(JETAE) Journal of Emerging Trends in Applied Engineering (ISSN 2518-4059),

Vol. 1, No. 2, 2016

Analysis and Optimization of IEEE 33 Bus Radial

Distributed System Using Optimization Algorithm
Arif Wazir, Naeem Arbab
University of Engineering and Technology, Peshawar, Pakistan

Loss shows overall loss in power

Abstract—This paper mainly focusses on the impact of dis- M1 No’s of load levels
tributed generation and best feeder reconfiguration of distribu- N No’s of feeders
tion system, in order to improve the quality of power in the Ix current flow in branch x
distribution system. Primarily the goal of this paper is to mitigate Rx Resistance of branch x
as much as possible the losses in power system and improve
the voltage profile. The optimization of the system constrained
by feeder capability limit, radial configuration format, no load
point interruption and load-point voltage limits. Using Hybrid network use heuristic technique which are based on analytical
Genetic Algorithm Particle Swarm Optimization (HGAPSO) to
find the best configuration. This hybrid optimization or search
approach. Reconfiguration of the distribution will be done in
algorithm has more efficiency and accuracy. Proposed method- order to reduce the losses made use solely of heuristics to
ology comprises of demonstration of 33-bus radial distribution figure out the configuration of distributed system [6]. There
with and without distributed generation. Study finding consist of in a technique of reconfiguring the distributed network for
the best possible configuration of switches given by optimization the reduction of resistive line losses [7]–[9] in which Genetic
algorithm in order to minimize the losses but at the same time
respecting all the constrain mention above.
Algorithm (GA) were used Another optimization algorithm
Index Terms—Passive optical network; Protection; Deployment which is a combination of GA and particle swarm optimization
cost; Reliability block diagram (PSO) which is more stable effective were proposed [10]. This
paper emphasizes the advantage of network reconfiguration to
I. I NTRODUCTION the distribution system in the presence of DG units for loss
reduction and bus voltage improvement. The application of a
Configuration of distributed system are usually radial for the genetic algorithm is applied to determine the optimal on/off
significant management of the protective devices [1]. Gener- patterns of the switches to minimize the system loss subject
ally, the distribution system comprises two type of switches to system constraints. The effectiveness of the methodology is
for the configuration management and protection, one of them demonstrated by a practical sized distribution system consist-
which are normally closed known as sectionalizing switches ing of 33 buses. Major emphasis of this paper is on the benefits
and second normally opened know as tie switches [2]. of configuration of distribution networks with and without
Keeping the radial configuration intact, the configuration of the presence of Distributed generation (DG) units for the
distributed system can be varied by changing the statuses of minimization of losses and improvement of bus voltage. Using
sectionalizing and tie switches. That sort of configuration is HGAPSO algorithm to find out the optimal off/on patterns
known as feeder reconfiguration. Real power loss reduction, of switches to lessen the losses of system at the same time
improvement of bus voltage, load balancing of system, system maintaining the system constrains. To check the effectiveness
security and reliability and improvement of power quality are of the proposed methodology, methodology will be attested on
the benefits of feeder reconfiguration. [3], [4]. In the recent specific sized distributed system having 33 busses.
past, distributed generation trend substantially increase. Dis-
tributed generation may be connected to main grid or can be II. P ROBLEM F ORMULATION
used as stand-alone electric generation unit situated at or near
Fundamental objective of reconfiguration of feeder is the
to end user. Development in recent past DG includes solar,
reduction of all power losses as follow:
fuel cells, wind, hydrogen, and biomass. In the light of recent
past we can predict that penetration of distributed generation Ml X
N
will increase drastically in the coming time, which required
X 2
Minimize Lo = |Ix | Rx (1)
an extensive change in control, planning and operations of y=1 x=1
the network in order to do it in lucrative manner [5]. This
paradigm shift is an alarming situation for system operator where
in order to get most from the existing including dispatching The constrains for this optimization problem are formulated
of generators, transformers taps controlling, management of as: Equation for the flow of power are given below
reactive power, voltage regulators and reconfiguration of the Nb
system in a collective manner. Almost all the existing algo-
X
Px = |Zxy Vx Vy | cos (θxy + δx − δy ) (2)
rithm found in the literature on the reconfiguration of the y=1

17
 (JETAE) Journal of Emerging Trends in Applied Engineering (ISSN 2518-4059),
Vol. 1, No. 2, 2016

