Paper 13

International Journal of Advances in Science and Technology,
Vol. 3, No.6, 2011

Optimal Distributed Generator Placement for Loss Minimization of Unbalanced
Radial Distribution System Using Particle Swarm Optimization

Puthireddy Umapathi Reddy
1
, Sirigiri Sivanagaraju
2
, Prabandhamkam Sangameswararaju
3

1
Department of Electrical and Electronics Engineering, Sree Vidyanikethan Engineering College,
Tirupati, India.
moni_uma@yahoo.co.in

2
Department of Electrical and Electronics Engineering, JNT University College of Engineering Kakinada,
Kakinada, India.
sirigiri70@yahoo.co.in

3
Department of Electrical and Electronics Engineering, SV University College of Engineering, Tirupati,India.
raju_ps_2000@yahoo.com

Abstract

The Distributed Generators (DGs) have created a challenge and an opportunity for developing various novel
technologies in power generation. To minimize power losses, it is important to determine the location and size of local
generators to be placed in unbalanced power distribution systems. On account of some inherent features of
unbalanced distribution systems, such as radial structure, large number of nodes, a wide range of X/R ratios, the
conventional techniques developed for the transmission systems generally fail on the determination of optimum size
and location of distributed generators (DGs). The objective function formulated is to minimize the system power losses
while satisfying the voltage and power limits by placing the optimal size of DGs and location. Optimal location of DG
is determined by using Voltage Index Analysis (VIA) with Particle Swarm optimization (PSO) is used for finding the
optimal size of the DGs. This paper presents a simple method for investigating the problem of contemporaneously for
finding the optimal size and location for installation of DG in Unbalanced Radial Distribution Systems (URDS).
December Issue Page 126 of 140 ISSN 2229 5216
Vol. 3, No.6, 2011

Keywords: Unbalanced Radial Distribution Systems, Distributed Generators, Loss Minimization, Particle Swarm
Optimization

1. INTRODUCTION

Distributed or Dispersed Generator (DGs) is defined as a generating resource other than central generating station that is
placed close to load being served usually at a customer site. There is a wide range of terminologies used for distributed
generation, such as embedded generation, dispersed generation, or decentralized generation. Tsai-Hsiang Chen
proposed a method [1] DG means a small-scale power station different from a traditional or large central power plant. At
present, there are several technologies [2] ranging from traditional to non-traditional used in DG application. The former
is non-renewable technologies such as internal combustion engines, combined cycles, combustion turbines, and micro
turbines [3]. The latter technologies include fuel cells, storage devices, and a number of renewable energy-based
technologies [4,5] such as photovoltaic, biomass, wind, geothermal, ocean, etc. DG is also called as backup generator.
With the emergence of an era of deregulated power systems, several generators are installed on the consumer side [6]. It
may be connected to the supply side or demand side of meter. It can be either renewable sources based micro hydro, wind
turbines, photovoltaic, or fossil fuel based fuel cells, reciprocating engines, micro turbines etc. DG size range is from few
kilowatts to 100 megawatts [7, 8]. The planning of the electric system with the presence of DG requires the definition of
several factors [9,10] such as the best technology to be used, the number and the capacity of the units, the best location,
the type of network connection, etc. The share of DGs in power system has been continuously increasing worldwide [11].
The problem of DG allocation and sizing [12,13] has great importance in power system. The installation of DG units at
non-optimal places results increase in system losses which in turn results increase in costs [14] and therefore, has an
undesired effect. Energy policies and societies in the worldwide are encouraging the installation of DGs in both
transmission and distribution networks along with large scale power generating plants [15, 16].

