Loss minimization techniques used in distribution network: bibliographical survey

Loss minimization techniques used in distribution network:bibliographical survey

Shilpa Kalambe n, Ganga AgnihotriDepartment of Electrical Engineering, M.A.N.I.T., Bhopal 462051, M.P., India

a r t i c l e i n f o

Article history:Received 9 January 2013Received in revised form14 August 2013Accepted 24 August 2013

Keywords:Distribution lossNetwork restructuringCapacitor placementDG allocation

a b s t r a c t

Distribution system provides a link between the high voltage transmission system and low voltageconsumers thus I2R loss in a distributed system is high because of low voltage and high current.Distribution companies (DISCOs) have an economic enticement to reduce losses in their networks.Usually, this enticement is the cost difference between real and standard losses. Therefore, if real lossesare higher than the standard ones, the DISCOs are economically penalized or if the opposite happens,they obtain a profit. Thus loss minimization problem is a well researched topic and all previousapproaches vary from each other by selection of tool for loss minimization and thereafter either in theirproblem formulation or problem solution methods employed. Many methods of loss reduction exist likefeeder reconfiguration, capacitor placement, high voltage distribution system, conductor grading,Distributed Generator (DG) Allocation etc. This paper gives a bibliographical survey, general backgroundand comparative analysis of three most commonly used techniques (i) Capacitor Placement, (ii) FeederReconfiguration, (iii) and DG Allocation for loss minimization in distribution network based on over 147published articles, so that new researchers can easily find literature particularly in this area.

& 2013 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1842. Types of losses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1863. Capacitor placement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186

3.1. Loss equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1863.2. Note worthy contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186

4. Network reconfiguration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1935. DG allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1946. Comparative analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1977. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

1. Introduction

Losses in transmission and distribution networks represent thesingle largest consumption in any power system. Due to the rapidincrease in the demand for electricity, environmental constraints

and competitive energy market scenario the transmission anddistribution systems are often being operated under heavilyloaded conditions and the distribution system loss has becomemore and more of a concern. The requirement to provide accep-table power quality and enhanced efficiency to achieve all possibleeconomic benefits will create a very favorable climate for the needof loss minimization techniques and innovative operating prac-tices. The total power delivered to the distribution system hasbeen calculated according to the total power generation andpower loss of the transmission system. To enhance the efficiency

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/rser

Renewable and Sustainable Energy Reviews

1364-0321/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.rser.2013.08.075

n Corresponding author. Tel.: þ91 966 921 9959.E-mail addresses: [email protected] (S. Kalambe),

[email protected] (G. Agnihotri).

Renewable and Sustainable Energy Reviews 29 (2014) 184–200

www.sciencedirect.com/science/journal/13640321

www.elsevier.com/locate/rser

http://dx.doi.org/10.1016/j.rser.2013.08.075



http://crossmark.dyndns.org/dialog/?doi=10.1016/j.rser.2013.08.075&domain=pdf



mailto:[email protected]




of distribution system loss minimization is the only alternative.Thus it is found that, since last three decades research indistribution systems has been focused on line loss minimizationand voltage regulation. Various methods of loss minimization indistribution system are available in the literature but the basicthree methods such as (i) Capacitor Placement (generally applic-able in high voltage distribution systems) (ii) Feeder Reconfigura-tion (generally applicable in low voltage distribution systems) and(iii) DG Allocation (more focused to achieve interconnection whensmall generators exists. For instance, when isolated wind farms orsmall photovoltaic plants enter the distribution network) arediscussed here.

Traditionally, loss minimization has focused on optimizingnetwork reconfiguration or reactive power support through

capacitor placement. However, the evolution from passive distri-bution networks to active due to the insertion of DG presentsopportunities. Although planning issues, the regulatory frameworkand the availability of resources limit Distribution Network Opera-tors (DNOs) and developers in their ability to accommodatedistributed generation, governments are incentivizing low-carbon technologies, as a means of meeting environmental targetsand increasing energy security. This momentum can be harnessedby DNOs to bring network operational benefits through lowerlosses delivered by investment in DG [1]. This article incorporatesthe various approaches made by researchers in order to minimizethedistribution losses and the prolonged study shows that the lossminimization by DG allocation is providing enhanced prospects.

Nomenclature

L energy loss caused by quadrature current in feederCkVA capacitive kilovolt-amperesKVAr reactive kilovolt-ampereskVA kilo volt-ampRMS root mean squareEP evolutionary programmingULTC under-load tap changerCL capacitor location, per unit (pu) of total feeder

resistanceCR rated capacitor current, pu of maximum input quad-

rature current to feederT1, T2…Tn time, pu of period of load cycle, during which the

input quadrature current equalsX1,X2,….Xn ratio between input quadrature current during

time Tn and maximum input quadrature currentR line resistance in pui reactive current at feederX distance from feederLE energy lossx distance measured along the normalized equivalent

uniform feederF(x) feeder reactive current functionz control variableDPSO discrete particle swarm optimizationIS peak reactive current injected into the feederr P.U. length resistance of lineIcj sizes of shunt capacitors (i¼1,2……n)t time required to on/off fixed capacitorsOF objective function ($)OD present worth of revenue requirements of one dollar

investment, ($/$)BPSO binary particle swarm optimizationRC cost of released kVA capacity ($)CC construction cost of all related capacitor

installations ($)τ power loss reduction timeSC annual marginal cost of system capacity ($/kW)PL power loss savings at time of system peak (kW)EC marginal cost of energy losses ($/kWh)EL annual energy loss savings (kWh)Fp, FE present worth factors for power and energy losses,

which are functions of discount rate, comparison timeinterval and appropriate cost escalation.

Xline aggregate reactance of the line connecting the load tothe feeding substation

SLoss ij power loss in line i–j

Sij complex power from bus i to jSji complex power from bus j to iPq(τ) power loss reduction at time τIc capacitor currenta(τ) the substation load time variationH matrix to calculate distance between substation to

capacitorbT reactive load current distributionI2 end load currentS savings ($/yr)LP reduction in peak power lossesCC total cost of capacitorsKp factor to convert peak power losses to dollarsKe factor to convert energy losses to dollarsδ load anglesθ voltage anglem, q variables to represent number of busesF objective functionKr annual cost per unit of the real power loss ($/kW/yr)Pload active power of loadQload reactive power of loadTPC Taiwan power companyKc annual cost per unit of the reactive power injection at

bus i ($/kVAR/yr)Qc reactive power injection at respective busPB, QB active and reactive power flow in the branchPL, QL active and reactive load powerQC capacitor powerCPU central processing unitR resistance of branchJ complex current flow in branchEm component of' E¼RbusIbus’ corresponding to bus m.

Rbus is the “bus resistance matrix” of feeder before theload transfer which is found using the substation busas reference.

Ibus the vector of bus currents for feederEn similar to Em, but defined for bus n of next Feeder.PDG active power of DGRe real part of impedance.VL load voltagePrloss real power loss of the system.kp, ke and kq the cost of power loss at peak-load time (in

$/kW), the cost of fuel served for energy losses (in$/kWh) and the cost of reactive sources (in$/kVAr), resp.

τa fraction of time in load curveQDG reactive power of DGSCADA supervisory control and data acquisition

S. Kalambe, G. Agnihotri / Renewable and Sustainable Energy Reviews 29 (2014) 184–200 185

In this bibliography, collection of selected literature from IEEEtransactions, IEE proceedings, proceeding of conferences, andleading technical journals such as International Journal ofElectrical Power and Energy Systems, Power System Research,Energy etc. is included. Exceptions were made to include publica-tion in other resources if a publication offered a significantlyunique, technical viewpoint on the relevant issue.

2. Types of losses

The studies found in the literature can be classified into twoapproaches:

1. minimization of power losses and2. minimization of energy losses.

Energy loss deals with fuel burned to produce that energywhereas the peak power loss is important because it has a definitebearing on the equipment ratings to carry the peak load.

3. Capacitor placement

This section contains those publications which are related toDistribution Network Loss Reduction by using Capacitor PlacementTechnique which was found to be feasibly applicable in highvoltage distribution systems. The application of capacitors toelectric power systems can be used for:

1. the control of power flow,2. stability improvement,3. voltage profile management,4. power factor correction, and5. power and energy loss reduction.

The capacitor is a source of reactive power as by reducing theinductive reactance portion of the line loading, it can reduce thereactive losses which can be done by the addition of shuntcapacitors. Various authors have done a perceptible work incapacitor placement technique for voltage control and thereafterfor loss minimization [1].

The main challenges in this technique are:

1. selection of an appropriate number of capacitor units,2. allocation of capacitors, and3. sizing of capacitors to achieve a required result i.e.,

a. loss reduction,b. voltage regulation, andc. power flow control.

The advantages of varying the capacitive volt-amp reactive(VARs) in response to the load change have been recognized since1940s. Before 50s the trend of loss minimization by capacitorplacement at the substation was prevailing but from the decade of50s the trend has been started to install capacitors out on theprimary distribution feeders closer to the loads rather than at thesubstation due to both the availability of pole-mounted equipmentand because of the economics favoring the arrangement [2].

3.1. Loss equation

Consider a single line diagram of a uniformly loaded feeder asshown in Fig. 1.

The basic reactive loss given by equation

Total Losses ðdue to reactive currentÞ

¼ minZ

i2Rdx¼ minZ

½Ið1�xÞ�2Rdx ð1Þ

where i is the reactive current at feeder; R is the total resistanceand x is the distance from source.

This formula is used in the literature, in various forms for sitingand sizing the capacitors so as to achieve minimum losses.

3.2. Note worthy contributions

In 1956 Neagle and Samson [3] has presented rules-of-thumbfor capacitor allocation. In 1961 Cook [4] has shown an out-standing work by considering the effects of Energy and PeakPower Losses as well as released capacity or reduction inkilovolt-amp (kilo Volt-Amp) demands. In 1968 Duran [6] devel-oped the method which determines conditions when the capaci-tors are not economically justified. All those approaches sufferfrom following drawbacks:

1. very restricted reactive-load distribution,2. wire size of the feeder has been usually assumed to be uniform,3. voltage control problem was not considered,4. consideration of limited number of capacitors at a time,5. solutions obtained under these assumptions may be far from

real circumstances, and6. methods were practically not compatible

In 1981, Grainger and Lee [7] by eliminating many previousshortcomings set forth procedures for sizing and locating fixedshunt capacitors on primary distribution single radial feederhaving possibly many sections of different wire sizes and for anyknown reactive-load distribution along the feeder which may notnecessarily either uniformly distributed or concentrated feeders.Whereas in [8] they extend their concepts and modeling proce-dures to account for both fixed and switched capacitors. After thatin 1982 [9] they had proposed voltage dependent methodology forshunt capacitors which allows incorporation of an a. c. load flowprogram into the procedure for radial feeder compensation. In1983 Grainer et al. [10] presented continuously controllable,capacitive compensation scheme for primary distribution feederswhich assist distribution automation schemes being consideredfor implementation by electric utilities which rely heavily onsubstation-based computers for control of Reactive Power onprimary feeders. But none of the above approaches have acc-ounted for the benefits due to either capacity release, effects of thegrowth factors, load growth, growth in load factor, voltage riseproblems during off-peak period and change in cost of energy. Butall those methodologies were not capable of determining theoptimal number of capacitors, their type and the optimum banksize selected was not necessarily a standard industry size. In 2008Ulinuha et al. [39] proposed Evolutionary-Based Algorithms whichwere capable of optimizing large distribution systems with differ-ent types of nonlinear loads. In this algorithm the optimalscheduling of Load Tap Changers and switched shunt capacitorsfor simultaneously minimizing energy loss and improving thevoltage profile while taking harmonics into account was

l

x dx

i(l-x)

i

Sub-station

Feeders

Fig. 1. Single line diagram of uniformly loaded feeder.

