Optimal Power Flow Paper 7

Staff paper by
Paula A. Lipka
Richard P. O’Neill
Shmuel Oren
April 2013

Page 1
 Developing Line Current magnitude Constraints for
IEEE Test Problems
Optimal Power Flow Paper 7
Paula A. Lipka, Richard P. O’Neill, and Shmuel Oren
[email protected]; [email protected]; [email protected]
April 2013

Abstract and Executive Summary

All power system operators ensure their systems adhere to thermal limits on
transmission lines in order to avoid line deformation. Also, thermal limits are used
as surrogates for voltage stability. The IEEE test problems do not include data on
these limits. The purpose of this paper is to present a simple method for
constructing current magnitude constraints and to report on the computational
properties in solving the resulting problems. This paper finds limits on the
maximum allowable current magnitude that result in a feasible solution for the 14‐
bus, 30‐bus, 57‐bus, and 118‐bus IEEE test problems. For each test problem, one
single limit is applied to all lines that makes the optimal solution without these
limits infeasible. For each problem we develop a ‘tight’ and a ‘loose’ constraint. We
solve the resulting problem using the current voltage formulation. Different test
problems exhibit different characteristics. Including these constraints in the ACOPF
increases the solution time between 2 to 20 times and costs objective function up
to 25 percent.

Disclaimer: The views presented are the personal views of the authors and not the
Federal Energy Regulatory Commission or any of its Commissioners.

Page 2
 Table of Contents
1. Introduction .................................................................................................... 4

2. Notation ............................................................................................................ 4

3. Current‐Voltage ACOPF Model ................................................................ 5

4. Methods ............................................................................................................ 6

5. Computational Analysis .............................................................................. .7

6. Conclusions...................................................................................................... 17

References ............................................................................................................ 18

Appendix ............................................................................................................... 21

Page 3
1. Introduction
The amount of current that can flow through power system transmission
assets referred to here as lines is limited by thermal restrictions. The thermal
ratings of the transmission are functions of the materials that compose the assets.
The excessive heat caused by the line current can deform and degrade transmission
lines and cause them to sag. Heat losses are proportional to current magnitude
squared. In addition, current magnitude constraints are often used as surrogates for
other constraints such as voltage stability. However, most IEEE test problems do
not include current magnitude limits on the transmission lines even though they are
an important aspect in model testing.
In the absence of these constraints, one approach is to put in constraints
based on physical characteristics. Often, there is little information about the lines. It
takes considerable time to develop constraints based on physical characteristics,
and the result may not be binding constraints.
The purpose of this paper is to develop a methodology for creating line
current magnitude constraints using a set of the IEEE test problems. We are
interested in creating binding constraints based on maximum current magnitude
rather than on apparent power on the lines or on voltage angle differences. The
approach we employ is to create constraints from the optimal solution without
these constraints. With these constraints, the resulting power flow problem has a
feasible solution. Subsequent testing helps to understand how constraining the line
flows affects the resulting power flow solution and solution time.

2. Notation
Variables and parameters are indexed over buses denoted by subscripts n
and m. Transmission lines are indexed by terminal buses n and m and k. For a
complex variable or parameter, the superscript r denotes the real portion and the
superscript j denotes the imaginary portion. For example, if x a jb, xr a, xj b
where j ‐1 1/2.
Variables
pn real power injected at bus n
qn reactive power injected at bus n
vrn real part of the voltage at bus n
vjn imaginary part of the voltage at bus n
vn the voltage magnitude at bus n
irn real part of the current injected at bus n
ijn imaginary part of the current injected at bus n
in the current magnitude of injection at bus n
i nmk real part of the current on line k at bus n to bus m
r