The general belief is that the future of the power generation [17, 18] will be of DGs. But the fact is that the distribution
systems are not planned to support the installation of active power generating units in them and hence DGs come with
opportunities as well as challenges [19, 20]. T.N. Shukla, S.P. Singh and K.B. Naik [22] proposed the optimal location to
place a DG with unity power factor in a radial or loop system while minimizing the losses. This technique is basically
concerned with finding the optimal location but not the optimal size. Energy cost of renewable-based distributed
generation [23], when compared to the conventional generating plants is generally high because the social and
environmental benefits could not be included in the cost account.Cheng and Shirmohammadi [4] proposed PV buses for
three phase unbalanced system. S. Kumar Injeti, et.al [24] developed optimal planning of distributed generation for
improved voltage stability and loss reduction as PQ or PV bus. PSO based distributed generator placement technique
reduce the overall power losses in distribution system.
Vol. 3, No.6, 2011

2. DG MODEL IN AN UNBALANCED RADIAL DISTRIBUTION SYSTEM
While most of the research has been done on DG placement in radial balanced distributions systems, very little research
is related to DG placement in unbalanced radial distribution systems. The DG can be considered as PV or PQ models in
radial distribution system. The PV model delivers power at a specific terminal voltage, while PQ model DG delivers
power irrespective of the bus voltage. Such source will be modeled as a negative load delivering a real and reactive
power to the distribution system. While determining the size of DG, which can be added to distribution system is
presented without increase in loss as well as operational constraints violation.
3. MATHEMATICAL FORMULATION
In this paper, voltage sensitive buses are first identified by penetrating DG with 20% of the total feeder loading at each
bus, one bus at a time except source bus and the bus with least voltage index for placing DG has been picked as the best
location for the DG placement. PSO algorithm is used to find the optimal size of the DG. The objective function
formulated is to minimize the system power loss while satisfying the voltage and power limits by placing the optimal size
of the DG is stated as:
=
' =
nb
1 j
f Minimize
abc
j loss
P (1)
Where,
abc
j loss
P' is the active power loss in the j
th
branch after DG placement.
nb is the number of branches in the system.
The equality constraints are the power flow equations and the inequality constraints are:
- Voltage operational tolerance limits at all buses
max
q
V
abc
q
V
min
q
V s s
(2)

- Limit on real power loss in the distributed generator
Vol. 3, No.6, 2011

s
nb
j
abc
j
loss
P
ndg
i
abc
i
loss
PG
(3)
Where,
abc
i
loss
PG is the active power loss with DG
abc
j loss
P is the active power loss in the j
th
line without DG
ndg is the number of distributed generators
- Limit on total power generated by DG, subject to a penetration level of 20% (i.e., it must not exceed 20% of
feeder load).
=
s
=
n
1 q
abc
q
PL 0.2
ndg
1 i
abc
i
PG
,
=
s
=
n
1 q
abc
q
QL 0.2
ndg
1 i
abc
i
QG
(4)
Where,
abc
q
PL
is the real power phase loads at bus i
abc
q
QL
is the reactive power phase loads at bus i
- Branch flow limits (e.g., they must remain below thermal limits)
max
S
abc
j
S s
(5)

3.1. Algorithm for Minimum Voltage Index Bus for Location of DG

The algorithm for finding the minimum voltage index bus for the best suitable location for DG placement is:
Step 1: Read the given data for unbalanced radial distribution system.
Step 2: Calculate the voltage index at each bus by penetrating 20% DG capacity of feeder loading at each bus, one bus
at a time except the source bus by using eqn. (6).
Vol. 3, No.6, 2011

Voltage index (
index
V ) is defined as

( )
n
V
V
i
index

=
2
1
(6)
Where

i
V is sum of phase voltage at i
th
bus
n is the number of buses
Step 3: Select the minimum voltage index bus for the best location of DG placement.
Step 4: Stop.
3.2 Illustration
The location of bus at which distributed generator has to be placed is varied from 2 to 25 (except source bus) for given 25
bus URDS whose single line diagram is shown in Fig. 1. The line and the load data of this system are given in [23]. The
voltage index at each bus is calculated by using eqn. 6.
The voltage index value by penetrating the 20% DG capacity of feeder loading at each bus, one bus at a time except
source bus for different locations for identifying the best location of DG. From Fig.2, it can observe that 13
th
bus shows
the minimum voltage index. Hence, it is the best suitable location for the DG placement for 25 bus URDS.

