814 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 25, NO.

3, SEPTEMBER 2010

Analytical Expressions for DG Allocation

in Primary Distribution Networks
Duong Quoc Hung, Nadarajah Mithulananthan, Member, IEEE, and R. C. Bansal, Senior Member, IEEE

Abstract—This paper proposes analytical expressions for finding are placed properly and appropriately sized, they could also be
optimal size and power factor of four types of distributed gener- considered as an effective way to reduce losses, improve voltage
ation (DG) units. DG units are sized to achieve the highest loss profiles and increase reliability.
reduction in distribution networks. The proposed analytical ex-
pressions are based on an improvement to the method that was There is a wide range of terminologies used for “distributed
limited to DG type, which is capable of delivering real power only. generation,” such as “embedded generation,” “dispersed gener-
Three other types, e.g., DG capable of delivering both real and re- ation,” or “decentralized generation” [3]. DG essentially means
active power, DG capable of delivering real power and absorbing a small-scale power station different from a traditional or large
reactive power, and DG capable of delivering reactive power only, central power plant. At present, there are several technologies
can also be identified with their optimal size and location using the
proposed method. The method has been tested in three test dis- ranging from traditional to nontraditional used in DG applica-
tribution systems with varying size and complexity and validated tion. The former is nonrenewable technologies such as internal
using exhaustive method. Results show that the proposed method combustion engines, combined cycles, combustion turbines, and
requires less computation, but can lead optimal solution as verified microturbines. The latter technologies include fuel cells, storage
by the exhaustive load flow method. devices, and a number of renewable energy-based technologies
Index Terms—Analytical expressions, distributed generation such as photovoltaic, biomass, wind, geothermal, ocean, etc.
(DG), loss reduction, optimal location, optimal power factor, When renewable energy-based DG units are placed for loss
optimal size. reduction, both aspects of sustainable energy, i.e., renewable en-
ergy and energy efficiency are addressed. The challenges in DG
I. INTRODUCTION applications for loss reduction are proper location, appropriate
LECTRICAL energy is continuously lost due to resistance sizes, and operating strategies. Even if the location is fixed due
E in power system networks, and distribution system loss
accounts for more compared to transmission system [1]. More-
to some other reasons, improper size would increase the losses
in the system beyond the losses for case without DG. Optimal
over, distribution systems are well known for a higher R/X ratio sizing and location depend on the type of DG as well. Hence,
compared to transmission systems and significant voltage drops in this paper, an attempt is made to develop simple analytical
can result in substantial power and energy losses along distribu- expressions for sizing, which can be easily calculated.
tion feeders. As a result, loss reduction in distribution systems Rest of the paper is organized as follows: Section II presents
is one of the greatest challenges to many utilities around the a brief summary of literatures on loss reduction techniques.
world. Reconfiguration and capacitors placement are two major Proposed analytical expressions for finding optimal sizes of
methods for loss reduction in distribution systems. various DG types are introduced in Section III. Optimal place-
In recent years, penetration of DG into distribution systems ment and operating power factor of DG is also addressed in
has been increasing around the world. Major reasons for this the section. Section IV presents numerical results of application
trend are liberalization of electricity markets, constraints on of developed analytical expression in three test systems, inter-
building new transmission and distribution lines and environ- esting observations along with discussions. Finally, the major
mental concerns [2], [3]. For instance, a research by the Electric contributions and conclusions are summarized in Section V.
Power Research Institute (EPRI) estimates that DG will be about
25% of the new generation by 2010, while a similar study by the II. LOSS REDUCTION TECHNIQUES
National Gas Foundation shows that this figure could be even There have been many studies on the reconfiguration of distri-
higher, account for nearly 30% [3]. Given this trend, if DGs bution systems for loss reduction. A switch exchange algorithm
was proposed in [4]. In [5], an approximate power-flow tech-
nique was developed for analyzing loss reduction from network
reconfiguration. In [6], Fan et al. formulated the reconfigura-
Manuscript received July 30, 2009; revised December 10, 2009; accepted
February 8, 2010. Date of current version August 20, 2010. Paper no. TEC- tion problem as a linear programming problem and applied a
00309-2009. single-loop optimization method to solve network reconfigura-
D. Q. Hung is with the Power Company No. 2, Electricity of Vietnam, Ho tion. Other techniques such as the genetic algorithm (GA) [7],
Chi Minh City 70000, Vietnam (e-mail: [email protected]).
N. Mithulananthan and R. C. Bansal are with the School of Information Tech- simulated annealing (SA) [8], improved Tabu Search (TS) [9],
nology and Electrical Engineering, The University of Queensland, Brisbane, and ant colony search (ACS) algorithm [10] have been used for
Qld. 4072, Australia (e-mail: [email protected]; [email protected]). the purpose of network reconfiguration for reducing losses.
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org. For optimal capacitor placement for loss reduction, a well
Digital Object Identifier 10.1109/TEC.2010.2044414 known “Golden Rule,” or “2/3 rule” is presented in [11]. This
0885-8969/$26.00 © 2010 IEEE
HUNG et al.: ANALYTICAL EXPRESSIONS FOR DG ALLOCATION IN PRIMARY DISTRIBUTION NETWORKS 815

