Professional Documents
Culture Documents
Optimal Allocation of Combined DG and Capacitor For Real Power Loss Minimization in Distribution Network
Optimal Allocation of Combined DG and Capacitor For Real Power Loss Minimization in Distribution Network
a r t i c l e i n f o a b s t r a c t
Article history: Optimal siting and sizing of Distributed Generation (DG) and shunt capacitor at the distribution networks
Received 26 November 2012 for the purpose of real power loss minimization is drawing much attention of electric power utilities in
Received in revised form 1 June 2013 the present days. Some inherent benefits of power loss minimization includes: reduction of power flow in
Accepted 6 June 2013
feeder lines, releases stress on feeder loading, and hence increases their life time, adds opportunity to
using the existing facility to serve any increased load demand, avoidance of power purchased from the
grid and also the cost of loss compensating devices, reduction in customer bill, etc. In this paper, a
Keywords:
method based on analytical approach for optimal allocation (sizing and siting) of DG and capacitor with
Power distribution networks
Distributed generation
the objective to minimize the total real power loss subjected to equality and inequality constraints in the
Shunt capacitors distribution network is presented. A sensitivity analysis technique has utilized to identify the optimal
Optimal siting candidate locations for DG and capacitor placement and the heuristic curve fitting technique is used to
Optimal sizing determine their optimal capacity in the networks. To validate the suitability of the proposed method,
Loss minimization it has been applied to 12-bus and IEEE 33-bus test distribution systems. The obtained simulation results
and comparison of different cases considered reveals that allocation of DG and capacitor combination
results in significant loss reduction with good voltage profile and also release in the line loading in the
power distribution networks.
Ó 2013 Elsevier Ltd. All rights reserved.
0142-0615/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijepes.2013.06.008
968 S. Gopiya Naik et al. / Electrical Power and Energy Systems 53 (2013) 967–973
of DG at optimal power factor has been proposed to minimize Here Ppq(Qpq) are the sending end active (reactive) power flows and
the power loss [18]. DG
Rpq(Xpq) are the series resistance (reactance). P DG
q ðQ q Þ are the active
Most of the approaches presented so far model the optimal allo-
cation of DG only. However, very few approaches [27–29] consid- (reactive) power injections by DG; Q Cq is the reactive power injec-
ered the optimal allocation of combined DG and capacitor, but tion by capacitor and P Lq ðQ Lq Þ are the total active (reactive) demand
release of line loading is not being considered and therefore needs load at bus q. PFq ðQ Fq Þ are the sum of active (reactive) power flows
further attention. The present paper considers the optimal alloca-
through all the downstream branches connected to bus q. Vq is
tion of DG and capacitor considering that DG is also capable to sup-
the magnitude of voltage at bus q. SB is a set of buses containing
ply reactive power in addition to real power. The optimal
all the buses in the system. The value of current flowing through
allocation of DG and capacitor with the key objective to minimize
a branch connected between nodes p and q is given as:
the real power loss besides satisfying the network constraints of vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u 2
the distribution network is presented. Heuristic sensitivity analy- uPpq þ Q 2pq
sis, i.e. both real and reactive power loss is being utilized to iden- Ipq ¼t ð4Þ
V 2p
tify the optimal candidate locations for DG and capacitor
placement, and the quadratic curve fitting technique is employed Mathematically, the objective function is given as:
to determine their optimal capacity in the distribution networks. X
Min PL ¼ I2pq Rpq ð5Þ
This paper is organized as follows: followed by the brief intro-
8p;qjp;q2SB
duction in Section 1, Section 2 gives the problem formulation,
and the methodology proposed is discussed in Section 3. In Sec- The above objective function is subjected to the set of equality and
tion 4, results and discussions are given for single DG capacitor inequality constraints as given below:
S. Gopiya Naik et al. / Electrical Power and Energy Systems 53 (2013) 967–973 969
1. Equality constraints 1. Once the optimal locations for DG and capacitor are identified,
– Power balance: The flow of active and reactive power in all the the maximum DG capacity is fixed as 50% of the total feeder
branches of the system must satisfy Eqs. (1) and (2), respectively. loading capacity and the capacitor capacity is fixed as 100% of
– Voltage equation: For all branches of the system, the voltage total MVAr loading of the network under consideration. For
magnitudes at sending and receiving end nodes must satisfy the combined DG–capacitor case both the DG and capacitor
Eq. (3). are varied in small incremental steps simultaneously. At each
– DG power factor: The power factor of DG connected at bus q step perform the load flow using Eqs. (1)–(3) and get the total
must satisfy the following eqn. real power loss using Eqs. (4) and (5) and check the constraints
described in Eqs. (7) and (8). If, at least one of the constraints is
PDG not satisfied, decrease/increase the capacity by 5% (one at a
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiq ¼ cos uq ð6Þ time) and again perform the load flow and after ensuring both
2 DG 2
ðPDG q Þ þ ðQ q Þ constraints are satisfied the corresponding capacity gives the
optimal size(s).
