Professional Documents
Culture Documents
a r t i c l e i n f o a b s t r a c t
Article history: This paper proposes a fuzzy set optimization approach for capacitor allocation in radial distribution
Received 26 April 2009 system. In this approach, a membership function for voltage profile constraint has been used. Moreover,
Received in revised form 1 June 2013 another membership function incorporating feeder section active power losses and total power losses
Accepted 22 August 2013
constraints has been proposed. This membership function indirectly imposes thermal capability of the
feeder on the optimization process. The proposed approach has been applied to 9-bus and 34-bus radial
Keywords:
distribution systems. The results have been compared with those of two fuzzy approaches in literature.
Distribution systems
The comparison showed the effectiveness of the proposed approach for optimizing the sizes and locations
Capacitor allocation
Fuzzy sets theory
of the capacitor with running and total cost reduction.
© 2013 Elsevier B.V. All rights reserved.
0378-7796/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.epsr.2013.08.019
H.A. Ramadan et al. / Electric Power Systems Research 106 (2014) 232–240 233
P1 , Q1 P2 , Q2 Pi , Q i Pi 1, Qi 1 Pi 2 , Q i 2
Pn -1 , Q n -1 Pn , Q n
Fig. 1. A single line diagram of a radial distribution (i) feeder.
In addition, the method in [10] also accounts for varying loads of 2. Problem formulation
the distribution system considered. Hamada et al. [11] introduced a
new strategy for capacitor allocation handling the reduction in the The purpose of placing compensating capacitors along the dis-
section losses by adding a new constraint to the well-known con- tribution feeders is to lower the total power loss and bring the bus
straint (allowed voltage violation constraint). The new constraint voltages within specified limits while minimizing the total power
is the sectional ohmic losses in each branch of the feeder. Ref. [12] cost. The total power loss PTloss is given by
presents a fuzzy-based approach for capacitor placement for the 9-
bus feeder. Two membership functions for total real power losses
n−1
PTloss = Ploss(i,i+1) i = 1, 2, . . ., n (1)
and voltage sensitivity have been defined to reduce the effort of
i=1
finding the optimal locations. The whole problem has been pre-
sented as a fuzzy-set optimization problem to minimize the total where, i is the bus number and n is the total number of buses
real losses and capacitor cost with voltage limit constraints. They as shown in Fig. 1. Considering investment cost, there is a finite
used the intersection principle in fuzzy as the fuzzy decision to find number of standard capacitor sizes that are integer multiples of
the capacitor location, then a variational method has been used to Q0c . The cost per kVAr varies from one size to another. Generally,
find capacitor sizes to attain minimum cost without violating the large capacitor sizes are cheaper than smaller ones. The available
voltage constraints. capacitor size is usually limited to [13]:
In Ref. [13], exactly the same procedures using the same feeder c
have been implemented, but with two different membership func- Qmax = L × Q0c (2)
tions. In fact, their membership functions for real power losses and where Q0c is the smallest capacitor size in Table A1 and L is an
voltage are the fundamental part of the membership functions that
integer. Therefore
for each installation are L capaci-
location, there
have been used in [12]. However, they have relatively achieved tor sizes Q0c , 2Q0c , . . ., LQ0 c to choose from. Let K1c , K2c , . . ., KLc
better results by introducing a certain constant in the real losses be their corresponding equivalent annual cost per kVAr. The objec-
membership function depending on their experiences. In [14], the tive (Cost) function can be expressed as
authors used membership functions forms of [12], but replaced the
real losses by reactive losses and the intersection decision (using
J
min operator) by product decision. They used the product fuzzy Cost = Cp × PTloss + Kjc Qjc (3)
decision to determine the location of the capacitors. To find the j=1
capacitor sizes, they used their analytical method that has been
where Cp is the cost per power loss ($/kW) [13] and j = 1,2,...,j repre-
explained in [15]. This analytical method is based on differentiat-
sents the selected buses for compensation. The objective function
ing a well-defined net saving function of power and energy losses
Eq. (3) is to be minimized subjected to two constraints:
with respect to capacitor size, thus obtaining the optimum capac-
The first constraint is the bus voltage constraint.
itor size. Their method has been applied on the 34-bus feeder
[15]. Vmin ≤ Vi ≤ Vmax , i = 1, 2, . . ., n (4)
Ref. [16] presented fast power loss computation using super-
The second constraint is the sectional ohmic loss constraint (pro-
visory control and data acquisition system (SCADA), which a
posed in the author’s previously published paper).
