ECEN 615

Methods of Electric Power

Systems Analysis
Lecture 19: Equivalents, Voltage Stability

Prof. Tom Overbye

Dept. of Electrical and Computer Engineering
Texas A&M University
[email protected]
Announcements

• Read Chapters 3 and 8 from the book

• Homework 4 is due today
• Homework 5 is due on Tuesday November 12

1
QR Factorization

• Used in in SE since it handles ill-conditioned m by n

matrices (with m >= n)
• Can be used with sparse matrices
• As before we will first split the R-1 matrix
1 1
1
H R HH R
T T 2
R 2
H  H T H

• QR factorization represents the m by n H' matrix as

H  Q U
with Q an m by m orthonormal matrix and U an upper
triangular matrix (most books use Q R but we use U to
avoid confusion with the previous R) 2
QR Factorization

• We then have H T H  U T Q T Q U
• But since Q is an orthonormal matrix, Q T Q  I
• Hence we have H T H  U T U
1
Originally x   H T R 1H  H T R 1  z meas  f (x) 

With H T R 1H  H T H  H T H  U T U
Q is an m by m
1 matrix
Let z  Q T R 2  z meas  f (x) 

1 1 1 1 1
x   U T U  H T R 2
R 2  z meas  f (x)    U T U  U T Q T R 2  z meas  f (x) 

U T Ux  U T z  x  U 1z 
3
QR Factorization

• Next we’ll briefly discuss the QR factorization

algorithm
• When factored the U matrix (i.e., what most call the R
matrix ) will be an m by n upper triangular matrix
• Several methods are available including the
Householder method and the Givens Method
• Givens is preferred when dealing with sparse matrices
• All good reference is Gene H. Golub and Charles F.
Van Loan, “Matrix Computations,” second edition,
Johns Hopkins University Press, 1989.

4
Givens Algorithm for Factoring a
Matrix A
• The Givens algorithm works by pre-multiplying the
initial matrix, A, by a series of matrices and their
transposes, starting with G1G1T
– If A is m by n, then each G is an m by m matrix
• Algorithm proceeds column by column, sequentially
zeroing out elements in the lower triangle of A, starting
at the bottom of each column
If A is sparse,
G1 G p G ...G A  QU
T
p
T
1
then we can take
advantage of
G1 Gp  Q sparsity going up
the column
G Tp ...G1T A  U
5
Givens Algorithm

• To zero out element A[i,j], with i > j we first solve

with a=A[k,j], b= A[i,j] To zero out an element
c  s   a   r  we need a non-zero
 s c  b   0 pivot element in column j;
     assume this row as k.
r

r  a 2  b2
• A numerically safe algorithm is
If b=0 then c=1, s=0 // i.e, no rotation is needed

Else If |b > a| then  a / b; s  1/ 1   2 ; c  s

Else   b / a; c  1/ 1   2 ; s  c
6
Givens G Matrix

• The orthogonal G(i,k,q) matrix is then

1 0 0 0 To zero out an element
  we need a non-zero
  pivot element in column j;
0 c s 0
  assume this row as k. Row
G (i, k,q )    k here is the first non-zero
0 s c 0 above row i.
 
 
0 1 
 0 0

• Premultiplication by G(i,k,q)T is a rotation by q

radians in the (i,k) coordinate plane
7
Small Givens Example

• Let 4 2 First start in column j=1; we will

1 0  zero out A[4,1] with i=4, k=2
A
0 5
 
2 1
• First we zero out A[4,1], a=1, b=2 giving s= 0.8944,
c=-0.4472
1 0 0 0   4 2 
0 0.4472 
0 0.8944   
2.236 0.8944
G  G A
T 
1 0 0 1 0  1  0 5 
   
0 0.8944 0 0.4472  0 0.4472
8
Small Givens Example

• Next zero out A[2,1] with a=4, b=-2.236, giving

c= -0.8729, s=0.4880
 0.873 0.488 0 0   4.58 2.18 
 0.488 0.873 0 0   0 0.195 
G   G TG T A   
2  0 0 1 0 2 1  0 5 
   