method would yield good solutions in system where the loads Qi and Qj the reactive power injections at the ith and
are uniformly distributed. Many researchers have applied other jth buses, respectively;
techniques such as dynamic programming [12], fuzzy expert N the number of buses.
system [13], TS [14], and GA [15] for finding best locations for
capacitors to reduce losses. B. Types of DG
In distribution systems, DG can deliver a portion of real and/or
DG can be classified into four major types based on their
reactive power so that the feeder current is reduced and voltage
terminal characteristics in terms of real and reactive power de-
profile can be improved with reduction in losses. However, stud-
livering capability as follows:
ies indicate that poor selection of location and size would lead
1) Type 1: DG capable of injecting P only.
to higher losses than the losses without DGs [16], [17]. A tech-
2) Type 2: DG capable of injecting Q only.
nique for DG placement using “2/3 rule” has been presented
3) Type 3: DG capable of injecting both P and Q.
in [17]. Although the 2/3 rule is simple and easy to apply, this
4) Type 4: DG capable of injecting P but consuming Q.
technique may not be effective in distribution with not uniformly
Photovoltaic, micro turbines, fuel cells, which are integrated
distributed loads. Besides, if a DG is capable of delivering real
to the main grid with the help of converters/inverters are good
and reactive power, applying the method that was developed
examples of Type 1. Type 2 could be synchronous compensators
for capacitor placement may not work. In [18], an analytical
such as gas turbines. DG units that are based on synchronous
approach has been presented to identify appropriate location to
machine (cogeneration, gas turbine, etc.) fall in Type 3. Type 4
place single DG in radial as well as loop systems to minimize
is mainly induction generators that are used in wind farms.
losses. But, in this approach, optimal sizing is not considered.
GA was applied to determine the size and location of DG in [19]
C. Sizing at Various Locations
and [20]. Though GA is suitable for multiobjective problems
and can lead to a near optimal solution, they demand compu- Assuming a = (sign)tan(cos−1 (PFDG )), the reactive power
tational time. Recently, an analytical approach based on exact output of DG is expressed by (2)
loss formula was presented to find the optimal size and location
QDG i = aPDG i (2)
of DG [16]. In this method, the load flow is required to be con-
ducted only twice. The first load-flow calculation is needed to in which
calculate the loss of base case. The second load-flow solution sign = +1: DG injecting reactive power;
is required to find the minimum total loss after DG placement. sign = −1: DG consuming reactive power;
The technique requires less computation. However, the analyt- PFDG is the power factor of DG.
ical approach can be applied to DG capable of delivering only The active and reactive power injected at bus i, where the DG
real power. located, are given by (3) and (4), respectively,
Most of the approaches presented so far model DG as a ma-
Pi = PDG i − PD i (3)
chine that is capable of delivering only real power. However,
there are other types of DG being integrated into distribution Qi = QDG i − QD i = aPDG i − QD i . (4)
systems. This paper develops a comprehensive formula by im-
proving the analytical (IA) method proposed in [16] to find the From (1), (3), and (4), the active power loss can be rewritten
optimal sizes, optimal locations of various types of DG. The as
 