Page 4
 ijnmk imaginary part of the current on line k at bus n to bus m
inmk the current magnitude on line k at bus n to bus m
Parameters
cpn pn cost function of real power injected by a generator at bus n
cqn qn cost function of reactive power injected by a generator at bus n
cpln pn linear cost function of real power by a generator at bus n
cqln qn linear cost function of reactive power by a generator at bus n
bnmk imaginary part of the admittance matrix for line k between n and m
gnmk real part of the admittance matrix for line k between n and m
pminn minimum required real power at bus n
pmaxn maximum allowed real power at bus n
qminn minimum required reactive power at bus n
qmaxn maximum allowed reactive power at bus n
vminn minimum required voltage magnitude at bus n
vmaxn maximum allowed voltage magnitude at bus n
imaxnmk maximum allowed current magnitude on line k from bus n to bus m

3. Current‐Voltage ACOPF Model

The current‐voltage IV ACOPF formulation is used to find a set of voltages and
currents at each bus and currents on each transmission line that minimize the
objective function in terms of real and reactive power. More detail can be found
in Cain et al 2012 and O’Neill et al 2012 . The IV‐ACOPF model is:
Minimize ∑n cpn pn cqn qn
Subject to
irnmk gnmk vrn ‐ vrm ‐ bnmk vjn ‐ vjm
ijnmk bnmk vrn ‐ vrm gnmk vjn ‐ vjm
irn ∑mki rnmk
ijn ∑mki jnmk
pn vrnirn vjnijn
pminn pn pmaxn
qn vjnirn ‐ vrnijn
qminn qn qmaxn
vrn 2 vjn 2 vmaxn 2

vminn 2 vrn 2 vjn 2

irnmk 2 ijnmk 2 imaxnmk 2

Page 5
The objective function for test problems is generally presented as a quadratic
function, but many applications require linear bid functions. The IV‐ACOPF with
linearized objective function program has the same constraints as the nonlinear
model, but the objective function is a step function approximation of the nonlinear
objective function. In solving the ACOPF models, we allow for an infeasible answer
that is penalized by the amount the system is infeasible. We add cPenalty vr, vj, ir, ij, p,
q to the linearized objective function, where the quantity x is equal to max x, 0 ,
that is, if x is positive, x x; if x is negative, then x 0. For example, if the real
power is greater than the max, the objective function is penalized by that quantity
times some cost; if it is less than or equal to the maximum, there is no penalty.
cPenalty vr, vj, ir, ij, p, q ∑nmk cpenn vn ‐ vminn cpenn vmaxn ‐ v
cpenn inmk ‐ imaxnmk cpenn pn ‐ pmaxn cpenn pminn ‐ pn
cpenn qminn ‐ qn cpenn qn ‐ qmaxn
The system infeasibilities could possibly occur because the voltage is above the
maximum or below the minimum levels, because the current is above the maximum
level, or because real or reactive power violate maximum or minimum limits, as
detailed in the penalty cost. Here we set cpenn 105. A problem is declared
infeasible if the solution sets cPenalty to be greater than 10‐3.

4. Methods
We considered two methods for creating line current magnitude constraints.
In both these methods, we solved the alternating current optimal power flow model
without line constraints and extracted the optimal current magnitudes of each line
from the optimal solution.
In the first method, we constrained the current magnitude on each line to be a
fraction less than 1 of its optimal current magnitude in the unconstrained
problem. However, this method did not always return feasible solutions. This is
likely because the unconstrained optimization minimizes losses by lowering current
magnitudes subject to all other constraints in the model. Therefore, it may not be
feasible for all line currents to be restricted to something lower than in the
unconstrained solution.
In the second method, we constrained the current magnitude on each
transmission line to some fraction of the highest optimal current magnitude over all
lines in the unconstrained problem. This method returned local optimal solutions
since the problem is non‐convex for some limits, but, if the current magnitude
limit was too low, the solver could not find a feasible point. This current magnitude
limit level varied widely depending on the test problem.
Procedure:

Page 6
 1. Solve the ACOPF without any thermal line constraints
2. Extract the optimal line current magnitudes inmk* for current magnitude
for line k at bus n to m.
3. Let imax* maximum inmk* over all n, m, and k.
4. Solve the thermally constrained problem by including the constraint
inmk f *imax* for all combinations of n, m, and k where f is a parameter
with 0 f 1.

