a( ) + aI a( ) + a 2 I a( )
0 1 2

= − j6.525 + a ( − j1.123 ) ( − j1.123)
+ a 2
= − j5.402
per unit

Since the base current in the machine circuit is:

100,000
base = 2886.751
I = amps
3 × 20

I c = 15,994
amps


Line-‐to-‐Line Faults

To represent a line-‐to-‐line fault through impedance Zf the
hypotheHcal stubs on the three lines at the fault are connected as
shown. Bus k is again the fault point P, and without any loss of
generality, the line-‐to-‐line fault is regarded as being on phases b
and c. Clearly:
k
a
I fa I fa = 0
k
b
I fb Zf
Vkb − Vkc = Z f I fb
k
c
I fc I fb = − I fc

SubsHtuHng: ⎡ I (0) ⎤ ⎡ 0 ⎤
⎢ fa ⎥ ⎡ 1 1 1 ⎢ ⎤
⎢ (1) ⎥ 1 ⎢ ⎥ ⎥
I = 1 a a 2
⎢ I fb ⎥
⎢ fa ⎥ 3 ⎢ ⎥
⎢ ( 2) ⎥ ⎢⎣ 1 a 2 a ⎥⎦ ⎢ − I fc ⎥
⎢⎣ I fa ⎥⎦ ⎢⎣ ⎥⎦
(0)
⇒ I fa
=0
(1) ( 2)
⇒ I fa = − I fa

The voltages throughout the zero-‐sequence network must be zero
since there are no zero-‐sequence sources, and because I (fa ) =
0
0 ,
current is not being injected into that network due to the fault.
Hence, line-‐to-‐line fault calculaHons do not involve the zero-‐
sequence network, which remains the same as before the fault -‐ a
dead network.

(1) ( 2)
To saHsfy the requirement that I fa = −
I fa we connect the Thévenin
equivalents of the posiHve-‐ and negaHve-‐sequence networks in
parallel, as shown:
Z kk( ) Z kk( )
1 2
k Zf k
(1) + + I ( 2)
+ I fa fa
Vf (1) ( 2)
− Vka Vka
− −
Reference


To show that this connecHon of the networks also saHsfies the
voltage equaHon:
Vkb − Vkc = Z f I fb

we now expand each side of that equaHon separately as follows:

V kb
− V kc
= V ( (1)
kb
+ V ( 2)
kb ) (
− V (1)
kc
+ V ( 2)
kc )

( 1 1
) (
= Vkb( ) − Vkc( ) + Vkb( ) − Vkc( )
2 2
)
( ) (
= a 2 − a Vka( ) + a − a 2 Vka( )
1 2
)

= a(2
− a)( V (1)
ka
− V ( 2)
ka )
and:
( ) (
Z f I fb = Z f I (fb) + I (fb) = Z f a 2 I (fa) + aI (fa)
1 2 1 2
)

Equate both terms and set: I (fa1) = − I (fa2)

V kb
− V (
kc
= a 2
)(− a V (1)
ka
− V ( 2)
ka
=) Z I
f fb
= Z f
a 2 (1)
I fa
+ aI ( 2)
fa ( )

⇒ ( a 2
)(
− a V ka
(1)
− V )
( 2)
ka
= Z f (a 2
− a I (1)
fa )

(1) ( 2) (1)
⇒ V ka
− V ka
= Z I
f fa

But this is precisely the voltage drop across Zf in the figure on Slide 66.


Thus, all the fault condiHons are saHsfied by connecHng the
posiHve-‐ and negaHve-‐sequence networks in parallel through
impedance Zf as was shown.

The zero-‐sequence network is inacHve and does not enter into the
line-‐to-‐line calculaHons. The equaHon for the posiHve-‐sequence
current in the fault can be determined directly from the circuit as:

Example – Line-‐to-‐Line Faults

The same system as before is operaHng at nominal system voltage
without prefault currents when a bolted line-‐to-‐line fault occurs at
bus 3. Determine the currents in the fault, the line-‐to-‐line voltage at
the fault bus, and the line-‐to-‐line voltages at the terminals of
machine 2.


Note that we will not need the zero-‐sequence bus impedance
matrix since the fault is line-‐to-‐line (see Slide 71).

The Thévenin equivalent circuit for the posiHve-‐ and negaHve-‐
sequence is: (1)
Z33 ( 2)
3 Zf = 0 3 Z
33

⎡ j0.1437 j0.1211 j0.0789 j0.0563 ⎤ (1) + + ( 2)
⎢⎢ j0.1211 j0.1696 j0.1104 j0.0789 ⎥⎥ I fA I fA
⎢ j0.0789 j0.1104 j0.1696 j0.1211 ⎥⎥ ± V f V3(A) V3(A)
( )
1,2
1 2
Z =⎢
bus
⎢⎣ j0.0563 j0.0789 j0.1211 j0.1437 ⎥⎦

− −

Reference
From the circuit:
Vf 1.0
(1) ( 2)
I fA = − I fA = (1) ( 2 ) = = − j2.9481
per unit
Z33 + Z33 2 × j0.1696

Since I (fA) = 0 (Uppercase since the fault is in the high-‐voltage
0
transmission line)

Then I fA = I (fA1) + I (fA2)

and since I (fA1) = − I (fA2)

Then I fA = 0

2 (1) ( 2)
Also I fB = a I fA + aI fA = −5.1061
per unit

and I fC = − I fB = 5.1061
per unit

MulHply these by Ibase = 167.35 A to get the actual currents.

