2017 - 09 - 19 - Power System Modeling and Fault Analysis Using Symmetrical Components - Dean Sorenson - Technical Meeting PDF

Symmetrical Components
IEEE PES Boston Chapter

Technical Meeting
September, 19 2017
National Grid, Waltham, MA
Dean.Sorensen@NationalGrid.com
Symmetrical Components
Discussion Topics
• History and Description
• The General Method of Symmetrical Components

– N-Phase Systems
– 3-Phase Systems
• Circuit Element Sequence Representations
• Fault Analysis Using Symmetrical Components
_______________________________________
1J. Lewis Blackburn and Thomas J. Domin, Protective Relaying Principles and Applications, 3rd Ed., CRC Press, 2007
.
2John, A Horak, Derivation of Symmetrical Component Theory and Symmetrical Component Networks, Georgia Tech
protective Relaying Conference, Atlanta, GA, April 2005, http://www.basler.com/downloads/Symmcomp.pdf
IEEE-PES-Boston-2017-Fall Symmetrical Components 2

Symmetrical Components – History and Description
• The method of symmetrical components provides a tool to study systems

with unbalanced phasors.
• Developed by Charles Fortescue in 1913, who presented a paper entitled’

“Method of Symmetrical Co-ordinates Applied to the Solution of Polyphase
Networks.”3
• In mathematics terms, a linear transformation4 is a mapping of quantities

between two vector spaces, in this case a physical domain (ABC) and a
sequence domain (012).
• The method simplifies circuit analysis of a three-phase mutually coupled

circuit by transforming it into 3 single phase circuits with no mutual
coupling.
_______________________________________
3Fortescue’s paper is available from the University of Waterloo from
http://thunderbox.uwaterloo.ca/~ccanizar/papers/classical/Fortescue.pdf.
4Rowland,Todd and Weisstein, Eric W. "Linear Transformation." From MathWorld—A Wolfram Web
Resource http://mathworld.wolfram.com/LinearTransformation.html

A Tool for Simplifying Fault Analysis
• A balanced system is
easily analyzed because
only one phase needs to
be considered.
• Unbalanced systems
require a full circuit
analysis of all three
phases, neutral and
ground elements.
• Therefore transforming an
unbalanced system into
balanced systems
promises to simplify our
analysis.

The General Method of Symmetrical Components
• Method for 3-phase power system is a subset of the more general transformation
method. The general method resolves N unbalanced phasors which share the
same reference plane into N sets of balanced phasors, with each set having N
members.
• Within each set, each of the phasors has the same magnitude and successive
phases have the same phase angle separation between them.
• Let α be the angle shift between successive phases in an N-phase system. Then
let’s define a useful operator dubbed the “a” operator. The “a” operator is a unit
phasor (magnitude = 1) with an angle equal to α. Multiplying a phasor by a simply
rotates that phasor by α degrees in the counterclockwise direction.
360
 a 1 Rotation operator
N
• Within each sequence network, the angular displacement of successive phasors is

-α∙n where n is the phase sequence network number and where n = 0, 1, 2, …N-1.

3-Phase System Example
Physical System Seq #0 System (Zero Seq)
c
a0,
α a b0,
c0
N=3
Seq #1 System (Pos Seq)
n = 0, 1, 2, …N-1
α = 120 deg c1
a1
b1
Note: All system rotation,
even “negative-sequence”
Seq #2 System (Neg Seq)
phasors, is counterclockwise!
b2
a2
c2

Physical System Seq #0 System Seq #3 System
d b3
a0,
b0,
c a c3 a3
α c0,
d0
b d3
N=4
Seq #1 System
n = 0, 1, 2, …N-1
α = 90 deg d1
c1 a1
b1
Seq #2 System
b2, a2,
d2 c2

e d3
d
b3
a0,
b0,
α a c0, a3
d0,
e0
c e3
b c3
N=5
Seq #1 System Seq #4 System
n = 0, 1, 2, …N-1
α = 72 deg e1 b4
d1 c4
a1 a4
c1 d4
b1 e4
Seq #2 System
c2
e2
a2
b2
d2

e f
a0,
b0,
b3, a3,
c0,
d a d3, c3,
d0,
f3 e3
e0,
f0
c b
N=6
n = 0, 1, 2, …N-1
α = 60 deg e1 f1 b4,
e4
A more detailed development of 6-phase sequence network
applications can be found in the following paper: a4,
d1 a1
d4
5Bhatt, Navin B., Six-Phase (Multi-Phase) Power Transmission
Systems: Fault Analysis, IEEE Transactions on Power c4,
c1 b1 f4
Apparatus and Systems, Vol. PAS-96, No. 3, May/June 1977,
http://www.libsou.com/pdf/01601991.pdf
c2,
c5 b5
f2
Trivia Factoid:
When N is prime, each set where n>0 will form a regular N-
sided polygon. Consider n=0 as being a 1-sided polygon. a2,
d5 a5
For N=6, prime factors are 1, 2, 3 and 6. Can you find a 1- d2
sided, 2-sided, 3-sided and 6-sided polygon among these

sets? Try this for any value of N. b2,
e5 f5
e2

Back to the 3-Phase System Example
Physical System Seq #0 System (Zero Seq)
c
We can relate physical domain quantities to
sequence domain quantities by superposition. a0,
α a b0,
c0
Va  Va 0  Va1  Va 2
Vb  Vb 0  Vb1  Vb 2 b
Vc  Vc 0  Vc1  Vc 2
N=3
Seq #1 System (Pos Seq)
n = 0, 1, 2, …N-1
α = 120 deg c1
This transforms 3 quantities into a total of 9 quantities. Simplified?

