International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 6 (2017) pp.

969-975
© Research India Publications. http://www.ripublication.com

Application of MATPOWER for the Analysis of Congestion in Power

System Network and Determination of Generator Sensitivity Factor

Kaushik Paul1
Electrical & Electronics Engineering Department, National Institute of Technology,
Jamshedpur, Jharkhand,831014, India.

Niranjan Kumar2
Electrical & Electronics Engineering Department, National Institute of Technology
Jamshedpur, Jharkhand,831014, India.

Abstract technology. The simulation and the analysis of the large

power system network can be done easily with the help of the
The congestion management is considered as one of the key
modern computer technology. The modern computer can
feature of the independent system operator in deregulated
efficiently react and provide useful data and information
power system environment. The rescheduling of the generator
regarding the various responses which have made it feasible
is one of the most vital method for congestion management.
for the integration of these computers in the modern power
The congestion in transmission line is caused due to increase
system courses [1].
in power flow should be identified and proper measures
should be taken by the system operator for relieving the There have been a large number of software packages like
congestion. The intent of this paper is to demonstrate the ETAP, ERACS, PSCAD, Power World Simulator which are
application of MATPOWER, an open source MATLAB used to analyze and solve power system problems but these
simulation software, which is much efficient in solving the software packages require a pretty decent knowledge of
rigorous task of power flow for the identification of the modeling and simulation for solving the problems related to
congested line in the power system network and the power system analysis. Moreover these software packages do
determination of the generator sensitivity factor by changing not give the facility to add new algorithms to it nor allow the
the MATPOWER source code. In this paper one illustrative users to change the source code [2]. This limitation on the
example has been presented using Newton Raphson power software tools is not desirable for the researchers. Moreover, it
flow for 5 bus and 30 bus cases from the MATPOWER is highly time consuming and difficult for students to design
library. Moreover the paper also presents MATPOWER as their own codes for the power flow analysis. This problem can
one of the most important research tool which will facilitate be easily solved by the use of the MATPOWER. There are
the study of the power system problem for both the several other power system software tools [3-5] which uses
undergraduate and the research scholars. MATLAB as its common platform and provide the facility for
the modification in their codes. Moreover several M-file
Keyword: MATPOWER, MATLAB, Congestion, Generator
functions which also use MATLAB have been designed by
sensitivity, Jacobian
Sepasian and Seifi [1] for the better understanding of the
power system analysis problem for the students and research
scholars. However in addition to these, MATPOWER is very
INTRODUCTION simple and robust. It is much more efficient and flexible in
The restructuring in the power system industry have led to the computing and displaying much detailed results.
increase in the power trade. This increase in the power trading Since power flow studies is considered as one of the most
in the deregulated power environments have led to the essential tool of the power system courses [6], the cases
extensive use of the transmission line hence leading to the
illustrated in this literature is only confined to the analysis of
congestion in the transmission line. The insufficient the power flow in the power system network for the analysis
transmission capacity of the transmission lines results in the of the congestion in the power system network and
congestion of the lines. The congestion in the power
determination of Generator Sensitivity Factor (GSFG). As
transmission line is not desirable for the reliable operation of MATPOWER uses the MATLAB platform, a good
the power system network, so the identification of the introduction to the MATLAB is necessary from the electrical
congested line is important for the reliable operation of the
engineering point of view. A good introduction to the
power system network. The amount of the power flow in the MATLAB can be found in [7]. In this paper section II
transmission line can be determined by the analysis of the discusses the definition and features of MATPOWER. The
power flow in the lines. The analysis of the power flow analysis of congested line and evaluation of generator
involves a lot of mathematics and complex calculation which sensitivity factor is illustrated with 5 bus and 30 bus test case
may be quite time consuming, laborious and also erroneous
in section III and section IV respectively by the use of
for the students and research scholars This situation has MATPOWER. Section V concludes the paper.
changed drastically with the advancement in the computer