paper also presents the importance of operating DGs that are
N
N
αij [(PDG i − PD i )Pj + (aPDG i − QD i )Qj ]
capable of delivering both real and reactive power at the proper PL = .
power factor to achieve minimum losses. i=1 j =1 +βij [(aPDG i − QD i )Pj − (PDG i − PD i )Qj ]
(5)
III. PROPOSED METHODOLOGY The total active power loss of the system is minimum if
the partial derivative of (5) with respect to the active power
A. Problem Formulation injection from DG at bus i becomes zero. After simplification
The total real power loss in power systems is represented by and rearrangement, (5) can be written as
(1), popularly known as “exact loss formula” [21]
N
∂PL
=2 [αij (Pj + aQj ) + βij (aPj − Qj )] = 0. (6)


N

N
∂PDG i
PL = [αij (Pi Pj + Qi Qj ) + βij (Qi Pj − Pi Qj )] (1) j =1

i=1 j =1 Equation (6) can be rewritten as

where

N
rij rij
αij = cos (δi − δj ), βij = sin (δi − δj ); αii (Pi + aQi ) + βii (aPi − Qi ) + (αij Pj − βij Qj )
Vi Vj Vi Vj
j =1
Vi
δi the complex voltage at the bus ith; j
= i
rij + jxij = Zij the ijth element of [Zbus] impedance
matrix;

N
+a (αij Qj + βij Pj ) = 0 (7)
Pi and Pj the active power injections at the ith and
j =1
jth buses, respectively; j
= i
816 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 25, NO. 3, SEPTEMBER 2010



n

 (αij Pj − βij Qj )


Xi =




j =1
j
= i
let (8)


n



 Yi = (αij Qj + βij Pj ).


 j =1
j
= 1 Fig. 1. Simple distribution system with DG.
From (3), (4), (7), and (8), (9) can be developed
 distribution systems as trend of loss reduction can be captured
αii PDG i − PD i + a2 PDG i − aQD i
with α and β coefficients from the bases case.
+ βii (QD i − aPD i ) + Xi + aYi = 0. (9)
E. Optimal Power Factor
From (9), the optimal size of DG at each bus i for minimizing
loss can be written as Consider a simple distribution system with two buses, a
source, a load and DG connected through a transmission line as
αii (PD i + aQD i ) + βii (aPD i − QD i ) − Xi − aYi
PDG i = . shown in Fig. 1.
a2 αii + αii
(10) The power factor of the single load (PFD 2 ) is given by (13)
The power factor of DG depends on operating conditions and PD 2
type of DG. When the power factor of DG is given, the optimal PFD 2 =  . (13)
PD2 2 + Q2D 2
size of DG at each bus i for minimizing losses can be found in
the following way. It can be proved that at the minimum loss occur when power
1) Type 1 DG: For Type 1 DG, power factor is at unity, i.e., factor of DG is equal to the power factor of load as given by
PFDG = 1, a = 0. From (10), the optimal size of DG at (14).
each bus i for minimizing losses can be given by reduced
equation (11) PDG 2
PFD 2 = PFDG 2 =  . (14)
  2
PDG + Q2DG 2
2

1 
N

PDG i = PD i − βii QD + (αij Pj − βij Qj ) In practice, a complex distribution system includes a few
αii  i .
j =1 sources, many buses, many lines and loads. The power factors
j
= i of loads are different. If each load is supplied by each local DG,
(11) at which the power factor of each DG is equal to that of each
2) Type 2 DG: Assuming PFDG = 0 and a = ∞, from (2) to load, there is no current in the lines. The total line power loss
(10), the optimal size of DG at each bus i for minimizing is zero. The transmission lines are also unnecessary. However,
losses is given by reduced equation (12) that is unrealistic since the capital investment cost for DG is too
  high. Therefore, the number of installed DGs should be limited.