S. Kalambe, G. Agnihotri / Renewable and Sustainable Energy Reviews 29 (2014) 184–200186

Table 1Gradual evolution in capacitor placement method.

Author Type of loss/application

Specific feature if any Optimizationapproaches/techniques

Optimization function/basic principle Network Achievements/merits De-merits

Neagle andSamson [3]

Power loss/shuntcapacitors.

Maximum loss reduction for asingle capacitor bank can beobtained when Ckva of the bank isequal to two thirds of the KVAr loadon the feeder.

Rules-of-thumb Total losses ¼ Ri2Rdx¼ R ½Ið1�xÞ�2Rdx Radial circuits with

distributed loadStarted the trend of moving thecapacitors from the substationout to the load areas.

Only peak kilowatt losssavings are considered.Not suitable for largersystems.No guarantee of globaloptimality

Cook [4] Power and energyloss/fixed andswitched shuntcapacitors

Considered the effects of losses aswell as released capacity orreduction in KVA demands.

Digitalprogramming L¼

Z CL

0

½ðX1�X1R�CRÞ2T1

þðX2�X2R�CRÞ2T2

þ…þðX1�XnR�CRÞ2Tn�

þZ 1

CL

½ðX1�X1RÞ2T1þðX2�X2RÞ2T2

þ…þðXn�XnRÞ2Tn�dR

Radial circuits withdistributed load.

Economic location, optimumsize, location, and switchedcapacitors are determined

Fixed cost per capacitor isassumed.Incremental cost isneglected.Reactive losses areneglectedSeparate optimization offixed and switchedcapacitors

Schmill [5] Power loss/shuntcapacitors

Build a search routine in which losssavings are compared for differentlocations of the capacitor banksuntil optimum savings are obtained

Load flow Total losses ¼ Ri2Rdx¼ R ½Ið1�xÞ�2Rdx Uniformly

distributed loadsand randomlydistributed variablespot loads.

Optimum sizes of capacitorbanks for specific location.

Uniform wire size.Method ofcalculus

Ignored voltage controlproblem.Cost saving.

Duran [6] Power loss/fixed typeof Shunt capacitors

Developed the method whichdetermines conditions when thecapacitors are not economicallyjustified.

Dynamicprogramming

Modified form of objective function usedby Cook

Radial distributionfeeder withdiscrete lumpedloads

Optimum number, location, sizeand cost saving.

Voltage control problem isnot assumed.

Grainger andLee [7–10]

Power and energy/fixed and switchedcapacitors

By eliminating many previousshortcomings set forth proceduresfor sizing and locating fixed shuntcapacitors. Proposed voltagedependent methodology for shuntcapacitors which allowsincorporation of an a. c. load flowprogram into the procedure forradial feeder compensation

Computer-basedoptimizationtechniques

Equal area criterion Primarydistribution singleradial feederhaving possiblymany sections ofdifferent wire sizes

Optimum sizing, siting for anyknown reactive-loaddistribution along the feederwhich may not necessarilyeither uniformly distributed orconcentrated feeders, numberof units and cost of capacitors.

Not considered the benefitsdue to capacity release,effects of the growthfactors, load growth, growthin load factor, voltage riseproblems during off-peakperiod and change in cost ofenergy.

Ponnavaikkoand Rao[11]

Energy loss/fixed andswitched capacitors

Suggested a mathematical modelwhich represent cost saving due toenergy loss reduction.

Method of LocalVariations anddynamicprogramming

LE¼ 3R fR ðIsðtÞFðxÞÞ2rdx

�Z

ðIsðtÞFðxÞ�∑Z

ðISðtÞFðxÞ�∑IcjÞ2rdx

þZ

ðIsðtÞFðxÞÞ2rdxgdt

Radial distributionfeeders

Considered load growth,growth in load factor andincrease in cost of energy

Consideration of mainfeeder branch onlyArbitrary number ofcapacitor banks.

Kaplan [12] Power and energyloss/fixed andswitched shuntcapacitors

Developed a method for costoptimization associated withreleased system capacity

Heuristics basedcomputerizedmethod

OF¼OD(RC�CC)þSCFP PLþEC FE EL Radial distributionfeeders

Introduced the method whichconsiders the installation costof capacitor banks also.

But the procedures aremainly based on heuristics

Grainger andCivanlar[13–15]

Power and energyloss/fixed andswitched shuntcapacitors

Modeled the distribution systemsencountered in practice.Formulated, simplified and solvedthe problem of volt/var control ongeneral radial distribution systemswith lateral branches by properallocation and sizing of voltageregulators and shunt capacitors inthree companion papers.

Dynamicprogramming

PQ ðτÞ ¼ rð�ITC ðHÞIC þ2aðτÞbT IC Þ Modeled thedistributionsystemsencountered inpractice.

Nonlinear costs of installationof the capacitors areincorporated into the capacitorproblem.

Not capable of determiningthe optimal number ofcapacitors and their typeand the optimum bank sizeselected was not necessarilya standard industry size.

S.Kalam

be,G.A

gnihotri/Renew

ableand

SustainableEnergy

Review

s29

(2014)184

–200187

Table 1 (continued )




Salama et al.[16]

Power and energyloss/shunt capacitors

Presented a mathematical analysisof shunt capacitor application inuniform feeder with an end-loadcondition by considering therequirements in practice.

Dynamicprogramming

Total losses¼ 3xIc ½2�xð1�I2Þ�Ic � Feeders withuniformlydistributed loadsand end-load aredeveloped.

Conditions of fixed load, growthin the load and the presence ofend-load in the feeder.

Becomes complicated forlarge systems

Bishop et al.[17]

Power loss/shuntcapacitors

Gave an examination of sevenfeeders yielding a significantreduction in real line losses,formulating the justification forfuture capacitor application efforts.

AnalyticalApproach

min∑RjJj2 Feeders withuniformlydistributed loads.


The optimum bank sizeselected was not necessarilya standard industry size

Suat Ertemet al. [18]

Energy loss/shuntcapacitors

Developed a nonlinear model basedon the bus impedance referenceframe, by taking into account theuncertainty of load

PatternRecognitionTechniques

Sensitivity analysis Transmission aswell as distributionsystem.

Method suitable for bothtransmission as well asdistribution system.

No assurance of optimality.

Lee [19] Power loss/Fixed andswitched shuntcapacitors

Developed a straightforwardmethod, capable of examining theeffects of continuoussystem reconfiguration of switchesand capacitors with automateddistribution control.

Dynamicprogramming

Min f ¼ minðPr; lossÞ Real timeDistributionSystem


Separate optimization offixed and switchedcapacitors

Baran et al.[20]


A solution method has beenimplemented that decomposes theproblem into a master problem anda slave problem. The masterproblem is used to determine thelocation of the capacitors. The slaveproblem is used by the masterproblem to determine the type andsize of the capacitors placed on thesystem.

Mixed integerprogramming

Based on probability Radial distributionsystem

Considered discrete nature ofcapacitor sizes and locations


Ertem et al.[21]


Developed a mathematical modeland presented a solution methodfor reactive power compensation ofdistribution circuits havingnonlinear loads

Dynamicprogramming

Min f ¼ minðPr; lossÞ Radial distributionsystem

Increase in RMS voltages due toharmonic current flows isincorporated in the problemformulation.

Lengthy and complicatedapproach.

Baldick et al.[22]


Theory is compared to load flowcalculations and a simple quadraticfunction was shown to be accuratein estimating the losses

Load Flow analysis Examined the distribution system lossfunction and voltage dependence andderived useful approximations analogousto the loss formulae used in thetransmission system

Radial distributionsystem

Considered the conditions ofvarying the capacitivesusceptance ofcapacitors installed inthe system

Iterative approach makesprocess lengthy.

Salama andChikhani[23]


The distribution system voltageprofile was kept within the desiredlimits by proper choice of bothshunt capacitors and voltageregulators which provided thevalues of the reactive power levelsto be injected into the distributionsystem so as to minimize thesystem losses

Human expertsheuristic rules

Based on Gauss-Sidel Method Radial distributionsystem

Along with Heuristicsknowledge of experiencedsystem is used to obtainoptimum saving.

Based on Heuristics.

Hsu et al. [24] Power Loss/ Shuntcapacitors

Investigatedoptimal capacitor dispatchingschedule based on the forecast ofhourly loads for the next day, such

Dynamicprogramming

Min f ¼ minðPr; lossÞ Radialdistribution system

Considered the variable loadnature for optimization.

Energy loss and voltageregulation not considered.

S.Kalam

be,G.A

gnihotri/Renew

ableand

SustainableEnergy

Review

s29

(2014)184

–200188

that the total feeder loss in a daycan be minimized.

Abdel-Salamet al. [25]


Depends on an identification of thesensitive nodes that have a verylarge impact on reducing the lossesin the distribution systems.

Analyticalapproach.

Sensitivity analysis Radial distributionsystem

The method is relatively fast,very effective and givesconsiderable saving both inenergy and in net dollar savingswhen the costs of the capacitorsand their installations are takeninto account.

Assumed linear systemload.

Shao et al.[26]


Developed heuristic based rules forloss minimization.

Heuristic rules Graph search technique Radial distributionsystem

Determine the proper size,location of capacitors and shuntcapacitor cost.

Based on heuristics

Laframboiseet al. [27]

Power loss/shuntcapacitors andvoltage regulators

Developed an expert system whichbecomes part of the SCADA systemfor short and long-term planning ofvoltage control and loss reductionin distribution systems.

Expert system Objective function is to minimize powerloss.


Determine the proper size,location of capacitors andvoltage regulators.

Do not get exact solutiononly approximate solutionis obtained

Cost saving.

Dan Jianget al. [28]

Provided a single comprehensivealgorithm for distribution systemswitch reconfiguration andcapacitor control.

Simulatedannealing- tooptimize theswitchconfiguration.

Minimization of power and feeder losses Feeders withuniformlydistributed loads

Benefits due to the optimalswitch configurationand capacitor control wereevaluated, both in termsof loss reduction and decreasedvoltage bandwidth

But it has to suffer fromdrawbacks of both methods.

DiscreteOptimizationAlgorithm-optimalcapacitor control

Cho et al. [29] Power and Energyloss/fixed andswitched shuntcapacitors

Developed a capacitor operationstrategy according to the reactiveload duration curve of the feeders.

Analyticalapproach.

Minimize the average power losses Real timedistribution system

determine the proper size,location and switching time ofcapacitors and shunt capacitorcost


Ramakrishnaet al. [30]

Power loss/shuntcapacitors.

Proposed a fuzzy inference systemto assist the operator for voltagecontrol and Power Lossminimization.

Fuzzy logic. Fuzzy expert system Feeders withuniformlydistributed loads.

Simplicity of approach andspeed makes it a viable optionfor online VAR control.

The reactive power controlproblem was decoupledfrom the voltage regulatorproblem.