969
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 6 (2017) pp.969-975
© Research India Publications. http://www.ripublication.com

MATPOWER
MATPOWER is considered as an open source power system
analysis programming tool for solving the problems related to
power flow and optimal power flow. The MATPOWER runs
on the MATLAB platform best suited with the version of
MATLAB 6 and above. The MATPOWER package was
developed at the CORNELL University and can be
downloaded from their website [8]. The basic purpose of this
software package is to provide facilities to the research
scholars, education professionals and also to the industry
related issues [9]. The codes of the MATPOWER are simple Figure 1: M-File of branch data for 5 bus system
and the user can modify the codes according to their needs.
After downloading the MATPOWER, the MATLAB platform
needs to be opened and the core MATPOWER function and
test script to be added to the MATLAB path. The installation
of MATPOWER 5.1 version is shown as
C:\ProgramFiles\MATLAB\2013b\matpower5.1. After
selecting the path, the command window has to be opened and
the required M-File name to be typed followed by pressing
enter. In order to run a power flow for 30 bus system for
example type:
>>runpf(‘case30’)
In the command window followed by enter button. This will
give the results of the power flow for a 30 bus system which is Figure 2: Power flow under normal operating condition of 5
already present in the MATPOWER. The MATPOWER uses bus system
Gauss Seidal, Newton, DC power flow, XB fast-decouple and
BX fast-decouple as solvers for solving the power flow The congestion in the line may be due to several cases like
analysis. line outage, increase in the load, tripping of the generator etc.
Here the increase in the load is taken as the case study for the
The MATPOWER provides us with the advantage of adding analysis of the congested line. The load is increased in every
new algorithm or to modify its source code as per the bus by a factor 1.5 and line 4-5 is outage. Power demand in
requirement of the research problem. The other advantage is each load bus can be changed by the following steps,
that it is an open source package. The MATPOWER can be
used to find the congested line and the generator sensitivity Step1: Loading the case data. In this the load flow input data
index for the generators which are to be rescheduled for the is prepared.
purpose of the congestion management. Step2: Changing the real and the reactive power demand. The
real and reactive power demand data are changed from a
previous value to a new value.
DETERMINATION OF CONGESTED LINE
Step3: Running the power flow
The congestion management is one of the most vital technical
issues in deregulated power system environment. In mpc=loadcase('case5'); % read the load flow input data
deregulated environment the transmission congestion occur % increase the real power demand at bus
when there is an insufficient transmission capacity to mpc.bus(2,PD)=450;
2 to 450 MW, PD load demand (MW)
simultaneously accommodate all the transmission constraints.
% increase the reactive power demand at
The determination of the congested line in the power system
mpc.bus(2,QD)=147.91; bus 2 to 147.91 MVAr, QD load demand
network is analyzed by taking the example of 5 bus system of (MVAr)
the MATPOWER 5.1. The normal power flow is run by using
the MATPOWER code: % increase the real power demand at bus
mpc.bus(3,PD)=450;
3 to 450 MW
>>runpf(‘case5’)
% increase the reactive power demand at
mpc.bus(3,QD)=147.91;
This MATPOWER code runs the power flow analysis for 5 bus 3 to 147.91MVAr
bus system. The M file of the branch data for 5 bus system is % increase the real power demand at bus
represented in figure 1 in which rate represents the MVA limit mpc.bus(4,PD)=600;
4 to 600 MW
of the line. The normal power flow status of the 5 bus system
is given in figure 2. % increase the reactive power demand at
mpc.bus(4,QD)=197.20;
bus 4 to 197.20MVAr
runpf(mpc); % run AC power flow

970
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 6 (2017) pp.969-975
© Research India Publications. http://www.ripublication.com

The generator sensitivity factor helps to determine which

generator will take part in the rescheduling in order to remove
congestion. A positive value of the generator sensitivity factor
indicates there is a decrease in the power flow in the
congested line with the decrease in the generation and on the
other hand the negative value of generator sensitivity factor
represents a decrease in the power flow in the congested line
with the increase in the generation. The generators having
non-uniform and large magnitude of GSFG will take part in
the rescheduling for removal of congestion. The generator
sensitivity factor (GSFG) is computed using the following
Figure 3: Bus data of 5 bus system with increase in load procedure [10].
2
Pij  ─Vi Gij  ViV j Gij cos ( i ─ j )  ViV j Bij sin ( i ─ j ) (2)

Where 𝜃i and 𝜃j = bus voltage angle at bus i & j respectively.