 0 0 0 1   0  0.447 

• Next zero out A[4,2] with a=5, b=-0.447, c=0.996,

s=0.089 1 0 0 0   4.58 2.18 
0 1 0 0   0 0.195
G   G TG TG T A  
3 0 0 0.996 0.089  3 2 1  0 5.020 
   
 0 0 0.089 0.996   0 0 
9
Small Givens Example

• Next zero out A[3,2] with a=0.195, b=5.02,

c=0.996, s=0.089
1 0 0 0  4.58 2.18 
0 0.039 0.999 0   0 5.023
G  G TG TG TG T A  U   
4 0 0.999 0.039 0 4 3 2 1  0 0 
   
0 0 0 1  0 0 

• Also we have
0.872 0.019 0.487 0 
 0.218 0.094 0.387 0.891 
Q  G1G 2G 3G 4   
 0 0.995 0.039 0.089 
 
 0.436 0.009 0.782 0.445 
10
 Start of Givens for SE Example
• Starting with the H matrix we get  0 10 0 
 10 0 10 
 
1  0 10 0 
H  R H  100  
2

 10 0 10 
 1 0 0 
 
• To zero out H'[5,1]=1 we have  0 0 1 

b=100, a=-1000, giving 1 0 0 0 0 0

0 0 
c=0.995, s=0.0995  1 0 0 0
0 0 1 0 0 0
G 
Here the column (j) is 1, while 
0 0 0 0.995 0.0995 0 
1

i is 5 and k is 4. 0 0 0 0.0995 0.995 0 

 
0 0 0 0 0 1
11
 Start of Givens for SE Example
• Which gives
 0 10 0 
 10 0 10 
 
 0 10 0 
G H  100  
T

1
10.049 0 9.95 
 0 0 0.995 
 
 0 0 1 

• The next rotation would be to zero out element

[4,1], continuing until all the elements in the lower
triangle have been reduced
12
Givens Comments

• For a full matrix, Givens is O(mn2) since each

element in the lower triangle needs to be zeroed
O(nm), and each operation is O(n)
• Computation can be drastically reduced for a
sparse matrix since we only need to zero out the
elements that are initially non-zero, and any that
become non-zero (i.e., the fills)
– Also, for each multiply we only need to deal with the
nonzeros in the impacted row
• Givens rotation is commonly used to solve the SE

13
Power System Equivalents

• No electric grid model is ever going to completely

represent a real electric grid
– “All models are wrong but some models are useful”
• A key modeling consideration is how much of the
electric grid to represent
– For large-scale systems the distribution system is usually
equivalenced at some point; this has few system level
ramifications if it is radial; if it is networked then there are
potential issues
– At the transmission level either the full interconnect is
represented or it is equivalenced
– In an SE model in large grids (like the Eastern Interconnect) it
is always an electrical equivalent 14
Kron Reduction, Ward Equivalents

• For decades power system network models have been

equivalenced using the approach originally presented
by J.B. Ward in 1949 AIEE paper “Equivalent Circuits
for Power-Flow Studies”
– Paper’s single reference is to 1939 book by Gabriel Kron,
so this is also known as Kron’s reduction or a Ward
equivalent
• System buses are partitioned into a study system (s) to
be retained and an external system (e) to be eliminated;
buses in study system that connect to the external are
known as boundary buses
15
Ward Equivalents

• The Ward approach is based on the below relationship

• No impact on study, non-boundary buses

• Equivalent is created by doing a partial factorization of
the Ybus
– Computationally efficient

16
Other Types of Equivalents

• There are many different methods available for creating

power system equivalents
– A classic paper is by S. Deckmann, et. al., “Studies on Power
System Load Flow Equivalencing,” IEEE Transactions Power
App. And Syst., Nov/Dec 1980
– Companion paper covers numerical testing of equivalents
• The major equivalencing types are
– Ward-Type Equivalence: this is what we’ll be covering, with
the major differences associated with how the generator buses
and equivalent loads are represented
– REI Equivalents: All boundary buses connect to one “REI” bus
– Linearized Methods: Linearize about an operating point
– Others: PTDF-based, backbone type 17
Equivalent System Properties

• An equivalent is usually created from a larger model

– In the Eastern Interconnect there are full grid models that are
used for wide-area planning, these are equivalenced for real-
time usage or more specialized studies
• The equivalent is usually smaller and less detailed
– Solves quicker
– Requires less storage
– Requires less up-to-date data
• Equivalences contain fictitious elements
– This can make modeling/updating more difficult
• The equivalent only approximates the behavior of the
original 18
Study vs External System