a1
Let’s define quantities V0, V1 and V2 in the sequence domain using

a-phase (or specifically Va) as a reference and along with the “a” b1
operator substitute the new “sequence” quantities into b and c-

Seq #2 System (Neg Seq)
phases.
b2
V0  Va 0 Va  V0  V1  V2
V1  Va1 Vb  V0  a 2V1  aV2
a2
V2  Va 2 Vc  V0  aV1  a 2V2 c2

3-Phase System Example Using Matrix Notation
Expressing the transformation in matrix notation makes the transformation format easier
to remember.
Va  V0  V1  V2 Va  1 1 1  V0 
V   1 a 2 a   V 
Vb  V0  a 2V1  aV2  b    1 Vabc   A V012
Vc  V0  aV1  a 2V2 Vc  1 a a 2  V2 
Matrix notation also lets us easily derive an inverse transform.
V0  1 1 1  Va 
V012  A1  Vabc V   1 1 a a 2   Vb 
 1 3 
V2  1 a 2 a  Vc 

3-Phase System Example Using Matrix Notation
Naturally, other phasor quantities such as current or flux have the same transformation
form as voltage phasors.
I a  I 0  I1  I 2  I a  1 1 1   I 0 
I b  I 0  a 2 I1  aI 2  I   1 a 2 a    I 
 b    1 I abc   A I 012
I c  I 0  aI1  a I 22
 I c  1 a a 2   I 2 
I0  1 1 1  I a 
I 012  A  I abc 
1
 I   1 1 a a 2    I b 
 1 3 
 I 2  1 a 2 a   I c 

Symmetrical Component Transformation
Worked Example
6This example is from Washington State University’s March 2011 Hands-On Relay School

Worked Example (Continued)




Symmetrical Component Transformation – Visualization
• 7Steven Blair at the University of Strathclyde has a good interactive
tool for visualizing symmetrical components:
http://personal.strath.ac.uk/steven.m.blair/seq/

Circuit Element Sequence Representations

Sequence Networks
• Elements of a power system are represented by impedances in
each of sequence networks. For a 3-phase power system:
– Zero sequence represents impedances of the system to equal

(in-phase) currents in all three phases.
– Positive sequence represents impedances of the system to

normal (balanced) currents in all three phases.
– Negative sequence represents impedances of the system to

currents with reversed phase sequence.

Sequence Network Independence
• Each of the sequence networks is independent of the others.
• For a balanced network:

– Sequence currents produce only like sequence network voltage drops.
• Thus the sequence networks are not connected to each other.
• Unbalanced sources resolve to zero, positive and negative sequence
sources.
• For unbalanced network:

– Sequence currents can produce voltage drops in any or all three
sequence networks.
• Thus, we model unbalances by setting up independent sequence networks
and interconnecting them at the point of the unbalance (i.e. the fault
location).
• For these studies we assume the rest of the system and all the sources are
balanced (i.e. no sources in the zero and negative sequence networks).

3-Phase System with Sending and Receiving End Voltages
IA RA jXAA
RB jXBB jXAB jXCA

IB
RC jXCC jXBC
IC
VAS VBS VCS jXAG jXCG VAR VBR VCR
jXBG
RG

Self Impedances
• The voltage drop equations in matrix form for our system are
VAS  VAR   Z AA Z AB Z AC   I A 
V   V    Z Z BB Z BC    I B 
 BS   BR   BA
VCS  VCR   Z CA Z CB Z CC   I C 
• Starting with KVL around the loop, the self impedance of loop A (i.e. the voltage drop
in loop A resulting from the current in loop A) is
VAS  VAR I A
 RA  RG I A  j  X AA  X AG I A
VAS  VAR I
Z AA  A
 RA  RG   j  X AA  X AG 
IA
• ZBB and ZCC in the impedance matrix are similarly defined.

