Ain Shams Engineering Journal (2018) 9, 187–201

Ain Shams University

Ain Shams Engineering Journal

www.elsevier.com/locate/asej
www.sciencedirect.com

ELECTRICAL ENGINEERING

Optimal selection of distributed generating units

and its placement for voltage stability enhancement
and energy loss minimization
Pankita Mehta a, Praghnesh Bhatt a,*, Vivek Pandya b

a
Department of Electrical Engineering, C.S. Patel Institute of Technology, CHARUSAT, Changa, Gujarat, India
b
Department of Electrical Engineering, School of Technology, PDPU, Gandhinagar, Gujarat, India

Received 31 July 2015; revised 24 October 2015; accepted 31 October 2015

Available online 11 December 2015

KEYWORDS Abstract The integration of distributed generation (DG) in distribution network may significantly
Continuation power flow affect its performance. Transmission networks are no longer accountable solely for voltage security
(CPF); issues in distribution networks with penetration of DGs. The reactive power support from the DG
Distributed generation (DG); sources greatly varies with the type of DG units and may potentially distress the larger portion of
Eigenvalue; the network from the voltage stability aspects. This paper presents the analysis for the selection of
Modal analysis; the best type of DG unit among different categories and its optimal location that can enhance the
Participation factor voltage stability of distribution network with simultaneous improvement in voltage profile. Voltage
sensitivity index and bus participation factors derived from continuation power flow and Modal
Analysis, respectively, are used together for voltage stability assessment and placement of DGs.
Changes in mode shapes and participation factors with the placement of DGs are comprehensively
analyzed for 33 and 136 nodes radial distribution network.
Ó 2015 Ain Shams University. Production and hosting by Elsevier B.V. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction occur in a conventional system without generation directly

connected at the distribution level [1]. Attention must be paid
The integration of DG is expected to play an increasingly to effectively eliminate the potentially adverse impacts that
important role in the electric power system infrastructure plan- DG penetration could impress on the electric delivery system.
ning and market operations. The integration of DG at the dis- Inappropriate DG placement and sizing may increase system
tribution network results in operating situations that do not losses and network capital and operating costs. Hence, system-
atic studies and planning are required to locate and operate the
* Corresponding author. Tel.: +91 7567513461. DGs at the distribution level in order to improve voltage pro-
E-mail address: [email protected] (P. Bhatt).
file, to reduce system losses, and to enhance the stability [1–3].
Peer review under responsibility of Ain Shams University.
Power systems are expected to become more heavily loaded
in the forthcoming years as the demand for electric power rises
sharply. The economic and environmental concerns limit the
construction of new transmission and generation capacity.
Production and hosting by Elsevier This may result in deficit of local reactive power support in
http://dx.doi.org/10.1016/j.asej.2015.10.009
2090-4479 Ó 2015 Ain Shams University. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
188 P. Mehta et al.