5. Computational Analysis
The input model data used is from the 14, 30, 57, and 118‐bus IEEE test system
data at http://www.ee.washington.edu/research/pstca/index.html. The generator
costs come from MATPOWER see Zimmerman et. al. 2011 . The quadratic cost
parameters are shown in the appendix. Where there are multiple transmission lines
between two nodes, the lines are aggregated into an equivalent single line between
the two nodes. The current magnitude measurement is per unit. Table 1
summarizes the test problem characteristics.

Table 1: IEEE Test System Data

Buses Lines Generators Demand Vmax Vmin Best
Number Total Capacity Known Value*
14 20 5 7.72 2.59 1.06 0.94 80.81
30 41 6 326.80 42.42 1.10 0.94 5.745
57 80 7 326.78 235.26 1.06 0.94 417.4
118 186 54 99.66 42.42 1.06 0.94 1297
*Quadratic Objective, Nonlinear Constraints
Hardware. The problems were solved on an Intel Xeon E7458 server with 8 64‐bit
2.4GHz processors and 64 GB memory. Minor differences in solution times were
recorded when the same problems were run at different times, but the differences
were small enough to be considered background noise.
Software. The procedures were formulated in GAMS 23.6.2. The nonlinear solver
used was IPOPT version 3.8.
Approach. The IV‐ACOPF formulation, as described above, was used to solve the test
cases shown below. For each test problem, a figure shows the impact on the line
current magnitudes on test systems were examined with the current magnitude
level at 75%, 80%, 90%, and 100% of the maximum current magnitude level in the
test problem without line current magnitude constraints. The optimal solution may
be infeasible, that is, have a penalty function greater than zero. A second figure
shows the line index versus the line current magnitude where the line index is
sorted from lowest to highest current magnitude level. The red solid line in each
graph is the chosen tight current magnitude limit and the dashed green line is the

Page 7
chosen loose current magnitude limit. A cross walk from line number to the buses
the line is connecting is in the appendix.
Testing Current Magnitude Constraints. From the testing described in the previous
sections, the uniform rule restricting current magnitude to the same percentage of
maximum current magnitude has a different impact on the different test cases.
Therefore, we picked different percentages of current magnitudes for each case and
created two sets of constrained problems: tight and loose. The tight current
magnitude level is near the current magnitude level where the problem becomes
infeasible. The loose current magnitude level is a level that constrains the problem
but is farther away from the point of infeasibility. These levels are shown in Table 2.

Table 2: Tight and loose testing levels of current magnitude constraints

14‐bus 30‐bus 57‐bus 118‐bus
Low Level Tight Constraint :
% of Maximum Optimal Current magnitude: 18.5 93.3 76.2 24.3
High Level Loose Constraint :
% of Maximum Optimal Current magnitude: 51.1 95.4 77.0 71.9
Low Tight Constraint Current magnitude Level 0.2264 0.3092 1.4027 0.9294
High Loose Constraint Current magnitude Level 0.6246 0.3162 1.4168 2.7536

Page 8
14‐Bus Problem. In the 14‐bus system when current is not constrained, one line has
nearly twice the current magnitude as the next highest current line. As the current
magnitude constraint is decreased, the effect on other lines is shown in figures 1 and
2. The current magnitudes that change the least under restriction are the lines with
comparatively lower current magnitudes. For the loose constraint, only one current
has a binding constraint. For the tight constraint, four constraints are binding or
near binding.
1.3
1.2 14‐bus Line Current Magnitude as a Function of
Line Current Magnitude Per Unit

1.1 % Restriction on the Line

1
100% Max Current
0.9
0.8
90% Max Current
0.7
0.6
80% Max Current
0.5
0.4
75% Max Current
0.3
0.2
0.1
0
0 2 4 6 8 10 12 14 16 18 20
Line Index
Figure 1. ‘Optimal’ 14‐Bus Current Magnitudes per Line under Percent Restrictions