Haque [31] Power loss/shuntcapacitors

Developed a method of minimizingthe loss associated with thereactive component of the branchcurrents

Dynamicprogramming

The optimal size of the capacitor of thecompensated nodes is then determined byoptimizing the loss saving equation withrespect to the capacitor currents.

Radialdistribution system

Simple and easy to use, quickresults

May cause erroneoussolution for real life system,

Carlisle et al.[32]

Power and energylosses/fixed andswitched shuntcapacitors

Used a Graph Search Algorithm forthe capacitor placement problemand determined a near-optimalsolution.

Graph searchalgorithm

max S¼ KPLPþKeLE�CC Radialdistribution system

Considered capacitor sizes asdiscrete variables and usesstandard sizes and exactcapacitor costs.

Only the balanced loadconditions are assumedwith the real power flow asconstant.

Deng et al.[33]

Real power losses/shunt capacitors.

Proposed an approach with fuzzyvariables which incorporates theload uncertainty inoptimizing capacitor on/off status

Fuzzy logic Objective function is to determine theoptimal combinations of adjustablecapacitor banks to compensate the reactiveload demand.


Uncertain load values wererepresented as trapezoidalfuzzy numbers via fuzzy sets.

Do not meet robustnessrequirements.

Fast speed of calculationsR. H. Liang2003 [34]

Real Power Losses/Fixed and switchedcapacitors.

Presented fuzzy-based reactivepower and voltage control in adistributed system

Fuzzy logic To find the combination of maintransformer LTC positions and capacitoron/off switching operations in a day, suchthat the voltage deviations at thesecondary bus of main transformer becomeas small as possible, while the reactivepower flows through the main transformerand Losses at feeders become as little aspossible.


Fast and accurate indetermining the size andlocation


de Souzaet al. [35]

Energy loss/fixed andswitched capacitors.

Proposed a Micro GeneticAlgorithm (MGA) in conjunctionwith Fuzzy Logic (FL) for solvingthe capacitor placement problem.

Micro GeneticAlgorithm inconjunction withFuzzy Logic

The objective function includes economicsavings obtained by an Energy LossReduction in contrast with acquisition and


Cost saving, capacitorplacement, loss minimization.

Complicated methodology.

S.Kalam

be,G.A

gnihotri/Renew

ableand

SustainableEnergy

Review

s29

(2014)184

–200189





installation costs of fixed and switchedcapacitors.

Carpinelliet al. [36]


Developed capacitor placementprocedure to minimize the PowerLosses

Dynamicprogramming

Loss minimization Radial distributionsystem

Capacitor placement,Considered the impact on theharmonic distortion of busvoltages.

Cannot be directly appliedgiven the size, complexityand the specificcharacteristics ofdistribution systems

Venkateshet al.[37]


Proposed a single dynamic datastructure for an EP Algorithm thathandles the problems of siting andsizing of newshunt capacitors simultaneouslywhile considering transformer taps,existing reactive-power sourcesand reconfiguration.

Evolutionaryprogramming.

Fuzzy based objective function. Radialdistribution system

Considered different load levelsand time durations

Energy loss is notconsidered

HaghifamandModares[38]

Energy and peak loadloss/fixed andswitchable capacitors

Developed an efficient GeneticAlgorithm with a new coding astwo rows of chromosome used foroptimization in simultaneousallocation of fixed andswitchable capacitors

Genetic Algorithm Capacitor cost was considered in the costfunction.


Time variation and uncertaintyof load were also involved inproblem formulation


Ulinuha et al.[39]

Energy and peak loadloss/fixed andswitched shuntcapacitors

Proposed EAs which were capableof optimizinglarge distribution systems withdifferent types of nonlinear loads.

Evolutionary-Based Algorithms(EAs)

Optimal scheduling of LTCs and switchedshunt capacitors for simultaneouslyminimizing Energy Loss and improving thevoltage was performed.


Considered Time variation anduncertainty of load.

Separate calculations arerequired for power andenergy loss.Improving the voltage profile

while taking harmonics intoaccount.

Jong-YoungPark et al.[40]

Energy Loss/Fixedand switched ShuntCapacitors

The effect of device lifetime on theswitching operation of devices suchas ULTC transformers or switchedcapacitors were added in theobjective function as a function ofthe annual number of operations ofthe device.

Genetic Algorithm PL ¼∑∑VmVqYmq cos ðθmθq�δmqÞ Large radialdistribution system

Expected device lifetimes areincluded in the formulation,


Eajal et al.[41]


Minimized the overall cost of thetotal Real Power Loss and that ofshunt capacitors while satisfyingoperating and power qualityconstraints

Particle SwarmOptimization

F ¼ KrPlossþ∑KCQC Small radialdistribution system

Developed Hybrid PSOAlgorithm to enhance theefficiency in achieving solution.

Not Suitable for Largesystem.

VahidFarahaniet al. [43]

Energy loss/ shuntcapacitors

presents a joint optimizationalgorithm for both conductorreplacement of overhead lines andcapacitorplacement to minimizeenergy losses in the presence ofharmonics throughout thedistribution system, such that bothindividual and total harmonicdistortion of bus voltages are keptwithin the acceptable levels.

Discrete geneticalgorithm

PBþ jQB ¼∑ðPLþ jQL�QC Þ 20-kV realisticoverheaddistributionnetwork in Sirjancity-center in Iran.

Application Capacitorplacement and conductorreplacement simultaneously.

Not suitable for largesystem.

S.Kalam

be,G.A

gnihotri/Renew

ableand

SustainableEnergy

Review

s29

(2014)184

–200190

Table 2Gradual evolution in network reconfiguration method.

Author Type ofloss

Reconfiguration approach Optimizationtechniques

Objective function/basic principle Network Achievements/merits De-merits

Merlin and Back[57]

Powerloss

Starts with a meshed network by initiallyclosing all switches in the network. Theswitches are then opened one at a time until anew radial network is reached.

Exhaustive searchtechniques.

Branch and bound method Radial topology,typical of urbandistribution systems.

Final network configuration isindependent of the initial status ofthe switches.

Method involvesapproximations.Network voltage anglesare assumed as negligible.Solution process leads to the

optimum or a near-optimumsolution.

Losses associated withline equipment are notconsidered

Shirmohammadiand Hong [58]

Resistivepowerloss

Starts by closing all the network switcheswhich are then opened one after another bydetermining the optimum flow pattern in thenetwork.

Heuristicapproach

min∑RjJj2 Realistic distributionnetworks

Avoids approximations, The environments underwhich the calculations aremade do not correspondto the actual operatingconditions.

Convergence to the optimum or anear-optimum solution,Final network configuration isindependent of the initial statusof the switches.

Civanlar et al.[59]

Powerloss

Determination of the loss change due to aswitch exchange.

Load flow basedapproach

Power loss¼ Ref2ð∑IiÞðEm�EnÞ2gþRloopj∑Iij2

Realistic distributionnetworks

Suggests a filtering mechanism foreliminating those switchingoptions which would not yield lossreduction

Multiple switchingoperations were beyondthe scope of his scheme

Sarfi et al. [60] Powerloss

Commences the search for a minimal lossnetwork configuration while the distributionsystem is partitioned into subsystems.

Dynamicprogramming.

Based on network partitioningtheory.

Meshed distributionnetworks

Overcomes the size restrictionsimposed by previously describedreconfiguration techniques.

Algorithm gives a nearoptimal solution butmethod does not work sowell in the case of loadvariation.

Ji-Yuan Fan et al.[61]

Powerloss

Determined the switch exchanges within aloop for minimum line losses, and utilized aheuristic scheme to develop the optimal switchplan with minimum switch operations toachieve transition from the initial configurationto optimal configuration.

Heuristicapproach

Single-loop optimization technique. Meshed distributionnetworks

Loss reduction. Simultaneous switchingof the feederreconfiguration is notconsidered.

Installation and switching costsReduction.Simple and effective scheme toefficiently determine the switchexchanges within a loop forminimum

Taleski et al. [62] Powerandenergyloss

Reconfiguration is performed by closing theopen tie switch that defines the loop, andopening the switch in the branch that producesmaximum energy loss saving.

Heuristicapproach

Branch exchange techniques Radial networkanalysis

Provides an alternative to thepower minimization methods foroperations and planning purposes

Complicatedmathematical techniquesdue to the combinatorialnature of the problem.Large computational timeis needed

Kashem et al.[63]

Realpowerloss.

The trained ANN models determine theoptimum switching status of the dynamicswitches along the feeders of the network,which thereby reduce real power loss bynetwork reconfiguration.

Artificial neuralnetwork (ANN)-approach

The objective is to determine theswitching status of the dynamicswitches in the feeders that providea minimum loss configuration tonetwork.

Small radialdistribution system

Proposed method is not dependenton the system size.

Applied on Dynamicswitches only.

The proposed ANN model canprovide a fast prediction of optimalor near-optimal systemconfiguration.

No assurance of optimumsolution

McDermott et al.[64]

Powerloss

Starts with all operable switches open, and ateach step, closes the switch that results in theleast increase in the objective function.

Heuristicnonlinearconstructivemethod

The objective function is defined asincremental losses divided byincremental load served

Radial network. Algorithm differs from most others,by constructing the system fromscratch, rather than performingswitch exchanges or sequentialswitch openings.

Fast but may trap intolocal minima.Do not guarantee to findout the global optimumin finite running time

Kashem et al.[65,66]

Powerloss

Starts with generation of limited number ofswitching combinations and the best switchingcombination is determined then an extensivesearch is employed to find out any otherswitching combination that may give rise tominimum loss compared to the loss obtained inthe first stage.

Dynamicprogramming.

Total loss¼∑ðrLðP2L þ jQ2

L Þ=V2L Þ Medium Radial

distribution system.Efficient for continuousreconfiguration for loss reduction.

Not suitable for largesystems

Requires very less computer timeas well as memory.

S.Kalam

be,G.A

gnihotri/Renew

ableand

SustainableEnergy

Review

s29

(2014)184

–200191


Author Type ofloss

Reconfiguration approach Optimizationtechniques

Objective function/basic principle Network Achievements/merits De-merits

Ramos et al. [67] Powerloss

It takes into account the status of all branchesin the network and concurrently introducedsome approximations in order to reducecomputation time.

Linearprogramming

min∑RjJj2 Small, medium andlarge distributionsystem.

Provides effective solution whichdoes not require load flowthroughout the process,

Solution depends onapproximations

Simple technique,Requires less computation time

Jeon et al. [68] Powerloss.

The proposed method augment the costfunction with the operation condition ofdistribution systems, improve the perturbationmechanism with system topology, and use thepolynomial time cooling schedule, which isbased on the statistical calculation during thesearch for solution

Simulatedannealing

Total loss¼∑ðrLðP2L þQ2

L Þ =V2L Þ Real distribution

system of the KoreaElectric PowerCorporation (KEPCO).

Well suited for a largecombinatorial optimizationproblem.

It often requires aelaborate coolingschedule and a specialstrategy to obtainsolutionApplicable only forsymmetrical systems andconstant loads.

It can avoid local minima byaccepting improvements in cost.

Augugliaro et al.[69]

Powerloss

Proposed control strategy of the open-closedstatus of the tie-switches.

Based on neuralnetworks and adeterministicAlgorithm.

The objective can be fulfilled byperforming a centralized controlstrategy in which the status of eachtie-switch depends on the currentloading condition of the network.

