t

1s IEEE International Conference on Power Electronics. Intelligent Control and Energy Systems (ICPEICES-2016)

An Analytical Approach for Optimal Sizing and

Placement of Distributed Generation in
Radial Distribution Systems
Prem Prakashl and Dheeraj K. Khatod2
1,2Alternate Hydro Energy Centre, Indian Institute of Technology, Roorkee, Roorkee, India
E-mail:1ppyadavI974@gmail.com.2dheerfah@iitr.ac.in

Abstract -In present study, an analytical approach voltage profile enhancement of system voltage stability
based technique is presented for the optimum installation margin and system loadability [2], [3], [11], [12] and [17].
and sizing of distributed generation (DG) units for balanced
Furthermore, the investment or cost minimization or
radial distribution system. The formulations of developed
technique are based on to minimize active and reactive
maximization of benefit to cost ratio, reduction of energy
components of system losses related to branches current. In losses [4], and [10], moreover, the objective of the
the presented technique the magnitude of branches current is investigators to formulate the methodology for optimum
reduced by installation of DG units at various positions installation and sizing of DG units in distribution network
(buses) in distribution system. According to proposed
and for optimization they applied several optimization
technique, it identities a group of buses wherever DG units
techniques, like direct mathematical analysis by c1assical
are to be installed. Further, the loss saving derivations is
used to obtain optimum size of DG units by c1assical optimization approach [1], [3], [9], [10], [12], [13], and [17].
optimization technique. An algorithm is also developed to Index based approaches (various indices are applied for
optimum installation of DG units in distribution system. The optimum placement and sizing of DG units such as voltage
developed method is relatively simple since it needs solution sensitivity index, power loss sensitivity index, power
of base case load flow only. The presented technique has been
stability index and index of voltage stability etc. [11], the
applied on a 33-bus radial test distribution network. The
developed technique is very much effective as far as system
author of [4] developed a software tool based on multi­
loss reduction is concern as these are noticed from results. period AC optimum power tlow and continuation power
Keywords-Distribution System; Loss Minimiztion; DG; tlow for installation of DG units. The inherent property of
Analytical Technique; Optimal Sizing and Siting distribution system i.e. RlX ratio is higher as compare to
transmission network, due to this inherent property, the
I. INTRODUCTION
losses of power and electrical energy are more as compare to
The small scale electric power generation units, transmission networks as explained in [7] and [9], as a
directly connected to loads or customer side of meter are consequence, the reduction of these losses is one of the great
known as DG. The range of output power of DG units issue in front of the electric distribution companies,
typically varies from a few kWs to a few MWs [6].The researchers and academician across the globe [7].
use of DG technologies in existing distribution network is According to concern literature described above, the
increasing rapidly because of their small sizes, local most of the investigators assumed DG as an active power
availability and environmental friendliness. Since DGs source only. In addition to that it is more important and
may be derived by renewable and conventional energy advantageous to enhancement the performance of
resources, DG would contribute nearly 20% of total power distribution network by considering other types of DGs
generation in upcoming days [1] and [17]. The insertion of are to be installed. Furthermore, the types of DG itself
DG in distribution network is always committed to reduce provide information about capability of injection and
the network losses, improvement of voItage profile, absorption of active and reactive power ofDG units [14].
enhancement of voItage stability margin, improvement of Moreover, the majority of the researchers applied exact
system reliability and power quality of power supplied and loss formula based analytical technique in their formulation
[17], ifDG placement is planned strategically. development for optimum installation and sizing ofDG units
On the basis of available researches contributed by in distribution networks for reduction of system losses [1]
earlier researchers on optimum placement and sizing ofDG and [15]. It is more important to discuss that the usage of
in distribution network it is observed that their main focus on exact loss formula for development of any methodology
to discuss various issues like reduction of system losses [1- requires evaluation of impedance Zausmatrix, those are time
4], [9], [10], [12], [14] and [15], enhancement of system consuming processes.

