2019 4th International Conference on Electrical, Electronics, Communication, Computer Technologies and Optimization Techniques (ICEECCOT)

Multiple Contingency Analysis for optimal

placement and estimate the value of SVC for power
loss reduction employing Particle Swarm
Optimization
Mohammed Shuaib Iqbal Sujatha B.C
Dept of Electrical Engg Dept of Electrical Engg
UVCE, Bengaluru, Karnataka, India UVCE, Bengaluru, Karnataka, India
[email protected] [email protected]

Abstract— The development of FACTs technology has been a The main objective of utilizing a Static VAR compensator
great help in enhancing the function of power systems as it in a power system is to enhance the power transmission
minimizes the power system instability problem, minimize efficiency, with a given transmission structure, from the
losses and voltage profile. Due to high capitals involved with generators to the loads. The voltage control is influenced
Flexible AC transmission systems (FACTs) devices, proper
alternately by SVC which cannot generate or absorb true
planning in the early phase of commissioning is essential so as
to achieve maximum feasible use of them. Positioning FACTs power in power transmission structure. In order to have
devices in proper place can assist the objective of enhancing desired power flow under feasible system disruption and
voltage levels and minimizing losses in the structure. In this contingencies, the output reactive power of compensator is
paper, Particle Swarm Optimization (PSO) is used to discover modified to control the voltage at given terminals of
the ideal place of Static VAR Compensator (SVC) for the transmission structure. For mitigating power structure
increase in load condition and loss of generation. An losses, SVC is placed at ideal position. In [2], PSO is
appreciable reduction in active power loss and improvement in utilized for ideal placement of SVC for mitigating power
voltage profile are observed. The MATLAB program with loss and reducing size of SVC.
MATPOWER is used to demonstrate the procedure on IEEE
30 bus test system.
In this paper, PSO is used for finding an ideal position and
value of SVC for mitigating true power loss, decreasing
Keywords— Static VAR Compensator (SVC), Particle Swarm voltage distortion and to adjust the value of SVC. PSO is
Optimization (PSO), Contingency Analysis. used for figuring out an optimized problem for the position
of SVC. To demonstrate the efficiency of the technique
I. INTRODUCTION IEEE 30 bus structure is employed. Typically, the
computational burden in this area has reduced either by
Power system structures are complex systems that are
decreasing the required time to compute contingency [9-12]
nonlinear, non-stationary, and disposed to disruption and
or by minimizing the number of cases to be computed.
faults. By enhancing voltage profiles, increasing the
transmission size and loss reduction and others power
system can be strengthened. II. STATIC VAR COMPENSATOR (SVC)

Increase in power required due to increase in inhabitants

The SVC is a static VAR generator/load whose output is
demands the existing transmission structures to transfer
arranged to switch capacitive or inductive current so as to
more power. Therefore, in designing and functioning of the
vary power system variable [3, 4]. Conventionally, the
power system the power flow managing in the network is an
terminal bus voltage is the power system control variable.
important problem. By using FACTs devices system can be
SVC’s have two conventional forms. One form is solely of a
enhanced for the power flow controlling and transmission
Thyristor-Controlled Reactor (TCR) and the other is solely
size [1].
of a Thyristor-Switched Capacitor (TSC). In order for SVC
to behave as a fixed capacitor or an inductor unit,
Classification of FACTs devices are based on power
susceptance limit has been varied from minimum or
electronic converters which are known for its efficiency to
maximum.
control several variables of power structure in electric
transmission circuits for stable power flow as well as
dynamic stability control. In FACTs devices static VAR
Compensator (SVC) is employed because of its lower cost
in contrast with STATCOM.

978-1-7281-3261-7/19/$31.00 ©2019 IEEE 129

Authorized licensed use limited to: University of Exeter. Downloaded on June 13,2020 at 14:32:18 UTC from IEEE Xplore. Restrictions apply.
 Where, NB represents bus number. PGm and QGm are true
and reactive powers of the generator at bus m and n; PDm
and QDm are the load true and reactive power at bus m and n
respectively.
 Inequality Constraints:
Generator reactive power for lower and upper limits in
the inequality constraints can be written as:
Fig 1. Schematic diagram of SVC connected to Qgmmin ≤ Qgm ≤ Qgmmax ; ---(6)
electric utility.
Vmmin ≤ Vm ≤ Vmmax ; ---(7)
m = 1,2,3…………NG

IV. PARTICLE SWARM OPTIMIZATION

A. Summary of PSO
Fig 2. Illustrates the current injected model by SVC. The PSO is a population-based search algorithm in which
individuals called particles change their positions.
Injected reactive power into the bus-bar is given by According to its own experience, each particle moves to the
the equation (1) below: new position using velocity, called as Pbest. Global best
2
value of the population obtained so far by any particle is the
QSVC = BSVC *V --- (1) Gbest. In each step, each particles velocity change towards its
Pbest and Gbest. Each particles velocity and current position
are modified between its distance from current position and
Pbest, and the current position and Gbest. The velocity and
position are updated after finding the best values of the
particle. Each particles velocity can be modified.