Mutual Impedances
• Again, our voltage drop equations in matrix form are
V   V    Z Z BB Z BC    I B 
 BS   BR   BA
• The mutual impedance from Loop A to Loop B (i.e. the voltage drop in loop A resulting
from the current in Loop B) is
(VAS  VAR ) I B
Z AB   RG  j  X AB  X AG 
IB
• The XAG term might look like a typo but recall the following for flux linkage in Phase A:
 A (t )  LAAi A (t )  LAB iB (t )  LAC iC (t )  LAG iG (t )

where iG (t )  i A (t )  iB (t )  iC (t )
 A (t )  LAA  LAG i A (t )  LAB  LAG iB (t )  LAC  LAG iC (t )

ABC-to-012 Conversion Using Voltage Drop Expressions
• Start with our voltage drop equations in matrix form
V   V    Z Z BB Z BC    I B 
 BS   BR   BA
• Multiply both sides of the expression by [A-1] which preserves the equality and multiply
[ZABC] by [A-1]·[A] which is the same as multiplying it by the identity matrix [I].
VAS  VAR   Z AA Z AB Z AC  I A 
A1 VBS   A1 VBR   A1  Z BA Z BB Z BC A A  I B 
1
• This leaves us with the voltage drop expressions in the sequence domain.
V0 S  V0 R   Z 00 Z 01 Z 02   I 0 
V   V    Z Z11 Z12    I1 
 1S   1R   10
V2 S  V2 R   Z 20 Z 21 Z 22   I 2 

Sequence Independence and Sequence Coupling
• Let’s examine each of the impedance terms of Z012.
V0 S  V0 R   Z 00 Z 01 Z 02   I 0 
V   V    Z Z11 Z12    I1 
 1S   1R   10
V2 S  V2 R   Z 20 Z 21 Z 22   I 2 
• See how the diagonal terms indicate impedances with no coupling between
the sequence networks whereas the off-diagonal terms indicate impedances
with coupling between the sequence networks.
(V0 S  V0 R ) I 0 (V0 S  V0 R ) I1 (V0 S  V0 R ) I 2

Z 00  Z 01  Z 02 
I0 I1 I2
(V1S  V1R ) I 0 (V1S  V1R ) I1 (V1S  V1R ) I 2
Z10  Z11  Z12 
I0 I1 I2
(V2 S  V2 R ) I 0 (V2 S  V2 R ) I1 (V2 S  V2 R ) I 2
Z 20  Z 21  Z 22 
I0 I1 I2
Cases of Impedance Matrices with High Symmetry
Case 1 – Symmetrical Passive Elements
• In the case of transmission lines, common assumptions include symmetrically spaced
phase conductors with regular conductor transposition. This yields the following
equalities:
RP  RA  RB  RC
X S  X AA  X BB  X CC
X M  X AB  X BC  X CA  X BA  X CB  X AC
X G  X AG  X BG  X CG
• This in turn enables the introduction of new

variables representing self and mutual
impedances.
Z S  Z AA  Z BB  Z CC
 RP  RG  j ( X S  X G )
Z M  Z AB  Z BC  Z CA  Z BA  Z CB  Z AC
 RG  j ( X M  X G )

Case 1 – Symmetrical Passive Elements
• Now we substitute ZS for the self impedance and ZM for the mutual impedances in the
ZABC impedance matrix.
 Z AA Z AB Z AC   Z S ZM ZM 
Z ABC   Z BA Z BB Z BC    Z M ZS Z M 
 Z CA Z CB Z CC   Z M ZM Z S 
• And convert to Z012.
 ZS ZM ZM  Z S  2Z M 0 0 
Z 012  
 A1   Z M ZS Z M   A   0 ZS  ZM 0 
 Z M ZM Z S   0 0 Z S  Z M 
• Notice that Z012 contains only diagonal elements (i.e. no coupling between sequence
networks) and that Z11 = Z22. The more common form of expression is that Z1 = Z2.

Case 2 – Rotating Machines
• In the case of rotating machines (motors and generators), mutual coupling between
phases includes rotation as well as physical geometry. This differs from non-rotating
circuit elements because now ZAB ≠ ZBA, ZBC ≠ ZCB and ZCA ≠ ZAC.
RP  RA  RB  RC
X S  X AA  X BB  X CC
X M   X AB  X BC  X CA
X M   X BA  X CB  X AC
X G  X AG  X BG  X CG
• This in turn enables the introduction of new variables
representing self and mutual impedances.
Z S  Z AA  Z BB  Z CC
 RP  RG  j ( X S  X G )
Z M   Z AB  Z BC  Z CA Z M   Z BA  Z CB  Z AC
 RG  j ( X M   X G )  RG  j ( X M   X G )
Case 2 – Rotating Machines
• Now we substitute ZS for the self impedance and ZM for the mutual impedances in the
ZABC impedance matrix.
 Z AA Z AB Z AC   Z S ZM  ZM  
Z ABC   Z BA Z BB Z BC    Z M  ZS Z M  
 Z CA Z CB Z CC   Z M  ZM  Z S 
• And convert to Z012.
 ZS ZM  ZM   Z S  Z M   Z M  0 0 
Z 012  
 A1   Z M  ZS Z M    A   0 Z S  a 2 Z M   aZ M  0 

 Z M  ZM  Z S   0 0 Z S  aZ M   a Z M  
2
• Notice that Z012 still contains only diagonal elements (i.e. no coupling between
sequence networks) but this time Z11 ≠ Z22 or more commonly Z1 ≠ Z2.