fuzzy-based decision maker used to compute the suitable shunt
capacitor required to improve the power factor according to the (K+1) (K)
PSec,loss ≤ PSec.loss (5)
measured parameters. Most of these studies consider the loss
reduction for the capacitor allocation problem, but these studies where (K + 1) is the case after the capacitor placement and (K) is the
take in consideration the reduction of total losses not the reduction case before the capacitor placement.
in the individual sectional losses. The general capacitor placement
problem in distribution feeders consists of determining the optimal 3. Application of fuzzy set theory
location, type (fixed or switched), and size of capacitors; such that
power and energy losses are minimized while taking the cost of the In the conventional methods, the large number of combinations
capacitor into account. For simplifying the problem fixed capacitors in the solution space makes the solution searching process time
are only considered in this paper. consuming. The problem is formulated as a fuzzy reasoning opti-
This paper presents a fuzzy-based approach for capacitor alloca- mization model to minimize the cost and power loss subjected to
tion in radial distribution systems. Two membership functions are voltage limit and sectional losses constraints. Minimum operation
defined in this paper, one for the voltage sensitivity and another of fuzzy sets is employed to find the optimal locations and sizes of
proposed one for the real sectional ohmic loss constraint. This con- capacitors.
straint has been initially introduced by the authors in a previous
published paper, Ref. [11]. The problem is formulated as a fuzzy-set 3.1. The proposed method
optimization problem to minimize the real power loss and capaci-
tor cost with sectional loss and voltage limit constraints. Moreover, Two membership functions are defined in this method, one for
this paper presents a comparison between the proposed approach the voltage sensitivity and another for the active sectional ohmic
and previous works to show the validity of the proposed approach. loss. This method suggests a new membership function for the
234 H.A. Ramadan et al. / Electric Power Systems Research 106 (2014) 232–240
P P
1 .0 1 .0
0 .8
0 .6
0 .4
0 .2 0 .0
LP Ploss ( i , i 1)
0 .0
0 .0 0 .1 0 .3 0 .4 0 .5 0 .6 Ploss (i, i 1) / PTloss Fig. 6. Membership function P for real power loss for method (2) [13].
Fig. 4. Membership function P for real power loss for method (1) [12].
a low loss branch gives a high value. The membership function
defined for the power loss is given as:
V
Ploss (i, i + 1)
1.0 1− , i = 1, 2, . . ., n ; Ploss (i, i + 1) ≤ LP
P (i) = LP (11)
0.8 0, i = 1, 2, . . ., n ; Ploss (i, i + 1) > LP
1.15
Vmax
1.1
1.05
1
0.95
Vmin
Voltage in p.u.
0.9
0.85
0.8
0.75
0.7
0.65
0.6
0.55
0.5
1 2 3 4 5 6 7 8 9
Bus No
Without Qc Method1
proposed method Method 2
1.15
Vmax
1.1
Voltage in p.u
1.05
0.95
Vmin
0.9
0.85
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
Bus No
Table 1 Table 4
Capacitor values and locations for the 9-bus test system. Capacitor values and locations for the 34-bus test system.
Method no. The proposed method Method (1) [12] Method (2) [13] Method no. The proposed method Method (1) [12] Method (2) [13]
Bus no. Qc in MVAr Qc in MVAr Qc in MVAr
Bus no Qc in MVAr Qc in MVAr Qc in MVAr
1 0 0 0
1 0 0 0
2 0 0 0
2 0 0 0
3 0 0 0
3 3.15 0 1.05
4 0.6 2.1 0
4 1.5 2.1 1.05
5 0 1.05 0
5 0.9 2.5 1.95
6 1.35 0.3 0
6 0 0 0
7 0.15 0 0
7 1.5 0 0
8 0.15 0 0
8 0 0 0
9 0 0 0
9 0.45 0.9 0.9
10 0.15 0 0
Summation of MVAr 7.5 5.5 4.95
11 0 0 0
12 0 0 0
13 0 0 0
14 0 0 0
proposed method and for method (1) are close. The voltage pro- 15 0 0 0
file for method (2) at some buses is lower than that obtained by the 16 0 0 0
17 0 0.45 0
proposed method and method (1). All the methods including the 18 0 0.15 0
uncompensated case produced voltage profiles within the specified 19 0 0.15 0
maximum and minimum voltage limits. 20 0 0.15 0.15
Therefore, it can be concluded that the effect of the proposed 21 0.3 0.15 0.15
22 0 0.15 0.15
active power losses membership function and constraints may
23 0.15 0 0
impose a noticeable effect on the capacitor locations and MVArs 24 0 0.15 0.15
choices in such systems. The sectional active power losses for the 25 0 0 0.15
34-bus system for all methods are shown in Fig. 10. From this figure 26 0 0 0
it can be seen that, method (2) increased the active power losses in 27 0.45 0 1.05
28 0.15 0 0
sections no. (10–11), (22–23), (23–24), (24–25), (25–26), (31–33) 29 0 0 0
and (32–33). However, this increase in active power losses has not 30 0 0 0
been obtained in the proposed method, which validates the proper 31 0 0 0
behavior of the proposed active power loss membership function 32 0 0 0
33 0 0 0
(Figs. B1 and B2).