Figure 1. Line diagram of 25 node URDS. Figure 2. Voltage index for 25 bus URDS
Vol. 3, No.6, 2011

4. IMPLEMENTATION OF PSO

4.1. Initialization of PSO Parameters

Initialize the PSO parameters like Swarm size(P), number of generations(NG), Initial weight of the weighting function
(W
max
), Final weight of the weighting function (W
min
), weighting factors(C
1
,C
2
), input system data, control parameters
such as lower and upper bounds of bus voltage and maximum number of capacitor banks to be installed at a
compensation node. Randomly generate an initial swarm (array) of particles with random positions and velocities.

4.2. Evaluation of Fitness Function

The fitness function should be capable of reflecting the objective and directing the search towards optimal solution. Since
the PSO proceeds in the direction of evolving best-fit particles and the fitness value is the only information available to
the PSO, the performance of the algorithm is highly sensitive to the fitness values. For each particle or swarm, the
calculated DGs are placed at the sensitive nodes and the load flow method is run and the losses are calculated by using
PSO.

4.3 Optimal Solution

The optimal solution to the target problem is the best position and the corresponding fitness value. The best position
information includes the optimal locations and numbers of DGs at each load level, and the corresponding fitness value
represents the minimizing the total losses in the system. Accordingly, the optimal types and sizes of DGs to be placed at
each compensation node can be determined: if number of DGs during all load levels is identical, then the DG at this node
can be regarded as the fixed type; if number of DGs at different load levels is not all the same. The size of DG placed at
each compensation node is the product of the maximum number of DGs and the standard size per DG.
This modification can be represented by the concept of velocity (modified value for the current positions). Velocity of
each particle can be modified by the following equation.
] [ ] [
2 2 1 1
1 k
i
k
i i
k
i
k
i
X Gbest rand C X Pbest rand C WV V + + =
+
(7)
where,

k
i
V : Velocity of particle i at iteration k,
Vol. 3, No.6, 2011

1 + k
i
V
: Modified velocity of particle i at iteration k+1,
W : Inertia weight,
C
1
,C
2
: Acceleration Constants,
rand
1
,rand
2
: Two random numbers
k
i
X

: Current position of particle i at iteration k,
Pbest
i
: Pbest of particle i,
Gbest : Gbest of the group.
In the equation (7),
The term
) X - (Pbest rand
k
i i 1
is called particle memory influence
The term
) X - (Gbest rand
k
i 2
is called swarm influence
The rand
1
, rand
2
are the two random numbers with uniform distribution with range of {0 to 1.0 }
The following inertia weight is usually utilized in equation (7):
max min
max
max
W -W
W=W - iter
iter
(8)
where,
max
W

: initial value of the Inertia weight,
min
W

: final value of the Inertia weight,
max
iter : maximum iteration number,
iter : current iteration number.
Accordingly, the optimal location and sizes of DGs to be placed at each node can be determined. The size of DG placed
at each node is such that voltage and power limit must be satisfied.

4.4 Algorithm for Optimal Location and Size of the DG

The detailed algorithm to determine optimal location and size of the DG is given below.
Step 1: Read system data such as Line data and Load data of the Distribution system.
Vol. 3, No.6, 2011

Step 2: Initialize the PSO parameters such as Swarm size(P), initial weight & final weight of the weighting function
(W
max
& W
min
), weighting factor(C
j
), number of generations (NG).
Step 3: Obtain the optimal location of DG by using Voltage Indices as input.
Step 4: Randomly create initial particles where each particle is a solution to the optimal location and size of the DG.
Step 5: Initialize the velocities for the particles.
Step 6: Run the load flow for each particle and Compute the value of voltages, total active and reactive power losses at
all the nodes.
Step 7: Set the local best values for each particle and global best value for the current iteration.
Step 8: Update the particle positions & velocity using eqns (7) & (8).
Step 9: If the particle position is out of boundary then it is brought back to its nearest value using eqns (7).
Step 10: Execute steps 6-9 in a loop for maximum number of generations (NG).
Step 11: Stop the execution and display the global best values as the final result of optimal size of the DG.