Radial distributionnetworks

The approach represents the firststep towards a completeautomation of distribution system.

The local control is lessaccurate than thecentralized control in theidentification of theoptimal solution.

Low cost and the guarantee, innormal operating conditions ofradial operation while supplyingcomplete loads.

Ching-Tzong Su[70]

Powerloss

Method is simple and based on stochasticsearches, in which function parameters areencoded as floating-point variables.

Mixed-IntegerHybridDifferentialEvolution(MIHDE) method.

The objective function isMin f ¼ minðPr; lossÞ

Practical distributionnetwork of the TaiwanPower Company (TPC).

Proposed method provides bothPower loss and enhanced voltagestability.

May give a optimumsolution butcomputationally veryexpensive and theoperations will be muchslower than heuristicmethods.

It determines system topologythat reduces the power lossaccording to a load pattern.

Hernan PrietoSchmidt et al.[71]

Powerloss.

The integer variables represent the state of theswitches and the continuous variablesrepresent the current flowing through thebranches. The standard Newton method is usedto compute branch currents at each stagewithin the integer search

Mixed-integernonlinearoptimization

Branch exchange type algorithm Real-size largedistribution system

Suitable for real size distributionsystems in an attempt to achievethe advantages of both networkrestructuring as well as capacitorplacement to minimize the losses.

Method does notguarantee the optimality.

The search procedure is fast.Golshan andArefifar [72]

Energyandpowerlosses

Method is based on Tabu Search to solve morecomprehensive distributed-generationplanning problem including simultaneousdistributed generation sourcesand network configuration planning.

Tabu Search Cost function¼ kpPoðzoÞþke∑τaPðzÞþkq∑jqj

Medium and largedistribution system.

Cost saving. It could not give anyguarantee for theconvergence property.

Technique allows the search togo beyond the local optimalpoints while still making the bestpossible move in each iteration.Makes extensive use of memorystructures.

Siti et al. [73] Powerloss.

Method uses the neural network in conjunctionwith a heuristic method which enablesdifferent reconfiguration switches to be turnedon/off and connected consumers to beswitched between different phases to keep thephases balanced.

Heuristic method,Neural Network isadopted for thenetworkreconfigurationproblem.

Total loss¼∑ðrLðP2L þ jQ2

L Þ=V2L Þ Low-voltage and

medium-voltage levelsofa distribution network

Phase balancing, Not suitable for largesystems.Loss minimization problem

Chung-Fu Chang[74]

Powerloss.

Method solves the optimal feederreconfiguration problem, the optimal capacitorplacement problem and the problem with acombination of the both.

Ant Colony SearchAlgorithm

The objective function is,Min f ¼ minðPr; lossÞ

Practical distributionnetwork of TaiwanPower Company (TPC)

Shows significantly enhancedoptimization capability.

Computational timerequired is more.

Queiroz and Lyra[75]

Powerloss andenergyloss.

Solved larger optimization problem whichwould be solving network reconfigurationproblem to suit each of the significant loadvariations.

Adaptive hybridgenetic algorithm

Total loss¼∑ðrLðP2L þQ2

L Þ=V2L Þ Small, medium and

large radialdistribution systems.

Method deals with energy flowsinstead of only instantaneouspower flows.

Considered only fixedtopologies for either peakor average loads.

S.Kalam

be,G.A

gnihotri/Renew

ableand

SustainableEnergy

Review

s29

(2014)184

–200192

performed. Again in 2011 [42] they proposed a Hybrid Genetic-Fuzzy Algorithm (GA-Fuzzy) for optimal volt/var/Total HarmonicDistortion (THD) control in distorted distribution systems servingnon-linear loads. Load interval division (over a 24 h period) andoptimal scheduling of LTC and switched shunt capacitors forsimultaneously minimizing Energy Losses and improving thepower quality were performed using Genetic Algorithms withFuzzy Reasoning and the non-linear load flow was solved by usinga Decoupled Harmonic Power Flow Algorithm.

In this way the Gradual evolution in the direction of makingloss minimization by Capacitor Placement superior has beenrevealed in Table 1.

Thus it can be seen that researchers have included varioustechniques to make the work of loss reduction by CapacitorPlacement more effective. Such as:

1. Mixed-Integer Programming [44].2. Linear Programming [45].3. Nonlinear Programming [7–9,12–14,46].4. Genetic Algorithms [47].5. Ant Colony Search [50].6. Artificial Neural Networks [52].7. Tabu Search [49,51,53].8. Particle Swarm Optimization [41,48,53–55].9. Fuzzy Set Theory [56].

10. Simulated Annealing [64,65].

It is found that in this method of loss reduction along with size,location and methodology, one has to also deal with cable size,number of capacitors etc. and utility has to bare extra cost ofcapacitors and for capacitor placement techniques. Apart fromthis, losses due to the in-phase component of current are notappreciably changed by the application of capacitors; it only dealswith reactive component of current.

4. Network reconfiguration

This section contains those publications which are related to lossreduction by using Network Reconfiguration which was found to begenerally applicable in low voltage Distribution networks. It is a veryimperative approach to save the electrical energy. Distributionsystems consist of groups of interconnected radial circuits. Theconfiguration may be varied via switching operations to transferloads among the feeders. Two types of switches are used in primarydistribution systems. They are normally closed switches (sectionaliz-ing switches) or normally open switches (tie switches). Both typesare designed for protection and configuration management. Networkreconfiguration is the process of changing the topology of distribu-tion systems by altering the open/closed status of these switches.Reconfiguration is applied for:

1. service restoration under faulty conditions,2. load balancing to

a. relieve overload on networks andb. improve voltage profile

3. planning outages for maintenance and4. loss minimization.

The switching operation is the basic control action in networkreconfiguration. A switching operation consists of closing theswitch in an opened branch and opening the switch in a closedone keeping the network configuration radial. However, sincethere are many candidate switching combinations in the system,the feeder reconfiguration is a complicated problem. The discretenature of the switch values makes it a discrete optimization

Prov

ides

ago

odtrad

e-off

betw

eenloss

reduc

tion

and

switch

ingop

erations.

Yuan

-Kan

gW

uet

al.[76

]Po

wer

loss

Itaimsat

achievingtheminim

um

pow

erloss

andincrem

entalload

balance

factor

ofradial

distribution

networks

withdistributed

generators.

AntColon

yAlgorithm

Power

loss

¼∑ðr L

ðP2 LþjQ

2 LÞ=V2 LÞ

Distribution

network

ofTP

C.

Avo

idspremature

conve

rgen

ce.

Becom

esco

mplic

ated

for

largesystem

sStrongglob

alsearch

ability.

Rao

etal.[77

]Po

wer

loss

Algorithm

usesastoc

hasticrandom

search

instea

dof

agrad

ientsearch

whichelim

inates

thenee

dforderivativeinform

ation.

Simulation

s

Harmon

ysearch

algo

rithm

Theob

jectivefunctionis,

Min

f¼

minðP

r;lossÞ

Med

ium

andlarge-

distribution

networks

Con

vergeto

optimal

solution

very

fast

even

foralargesystem

.Guaran

teed

aglob

ally

optimal

solution

but

nee

ded

anex

hau

stive

search

.Ta

her

Niknam

etal.[78

]Po

wer

loss

Afuzzysystem

basedon

someheu

risticsis

designed

toad

aptive

lyad

just

theparam

etersof

DPS

Oan

dBPS

Oduringtheop

timization

process

toim

prove

theov

erallperform

ance.

TheNelder–Mea

dis

asimplexsearch

method

that

has

been

widelyusedin

unco

nstrained

optimizationproblem

New

fuzzy

adap

tive

particle

Power

loss

¼Rloopj∑

I ij2

Med

ium

andlarge-

scaledistribution

networks.

Show

shighaccu

racy

and

conve

rgen

cerate.

Method

doe

snot

guaran

teetheop

timality.

Swarm

optimization

(NFA

PSO)

Indep

enden

ton

theinitialstatus

ofnetworksw

itch

es.

Algorithmsruntimeincrea

ses

when

thenumbe

rof

control

variab

lesincrea

se.

Akb

arBay

at[79]

Power

loss

Thealgo

rithm

starts

withex

pan

dingasu

bnetworkthroug

htracingmax

imum

busvo

ltag

ean

dperform

ingaseries

ofbran

chex

chan

geop

erationco

ncu

rren

tly

Heu

ristic

method

Branch

exch

ange

typealgo

rithm

Large-scale

distribution

networks

Con

vergen

cesp

eedis

accelerated.

Method

isap

plic

able

only

fortheradialnetworks.

Dep

endson

load

flow

.Su

itab

leforlargedistribution

system

s.


problem. The discrete nature of the switch values and radiallityconstraint prevent the use of classical optimization techniquesto solve the distribution reconfiguration problem. Therefore,most of the algorithms in the literature are based on heuristicsearch techniques by using either analytical or knowledge-basedapproaches. As per the method to handle the reconfigurationprocess one can categorize the algorithms in two types:

1. Branch Exchange Type—The system operates in a feasible radialconfiguration and the algorithm opens and closes candidateswitches in pairs.

2. Loop Cutting Type—The system is completely meshed and thealgorithm opens candidate switches to reach a feasible radialconfiguration.

Most of the literatures reveal the branch exchange type algo-rithms to minimize the losses by various methods. Loss reductionin the distribution system by reconfiguration was first proposed in1975 by Merlin and Back [57] by a branch-and-bound methodfor distribution systems which was later in 1989 modified byShirmohammadi and Hong [58]. Thereafter considerable resear-ches have been conducted in this field which is summarized inTable 2.

So after going through a deep bibliographical Survey it can beobserved that to make the Network Restructuring more effectivevarious techniques such as Particle Swarm Optimization [80–84],Constrained Decision Problems Approach (CDPs) [85], GeneticAlgorithm [86,87], Tabu Search [88,89], knowledge based expertsystem [90–92], Simulated Annealing [93–96], Single Loop Opti-mization [97], Ant Colony Search Method [98–100], HarmonySearch Algorithm [101,102], Mixed-Integer Convex Programming[103], Mixed Integer Linear Programming [104,105] have beenused by researchers.

The method of Network Restructuring was found to be acomplex decision-making process for dispatchers to follow, itoften requires extensive numerical computation and it also affectsthe coordination of the protective devices. In the aforementionedwork on feeder reconfiguration, it is usually assumed that theprotective devices are still properly coordinated when the feederconfigurations are changed by switch operations but actuallyprotective device planning and coordination are usually carriedout for a fixed configuration [106]. Frequent changes in configura-tion can trigger outages or cause transient problems. Apart fromthis although the methods mentioned above do not have goodconvergence property most of them are used due to less computa-tion time requirement for smaller systems. Despite the fact that allmethods used above guaranteed that the optimal solution will beachieved, they do provide high-quality precise solutions and maylack in accuracy. Maximum techniques suffer to a certain extent,from scaling the problem to a finite number of switches, typicallyfound on realistic distribution systems for larger systems, thecomputation time is prohibitively high and may not be suitable forreal-time operation.

5. DG allocation

This section contains those publications, which are related toloss reduction by using DG allocation which depends on theavailability of Distributed sources (if it is a Renewable Source). DGcan be defined as “The generation of electricity from facilities thatare sufficiently smaller than central generating plants so as to allowinterconnection at nearly any point in a power system.” [107]

Recently the penetration of Distributed Generators (DG's) intodistribution systems has been increasing rapidly in many parts of

the world. Integration of DG into an existing utility can result inseveral benefits such as:

� reduced environmental impacts,� increased overall energy efficiency,� relieved transmission and distribution congestion,� voltage support,� exploitation of the renewable resources, such as wind, solar,

hydro, biomass, geothermal and ocean energy and� line loss reduction.