34 0 0 0.45
A summary of the results obtained for the 9-bus system is given Summation of MVAr 3.45 4.8 2.25
in Tables 1–3. Table 1 shows the locations and MVAr for the three
methods and the total MVArs required in each case. The proposed
method requires larger MVAr values than the other two methods.
The locations for capacitors are not the same for the three methods. method. In the proposed method, the fixed cost is higher than the
Table 2 explores the active losses in each section for each method. other two methods.
Table 3 shows that the proposed method achieved lower total active For the 34-bus system, the summary of results is given in
power losses and higher reduction than the other two methods. Tables 4–6. Table 4 shows the locations and total MVArs for the
This has been reflected on the lowest running cost of the proposed system. It is shown that the total MVArs of the proposed method
Table 2
The losses in each section for the 9-bus feeder in MW.
Without Qc The proposed method Method (1) [12] Method (2) [13]
Table 3
Summary of results for the 9-bus test system.
Method no. Without Qc The proposed method Method (1) [12] Method (2) [13]
Max voltage 1 1 1 1
Min voltage 0.8375 0.9 0.9 0.9
Total power loss in MW 0.783 0.680 0.707 0.705
Running cost $ 131,675 115,250 118,776 118,550
Fixed cost $ 0 1195.3 1006.8 1055.95
Total cost $ 131,675 115,535.3 119,782.8 119,595
Loss reduction in kW 0 103.8 76.8 78.8
Cost saving in $ 0 16250.7 11892.2 12,180.05
238 H.A. Ramadan et al. / Electric Power Systems Research 106 (2014) 232–240
are higher than that obtained by method (2). Method (2) and the losses in each section for each method. Table 6 shows that the
proposed method have the same voltage membership function. proposed method offers less running cost than the other two meth-
The effect of unincreased sectional active power losses appeared in ods. The fixed cost for the proposed method is higher than that
large MVArs for the proposed method. Table 5 explores the active of method (2) and is lower than that of method (1). The proposed
Table 5
The losses in each section for the 34-bus feeder in MW.
Without Qc The proposed method Method (1) [12] Method (2) [13]
Table 6
Summary of results for the 34-bus test system.
Method no. Without Qc The proposed method Method (1) [12] Method (2) [13]
Max voltage 1 1 1 1
Min voltage 0.9416 0.951 0.951 0.952
Total power loss in MW 0.2215 0.162 0.179 0.171
Running cost $ 37,212 27,216 30,072 28728
Fixed cost $ 0 1005.3 1277.85 795.45
Total cost 37,212 28,221.3 31,349.85 29,523.45
Loss reduction in kW 0 59.128 42.38 50
Cost saving in $ 0 8990.7 5862.15 7688.55
H.A. Ramadan et al. / Electric Power Systems Research 106 (2014) 232–240 239
method achieved the higher total cost reduction than the other two Appendix B.
methods.
See Tables B1 and B2 and Figs. B1 and B2.
6. Conclusion
Table B1
A fuzzy-based approach for optimal capacitor allocation and The load data and the feeder data of the 9-bus test system [3]. Base impedance
sizing has been presented. A new membership function for active = 529 .
sectional and total active power losses has been proposed and con-
Bus no. Load Sectional parameters
stant active sectional power loss concept has been implemented.
A voltage membership function has been used. The application of P Q Bus no. Ri , i + 1 () Xi , i + 1 ()
the proposed approach on 9-bus and 34-bus systems led to the (kW) (kVAr) From To
following conclusions: 1 1840 460 0 1 0.123 0.4127
2 980 340 1 2 0.014 0.6051
(1) The proposed membership function proved its proper behavior. 3 1790 446 2 3 0.746 1.205
(2) The increase in fixed cost resulted from the proposed method is 4 1598 1840 3 4 0.698 0.6084
5 1610 600 4 5 1.983 1.7276
compensated by the decrease in the running cost and generally 6 780 110 5 6 0.905 0.7886
lower total cost has been obtained. 7 1150 60 6 7 2.055 1.164
(3) The proposed approach is characterized by simple capacitor 8 980 130 7 8 4.795 2.716
allocation and sizing technique. 9 1640 200 8 9 5.343 3.0264
7. List of symbols
2 2 2 3 2 4 6 2 6 3
2 1 6 1
S 6 10
1 2 3 4 5 7 8 9 11
5 1 9 1
5 2
9 29 3 9 4
5 3
5 4 5 5 5 6 5 7 5 8 5 9 5 10 5 11
Fig. B2. The single line diagram of the 34 bus test system [13].