1.3 14‐bus Line Current Magnitude with no, loosely, and

1.2 tightly constricting current limits
Line Current Magnitude Per Unit

1.1 No Limit
1
0.9 Loose Limit
0.8 Tight Limit
0.7
0.6 Loose Constraint
0.5
Tight Constraint
0.4
0.3
0.2
0.1
0
0 5 10 15 20
Line Index
Figure 2. 14‐Bus Optimal Current Magnitudes under the Loose and Tight Constraints.

Page 9
30‐Bus Problem. For the 30‐bus system, the current level could not be restricted
much lower than the maximum optimal current level 0.3314 before the system
became infeasible. Restricting the current shifts the line current magnitudes more
than in the 14‐bus problem; there are many lines where current actually increases.
While the difference between the tight 0.3092 and loose 0.3162 constraints is
small, these restrictions do have a large impact on the resulting current through
some of the lines. In both cases, only two lines have binding constraints
30‐bus Current per Line as a Function of % Restriction
0.35 on the Line
75%, 80%, and 90% are Infeasible
0.3
100% Max Current
0.25 90% Max Current
80% Max Current
Current Magnitude

0.2 75% Max Current

0.15

0.1

0.05

0
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42
Line Index

Figure 3. ‘Optimal’ 30‐Bus Current Magnitudes per Line under Percent Restrictions

0.35
30‐bus Line Current Magnitude with no, loosely, and tightly
constricting current limits
0.3 No Limit
Line Current Magnitude Per Unit

Loose Limit
0.25
Tight Limit
0.2 Loose Constraint
Tight Constraint
0.15

0.1

0.05

0
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42
Line Index

Figure 4. 30‐Bus Optimal Current Magnitudes under the Loose And Tight Constraints.

Page
10
57‐Bus Problem. For the 57‐bus system, the optimal solution has most of current
magnitudes of half or less the highest current magnitude, similar to the 14‐bus
problem. While about half of the lines exhibit minimal change, the other half have
currents that change significantly with the restriction. Like the 30‐bus problem, the
two current restrictions are close together, with 1.4027 for the loose limit and
1.4168 for the tight limit. In both cases, only two lines have binding constraints.
However, even this small difference in limits greatly impacts the current on some
lines.
2 57‐bus Current per Line as a Function of % Restriction
on the Line
Line Current Magnitude Per Unit

75% is Infeasible
1.5
100% Max Current
90% Max Current
1
80% Max Current
75% Max Current
0.5

0
0 20 40 60 80
Line Index
Figure 5. ‘Optimal’ 57‐Bus Current Magnitudes per Line under Percent Restrictions
2
57‐bus Line Current Magnitude with no, loosely, and tightly
Line Current Magnitude Per Unit

1.8 constricting current limits

1.6
1.4 No Limit
1.2 Loose Limit
1 Tight Limit
0.8 Loose Constraint
0.6 Tight Constraint
0.4
0.2
0
0 20 40 60 80
Line Index

Figure 6. 57‐Bus Optimal Current Magnitudes under the Loose and Tight Constraints.

Page
11
118‐Bus Problem. The 118‐bus system has several lines with higher currents,
similar to the 30‐bus system. The current can be restricted greatly before the
problem becomes infeasible. Reducing current by up to 75% of its original optimal
maximum does not have much impact on line currents except the highest ones, as
seen in Figure 7. Constraining the current with the loose limit impacts the currents
slightly; however, constraining the current with the tight limit has a major impact on
the current magnitudes, as in Figure 8. Under the loose constraint, only one line
constraint is binding. Under the tight current magnitudes, 22 of the lines have the
highest permissible current under this restriction, and even lines with lower
currents exhibit big differences from the unrestricted and loose current cases, both
higher and lower than before.
4.5
118‐bus Current per Line as a Function of % Restriction on
Line Current Magnitude Per Unit