Fig 3. Ideal static reactive compensator Each particle in the problem space is affiliated with the
better solution (fitness) as it has gained so far by keeping
III. PROBLEM FORMULATION track of its coordinates and also preserving the fitness value.
This value is known as Pbest. The global best is the best
A. Reduction of true power loss value obtained and is called Gbest, when a particle and its
F1 (u, v) of the objective function is specified as the true adjacent particles completes all the population. The particles
power loss (PL): velocity and positions are updated with the following
equations after finding the two best values.
--- (2)
Where, NL and gk depict total transmission lines and line
---(8)
conductance; the end buses magnitude and phase angles of
the voltages are Vm , δm and Vn , δn respectively.
---(9)
B. Reduction of value of SVC
F2 (SVC value) of the objective function is for reduction of h=1,2,……..j,
SVC value. i =1,2,……..k.
F2 (SVC value) = Rating of SVC in p.u. --- (3)
Where,
C. Objective function = best particle of agent i.
The global optimized problem can be acquired by uniting
the above two objectives mentioned which is: = global best.
fF(u,v) = F1(u,v) + F2 ( SVC size )
j = particle number in a group.
Now, by reducing the objective function fF (u, v) the k = number of member particles.
optimization can be carried out as subject to different r = indicator of iterations (generations).
equality and inequality constraints. u = weight factor of inertia.
 Equality Constraints: a1 and a2 = two acceleration constants.
The equations of conventional power flow rate equality rand0 and rand1 = two random uniform values in
constraints are as follows: the range [0,1].

--- (4) ith particle velocity at iteration r.

For m=1…………………..NB

--- (5)
For m=1…………………..NB
130

Authorized licensed use limited to: University of Exeter. Downloaded on June 13,2020 at 14:32:18 UTC from IEEE Xplore. Restrictions apply.
 ith particle current position at iteration r. V. RESULTS
The PSO is evaluated to find ideal position and estimate the
value of the SVC in IEEE 30 bus system [6]. For
B. Implementation of PSO Parameters
contingency as well as base load conditions the test is done.
1) Number of Particles: It consists of 6 generators with a generation capacity
Using a power flow solution, the fitness value of each of 192.06 MW and 105.08 MVAR, 41 lines and real
particle has to be computed at each iteration through a and reactive power load of 189.2 MW and 107.20
balance between the number of particles and iterations of MVAR and SVC with a value of 100 MVAR.
the swarm and, the computational effort dramatically
increases if the particle size is too large. Therefore swarm The ideal settings of the PSO control variables are given
population size of 5 and 25 particles are chosen. below:
2) Inertia Weight: Number of particles: 20.
To advance the fast convergence of the results, as the aim is Individual acceleration constant: 2.
to decrease the weight of inertia from an initial value of 0.9 Maximum velocity: 3.
to 0.1, as inertia weight decreases linearly. The maximum Minimum velocity: 1.
number of iterations is changed in even steps as shown in Inertia weight: 0.1.
the equation.
Voltage profile and power loss are examined for
----(10) different cases for increase in load and loss of
generation at bus 13 as shown in Table I and Table
Where, II. By examining these figures below, it can be
ur = rth iteration inertia weight . concluded that the position of SVC enhances the
iter = iteration count. voltage stability and decreases the power loss.
maxiter = Maximum number of iterations.

Fig 5. IEEE 30 bus test system.

Case1: Base case without Contingency

Fig 4. Flowchart of PSO Implementation

Fig 6. Voltage profile for base case.

131

Authorized licensed use limited to: University of Exeter. Downloaded on June 13,2020 at 14:32:18 UTC from IEEE Xplore. Restrictions apply.
Case2: 20% increase in loading with and without SVC Case4: 100% increase in loading and removal of generator
with and without SVC

Fig 9. Voltage profile for 100% enhanced load & removal of

generation at bus 13 (with & without SVC).