Vi and Vj are the bus voltages at ith and jth bus respectively. Gij
and Bij are the conductance and susceptance of the line
connected between i and j. Neglecting the PV coupling the
GSFG Expression can be written as
Pij θ Pji θ j
i
GSFg   (3)
θi PG θ j PG
g g

Figure 4: Congestions due to increase in the power flow Where,

Pij
 ViV j Gij sin ( i ─ j )  ViV j Bij cos ( i ─ j ) (4)
From the branch data in figure 4 it is observed that due to θi
increase in load to 1.5 times and line 4-5 outage, line 1-2 and
line 1-5 gets congested
Pij
TABLE I: POWER FLOW IN CONGESTED LINES IN 5 BUS ViV j Gij sin ( i ─ j )  ViV j Bij cos ( i ─ j ) (5)
θ j
SYSTEM

Line Power flow Line limit Pij Pij

i.e 
θi θ j
1-2 429.86 400
1-5 465.11 400
The injection of active power Pi at bus s can be represented as
N
The analysis of the congested line with a test case of 30 bus Pi  Vi  V j (Gij cos ( i   j )  Bij sin ( i   j )) (6)
system from the MATPOWER library is also shown with j
increase in the load up to 1.5 times in table II.
2 N
TABLE II: POWER FLOW IN CONGESTED LINES IN 30 BUS Pi  Vi Gij  Vi  V j ( Gij cos ( i   j )  Bij sin ( i   j ))
j
SYSTEM

Line Power flow(MW) Line limit(MW) Now differentiating the equation with respect to 𝜃i and 𝜃j, we
obtain
6-8 40.39 32
Pij
V V G sin ( i ─ j )  ViV j Bij cos ( i ─ j ) (7)
θi i j ij
GENERATOR SENSITIVITY FACTOR
Generator sensitivity factor defined as the ratio of change in Pij
 ViV j Gij sin ( i ─ j )  ViV j Bij cos ( i ─ j ) (8)
real power flow in a transmission line k which is connected θ j
between bus i and bus j to the change in the power generation
of the generator G. Mathematically generator sensitivity By neglecting the PV coupling the relation between
(GSFg) of a line can be represented as following. incremental change in active power at system bus and phase
angle of voltage can be written in matrix form as
GSFG=∆Pij/∆PGg (1)
∆Pij= change of real power flow of congested line k [ P ]n*1  [J11 ]n*n * [  ]n*1 (9)

∆PGg=change of real power of gth generator

971
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 6 (2017) pp.969-975
© Research India Publications. http://www.ripublication.com

1
[  ]n*1  [ H ]n*n * [ P ]n*1 (10)
A. Generator Sensitivity Factor Determination with
[ H ]n*n  [ J11 ]n*n (11) MATPOWER
The calculation of G and B matrices involves the computation
n= number of buses of the Y bus for the system. The Y bus matrix can be obtained
by altering the coding in the MATPOWER codes: the
[  ]n*1  [ M ]n*n * [ P ]n*1 (12) alteration in the MATPOWER code is given below
1
M  n*n   H  n*n (13)

0 0 0 
 
 
 θ2 θ2 
0 
 H    P2 Pn  (14)
n*n  
 
 θn θn 
0 
 P2 Pn 
 nn

[  1 ]  [ M 1 ] * [ P1 ] (15)