System Equivalent Sources
• You can often get Thevenin impedances (Z0 and Z1) directly from
short circuit programs like CAPE or Aspen. An equivalent source is
nonrotating so Z2 = Z1.
• Otherwise, you can calculate Z0 and Z1 from fault duties, expressed

in MVA or kVA based on the system base voltage and the available
fault current. You need both 3-phase and SLG fault duties to
calculate Z0 (see next slide). With this method assume impedances
to be all reactive.
• For hand calculations you can often assume an ideal voltage

source, a.k.a. an infinite bus, (i.e. Z0 = Z1 = Z2 = 0).

System Equivalent Sources
Following is a calculation method for Z0, Z1 and Z2 from given 3ph (S3ph) and
SLG (S1ph) fault duties in MVA. Resulting impedances are in per unit.
S Base 3S Base
Z1  Z 2  Zg   Z 0  Z1  Z 2
S3 ph S1 ph
Z 0  Z g  Z1  Z 2
 3S Base   S Base 
   2 
 S  S 
 1 ph   3 ph 
See Blackburn Appendix 4.1 for a detailed derivation.

Synchronous Generators
For a sustained fault on the terminals of an unloaded generator, the
armature current has a decrement as shown below. A piece-wise
linear approximation of this nonlinear response is achieved using
separate positive sequence reactances: Xd”, Xd’ and Xd.

Synchronous Generators
• Reactance increases with time after a short circuit because of the

demagnetizing effect of the fault current on the air-gap flux. For round
rotor machines, typical positive sequence reactances are as follows:
0.95  Xd 1.45 
0.12  Xd’ 0.28  pu on generator base
0.07  Xd’’ 0.17 
• Negative-sequence reactance is often approximated by equating it to

the subtransient reactance.
X2 = Xd’’

Generators – Induction Machines
• Induction machines (motors or generators) are often ignored as
fault current sources for relaying purposes.
• The fault current contribution from an induction machine typically

decays in a few cycles.
• Induction machine contribution to fault current (subtransient

reactance) may be considered when performing maximum
instantaneous fault current studies for bus and switchgear rating
purposes.

Balanced 3-Phase Sources
Zero-sequence impedance depends on the manner in which the
generator is grounded. Any impedance in the neutral circuit (Zn) is
represented as three times this value (3Zn) in the zero-sequence
model since 3I0 flows through the neutral.
3Zn
Z1 External neutral
Z2 impedance.
E Z0 Z0
For solidly
REF REF REF grounded neutral REF
3Zn= 0.
Pos. Seq. Neg. Seq. Zero Seq. Zero Seq.
(Wye-Connected) (Delta-Connected)
For ungrounded
neutral 3Zn = ∞

Load Impedances
• Positive and negative sequence impedances of lumped loads are generally
equal. These are shown on a reference phase basis in sequence networks.
 ZS ZM ZM  Z S  2Z M 0 0 
Z 012  
 A1  Z M ZS Z M   A   0 ZS  ZM 0 
Z M ZM Z S   0 0 Z S  Z M 
• For synchronous motors, particularly those that are designed with salient
poles, the negative sequence impedance generally lies between Xd’ and Xd’’.
 ZS ZM  ZM   Z S  Z M   Z M  0 0 
Z 012  
 A1   Z M  Z S Z M    A   0 Z S  a 2 Z M   aZ M  0 

Z M  Z M  Z S   0 0 Z S  aZ M   a Z M  
2
• Zero sequence impedance of loads depends on the manner in which they are
connected and grounded as shown on the following page.

Zero Sequence Models for Different Load Connections
Ia0 = I0
I0
Ib0 = I0 Z Z
Z 3I0
N
Z
Ic0 = I0
REF
Ia0 = I0
I0
Ib0 = I0 Z 3Zn
3I0
Z N
Zn Z
Z
Ic0 = I0
Connection REF
Zero Sequence
Arrangement Equivalent Circuits
Z
Z N
Z Z
REF
Z0
REF
Transmission Lines and Cables
• In sequence network calculations, the per-phase values of line
resistance, reactance and shunt capacitive susceptance are used.
– These values can be found from conductor tables or calculated by hand
using Carson’s equations.
– There are also several software apps available for deriving the per-
phase line constants from physical dimensions of the line or cable.
• Recall from earlier expressions that for lines and cables, ZL1 = ZL2
while ZL0 > ZL1.
 ZS ZM ZM  Z S  2Z M 0 0 
Z 012  
 A1  Z M Z S Z M   A   0 ZS  ZM 0 
Z M Z M Z S   0 0 Z S  Z M 

ZL0 for Overhead Transmission Lines
• The zero sequence impedance, ZL0, of an overhead line depends on

several factors which can result in wide variation.
– The use of overhead shield wires and the type of tower grounding and
counterpoise.
– Ground resistivity.
– Zero-sequence impedance is usually 2 to 3.5 times the

positive-sequence impedance.
Check out the Z0,
Z1 and Z2 values in
the sample EMTP
line model data.