heavily loaded system which in turn causes the progressive to voltage collapse in radial networks. A method for DG
voltage decline [4–6]. The decline in voltage level is one of placement in radial distribution network based on Voltage
important factors which restrict the increase of load served security margin (VSM) enhancement and loss reduction is
by distribution companies. Therefore, it is necessary to con- presented in [22], which use CPF to identify the most sensitive
sider voltage stability constraints for planning and operation bus to voltage collapse. The work in [23] is suggested that
of distribution systems. The shortage of reactive power can voltage magnitude is not a suitable indicator for the proximity
be compensated by an increased share of DGs in distribution to voltage collapse. Author [24] has described a technique for
systems to improve voltage stability. These DG sources possess selection of buses in a subtransmission system for location of
different reactive power generation capability and their impact DG and determination of their optimum capacities by
on the voltage stability should be studied distinctly. However, minimizing transmission losses and improvement in voltage
these types of studies are least reported in the literature. profile.
The reactive power support using capacitor is one of the A method is presented for locating and sizing of DGs to
methods to reduce the power losses [7]; it is also shown that enhance voltage stability and to reduce network losses simulta-
placement of many capacitors at the incorrect positions will neously in [25], where Modal analysis and continuous power
further increase line losses. The minimization of losses does flow are used for DG placement in the event of reactive power
not give any assurance of the maximization of benefits unless deficit. The enhancement of voltage profile, loadability with
the problem is well formulated [7–9]. However, the active simultaneous reduction distribution losses considering voltage
power support from the capacitors cannot be possible which related constraint is discussed in [26].
have significant impact on system performance. Like capacitors Based on the literature review discussed above, it has been
placement, higher penetration of the DG at the suboptimal observed that the different issues in the field of distributed gen-
location may increase losses in distribution lines. Hence, it is eration are covered which are based on optimal location and
significantly essential to decide the size, location and type of sizing of DG, loss minimization, voltage stability enhancement
DG unit to be placed. Authors in [10,11] revealed that subopti- and reconfiguration of distribution network after the integra-
mal location for DG placement and improper sizing, would tion of DG resources. This paper is mainly focused on voltage
lead to higher losses as compared to the system without DG. stability analysis with different types of distributed generating
The issue of loss minimization in the distribution network is units which were not addressed earlier in the literature. The
reported in [12] where an analytical expression and the method- impact of DG operating in a voltage regulation mode with
ology based on the exact loss formula are being used for the varying reactive power support has been analyzed for the volt-
estimation of the optimal size and location of DG for reducing age stability enhancement. Further, the optimal location of
the total power losses in primary distribution systems. How- DG is ascertained at the weakest bus which is determined by
ever, the proposed methodology is limited to DG type which CPF and modal analysis. The variation in mode shapes with
is capable of delivering real power only. Multiple DG units the variation in DG types and its locations have been compre-
placement using an improved analytical (IA) method to iden- hensively studied for the analysis of voltage stability. The
tify the best location for DG for high loss reduction in large- mode shapes and participation factors are used to identify
scale primary distribution network is given in [13]. The tradi- the weakest region where placement of DG may have the ben-
tional problem of minimizing losses in distribution networks eficiary effect to enhance the voltage stability. The objectives
has been investigated using a single, deterministic demand level of this paper are as follows:
with the help of OPF in [14]. A new long term scheduling for
optimal allocation and sizing of different types of DG units (1) To perform power flow analysis on a Test systems (33-
in the distribution networks in order to minimize power losses nodes radial distribution network [18] and 136-nodes
are presented in [15]. In [16], author discussed the iterative radial distribution [35]) without considering any DG
search method for optimal sizing and siting of single DG for units. These systems are considered as base case test
minimization of cost and loss. The issue of voltage stability in system.
distribution networks and the impact of DG units on it draw (2) To identify the weakest bus for the base case test systems
attention in recent period. A meta-heuristic technique based by performing continuation power flow and Modal
on backtracking search optimization algorithm (BSOA) is pre- analysis.
sented in [17] with objectives to reduce power loss and improve (3) To determine and rank the buses which are more prone
voltage profile in RDN considering the DG type which can to voltage instability based on the mode shapes partici-
inject active and reactive power both. In [18], BSOA and set pation factors obtained through the Modal analysis.
of fuzzy expert roles using loss sensitivities factors and bus volt- (4) To ascertain the type of DG units and its placement
ages are used to identify the DG locations. The Optimal size methodology in a system to enhance voltage stability
and location of multiple DGs are determined with particle margin.
swarm optimization using constriction factor approach in [19] (5) To compare the impact of different types of DG units to
considering predetermines annual load growth with voltage improve the performance of distribution system.
regulation as constraint. Simultaneous placement of DG and
fixed capacitor in radial distribution network with time varying
2. DG technologies and their impact on voltage stability
load in order to decrease reactive power loss, energy and power
loss reduction, improvement of voltage profile and voltage sta-
bility has been discussed in [20]. DG technologies are classified based on their capability of
A new voltage stability index is introduced in [21] by injecting real and/or reactive power in the system. Accord-
simplified load flow equations to seek the most sensitive buses ingly, DG technologies are grouped in the following manner.
Optimal selection of generating units and its placement 189