5. RESULTS AND ANALYSIS
The performance is evaluated for test system of 25-node URDS. The PSO control Parameters for DG placement:
Number of Particles (N)=20, Number of Iterations (K
max
)=100, Initial value of the inertia weight (W
max
)=0.9, final value
of the inertia weight (W
min
)=0.4, acceleration constants (C
1
& C
2
)=2 to find size and location of DG.
5.1 Example-1
The proposed method for DG placement is tested for 25 bus URDS. The line and the load data of this system are given in
[23]. Single line diagram of 25 bus URDS after DG placement is shown in Fig.3. Total Active Power loss Vs Generation
number of 25 bus URDS is shown in Fig.4. From the graph it is observed that the active power loss is reduced from
excising method to proposed method with PSO Technique.
Vol. 3, No.6, 2011

Figure 3.Single line diagram of 25 bus URDS Figure 4. Total Active Power loss Vs
after DG placement Generation number of 25 bus URDS

The total active power loss Vs generation number of 25 bus URDS after DG placement is shown in fig.4.Which gives clear
difference between power loss reduction. Voltage values of 25 bus URDS for DG placement (DG Size at 0.85 p.f (kW)),
active and reactive power losses in each branch of 25 bus URDS after DG Placement is shown in fig.5, fig.6 and fig.7
respectively.

The summary of test results of 25 bus URDS for DG placement are given in tables 1. From table1, it is observed that the
minimum voltage in phases A, B and C are improved from 0.9642, 0.9645and 0.9685 p.u (with DG using GA) to 0.9651,
0.9653and 0.9692p.u (with DG using PSO) respectively and the active power loss in phases of A, B and C is reduced from
21.05, 21.46 and 16.45kW to 20.78, 21.09and 16.31kW respectively. Hence, there is an improvement in the minimum
voltage and reduction in power loss when compared with the before DG placement and after DG placement with existing
and proposed methods.

Vol. 3, No.6, 2011

Figure 5. Voltage values of 25 bus URDS for DG placement (DG Size at 0.85 p.f (kW))

Figure 6. Power Losses in each branch of 25 bus URDS for DG Placement
Vol. 3, No.6, 2011

Table 1. Summary of test results before and after DG placement of 25 node URDS

Description
Before DG Placement
After DG Placement by
using GA
After DG Placement by
using Proposed PSO
Phase A Phase B Phase C
Phase
A
Phase
B
Phase C
Phase
A
Phase
B
Phase
C
Distributed
generators placed
nodes and size of
DG (in kW)
10
- - -
175 175 175 215 215 215
11 185 185 185 215 215 215
12 55 55 55 45 45 45
13 55 55 55 30 30 30
Minimum Voltage 0.9284 0.9284 0.9366 0.9642 0.9645 0.9685 0.9651 0.9653 0.9692
Max.Volt. regulation (%) 7.16 7.16 6.34 3.58 3.55 3.15 3.49 3.47 3.08
Improvement of Max.
Voltage regulation (%)
- - - 50.00 50.42 50.32 51.26 51.54 51.42
Total Active Power Loss
(kW)
52.82 55.44 41.86 21.05 21.46 16.45 20.78 21.09 16.31
Total Active Power Loss
reduction (%)
- - - 60.14 61.29 60.71 60.67 61.96 61.05
Total Reactive Power
Loss (kVAr)
58.32 53.29 55.69 23.56 21.14 22.33 22.74 20.37 21.50
Loss reduction (%)
- - - 59.60 60.34 70.47 61.01 61.78 61.39
Total Demand (kW) 1126.12 1138.74 1125.16 627.92 639.16 632.67 591.00 602.16 596.13
Total Released Demand
(kW)
- - - 498.20 499.58 492.49 535.13 536.58 529.03
Demand (kVAr)
850.32 854.29 855.69 527.14 533.28 534.31 503.25 509.56 510.55
Total Released Reactive
Power Demand (kVAr)
- - - 323.18 321.01 321.38 347.07 344.73 345.14
Total Feeder Capacity
(kVA)
1411.09 1423.57 1413.57 819.85 832.42 828.10 776.23 788.82 784.88
Total Released Feeder
Capacity (kVA)
- - - 593.84 593.82 588.08 637.82 637.78 631.66
Execution time (in sec) - 16.84 15.53
Vol. 3, No.6, 2011