978-1-4673-8587-9/16/$31.00 ©2016 IEEE [1]

 t
1s IEEE International Conference on Power Electronics. Intelligent Control and Energy Systems (ICPEICES-2016)

To mitigate hurdles of previously developed magnitude of current in the branches Iying between bus 1
techniques and inspired by research work of [9] and [12], to bus k. However, the currents in the rest of the branches
the present study proposes a new and simple analytical remain unchanged with DG installation if voltage
technique for placement of DG units in distribution improvement is negligible. The modified current in the
networks. This paper proposed a technique for reduction 'itl" branch can be given as:

of system losses. The presented technique is based on the

Sub-station
fact that magnitude of active and reactive components of
branch current is reduced by placement of DG units at
various positions or buses. The proposed technique needs
only the base case load tlow solution. The developed
approach has been applied on a 33-bus radial test Fig. 2: n-Bus Test Radial Distribution System with DG

distribution system.
The rest of paper is organized as: section 2 describes
I;
N
=I; - D/dg (4)
the proposed methodology for placement of single DG; Where Idg and D; can be given as
section 3 discusses the algorithm for placement of DG S;g
Idg =---;;­
along with tlow chart. Section 4 presents the result and Vk
discussion and finally section 5 highlights the conc1usions
of the study.
D; = {I
o
if i'" branch is between source bus and bus at which k'" DG is placed
othefWIse

11. PROPSED METHODOLOGY

Further net loss saving can be obtained by subtracting
system active power losses with DG from system active
This section presents the mathematical formulations power loss withoutDG and can be expressed as
of developed analytical approach. The following �I {
Sp= �(Jf+Q' ) - (q-D,P",) +(Q-D;{2,,) R;
2

(5)
2 }]
preassumptions are made:
a. It is considered that system is balanced test radial And the size ofDG can be retrieved by differentiating
distribution network and fed by a sub-station. (5) W.r.t. Pdg and Qdg respectively partially and equate to
b. The voltage at all the buses is c10se to 1 pu zero, therefore finally Pdg and Qdg can be given as
c. All loads are constant power load. 11-1

d. Maximum DG capacity for present test system is LP;D;Ri

not exceeds from system connected load. Pdg -_-'.i==:"n I I
-
___
(6)
Consider an 'n' bus radial distribution network shown
in Fig. 1. Here it is preassumed that I; is the phasor current LD7Ri
in branch 'i'. PLn, QLn, Pn-I and Qn-I are the load at nth bus
i=1
11-1
and sending end power tlow in (n-ll branch respectively
LPiDiRi
SUD-station
n
;.l Q dg
...:.i=
....:I (7)
1
___

p,,->jQ, �->jQ = n-I

Ik
-
10
LD?Ri
i=1
Pt.J<-+iQk PLn+jOtn
Therefore, fmal optimum size ofDG will be as
Sdg=Pdg+jQdg (8)
Fig. I: n-Bus Test Radial Distribution System

The current in ,/17, branch can be expressed as follows III. ALGORITHM FOR SIZlNGAND PLACEMENT OF DG

1= P; - jQ; (1) The following algorithm is adopted to evaluate the

I
*
Vi optimum size and placement of DG for minimization of
Further total active power loss (PL) of the system system losses. The related explanations of the base case
withoutDG may be given as
,,-I 2 "-I p2+Q2 ( load tlow are summarized in tlow chart for proposed
analytical method in Fig. 3.
�=L/;R;=I i=1
-- J
' �2 -' R; i=1
(2)

Since it is assumed that the voltage remains c10se to

Step 1: Read the data of stipulated radial distribution
system and run the load tlow of base case and
pu at every bus, the expressions of PL can be evaluate sending and receiving end real and
approximated as reactive powers and currents, bus voltages
etc. The required data is taken from [5].
�=I(p'2+QnR;
1/-1
(3) Step 2: Determine loss saving and subsequent size of
i=!