To find the optimal power factor of DG for a radial complex
1  
N
QDG i = QD i + βii PD − (αij Qj + βij Pj ) distribution system, fast and repeated methods are proposed. It
αii  i .
j =1 is interesting to note that in all the three test systems the optimal
j
= i
power factor of DG (Type 3) placed for loss reduction found
(12) to be closer to the power factor of combined load of respective
3) Type 3 DG: Assuming 0 < PFDG < 1, sign = +1 and “a” system.
is a constant, the optimal size of DG at each bus i for the 1) Fast Approach: Power factor of combined total load of
minimum loss is given by (10) and (4), respectively. the system (PFD ) can be expressed by (13). In this condition, the
4) Type 4 DG: Assuming 0 < PFDG < 1, sign = −1 and “a” total active and reactive power of the load demand are expressed
is a constant, the optimal size of DG at each bus i for the as
minimum loss is given by (10) and (4), respectively.


N
PD = PD i (15)
D. Optimal Location i=1

For optimal location, the method proposed in [16] is used.

N
Based on this method, first the optimal sizes at various locations QD = QD i . (16)
have been calculated for different types of DG and the losses i=1
were calculated with optimal sizes for each case. The case with The “possible minimum” total loss can be achieved if the
minimum losses is selected as the optimal location for each type power factor of DG (PFDG ) is quickly selected to be equal to
of DG. that of the total load (PFD ). That can be expressed by (17)
Based on this method proposed in [16], one can avoid ex-
haustive computation and save time, especially for large-scale PFDG = PFD . (17)
HUNG et al.: ANALYTICAL EXPRESSIONS FOR DG ALLOCATION IN PRIMARY DISTRIBUTION NETWORKS 817

2) Repeated Approach: In this method, the optimal power

factor is found by calculating power factors of DG (change
in a small step of 0.01) near to the power factor of combined
load. The sizes and locations of DG at various power factors
with respect to losses are identified from (14). The losses are
compared together. The optimal power factor of DG for which
the total loss is at minimum is determined.

F. Computational Procedure
When power factor of DG is set to be equal to that of combined
total loads, computational procedure to find optimal size and
location of one of four types of DGs is described in the following.
Step 1: Run load flow for the base case. Fig. 2. Optimal sizes of DG at 0.98 lagging load power factor at various
Step 2: Find the base case loss using (1). locations with respect to losses for 16-bus system.
Step 3: Calculate power factor of DG using (17).
Step 4: Find the optimal size of DG for each bus using (4)
and (10).
Step 5: Place DG with the optimal size obtained in step 4 at
each bus, one at a time. Calculate the approximate
loss for each case using (1) with the values α and β
of base case.
Step 6: Locate the optimal bus at which the total loss is mini-
mum corresponding with the optimal size at that bus.
Step 7: Run load flow with the optimal size at the optimal
location obtained in step 6. Calculate the exact loss
using (1) and the values α and β after DG placement.
It is noted that when the type and power factor of DG is
given; the computational procedure to find the optimal size and
location of DG is as described earlier apart from step 3. At this Fig. 3. Optimal sizes of DG at 0.85 lagging load power factor at various
step, the power factor of DG is entered rather than using (17). locations with respect to losses for 33-bus system.
In exhaustive load flow (ELF) method, optimal sizes, optimal
location, and power factor are obtained with a number of load
flow solutions. 2) Maximum active power limit of DG for different test sys-
tems is assumed to be equal to the total active load of the
system.
IV. NUMERICAL RESULTS
3) The lower and upper voltage thresholds for DG with the
A. Test Systems optimal size, location and power factor are set at 0.95–
Three test systems have been used to test and validate the 1.05 pu.
proposed analytical expressions for optimal sizing, placement,
and power factor. The first system is 16-bus radial test distri- C. Simulation Results
bution system with a total real and reactive load of 28.7 MW
1) Sizing Allocation: Figs. 2–4 show optimal sizes and total
and 5.9 MVAr, respectively [22]. The second system is 33-bus
losses with the optimal sizes of DG one at a time at respective
test radial distribution system with a total real and reactive load
location for 16, 33, and 69 bus test systems, respectively. In all
of 3.7 MW and 2.3 MVAr, respectively [23]. The last system
the figures, lower half shows the optimal sizes obtained using
is 69-bus test radial distribution system with a total real and
proposed method (IA) and ELF method using exact loss formula.
reactive load of 3.8 MW and 2.69 MVAr, respectively [24].
Notice in all the cases, optimal sizes of DG obtained from the
An analytical software tool has been developed in MATLAB
proposed method is closer to actual optimal size obtained from
environment to run load flow and calculate losses and optimal
exhaustive load flow solutions. Similarly, total losses obtained
sizes of DG. Though the proposed analytical methods can han-
from the method without updating α and β for the purpose of
dle four different types of DG, results of Type 3-DG, i.e., DG
identifying best location and exhaustive methods give a compa-
capable of delivering real and reactive power is presented in this
rable result.
paper.
Observe that when DG with a large size is placed at any
locations near a swing bus, the loss reduction of each system
B. Assumptions and Constraints
is quite low. In contrast, only a small size of DG is added at
1) The sizing and location of DG is considered at the peak the other locations, a higher loss reduction can be achieved. As
load only. a result, finding a location at which the total loss at minimum
818 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 25, NO. 3, SEPTEMBER 2010