Qx, Px Reactive and acive power at bus x

Nb No’s of busses
One advantage of PSO over GA is its algorithmic simplicity.
Zxy element (x,y) in bus x and y Another clear difference between PSO and GA is the ability to
Vx, Vy Voltage of bus x and y control convergence. Crossover and mutation rates can subtly
θxy Angle of Zxy affect the convergence of GA, but these cannot be analogous
δx, δy Bus x and bus y voltage angle
to the level of control achieved through manipulating of the
inertia weight. In fact, the decrease of inertia weight dramat-
ically increases the swarms convergence. The main problem
with PSO is that it prematurely converges (Van den Bergh and
Engel Brecht 2004) to stable point, which is not necessarily
maximum. To prevent the occurrence, position update of the
global best particles is changed. The position update is done
through some hybrid mechanism of GA. The idea behind
GA is due to its genetic operators crossover and mutation.
By applying crossover operation, information can be swapped
between two particles to have the ability to fly to the new
search area. The purpose of applying mutation to PSO is to
increase the diversity of the population and the ability to have
the PSO to avoid the local maxima. In this model the initial
population of PSO is assigned by solution of GA. The total
numbers of iterations are equally shared by GA and PSO. First
half of the iterations are run by GA and the solutions are given
as initial population of PSO. Remaining iterations are run by
PSO.

IV. M ETHODOLOGY
Appling HGAPSO algorithm in order to find the optimal
solution of the network configuration to minimize the loss
problem, following steps being followed in order to imple-
Fig. 1: Flow chart of HGAPSO ment:
1) Loop counter and counter of global best improvement
are set to zero.
Nb
X 2) Find the possible matches from the bidding.
Qx = − |Zxy Vx Vy | cos (θxy + δx − δy ) (3) 3) Initial population is generated.
y=1 4) Calculate the fitness (utility) for each chromosome and
where particle variables (velocity and position are set to zero
It must be observed that the limit of voltage busses must be while local and global best positions are set.
within the rage 5) Preserve global best, to be compared to the new one, in
V min ≤ V i ≤ V max order to decide whether it is improved or not.
Moreover, capability of feeders has the limits |Ik | ≤ 6) Selection operation is conducted to find best chromo-
Ikmax k {1, 2, 3, ..., l} some in population and copy the best chromosome in
Format of the configuration must ne radial No load-point the new population Pn.
interruption 7) Population P is mutated and crossed-over based on the
where V min , V min and Ikmax represent the minimum probability of crossover and mutation and then put into
voltage, maximum voltage and maximum capability of current Pn.
that k branch can have 8) Calculate Pn fitness and particles values
9) Update Count according to the value of Pg . If the global
III. H YBRID G ENETIC A LGORITHM PARTICLE S WARM best did not improve after the selection, between original
O PTIMIZATION HGAPSO P and Pn , then the counter will be increased. This step
The hybrid of GA and PSO named HGAPSO shown in takes advantage of PSO in order to reduce the time.
Fig. 1S, originally presented by Juang [11], is used. This 10) Increase the loop counter.
algorithm consists of four major operators: enhancement, 11) Check the termination condition. If the maximum gen-
selection, crossover, and mutation. eration number is reached or when the global best is not
To overcome the limitations of PSO, hybrid algorithms with improved for a specific number of times (i.e., Count Pg
GA are proposed. The basis behind this is that such a hybrid variable equals quarter maximum generation number),
approach is expected to have merits of PSO with those of GA. then stop the algorithm and put P into W, which have the
18
 (JETAE) Journal of Emerging Trends in Applied Engineering (ISSN 2518-4059),
Vol. 1, No. 2, 2016