Now consider that a DG injecting Sdg power in the DG by using equations (5) and (8) by
system is placed at bus 'k' shown in Fig. 2. This changes assuming each bus as candidate bus

[2]
 t
1s IEEE International Conference on Power Electronics. Intelligent Control and Energy Systems (ICPEICES-2016)

Step 3: IdentifY the node that has highest loss saving A. 33-Bus Test Radial Distribution System
from Fig. 6 and find the size of DG
A 12.66 kV, 33-bus radial test distribution system has
corresponding to that bus from Fig. 5. Further
been considered to test the developed analytical method.
choose that bus for placement of DG of size
The present test system has a total load of (3.715 + )2.300)
selected from Fig. 5. This is the case for
MVA [5].
single DG placement
Firstly, the optimum loss saving and subsequent size
of DG calculated on the basis of equations (5) and (8)
respectively for each bus except reference bus. Fig. 5 and
Read data of the systetn (branch and load Fig. 6 illustrate the DG sizes (kW) and corresponding loss
data)
savings (MVA) for all nodes in the network except
reference node. The Fig. 6 provides an information about
highest loss saving of 130.71 kW at bus number 6 by
Estirnate active and reactive po'\-Ver and placing a DG of size (2.569+)l. 771) MVA at bus number
ClUTent by bacbvard fonvard JX)vver flovv
technique 6. The placement of DG of size (2.569+)l. 771) MVA at
bus number 6 further provides loss saving of 13.79 kW
and next DG of size (4015+)0.444) MVA at bus 31 and
Estirnate voltage magnitude along vvith
simultaneously active power loss becomes 6l.38 kW with
phasor angle of each bus
reduction of 69.72%age in same iteration. Further, a
second DG of size (0.415+)0.444) MVA at bus number
31provides loss saving of 8.17 kW and DG of size
(0.632+)0.252) MVA at bus 25 with loss reduces to 46.57
No
Is tolerance greater than equa kW with 77.02%age.
to specified tolerance?

Yes

Stop and obtain the results

Fig. 3: Flow Chart for Proposed Analytical Technique for

Placement and Sizing ofDG

IV. RESULTAND DISCUSSION

Fig. 5: Size ofDG (in MVA) at Different Buses for 33-Bus Test Radial
The 33-bus test radial distribution network is Distribution System
presented in Fig. 4 and for relevant the data refer [5]. An
analytical method has been derived and its relevant
programming is developed in MATLAB environment to
implement the proposed technique. The present test
system has total system active power and reactive power
losses are 202.68 kW and 135.14 kVAR correspondingly
without DG placement. Further, the description of change
in active power loss with DG placement is given as under

19 20 21 22
Fig. 6: Loss Saving (in kW) at Various Buses for 33-Bus Test
Fig. 4: Single Line Diagram of 33-bus Test Distribution System RadialDistribution System

[3]
 t
1s IEEE International Conference on Power Electronics. Intelligent Control and Energy Systems (ICPEICES-2016)

The Fig. 7 represents system loss saving (in kW) and It is evident from Table. 11, the minimum and
subsequent size of DG (in MVA) for different buses maximum values of system voltage profile under base
according to developed technique for 33-bus test radial case are 0.9131 pu is at bus 18 and 1.0000 p.u. at bus 1
distribution network respectively and that of afterDG placement are 0.9677 pu
at bus 18 and 1.0034 p.u. at bus 31.
TABLE 2: VOLTAGE PROFILE SUMMARY FOR 33-Bus TEST RADIAL
DISTRIBUTION SYSTEM