Figure 7. Power Losses in each branch of 25 bus URDS for DG Placement

6. CONCLUSIONS
This paper presents a methodology for solving the optimal allocation and sizing of DG problem in Unbalanced Radial
Distribution Systems through voltage index analysis to minimize power losses while satisfying the voltage and power
limits using PSO algorithm. The sizing of DG has been obtained with the view of objective function of reducing power
loss by using PSO. The proposed PSO based method was applied to 25 bus URDS. The obtained solution has succeeded
in reducing the total average active power loss and improved minimum voltage for a given test system of 25 bus URDS.
Vol. 3, No.6, 2011

7. REFERENCES

[1] Chen,T.H. Distribution System Power Flow Analysis-A Rigid Approach, IEEE Transactions on Power Delivery,
Vol. 6, No. 3, pp.1146 1152, July 1991.
[2] Ray, D. Zimmerman and Chiang ,H.D., Fast Decoupled Power Flow for Unbalanced Radial Distribution Systems
IEEE Transactions on Power Systems, Vol.10, No.4, pp. 2045-205, Nov. 1995.
[3] Chen T. H, Chen M. S, Hwang K. J, Kotas. P, Chebli E. A, Three phase Co- generator and transformer models for
distribution analysis IEEE Trans. Power Del., vol. 6, no. 4, pp- 1671-1681, Oct. 1991.
[4] Cheng,C.H. Shirmohammadi,D .A Three Phase Power Flow Method for Real time Distribution System Analysis,
IEEE Transactions on Power Systems, vol. 10, no. 2 pp 671- 679, May 1995.
[5] Vu,H. Pruvot, P. Launay, C. Harmand.Y. An improved voltage control on large-scale power systems, IEEE Trans.
Power Syst., vol. 11, no. 3, pp. 12951303, Aug. 1996.
[6] Chiradeja,P. Ramakumar,R. Benefits of distributed generation-a simple case study, in Proc. 32nd Annual
Frontiers of Power Conf., Stillwater, UK, pp. X 19,1999.
[7] I. Roytelman , Ganesan,V. Modeling of local controllers in distribution network application, IEEE Trans. Power
Del., vol. 15, no. 4,pp. 12321237, Oct. 2000.
[8] Teng JH. A direct approach for distribution system load flow solutions, IEEE Trans Power Deliver 18; 882-887,
2003.
[9] Kim, K. H. Lee, Y. J. Rhee, S. B. Lee, S. K. & You, S. K. Dispersed Generation Placement Using Fuzzy-GA in
Distribution System, Proceedings of IEEE Power Engineering Society Summer Meeting, Chicago, Vol. 3, pp.
1148-1153, 2002.
[10] Dugan,R.C. Mc dermont,T.E. Distributed generation, IEEE Industrial Application Magazine, pp. 1925, Mar.
/Apr. 2002.
[11] Choi, J.H Kim,J.C Moon,S. Integration operation of dispersed generations to automated distribution networks for
network reconfiguration, in Proc. IEEE Power Technology Engineering Conf., Bologna, Italy, vol. 4, pp. 2363
2367, 2003.
[12] Chiradeja, P. Ramakumar,R. Voltage Profile Improvement with Distributed Wind turbine Generation - A Case
Study, IEEE Power Engineering Society General Meeting, Vol. 4, pp 2331-2336, 2003.
[13] Hegazy,Y.G. Salama M.M.A. Chikhan. Y.: Adequacy Assessment of Distributed Generation Systems Using Monte
Carlo Simulation, IEEE Transactions on Power Systems,Vol. 18, No. 1, pp.48-52, February 2003.
[14] Khattam, W. El. Bhattacharya,K.. Hegazy,Y. Salama,M.M.A.: Optimal investment planning for distributed
generation in a competitive electricity market, IEEE Trans. Power systems, vol. 19 pp. 1674-1684, 2004.
[15] Koen J. P. Macken, Math H. J. Bollen, Ronnie J. M. : Mitigation of Voltage Dips Through Distributed Generation
Systems, IEEE Transactions on Industry Applications, Vol. 40, No. 6, pp. 1686- 1693, November/December 2004.
Vol. 3, No.6, 2011