• The key decision in creating an equivalent is to divide

the system into a study portion that is represented in
detail, and an external portion that is represented by the
equivalent
• The two systems are joined at boundary buses, which
are part of the study subsystem
• How this is done is application specific; for example:
– for real-time use it does not make sense to retain significant
portions of the grid for which there is no real-time information
– for contingency analysis the impact of the contingency is
localized
– for planning the new system additions have localized impacts
19
Ward Type Equivalencing

20
Ward Type Equivalencing
Considerations
• The Ward equivalent is calculated by doing a partial
factorization of the Ybus
– The equivalent buses are numbered before the study buses
– As the equivalent buses are eliminated their first neighbors are
joined together
– At the end, many of the boundary buses are connected
– This can GREATLY decrease the sparsity of the system
– Buses with different voltages can be directly connected

21
Ward Type Equivalencing
Considerations
• At the end of the Ward process often many of the new
equivalent lines have high impedances
– Often there is an impedance threshold, and lines with
impedances above this value are eliminated
• The equivalent lines may have unusual values, including
negative resistances
• Load and generation is represented as equivalent current
injections or shunts; sometimes these values are
converted back to constant power
• Consideration needs to be given to loss of reactive
support
• The equivalent embeds the present load and gen values2222
B7Flat_Eqv Example

• In this example the B7Flat_Eqv case is reduced,

eliminating buses 1, 3 and 4. The study system is
then 2, 5, 6, 7, with buses 2 and 5 the boundary buses
0 MW 0 MW 0 MW 0 MW 0 MW
A A

1.04 pu MVA 1.04 pu MVA

Bus 1 Bus 3 Bus 4 1.04 pu For ease of

0 MW 0 MW 0 MW 0 MW 0 MW
AGC OFF
comparison
A A
0 MW
MVA

0 MW
MVA
MVA
A
Case Hourly Cost
3454 $/h
A
MVA AGC OFF system is
0 MW
0 MW
1.04 pu
0 MW
0 MW
A
MVA
0 MW 1.04 pu
modeled
Bus 2 Top Area Cost

0 MW
0 MW
2122 $/h
Bus 5

0 MW 0 MW
unloaded
A
A 0 MW AGC OFF
MVA
MVA 0 MW
0 MW 0 MW 0 MW
A

1.04 pu MVA 1.04 pu

Bus 6 0 MW 0 MW Bus 7
A
MVA
0 MW Left Area Cost Right Area Cost 0 MW
slack

0 Mvar 832 $/h 500 $/h 0 Mvar

0 MW AGC OFF -0 MW AGC OFF

23
B7Flat_Eqv Example

• Original Ybus
 20.83 16.67 4.17 0 0 0 0 
 16.67 52.78 5.56 5.56 8 .33 16 .67 0 
 
 4.17 5.56 43.1 33.3 0 0 0 
 
Ybus  j 0 5.56 33.3 43.1 4.17 0 0 
 0 8.33 0 4.17 29.17 0 16.67 
 
 0 16.67 0 0 0 25 8.33 
 0 25 
 0 0 0 16.67 8.33
 20.833 4.167 0 
Yee  j  4.167 43.056 33.333 
 0 33.333 43.056 

24
B7Flat_Eqv Example
16.667 5.556 5.556 
16.667 0 0 0  0 
Yes  j  5.556 0 0 
0 4 .167
0 Yse  j  
 0 0 0 
 5.556 4.167 0 0   
 0 0 0 
 52.778 8.333 16.667 0  Note Yes=Yse'
 8.333 29.167 0 16 .667  if no phase
Yss  j  
 16.667 0 25.0 8.333  shifters
 
 0 16 .667 8 .333 25 .0 
 28.128 11.463 16.667 0 
 11.463 28.130 
 Yss  Yse Yee1Yes   j  16.667
0 16 .667 
0 25.0 8.333 
 
 0 16.667 8 .333 25 .0 
25
Equivalencing in PowerWorld

• Open a case and solve it; then select Edit Mode, Tools,
Equivalencing; this displays the Power System
Equivalents Form
Next step is then to
divide the buses into
the study system and
the external system;
buses can be loaded
from a text file as
well

26
Equivalencing in PowerWorld

• Then go to the Create The Equivalent page, select the

desired options and select Build Equivalent System
Maximum
impedance
lines to retain

’99’ or ‘EQ’
Click to create the equivalent
are common
circuit values
for equivalent
lines