The symmetrical components of phase-‐A voltage to ground at bus 3
are:

V3(A) = 0
0

V3(A) = V3(A) = 1− Z kk( ) I (fA) = 1− j0.1696 − j2.9481
1 2 1 1
( )( )
= 0.5 + j0
per unit


The line-‐to-‐ground voltages at fault bus 3 are:

(0) (1) ( 2)
V3 A = V3 A + V3 A + V3 A = 0 + 0.5 + 0.5 = 1.0

(0) 2 (1) ( 2)
V 3B
= V 3B
+ a V 3B
+ 2V 3B
= 0 + a 2
× 0.5 + a × 0.5 = −0.5

V3C = V3B = −0.5

all per unit


The line-‐to-‐line voltages at fault bus 3 are:

(
V3,AB = V3 A − V3B = 1+ j0 − −0.5 + j0 = 1.5 ) ( )

(
V3,BC = V3B − V3C = −0.5 + j0 − −0.5 + j0 = 0 ) ( )

(
V3,CA = V3C − V3 A = −0.5 + j0 − 1+ j0 = −1.5 ) ( )

all per unit. MulHply these by 345/31/2 to obtain the actual voltage.


For the moment, let us avoid phase shiLs due to the Δ-‐Y
transformer connected to machine 2 and calculate the sequence
voltages of phase A at bus 4 using the the results from Slide 21 with
k = 3 and j = 4:

V ja( ) = −Z (jk ) I (fa) , V ja( ) = V f − Z (jk) I (fa) , V ja( ) = −Z (jk ) I (fa)
0 0 0 1 1 1 2 2 2

V4a( ) = −Z 43 (0) (0)
0
I fa = 0

(1) (1) (1)
(
V4a = V f − Z jk I fa = 1− j0.1211 − j2.9481 = 0.643 )( )

V4a( ) = −Z 43
2 ( 2) ( 2)
( )(
I fa = − j0.1211 − j2.9481 = 0.357 )
per unit

To account for phase shiLs in stepping down from the high-‐voltage
transmission line to the low-‐voltage terminals of machine 2, we
must retard the posiHve-‐sequence voltage and advance the
negaHve-‐sequence voltage by 30°. At machine 2 terminals,
indicated by lowercase a, the voltages are:
(0) (0) (0)
V4a = −Z 43 I fa = 0
V4a( ) = V4a( )∠ − 30° = 0.643∠ − 30° = 0.5569 − j0.3215
1 1
V4a( ) = V4a( )∠30° = 0.357∠30° = 0.3092 + j0.1785

2 2
V4a = V4a( ) + V4a( ) + V4a( ) = 0.8861− j0.1430 = 0.8778∠ − 9.4°

0 1 2


Phase-‐b voltages at the terminals of machine 2 are:
V4b( ) = V4a( ) = 0
0 0
V4b( ) = a 2V4a( ) = 1∠240° × 0.643∠ − 30° = −0.5569 − j0.3215

1 1
V4b( ) = aV4a( ) = 1∠120° × 0.357∠30° = −0.3092 + j0.1785

2 2
V4b = V4b( ) + V4b( ) + V4b( ) = −0.8861− j0.1430 = 0.8778∠ − 170.6°

0 1 2


Phase-‐c voltages at the terminals of machine 2 are:
V4c( ) = V4a( ) = 0
0 0
V4c( ) = aV4a( ) = 1∠120° × 0.643∠ − 30° = 0.643∠90°

1 1
V4c( ) = a 2V4a( ) = 1∠240° × 0.357∠30° = 0.357∠ − 90°

2 2
V4c = V4c( ) + V4c( ) + V4c( ) = j0.286

0 1 2


Line-‐to-‐line voltages at the terminals of machine 2 are:

(0) (0) (0)
V 4,ab
= V 4a
− V4b
= 1.7322
(0)
= V4b( ) − V4c( ) = −0.8661− j0.429
0 0
V4,bc

(0) (0) ( 0) per unit
V4,ca = V4c − V4a = −0.8661+ j0.429

For the voltage, mulHply by 20 kV/31/2

(0) 20
V4,ab = 1.7322 × = 20

3
0 ()
V 4,bc = 11.2∠
− 153.65°
kV
(0)
V4,ca = 11.2∠153.65°

Double Line-‐to-‐Ground Faults

Again it is clear, with fault taken on phases b and c, that the
relaHons at fault bus k are:
k
a
I fa I fa = 0
k
b
I fb Zf
k
c
I fc
(
Vkb = Vkc = Z f I fb + I fc )

(
(0) 1
3
1
Since I fa = 0 ⇒ I fa = I fa + I fb + I fc = I fb + I fc
3
) ( )

( ) (0)
Thus Vkb = Vkc = Z f I fb + I fc = 3Z f I fa

SubsHtuHng into:
⎡ V (0) ⎤ ⎡ V ⎤ ⎡ V ⎤
⎢ ka ⎥ ⎡ 1 1 ⎤
1 ⎢ ka ⎡ 1 1 ⎤
1 ⎢ ka
⎥ 1⎢ ⎥
⎢ V (1) ⎥ = 1 ⎢ 1 a 2 ⎥⎢ V
a ⎥ ⎥= ⎢ 1 a 2 ⎥⎢ V
a ⎥ ⎥
⎢ ka ⎥ 3 ⎢ ⎢
kb
⎥ 3⎢ ⎢
kb
⎥
⎢ V ( 2) ⎥ ⎢⎣ 1 a 2 ⎥
a ⎦ ⎢ Vkc ⎣ 1 a 2 ⎥
a ⎦ ⎢ Vkb
⎢⎣ ka ⎥⎦ ⎣ ⎥⎦ ⎣ ⎥⎦


Expanding the second and third equaHons:
⎡ V (0) ⎤ ⎡ V ⎤
⎢ ka ⎥ ⎡ 1 1 ⎤
1 ⎢ ka ⎥
⎢ V (1) ⎥ = 1 ⎢ 1 a 2 ⎥⎢ V
a ⎥ ⎥
⎢ ka ⎥ 3 ⎢ ⎢
kb
⎥
⎢ V ( 2) ⎥ ⎢⎣ 1 a 2 ⎥
a ⎦ ⎢ Vkb
⎢⎣ ka ⎥⎦ ⎣ ⎥⎦
(1) 1
(
Vka = a + a 2 Vkb
3 (1) ( 2)
)
⇒ Vka = Vka
( 2) 1 2
(
Vka = a + a Vkb
3
)