Sample ATP-EMTP Output for OH Transmission Lines
Calculated impedances are for a typical flat profile 345kV transmission line (left-most
tower configuration) with two bundled conductors per phase and no transposition.
Impedance matrix, in units of [ohms/mile ] for the system of equivalent phase conductors.
Rows and columns proceed in the same order as the sorted input.
ZAA
1 2.780137E-01
1.099286E+00 Notice that ZAA = ZCC ≠ ZBB
ZBA ZBB and ZBA = ZCB ≠ ZCA.
2 2.305740E-01 2.910487E-01
5.180598E-01 1.085324E+00
ZCA ZCB ZCC
3 2.230820E-01 2.305750E-01 2.780138E-01
4.412611E-01 5.181495E-01 1.099286E+00
Impedance matrix, in units of [ohms/mile ] for symmetrical components of the equivalent phase conductor
Rows proceed in the sequence (0, 1, 2), (0, 1, 2), etc.;
Columns proceed in the sequence (0, 2, 1), (0, 2, 1), etc. Z AA  RA  RG   j  X AA  X AG 
Z00
0 7.385127E-01 Z AB  RG  j  X AB  X AG 
2.079613E+00
Z10 Z12
1 -2.156088E-02 -4.804473E-02  Z S  2Z M 0 0 

Z 012   ZS  ZM 
-4.577000E-03 2.854904E-02 0 0 
Z20 Z22 Z21  0 0 Z S  Z M 
2 1.471788E-02 5.428174E-02 4.869572E-02
-1.642828E-02 6.021421E-01 2.742369E-02
ZL0 for Transmission Line Cables
• For cables, the type of sheath or pipe used in the construction is a

major factor in the zero-sequence impedance.
• In addition, the placement of the phase conductors relative to each

other affects the amount of current flow in the cable sheaths or pipe
and thus has a significant impact on the zero-sequence impedance.

Pipe-type Cables – R0 and X0
• R0 and X0 in pipe type cables vary with current loading.
• Most fault studies will choose a current level at which to study the
cable, usually a level corresponding with the available fault current.
8
40
7
35
30 6
25 5
X0/X1
R0/R1
20
4
15
3
10
5 2
0 1
10 100 1000 10000 100000 10 100 1000 10000 100000
Total Pipe Current - Amperes Total Pipe Current - Amperes

Transformers – Core Type vs Shell Type
Single Phase Units
3-Phase Units
Core-type
(3-legged) Shell-type

Transformers – Positive and Negative Sequence
Regardless of whether the transformer is a bank of three (3) single-
phase units or a 3-phase transformer having either a shell-form or a
core-form core:
• Z1 and Z2 are equal and are the leakage impedance of the

transformer.
• Positive-sequence voltages and currents shift by  30 when

passing through a Delta - Wye transformer bank. The sign of the
phase shift angle is determined by the transformer connections and
is equal to that of the physical voltage and current phase shift angle.
• The sign of the negative sequence voltage and current phase shift
angle is opposite that of the positive sequence phase shift angle.
• No phase shift occurs in the zero sequence network.

Transformers – Zero Sequence
• Where a 3-phase transformer bank is arranged
without interlinking magnetic flux (three-phase shell
type, or three single-phase units) and provided there
is a path for zero sequence currents, then Z0 = Z1.
• With 3-phase core type units, the zero sequence

fluxes produced by zero sequence currents can find a
high reluctance path through the tank, the effect being
to reduce Z0 to about 85-90% of Z1. With a grounded
wye-wye configuration, this appears as a phantom or Shell-type
“tank” tertiary winding in the zero sequence network.
• With hand calculations, it’s not uncommon to ignore

this variation and consider the positive and zero
sequence impedances to be equal.
• When using short circuit software, it’s common to

assume a zero-sequence impedance as indicated
above, if manufacturer provided zero sequence
values are not available. Core-type
(3-legged)
Source: 8Areva Network Protection and Automation Guide, Chapter 5, page 57.
2-Winding Transformers – Zero Sequence
This represents a solidly grounded

The connection from internal neutral (i.e. 3Zn = 0).
impedance to the reference
and the open connection to the Replace with 3Zn for an impedance
external terminal represents a grounded neutral.
delta connection.
Replace with an open (i.e. 3Zn = ∞) for
an ungrounded neutral.
3-Winding and Autotransformers
A 3-winding transformer can be represented with a T-model.
Recall that impedances derived from measurements (i.e test reports) are
those between pairs of windings with the third winding being an open circuit.
Thus we can relate the values ZPS, ZPT and ZST to effective individual winding
impedances ZP, ZS and ZT as follows:
With some algebraic manipulation we can arrive at the following:

3-Winding and Autotransformers – Positive and Negative Sequence
The positive and negative sequence representations are the same and simply
include solid connections from the internal impedances ZP, ZS and ZT to the
external terminals P, S and T as shown below right.
Ref.

3-Winding and Autotransformers – Zero Sequence
Zero sequence representations use the same internal connection rules for
each winding configuration as seen in the 2-winding example.
Note for N-Winding Transformers:

Representation of transformers with
more than 3 windings (i.e. N-winding
transformers) follow the same process
and connection rules as 3-winding
transformers.