Removes equivalenced objects from the oneline 27

Small System Equivalent Example

• Example shows the creation of an equivalent for

Aggieland37 example First example is
Aggieland Power and Light A

33%
MVA

52%
MVA
A
SLACK345
simple, just removing
Total Load 1420.7 MW 1.02 pu HOWDY345
641 MW
1.02 pu TEXAS345
A

79%
MVA
Total Losses: 24.72 MW
A

22%
MVA
69%

0.99 pu
A

MVA

A
A

69%
MVA

HOWDY138
A

26%
SLACK138
A

52%
MVA

1.03 pu
slack

WHITE138 (bus 3);

0.98 pu TEXAS138 53 MW 85% MVA

A
1.01 pu

37 MW
A

54%
MVA
1.0375 tap
TEXAS69

44%
A
0.99 pu
21 Mvar
0.0 Mvar
BATT69

0.99 pu
29 MW
8 Mvar
A

64%
MVA

NORTHGATE69
1.02 pu
MVA

74%
HOWDY69

27 MW
4 Mvar
A

72%
MVA
1.01 pu

100 MW

30 Mvar

16%
note TEXAS138 is
14 Mvar MVA

21% 12MAN69
MVA

now directly joined to

MVA
A
MVA 0.98 pu BONFIRE69
A 29% 34 MW
WHITE138 69% MVA
0 Mvar
MVA A
A
78%
20 MW 35% CENTURY69
0.0 Mvar
MVA
MVA A
8 Mvar 31 MW 78% PLUM138
13 Mvar 0.957 pu
A
WEB138 MVA
40% 0.98 pu
MVA GIGEM69 A
0.98 pu A

MAROON69
82 MW
REVEILLE69

27 Mvar

0.0 Mvar
0.0 Mvar
0.98 pu
A
93 MW
65 Mvar

TREE69
59 MW
17 Mvar
0.97 pu

16%
MVA
A
53%
MVA

34%
A

MVA
0.999 pu
96%
MVA
1.0875 tap

WEB69

100 MW

30 Mvar
49 MW
17 Mvar

1.01 pu
RELLIS138..
A
A

64%
MVA 0.0 Mvar FISH69
64% MVA
A
MVA 93 MW
1.005 pu A
KYLE138 63%
SPIRIT69 58 Mvar 51% 0.961 pu MVA

1.01 pu
MVA

A
A

51%
A

63%
1.0000 tap
YELL69 24% 35 MW
MVA
MVA MVA
0.0 Mvar A

11 Mvar 0.98 pu A

KYLE69 67%
0.985 pu
MVA
A MVA

25 MW 41%
A
MVA
61 MW
10 Mvar A
17 Mvar
MVA
1.00 pu 50%
A

MVA 57%
A
A
58 MW A MVA
0.99 pu
39%
13% 0.0 Mvar 96 MW MVA
BUSH69 17 Mvar 0.0 Mvar 47%
MVA
A
MVA
20 Mvar 21%
0.97 pu 0.97 pu MSC69 MVA A
A

55% 70 MW 59 MW RING69
RELLIS69 MVA
MVA

0.97 pu RUDDER69 0 Mvar 6 Mvar

1.000 pu
36 MW
1.0213 tap
A A
22 MW A

66%
MVA
70%
MVA
1.0213 tap 24 Mvar
0 Mvar
46%
MVA

38 MW 0.98 pu AGGIE138 AGGIE345

RELLIS138 10 Mvar 0.990 pu REED69
0.982 pu
0.99 pu HULLABALOO138
0 deg A

54%
A MVA A

14% 59%
MVA A 0.989 pu REED138 A MVA

34% 57%
MVA MVA

45%
MVA
1.03 pu
1.01 pu A

59%
MVA

Case is Aggieland37_HW5
28
 Small System Equivalent Example
Aggieland Power and Light Only bus 3 was
A
SLACK345
33%
MVA
A

removed; the new

52%
MVA

Total Load 1420.7 MW 1.02 pu HOWDY345

slack
641 MW
Total Losses: 24.70 MW
A A A

69% 69% 52%

1.02 pu TEXAS345 MVA MVA SLACK138 MVA

0.98 pu
A
A

79%
MVA

TEXAS138 53 MW
21 Mvar
A

22%
MVA

0.0 Mvar
29 MW
0.99 pu

85%
A

MVA
HOWDY138
A

26%
MVA

1.01 pu
1.03 pu
equivalent line was
54%
MVA
1.0375 tap 8 Mvar
1.02 pu HOWDY69
100 MW