Fig. 3: Voltage profile and current flows for different cases

The numerical results for the four cases are summarized in

Table I
TABLE I: Numerical results for different cases

Case 1 Case 2 Case 3 Case 4

Fig. 2: Single line diagram of 33-bus radial distribution Opened section- – 10,7,1,4,32,3,7 – 9,7,14,32,37
alizing switches
Closed tie – 33,34,35 – 33,3,4,35
switches
fittest matches conducted from the algorithm. Otherwise, Vmin p.u. 0.9108 0.9423 0.9584 0.9734
go to step (4). Total power loss 209 140 84 57
KW
A. Test Case
The test system for the case study is radial distribution It is noticed that there is a considerable decrease in the
system with 33 buses, 5 tie-lines (looping branches), as shown power loss values when the distributed generator is placed in
in Fig. 2. The load data are given in Table A1 and branch the distribution system. It is confirmed from case 3 that the
data in Table A2 in Appendix [10]. The initial statuses of distributed generation helps to reduce the system loss from
all the sectionalizing switches (switches No. 1-32) are closed 208.kW to 84kW, or 59% of those configurations without
while all the tie switches (switch No.33-37) are open. The distributed generation. When comparing case 2 with case
total loads for this test system are 3,715 kW and 2,300 kVAr. 3, the former sees a higher power loss. The minimum loss
The current carrying capacity of branch No.1-9 is 400 A, and is observed in case 4, where there are changes in branch
the other remaining branches including the tie lines are 200 currents after the reconfiguration. Distributed generation, from
A. The minimum and maximum voltages are set at 0.95 and the perspective of losses, impacted positively in the analyzed
1.05 p.u., respectively. The maximum iteration for the genetic distribution network, achieving values of 73%. This confirms
algorithm is 100. Four cases are examined as follows: that DG units can normally, although not necessarily, help to
1) Case 1: The system is without distributed generation reduce the current flow in the feeders; and hence contributes
and feeder reconfiguration. to the power loss reduction, mainly because they are usually
2) Case 2: The same as case 1 except that the feeders can placed near the load being supplied. In case 4, where the
be reconfigured by the available sectionalizing switched and feeders are reconfigured and the voltage constraint is imposed
the tie switches. in the optimization process, no bus voltage is found violated.
3) Case 3: The same as case 1 except that there are 3 small The bus voltage profile and current flows in branch for case
power producers who can provide only firm active power to 1 and case 4 are depicted in Fig. 3.
the system by their DG units. The producers are located at
buses 16, 22, and 30 with capacities of 250, 250 and 500 kW, V. C ONCLUSION
respectively. has been presented in this paper to find the most appro-
4) Case 4: The same as case 3 but with feeder reconfigu- priate topology of the distribution system in the presence of
ration. distributed generation. A 33-bus distribution system with three
19
 (JETAE) Journal of Emerging Trends in Applied Engineering (ISSN 2518-4059),
Vol. 1, No. 2, 2016

distributed generation is used to demonstrate the effectiveness A PPENDIX

of the proposed technique. Although the distributed generation
TABLE A1: 33-bus distribution system load data
contributes to loss reduction, some bus voltages violate the
minimum voltage constraint. Such a problem can be remedied
by network reconfiguration. The results show that the optimal
on/off patterns of the switches can be identified which give the
minimum power loss while keeping bus voltage magnitudes
within the acceptable limits.
R EFERENCES
[1] M. Kashem, V. Ganapathy, and G. Jasmon, “Network reconfiguration for
load balancing in distribution networks,” IEE Proceedings-Generation,
Transmission and Distribution, vol. 146, no. 6, pp. 563–567, 1999.
[2] C.-T. Su and C.-S. Lee, “Network reconfiguration of distribution sys-
tems using improved mixed-integer hybrid differential evolution,” IEEE
Transactions on Power Delivery, vol. 18, no. 3, pp. 1022–1027, 2003.
[3] M. E. Baran and F. F. Wu, “Network reconfiguration in distribution
systems for loss reduction and load balancing,” IEEE Transactions on
Power Delivery, vol. 4, no. 2, pp. 1401–1407, 1989.
[4] J. Savier and D. Das, “Impact of network reconfiguration on loss
allocation of radial distribution systems,” IEEE Transactions on Power
Delivery, vol. 22, no. 4, pp. 2473–2480, 2007.
[5] H. A. Gil and G. Joos, “Models for quantifying the economic benefits of
distributed generation,” IEEE Transactions on Power Systems, vol. 23,
no. 2, pp. 327–335, 2008.
[6] S. Civanlar, J. Grainger, H. Yin, and S. Lee, “Distribution feeder
reconfiguration for loss reduction,” IEEE Trans. Power Del.;(United
States), vol. 3, no. 3, 1988.
[7] D. Shirmohammadi and H. W. Hong, “Reconfiguration of electric distri-
bution networks for resistive line losses reduction,” IEEE Transactions
on Power Delivery, vol. 4, no. 2, pp. 1492–1498, 1989.
[8] K. Nara, A. Shiose, M. Kitagawa, and T. Ishihara, “Implementa-
tion of genetic algorithm for distribution systems loss minimum re-
configuration,” IEEE Transactions on Power Systems, vol. 7, no. 3, pp.
1044–1051, 1992.
[9] Y. Song, G. Wang, A. Johns, and P. Wang, “Distribution network
reconfiguration for loss reduction using fuzzy controlled evolutionary
programming,” IEE Proceedings-Generation, Transmission and Distri-
bution, vol. 144, no. 4, pp. 345–350, 1997.
[10] X. Wu, Y. Wang, and T. Zhang, “An improved gapso hybrid pro-
gramming algorithm,” in 2009 International Conference on Information
Engineering and Computer Science. IEEE, 2009, pp. 1–4.
[11] C.-F. Juang, “A hybrid of genetic algorithm and particle swarm opti-
mization for recurrent network design,” IEEE Transactions on Systems,
Man, and Cybernetics, Part B (Cybernetics), vol. 34, no. 2, pp. 997–
1006, 2004.

20
 (JETAE) Journal of Emerging Trends in Applied Engineering (ISSN 2518-4059),
Vol. 1, No. 2, 2016

TABLE A2: 33-bus distribution system data

21