4 the Line
3.5
100% Max Current
3
2.5 90% Max Current
2 80% Max Current
1.5
75% Max Current
1
0.5
0
0 50 100 150
Line Index
Figure 7. ‘Optimal’ 118‐Bus Current Magnitudes per Line under Percent Restrictions
4.5 118‐bus Line Current Magnitude with no, loosely, and
4 tightly constricting current limits
Line Current Magnitude Per

3.5 No Limit
3 Loose Limit
2.5 Tight Limit
Unit

2 Loose Constraint
1.5 Tight Constraint
1
0.5
0
0 50 100 150
Line Index
Figure 8. 118‐Bus Optimal Current Magnitudes under Loose and Tight Constraints.

Page
12
Normalized Objective Function Value as a Function of the Current Magnitude Limit
When the current magnitude is restricted, this limit is binding on at least one
line. As shown in Figure 9, the different IEEE test cases exhibit infeasibility
according to the solver at different percentages of their maximum current
magnitude. The denominator of the normalized objective function is the objective
function without constraints.
1.20

14 Bus System
Normalized Objective Function

1.15 30 Bus System

57 bus System
118 Bus System

1.10

1.05

1.00
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Fraction Restriction on Maximum Optimal Line Current

Figure 9. Normalized Objective Function versus Current Magnitude Limit

Normalized Solution Time as a Function of the Current Magnitude Limit

As shown in Figure 10, the computational time to solve find a local optimal
solution or declare the problem infeasible the constrained problem only seems to
grow greatly when the current magnitude is restricted to 5‐15% of its original value.
When it is restricted to 20% or more of its original value, the added constraints do
not seem to increase or decrease the computational time predictably.

Page
13
 100

14 Bus
30 Bus
57 Bus
Normalized CPU Time 10 118 Bus

0.1

85

90

95

00
05

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80
0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

1.
Fraction Restriction on Maximum Optimal Line Current

Figure 10: Impact of Current magnitude Constraints on CPU Time

Effect of Current Magnitude Constraints on the Optimal Value. When binding current
magnitude constraints are added to the model, the feasible region shrinks and the
objective function value increases, as shown in Figures 11 and 12 and Table 3. The
objective function value of the 14‐bus problem with the tight constraint is 30
percent greater. For other problems, the increases in objective function value with
current magnitude constraints were 6 percent or less.

Effect on Objective Value of Current Line Constraints on

Fully Nonlinear IV ACOPF

1.32
Ratio of Objective Function to Unconstrained

1.28

1.24
objective Function

1.20
Tight Constraint
1.16
Loose Constraint
1.12

1.08

1.04

1.00
14 Bus 30 Bus 57 Bus 118 Bus

Figure 11. Value of the quadratic objective function with current magnitude
constraints

Page
14
 Effect on Objective Value of Current Line Constraints on
Nonlinear IV ACOPF with Linear Objective Function

Ratio of Objective Function to Unconstrained 1.30

1.25
Objective Function

1.20
Tight Constraint
1.15 Loose Constraint

1.10

1.05

1.00
14 Bus 30 Bus 57 Bus 118 Bus

Figure 12. Value of the linear objective function with current magnitude constraints

Table 3: Objective function value for each test system and constraint type
Problem Type Constraint type 14‐bus 30‐bus 57‐bus 118‐bus
Tight Constraint 105.4 5.89 421.5 1364.9
Quadratic Objective
Loose Constraint 85.3 5.79 419.2 1300.1
Nonlinear Constraints
Unconstrained 80.8 5.75 417.4 1296.6
Tight Constraint 107.4 6.10 432.2 1388.4
Linear Objective
Loose Constraint 86.5 5.98 425.5 1315.5
Nonlinear Constraints
Unconstrained 82.8 5.92 423.8 1311.5

Effect of Current Magnitude Constraints on the Solution Time. Current magnitude

constraints also impact the solution time of each test system as shown in Figures 13
and 14 and Table 4. Adding constraints at least doubles the solution time of the
problem. For the 30‐bus and 118‐bus problems, the more tightly constrained
problem solves faster than the less constrained problem. For the 14‐bus and 118‐
bus problems, the less constrained problem solves faster than the more constrained
problem. The 57‐bus and 118‐bus solution time took significantly longer than the
14‐bus and 30‐bus systems, but solution time is not solely based on the number of
buses. The linear objective function required considerably more time to solve.