auto-inserted.
A
27 MW
A

1.01 pu TEXAS69 BATT69 64% 72% 30 Mvar

0.99 pu MVA 4 Mvar MVA
A

37 MW 44% 0.99 pu NORTHGATE69 A A

74% 16%
14 Mvar MVA

12MAN69
MVA MVA
A
0.98 pu BONFIRE69
A 29% 34 MW
69% MVA
0 Mvar
MVA A
A
78%
35% CENTURY69
0.0 Mvar
MVA
MVA A
31 MW 78% PLUM138
13 Mvar 0.957 pu
A
WEB138 MVA
40% 0.98 pu
MVA GIGEM69 A A

MAROON69 REVEILLE69 93 MW 53% 96% 1.0875 tap 49 MW

82 MW 65 Mvar MVA MVA

59 MW 17 Mvar
27 Mvar
17 Mvar 34%
A
0.999 pu
WEB69
TREE69 0.97 pu MVA
1.01 pu
0.0 Mvar 0.98 pu
A 100 MW
0.0 Mvar A
16%
MVA
30 Mvar

A
A

64%
MVA 0.0 Mvar FISH69
64% MVA
A
MVA 93 MW
1.005 pu A
KYLE138 63%

Don’t save the

SPIRIT69 58 Mvar 51% 0.961 pu MVA

1.01 pu
MVA

A
A

51%
A

63%
1.0000 tap
YELL69 24% 35 MW
MVA
MVA MVA
0.0 Mvar A

11 Mvar 0.98 pu A

KYLE69 67%
0.985 pu
MVA
A MVA

25 MW 41%
A
MVA
61 MW
10 Mvar A
17 Mvar

equivalent with the

MVA
1.00 pu 50%
A

MVA 57%
A
58 MW A MVA
0.99 pu
39%
0.0 Mvar 96 MW MVA
BUSH69 17 Mvar 0.0 Mvar 47%
MVA
A
20 Mvar 21%
0.97 pu 0.97 pu MSC69 MVA A
A

55% 70 MW 59 MW RING69
RELLIS69

same name as the

MVA
MVA
0.97 pu RUDDER69 0 Mvar 6 Mvar
1.000 pu
36 MW
1.0213 tap
A A
22 MW A

66%
MVA
70%
MVA
1.0213 tap 24 Mvar
0 Mvar
46%
MVA

38 MW 0.98 pu AGGIE138 AGGIE345

RELLIS138 10 Mvar 0.990 pu REED69
0.982 pu
A

14%
MVA
0.99 pu HULLABALOO138

34%
MVA
0.989 pu
0 deg A

54%
MVA

REED138
57%
A

MVA
A

59%
MVA
original, unless you
45%
A

MVA

1.01 pu A

59%
MVA
1.03 pu
want to lose the
original

29
Small System Equivalent Example

• Now remove buses at WHITE138 and TEXAS and

RELLIS (1, 3, 12, 40, 41, 44); set Max Per Unit
Impedance for Aggieland Power and Light
SLACK345

Equivalent Total Load 1376.0 MW 1.02 pu HOWDY345

52%
MVA
A

slack
641 MW
Total Losses: 24.61 MW
A A A

69% 69% 52%

MVA MVA SLACK138 MVA

Lines to 53 MW
21 Mvar
0.0 Mvar
29 MW
8 Mvar
0.99 pu

1.02 pu
85%
A

MVA

HOWDY69
HOWDY138
A

26%
MVA

1.01 pu
1.03 pu

100 MW
A
27 MW
A

99 (per unit) 0.99 pu

BATT69

29%
MVA
0.99 pu
64%
MVA

NORTHGATE69 A

74%
MVA
4 Mvar

12MAN69
34 MW
0 Mvar
72%
MVA

0.98 pu
30 Mvar

A
BONFIRE69
A

16%
MVA

78%

to retain all
CENTURY69
0.0 Mvar
MVA
A
31 MW 78% PLUM138
13 Mvar 0.957 pu
A
WEB138 MVA
40% 0.98 pu
MVA GIGEM69 A A

MAROON69 REVEILLE69 93 MW 53% 96% 1.0875 tap 49 MW

82 MW 65 Mvar MVA MVA

59 MW 17 Mvar
27 Mvar
17 Mvar 34%
A
0.999 pu
WEB69

lines. Again
TREE69 0.97 pu MVA
1.01 pu
0.0 Mvar 0.98 pu
A 100 MW
0.0 Mvar A
16%
MVA
30 Mvar