2.1. Type 1: DG capable of injecting both active power (P) and information from a system wide perspective and clearly identi-
reactive power (Q) fies the areas that are prone to voltage instability. In addition,
it gives the information about the voltage instability character-
DG units that are based on synchronous machine for small istic, what are the key factors influencing the voltage instability
hydro, geothermal, and combined cycles fall in Type 1 cate- and which measures are the most effective to enhance the volt-
gory. The DG with the synchronous generators may be mod- age stability [30–32].
eled either with constant terminal voltage control (voltage Continuation power flow (CPF) is a tool for steady state
control mode) or with constant power factor control (power voltage stability analysis which reformulates the conventional
factor control). The DGs with the voltage control mode are power flow equations so that they remain well-conditioned at
considered as PV nodes and DGs with the power factor con- all possible loading conditions. The CPF can successfully
trol mode are considered as PQ nodes [27]. In this work, the determine a distance from an operating point to a voltage col-
DG with synchronous generator of Type 1 category is modeled lapse which may be measured in terms of physical quantities
as PV nodes. such as load level, active power flow and reactive power gener-
ation/absorption capability of network [33].
2.2. Type 2: DG capable of injecting active power (P) only
3.1. Modal analysis
Photovoltaic (PV), micro turbines, fuel cells, which are inte-
grated to the main grid with the help of converters/inverters Incremental voltage of the buses from voltage stability view
[28,29] are the examples of Type 2 category. In this work, it point is of great importance, and also useful to identify the
is assumed that DG units in this category neither absorb nor buses for addition of DG. The incremental power flow equa-
deliver reactive power to system and operate with unity power tions are used for obtaining incremental voltage sensitivities
factor only. [30,34] as given in (1) where elements of Jacobian matrix gives
the sensitivity between power flow bus voltage changes.




2.3. Type 3: DG capable of injecting reactive power (Q) only DP JPh JPV Dh
¼ ð1Þ
DQ JQh JQV DV
The DG units equipped with synchronous compensator are
where DP = incremental change in bus real power;
considered as Type 3 category.
DQ = incremental change in bus reactive power injection;
Dh = incremental change in bus voltage angle; DV = incre-
2.4. Type 4: DG capable of injecting active power (P), but
consuming reactive power (Q) mental change in bus voltage magnitude. JPh ¼ @P @h
@P
, JPV ¼ @V ,
@Q @Q
JQh ¼ @h and JQV ¼ @V are the elements of Jacobian. Voltage
Fixed speed squirrel cage induction generator (SQIG) used for stability of system is affected by both active and reactive
wind turbine generating (WTG) system falls under this cate- power. However at each operating point active power P is kept
gory. SQIG in super-synchronous mode is capable of injecting constant and voltage stability is evaluated by considering the
real power in the system whereas it demands reactive power incremental relationship between reactive power Q and bus
from the system. voltage magnitude V.
Thus, it is worthwhile to note that the type of DG technol- Considering, DP ¼ 0; the incremental relationship between
ogy adopted will have a significant bearing on the performance reactive power Q and bus voltage magnitude V is given as (2):
of distribution network. The installation of synchronous DV ¼ J1
R DQ ð2Þ
machine-based DG units that are close to the loads can lead
to beneficial impact on system voltage stability margin; on where the reduced Jacobian Matrix ðJR Þ is given by (3):
another end, the case with an induction generator may have JR ¼ ½JQV  JQh J1
Ph JPV  ð3Þ
detrimental impact on the system stability margin. Therefore,
it is an utmost requirement to analyse the effect of different Let,
types of DG technologies on the voltage stability to enjoy JR ¼ nKg ð4Þ
the system wide benefits.
where n and g are right and left eigenvector matrices of JR,
respectively. K is diagonal eigenvalue matrix of JR; Inverting
3. Modal analysis and continuation power flow for voltage (4) yields
stability analysis 1
J1
R ¼ nK g ð5Þ
The penetration of different types of DGs at the distribution and substituting (5) in (2) results in
network greatly affects the voltage profile and voltage stability
DV ¼ nK1 gDQ ð6Þ
limits due to their different reactive power generation capabil-
ity. While integrating DGs in the system, it is not only impor- X ni gi
tant to maintain acceptable voltage magnitudes, but it is also DV ¼ DQ ð7Þ
i
ki
crucial to ensure that the steady-state loading should not reach
the critical state at which voltage collapse may occur. where gi is the ith row of the left eigenvector of JR, and ni is
Modal analysis is a static approach that captures the snap- the ith column of the right eigenvector. The ith mode of the
shots of the system conditions at various time frames along the Q–V response is defined by the ith eigenvalue ki and the corre-
time domain trajectories. It gives the voltage stability related sponding left and right eigenvectors gi &ni .
190 P. Mehta et al.