[16] Aleksandar Pregelj, Miroslav Begovic , Ajeet Rohatgi.: Quantitative Techniques for Analysis of Large Data Sets
in Renewable Distributed Generation, IEEE Transactions on Power Systems, Vol. 19, No. 3, 1277- 1285, August
2004.
[17] Caisheng Wang, M. Hashem NehrirV, : Analytical Approaches for Optimal Placement of Distributed Generation
Sources in Power Systems, IEEE Transactions on Power Systems, Vol. 19, No. 4, pp. 2068 2076, November
2004.
[18] Pathomthat Chiradeja, Ramakumar.R.: An Approach to Quantify the Technical Benefits of Distributed Generation,
IEEE Transactions on Energy Conversion, Vol. 19, No. 4, pp. 764-773, December 2004.
[19] Acharya, N. Mahat, P. Mithulananthan. N,: An analytical approach for DG allocation in primary distribution
network, International Journal of Electrical Power and Energy Systems, Vol. 28, pp. 669-678, 2005.
[20] Benemar A. de Souza, M. C. de Albuquerque.: Optimal placement of distribution generators networks by
evolutionary programming, Proc. IEEE PES Transmission and Distribution Conference and Exposition Latin
America, Venezuela pp 01-06, 2006.
[21] Vctor H. Mndez Quezada, Juan Rivier Abbad, and Toms Gmez San Romn.: Assessment of Energy
Distribution Losses for Increasing Penetration of Distributed Generation, IEEE Transactions on Power Systems,
Vol. 21, No. 2, pp. 533- 540, May 2006.
[22] Shukla, T.N Singh, S.P. Naik K.B. : Allocation of Optimal Distributed Generation using GA for Minimum System
Losses, Proceedings of Fifteenth National Power Systems Conference (NPSC), IIT Bombay,pp 113-118,
December 2008.
[23] Subrahmanyam, J. B. V. Radhakrishna,C.: Distributed Generator Placement and Sizing in Unbalanced Radial
Distribution System, World Academy of Science, Engineering and Technology 52, pp 737-744, 2009.
[24] Kumar Injeti,S Navuri Kumar,P.:Optimal Planning of Distributed Generation for improved Voltage Stability and
Loss Reduction International Journal of Computer applications (0975 8887),Volume 15 No.1, February 2011.

Authors
P.UMAPATHI REDDY: He Received B.E from Andra University and M.Tech.,(Electrical Power
Systems) from Jawaharlal Nehru Technological University, Anantapur, India in 1998 and 2004
respectively, Now he is pursuing Ph.D. degree. Currently he is with Department of Electrical and
Electronics Engineering, Sree Vidyanikethan Engineering College, Tirupati, India. His research
interest includes Power distribution Systems and Power System operation and control. He is Life
Member of Indian Society for Technical Education.

Vol. 3, No.6, 2011

Dr.S.Sivanaga Raju: He received B.E from Andra University and M.Tech.degree in 2000 from
IIT, Kharagpur and did his Ph.D from Jawaharlal Nehru Technological University, Anantapur, India
in 2004. He is presently working as Associate professor in J.N.T.U.College of Engineering
Kakinada,(Autonomous) Kakinada, Andrapradesh, India. He received two national awards (Pandit
Madan Mohan Malaviya memorial Prize and best paper prize award from the Institute of Engineers
(India) for the year 2003-04. He is referee for IEEE journals. He has around 75 National and
International journals in his credit. His research interest includes Power distribution Automation
and Power System operation and control.

Dr.P.Sangameswara Raju: He is presently working as professor in S.V.U.College Engineering,
Tirupati. Obtained his diploma and B.Tech in electrical Engineering, M.Tech in power system
operation and control and Ph.d in S.V.University,Tirupati. His areas of interest are power system
operation, planning and application of fuzzy logic to power system, application of power system
like non-linear controllers.


Paper 13

Diunggah oleh

Informasi Dokumen

Deskripsi Asli:

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Paper 13

Diunggah oleh

Hak Cipta:

Format Tersedia

International Journal of Advances in Science and Technology,

Vol. 3, No.6, 2011

Anda mungkin juga menyukai