A
A

64%
MVA 0.0 Mvar FISH69
64% MVA
A
MVA 93 MW
1.005 pu A
KYLE138 63%
SPIRIT69 58 Mvar 51% 0.961 pu MVA

to an auto-
1.01 pu
MVA

A
A

51%
A

63%
1.0000 tap
YELL69 24% 35 MW
MVA
MVA MVA
0.0 Mvar A

11 Mvar 0.98 pu A

KYLE69 67%
0.985 pu
MVA
A MVA

25 MW 41%
MVA
61 MW
10 Mvar A
17 Mvar
1.00 pu 50%
A

57%

insert to show
MVA

58 MW A MVA
0.99 pu
BUSH69 17 Mvar 0.0 Mvar 47%
MVA
A

0.97 pu 0.97 pu MSC69 21%

MVA A

70 MW 59 MW RING69 MVA

0.97 pu RUDDER69 0 Mvar 6 Mvar

the equivalent
36 MW
22 MW A

24 Mvar 46%
0 Mvar MVA

38 MW 0.98 pu AGGIE138 AGGIE345

10 Mvar 0.990 pu REED69
0.99 pu HULLABALOO138
0 deg A

54%
MVA A

59%
0.989 pu REED138

lines.
A A MVA

34% 57%
MVA MVA

1.03 pu
1.01 pu A

59%

30
MVA
Small System Equivalent Example

• Now set the Max Per Unit Impedance for

Equivalent Lines to 2.5.
Aggieland Power and Light
SLACK345

52%
MVA

Total Load 1376.7 MW 1.02 pu HOWDY345

slack
641 MW
Total Losses: 23.85 MW
A A A

69% 69% 52%

MVA MVA SLACK138 MVA

0.99 pu HOWDY138
A
1.03 pu
A 26%
53 MW 85%
MVA

21 Mvar MVA 1.01 pu

29 MW
0.0 Mvar 8 Mvar
1.02 pu HOWDY69
100 MW
A
27 MW
A

0.99 pu BATT69 64% 72% 30 Mvar

MVA 4 Mvar MVA

0.99 pu NORTHGATE69 A A

74% 16%
12MAN69
MVA MVA
A
0.98 pu BONFIRE69
29% 34 MW
MVA
0 Mvar
A

78%
CENTURY69
0.0 Mvar
MVA
A
31 MW 78% PLUM138
13 Mvar 0.957 pu
A
WEB138 MVA
40% 0.98 pu
MVA GIGEM69 A A

MAROON69 REVEILLE69 93 MW 53% 96% 1.0875 tap 49 MW

82 MW 65 Mvar MVA MVA

59 MW 17 Mvar
27 Mvar
17 Mvar 34%
A
0.999 pu
WEB69
TREE69 0.97 pu MVA
1.01 pu
0.0 Mvar 0.98 pu
A 100 MW
0.0 Mvar A
16%
MVA
30 Mvar

A
A

64%
MVA 0.0 Mvar FISH69
64% MVA
A
MVA 93 MW
1.005 pu A
KYLE138 63%
SPIRIT69 58 Mvar 51% 0.961 pu MVA