Since, n1 ¼ g, equation (6) may be written as follows: PL ¼ kPL0 ð12Þ

1
gDV ¼ K gDQ ð8Þ QL ¼ kQL0 ð13Þ
Define v ¼ gDV as the vector of modal voltage variations where k is the loading parameter which ranges from
and q ¼ gDQ as the vector of modal reactive power variations, 0 6 k 6 kcritical . PL0 and QL0 are active and reactive loads,
k
so (8) can be written as follows: respectively, in the base case. The generated active power
v ¼ K1 q ð9Þ (PG ) by the generators is also increased simultaneously with
the same scaling factor to cater the increased load demand
Thus for the ith mode, the vector for the modal voltage and it is given by (14):
variations can be written as
PG ¼ kPG0 ð14Þ
1
vi ¼ q ð10Þ
ki i
4. Optimal placement of DG units based on voltage stability
The magnitude of ki determines the degree of stability of
enhancement
the ith modal voltage. The smaller the magnitude of positive
ki , ith modal voltage is much closer to be unstable. The relative
contribution of the power at bus k in mode i is given by the bus The 33- and 136-node radial distribution systems given in
participation factor: [18,35], respectively, are used for the power flow analysis and
to propose a methodology for placement of the different types
Pki ¼ nki gki ð11Þ of DG units with the objectives to reduce power losses,
Pki indicates the contribution of the ith eigenvalue to the V-Q improve voltage profile and enhance the voltage stability mar-
sensitivity at bus k. The participation factors determine the gin of the system. The total active and reactive loads for 33-
most critical areas which lead the system to instability. node RDN are 3715 kW and 2300 kvar, respectively. For
136-node RDN, the active and reactive loads are 18214 kW
3.2. Continuation power flow and 7933 kvar, respectively.

4.1. Steps for DG placement

The continuation power flow is an iterative process that
employs a predictor–corrector scheme as shown in Fig. 1a to
find a solution path of a set of power flow equations that have The placement of different types of DG in distribution net-
been reformulated to include a load parameter [33,34]. The work greatly affects the voltage profile at the different buses
process starts from a known solution (i.e. from Point A corre- and power flow in feeders, thus alters the active and reactive
sponding to base case loading kk ) and uses a tangent predictor power losses in a system. The variation in voltage profile and
to estimate a subsequent solution corresponding to a different losses in a system also varies with the different types of DG
value of the load parameter (i.e. point B). This estimate is then technologies discussed in Section 2. Hence, the attention must
corrected using a conventional power flow method with a sys- be paid not only to decide the location for DG placement but
tem load assumed to be fixed (i.e. Point C). The CPF method the types of DG technologies need to be considered. In this
stresses the system up to kcritical which corresponds to critical work, voltage stability enhancement is considered to be the
k
loading in the system where the voltage instability occurs. In major criteria for the DG placement to ensure the stable oper-
this paper, the increase in active and reactive load takes the ation of the system with acceptable voltage levels at the con-
form as given in (12) and (13): sumer nodes. The procedure adopted to find out the optimal
locations for DG placement along with selection of different

Predictor Base Case

A Loading
B Corrector
A Stressed
B
C Loading
Exact Solution D
Bus Voltage

Bus Voltage

E
C
Critical Point Critical Point
F
G

Loading, λ Loading, λ

Figure 1a Sequence of calculations in CPF [5]. Figure 1b Typical PV curve.

Optimal selection of generating units and its placement 191

types of DG technologies in a given test system is shown in

Table 1 Indices for the performance measurement of test
Fig. 2.
system.