2. Inequality constraints
2. Step 1 is repeated for all sizes and finally, the minimum loss
– Bus voltage: The voltage at each bus must lie within the pre-
point is obtained.
scribed limits
3. The DG and/or capacitor size corresponding to the minimum
V min
q 6 V q 6 V max
q q 2 SB ð7Þ loss point are the optimal capacities.
4. Place the optimal sizes obtained in step 3 at optimal locations
– Line current: The flow of current through each branch should and once again perform load flow and obtain the total real
not exceed its thermal limit power loss, bus voltage values and line current value in the
Ipq 6 Irated 8pandq 2 SB ð8Þ network.
pq
Bus Number
0
Table 1
Optimal results of 12-bus network.
Particulars Total real power % Loss reduction Capacity/optimal location Minimum voltage
loss (kW) in the system (p.u.)
Base case 198.9 – – 0.94414 at bus 12
Case 1: Only capacitor 134.3 32.47 0.16 MVAr capacitor at bus 12 0.95596 at bus 11
Case 2: Only DG at upf 109.2 45.09 0.2 MW DG bus 12 0.98032 at bus 8
Case 3: DG with upf and capacitor 71.93 63.8 0.12 MW DG at bus 12 and 0.24 MVAr capacitor at bus 12 0.9815 at bus 8
Case 4: DG at 0.9 Pf with Qc = 0 57.41 71.14 0.2 MW and 0.1 MVAr DG at bus 12 0.9847 at bus 7
Case 5: DG at 0.75 pf with Qc = 0 47.56 76.09 0.2 MW and 0.16 MVAr DG at bus 12 0.9867 at bus 7
By sensitivity analysis as shown in Fig. 2, the optimal bus to The optimal DG and capacitor sizes obtained for the different
place the DG and capacitor was found to be 12 where the real cases considered are 0.16 MVAr, 0.2 MW, (0.12 MW, 0.24 MVAr),
power loss sensitivity w. r. t real and reactive power injection is (0.2 MW, 0.1 MVAr with Qc = 0) and (0.2 MW, 0.16 MVAr with
high. Qc = 0), respectively. The results of comparison of various cases
considered are given in Table 1. From the table, it may be noted
Case 1: Only the capacitor is placed at the optimal location. that more loss reduction is possible if DG capable to supply reac-
In this case the maximum capacity of capacitor is taken equal to tive in addition to real power (i.e. when the DG power factor is
total MVAr loading of the network, i.e. 0.395 0.4 MVAr and 0.75) and is placed at the optimal location with optimal sizes in
varied in small incremental steps from 0 to 0.4 MVAr. the network compared to other considered cases.
Case 2: Only DG operating at unity power factor (upf) is placed The comparisons of voltage profile at various buses for the dif-
at the optimal location ferent cases considered in a 12-bus radial distribution network are
In this case the maximum capacity of the DG is taken equal to depicted in Fig. 3. It can be seen from the figure that improvement
50% of the total MW loading of the network, i.e. in the voltage profile by optimal placement of combined DG oper-
0.5 0.435 0.25 MW and is varied in small incremental steps ating at upf and capacitor (Case 3) is somewhat superior compared
from 0 to 0.25 MW. to DG alone. Alternatively, if DG capable of supplying reactive
Case 3: Combined DG operating at upf and capacitor were power in addition to real power is placed at optimal location where
placed at their optimal location improvement in the voltage profile is found to be better than pre-
In this case the maximum capacity of DG and capacitor are vious cases.
selected same as in Case 1&2 and are simultaneously varied
in small incremental steps from 0 to 0.25 MW and 0 to
1.01
0.4 MVAr, respectively.
1
Case 4: Combined DG operating at 0.9 pf and the capacitor were
Voltage Profile (p.u)
0.99
placed at their optimal location.