1.01 pu
MVA

A
A

51%
A

63%
1.0000 tap
YELL69 24% 35 MW
MVA
MVA MVA
0.0 Mvar A

11 Mvar 0.98 pu A

KYLE69 67%
0.985 pu
MVA
A
MVA

25 MW 41%
MVA
61 MW
10 Mvar A
17 Mvar
1.00 pu 50%
A

MVA 57%
58 MW
MVA
A
0.99 pu
BUSH69 17 Mvar 0.0 Mvar 47%
MVA
A

0.97 pu 0.97 pu MSC69 21%

MVA A

70 MW 59 MW RING69 MVA

0.97 pu RUDDER69 0 Mvar 6 Mvar

36 MW
22 MW A

24 Mvar 46%
0 Mvar MVA

38 MW 0.98 pu AGGIE138 AGGIE345

10 Mvar 0.990 pu REED69
0.99 pu HULLABALOO138
0 deg A

54%
MVA A

59%
A
0.989 pu REED138 A MVA

34% 57%
MVA MVA

1.03 pu
1.01 pu
A

59%
MVA

31
Large System Example: 70K Case

• Original System has 70,000 buses and 71,343 lines

32
Large System Example: 70K Case

• Just retain the Oklahoma Area; now 1591 buses and

1745 lines (deleting ones above 2.5 pu impedance)
TYRONE 6

S COFFEYVI LLE

COPAN
TURPI N 1
FORGAN
NEWKI RK
BUFFALO 77
ALVA 1
BLACKWELL 3
PONCA CI TY 1 MI AMI 81
M EDFORD 18 SHI DLER DEWEY 1
GOODWELL BUFFALO 76 PAWHUSKA 1
HOOKER PAWHUSKA 3
DEWEY 2
PONCA CI TY 3
BARTLESVI LLE 2
KEYES ALVA 3
BLACKWELL 2
GUYM ON 2
GUYM ON 1
CHEROKEE 7 BLACKWELL 1
BARTLESVI LLE 3
BOI SE CI TY GUYM ON 3 BEAVER 8 PONCA CI TY 5
PONCA CI TY 9
TURPI N 2
PAWHUSKA 2
LAVERNE 2 PONCA CI TY 2 PONCA CI TY 7 PONCA CI TY 10
VI NI TA 3
PONCA CI TY 4
WOODWARD 7
POND CREEK TONKAWA 2
PONCA CI TY 8
PONCA CI TY 6 BARTLESVI LLE 5
TEXHOM A 2
LAVERNE 3
TONKAWA 1
LAVERNE 1
TEXHOM A 1
HUNTER 2
FAI RFAX 22
WOODWARD 8

FORT SUPPLY
WAYNOKA
GUYM ON 5 ALI NE HELENA 7
GUYMON 4
WOODWARD 6
GARBER 1

WOODWARD 4
ENI D 6 HOM I NY 2
ENI D 5
TEXHOM A 3
WOODWARD 3 M ORRI SON 4
M OORELAND 2
GAGE LAHOM A 1
MOORELAND 3 OOLOGAH 2
FAI RVI EW 15 ENI D 7
ENI D 4
MORRI SON 5
WOODWARD 5
RI NGWOOD 3
LANGLEY 2
ENI D 1 PAWNEE 3
OOLOGAH 3
LAHOM A 2
ENI D 3 PERRY 23
GARBER 2
WOODWARD 9

WOODWARD 10 FAI RVI EW 16

WAUKOM I S
M ARSHALL 44
STI LLW ATER 11 GLENCOE 6 OWASSO 1 SALI NA 4
VI CI FAI RVI EW 17
SHATTUCK

STI LLW ATER 9 TULSA 7 PRYOR 5

CLEVELAND 95
SEI LI NG 1 OKEENE 1
SEI LI NG 3 STI LLW ATER 10 YALE 3 CHOUTEAU 3
ORLANDO 58
STI LLW ATER 16
SAND SPRI NGS 5
SEI LI NG 2 OKEENE 2
STI LLW ATER 12
HENNESSEY
DRUM RI GHT 2 BROKEN ARROW 13
ARNETT STI LLWATER 14
CANTON 58
TALOGA
DRUM RI GHT 1

CANTON 59 BROKEN ARROW 14

OAKWOOD 5
CRESCENT CUSHI NG 5
JENKS 5
TAHLEQUAH 3
WATONGA 1

PUTNAM 3 WATONGA 2

LEEDEY

FORT GI BSON 4

THOM AS 2
EDMOND 17
GEARY 1
CHEYENNE 1 MUSKOGEE 9
LUTHER 4
OKLAHOMA CI TY 67
ARAPAHO GEARY 2
CALUMET 5
CLI NTON 63 GORE 2
HAMMON
BOLEY
CALUM ET 4
ELK CI TY 3
WEATHERFORD 1
CHEYENNE 2
HARRAH 4 OKM ULGEE 1
HI NTON 7 YUKON 5
ELK CI TY 1 CLI NTON 64
WEATHERFORD 2
SWEETWATER 4 ELK CI TY 2
BURNS FLAT 2
HI NTON 5 OKEM AH 1
OKEM AH 3
HENRYETTA 1
HYDRO