Fig. 4. Optimal sizes of DG at 0.82 lagging load power factor at various

locations with respect to losses for 69-bus system. Fig. 6. Plot of optimum sizes and locations of DG at various power factors
versus losses for 33-bus test system.

Fig. 5. Plot of optimum sizes and locations of DG at various power factors

versus losses for 16-bus test system. Fig. 7. Plot of optimum sizes and locations of DG at various power factors
versus losses for 69-bus test system.

is important. That can be implemented by the support of IA as TABLE I

presented in the following. POWER FACTORS OF DG AND COMBINED LOADS
2) Site Selection: Figs. 2–4 also show (upper half) the total
losses with the optimal DG at various buses, one at a time using
both methodologies. The total losses by IA are slightly higher
than the exact losses at by ELF. However, the trend of losses with
IA and ELF are completely coinciding and for the purpose of
identifying location, a fast method would be sufficient. In other
words, IA can find the optimal locations with less computational
efforts.
In 16-bus, 33-bus, and 69-bus test distribution systems used loss reduction. The range of power factors good for minimizing
in the paper, the best locations for DG for minimizing losses losses at bus 9 is 0.87 leading to 0.46 lagging with optimal sizes
found to be buses 9, 6, and 61, respectively. ranging from 7.46 to 13.12 MVA.
3) Power Factor Selection: Figs. 5–7 show the optimal sizes Though in reality the sizes will be fixed and power factor
and locations of DGs at various power factors and total losses can be allowed to vary to observe the impact of DG power
by IA method for 16, 33, and 69-bus test systems, respectively. factor on loss reduction. If the power factor is fixed due to
The power factors of DG varied from zero to 1.00, both in the limitation of the technology, it is important to select the
leading and lagging operation in small steps of 0.01. This study appropriate size and location to achieve minimum losses. If the
is carried out to see the optimal power factor of DG that would technology available for DG to be a Type I, i.e., around unity
give minimum losses. power factor, the best location is bus 9. On the other hand if
In 16-bus system, buses 8, 9, and 12 found to be good can- DG type is Type II, i.e., zero power factor then the best location
didate locations for distributed generator placement for mini- is at bus 8, for reducing losses and so on. In this test system,
mizing losses. These results are in agreement with the results the best power factor found to be closer to unity (i.e., 0.99) and
obtained from previous section “site selection.” Notice at bus corresponding size and losses are 13.12 MVA and 1642.58 kW,
9, the best location, power factor plays an important role in respectively. Unity power factor means this system needs only
HUNG et al.: ANALYTICAL EXPRESSIONS FOR DG ALLOCATION IN PRIMARY DISTRIBUTION NETWORKS 819