Distributed generators are very much beneficial in reducing thelosses effectively compared to other methods of loss reduction.The main reasons for continuous growth in the penetration ofDG in power network are:

� the environmental concerns,� constraints on building new transmission and distribution lines,� technological advances in small generators,� power electronics and energy storage devices for transient

backup and� increasing public desire to promote “green” technologies based

on renewable-energy sources.

It is also observed that improper allocation or sizing of DG cancounter effect the system. The rise of DG is shifting the gridarchetype away from the traditional centralized systems that havelong been the basis for grid operation. As per the type, DG willhave both positive and negative impacts on distribution networksas mentioned in Table 3.

Thus the problem of DG planning has recently received muchattention by power system researchers so as to garner themaximum benefit from this upcoming power generation technol-ogy without violating the existing power system infrastructure.Various efforts taken by researchers in this method of lossminimization are summarized in Table 4.

In this technique of loss reduction by DG allocation alsoresearchers have included various advanced techniques suchas BEE Colony Algorithm [127], Mixed Integer Non-Linear Pro-gramming [128], Exhaustive Load Flow (ELF) Method [129],Particle Swarm Optimization [130–138], Fuzzy-Genetic algorithm[139,140], Hereford Ranch Algorithm [114,141], Ant colony search[76,142–145], Differential Evolution Approach [146], HeuristicCurve-Fitted Technique [147].

From the literature it may be considered that this promisingmethod of Loss Minimization is getting wide attention due to itsvery important benefit that it minimizes the network loss alongwith the provision of electrical energy supply to fulfill the crises ofdemand. So researchers are trying to investigate new techniquesto implement this method with maximum benefit.

Table 3Impacts of DG allocation on distribution system.

Type of impact Non-renewable DG Renewable DG

Environmental ✓

Voltage stability ✓ ✓

Reverse power flow ✓ ✓

Reliability ✓

Deferring upgrades of power system ✓ ✓

Reduction of electricity tariff ✓ ✓

Reduction of green house gases ✓

Cost saving ✓ ✓

DG allocation flexibility ✓

Loss minimization ✓ ✓


Table 4Gradual evolution in DG placement method.

Author Type of loss Prominent feature Optimizationapproaches/techniques


Rau and Wan[108]

Minimizationof losses, VArlosses, or

Proposed second order algorithm tocompute the amount of resources inselected nodes to achieve the desiredoptimizing objectives.

Analyticalapproach

Objective function¼ min ∑ power loss Meshed Distributionnetwork

The objective is reduction of networklosses, VAr losses, or loadings onselected lines.

Lengthy iterativeprocedure.No assurance ofconvergence.Loadings in

selectedlines.

Not suitable forlarger realisticsystem in terms ofcomputationalrequirements(memory andspeed).

Willis [109] Active andreactivepower losses

Analyzed the impact on feeder losses byDG installation by using 2/3 rule oftenused in capacitor placement ondistribution systems.

AnalyticalApproach basedon 2/3rd rule forLossminimization

Zero point analysis Radial Distributionnetwork

Analysis was undemanding and easy toapply

Approach may notbe functional incase of variableload conditions.

Mutale et al.[110]

Active andreactivepower losses

Proposed two new loss allocationschemes one based on the allocation ofmarginal losses and the other on theallocation of total losses.

The substitutionmethod

The objective of method is to derive arelationship such that losses can beexpressed directly in terms of injections.

Real network based265-node genericdistribution systemmodel.

Proposed loss allocation coefficients canbe positive or negative therefore canrecognize the presence of counter-flowsdue to the presence of DG.

Fixed componentsof loss are notconsidered.

Méndez et al.[111]


Provides annual losses variation due toDG

Analyticalapproach

Objective of the study was to quantifythe impact of that DG allocation byconsidering aspects such as DGpenetration level, DG dispersion, DGtechnologies, Generation mix by locatingoptimal DG penetration from lossespoint of view

Radial Distributionnetwork

Loss minimization. Complicated forlarge distributionsystems.

Yiming Maoet al. [112]


Provides switch placement scheme toimprove the system reliability and thelosses by DG placement

Graph-basedalgorithms,whichincorporatedirect loadcontrol, aredeveloped tolocate switches.

The switch placement problem isformulated as a non differentiable, multiobjective optimization problem.

394-bus radialdistribution systems

Customer priority is also considered inproblem.

Computationallyinexpensive interms of memoryand speed, robust,with moderatecomputerrequirements.

Applicable for unbalanced distributionnetworks with single or multipledistributed generators.

Caisheng Wanget al. [113]

Power losses Determine the optimal location to placea DG in radial as well as meshed systemsto minimize the power loss ofthe system considering total powerpenetration from DG units.

analyticalmethod

To find the optimal bus for placing DG ina networked system based on busadmittance matrix, generationinformation and load distribution of thesystem.

Radial as well asmeshed systems

Fast. Not suitable forlarge systems.No convergence problems involved.DG sizeoptimization is notconsidered.Ignored economicand geographicconsiderations.

Mithulananthanet al. [114]

Power losses Both the optimal size and location areobtained as outputs from the geneticalgorithm toolbox.

Geneticalgorithm

Objective function is based on the powerloss in line i� j Slossij ¼ SijþSji


Considered discrete as well ascontinuous parameters.

Lack of accuracywhen high-qualitysolution is required

Gandomkaret al. [115]

Power losses HRA uses sexual differentiation andselective breeding in choosing parentsfor genetic string.

Hereford ranchalgorithm

The objective function is,Min f ¼ minðPr; lossÞ

34 –bus radialdistribution system

Less number of iterations. Prematureconvergence,excessiveconvergence time

Robustness

Acharya et al.[116]


To calculate losses and to find out theoptimal location of DG

AnalyticalMethod

The objective function is based onsensitivity factors of active power

16, 33,69 bus radialdistribution system

No Convergence problem, fast incalculation of the losses

Do not meetrobustnessrequirements

S.Kalam

be,G.A

gnihotri/Renew

ableand

SustainableEnergy

Review

s29

(2014)184

–200195


Author Type of loss Prominent feature Optimizationapproaches/techniques


Víctor et al. [117] Power losses Provides an approach to compute annualenergy losses variations when differentpenetration and concentration levels ofDG are connected to a distributionnetwork.

Analyticalmethod.

DG impact on losses was measured asthe difference between losses in theconsidered scenario and losses in thebase case (without DG).


Evaluated the losses on an annual basis, No-load losses,nontechnical andcommercial lossesare not considered.

Results show that energy lossesvariation, as a function of the DGpenetration level, presents acharacteristic U-shape trajectory.

Fracisco et al.[118]


Defined new relevant Line Loss Index(LLI) and observed the effect withrespect to various DG penetration levels,DG dispersion and DG technologies aswell as different power factors.

Analyticalmethod.

Line Loss Index (LLI) Radial distributionsystem

He pointed out some results like (i) Forlow DG penetration level, lossesdecrease but for higher penetration levellosses marginally increase (ii) Reactivepower loss curve is convex and (iii)Minimum active power loss levels arereached if DG is sufficiently dispersed

Considered onlysteady statecapacity of DG

Hedayati [119] Power losses This method is based on the analysis ofpower flow continuation anddetermination of most sensitive buses tovoltage collapse.

Continuationmethod and adirect method

Q loss ¼ ðPload�PDG Þ2V2 þðQ load�QDGQDG Þ2

V2

� �Xline

34-bus test system Improvement of voltage profile. ComputationallydemandedReduction of power Losses,

Permit an increase in power transfercapacity, maximum loading, andvoltage stability margin.

Thukaram et al.[120]

Power andenergy losses

Identify the best place to install a newDG for a projected load increase

Analyticalapproach

Relative Electrical Distances (RED) 20 node, 66 kVsystem, a part ofKarnataka Transco.

This approach will help to identify thenew DG location(s), without thenecessity to conduct repeated powerflows.

Considered onlyreal power losses.

Tuba Gozel et al.[121]

Power losses Sensitivity factor is employed for thedetermination of the optimumsize andlocation of distributed generation so asto minimize total power losses by ananalytical method without use ofadmittance matrix, inverse ofadmittance matrix or Jacobian matrix.

AnalyticalApproach

Formulated loss sensitivity factor basedon the equivalent current injection


Method is in close agreement with theclassical grid search algorithm based onsuccessive load flows.

Computationallydemanded

Atwa et al. [122] Power andenergy losses

Determining the optimal fuel mix ofdifferent types of renewable DG units inorder to minimize the annual energylosses in the distribution system.

Probabilistic-based planningtechnique

The planning problem is formulated asmixed integer nonlinear programming(MINLP), with an objective function forminimizing the system's annual energylosses

Typical ruraldistribution systemwith differentscenarios, includingall possiblecombinations of therenewable DG units.

Guarantees the optimum allocation ofthe renewable DG units for all possibleoperating conditions.

Lack of accuracywhen high qualitysolution is required.

The objective function can beexpanded to accommodate additionalterms, such as capital and runningcosts of the DG units.

Karar Mahmoudet al. [123]

Power losses Best location and size for systemstability improvements.

Analyticalapproach

Stability index criterion. 90 bus test system Proved that calculating minimumsystem losses is not necessary to achievecoherence improvement for the voltagestability problem.

Not suitable forcomplicatedsituations.

The objective function is formulatedwith full consideration of both qualityand inequality constraints.

Akorede et al.[124]

Power loss These objective functions were firstfuzzified to evaluate their imprecisenature and then transformed into asingle multi-objective function, before itwas finally solved

Genetic Algorithm(GA).

Objective functions considered in thestudy were maximization of the systemloading margin as well as the DISCO'sprofit.

Meshed powersystems

Determining the optimal capacity andlocation of DG units

GA is known to bevery good at findinggood globalsolutions but not soefficient indetermining theabsolute optimum

Hung et al.[125,126]

Active andreactivepower loss.

Single and multiple DG allocation Analytical Method Exact Loss Formula 16, 33,69 bus radialdistribution system

No Convergence problem, fast incalculation of the losses

Do not meetrobustnessrequirements

S.Kalam

be,G.A

gnihotri/Renew

ableand

SustainableEnergy

Review

s29

(2014)184

–200196

6. Comparative analysis

The prolonged study of all above mentioned Loss MinimizationMethods reveal the comparative analysis given in Table 5 and thevarious impacts are summarized in Table 6.

7. Conclusion

This paper provides a bibliographical survey of the state-of-the-art of three basic loss minimization techniques such as (i) FeederReconfiguration, (ii) Capacitor Placement and (iii) DG Allocationused in distribution system, with a profound review of pertinentbackground, realistic requirements, the current status and compe-tent techniques. It is based on many research articles publishedfrom the past 3 decades so that the detailed progressive researchin this field can be summarized. The citations listed in thisbibliography provide a representative sample of contemporarytechnical assessment pertaining to the improvement in efficiencyof the distribution system by achieving power saving through lossminimization. Periodic Bibliographic updates on this topic will beuseful as the power system continues to evolve.