Now For a Word About Textbook Lookup Tables …
• Textbook diagrams often present transformer zero sequence
diagrams in a misleading way.
• They often put neutral connections

where delta connections go.
• They often depict that which is

physically outside the transformer
as being inside the transformer.
• This can make representing

You can also refer to Figures A4.2-1 and A4.2-3 in
shared neutrals (or neutral buses) Blackburn (Reference #1).
seem harder than it really is.

A Common but Incorrect Mapping of I0 to IN
Current across shunt elements in the zero sequence representation corresponds with
neutral current.
This is just
wrong.
3I0
If the zero sequence is open, as for a delta-delta transformer or ungrounded wye

winding, clearly there is no neutral current. When there is only a series path there is
zero sequence current flow but not through the neutral.
This is just
wrong.
3I0 ??!

The Correct Mapping of I0 to IN
The connection to reference, making Zt0 a shunt element, models the delta connection.
I0 flowing through the shunt connection models I0 inside the delta while I0 leaving Zt0
models I0 in the wye winding.
I0
I0
I0
Consider a similar mapping for wye-wye windings. Notice that when there is only a
series path there is zero sequence current flow through the windings but the net zero
sequence current through the neutrals is zero.

Shared Neutral Bus Example
• If remember these simple connection rules for wye grounding and
delta arrangements, you may never need a lookup table again.
• Always draw what’s physically outside the transformer as being

outside the transformer in your sequence representations.
H jXH jXX X
H T X
H0 T0 X0
jXX
3RG
H T X
jXH jXX X
H0 T0 X0
RG jXX
T
See the problem?
Shared Neutral Bus Example
Station A Station A Station B Station B
115kV Bus T1-A 23kV Bus 1 23kV Bus 1 T1-B 115kV Bus
Line 1
RG1B
Station A Station B
T2-A 23kV Bus 2 23kV Bus 2 T2-B
Line 2
RG2B
RGA
I0
Zero Sequence Reference
Station A
Station A Station A Station B
Neutral Bus Station B
115kV Bus 23kV Bus 1 23kV Bus 1
Line 1 3RG1B T1-B 115kV Bus
jXGA T1-A 3RGA jXGB

I0
Station A Station B
23kV Bus 2 23kV Bus 2
Line 2 3RG2B T2-B
T2-A

Zero Sequence
Reference
Station A
Neutral Bus
115kV Bus 23kV Bus 1 Line 1 23kV Bus 1 3RG1B T1-B 115kV Bus
jXGA T1-A 3RGA jXGB
Station A Station B
23kV Bus 2
Line 2 23kV Bus 2 3RG2B T2-B
T2-A
I0
Positive Sequence
Reference

115kV Bus 23kV Bus 1 Line 1 23kV Bus 1 T1-B 115kV Bus
jXGA T1-A jXGB
Station A Station B
23kV Bus 2 Line 2 23kV Bus 2 T2-B
T2-A
I1
Negative Sequence
Reference

115kV Bus 23kV Bus 1 Line 1 23kV Bus 1 T1-B 115kV Bus
T1-A
I2
jXGA jXGB
Station A Station B
23kV Bus 2 Line 2 23kV Bus 2 T2-B
T2-A

Fault Analysis Using Symmetrical Components

Example System (Fault at Bus C)

3-Phase (3PH) Fault
a
At the fault point Va = Vb = Vc = 0 and Ia = Ib = Ic
Since Va = Vb = Vc = 0 b
From the definitions: V0 = V1 = V2 = 0
c
Since Ia = Ib = Ic
From the definitions: I0 = 0 and I2 = 0
As expected, these relationships
suggest only the positive sequence
network to be connected at the
fault point.
REFERENCE EQUATIONS
V0 
1
Va  Vb  Vc  I0 
1
I a  I b  I c 
3 3 Va  V0  V1  V2 I a  I 0  I1  I 2
V1  Va  aVb  a 2Vc  I1  I a  aI b  a 2 I c 
1 1
3 Vb  V0  a 2V1  aV2 I b  I 0  a 2 I1  aI 2
3
V2  Va  a 2Vb  aVc  Vc  V0  aV1  a 2V2 I c  I 0  aI1  a 2 I 2
I 2  I a  a 2 I b  aI c 
1 1
3 3

3-Phase (3PH) Fault
a Va  Vb  Vc  0
Ia  Ib  Ic
b
c
V
I a1 
Z1

3-Phase (3PH) Fault
I1
Z1
Ia  Ib  Ic
Va  Vb  Vc  0 V POS V1
I2
Z2
V1  0
V0  V2  0 NEG V2
I0  I2  0
I0
Z0
Notice we observe an extra
component of V1 and I1
because of load current and
source impedance. ZERO V0
The above fault record is from Washington State University’s March 2011 Hands-On Relay School

Single-Line-to-Ground (SLG) Fault
a
At the fault point Va = 0 and Ib = Ic = 0
b
Since Va = 0 V0 + V1 + V2 = 0
c
Since Ib = Ic = 0 I0 + a2I1 + aI2 = 0 and
I0 + aI1 + a2I2 = 0 These relationships suggest that
all three sequence networks are
Which simplifies to I0 = I1 = I 2 connected in series at the fault
point.
REFERENCE EQUATIONS
V0 
1
Va  Vb  Vc  I0 
1
I a  I b  I c 
3 3 Va  V0  V1  V2 I a  I 0  I1  I 2
1 1
3 Vb  V0  a 2V1  aV2 I b  I 0  a 2 I1  aI 2
3
I 2  I a  a 2 I b  aI c 
1 1
3 3