1
0 0    
[  ]n*1    * P 
n*1  1 (16)
0 [ M 1] 1


The second term of the sum in (16) vanishes as being the

change in phase angle of slack bus is zero. Accordingly,
reduces to
0 0   
[  ]n*1    * P 
n*1 (17)
0 [ M 1]

Thus value of the generator sensitivity factor is obtained from

the above method with respect to the slack bus.
GSFG denotes how much active power flow over a
transmission line connecting bus-i and bus-j would change
due to active power injection by generator g. The system
operator selects the generators having non uniform and large
magnitudes of sensitivity values as the ones most sensitive to
the power flow on the congested line and to participate in Figure 5: Modification in MATPOWER Source Code
congestion management by rescheduling their power outputs. The real and the imaginary part of the admittance matrix are
represented by G and B matrices respectively. The values of G
matrix and the B matrix obtained for a 5 bus system computed
by the altering the MATPOWER coding is given below

 22.2507 3.5235 0 3.2569 15.4703

 3.5235 12.6911 9.1676 0 0 
 
G   0 9.1676 12.5013 3.3337 0 
 
 3.2569 0 3.3337 9.9242 3.3337 

 15.4703 0 0 3.3337 18.8040 


972
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 6 (2017) pp.969-975
© Research India Publications. http://www.ripublication.com

 222.4844 35.2348 0 32.5690 154.7030 

 35.2348 126.8979 91.6758 0 0 
 
B 0 91.6758 124.9999 33.3367 0 
 
 32.5690 0 33.33367 99.2324 33.3367 

 154.7030 0 0 33.3367 188.0206 


The generator sensitivity factor is obtained by substituting the  i  j

Pij Pij  j
obtaining the values of and from the inverse of J11.
 i Pg Pg
values of , , and in equation 3. The values of
θi θ j Pg Pg
The modification in the coding is shown in the figure 5. In
Pij Pij order to obtain the generator sensitivity factors the values of
and is calculated from the equation 4. The values of  j
θi θ j  i
and is substituted from M matrix in equation 3. M
Pg Pg
 i  j
the other half of the equation 3 i.e. the and can be
Pg Pg matrix is the inverse of the J11 matrix. J11 is a component of
obtained from the inverse of J11 matrix from the Jacobian the Jacobian Matrix. The value of M matrix for the 5 bus
matrix. The MAPOWER built in source code is modified for system is given below

 0.0212 0.0093 0.0211 0.0127 

 0.0090 0.0212 0.0090 0.0180 
M   
 0.0212 0.0093 0.0275 0.0127 
 
 0.0124 0.0180 0.0123 0.0246 

The value of M matrix is required to determine the value of The generator sensitivity factors for the 5 bus system are
 i  j shown in the figure 6 and figure 7 for the congested line 1-2
and . Finally we get GSFG values.
Pg Pg and 1-5. Same method has been followed to determine the
GSFG along the congested line of a 30 bus system. The
TABLE III: GENERATOR SENSITIVITY FACTOR FOR 5 BUS generator sensitivity values obtained are with respect to the
SYSTEM slack bus as the reference. So the sensitivity of the slack bus
generator to any congested line in the system is always zero.
Congested GSF1 GSF2 GSF3 GSF5 The generators having large and non uniform values of
Line sensitivity factor will take part in the congestion management.
1-2 0.2670 -0.0468 0.4580 -0.2745 It is observed from Figure 6 and Figure 7 that all the
1-5 0.7140 -0.1520 0.1580 0.1892 generators are having non-uniform GSFG values, so all the
generators will take part in congestion relief. The generators
with positive GSFG values will decrease their generation and
TABLE IV: G ENERATOR SENSITIVITY FACTOR FOR 30 the generators with negative GSF values will increase their
BUS SYSTEM generation. In case of the 30 bus system in figure 8 the
generators G2, G22 and G27 having high sensitivity index
Congested GSF2 GSF13 GSF22 GSF23 GSF27 will take part in the congestion management scheme.
Line
6-8 -0.0231 -0.0375 -0.0831 -0.0410 0.0823