OKEM AH 2

CANUTE HI NTON 6
SHAWNEE 2
MI NCO 5 HENRYETTA 3
SAYRE 4 CORDELL 2 MUSTANG 3
SAYRE 3
HENRYETTA 2 EUFAULA 5
EUFAULA 6 PORUM 3 SALLI SAW 5
BURNS FLAT 1
WETUM KA 1
WELEETKA 2
SAYRE 5 MI NCO 3 WELEETKA 1 STI GLER 3

MUSTANG 2 OKLAHOMA CI TY 17 STI GLER 1

EUFAULA 4
SENTI NEL
WETUM KA 2
ERI CK PANAMA
CARTER
FORT COBB
QUI NTON 4
FLETCHER 4
CORDELL 1
CARNEGI E 6
ANADARKO 3 HOLDENVI LLE 1

ANADARKO 4 M CALESTER 1
CARNEGI E 5
HOBART 6 HOLDENVI LLE 2 M CALESTER 4
RED OAK 5

GRANI TE
KONAWA 1
MCALESTER
M CALESTER 2
3
WI STER 1
M ANGUM 2 HOBART 7 KONAWA 3
WI LBURTON 2
WI STER 2
KONAWA 4
WI LBURTON 1
M ANGUM 1
CARNEGI E 8
ADA 8 KI OWA 2
HARTSHORNE
BLAI R 2 HEAVENER
ALLEN 1
CALVI N
TALI HI NA 2
ADA 11

ADA 9

ALTUS 3
PAULS VALLEY 2
STRATFORD 14
ALLEN 2
ADA 10
M ARLOW 2 KI OWA 1 HODGEN 1
ALTUS 5 HODGEN 2

HOLLI S 2 I NDI AHOM A M ARLOW 1 ELM ORE CI TY 1

PAULS VALLEY 1
KI OWA 3 CLAYTON 18 TUSKAHOM A
SNYDER COALGATE 3
STONEWALL 2
ELM ORE CI TY 2
TALI HI NA 1
M ARLOW 3
2 4
ALTUSALTUS COALGATE 1
DAVI S 7
DUNCAN 4
ELM ORE CI TY 3
ROFF SM I THVI LLE 9
ELDORADO 3
SULPHUR 9

LAWTON 17
DUNCAN 3
DUNCAN 6
SULPHUR 10
COALGATE 2
CACHE 3 M OYERS
DUNCAN 5 DAVI S 6 STONEWALL 3

FREDERI CK 16

RATLI FF CI TY HENNEPI N 2 WATSON 3

WALTERS 1 DAVI S 5
ATOKA 2
WALTERS 2
COM ANCHE 1
BETHEL 13
ATOKA 3
FREDERI CK 15
ATOKA 4
HEALDTON 2
M ANNSVI LLE
WALTERS 3 TI SHOM I NGO 1 RATTAN
GRANDFI ELD
COLEM AN 5
ARDM ORE 4

RANDLETT WAURI KA 2 TI SHOM I NGO 3 ANTLERS 2

TEM PLE 4
ARDM ORE 6
ANTLERS 1
ARDM ORE 3
LONE GROVE CADDO
ARDMORE 5 WRI GHT CI TY 3
BROKEN BOW 1
WAURI KA 1
BROKEN BOW 2 BROKEN BOW 4
RI NGLI NG M ADI LL 2
BROKEN BOW 3
KI NGSTON 24
M ADI LL 3 SOPER HUGO 5 FORT TOWSON 1
WI LSON 12 DURANT 6
EAGLETOWN
DURANT 7
DURANT 4 HUGO 7 VALLI ANT
M ARI ETTA 19 M EAD
CALERA 5 BOSWELL 3
FORT TOWSON 2
BENNI NGTON 4

GARVI N
TERRAL
DURANT 5 I DABEL 2
BURNEYVI LLE BOKCHI TO 2
KI NGSTON 25 CALERA 4
I DABEL 3
HAWORTH
COLBERT 2 I DABEL 1

BOKCHI TO 1

THACKERVI LLE

HENDRI X

33
Grid Equivalent Examples

• A 2016 EI case had about 350 lines with a circuit ID of

’99’ and about 60 with ‘EQ’ (out of a total of 102,000)
– Both WECC and the EI use ’99’ or ‘EQ’ circuit IDs to
indicate equivalent lines
– One would expect few equivalent lines in interconnect wide
models
• A ten year old EI case had about 1633 lines with a
circuit ID of ’99’ and 400 with ‘EQ’ (out of a total of
65673)
• A ten year old case with about 5000 buses and 5000
lines had 600 equivalent lines
34