From the first equaHon: recall: Vkb = Vkc = Z f I fb + I fc ( )
(0)
3I fa = I fb + I fc
⎡ V (0) ⎤ ⎡ V ⎤
⎢ ka ⎥ ⎡ 1 1 ⎤
1 ⎢ ka ⎥
⎢ V (1) ⎥ = 1 ⎢ 1 a 2 ⎥⎢ V
a ⎥ ⎥
⎢ ka ⎥ 3 ⎢ ⎢
kb
⎥
⎢ V ( 2) ⎥ ⎢⎣ 1 a 2 a ⎥⎦ ⎢ Vkb
⎢⎣ ka ⎥⎦ ⎣ ⎥⎦
0
(
3Vka( ) = Vka + 2Vkb = Vka( ) + Vka( ) + Vka( ) + 2 3Z f I (fa)
0 1 2
) ( 0
)


Collect zero-‐sequence terms and recall that Vka( ) = Vka( )
1 2

3Vka( ) = Vka( ) + 6Z f I (fa) + 2Vka( )
0 0 0 1

Now solve:
(1) (0) (0)
V ka
= V ka
− 3Z I
f fa
Thus:

(1) ( 2) (0) (0)
Vka = Vka = Vka − 3Z f I fa

and since Ifa = 0
(0) (1) ( 2)
I fa + I fa + I fa = 0

These last two results characterize the double line-‐to-‐ground fault.

These can be realized by puxng all three sequence networks in
parallel as follows:

I (fa) ( 2) ( 0)
1

k I fa k I fa k
+ (1) + (1) ( 2) + ( 2) (0) + (0)

Vf Z Vka Z V Z V
− kk
−
kk
−
ka kk
−
ka

3Z f

Clearly: (1) Vf Vf
I fa = 1 =
()
( ()
0
Z kk + Z kk + 3Z f  Z kk ()
2
) (1)
Z kk + (0)
(0)
Z kk + 3Z f Z kk (
( 2)
)
( 2)
Z + Z + 3Z kk kk f

I (fa) ( 2) (0)
1

k I fa
k I fa k
+ (1) + (1) ( 2) + ( 2) (0) + (0)

Vf Z V Z V Z V
−
kk ka kk
−
ka kk
−
ka
−

3Z f

By current divider:
(0) ( 2)
( 2) (1) Z + 3Z (0) (1) Z kk
I fa = − I fa (0) kk
( 2)
f
, I fa = − I fa (0) ( 2)
Z kk + Z kk + 3Z f Z kk + Z kk + 3Z f


For a bolted fault Zf is set equal to 0. When Zf = ∞, the zero-‐
sequence circuit becomes an open circuit, no zero-‐sequence current
an flow, and the equaHons revert back to those for the line-‐to-‐line
fault.

Again we observe that the sequence currents, once calculated, can
be treated as negaHve injecHons into the sequence networks at
the fault bus k and the sequence voltage changes at all buses of the
system can then be calculated from the bus impedance matrices, as
we have done in all along.

Example Double Line-‐to-‐Ground Faults

Find the subtransient currents and the line-‐to-‐line voltages at the
fault under subtransient condiHons when a double line-‐to-‐ground
fault with Zf = 0 occurs at the terminals of machine 2 in the system
below. Assume that the system in unloaded and operaHng at rated
voltage when the fault occurs.
1 T1 2 3 T2 4
Machine 1 Machine 2


Model (the fault is a bus k = 4):
(1) ( 2)
I fa I (fa)
0
k I fa k k
+ Z kk( )
1 + (1) Z kk( )
2 + ( 2) Z kk( )
0 + (0)
Vf Vka Vka Vka
− − − −
3Z f
⎡ j0.19 0 0 0 ⎤ ⎡ j0.1437 j0.1211 j0.0789 j0.0563 ⎤

⎢ ⎥ ⎢ ⎥
(0) ⎢ 0 j0.08 j0.08 0 ⎥ (1,2) ⎢ j0.1211 j0.1696 j0.1104 j0.0789 ⎥
Z bus =⎢ ⎥ , Z bus =⎢
⎢ 0 j0.08 j0.58 0 ⎥ ⎢ j0.0789 j0.1104 j0.1696 j0.1211 ⎥⎥
⎢ 0 0 0 j0.19 ⎥⎦ ⎢ j0.0563 j0.0789 j0.1211 j0.1437 ⎥⎦
⎣ ⎣


Model (the fault is a bus k = 4):

(1) ( 2) (0)
I fa 4 I fa
4 I 4
fa

1.0 + j0.1437 + V (1) j0.1437 + V ( 2) j0.19 + V (0)
4a 4a
− −
4a
− −

Vf 1.0 1.0
I (fa1) = = = = − j4.4342
(1) ( kk
Z + 3Z f ) Z kk ( j0.19)( j0.1437 )
(0) ( 2) (0) ( 2)
(1) Z 44 Z 44
Z kk + (0) Z 44
+ (0) ( 2)
j0.1437 +
j0.19 + j0.1437
( 2) Z + Z
+ Z +
kk kk f per unit
Z 3Z 44 44


The sequence voltages at the fault are:

(1) ( 2) (0)
I fa 4 I fa
4 I 4
fa

1.0 + j0.1437 + V (1) j0.1437 + V ( 2) j0.19 + V (0)
4a 4a
− −
4a
− −

(1) ( 2) (0) (1) (1)
V4a = V4a = V4a = V f − Z 44 I fa

(
= 1.0 − j j0.1437 )(
− j4.4342
= 0.3628 )
per unit

(1)
Now that we have I fa

(1) ( 2) (0)
I fa 4 I fa
4 I 4
fa

1.0 + j0.1437 + V (1) j0.1437 + V ( 2) j0.19 + V (0)
4a 4a
− −
4a
− −

(0)
(2) (1) kk + 3Z
Z f (1 ) Z kk(0) ( 1) Z 44( 0 )
I fa = − I fa (0) ( 2) = − I fa (0) ( 2) = − I fa (0) ( 2)
Z kk + Z kk + 3Z f Z kk + Z kk Z 44 + Z 44