Each method mentioned in this survey has been utilized tosolve problem with various and limited objectives and constraints.Distribution system Loss Minimization techniques discussed inthis literature as summarized in Tables 1, 2, 4 and 5 are leading tothe following conclusions:

(i)Capacitor Placement Method which is especially applicable inHigh Voltage networks is most reliable and simple to imple-ment. However apart from loss minimization, shows limitedadvantages as compared to DG allocation Method.

(ii) Network Reconfiguration which is found to be applicable inlow voltage distribution systems is most economic butrequires complicated controlling techniques. The reconfigura-tion involves many candidate-switching combinations whichmakes it a complicated combinatorial, non-differentiable,constrained optimization problem. It often requires extensivenumerical computation and it also affects the coordination ofthe protective devices. Method is efficient but shows limitedpayback.

(iii) DG allocation is more focused to achieve interconnectionwhen small generators exist (for instance, when isolated windfarms or small photovoltaic plants enter the distributionnetwork) is most efficient and Shows perceptible potentialtowards loss minimization as along with loss minimization itserves many other benefits as explained in Section 5. Howeverit faces problems towards implementation and installation.Due to the lake of effective benefit harnessing techniques thismethod is found to be least reliable. However researchers aretrying hard to locate superior techniques to exploit allpossible benefits of DG Allocation.

For the Distribution system operators the main hurdles are theimplementation and reliability of the loss minimization technique.However with the view of environmental concerns and energycrises DG allocation is found to the best solution for the Distribu-tion Loss Reduction.

References

[1] Ochoa LF, Harrison GP. Minimizing energy losses: optimal accommodationand smart operation of renewable distributed generation. IEEE Transactionson Power Systems 2011;26(1):198–205.

[2] Bortignon GA, El-Hawary. M. A review of capacitor placement techniques forloss reduction in primary feeders on distribution systems. In: Proceedings ofthe Canadian conference on electrical and computer engineering, vol. 2:Montreal, Quebec; September 1995, p. 684–7.

[3] Neagle NM, Loss Samson. Reduction from capacitors installed on primaryfeeders. AIEE Transaction 1956;PAS-75(III):950–9.

[4] Cook RF. Optimizing the application of shunt capacitors for loads in certainareas reactive-volt-ampere control and loss reduction. AIEE Transactions1961;PAS-80:430–44.

Table 5Summarized comparison of three loss minimization methods.

LossMinimizationMethod

Benefits Drawbacks

Capacitorplacement

Line loss reduction. Losses due to only out of phase components of currents i.e. reactive currents can bereduced.Control of power flow.

Improvement of stability. Along with size, location and methodology one has to also deal with cable size, numberof capacitors etc. Utility has to bare extra cost of capacitors and for capacitor placementMethods.

Voltage profile management.Power factor correction.Source of reactive power Loss minimization.

Networkrestructuring

Line loss reduction. Complex decision-making process.Load balancing.Provides protection to isolate a fault. Requires extensive numerical computation Affects the coordination of the protective

devices.May not be suitable for real-time operation.

DG allocation Line loss reduction. Improper allocation or sizing of DG can counter effect the system.Source of Energy.Reduced environmental impacts.Increased overall energy efficiency. DG planning.Relieved transmission and distribution congestion. Causes reverse power flow in Distribution System which is traditionally designed for

unidirectional power flow.Voltage support.Deferred investments to upgrade existing generation,transmission and distribution systems.

Table 6Impacts of three loss minimization methods.

Impacts of methods Capacitorplacement

Networkrestructuring

DGallocation

Loss minimization ✓ ✓ ✓

Reliability ✓ ✓

Cost saving ✓ ✓

Voltage support ✓ ✓ ✓

Demand sidemanagement

✓ ✓

Affects protectionsystem

✓ ✓

Green power ✓

Load balancing ✓ ✓


http://refhub.elsevier.com/S1364-0321(13)00615-1/sbref1



http://refhub.elsevier.com/S1364-0321(13)00615-1/othref0005









[5] Schmill JV. Optimum size and location of shunt capacitors on distributionfeeders. IEEE Transactions on Power Apparatus and Systems 1965;PAS-84(9):825–32.

[6] Duran H. Optimum number, location, and size of shunt capacitors in radialdistribution feeders—a dynamic programming approach. IEEE Transactionson Power Apparatus and Systems 1968;PAS-87(9):1769–74.

[7] Grainger JJ, Lee SH. Optimum size and location of shunt capacitors forreduction of losses on distribution feeders. IEEE Transaction on PowerApparatus and Systems 1981;PAS-100(3):1105–18.

[8] Lee SH, Grainger JJ. Optimum placement of fixed and switched capacitors onprimary distribution feeders. IEEE Transactions Power Apparatus andSystems 1981;100:345–51.

[9] Grainger JJ, Lee SH. Capacity release by shunt capacitor placement ondistribution feeders: a new voltage-dependent model. IEEE Transactions onPower Apparatus and Systems 1982;PAS-101(5):1236–44.

[10] Grainger JJ, Civanlar S, Lee SH. Optimal design and control scheme forcontinuous capacitive compensation of distribution feeders. IEEE Transac-tions on Power Apparatus and Systems 1983;PAS-102(10):3271–8.

[11] Ponnavaikko M, Prakasa Rao KR. Optimal choice of fixed and switched shuntcapacitors on radial distributors by the method of local variations. IEEETransactions on Power Apparatus and Systems 1983;PAS-102(6):1607–15.

[12] Kaplan M. Optimization of number, location, size, control type, and controlsetting of shunt capacitors on radial distribution feeders. IEEE Transactionson Power Apparatus and Systems 1984;PAS-103(9):2659–65.

[13] Grainger JJ, Civanlar S. Volt/Var control on distribution systems with lateralbranches using shunt capacitors and voltage regulators. Part I: the overallproblem. IEEE Transactions on Power Apparatus and Systems 1985;PAS-104(11):3278–83.

[14] Civanlar S, Grainger JJ. Volt/Var control on distribution systems with lateralbranches using shunt capacitors and voltage regulators. Part II: the solutionmethod. IEEE Transactions on Power Apparatus and Systems 1985;PAS-104(11):3284–90.

[15] Civanlar S, Grainger JJ. Volt/Var control on distribution systems with lateralbranches using shunt capacitors and voltage regulators. Part III: the numer-ical results. IEEE Transactions on Power Apparatus and Systems 1985;PAS-104(11):3291–7.

[16] Salama MMA, Chikhani AY, Hackam R. Control of reactive power indistribution systems with an end load and fixed load condition. IEEETransactions on Power Apparatus and Systems 1985;PAS-104(10):2779–88.

[17] Bishop MT, Lee RE. Distribution system line loss reduction through enhancedcapacitor location techniques. IEEE Transactions on Power Delivery 1986;1(2):190–7.

[18] Ertem Suat, Tudor RJames. Optimal shunt capacitor allocation by nonlinearprogramming. IEEE Transactions on Power Delivery 1987;PWRD-2(4):1310–6.

[19] Lee RE. A method and its application to evaluate automated distributioncontrol. IEEE Transactions Power Delivery 1988;3(3):1310–6.

[20] Baran ME, Wu FF. Optimal capacitor placement on radial distributionsystems. IEEE Transactions on Power Delivery 1989;4(1):725–34.

[21] Ertem S, Baghzouz Y. Optimal shunt capacitor sizing for distribution systemswith multiple nonlinear loads. In: Proceedings of industrial and commercialpower systems technical conference; May1990, p. 51–6.

[22] Baldick R, Berkeley CA. Approximation formulas for the distribution system:the loss function and voltage dependence. IEEE Transactions on PowerDelivery 1991;6(1):252–9.

[23] Salama MMA, Chikhani. AY. An expert system for reactive power control of adistribution system. Part 1: System configuration. IEEE Transactions onPower Delivery 1992;7(2):940–5.

[24] Hsu YY, Kuo HC. Dispatch of capacitors on distribution system using dynamicprogramming. In: IEE proceedings generation, transmission and distribution,vol. 140(6); November 1993, p. 433–8.

[25] Abdel-Salam TS, Chikhani AY, Hackam. R. A new technique for loss reductionusing compensating capacitors applied to distribution systems with varyingload condition. IEEE Transactions on Power Delivery 1994;9(2):819–27.

[26] Shao J, Rao NDM, Zhang Y. A capacitor placement expert system. Interna-tional Journal of Engineering Intelligent Systems for Electrical Engineeringand Communications 1994;2(2):105–14.

[27] Laframboise JRPR, Ferland G, Salama MMA, Chikhani AY. An expert systemfor reactive power control of a distribution system. Part 2: system imple-mentation. IEEE Transactions on Power System 1995;10(3):1433–41.

[28] Jiang Dan, Baldick R. Optimal electric distribution system switch reconfi-guration and capacitor control. IEEE Transactions on Power Systems 1996;11(2):890–7.

[29] Cho MY, Chen YW. Fixed/switched type shunt capacitor planning ofdistribution systems by considering customer load patterns and simplifiedfeeder model. IEE Proceedings Generation, Transmission and Distribution1997;144(6):533–40.

[30] Ramakrishna G, Rao ND. Fuzzy inference system to assist the operator inreactive power control in distribution systems. IEE Proceedings Generation,Transmission and Distribution 1998;145(2):133–8.

[31] Haque MH. Capacitor placement in radial distribution systems for lossreduction. IEE Proceedings of Generation, Transmission and Distribution1999;146(5):501–5.

[32] Carlisle JC, El-Keib. AA. A graph search algorithm for optimal placement offixed and switched capacitors on radial distribution systems. IEEE Transac-tions on Power Delivery 2000;15(1):423–8.

[33] Deng Youman, Ren Xiaojuan. Fuzzy modeling of capacitor switching forradial distribution systems. In: Proceedings of IEEE power engineeringsociety winter meeting; 2001.

[34] Liang RH, Wang YS. Fuzzy-based reactive power and voltage control in adistribution system. IEEE Transactions on Power Delivery 2003;18(2):610–8.

[35] de Souza BA, Alves HdN, Ferreira HA. Microgenetic algorithms and fuzzylogic applied to the optimal placement of capacitor banks in distributionnetworks. IEEE Transactions on Power Systems 2004;19(2):942–7.

[36] Carpinelli G, Varilone P, Di Vito V, Abur A. Capacitor placement in three-phase distribution systems with nonlinear and unbalanced loads. IEEProceedings on Generation, Transmission and Distribution 2005;152(1):47–52.

[37] Venkatesh B, Ranjan R. Fuzzy EP algorithm and dynamic data structure foroptimal capacitor allocation in radial distribution systems. IEE ProceedingsGeneration, Transmission and Distribution 2006;153(1):80–8.

[38] Haghifam MR, Modares Tarbiat. Genetic algorithm-based approach for fixedand switchable capacitors placement in distribution systems with uncer-tainty and time varying loads. IET Generation, Transmission and Distribution2007;1(2):244–52.

[39] Ulinuha A, Surakarta Muhammadiyah, Masoum Surakarta. Optimal schedul-ing of LTC and shunt capacitors in large distorted distribution systems usingevolutionary-based algorithms. IEEE Transactions on Power Delivery2008;23(1):434–41.

[40] Park Jong-Young, Sohn Jin-Man, Park Jong-Keun. Optimal capacitor allocationin a distribution system considering operation costs. IEEE Transactions onPower Systems 2009;24(1):462–8.