TABLE II
COMPARISON OF RESULTS OF DIFFERENT TECHNIQUES

real power injections to reduce losses. However, this is not the TABLE III
VOLTAGES BEFORE AND AFTER DG
case in other systems as presented below. In 33-bus test system,
bus 6 found to be the best location. This result too coincides
with the result obtained from the previous section. The range
of power factors for that location is 0.89 leading to 0.7 lagging
and corresponding optimal sizes are 1.43–3.03 MVA. The best
power factor of DG located at 6 is 0.82 lagging with an optimal
size of 3.02 MVA. This means that the system needs both real
and reactive power injection to reduce losses. However, more
real power injection is needed compared to reactive power. The
runs so many load flows in calculation around the solution ob-
total losses corresponding to DG at the optimal location, optimal
tained for IA method which need only a couple of load flow
size, and best power factor is 67.98 kW.
runs. Hence, IA method requires shorter computational time.
In 69-bus test system, bus 61 is the obvious location for a
This technique is appropriate and useful for large-scale distri-
range of power factors compared to other locations. The best
bution systems.
power factor of DG located at bus 61 is 0.82 lagging with an
Selection of the power factor of DG equal to that of the load
optimal size of 2.2 MVA. In this case too, both real and reactive
is feasible, and it can lead to an optimal or near-optimal solution
power injections are needed to minimize the losses. The total
as shown in Table II.
loss corresponding to these optimal settings is 22.64 kW.
With DG the total losses can be reduced significantly while
Table I shows the optimal power factor of DG for different test
satisfying all the power and voltage constraints. When only one
systems using IA method with repeated approach and combined
DG is considered in terms of the optimal size, location and
load power factor of each system. It is interesting to see that the
power factor simultaneously; the loss reduction for the three
optimal power factor of DG for minimizing losses is in close
systems is achieved at the highest levels of 68.21%, 67.81%,
agreement with load power factor. This could be a guidance
and 89.68%, respectively.
to select the appropriate technological option for minimizing
Table III shows the minimum and maximum voltages before
losses in the system, given a choice. However, it is important
and after DG by IA for the three test systems. After DG, voltage
to notice that optimal power factor depends on the location as
profiles improve significantly. The voltages at various buses
well.
maintain within the acceptable constraints.
It should be noted that, in DG case leading power factor
means, DG is absorbing reactive power. On the other hand,
lagging means DG supplying reactive power and this is just V. CONCLUSION
opposite to load power factor conventions. The paper has proposed analytical expressions for finding
optimal size and power factor of different types of distributed
generators for minimizing losses in primary distribution sys-
D. Summary
tems. Validity of the proposed analytical expressions for finding
Table II shows a summary and comparison of the simulation optimal size are tested and verified on three test distribution sys-
results for 16, 33, and 69-bus test systems for two approaches tems with varying sizes and complexity using exhaustive load
for calculating power factor of DG, i.e., fast and repeated ap- flow solutions. Results show that locations, sizes, and operat-
proaches. In fast approach, the DG power factor is set at com- ing power factor of distributed generators are crucial factors in
bined load power factor, and in the repeated approach, the best reducing losses. If properly placed, appropriately sized and op-
power factor is calculated by checking all the power factors step erated distributed generators can reduce losses significantly. In
by step. The base case losses of all the test system too presented all the test systems used in this study, operating power factor of
to show the loss reduction by distributed generators. DGs for minimizing losses found to be closer to the power fac-
The loss reduction by IA and ELF is almost the same. How- tor of combined load in the respective system. This could be a
ever, ELF can demand excessively computational time since it good guidance for operating distributed generators that have the
820 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 25, NO. 3, SEPTEMBER 2010