j0.19
= j4.4342
= j2.5247 per unit
j0.19 + j0.1437

(1)
Now that we have I fa

(1) ( 2) (0)
I fa 4 I fa
4 I 4
fa

1.0 + j0.1437 + V (1) j0.1437 + V ( 2) j0.19 + V (0)
4a 4a
− −
4a
− −

(0) (1) Z (2) (1) Z (2)
I = − I kk
= − I 44
Z kk( ) + Z kk( ) + 3Z f (0) ( 2)

fa fa 0 2 fa
Z 44 + Z 44

j0.1437
= j4.4342 = j1.9095 per unit
j0.19 + j0.1437

The currents out of the system at the fault point are:
(0) (1) ( 2)
I fa
= I fa
+ I fa
+ I fa
= j1.9095 − j4.4342 + j2.5247 = 0
I fb = I (fa) + a 2 I (fa) + aI (fa) = −6.0266 + j2.8642
0 1 2
I fc = I (fa) + aI (fa) + a 2 I (fa) = 6.0266 + j2.8642

0 1 2

The current If into ground is:

I f = I fa + I fb + I fc = 2I (fa) = j5.7285
0

all in per unit


CalculaHng the a-‐b-‐c voltages at the fault bus:
(0) (1) ( 2)
V4a = V4a + V4a + V4a = 1.0884
V4b = V4c = 0
V4,ab = V4a − V4b = 1.0884

V4,bc = V4b − V4c = 0
V4,ca = V4c − V4a = −1.0884


Recall the base current in the circuit of machine 2 is:

100 × 103
I base = = 2887
amps
3 × 20

and the base line-‐to-‐neutral voltage in machine 2 is:

20
Vbase = kV
3

Phase Shias

The previous two examples show that phase shiLs due to Δ-‐Y
transformers do not enter into the calculaHons of sequence
currents and voltages in that part of the system where the fault
occurs provided Vf at the fault point is chosen as the reference
voltage for the calculaHons.

However, for those parts of the system which are separated by Δ-‐Y
transformers from the fault point, the sequence currents, and
voltages calculated by bus impedance matrix must be shiLed in
phase before being combined to form the actual voltages.

This is because the bus impedance matrices of the sequence
networks are formed without consideraHon of phase shiLs, and so
they consist of per-‐unit impedances referred to the part of the
network which includes the fault point.
Example Phase Shias

Solve for the subtransient voltages to ground at bus 2, the end of
the transmission line remote from the double line-‐to-‐ground fault
for the same system.
1 T1 2 3 T2 4
Machine 1 Machine 2
⎡ j0.19 0 0 0 ⎤ ⎡ j0.1437 j0.1211 j0.0789 j0.0563 ⎤

⎢ ⎥ ⎢ ⎥
(0) ⎢ 0 j0.08 j0.08 0 ⎥ (1,2) ⎢ j0.1211 j0.1696 j0.1104 j0.0789 ⎥
Z bus =⎢ ⎥ , Z bus =⎢
⎢ 0 j0.08 j0.58 0 ⎥ ⎢ j0.0789 j0.1104 j0.1696 j0.1211 ⎥⎥
⎢ 0 0 0 j0.19 ⎥⎦ ⎢ j0.0563 j0.0789 j0.1211 j0.1437 ⎥⎦
⎣ ⎣

Example Phase Shias

We just solved for the values of the fault-‐current components.

NeglecHng the phase shiL of the Δ-‐Y transformer for the moment
we have, from Slide 21:

V ja( ) = −Z (jk ) I (fa) , V ja( ) = V f − Z (jk) I (fa) , V ja( ) = −Z (jk ) I (fa)
0 0 0 1 1 1 2 2 2

At bus 2:
V2a = −Z 24 I fa = − ( 0 ) ( j1.9095) = 0
(0) (0) (0)
V2a( ) = V f − Z 24 I fa = 1.0 − ( j0.0789 ) ( − j4.4342 ) = 0.6501

() ()
1 1 1
V2a( ) = −Z 24 I fa = − ( j0.0789 ) ( j2.5247 ) = 0.1992

() ()
2 2 2

Example Phase Shias

AccounHng for phase shiL in stepping up to the transmission-‐line
circuit from the fault at bus 4 we have:

V2(A) = 0
0

V2(A) = V2(A)∠30° = 0.6501∠30°
1 1
( 2) ( 2)
V2A
= V 2A
∠ − 30° = 0.1992∠ − 30°
Now the required voltages can be calculated:
(0) (1) ( 2)
V 4A
= V 4A
+ V 4A
+ V4A
= 0.7355 + j0.2255
(0) 2 (1) ( 2)
V4 B = V4 B + a V4 B + aV4 B = −0.1275 j0.5535

(0) (1) ( 2)
V4C = V4C + aV4C + a 2V4C = −0.5656 + j0.1274

all per unit
Example Phase Shias

The per-‐unit values can be converted to volts by mulHplying by the
line-‐to-‐neutral base voltage of:

345
V base = kV
3
of the transmission line.


When one phase of a balanced three-‐phase system opens, an
unbalance is created and asymmetrical currents flow. A similar type
of unbalance occurs when any two of the three phases are opened
while the third phase remains closed.

These unbalanced condiHons are caused, for example, when one-‐ or
two-‐phase conductors of a transmission line are physical broken by
accident or storm.

In other circuits, due to current overload, fuses or other switching
devices may operate in one or two conductors and fail to operate in
other conductors. Such open-‐conductor faults can be analyzed by
means of the bus impedance matrices of the sequence networks, as
we now show.

Consider a secHon of a three-‐phase circuit in which the line currents
in the respecHve phases are la, Ib, and Ic, with posiHve direcHon from
bus m to bus n, with phase a open between points p and p’:
m p p’ n
a
Ia
m n
b
Ib
m n
c
Ic

Also consider the case where phases b and c are open between
points p and p’:
m p p’ n
a
Ia
m n
b
Ib
m n
c
Ic


The same open-‐conductor fault condiHons will result is all three
phases are first opened between points p and p’ and short circuits are
then applied in those phases which are shown to be closed in the
preceding figures.