973
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 6 (2017) pp.969-975
© Research India Publications. http://www.ripublication.com

Generator sensitivity factor along the congested line 1to2

0.6
Value of Generator Sensitivity

0.4
G1
0.2
Factor

G2
0
G3
-0.2 G5
-0.4
Generators

Figure 6: Generator sensitivity factor along the congested line 1-2 for 5 bus system

Generator sensitivity factor along the congested line 1to 5

0.8
Value of Generator Sensitivity

0.6
G1
0.4
G2
Factor

0.2
G3
0 G5
-0.2
Generators

Figure 7: Generator Sensitivity Factor along the congested line 1- 5 for 5 bus system

Generator sensitivity factor for congested line 6 to 8 of 30 bus system

Value of Generator Sensitivity

0.1
0.05
G2
0
Factor

G13
-0.05
-0.1 G22
-0.15 G23
-0.2 G27
-0.25
Generators

Figure 8: Generator Sensitivity Factor along the congested line 6-8 line for 30 bus system

CONCLUSION the huge amount of time invested in laborious efforts to

design the power flow programs and solve complex
This paper has presented the use of MATPOWER as an
calculations in the respective field of the power system
efficient tool for the analysis of the congestion in the power
studies. The inbuilt power flow solver is well efficient to deal
system network and the calculation of the generator sensitivity
with small as well as large system. The MATPOWER also
factor for the selection of the generators which has to be
displays several other details like the active and reactive
rescheduled for the elevation of congestion in the power
power losses, magnitude of the maximum and minimum
system network. Apart from this MATPOWER is a proficient
voltages etc. The ability of the user to change the source code
tool for high level power flow studies that will enhance the
according to their requirement gives an advantage to deal with
performance of the research scholars and students by reducing

974
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 6 (2017) pp.969-975
© Research India Publications. http://www.ripublication.com

complex problem like calculation of the generator sensitivity

factor and also the other contingency analysis like line outage.

REFERENCES
[1] H.Seifi and M.S Sepasian, 2011, "Electric Power
System Planing,”.Spinger (ISBN 978-3-642-17989-
1).
[2] F. Milano, 1. Vanfretti, and J. C. Morataya, 2008
"An open source power system virtual laboratory:
The PSAT Case and Experience," IEEE Transactions
on Education, vol. 51, pp. 17-23.
[3] C. A. Canizares and F. Alvarado, 1999
"UWPFLOW: continuation and direct methods to
locate fold bifurcations in AC/DC/FACTS power
systems," University of Waterloo.
[4] M. A. H. L. Chen, C. O. Nwankpa, H. G. Kwatny,
and X. Yu, 1996 "Voltage stability toolbox: an
introduction and implementation," Proc. of 28 th
North American Power Simposium.
[5] F. Milano, "An open source power system analysis
toolbox," IEEE Transactions on Power Systems, vol.
20, pp. 1199-1206,2005.
[6] J.Grainger and W.D Stevension .Jr, 2003 “POWER
SYSTEM ANALYSIS”, McGraw Hill Education; 1
edition.
[7] Stephen J. Chapman, “MATLAB Programming for
Engineers,” Wadsworth Publishing Co Inc; 4th
edition (8 November 2007)
[8] http://www.pserc.cornell.edu/matpower/
[9] R. D. Zimmerman and C. E. Murillo-Sanchez,2013
"MATPOWER 5.1 User's Manual, 2013. "
[10] FarzadVazinram, MajidGandomkar, Javad Nikoukar,
2013 “Optimal Active Power Rescheduling of
Generators for Congestion Management Based On
Big Bang-Big Crunch Optimization Using New
Definition of Sensitivity” International Journal of
Engineering and Advanced Technology (IJEAT),
Volume-3, Issue-2,pp. 44-52.

975