capability to deliver both real and reactive power for minimizing [24] M. E. Baran and F. F. Wu, “Optimum sizing of capacitor placed on radial
losses. distribution systems,” IEEE Trans. Power Del., vol. 4, no. 1, pp. 735–743,
Jan. 1989.
REFERENCES
[1] J. Federico, V. Gonzalez, and C. Lyra, “Learning classifiers shape reactive
power to decrease losses in power distribution networks,” in Proc. IEEE
Duong Quoc Hung received the M.Eng. degree in en-
Power Eng. Soc. General Meet., Jun. 2005, vol. 1, pp. 557–562.
ergy field of study (specialization on electric power
[2] IEA,Distributed Generation in Liberalized Electricity Markets. Paris,
system management) from Asian Institute of Tech-
France: OECD, 2002. Available: http://iea.org/textbase/nppdf/free/2000/
nology, Bangkok, Thailand, in May 2008.
distributed2002.pdf
Since 1999, he has been with the Power Company
[3] T. Ackermann, G. Andersson, and L. Soder, “Distributed generation: A
No. 2, Electricity of Vietnam, Ho Chi Minh City,
definition,” Electr. Power Syst. Res., vol. 57, no. 3, pp. 195–204, 2001.
Vietnam, where he is currently an Electrical En-
[4] S. Civanlar, J. J. Grainger, H. Yin, and S. S. H. Lee, “Distribution feeder
gineer in the Technical Department. His research
reconfiguration for loss reduction,” IEEE Trans. Power Del., vol. 3, no. 3,
interests include distribution system design and oper-
pp. 1217–1223, Jul. 1988.
ations, distributed generators, capacitors, and recon-
[5] M. E. Baran and F. F. Wu, “Network reconfiguration in distribution systems
figuration and application of optimization techniques
for loss reduction and load balancing,” IEEE Trans. Power Del., vol. 4,
in distribution systems.
no. 2, pp. 1401–1407, Apr. 1989.
[6] J. Y. Fan, L. Zhang, and J. D. McDonald, “Distribution network recon-
figuration: single loop optimization,” IEEE Trans. Power Syst., vol. 11,
no. 3, pp. 1643–1647, Aug. 1996.
[7] K. Nara, A. Shiose, M. Kitagawa, and T. Ishihara, “Implementation of
genetic algorithm for distribution system loss minimum re-configuration,”
Nadarajah Mithulananthan (M’02) received the
IEEE Trans. Power Syst., vol. 7, no. 3, pp. 1044–1051, Aug. 1992.
B.Sc. (Eng.) degree from the University of
[8] Y. J. Jeon, J. C. Kim, J. O. Kim, J. R. Shin, and K. Y. Lee, “An efficient
Peradeniya, Peradeniya, Sri Lanka and the M.Eng.
simulated annealing algorithm for network reconfiguration in large-scale
degree from the Asian Institute of Technology,
distribution systems,” IEEE Trans. Power Del., vol. 17, no. 4, pp. 1070–
Bangkok, Thailand, in May 1993 and August 1997,
1078, Oct. 2002.
respectively, and the Ph.D. degree in electrical
[9] D. Zhang, Z. Fu, and L. Zhang, “An improved TS algorithm for loss-
and computer engineering from the University of
minimum reconfiguration in large-scale distribution systems,” Electr.
Waterloo, Waterloo, ON, Canada, in 2002.
Power Syst. Res., vol. 77, no. 5–6, pp. 685–694, 2007.
He was an Electrical Engineer at the Genera-
[10] C. T. Su, C. F. Chang, and J. P. Chiou, “Distribution network reconfigu-
tion Planning Branch, Ceylon electricity Board, and
ration for loss reduction by ant colony search algorithm,” Electr. Power
a Researcher at Chulalongkorn University, Bangkok,
Syst. Res., vol. 75, no. 2–3, pp. 190–199, 2005.
Thailand. He is currently a Senior Lecturer at the School of Information Technol-
[11] J. V. Schmill, “Optimum size and location of shunt capacitors on dis-
ogy and Electrical Engineering, The University of Queensland (UQ), Brisbane,
tribution feeders,” IEEE Trans. Power App. Syst., vol. PAS-84, no. 9,
Qld., Australia. Prior to joining UQ, he was an Associate Professor at the Asian
pp. 825–832, Sep. 1965.
Institute of Technology. His research interests include integration of renewable
[12] T. H. Fawzi, S. M. El-Sobki, and M. A. Abdel-Halim, “A new approach
energy in power systems and power system stability and dynamics.
for the application of shunt capacitors to the primary distribution feeders,”
IEEE Trans. Power App. Syst., vol. PAS-102, no. 1, pp. 10–13, Jan. 1983.