The ensuing development follows this reasoning.

Opening the three phases is the same a removing the line m-‐n
altogether then adding appropriate impedances from buses m and n
to the points p and p’.


If line m-‐n has the sequence impedances Z0, Z1, and Z2, simulate the
opening of the three phases by adding the negaHve impedances –
Z0, – Z1, and – Z2 between buses m and n in the corresponding
Thévenin equivalents of the three sequence networks of the intact
system.


For example, consider the connecHon of – Z1 to the posiHve-‐
sequence Thévenin equivalent between buses m and n:
(1)
Z mm (1)
− Z mn m
+
Vm
−

−Z1
(1) (1)
− Z nn − Z nm n
Vn +

Z (1) = Z (1)
(1) (1) (1) (1)

mn nm
Z th,mn
= Z mm
+ Z nn
− 2Z mn
0 Reference

Voltages Vm and Vn are the normal (posiHve-‐sequence) voltages of
phase a at buses m and n before the open-‐conductor faults occur.

The posiHve-‐sequence impedances kZ1 and (1 – k)Z1, 0<k<1, is
added to represent the fracHonal lengths of the broken line m-‐n
from bus m to point p and bus n to point p’.
Z (pp)′
1
()
1 ()
1
Z mm − Z mn m kZ1 p
+
Vm
− +
(1)
−Z1 Va
(1) (1)
− Z −Z
nn nm
n (1− k ) Z 1 p’ −
Vn +
()1 ()
1
Z mn = Z nm () 1 () () 1 () 1 1
Zth,mn = Z mm + Z nn − 2Z mn
0 Reference


Simplify…
()
1 ()
1
+
Vm
− +
(1) Add the
−Z1 Va do-‐nothing
() ()
(1− k ) Z
1 1
− Z nn − Z nm n source
1 p’ −
Vn +
()1 ()
1
Z mn = Z nm
() 1
Zth,mn
0 Reference


Simplify…
Combine these
()
1 ()
1
+
Vm
− +
Va( )
1
−Z1
() ()
(1− k ) Z
1 1
− Z nn − Z nm n
1 p’ −
Vn +
()1 ()
1
Z mn = Z nm
() 1
Zth,mn
0 Reference


Simplify…
Perform a
()
1 ()
1 source conversion
Z mm − Z mn m Z1
p
+
Vm
− +
Va( )
1
−Z1
()
1 ()
1
− Z nn − Z nm n
p’ −
Vn +
()1 ()
1
Z mn = Z nm
() 1
Zth,mn
0 Reference


Simplify…
()
1 ()
1 Open circuit
Z mm − Z mn m
p
+
Vm
−
Va( )
1
−Z1 Z1
()
1
Z nn ()
1
− Z nm n
Z1
− p’
Vn +
()1 ()
1
Z mn = Z nm
() 1
Zth,mn
0 Reference


Final result (posiHve-‐sequence equivalent circuit):
() 1 () 1
Z mm − Z mn m
+
Vm
− (1)
Va
()
Z nn
1 ()
− Z nm
1
n
Z1
−
Vn +
()
1 ()1
Z mn = Z nm
() 1
Zth,mn
0 Reference


The above consideraHons for the posiHve-‐sequence network also
apply directly to the negaHve-‐ and zero-‐sequence networks, but we
must remember that the laber networks do not contain any internal
sources of their own.


NegaHve-‐sequence equivalent circuit:
Z (pp)′
2
()
2 ()
2
Z mm − Z mn m kZ 2 p
+
Va( )
2
−Z 2
() ()
(1− k ) Z
2 2
Z nn − Z nm n
2 p’ −
() 2 ()2
Z mn = Z nm () 2 () () 2 () 2 2
0 Reference


NegaHve-‐sequence equivalent circuit simplified:
() 2 () 2
Z mm − Z mn m
( 2)
Va
()
Z nn
2 ()
− Z nm
2
n
Z2
()
2 ()2
Z mn = Z nm
() 2
Zth,mn
0 Reference


Zero-‐sequence equivalent circuit:
Z (pp)′
0
()
0 ()
0
+
Va( )
0
−Z0
() ()
(1− k ) Z
0 0
Z nn − Z nm n
0 p’ −
() 0 ()0
Z mn = Z nm () 0 () () 0 () 0 0
0 Reference


Zero-‐sequence equivalent circuit simplified:
() 0 () 0
Z mm − Z mn m
(0)
Va
()
Z nn
0 ()
− Z nm
0
n
Z0
()
0 ()0
Z mn = Z nm
() 0
Zth,mn
0 Reference


a ( ) denote the phase-‐a posiHve-‐sequence
1
Let the voltage V
component of the voltage drops Vpp’a, Vpp’b, and Vpp’c from p to p' in
the phase conductors.

We will soon see that V ( 1)
a and the corresponding negaHve-‐ and
zero-‐sequence components V a ( ) and V a ( ) , take on different values
2 0
depending on which one of the open-‐conductor fauIts is being
considered.


In drawing the sequence equivalent circuits it is understood that the
currents sources owe their origin to the open-‐conductor fault
between points p and p' in the system.

If there is no open conductor, the voltages:

Va( ) , Va( ) , Va( )
0 1 2

are all zero and the current sources disappear.


Note that each of the sequence currents sources:
(0) (1) ( 2)
Va Va Va
, ,
Z0 Z1 Z2

can be regarded in turn as a pair of injecHons into buses m and n of
the corresponding sequence network of the intact system.

Hence, we can use the bus impedance sequence matrices of the
normal configuraHon of the system to determine the voltage
changes due to the open-‐conductor faults.


First we must find expressions for the symmetrical components of Va
(i.e., of the voltage drops across the fault points p and p’ for each
type of fault, (one or two open lines).