References [9] T.S. Abdel-Salam, A.Y. Chikhani, R. Hackam, A new technique for loss reduction
using compensating capacitors applied to distribution systems with varying
[1] H.N. Ng, M.M.A. Salama, A.Y. Chikhani, Classification of capacitor allo- load condition, IEEE Trans. Power Delivery 9 (2) (1994) 819–827.
cation techniques, IEEE Trans. Power Delivery 15 (1) (2000) 387–392, [10] M. Chis, M.M.A. Salama, S. Jayaram, Capacitor placement in distribution systems
Jan. using heuristic search strategies, in: Proc Inst. Elect. Eng. Gen. Transm. Dist., vol.
[2] S.H. Lee, J.J. Grainger, Optimum placement of fixed and switched capacitors 144, No. 2, 1997, pp. 225–230.
on primary distribution feeders, IEEE Trans. Power Apparatus and Systems [11] M.M. Hamada, M.A.A. Wahab, A.M. El-Sayed, H.A. Ramadan, A proposed strat-
PAS–100 (1) (1981) 345–352, Jan. egy for capacitor allocation in radial distribution feeders, in: Proc. 12th Middle
[3] J.J. Grainger, S.H. Lee, Capacity release by shunt capacitor placement on distri- East Power Systems (MEPCON) Conf, 2008, pp. 146–151.
bution feeders: a new voltage-dependent model, IEEE Trans. Power Apparatus [12] H. Chin, W. Lin, Capacitor placement for distribution systems with fuzzy algo-
and Systems PAS-101 (5) (1982) 1236–1244, May. rithm, in: Proc. IEEE Region 10‘s nineth Annu. Int. Conf.,Theme, Frontiers of
[4] M. Ponnavaikko, K.S. Prakasa Rao, Optimal choice of fixed and switched shunt Computer Technology, vol. 2, 1994, pp. 1025–1029.
capacitors on radial distributors by the method of local variations, IEEE Trans. [13] C. Su, C. Tsai, A new fuzzy-reasoning approach to optimum capacitor allocation
Power Apparatus Syst. PAS-102 (6) (1983) 1607–1615, June. for primary distribution system, in: Proc IEEE Int. conf. Industrial Technology,
[5] M.E. Baran, F.F. Wu, Optimal capacitor placement on radial distribution sys- 1996, pp. 237–241.
tems, IEEE Trans. Power Delivery 4 (1) (1989) 725–734. [14] H.N. Ng, M.M.A. Salama, A.Y. Chikhani, Capacitor placement in distribution
[6] T.M. Khalil, H.K.M. Yousssef, M.M. Abdel Aziz, Optimal capacitor placement on systems using fuzzy technique, in: Proc Can. Conf. Electrical and Computer
radial distribution feeders in presence of nonlinear loads using binary particle Engineering, vol. 2, 1996, pp. 790–793.
swarm optimization, in: 19th International Conference on Electricity Distribu- [15] M.M.A. Salama, A.Y. Chikhani, A simplified network approach to VAr control
tion, Vienna, 2007, pp. 119–122. problem for radial distribution systems, IEEE Trans. Power Delivery 8 (1993)
[7] S. Civanlar, J.J. Grainger, H. Yin, S.H. Lee, Distribution feeder reconfigura- 1529–1535.
tion for loss reduction, IEEE Trans. Power Delivery 3 (3) (1988) 1217–1223, [16] W.F. Mohammad, N. Tawalbeh, K.M. Al-Aubidy, Fast power loss computation
July. and shunt capacitor insertion using fuzzy logic technique, American Journal of
[8] T. Taylor, D. Lubkeman, Implementation of heuristic search strategies for distri- Applied Sciences 4 (1) (2007) 37–41.
bution feeder reconfiguration, IEEE Trans. Power Delivery 5 (1) (1990) 239–246, [17] Y. Baghzouz, S. Ertem, Shunt capacitor sizing for radial distribution feeders with
Jan. distorted substation voltages, IEEE Trans. Power Delivery 5 (1990) 650–657.