TABLE I. LOSSES WITH OPTIMAL LOCATION AND SIZE OF SVC:

Fig 7. Voltage profile for 20% enhanced load (with & without SVC)
Conve GA[14 PSO
ntional ]
Case3: 50% increase in loading and removal of generator metho
with and without SVC. d [13]
SVC Losses Losses Losse
Locatio (MW) (MW) s(M
n and W)
Size
Case With - 17.56 17.56 17.56
1 out
conti
ngenc
y
Case With - - 27.298 27.29
2 out 8
SVC
With Bus4 & - 26.607 26.60
SVC 63.129 6
(MVA
R)
Case With - - 47.824 47.82
3 out 4
SVC
With Bus4 & - 45.562 45.55
Fig 8. Voltage profile for 50% enhanced load & removal of generation
at bus 13 (with & without SVC). SVC 63.129 8
(MVA
R)
Case With - - 99.584 99.58
4 out 4
SVC
With Bus4 & - 93.762 93.53
SVC 63.129 3
(MVA
R)

132

Authorized licensed use limited to: University of Exeter. Downloaded on June 13,2020 at 14:32:18 UTC from IEEE Xplore. Restrictions apply.
 TABLE II. VOLTAGE PROFILE WITH & WITHOUT SVC FOR Further it is possible to improve by using D-STATCOM,
DIFFERENT CASES:
TCSC, UPFC and also multiple FACTs devices. For
practical power system it can also be employed.
REFERENCES
[1] Saravanan M.,Slochanal S.M. R. , Venkatesh P., Prince Stephen
Abraham J., "Application of artic1e Swarm optimization
technique for optimal location of FACTS devices considering
cost of installation and system loadability", Electric Power
System Research, Vol 77, pp. 276-283, 2007.
[2] Singh J.G., Singh S. N., Srivastava S. C, "An Approach for
Optimal Placement of Static VAR Compensators Based on
Reactive Power Spot Price", IEEE Trans. Power Syst, Vol
22,No.4, pp. 2021- 2029,2007.
[3] Thukaram D., Lomi A., Selection of static VAR compensator
location and size for system voltage stability
improvement,Electr Power Syst Res, 54 (2000): 139–150.
[4] Mithulananthan N., Cañizares C. A., Reeve J., Comparison of
PSS, SVC and STATCOM controllers for damping power
system oscillation, IEEE Trans Power Syst, 18 (2003),786–792.
[5] D.E. Goldberg, Genetic algorithms in search, optimization, and
machine learning. New York: Addison-Wesley, 1989.
[6] Hadi Saadat, Power System Analysis, Tata McGRAW- HILL \
Edition.
[7] Neelanbar Singh Chouhan, N.V. Srikanth, B. Vijay Kumar,
Optimal placement of SVC to minimize loss and improve the
voltage profile under power system contingency using GA,
Proceedings of 5th SARC-IRF International Conference, 25th
May-2014.
[8] J.W. Nims a, A.A. El-Keib b, R.E. Smith c, Contingency
ranking for voltage stability using a Genetic algorithm, Electric
Power Systems Research 43 (1997) 69-76.
[9] . Zhu, G. Xu, Approach to automatic contingency selection by
reactive type performance index, IEE Proc. C 138 (1) (1991)
65568.
[10] G. Jasmon, L. Lee, New contingency ranking technique
incorporating a voltage stabilitycriterion, IEE Proc. C 140 (4)
(1993) 87-90.
[11] A. Mohamed. G. Jasmon, Voltage contingency selection
technique for security assessment. IEE Proc. C 136 (1) (1989)
24428.
[12] S. Singh, P. Kalra, S. Srivastava, Novel non-iterative load flow
for voltage contingency analysis, Electr. Power Energy Syst. 16
(1) (1994) 11-16.
[13] Vishnu Suresh, “Comparison of Solvers Performance for Load
Flow Analysis”, TRANSACTIONS ON ENVIRONMENT
AND ELECTRICAL ENGINEERING, Vol 3, No 1 (2019).
[14] Mohammed Shuaib Iqbal, Sujatha B.C, “Multiple contingency
analysis for optimal placement and estimate the value of SVC
for power loss reduction employing Genetic Algorithm”,
International Journal of Electrical Engineering & Technology,
Vol 10, pp.60-69,2019.

VI. CONCLUSION
In this paper, the PSO is evaluated to find ideal position and
value of SVC to reduce loss and enhance the voltage profile.
For different cases of increased load demand and generation
loss at bus 13, SVC is utilized to analyze voltage profile and
loss of system. The suggested method is executed on IEEE
30 bus system to calculate ideal position and value of the
SVC.
For higher loading (100% increase in loading) and
removal of generator it is observed that around 6.051MW
power loss is obtained. It is observed that voltage profile at
bus number 26 is improved from 0.790 p.u to 0.817 p.u and
at bus number 30 is improved from 0.781 p.u to 0.806 p.u.
PSO is executed successfully and SVC value at ideal
position is acquired for different cases. PSO proposes value
of SVC for (100% increase in loading and loss of generation
at bus 13) 98.855MVAR. The performance of PSO has been
proved better in comparison with results obtained from
conventional method [13] and GA[14] as shown in Table I.

133

Authorized licensed use limited to: University of Exeter. Downloaded on June 13,2020 at 14:32:18 UTC from IEEE Xplore. Restrictions apply.