These voltage drops can be regarded as giving rise to the following
sets of injecHon currents into the sequence networks of the normal
system configuraHon:

POSITIVE NEGATIVE ZERO
SEQUENCE SEQUENCE SEQUENCE

Va( ) Z1 Va( ) Z 2 Va( ) Z0
1 2 0
At bus m:

(1) ( 2) (0)
At bus n: −Va Z1 −Va Z 2 −Va Z0

Recall Slide 87 from Network CalculaKons:
⎡ ΔV1 ⎤ ⎡ Z1 j ΔI j + Z1k ΔI k ⎤
⎢ ⎥ ⎢ ⎥
⎢  ⎥ ⎢  ⎥
⎢ ⎢ ⎥
ΔV j ⎥ ⎢ Z jj ΔI j + Z jk ΔI k ⎥
⎢ ⎥=
⎢ ΔVk ⎥ ⎢ Z kj ΔI j + Z kk ΔI k ⎥
⎢ ⎥ ⎢ ⎥
⎢  ⎥ ⎢  ⎥
⎢ ⎥ ⎢ ⎥
⎢⎣ ΔVN ⎥ ⎢ Z Nj ΔI j + Z Nk ΔI k ⎥
⎦ ⎣ ⎦


The changes in the symmetrical components of the phase-‐a voltage of
each bus i is::

(0) (0)
( 0) Z im − Z in (0)
Zero-‐sequence: ΔVi = Va
Z0

(1) (1)
(1) Z − Z (1)
PosiHve-‐sequence: ΔVi = im in
Va
Z1

( 2) ( 2)
( 2) Z − Z ( 2)
NegaHve-‐sequence: ΔVi = Va
im in
Z2


Before developing the equaHons for the sequence components of the
voltage for each type of open-‐ conductor fault, let us derive
expressions for the Thévenin equivalent impedances of the sequence
networks as seen from fault points p and p’.


Looking into the posiHve-‐sequence network between p and p’, we see
the impedance:
() 1
Z1Zth,mn −Z12
Z pp′ = kZ1 + Zth,mn  ( −Z1 ) + (1− k ) Z1 = Z1 − (1)
(1) (1)
= (1)
Zth,mn − Z1 Zth,mn − Z1
()
1 ()
1
+
Va( )
1
−Z1
() ()
(1− k ) Z
1 1
Z nn − Z nm n
1 p’ −
() ()
Z (pp)′
1 1 1
Z mn = Z nm
() 1 () () 1 () 1 1
0 Reference

(1) Vm − Vn
The open-‐circuit voltage from p to p’ is: V pp′ = −Z1 (1)
Zth,mn − Z1
(1)
−Z 2
Z
But: Z (p1p) ′ = (1) 1 Thus: V p(1p)′ = pp′ (Vm − Vn )
Zth,mn − Z1 Z1
Z (pp)′
1
()
1 ()
1
+
Vm
− +
Va( )
1
−Z1
() ()
(1− k ) Z
1 1
− Z nn − Z nm n
Vn +
1 p’ −
()
1 () 1
Z mn = Z nm () 1 () () 1 () 1 1
0 Reference

Before any conductor opens, the current Imn in phase a of the line m-‐n
is posiHve sequence and is given by:

Vm − Vn
I mn =
Z1
Thus: V p(p)′ = Z (pp)′ I mn
1 1

−Z 2
Where: Z (pp)′ = (1) 1

1
Zth,mn − Z1

( 2) −Z 2 2 (0) −Z0 2
Similarly: Z pp′ = (2 ) and: Z pp′ = (0)
Zth,mn − Z 2 Zth,mn − Z0

Now we can find: Va( ) , Va( ) , Va( )
0 1 2


Thévenin equivalents:
p p
(1) + +
( ) Z (pp)′ ( 2)
2
+ Z 1
p p ′
Ia (1) Ia ( 2)
(1)
I mn Z pp′ Va Va
− p’ − p’ −

PosiHve-‐Sequence NegaHve-‐Sequence

p

Z (pp)′
0 +
I a( )
0
Zero-‐Sequence:
Va( )
0
p’ −
Open-‐Conductor Faults – One Open Conductor

m p p’ n
a
Ia (0) (1) ( 2)
Ia = 0 ⇒ Ia + Ia + Ia = 0
m n
b
Ib V pp′ ,b = 0
m n
c
Ic V pp′ ,c = 0


⎡ (0) ⎤
⎡ V ⎤
⎢ Va ⎥ ⎡ 1 1 1 ⎤ ⎡ V pp′ ,a ⎤ ⎢ pp′ ,a
⎥
⎢ (1) ⎥ = 1 ⎢ ⎥ ⎢ ⎥ 1
Va ⎢ 1 a a 2
⎥ ⎢ 0 ⎥ = ⎢ V pp′ ,a ⎥
⎢ ⎥ 3
⎢ ⎥ 3⎢ ⎥
⎢ Va( 2) ⎥
⎢⎣ 1 a 2
a ⎥⎦ ⎣ 0 ⎦ ⎢ V pp′ ,a ⎥
⎢⎣ ⎥⎦ ⎣ ⎦

(0) (1) ( 2) V pp′ ,a
⇒ Va = Va = Va =
3

The open conductor in phase a causes equal voltage drops to appear
from p to p' in each of the sequence networks. We can saHsfy this
requirement by connecHng the Thévenin equivalents of the sequence
networks in parallel at the points p and p’.