0.98
In this case the maximum DG capacity is selected equal to 50%
0.97
of the total MVA loading of the network at 0.9 pf. So, the max-
0.96
imum DG capacity is 0.3 MW and 0.15 MVAr (with capacitor
0.95
value set to zero, i.e. Qc = 0), respectively, and are varied in small
0.94
incremental steps from 0 to 0.3 MW and 0 to 0.15 MVAr. 0.93
Case 5: Combined DG operating at power factor selected equal 0.92
to the total feeder load power factor and the capacitor were 0.91
placed at their optimal location. 1 2 3 4 5 6 7 8 9 10 11 12
In this case the maximum DG capacity is selected equal to 50% Bus Number
of the total MVA loading of the network at 0.75 pf lag (i.e. Base Case Case3
PFDG = PFD). So, the maximum DG capacity is 0.25 MW and Case 1 Case 4(0.9 pf)
0.2 MVAr (with capacitor value set to zero), respectively, and Case 2 Case5(0.75 pf)
are varied in small incremental steps from 0 to 0.25 MW and
0 to 0.2 MVAr. Fig. 3. Comparison of voltage profile for different Cases in 12-bus network.
S. Gopiya Naik et al. / Electrical Power and Energy Systems 53 (2013) 967–973 971
Fig. 3 also shows the comparison of voltage at various bus of 12- Case 4: Combined DG operating at 0.9 pf (considered as optimal
bus network for DG operating at 0.9 pf (Case 4) and 0.75 pf (Case pf) lag and the capacitor were placed at their optimal location
5), respectively. It can be seen that the voltage profile improve- In this case the maximum DG capacity is selected equal to 50%
ment with DG operating at 0.75 pf lag is better compared to the of the total MVA loading of the network at 0.9 pf. So, the max-
one at 0.9 pf lag, but the DG capacity required is more which re- imum MW and MVAr capacity of DG is 2.0 and 1.0, respectively,
sults in increased cost of the DG with additional loss reduction and the maximum capacitor capacity is selected as 2.3 MVAr
benefit of about 5%. and is simultaneously varied in small incremental steps from
0 to 2 MW, 0 to 1 MVAr and 0 to 2.3 MVAr.
4.2. 33-Bus network Case 5: Combined DG operating at the pf equal to total feeder
load pf and the capacitor were placed at their optimal location.
This is a 12.66 kV radial distribution system having 33 buses In this case the maximum DG capacity is selected equal to 50%
and 32 branches. The total load of the feeder is 3.72 MW and of the total MVA loading of the network at 0.85 pf lag (i.e.
2.3 MVAr. The line and load data are given in [35]. The combined PFDG = PFD). So, maximum MW and MVAr capacity of DG is
load power factor of this system is 0.85. 2.0 and 1.2, respectively, and the maximum capacity of capaci-
By sensitivity analysis given in Fig. 4, the optimal locations ob- tor is selected as 2.3 MVAr and is simultaneously varied in small
tained for DG and capacitor are 18 and 33 where the real power incremental steps from 0 to 2 MW, 0 to 1.2 MVAr and 0 to
loss sensitivity w. r. t real and reactive power injection are high, 2.3 MVAr.
respectively.
The optimal sizes obtained for different cases considered are
Case 1: Only capacitor is placed at the optimal location. 1 MVAr, 1 MW, (1 MW, 1 MVAr), (1.0 MW, 0.5 MVAr, 1 MVAr),
In this case the maximum capacity of capacitor is taken equal to and (0.8 MW, 0.4 MVAr, 0.8 MVAr), respectively. Table 2 compares
total MVAr loading of the network, i.e. 2.3 MVAr and varied in the results of different cases considered. It is seen from the table
small incremental steps from 0 to 2.3 MVAr. that if optimal size of DG that supplying both real and reactive
Case 2: Only DG operating at upf is placed at the optimal power at bus 18 and the capacitor at bus 33 were placed simulta-
location. neously, more loss reduction can be achieved compared to the
In this case the maximum capacity of the DG is taken equal other cases considered.
to 50% of the total MW loading of the network, i.e. The comparisons of voltage profile at different buses for the var-
0.5 3.72 2.0 MW and is varied in small incremental steps ious cases considered in a 33-bus network are depicted in Fig. 5. It
from 0 to 2.0 MW. can be seen from the figure that if combined DG and capacitor,
Case 3: Combined DG operating at upf and capacitor were where DG is capable to supply reactive power in addition to real
placed at their optimal location. power are optimally allocated can result in the substantial
In this case the maximum capacity of DG and capacitor are improvement in the voltage profile along the feeder compared to
selected same as cases 1 and 2 and are simultaneously varied
in small incremental steps from 0 to 2 MW and 0 to 2.3 MVAr,
respectively.