S.No. Voltage Profile before Voltage Profile after OG

OG Placement Placement
System Minimum Maximum Minimum Maximum
Oescription Volta!!;e Volta!!;e Volta!!;e Volta!!;e
1. 33- 0.9131 1.0000 0.9677 1.0034 p.u. at
bus p.u. at 18 p.u. at I p.u. at 18 31 bus
bus bus bus
From Flg. 8 It IS eVIdent the system voltage profile
before and after placement of DG and it is observed that
there is significant enhancement in system voltage profile
due to the magnitude of branch current decrease by
placement ofDG in distribution system.
Fig. 7: Loss Saving (in kW) and Corresponding DG size at Different Figure 9 represents the optimum power factor for all
Buses for 33-Bus Test Radial Distribution System
DG units for 33-bus test radial distribution network
Again, aDG of size (0.632+)0.252) MVA is placed at calculated on the basis of developed technique. Further,
bus number 25, it is observed that small loss saving is Fig. 9 indicate the optimum power factor of different DG
noticed and active power loss become and 38.04 kW with units at different buses for 33-bus test radial distribution
reduction of 81.23%age, with aDG of size (-0.68010.250) network. Further, DG unit associated to bus 6 having
MVA again at bus number 6. Further, this process of power factor ofO.8237 is the optimum power factor ofDG
iteration continue till loss saving approaches towards zero unit is to be installed
in this case proposed technique required 25 iteration for
convergence. Furthermore, it is worth mentioned that
initially bus number 6 was over compensated. To
determine optimum size of DG at bus 6 the amount of
active and reactive power component of load after third
iteration be subtracted from initially obtained size,
therefore, the resultant optimum size ofDG at bus number
6 will be (1.889+)1.192) MVA and again repeat same
process. Therefore, the sequence of buses whereDG units
may be placed will be 6, 31and 25. The total power loss PL
in above mentioned single DG placement reduced from
202.68 kW to 38.04 kW with percentage of 81.23% as
discussed in detail in Table. I.
T ABLE I: RESULTS SUMMARY FOR 33-Bus TEST RADIAL DISTRIBUTION
SYSTEM Fig. 8: Voltage Profile for 33-Bus Test Radial Distribution System before
and after Placement ofDG
Base Case Results after OG Installment

l>Jl
=
�- .=
.S '";)
e .e. '"
�'" Q = '"
... e �O '";:
rJJ
'" ...
- .-;, '"'"'" .oS '" iii'+�
'";;., '"'"''" - '" '"' � ��
=!:!
rJJ Q
'"' C
� ....l =Q. QC .;'� c,:.:
....l �
Base case for 202.7 6 2.57+jl.77 130.71
identitying bus DG
placement
(2.57+j 1.77) 61.4 31 0.42+jO.44 13.79
MVA, DG placed at
bus 6
(0.42+jO.44) MVA, 46.57 25 0.63+jO.25 8.17
at bus 31
(0.63+jO.25) MVA, 38.04 6 (-0.68-jO.58) -
Fig. 9: Optimum Power Factor of Different DG Units for 33-Bus
at bus 25
Radial Distribution System

[4]
 t
1s IEEE International Conference on Power Electronics. Intelligent Control and Energy Systems (ICPEICES-2016)