(1)
I ( 2) (0)
a
p I a
p I a
p
+
(1) + (1) ( 2) + ( 2) (0) + (0)
I Z (1) Z pp′ Va Z pp′ V Z pp′ Va

mn pp′
− −
a
−
−
p’ p’ p’

From this circuit the expression for the posiHve-‐sequence current is
found as:
(1)
I a = (1)
(1)
I mn Z pp′ I mn Z pp′ (1) 1
(
Z (pp)′ Z (pp)′ + Z (pp)′
0 2
)
(0) ( 2) = (0) ( 2) = I mn (0) (1) (1) ( 2) (0) ( 2)
Z pp′ + Z pp′  Z pp′ (1) Z Z
pp′ pp′
Z Z
pp ′ pp ′
+ Z Z
pp′ pp′
+ Z Z
pp′ pp′
Z pp′ + (0)
Z pp′ + Z (pp)′
2


I a( ) I a( ) (0)
1 2
p p Ia p
+
(1) + (1) ( 2) + ( 2) (0) + (0)
(1) Z pp′ Va Z Va Z V
I Z pp′ pp′ a
mn p p ′
− − −
−
p’ p’ p’

The sequence voltage drops are:
() () 0 2 ( ) () ( ) 0 1 2
(0) (1) ( 2) (1) Z Z
pp′ pp′
Z Z Z
pp′ pp′ pp′
Va = Va = Va = I a (0) ( 2) = I mn (0) (1)
Z pp′ Z pp′ + Z (pp)′ Z (pp)′ + Z (pp)′ Z (pp)′
1 2 0 2
Z pp′ + Z pp′
These terms are known from the impedance
parameters of the sequence networks and the
prefault current in phase a of the line m-‐n.
(0) (1) ( 2)
The currents: V V V
a
, a
, a
Z0 Z1 Z2
for injecHon into the sequence networks can now be determined.
(0,1,2) (0,1,2)
Z mm − Z mn m
Va( )
0,1,2
(0,1,2) (0,1,2) Z0,1,2

Z nn − Z nm n
(0,1,2) (0,1,2)
Z mn = Z nm
( ) 0,1,2
Zth,mn
0 Reference

Open-‐Conductor Faults – Two Open Conductors

Clearly:
m p p’ n
a
Ia
V pp′ ,a = Va( ) + Va( ) + Va( ) = 0
0 1 2
m n
b
Ib Ib = 0
m n
c
Ic Ic = 0


Resolving the current into its symmetrical components:

⎡ I (0) ⎤ ⎡ ⎤
⎢ a ⎥ ⎡ 1 1 1 ⎤ ⎡ Ia ⎤ I
1 ⎢ ⎥ ⎢ ⎥ 1 ⎢ a
⎥
⎢ I ⎥= ( 1)
1 a a 2 = ⎢ Ia ⎥
⎢ a ⎥ 3 ⎢ ⎥ ⎢ 0 ⎥ 3⎢
⎢ ⎥ ⎢ ⎥ ⎥
⎢ I ⎥ ( 2 ) ⎣ 1 a 2
a ⎦⎣ 0 ⎦ ⎢ Ia ⎥
⎢⎣ a ⎥⎦ ⎣ ⎦

(0) (1) ( 2) I a
⇒ Ia = Ia = Ia =
3
Along with:
(0) (1) ( 2)
V pp′ ,a = Va + Va + Va = 0

These can be saHsfied by connecHng the Thévenin equivalents of
the sequence networks in series between the points p and p’.

(1) (0) (1) ( 2)
Ia p Ia = Ia = Ia
+
Z (pp)′ + (1)
1
(1) Va
I mn Z pp′ p’
−
−
I a( )
2
p
Z (pp)′ + ( 2)
2
Va
p’ −
(0) p
Ia
(0) + (0)
Z pp′ Va
p’ −

(1) (0) (1) ( 2)
Ia p Ia = Ia = Ia
+
Z (pp)′ + (1)
1
(1) Va
I mn Z p’
pp′
−
−
I a( ) I a( ) = I a( ) = I a( )
2
p 0 1 2
( 2) + ( 2) (1)
Z pp′ Va I mn Z pp′
p’ − = (0) (1) ( 2)
Z pp′ + Z pp′ + Z pp′
I a( )
0
p
+ (0)
Z (pp)′
0
Va
p’ −

(1) (1) (1) (1)
(1) V = I Z − Z I
I p
a
a mn pp′ pp′ a
+
I mn Z (pp)′
1
(1) + (1)
I mn Z (pp)′ Z pp′ = I mn Z (pp)′ − Z (pp)′ (0)
1 1 1
Va
Z pp′ + Z (pp)′ + Z (pp)′
p’− 1 2

−
Z (pp)′ + Z (pp)′
0 2
( 2)
I p = I mn Z (pp)′ (0)
1
Z pp′ + Z (pp)′ + Z (pp)′

a 1 2
( 2) + ( 2)
Z pp′ Va
p’ −
I mn Z (pp)′ Z (pp)′
1 2
Va( ) = −Z (pp)′ I a( ) = − (0)

2 2 2
I a( ) Z pp′ + Z (pp)′ + Z (pp)′

0 1 2
p
+ (0)
Z (pp)′ I mn Z (pp)′ Z (pp)′
0
Va 0 1
p’ − Va( ) = −Z (pp)′ I a( ) = − (0)

0 0 0
Z pp′ + Z (pp)′ + Z (pp)′

1 2


In each of these equaHons the right-‐hand side quanHHes are all
known before the fault occurs.

The net effect of the open conductors on the posiHve-‐sequence
network is to increase the transfer impedance across the line in
which the open-‐conductor fault occurs.

For one open conductor this increase in impedance equals the
parallel combinaHon of the negaHve-‐ and zero-‐sequence networks
between points p and p’.

For two open conductors the increase in impedance equals the
series combinaHon of the negaHve-‐and zero-‐sequence networks
between points p and p'.
Open-‐Conductor Faults – EXAMPLE

For the system of Slide 39, consider that Machine 2 is a motor
drawing a load equivalent to 50 MVA at 0.8 power-‐factor lagging and
nominal system voltage of 345 kV at bus 3.
1 T1 2 3 T2 4
Machine 1 Machine 2

Determine the change in voltage at bus 3 when the transmission line
undergoes
a.  a one-‐open-‐conductor fault and
b.  a two-‐open-‐conductor fault along its span between buses 2 and 3.
Choose a base of 100 MVA, 345 kV in the transmission line.