1.04
1.02
Voltage Profile (p.u)
Bus Number 1
0 0.98
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
Real power loss sensitivity
-0.02 0.96
-0.04 0.94
-0.06 0.92
0.9
-0.08
0.88
-0.1
0.86
-0.12 0.84
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33
-0.14
-0.16 Bus Number
-0.18 Base Case Case 3
Real Power Loss Sensitivity Real Power Loss Sensitivity Case 1 Case 4 (0.9 pf)
w.r.t Real Power Injection w.r.t Reactive Power
Injection Case 2 Case 5 (0.85 pf)
Fig. 4. Real power loss sensitivity at various buses of a 33-bus network. Fig. 5. Comparison of voltage profile for different cases in 33-bus network.
Table 2
Optimal results of 33-bus network.
Before DG-Capacitor
Current magnitude (p.u)
After DG-Capacitor
Branch Number
Fig. 6. Feeder loading before and after DG–capacitor allocation for Case 5 of 33-bus network.
Table 3 Table 4
Comparison of optimal results of different approaches for 33-bus network. Comparison between single and multiple DG capacitor allocation.
Method [21] [20] [36] Proposed Particulars Single DG capacitor Multiple DG capacitor
allocation allocation
Optimal location(s) 6 6 6 18 and 33
DG capacity (MW) 2.49 3.38 2.4 0.76 DG capacity (MW) 0.76 0.855
Capacitor capacity (MVar) – – – 0.8 Capacitor capacity (MVar) 0.80 0.90
Percentage loss reduction 47.33 44.83 48.19 57.94 Power loss 89.72 84.28
% Loss reduction 57.94 60.49
1.01
1 [21,36], the DG units are limited to supply real power only where
as in the method [20] the DG unit is capable to supply both real and
Voltage profile (p.u)
0.99
reactive power. But, in the presented work the combined DG and
0.98 capacitor (Case 5) where DG unit is capable to deliver both real
0.97 and reactive power to the network is resulted in more loss reduc-
0.96 tion and better voltage profile in comparison to other methods.
0.95
0.94 4.3. Multiple DG capacitor allocation
0.93
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33
Optimal locations: From Fig. 4 it s clear that bus numbers 18
Bus Number and 17 have got highest values of real power loss sensitivity in
Single DG Capacitor Multiple DG Capacitor decreasing order with respect to real power injection and these
Allocation allocation
values are very close to each other. Therefore, bus number 18
Fig. 7. Voltage profile of single/multiple DG capacitor allocation. and 17 are considered as the candidate buses for DG siting. Further,
bus number 33 and 32 have got highest values of real power loss
sensitivity in decreasing order with respect to reactive power
the case when DG supplying only real power and capacitor combi- injection and are approximately same therefore, these buses are
nation was placed at their optimal locations. considered as the candidates for capacitor siting.
Fig. 5 also shows the comparison of voltage at various bus of 33- Optimal size(s): For determining the optimal size of multiple
bus network for 0.9 pf (Case 4) and 0.85 pf (Case 5). It can be seen DG and capacitor, simulation is conducted only for the case 5 of
that there is voltage limit violation with DG operating at 0.9 pf lag IEEE 33-bus network. Applying the algorithm detailed in Sec-
compared to that one at 0.85 pf lag, also the DG capacity required tion 3.1.2, optimal sizes of multiple DG and capacitor required
is more which results in increased cost of the DG with less loss for total possible minimum real power loss in the system satisfying
reduction of about 0.56% from Case 5. the system constraints are found to be 0.447 MVA, 0.559 MVA and
Feeder loading before and after DG–capacitor allocation for the 0.4 MVAR, 0.5 MVAR at locations 18, 17 and 33, 32, respectively.
Case 5 of 33-bus network is given in Fig. 6. It can be seen from the With this capacity the resulted total real power loss is 84.28 kW
figure that there results in significant release of (32.64%) line load- which is lesser than single DG capacitor allocation by 5.44 kW,
ing in comparison to base case (without DG). So, the life time of the but at the cost of more DG and capacitor size requirements. Com-
feeder is increased and the existing system can be used to supply parison of single and multiple-DG capacitor allocation is given in
any possible future load growth. Table 4.