Va  0
Ib  Ic  0
V
I a1  I a 2  I a 0 
Z1  Z 2  Z 0
3V
I a  I a1  I a 2  I a 0  3I a 0 
Z1  Z 2  Z 0

I1
Z1
POS
Va  0 V V1
Ib  Ic  0
I2
Z2
NEG
V0  V1  V2  0 V2
I 0  I1  I 2
I0
Z0
Notice we observe an
extra component of V1 ZERO
and I1 because of load
current and source V0
impedance.
Line-to-Line (LL) Fault
At the fault point Ia = 0 and Ib = -Ic and Vb = Vc a
Since Vb = Vc V0 + a2V1 + aV2 = V0 + aV1 + a2V2 b

Which simplifies to V1 = V2
c
Since Ib = -Ic I0 + a2I1 + aI2 = -(I0 + aI1 + a2I2)
Which simplifies to I1 = -I2
These relationships suggest the
Since Ia = 0 I0 + I1 + I2 = 0 where I1 = -I2 positive and negative sequence
Which simplifies to I0 = 0 networks are connected in
parallel at the fault point.
REFERENCE EQUATIONS
V0 
1
Va  Vb  Vc  I0 
1
I a  I b  I c 
3 3 Va  V0  V1  V2 I a  I 0  I1  I 2
1 1
3 Vb  V0  a 2V1  aV2 I b  I 0  a 2 I1  aI 2
3
I 2  I a  a 2 I b  aI c 
1 1
3 3

Ia  0
Ib  Ic
Vb  Vc
V
I a1   I a 2 
Z1  Z 2
Va1  Va 2

I1
Z1
Ia  0
Ib  Ic POS V V1
Vb  Vc
I2
Z2
extra component of V1
and I1 because of load NEG
current and source V2
impedance.
V1  V2 Z0
I0
I0  0
I1   I 2 ZERO V0
Double Line-to-Ground (DLG) Fault
a
At the fault point Ia = 0 and Vb = Vc = 0
Since Vb = Vc = 0 V0 + a2V1 + aV2 = 0 and b

V0 + aV1 + a2V2 = 0
Which simplifies to V0 = V1 = V2 c
Since Ia = 0 I0 + I1 + I 2 = 0
Or rewritten as I1 = -(I0 + I2) These relationships suggest the
zero, positive and negative
sequence networks are all
connected in parallel at the fault
point.
REFERENCE EQUATIONS
V0 
1
Va  Vb  Vc  I0 
1
I a  I b  I c 
3 3 Va  V0  V1  V2 I a  I 0  I1  I 2
1 1
3 Vb  V0  a 2V1  aV2 I b  I 0  a 2 I1  aI 2
3
I 2  I a  a 2 I b  aI c 
1 1
3 3

Ia  0
Vb  Vc  0
Ia  0
Ib  Ic
Vb  Vc
V
I a1   I a 2 
Z1  Z 2 Va1  Va 2  Va 0
Va1  Va 2
V
I a1 
 Z Z 
Z1   2 0 
 Z2  Z0 

I1
Z1
Ia  0
V POS V1
Vb  Vc  0
I2
Z2
V0  V1  V2 NEG V2
I1    I 0  I 2 
I0
Z0
extra component of V1
and I1 because of load
current and source ZERO V0
impedance.
Phase Shifting Across Delta-Wye Transformers

For this example HV leads LV by 30°. Let the system
voltage ratio (N) equal 1. Consequently, the turns
ratio (n) must be 1/3. We know that
Because the sequence networks are independent, we

can apply them individually adding the results by
superposition.
Starting with positive sequence values we get the
following for voltage and current.
Pictures and equations are from Appendix 4.3 in Blackburn.

Recall from before that
Next applying negative sequence values we get the

following for voltage and current.
2 2
Pictures and equations are from Appendix 4.3 in Blackburn.

• Blackburn Appendix 4.3 provides a second transformer,

Example (b), which has a high side delta, low side wye with
the HV side similarly leading the low side by 30°.
• The process for determining the positive and negative

sequence shift angles is the same as Example (a) as are the
actual shift angles.

Example System (Fault at Bus C)
1. Interconnect the sequence networks

to model the particular fault type and
solve the network to determine the
total fault current components I012
and convert to IABC.
2. Next, determine other internal values

for particular elements – say the
current through Generator GB and
the voltage behind its reactance – by
simply determining the current
through GB in each network (IGB0,
IGB1 and IGB2) as well as the 3
voltages across GB (VGB0, VGB1 and
VGB2).
3. Now convert IGB_012 and VGB_012 to

IGB_ABC and VGB_ABC.