Page
15
 Effect on CPU Time of Current Line Constraints on the Fully
Nonlinear IV ACOPF

Ratio of CPU Time to Unconstrained CPU Time

4.5

3.5

3
Tight Constraint
Loose Constraint
2.5

1.5

1
14 Bus 30 Bus 57 Bus 118 Bus

Figure 13. CPU Time with current magnitude constraints of quadratic objective
Effect on CPU Time of Current Line Constraints on the
Nonlinear IV ACOPF with Linear Objective Function
Ratio of CPU Time to Unconstrained CPU Time

18
17
16
15
14
13
12
11
10 Tight Constraint
9 Loose Constraint
8
7
6
5
4
3
2
1
14 Bus 30 Bus 57 Bus 118 Bus

Figure 14. CPU Time of the current magnitude constrained linear objective

Table 4. CPU Time for all bus systems, constraint types, and model types
Problem Type Constraint type 14‐bus 30‐bus 57‐bus 118‐bus
Tight Constraint 0.72 1.80 8.34 28.95
Quadratic Objective
Loose Constraint 0.58 1.94 7.77 33.33
Nonlinear Constraints
Unconstrained 0.28 0.78 1.95 8.13
Tight Constraint 1.16 1.55 8.42 32.20
Linear Objective
Loose Constraint 0.91 2.88 12.05 38.75
Nonlinear Constraints
Unconstrained 0.31 0.48 0.70 2.61

Page
16
 6. Conclusions
For each test problem, one single limit is applied to all lines that makes the
optimal solution without these limits infeasible. We solve the resulting problem
using the IV‐ACOPF formulation. For each problem we develop a ‘tight’ and a ‘loose’
constraint.
For the 14, 30, 57 and 118‐bus problems, creating line current magnitude
constraints for the ACOPF problem can result in infeasible problems. As one
tightens the current magnitude constraints, the objective function increases
gradually at first, then increases exponentially near the point of infeasibility.
Different test problems exhibit different characteristics in the line current
magnitude distribution and at what current magnitude level constraint the problem
becomes infeasible.
The current magnitude constraints also increase the solution time, although
stricter constraints do not necessarily increase the solution time more than looser
constraints. For problems like this case where the problem becomes infeasible
quickly, it may work well to restrict only some of the line current magnitudes rather
than all of them.
Including these constraints in the ACOPF increases the solution time between
2 to 20 times and objective function up to 25 percent.

Page
17
References
Mary B. Cain, Richard P. O’Neill, Anya Castillo, “History of Optimal Power Flow and
Formulations,” FERC Staff Technical Paper, December 2012
http://www.ferc.gov/industries/electric/indus‐act/market‐planning/opf‐papers/acopf‐1‐
history‐formulation‐testing.pdf

Richard P O’Neill, Anya Castillo and Mary Cain, “The IV Formulation and Linear
Approximations of the AC Optimal Power Flow Problem,” FERC Staff Technical
Paper, December 2012
http://www.ferc.gov/industries/electric/indus‐act/market‐planning/opf‐papers/acopf‐2‐
iv‐linearization.pdf

University of Washington, Dept. of Electrical Engineering, “Power System Test Case

Archive,” May 2010. Available:
http://www.ee.washington.edu/research/pstca/index.html.

R.D. Zimmerman, C.E. Murillo‐Sánchez, and R.J. Thomas, “MATPOWR Stead‐State

Operations, Planning and Analysis Tools for Power Systems Research and
Education,” Power Systems, IEEE Transactions on, vol. 26, no. 1, pp. 12‐19, Feb
2011.