[41] Eajal AA, El-Hawary ME. Optimal capacitor placement and sizing in un-balanced distribution systems with harmonics consideration using particleswarm optimization. IEEE Transactions on Power Delivery 2010;25(3):1734–41.

[42] Ulinuha A, Masoum Muhammadiyah Surakarta. Hybrid genetic-fuzzy algo-rithm for volt/var/total harmonic distortion control of distribution systemswith high penetration of non-linear loads. IET Generation, Transmission andDistribution 2011;5(4):425–39.

[43] Farahani Vahid, Sadeghi Seyed Hossein Hesamedin, Abyaneh HosseinAskarian, Agah Seyed Mohammad Mousavi. Energy loss reduction byconductor replacement and capacitor placement in distribution systems.IEEE Transactions on Power Systems 2013;28(3):2077–85.

[44] Adams RN, Laughton MA. Optimal planning of power networks using mixed-integer programming. Part 1: Static and time-phased network synthesis.Proceedings of the Institution of Electrical Engineers 1974;121(2):139–47.

[45] Bacher R, Glavitsch H. Loss reduction by network switching. IEEE Transac-tions on Power Systems 1988;3(2):447–54.

[46] Delfanti M, Granelli GP, Marannino P, Montagna M. Optimal capacitorplacement using deterministic and genetic algorithms. IEEE Transactionson Power System 2000:1041–615 2000:1041–6.

[47] Levitin G, Kalyuzhny A, Shenkman A, Chertkov M. Optimal capacitor allocation indistribution systems using a genetic algorithm and a fast energy loss computationtechnique. IEEE Transactions on Power Delivery 2000:623–815 2000:623–8.

[48] Prakash K, Sydulu M. Particle swarm optimization based capacitor place-ment on radial distribution systems. In: Proceedings of IEEE conference onpower engineering society general meeting; June2007, p. 24–8.

[49] Yang HT, Huang YC, Huang CL. Solution to capacitor placement in a radialdistribution system using tabu search method. Proceedings of InternationalConference on Energy Management and Power Delivery 1995;1:388–93.

[50] Chiou Ji-Pyng, Chang Chung-Fu, Su Ching-Tzong. Ant direction hybriddifferential evolution for solving large capacitor placement problems. IEEETransactions on Power Systems 2004;19(4):1794–800.

[51] Mori H, Ogita Y. Parallel tabu search for capacitor placement in radialdistribution systems. Proceedings of IEEE Power Engineering Society GeneralMeetings 2000;4:2334–9.

[52] Dash PK, Saha S, Nanda PK. Artificial neural net approach for capacitorplacement in power system. In: Proceeding of 1st international forum onapplication of neutral networks to power systems; 1991, p. 247–50.

[53] Huang Yann-Chang, Yang Hong-Tzer, Huang Ching-Lien. Solving the capacitorplacement problem in a radial distribution system using Tabu Search approach.IEEE Transactions on Power Systems 1996;11(4):1868–73.

[54] Ziari I, Ledwich G, Ghosh A. Optimal voltage support mechanism indistribution networks. IET Generation, Transmission and Distribution2011;5(1):127–35.

[55] Haidar AMA, Aziz HA, Noman KMG, Al-Jawfi RA. An intelligent placement ofdistributed capacitance based on loss minimization. In: IEEE conference onevolutionary computation (CEC); 5–8 June 2011, p. 1–5.

[56] Chin HC, Lin WM. Capacitor placements for distribution systems with fuzzyalgorithm. In: IEEE region 10's ninth annual international conference, vol. 2;1994, p. 1025–9.

[57] Merlin A, Bach A. Search for a minimal-loss operating spanning treeconfiguration in an urban power distribution system. In: Proceedings ofpower system computation conference (PSCC); 1975.

[58] Shirmohammadi D, Hong HW. Reconfiguration of electric distributionnetworks for resistive losses reduction. IEEE Transactions on Power Delivery1989;4:1392–488.

[59] Civanlar S, Grainger JJ, Lee SH. Distribution feeder reconfiguration for lossreduction. IEEE Transactions on Power Delivery. 1988;3:1217–23.












































































































































































[60] Sarfi RJ, Salama MMA, Chikhani AY. Distribution system reconfiguration forloss reduction: an algorithm based on partitioning theory. IEEE Transactionson Power Systems 1996;11:504–10.

[61] Fan Ji-Yuan, Zhang Lan, McDonald DJohn. Distribution networkreconfiguration: single loop optimization. IEEE Transactions on PowerSystems 1996;11(3):1643–7.

[62] Taleski R, Rajicic D. Distribution network reconfiguration for energy lossreduction. IEEE Transactions on Power Systems 1997;12(1):398–406.

[63] Kashem MA, Jlasmon GB, Mohamed A, Moghavvemi M. Artificial neuralnetwork approach to network reconfiguration for loss minimization indistribution networks. Electrical Power and Energy Systems 1998;20:247–58.

[64] McDermott TE, Drezga I, Broadwater RP. A heuristic nonlinear constructivemethod for distribution system reconfiguration. IEEE Transactions on PowerSystems 1999;14(2):478–83.

[65] Kashem MA, Ganapathy V, Jasmon GB, Buhari MI. A novel method for lossminimization in distribution networks. In: Proceedings of the internationalconference on electric utility deregulation, restructuring and power tech-nologies; April 2000.

[66] Kashem MA, Ganapathy V, Jasmon GB. Network reconfiguration for enhance-ment of voltage stability in distribution networks. IEE Proceedings Genera-tion, Transmission and Distribution 2000;147(3):171–5.

[67] Ramos ER, Martinez-Ramos JL, Exposito AG, Salado AJU. Optimal reconfi-guration of distribution networks for power loss reduction. In: IEEE portopower tech proceedings; 2001.

[68] Jeon Young-Jae, Kim Jae-Chul, Kim Jin-O, Shin Joong-Rin, Lee Kwang Y. Anefficient simulated annealing algorithm for network reconfiguration in large-scale distribution systems. IEEE Transactions on Power Delivery 2002;17(4):1070–8.

[69] Augugliaro A, Dusonchet L, Ippolito MG, Sanseverino ER. Minimum lossesreconfiguration of MV distribution networks through local control of tie-switches. IEEE Transactions on Power Delivery 2003;18(3):762–71.

[70] Su Ching-Tzong, Lee Chu-Sheng. Network reconfiguration of distributionsystems using improved mixed-integer hybrid differential evolution. IEEETransactions on Power Delivery 2003;18(3):1022–7.

[71] Prieto Schmidt Hernan, Ida Nathan, Kagan Nelson, Guaraldo Joao Carlos. Fastreconfiguration of distribution systems considering loss minimization. IEEETransactions on Power Systems 2005;20(3):1311–9.

[72] Golshan MEH, Arefifar SA. Distributed generation, reactive sources andnetwork-configuration planning for power and energy-loss reduction.IEE Proceedings Generation, Transmission and Distribution 2006;153(2):127–36.

[73] Siti MW, Nicolae DV, Jimoh AA, Ukil A. Reconfiguration and load balancing inthe lv and mv distribution networks for optimal performance. IEEE Transac-tions on Power Delivery 2007;22(4):2534–40.

[74] Chang Chung-Fu. Reconfiguration and capacitor placement for loss reductionof distribution systems by ant colony search algorithm. IEEE Transactions onPower Systems 2008;23(4):1747–55.

[75] Queiroz LMO, Lyra C. Adaptive hybrid genetic algorithm for technical lossreduction in distribution networks under variable demands. IEEE Transac-tions on Power Systems 2009;24(1):445–53.

[76] Wu Yuan-Kang, Lee Ching-Yin, Liu Le-Chang, Tsai Shao-Hong. Study ofreconfiguration for the distribution system with distributed generators.IEEE Transactions on Power Delivery 2010;25(3):1678–85.

[77] Rao RS, Narasimham SVL, Raju MR, Rao AS. Optimal network reconfigurationof large-scale distribution system using harmony search algorithm. IEEETransactions on Power Systems 2011;26(3):1080–8.

[78] Niknam Taher, Azadfarsani Ehsan, Jabbari Masoud. A new hybrid evolu-tionary algorithm based on new fuzzy adaptive PSO and NM algorithms forDistribution Feeder Reconfiguration. Energy Conversion and Management2012;54:7–16.

[79] Bayat Akbar. Uniform voltage distribution based constructive algorithm foroptimal reconfiguration of electric distribution networks. Electric PowerSystems Research 2013;104:146–55.

[80] Olamaei J, Gharehpetian G, Niknam T. An approach based on particle swarmoptimization for distribution feeder reconfiguration considering distributedgenerators. In: IEEE power systems conference: advanced metering, protec-tion, control, communication, and distributed resources: Clemson UniversityClemson, South Carolina; 2007.

[81] Wu Wu-Chang, Tsai Men-Shen, Hsu Fu-Yuan. A new binary coding particleswarm optimization for feeder reconfiguration. ISAP 2007. In: Internationalconference on intelligent systems applications to power systems; 5–8November 2007.

[82] Xiong Ning, Cheng Haozhong, Yao Liangzhong, Bazargan M.. Switch groupbased Tabu search algorithm for distribution network reconfiguration. In:Proceedings of third international conference on electric utility deregulationand restructuring and power technologies, DRPT; 6–9 April 2008.

[83] Yang, Hong-Tzer, Tzeng, Yi-Te, Tsai, Men-Shen. Loss-minimized distributionsystem reconfiguration by using improved multi-agent based particle swarmoptimization. In: Asia-pacific power and energy engineering conference(APPEEC); 28–31 March 2010.

[84] Hooshmand RA, Soltani S. Fuzzy optimal phase balancing of radial andmeshed distribution networks using BF-PSO algorithm. IEEE Transactions onPower Systems 2012;27(1):47–57.

[85] de-Oliveira CCB, Kagan N. Heuristic model for the selection and allocation ofshunt capacitors and voltage regulators in electrical power distribution

systems. In: Proceedings of international conference on intelligent systemsapplication to power system: Rio de Janeiro, Brazil; 1999.

[86] Prasad K, Ranjan R, Sahoo NC, Chaturvedi A. Optimal reconfiguration ofradial distribution systems using a fuzzy mutated genetic algorithm. IEEETransactions on Power Delivery 2005;20(2):1211–3.

[87] Naka S, Genji T, Yura T, Fukuyama. Y. A hybrid particle swarm optimizationfor distribution state estimation. IEEE Transactions on Power Systems2003;18(1):60–8.

[88] Zhang D, Fu Z, Zhang L. An improved TS algorithm for loss minimumreconfiguration in large-scale distribution systems. Electric Power SystemsResearch 2007;77(5):685–94.

[89] Hayashi Y, Matsuki J. Loss minimum configuration of distribution systemconsidering N-1 security of dispersed generators. IEEE Transactions onPower Systems 2004;19(1):636–42.

[90] Liu CC, Lec SJ, Venkata SS. An expert system operational aid for restorationand loss reduction of distribution systems. IEEE Trans. on Power Delivery1988;3(2):619–29.

[91] Liu CC, Lee SJ, Veu K. Loss minimization of distribution feeders: optimalityand algorithms. IEEE Transactions on Power Delivery 1989;4(2):1281–9.

[92] Taylor T, Lubkeman D. Implementation of heuristic strategies for distributionfeeder reconfiguration. IEEE Transactions on Power Delivery 1990;5(1):239–46.