V. CONCLUSIONS distributed generation," IEEE Trans. Power Syst., vol. 26, no. I,pp.
198-205,201I.
The paper presents an analytical technique for [5] D. Das, D. P. Kothari, and A. Kalam, "Simple and efficient method
optimum slzmg and placement of DG in distribution for load tlow solution of radial distribution networks," Int. 1.
Electr. Power Energy Syst., vol. 17, no. 5,pp. 335-346, 1995.
network with an objective to minimize the system losses. [6] T. Ackermann, G. Andersson, and L. Söder, "Distributed
The developed technique fIrst identifIes a order of buses to generation: a definition," Int. 1. Electr. Power Syst. Res., vol. 57,no.
be compensated, by obtain maximum power loss saving 3,pp. 195-204,200I.
with placement of single DG unit in distribution network. [7] M. E. Baran and F. F. Wu, "Network reconfiguration in distribution
systems for lass reduction and load balancing," IEEE Trans.,
In this context the basic and most suited analytical Power Deliv. vol. 4, no. 2,pp. 1401-1407, 1989.
expressions have been derived to compute the optimal size [8] M. EsmaiIi, E. C. Firozjaee, and H. A. Shayanfar, "Optimal
of DG unit and corresponding loss saving, by considering placement of distributed generations considering voltage stability
the variation in magnitude of branches current due to and power losses with observing voltage-related constraints," Int.
1.Appl. Energy, vol. 113,pp. 1252-1260,2014.
installation of DG unit. The present developed method is [9] M. H. Haque, "Capacitor placement in radial distribution systems
equally applicable to optimize the size and locations of for lass reduction," lEE Proc.-Gener. Transm. Distrib., vol. 146,
single as weil as multiple DGs units for balanced radial no. 5,pp. 501-505,1999.
distribution network. The proposed approach is applied on [10] D. Q. Hung, N. Mithulananthan, and K. Y. Lee, "Optimal
placement of dispatchable and nondispatchable renewable DG units
33-bus test radial distribution network. It is observed that in distribution networks for minimizing energy loss," Int. 1. Electr.
the developed technique is exclusively committed for the Power Energy Syst., vol. 55,pp. 179-186,2014.
purpose of optimal sizing and installation of DG units in [II] V. V. S. N. Murty and A. Kumar, "Optimal placement of DG in
radial distribution network. For 33-bus test radial radial distribution systems based on new voltage stability index
under load growth," Int. 1. Electr. Power Energy Syst., vol. 69, pp.
distribution network 6, 31 and 25 buses are the best suited 246-256,2015.
for placement of DG which provides signifIcant reduction [12] R. Viral and D. K. Khatod, "An analytical approach for sizing and
in system losses. siting of DGs in balanced radial distribution networks for lass
minimization," Int. 1. Electr. Power Energy Syst., vol. 67,pp. 191-
REFERENCES 201,2015.
[13] C. Wang, M. H. Nehrir, "Analytical Approaches for Optimal
[I] N. Acharya, P. Mahat, and N. Mithulananthan, "An analytical Placement of Distributed Generation Sources in Power Systems,"
approach for DG aHocation in primary distribution network," Int. 1. IEEE Trans. Power Syst., vol. 19, no. 4,pp. 2068-2076,2004.
Electr. Power Energy Syst., vol. 28, no. 10,pp. 669-678,2006. [14] S. Kansal, V. Kumar, and B. Tyagi, "Optimal placement of
[2] M. M. Aman, G. B. Jasmon, A. H. A. Bakar and H. Mokhlis, "A different type of DG sources in distribution networks," Int. 1.
new approach for optimum DG placement and sizing based on Electr. Power Energy Syst., vol. 53,pp. 752-760,2013.
voltage stability maximization and minimization of power [15] T. Gözel and M. H. Hocaoglu, "An analytical method for the sizing
10sses,"lnt. J.Energy Convers. Manag., vol. 70,pp. 202-210,20I3. and siting of distributed generators in radial systems," Int. 1. Electr.
[3] S Gopiya Naik, D. K. Khatod, and M. P. Sharma, "Analytical Power Syst. Res., vol. 79,pp. 912-918,2009.
approach for optimal siting and sizing of distributed generation in [16] M. H. Haque, "Load flow solution of distribution systems with
radial distribution networks," lET Gener Transm. Distrib., vol. 9, voltage dependent load models," Int. 1. Electr. Power Syst. Res.,
no. 3,pp. 209-220,20I5. vol. 36, no. 3,pp. 151-156,1996.
[4] L. F. Ochoa and G. P. Harrison, "Minimizing energy losses: [17] P. Prakash and D.K. Khatod, "Optimal sizing and siting techniques
Optimal accommodation and smart operation of renewable for distributed generation in distribution system: A review," Int. 1.
Renew. Sustain. Energy Rev vol. 57,pp. II 1-130,2016.
.•

[5]