To validate the results obtained by the proposed method, the Fig. 7 shows the voltage profile for single and multiple DG
proposed analytical method is compared with the solutions ob- capacitor allocation cases. It is observed that voltage profile is
tained by analytical method [21,36], and ABC algorithm [20] and somewhat improved in the case of multiple DG capacitor
the same is given in Table 3. In those methods proposed in allocation.
S. Gopiya Naik et al. / Electrical Power and Energy Systems 53 (2013) 967–973 973
5. Conclusions [13] Park JY, Sohn JM, Park JK. Optimal capacitor allocation in a distribution system
considering operation costs. IEEE Trans Power Syst 2009;24(1):462–8.
[14] Hung DQ, Mithulanathan N, Bansal RC. Multiple distributed generators
This paper was presented a method for optimal siting and sizing placement in primary distribution networks for loss reduction. IEEE Trans
of combined DG and capacitor in the distribution network. It can be Ind Electron 2011;99:1–8.
[15] Wang C, Nehir MH. Analytical approaches for optimal placement of distributed
concluded that allocation of capacitor alone can improve the volt-
generation sources in power system. IEEE Trans Power Syst
age but may not reduce the loss as expected. On the other hand, 2004;19(4):2068–76.
allocation of DG can reduce the power losses and also improves [16] Quezeda VH Mendez, Jua-Rivier Abbad, Gomez T. Assessment of energy
distribution losses for increasing penetration of DG. IEEE Trans Power Syst
the system voltage profile. If the DG, with both real and reactive
2006;21(2):533–40.
power generation capability is combined with the capacitor and [17] Mithulanathan N, Oo Than, Van Phu Lee. Distributed generator placement in
is optimally allocated, can result in substantial reduction in real power distribution system using genetic algorithm to reduce losses,
power loss of the distribution network and also improves the volt- Thammasat. Int J Sci Technol 2004;9(3):55–62.
[18] Hung DQ, Mithulananthan N, Bansal RC. Analytical expressions for DG
age profile along the feeder. In comparison to the conventional allocation in primary distribution networks. IEEE Trans Energy Convers
planning methods, the application of combined DG unit and capac- 2010;25(3):814–20.
itor can largely reduce the investment besides improving supply [19] Khatod DK, Pant Vinay, Sharma Jaydev. Evolutionary programming based
optimal placement of renewable distributed generators. IEEE Trans Power Syst
quality and reliability. The proposed methodology has been further 2013;28(2):683–95.
extended for multiple combined DG–capacitor (MCDGC) allocation [20] Abu-Mouti FS, El-Hawary ME. Optimal distributed generation allocation and
and it is found that some improvement in the loss reduction and sizing in distribution systems using artificial bee colony algorithm. IEEE Trans
Power Deliv 2011;26(4):2090–101.
voltage profile is possible with MCDGC compared to the single [21] Acharya Naresh, Mahat Pukar, Mithulananthan N. An analytical approach for
DG capacitor case. DG allocation in primary distribution networks. Electr Power Energy Syst
2006;28:669–78.
[22] Gandomkar M, Vakilian M, Ehsan M. A genetic based tabu search algorithm for
Acknowledgement optimal DG allocation in distribution networks. Electr Power Compon Syst
2007;33(12):1351–62.
[23] Choi JH, Kim JC. Network reconfiguration at the power distribution system
The authors would like to acknowledge AICTE’s Quality with dispersed generations for loss reduction. In: Proc of IEEE power
Improvement Programme (QIP) Scheme and the Management engineering society winter meeting, vol. 4; 2000. p. 2363–7.
P.E.T. (R), Mandya (Karnataka), for allowing pursuing the Ph.D. [24] Wu YK, Lee Ching-Yu, Liu Le-Chang, Tsai Shao-Hong. Study of reconfiguration
for the distribution system with DG. IEEE Trans Power Deliv
work. 2010;25(3):1678–84.
[25] Khorshidi R, Niknam T, Nayeripour M. Distribution feeder reconfiguration
considering distributed generators. World Acad Sci Eng Technol 2008:46.