[93] Cheng HC, Kou CC. Network reconfiguration in distribution systems usingsimulated annealing. Electric Power System Research 1994;29(3):227–38.

[94] Matos MA, Melo P. Multiobjective reconfiguration for loss reduction andservice restoration using simulated annealing. In: International conferenceon electric power engineering, power tech budapest 99; 1999.

[95] Sugang Xu, Sezaki K, Tanaka Y. A two-stage simulated annealing logicaltopology reconfiguration in IP over WDM networks. In: Proceedings of the11th international conference on telecommunications network strategy andplanning symposium, Networks; 13–16 June 2004.

[96] Parada V, Ferland JA, Arias M, Daniels K. Optimization of electrical distribu-tion feeders using simulated annealing. IEEE Transactions on Power Delivery2004;19(3):1134–40.

[97] Fan JY, Zhang L, McDonald JD. Distribution network reconfiguration: singleloop optimization. IEEE Transactions on Power Systems 1996;11(3):1643–7.

[98] Su CT, Chang CF, Chiou JP. Distribution network reconfiguration for lossreduction by ant colony search algorithm. Electric Power Systems Research2005;75(2–3):190–9.

[99] Abdelaziz AY, Osama RA, El-Khodary SM. Reconfiguration of distributionsystems for loss reduction using the hyper-cube ant colony optimisationalgorithm. IET Generation, Transmission and Distribution 2012;6(2).

[100] Falaghi H, Haghifam M-R, Singh C. Ant colony optimization-based method forplacement of sectionalizing switches in distribution networks using a fuzzymultiobjective approach. IEEE Transactions on Power Delivery 2009;24(1).

[101] Geem ZW, Kim JH, Loganathan GV. A new heuristic optimization algorithm.Harmony Search Simulation 2001;76(2):6 0–8.

[102] Mahdavi M, Fesanghary M, Damangir E. An improved harmony searchalgorithm for solving optimization problems. Applied Mathematics andComputation 2007;188(2):1567–79.

[103] Jabr RA, Singh R, Pal BC. Minimum loss network reconfiguration using mixed-integer convex programming. IEEE Transactions on Power Systems 2012;27(2).

[104] Baran ME, Felix FWu. Network reconfiguration in distribution systems forloss reduction and load balancing. IEEE Power Engineering Review 1989;9(4):101–2.

[105] Ajaja A, Galiana FD. Distribution network reconfiguration for loss reductionusing MILP. In: Proceedings of IEEE PES innovative smart grid technologies(ISGT); 16–20 January 2012.

[106] Hong HW, Sun CT, Mesa VM, Ng S. Protective device coordination expertsystem. IEEE Transaction on Power Delivery 1991;6(1):359–65.

[107] Ackermann T, Anderson G, Soder L. Distributed generation: a definition.Electric Power System Research 2001;57:195–204.

[108] Rau NS, Wan YH. Optimum location of resources in distributed planning. IEEETransactions on Power Systems 1994;9:2014–20.

[109] Willis HL. Analytical methods and rules of thumb for modelingDG-distribution interaction. In: Proceedings of IEEE power engineeringsociety summer meeting, vol. 3; 2000, p. 1643–4.

[110] Mutale J, Strbac G, Curcic S, Jenkins N. Allocation of losses in distributionsystems with embedded generation. In: Proceedings of institute of electricalengineers—generation, transmission and distribution, vol. 146(1); January2000, p. 7–14.

[111] Méndez VH, Rivier J,de la Fuente JI, Gómez T, Arceluz J, Marín J. Impact ofdistributed generation on distribution losses. In: Proceedings of med power:Athens, Greece; 2002.

[112] Mao Yiming, Miu Karen N. Switch placement to improve system reliability forradial distribution systems with distributed generation. IEEE Transactions onPower Systems 2003;18(4):1346–52.

[113] Wang Caisheng, Nehrir MH. Analytical approaches for optimal placement ofdistributed generation sources in power systems. IEEE Transactions on PowerSystems 2004;19(4):2068–76.

[114] Mithulananthan N, Oo Than, Van Phu Le. Distributed generator placement inpower distribution system using genetic algorithm to reduce losses.Thammasat International Journal of Science and Technology 2004;9(3):55–62.













































































































































































[115] Gadomkar M, Vakillan , M, Ehsan M. Optimal distributed generation alloca-tion in distributed network using Hereford Ranch algorithm. In: Proceedingsof the 8th international conference on electrical machines and systems—ICEMS; 2005, p. 916–8.

[116] Acharya N, Mahant P, Mitulananthan N. An analytical approach for DGallocation in primary distribution network. International Journal of ElectricalPower and Energy Systems 2006;28(10):669–78.

[117] Méndez Quezada Víctor H, Rivier Abbad Juan, San Román Tomás Gómez.Assessment of energy distribution losses for increasing penetration ofdistributed generation. IEEE Transactions on Power Systems 2006;21(2):533–40.

[118] Fracisco González-Longatt M. Impact of distributed generation over powerlosses on distribution system. In: Proceedings of 9th international conferenceon electrical power quality and utilization: Barcelona; 9–11 October 2007.

[119] Hedayati H, Nabaviniak SA, Akbarimajd A. A method for placement of DGunits in distribution networks. IEEE Transactions on Power Delivery2008;23:2673–81.

[120] Thukaram D, Vyjayanthi C, Surendra S. Optimal placement of distributedgeneration for a projected load increase using relative electrical distanceapproach. In: Proceedings of third international conference on powersystems: Kharagpur, India; December 2009, p. 27–9.

[121] Tuba Gozel, Hakan HTuba Gozel M, Hakan Hocaoglu M. An analytical methodfor the sizing and siting of distributed generators in radial systems. ElectricPower Systems Research 2009;79:912–8.

[122] Atwa YM, El-Saadany EF, Salama MMA, Seethapathy R. Optimal renewableresources mix for distribution system energy loss minimization. IEEETransactions on Power Systems 2010;25(1):360–70.

[123] Mahmoud Karar, Abdel-Akher Mamdouh, Ahmed Abdel-Fatah A. Sizing andlocating distributed generations for losses minimization and voltage stabilityimprovement. In: IEEE internal conference on power and energy (PECon2010);2010, p. 600–4.

[124] Akorede MF, Hizam H, Aris I, Ab Kadir MZA. Effective method for optimalallocation of distributed generation units in meshed electric power systems.IET Generation, Transmission and Distribution 2011;5(2):276–87.

[125] Hung DQ, Mithulananthan N, Bansal RC. Analytical expressions for DGallocation in primary distribution networks. IEEE Transactions on EnergyConversion 2010;3(3):814–20.

[126] Hung DQ, Mithulananthan N, Bansal RC. Multiple distributed generatorsplacement in primary distribution networks for loss reduction. IEEE Transac-tions on Industrial Electronics 2013;60(4):1700–8.

[127] Abu-Mouti SFahad, El-Hawary ME. Optimal distributed generation allocationand sizing in distribution systems via artificial bee colony algorithm. IEEETransactions on Power Delivery 2011;26(4):2090–101.

[128] Kumar A, Gao W. Optimal distributed generation location using mixedinteger non-linear programming in hybrid electricity markets. IET Genera-tion, Transmission and Distribution 2010;4(2):281–98.

[129] Keane A, Malley MO. Optimal allocation of embedded generation ondistribution networks. IEEE Transactions on Power Systems 2005;20(3):1640–6.

[130] Esmin AAA, Lambert-Torres G, Zambroni AC, de Souza. A hybrid particleswarm optimization applied to loss power minimization. IEEE Transactionson Power Systems 2005;20(2):859–66.

[131] Moradi A, Fotuhi-Firuzabad M. Optimal switch placement in distributionsystems using trinary particle swarm optimization algorithm. IEEE Transac-tions on Power Delivery 2008;23(1):271–9.

[132] Mantway AH, Al-Muhaini MM. Multi-objective BPSO algorithm for distribu-tion system expansion planning including Distributed Generation. In: Pro-ceedings of transmission and distribution conference and exposition 2008,IEEE/PES; April 2008, p. 1–8.

[133] Sedighi M, Igderi A, Parastar A.. Sitting and sizing of distributed generation indistribution network to improve of several parameters by PSO algorithm. In:IPEC conference proceedings; October 2010.

[134] Soroudi A, Afrasiab M. Binary PSO-based dynamic multi-objective model fordistributed generation planning under uncertainty. In: IET renewable powergeneration received on 7th February 2011 revised on 10th October; 2011.

[135] El-Zonkoly AM. Optimal placement of multi-distributed generation unitsincluding different load models using particle swarm optimization. IETGeneration, Transmission and Distribution 2011;5(7):760–71.

[136] Prabha DR, Mageshvaran R, Raghunath E, Raghuram G. Determining theoptimal location and sizing of distributed generation Unit using ParticleSwarm Optimization algorithm. In: Proceedings of international conferenceon computer communication and informatics (ICCCI); January 2012.

[137] Zou Kai, Agalgaonkar AP, Muttaqi KM, Perera S. Distribution system planningwith incorporating DG reactive capability and system uncertainties. IEEETransactions on Sustainable Energy 2012;3(1):112–23.

[138] Jamian JJ, Mustafa MW,Mokhlis H, Abdullah MN. Comparative study ondistributed generator sizing using three types of particle swarm optimiza-tion. In: Proceedings of third international conference on intelligent systems,modelling and simulation (ISMS); 8–10 February 2012.

[139] Kim KH, Lee YJ, Rhee SB, Lee SK, You SK. Dispersed generator placementusing fuzzy-GA in distribution systems. In: Proceedings of IEEE powerengineering society summer meeting, vol. 3; July2002, p. 1148–53.

[140] Silvestri A, Berizzi A, Buonanno S. Distributed generation planning usinggenetic algorithms. In: Proceedings of international conference on electricpower engineering (powertech Budapest); September 1999, p. 257.

[141] Kim JO, Park SK, Park KW, Singh C. Dispersed generation planning usingimproved Hereford Ranch algorithm. Electric Power Systems Research1998;47(1):47–55.

[142] Wang Lingfeng, Singh C. Population-based intelligent search in reliabilityevaluation of generation systems with wind power penetration. IEEETransactions on Power Systems 2008;23(3):1336–45.

[143] Sheidaei F, Shadkam M, Zarei M. Optimal distributed Generation allocation indistribution systems employing ant colony to reduce losses. In: Proceedingsof 43rd international universities power engineering conference UPEC;September 2008.

[144] Chang Chung-Fu. Reconfiguration and capacitor placement for loss reductionof distribution systems by ant colony search algorithm. IEEE Transactions onPower Systems 2008;23(4):1747–55.

[145] Zamboni L, Monteiro LHA. Optimization of the topology of electric energydistribution networks by using algorithm inspired on ant behavior. IEEE(Revista IEEE America Latina) Latin America Transactions 2009;7(1):85–91.

[146] Hussain Israfil, Roy Kumar Anjan. Optimal size and location of distributedgenerations using Differential Evolution (DE). In: Proceedings of 2nd nationalconference on computational intelligence and signal processing (CISP); 2–3March 2012.

[147] Abu-Mouti FS, El-Hawary ME. Heuristic curve-fitted technique for distributedgeneration optimization in radial distribution feeder systems. IET Generation,Transmission and Distribution 2011;5(2):172–80.














































































































Documents

Loss minimization techniques used in distribution network: bibliographical survey