[13] H. N. Ng, M. M. A. Salama, and A. Y. Chikhani, “Capacitor allocation by
approximate reasoning: fuzzy capacitor placement,” IEEE Trans. Power
Del., vol. 15, no. 1, pp. 393–398, Jan. 2000.
[14] R. A. Gallego, A. J. Monticelli, and R. Romero, “Optimal capacitor place-
ment in radial distribution networks,” IEEE Trans. Power Del., vol. 16, R. C. Bansal (SM’03) received the M.E. degree from
no. 4, pp. 630–637, Nov. 2001. Delhi College of Engineering, Delhi, India, in 1996,
[15] M. A. S. Masoum, M. Ladjevardi, A. Jafarian, and E. F. Fuchs, “Optimal the M.B.A. degree from Indira Gandhi National Open
placement, replacement and sizing of capacitor banks in distorted distri- University, New Delhi, India, in 1997, and the Ph.D.
bution networks by genetic algorithms,” IEEE Trans. Power Del., vol. 19, degree from the Indian Institute of Technology Delhi
no. 4, pp. 1794–1801, Oct. 2004. (IIT Delhi), Delhi, India, in 2003.
[16] N. Acharya, P. Mahat, and N. Mithulananthan, “An analytical approach From August 1989 to August 1998, he was with
for DG allocation in primary distribution network,” Int. J. Electr. Power the Civil Construction Wing, All India Radio, for nine
Energy Syst., vol. 28, no. 10, pp. 669–678, 2006. years. From June 1999 to December 2005, he was
[17] H. L. Willis, “Analytical methods and rules of thumb for modeling DG- an Assistant Professor in the Department of Elec-
distribution interaction,” in Proc. IEEE Power Eng. Soc. Summer Meet., trical and Electronics Engineering, Birla Institute of
Jul., 2000, vol. 3, pp. 1643–1644. Technology and Science, Pilani, India. From February 2006 to June 2008,
[18] C. Wang and M. H. Nehrir, “Analytical approaches for optimal placement he was with the School of Electrical and Electronics Engineering Division,
of distributed generation sources in power systems,” IEEE Trans Power School of Engineering and Physics, The University of the South Pacific, Suva,
Syst., vol. 19, no. 4, pp. 2068–2076, Nov. 2004. Fiji. He is currently a faculty Member at the School of Information Technol-
[19] K. H. Kim, Y. J. Lee, S. B. Rhee, S. K. Lee, and S. K. You, “Dispersed ogy and Electrical Engineering, The University of Queensland, Brisbane, Qld.,
generator placement using fuzzy-GA in distribution systems,” in Proc. Australia. He is the author or coauthor of more than 125 papers published in na-
IEEE Power Eng. Soc. Summer Meet., Jul. 2002, vol. 3, pp. 1148–1153. tional/international journals and conference proceedings. His research interests
[20] A. Silvestri, A. Berizzi, and S. Buonanno, “Distributed generation plan- include reactive power control in renewable energy systems and conventional
ning using genetic algorithms,” in Proc. Int. Conf. Electr. Power Eng. power systems, power system optimization, analysis of induction generators,
(PowerTech Budapest), Sep., p. 257. and artificial intelligence techniques applications in power systems. He is an
[21] D. P. Kothari and J. S. Dhillon, Power System Optimization. New Delhi: Editorial Board Member of the IET Renewable Power Generation Journal and
Prentice-Hall, 2006. Electric Power Components and Systems Energy Sources, Part B: Economics,
[22] S. Civanlar, J. J. Grainger, H. Yin, and S. S. H. Lee, “Distribution feeder Planning and Policy.
reconfiguration for loss reduction,” IEEE Trans. Power Del., vol. 3, no. 3, Dr. Bansal is the Editor of IEEE TRANSACTIONS OF ENERGY CONVERSION
AND POWER ENGINEERING LETTERS and the Associate Editor of IEEE
pp. 1217–1223, Jul. 1988.
[23] M. A. Kashem, V. Ganapathy, G. B. Jasmon, and M. I. Buhari, “A novel TRANSACTIONS OF INDUSTRIAL ELECTRONICS. He is a member of the Board
method for loss minimization in distribution networks,” in Proc. Int. Con. of Directors of the International Energy Foundation (IEF), Alberta, Canada. He
Electr. Util. Deregulation Restruct. Power Technol., Proc., Apr. 2000, is also a member of the Institution of Engineers (India) and a Life Member of
pp. 251–256. the Indian Society of Technical Education.