Page
18
Appendix
The generator costs take on the form cost aP bP2 .0001*|Q|, where P is the
real power and |Q| is the magnitude of reactive power q. We list all generator costs
used for each test system.
Table A1: Generator Cost Coefficients for the 14, 30 and 57‐bus Problems
14‐bus Cost Coefficient 30‐bus Cost Coefficient 57‐bus Cost Coefficient
Generator a b Generator a b Generator a b
1 0.04303 20 1 0.02000 2.00 1 0.07758 20
2 0.25000 20 2 0.01750 1.75 2 0.01000 40
3 0.01000 40 13 0.02500 3.00 3 0.25000 20
6 0.01000 40 22 0.06250 1.00 6 0.01000 40
8 0.01000 40 23 0.02500 3.00 8 0.02222 20
27 0.00834 3.25 9 0.01000 40
12 0.03226 20

Table A2: 118‐bus Generator Costs

Generator Cost Generator Cost Generator Cost
Coefficient Coefficient Coefficient
a b a b r a b
1 0.01000 40 46 0.52632 20 87 2.50000 20
4 0.01000 40 49 0.04902 20 89 0.01647 20
6 0.01000 40 54 0.20833 20 90 0.01000 40
8 0.01000 40 55 0.01000 40 91 0.01000 40
10 0.02222 20 56 0.01000 40 92 0.01000 40
12 0.11765 20 59 0.06452 20 99 0.01000 40
15 0.01000 40 61 0.06250 20 100 0.03968 20
18 0.01000 40 62 0.01000 40 103 0.25000 20
19 0.01000 40 65 0.02558 20 104 0.01000 40
24 0.01000 40 66 0.02551 20 105 0.01000 40
25 0.04545 20 69 0.01936 20 107 0.01000 40
26 0.03185 20 70 0.01000 40 110 0.01000 40
27 0.01000 40 72 0.01000 40 111 0.27778 20
31 1.42857 20 73 0.01000 40 112 0.01000 40
32 0.01000 40 74 0.01000 40 113 0.01000 40
34 0.01000 40 76 0.01000 40 116 0.01000 40
36 0.01000 40 77 0.01000 40
40 0.01000 40 80 0.02096 20
42 0.01000 40 85 0.01000 40

Page
19
 Table A3: 14 and 30‐bus Line Index Mapping
14‐bus 30‐bus
Line Buses Connected Line Buses Connected
Index To From Index To From
1 1 2 1 6 8
2 1 5 2 12 13
3 2 3 3 21 22
4 4 5 4 2 6
5 2 4 5 4 6
6 5 6 6 1 2
7 2 5 7 1 3
8 7 9 8 2 4
9 4 7 9 3 4
10 6 13 10 5 7
11 4 9 11 2 5
12 3 4 12 4 12
13 9 14 13 27 28
14 7 8 14 25 27
15 6 12 15 15 23
16 6 11 16 6 7
17 9 10 17 9 10
18 13 14 18 6 9
19 10 11 19 8 28
20 12 13 20 10 20
21 15 18
22 10 21
23 10 17
24 12 15
25 24 25
26 10 22
27 12 16
28 27 30
29 27 29
30 19 20
31 6 28
32 12 14
33 6 10
34 22 24
35 23 24
36 18 19
37 25 26
38 29 30
39 16 17
40 14 15
41 9 11
35 23 24
36 18 19
37 25 26
38 29 30
39 16 17
40 14 15
41 9 11