Prefault current (per unit) in line 2-‐3:
50 50
S = = = 0.5
Sbase 100
S = V I ⇒ I 23 =
* P − jQ
= S
( P − jQ ) S
3 23 *
V3 1.0
0.8 − j0.6
= 0.5 = 0.4 − j0.3
1.0


Recall: Z1 = Z 2 = j0.15, Z0 = j0.5

⎡ j0.19 0 0 0 ⎤ ⎡ j0.1437 j0.1211 j0.0789 j0.0563 ⎤
Also recall: ( ) ⎢⎢ 0 j0.08 j0.08 0 ⎥⎥ ( ) ⎢⎢ j0.1211 j0.1696 j0.1104 j0.0789 ⎥⎥
0 1,2
Z =⎢ ⎥ , Z = ⎢ j0.0789 j0.1104 j0.1696 j0.1211 ⎥
bus
⎢ 0 j0.08 j0.58 0 ⎥ ⎢
bus
⎥
⎢ 0
⎣ 0 0 j0.19 ⎥⎦ ⎢ j0.0563 j0.0789 j0.1211 j0.1437 ⎥
⎣ ⎦

−Z 2
−Z 2
From Slide 108: Z (pp)′ = (1) 1

1
Z (pp)′ = (0) 0
0 (1)
Zth,mn (1)
= Z mm (1)
+ Z nn (1)
− 2Z mn
Zth,mn − Z1 Zth,mn − Z0
− ( j0.15)
2
−Z12
Z (pp)′ = Z (pp)′ = (1)
1 2
= = j0.712
Z 22 + Z33 − 2Z 23 − Z1 j0.1696 + j0.1696 − 2 ( j0.1104 ) − j0.15
(1) (1)
− ( j0.5)
2
−Z0 2
Z (pp)′ = (0)
0
= =∞
Z 22 + Z33 − 2Z 23 − Z0 j0.08 + j0.58 − 2 ( j0.08) − j0.5
(0) (0)


Thus, if the line from bus 2 to bus 3 is opened, then an infinite
impedance is seen looking into the zero-‐sequence network between
points p and p' of the opening. The zero-‐sequence circuit confirms this
fact since bus 3 would be isolated from the reference by opening the
connecHon between bus 2 and bus 3.

Slide 47:
j0.04 1 j0.04 2 j0.5 3 j0.08 4 j0.04
Transformer Transformer
(1) Node Bus (3) (4) (5) Node Bus (6)
(2)
3X n = j0.15 j0.04 3X n = j0.15
Reference


One open conductor: From Slide 118,

(0) (0) (1) ( 2) (1) ( 2)
(1) ( 2) Z Z Z
pp′ pp′ pp′
Z Z
pp′ pp′
Va = Va = Va = I mn (0) (1) (1) ( 2) (0) ( 2) Z (→ I mn (1)
0)
Z pp′ + Z (pp)′
2
Z pp′ Z pp′ + Z pp′ Z pp′ + Z pp′ Z pp′ pp′ →∞

( j0.712 )
2
= ( 0.4 − j0.3) = 0.1068 + j1424

2 ( j0.712 )

From Slide 104:
(0) (0) (1) (1)
Z − Z Z − Z
ΔVi( ) = im Va( ) , ΔVi( ) = ΔVi( ) = im Va( )
0 in 0 1 2 in 1
Z0 Z1


One open conductor:
(0) (0) (1) (1)
Z − Z Z − Z
ΔV3( ) = 32 Va( ) , ΔV3( ) = ΔV3( ) = 32 Va( )
0 33 0 1 2 33 1
Z0 Z1
j0.08 − j0.58
ΔV3 = (0)
j0.5
( 0.1068 + j1424 ) = −0.1068 − j0.1424
( 2) j0.1104 − j0.1696
(1)
ΔV3 = ΔV3 =
j0.15
( 0.1068 + j1424 ) = −0.0422 − j0.0562
ΔV3 = ΔV3( ) + ΔV3( ) + ΔV3( ) = −0.1912 − j0.24548

0 1 2
V3new = V3 + ΔV3 = 1.0 − 0.1912 − j0.24548 = 0.8088 − j0.2548


Two open conductors: InserHng the infinite impedance of the zero-‐
sequence network in series between points p and p' of the posiHve-‐
sequence network causes an open circuit in the laber .

No power transfer can occur in the system, since power cannot be
transferred by only one phase conductor of the transmission line in
this case since the zero-‐sequence network offers no return path for
current. 1 2 3 4
(2) (3) (4)
(5)
j0.04 j0.04 Open j0.04 j0.04
j0.20 (1)
j0.20
+ +
Vf Vf
− Reference −

7-Unsymmetrical Faults PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

7-Unsymmetrical Faults PDF

Uploaded by

Copyright:

Available Formats

UNSYMMETRICAL

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 1

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 2

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 3

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 6

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 11

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 12

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 14

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 15

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 16

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 17

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 18

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 21

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 22

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 23

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 25

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 26

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 27

Z/2 Z/2 Z/2

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 29

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 30

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 31

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 36

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 37

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 41

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 42

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 43

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 44

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 45

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 46

(1) j0.19 j0.08 (2) j0.19 (5)

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 48

Add branch (1): Z1( ) = ⎡⎣ j0.19

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 49

Add branch (5):

⎡ j0.1437 j0.1211 j0.0789 j0.0563 ⎤

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 51

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 53

Solve for I (fa)

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 54

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 55

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 56

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 57

V ja( ) = −Z (jk ) I (fa)

(1) (1) (1)

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 58

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 59

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 60

j0.04 j0.08 j0.15 j0.08 j0.04

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 61

j0.04 j0.08 j0.15 j0.08 j0.04

⎡ j0.1553 j0.1407 j0.0493 j0.0347 ⎤

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 63

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 66

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 69

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 72

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 74

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 75

11/11/13 Unsymmetrical Faults (c) 2013 H. Zmuda 76

⎢⎣ j0.0563 j0.0789 j0.1211 j0.1437 ⎥⎦

11/12/13 Unsymmetrical Faults (c) 2013 H. Zmuda 79

11/12/13 Unsymmetrical Faults (c) 2013 H. Zmuda 80

11/12/13 Unsymmetrical Faults (c) 2013 H. Zmuda 81