References [26] Niknam T, Fard Abdollah Kavousi, Fard Alireza. Distribution feeder
reconfiguration considering fuel cell/wind/photovoltaic power plants. Renew
[1] Song YH, Wang GS, Johns AT, Wang PY. Distribution network reconfiguration Energy 2010;37:213–25.
for loss reduction using fuzzy controlled evolutionary programming. Proc Gen [27] Kai Zou, Agalgaonkar AP, Muttaqi KM, Perera S. Voltage support by distributed
Trans Distrib 1997;144:345–50. generation units and shunt capacitors in distribution systems. In: IEEE power
[2] Gallego RA, Monticelli AJ, Romero R. Optimal capacitor placement in radial and energy society general meeting; 2009. p. 1–8.
distribution networks. IEEE Trans Power Syst 2001;16(4):630–7. [28] Zou Kai, Agalgaonkar AP, Muttaqi KM, Perera S. Optimization of distributed
[3] Ng HN, Salama MMA, Chikhani AY. Classification of capacitor allocation generation units and shunt capacitors for economic operation of distribution
techniques. IEEE Trans Power Deliv 2000;15:387–92. systems. In: Australasian universities power, engineering conference; 2008. p.
[4] Chians HD, Wang JC, Cockings O, Shin HD. Optimal capacitor placement in 1–7.
distribution systems: Part I, Part II. IEEE Trans Power Deliv 1990;5(2):643–7. [29] Gunda Jagadeesh, Khan Nasim Ali. Optimal location and sizing of DG and shunt
[5] Pecas JA, Lopes, Hatziargyriou N, Mutale J, Djapic P, Jenkins N. Integrating capacitors using differential evolution. Int J Soft Comput 2011;6(4):128–35.
distributed generation into electric power systems: a review of drivers, [30] Quezeda VH Mendez, Abbad Jua-Rivier, Gomez T. Assessment of energy
challenges and opportunities. Electr Power Syst Res 2007;77:1189–203. distribution losses for increasing penetration of DG. IEEE Trans Power Syst
[6] Singh D, Singh D, Verma KS. Multi-objective optimization for DG planning with 2006;21(2):533–40.
load models. IEEE Trans Power Syst 2009;24(1):427–36. [31] Haque MH. Efficient load flow method for distribution systems with radial or
[7] Etemadi AH, Fotuhi-Firuzabad M. Distribution system reliability enhancement mesh configuration. IEE Proc Gen Trans Distrib 1996;143(1):33–8.
using optimal capacitor placement. IET Gener Trans Distrib 2008;2(5):621–31. [32] Khatod DK, Pant V, Sharma JD. A novel approach for sensitivity calculations in
[8] Gopiya Naik S, Khatod DK, Sharma MP. Planning and operation of distributed the radial distribution systems. IEEE Trans Power Deliv 2006;21(4):2048–57.
generation in distribution networks. Int J Emerg Technol Adv Eng [33] Abu-Mouti FS, El-Hawary ME. Heuristic curve-fitted technique for distributed
2012;9(2):381–8. generation optimization in radial distribution feeder systems. IET Gen Trans
[9] Haque MH. Capacitor placement in radial distribution systems for loss Distrib 2011;5(2):172–80.
reduction. IEE Proc Gen Trans Distrib 1996;146(5). [34] Das D, Nagi HS, Khotari DP. Novel method for solving radial distribution
[10] Sundhararajan S, Pahwa A. Optimal selection of capacitors for radial networks. IEE Proc Gen Trans Distrib 1994;141(4).
distribution systems using a genetic algorithm. IEEE Trans Power Syst [35] Baran M, Wu FF. Network reconfiguration in distribution systems for loss
1994;9(3):1499–507. reduction and load balancing. IEEE Trans on Power Deliv 1989;4(2):1401–7.
[11] Mekhamer SF, El-Hawary ME, Soliman SA, Moustafa MA, Mansour MM. New [36] Shukla TN, Singh SP, Srinivasarao V, Naik KB. Optimal sizing of distributed
heuristic strategies for reactive power compensation of radial distribution generation placed on radial distribution systems. Electr Power Compon Syst
feeders. IEEE Trans Power Deliv 2002;17(4):1128–35. 2010;38:260–74.
[12] Levitin G, Kalyuzhny A, Shenkman A, Chertkov M. Optimal capacitor allocation [37] Teng JH. Modeling distributed generations in three-phase distributed load
in distribution systems using a genetic algorithm and a fast energy loss flow. IET Gen Trans Distrib 2008;2(3):330–40.
computation technique. IEEE Trans Power Deliv 2000;15(2):623–8.