Page
20
 Table A4: 57‐bus Line Index Mapping
57‐bus 57‐bus
Buses Connected Buses Connected
Line Index To From Line Index To From
1 8 9 40 50 51
2 14 46 41 38 44
3 10 51 42 35 36
4 24 26 43 24 25
5 7 8 44 4 5
6 7 29 45 6 7
7 15 45 46 41 43
8 1 15 47 52 53
9 9 55 48 54 55
10 3 15 49 11 41
11 1 2 50 41 42
12 1 17 51 38 49
13 13 49 52 12 16
14 9 11 53 3 4
15 46 47 54 25 30
16 14 15 55 49 50
17 6 8 56 34 35
18 2 3 57 53 54
19 9 10 58 32 34
20 9 13 59 41 56
21 1 16 60 10 12
22 4 18 61 12 17
23 28 29 62 18 19
24 11 43 63 48 49
25 27 28 64 30 31
26 5 6 65 36 40
27 44 45 66 20 21
28 4 6 67 32 33
29 12 13 68 37 39
30 37 38 69 22 38
31 38 48 70 22 23
32 11 13 71 56 57
33 13 14 72 23 24
34 9 12 73 39 57
35 36 37 74 42 56
36 26 27 75 40 56
37 29 52 76 19 20
38 47 48 77 31 32
39 13 15 78 21 22

Page
21
 Table A5: 118‐bus Line Index Mapping
118‐bus 118‐bus 118‐bus 118‐bus
Line Buses Line Buses Line Buses Line Buses
Index To From Index To From Index To From Index To From
1 9 10 46 77 78 91 51 52 136 51 58
2 8 9 47 66 67 92 104 105 137 105 106
3 5 8 48 45 49 93 93 94 138 100 101
4 37 38 49 76 77 94 2 12 139 34 43
5 17 30 50 22 23 95 103 104 140 105 107
6 49 66 51 59 61 96 75 77 141 29 31
7 26 30 52 49 50 97 55 59 142 32 114
8 89 92 53 47 69 98 20 21 143 70 71
9 68 116 54 100 106 99 34 36 144 19 20
10 68 69 55 56 59 100 82 83 145 70 75
11 64 65 56 94 95 101 37 40 146 27 32
12 59 63 57 100 104 102 99 100 147 43 44
13 89 90 58 75 118 103 62 67 148 106 107
14 23 25 59 85 88 104 68 81 149 7 12
15 80 81 60 74 75 105 1 3 150 60 62
16 63 64 61 94 100 106 54 59 151 40 42
17 38 65 62 92 93 107 101 102 152 52 53
18 65 66 63 59 60 108 40 41 153 3 12
19 25 26 64 11 12 109 78 79 154 90 91
20 25 27 65 21 22 110 12 117 155 12 16
21 77 80 66 62 66 111 23 24 156 15 33
22 4 5 67 48 49 112 17 31 157 32 113
23 60 61 68 80 97 113 27 115 158 105 108
24 61 64 69 110 111 114 80 99 159 56 58
25 69 75 70 49 69 115 96 97 160 24 72
26 69 70 71 37 39 116 53 54 161 86 87
27 34 37 72 92 94 117 16 17 162 55 56
28 88 89 73 45 46 118 41 42 163 71 73
29 15 17 74 11 13 119 47 49 164 98 100
30 42 49 75 110 112 120 91 92 165 39 40
31 5 6 76 103 110 121 82 96 166 13 15
32 23 32 77 27 28 122 77 82 167 1 2
33 100 103 78 103 105 123 85 86 168 108 109
34 69 77 79 35 37 124 70 74 169 71 72
35 65 68 80 31 32 125 83 84 170 114 115
36 8 30 81 80 98 126 56 57 171 15 19
37 5 11 82 50 57 127 33 37 172 54 56
38 30 38 83 6 7 128 95 96 173 35 36
39 49 54 84 61 62 129 17 113 174 19 34
40 17 18 85 44 45 130 94 96 175 109 110
41 79 80 86 83 85 131 18 19 176 76 118
42 85 89 87 80 96 132 92 100 177 24 70
43 4 11 88 92 102 133 12 14 178 54 55
44 49 51 89 46 47 134 46 48 179 14 15
45 3 5 90 84 85 135 28 29

Page
22