Sequence Networks for a UG System With an SLG Fault
A B
G
L
No
Ground
Sources
• X0C is generally assumed equal

to X1C and X2C.
• X0C, X1C and X2C are very large

compared to the series
reactances of the system so X1C
and X2C are shorted by the
XFMR and GEN X1 and X2.
• Note that the location of the fault

has little effect on the magnitude
of the capacitive fault current
since the capacitance is
distributed throughout the
system and the values of the
series reactances are small.

Common Question From Projects Regarding
Open Delta VTs versus Wye-Wye VTs
Why do I have to reject proposals to use two (2) VTs in an open delta configuration instead
of three (3) VTs in a wye-wye configuration?
This is simply not

suitable on 3PH-4W
(grounded) systems
9Refer to IEEE Std. C37.230-2007, Guide for Protective Relay Applications to Distribution Lines, Figure 6-5.

Questions?

81

2017 - 09 - 19 - Power System Modeling and Fault Analysis Using Symmetrical Components - Dean Sorenson - Technical Meeting PDF

Uploaded by

Document Information

Original Description:

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

2017 - 09 - 19 - Power System Modeling and Fault Analysis Using Symmetrical Components - Dean Sorenson - Technical Meeting PDF

Uploaded by

Copyright:

Available Formats

Symmetrical Components

IEEE PES Boston Chapter

• History and Description

• The General Method of Symmetrical Components

• Circuit Element Sequence Representations

• Fault Analysis Using Symmetrical Components

protective Relaying Conference, Atlanta, GA, April 2005, http://www.basler.com/downloads/Symmcomp.pdf

IEEE-PES-Boston-2017-Fall Symmetrical Components 2

• The method of symmetrical components provides a tool to study systems

• Developed by Charles Fortescue in 1913, who presented a paper entitled’

• In mathematics terms, a linear transformation4 is a mapping of quantities

• The method simplifies circuit analysis of a three-phase mutually coupled

IEEE-PES-Boston-2017-Fall Symmetrical Components 3

IEEE-PES-Boston-2017-Fall Symmetrical Components 4

• Within each sequence network, the angular displacement of successive phasors is

IEEE-PES-Boston-2017-Fall Symmetrical Components 5

IEEE-PES-Boston-2017-Fall Symmetrical Components 6

IEEE-PES-Boston-2017-Fall Symmetrical Components 7

IEEE-PES-Boston-2017-Fall Symmetrical Components 8

sided, 2-sided, 3-sided and 6-sided polygon among these

IEEE-PES-Boston-2017-Fall Symmetrical Components 9

This transforms 3 quantities into a total of 9 quantities. Simplified?

Let’s define quantities V0, V1 and V2 in the sequence domain using

operator substitute the new “sequence” quantities into b and c-

IEEE-PES-Boston-2017-Fall Symmetrical Components 10

Matrix notation also lets us easily derive an inverse transform.

IEEE-PES-Boston-2017-Fall Symmetrical Components 11

IEEE-PES-Boston-2017-Fall Symmetrical Components 12

IEEE-PES-Boston-2017-Fall Symmetrical Components 13

IEEE-PES-Boston-2017-Fall Symmetrical Components 14

IEEE-PES-Boston-2017-Fall Symmetrical Components 15

IEEE-PES-Boston-2017-Fall Symmetrical Components 16

IEEE-PES-Boston-2017-Fall Symmetrical Components 17

IEEE-PES-Boston-2017-Fall Symmetrical Components 18

IEEE-PES-Boston-2017-Fall Symmetrical Components 19

– Zero sequence represents impedances of the system to equal

– Positive sequence represents impedances of the system to

– Negative sequence represents impedances of the system to

IEEE-PES-Boston-2017-Fall Symmetrical Components 20

• For a balanced network:

• For unbalanced network:

IEEE-PES-Boston-2017-Fall Symmetrical Components 21

RB jXBB jXAB jXCA

VAS VBS VCS jXAG jXCG VAR VBR VCR

IEEE-PES-Boston-2017-Fall Symmetrical Components 22

• ZBB and ZCC in the impedance matrix are similarly defined.

IEEE-PES-Boston-2017-Fall Symmetrical Components 23

 A (t )  LAAi A (t )  LAB iB (t )  LAC iC (t )  LAG iG (t )

 A (t )  LAA  LAG i A (t )  LAB  LAG iB (t )  LAC  LAG iC (t )

IEEE-PES-Boston-2017-Fall Symmetrical Components 25

(V0 S  V0 R ) I 0 (V0 S  V0 R ) I1 (V0 S  V0 R ) I 2

• This in turn enables the introduction of new

IEEE-PES-Boston-2017-Fall Symmetrical Components 27

• And convert to Z012.

IEEE-PES-Boston-2017-Fall Symmetrical Components 28

• And convert to Z012.

IEEE-PES-Boston-2017-Fall Symmetrical Components 31

• Otherwise, you can calculate Z0 and Z1 from fault duties, expressed

• For hand calculations you can often assume an ideal voltage

IEEE-PES-Boston-2017-Fall Symmetrical Components 33

IEEE-PES-Boston-2017-Fall Symmetrical Components 34

IEEE-PES-Boston-2017-Fall Symmetrical Components 35