5. Definition of indices for performance assessment of test Penetration level of DG % Penetration Level ¼ PPLoad
DG
 100%
Voltage sensitivity index Vbase case
Vcritical
systems VSI for bus k ¼ Pn k k

i¼1
ðVbase
i
case
Vcritical
i Þ
Active power loss indices P PDG
APLI ¼ LossPLossLoss  100%
To investigate the impact of placement of different types of
Reactive power loss indices Q QDG
DG units on the power losses, voltage profile and voltage sta- QPLI ¼ LossQLoss Loss  100%
P
bility, different performance indices have been defined and are Aggregate voltage deviation AVDI ¼ ni¼1 jVDG i  1j
summarized in Table 1 as follows. index
Voltage stability margin VSM ¼ kmax
j  kk
5.1. Penetration level of DG Active power transfer APTC ¼ PBase PDG þPL PDGL
PBase  100
capability
Reactive power transfer RPTC ¼ QBase QDG þQL QDGL
QBase  100
In a radial distribution system, the load demand is supplied capability
only by the source at one end (i.e. Sub-station). With the inte-
gration of DG at any node in a distribution network, the net
demand from the source is reduced and it affects the perfor-
mance of distribution network. Authors in [16] discussed the
effect of penetration level of DG on the voltage profile of minimization. The ratio of active power generated by DG
the buses whereas authors in [12] addressed the issue of size ðPDG Þ to total active power demand ðPLoad Þ is defined as pene-
selection of DG and its placement on the basis of loss tration level of DG in a distribution network and it is given by
(15):
PDG
Start % Penetration Level ¼  100% ð15Þ
PLoad

Run PF, CPF and Modal 5.2. Active and reactive power loss indices
Analysis without DG

The impact of DG placement on active power loss index

Ranking of weakest buses ðAPLIÞ and reactive power loss index ðQPLIÞ can be assessed
based on CPF and Modal with the use of (16) and (17), respectively [22]. Optimal DG
Analysis placement should result in higher loss indices.
Active power loss reduction; APLI
Place either Type1 or Type 2 or PLoss  PDG
Type 3 DG on Candidate Bus ¼ Loss
 100% ð16Þ
PLoss
Replace Next
DG Type candidate bus Reactive power loss reduction; QPLI
Run PF, CPF and Modal
QLoss  QDG
Analysis with DG ¼ Loss
 100% ð17Þ
QLoss

No 5.3. Voltage sensitivity index (VSI)

Is this a last candidate bus?

VSI gives the direct indication of maximum bus voltage devi-

Yes
ation at the point of voltage collapse. The bus with highest
No VSI could be identified as the ‘‘weakest bus” in a system.
Are all DG Type Considered?
Vbase case
 Vcritical
VSI for bus k ¼ Pn k k
ð18Þ
i¼1 ðVi  Vcritical Þ
base case
Yes i

Suggest location and type of DG

with highest VSM for voltage 5.4. Voltage stability margin (VSM)
stability enhancement
The VSM gives the relative distance from the current operating
point to the point of voltage collapse. Higher values of VSM
Calculate All Indices ensure that the system is capable of carrying more loading
and Stop without losing its stability. The VSM can be represented by
(19) as follows:
Figure 2 Flowchart for selection of DG type and location for VSM ¼ kcritical  kk ð19Þ
voltage stability enhancement.
192 P. Mehta et al.

Here, kk is the reference loading, which is assumed to be to critical loading of kcritical ¼ 3:40045. For 136-nodes RDN,
zero. kcritical is the loading at the point of voltage instability. Bus 118 is found to be the most critical bus with a voltage
of 0.5171 pu corresponding to critical loading of
5.5. Aggregate voltage deviation index (AVDI) kcritical ¼ 3:68643.
The results of the Modal analysis in the base case for both
AVDI gives the aggregate bus voltage deviation after the opti- test systems are given in Table 2 along with the eigenvalues for
mal DG placement in test system. The lowest value of AVDI the most associated buses corresponding to three operating
indicates more improvement in bus voltage profile at all the points A, B and C as shown in Fig. 1b. It is observed that with
buses after the placement of DG as compared to base case. the increment in loading, the eigenvalues start reducing and the
AVDI is given by (20): least eigenvalues are obtained for most stressed condition at
point C. The buses with minimum eigenvalue are ranked for
X
n
AVDI ¼ jVDG  1j ð20Þ both RDNs and they are determined as the most critical buses
i
i¼1 from voltage stability point of view. The top ten bus participa-
tion factors contributed in these most critical modes in base
5.6. Power transfer capacity (PTC) of distribution network case for both systems corresponding to point C are listed in
Tables 3 and 4. The bus participation factors present their per-
centage contribution to the respective modes; thus, they clearly
The increase in electric demand requires to enhance the power
reveal the order of buses which need more attention for voltage
transfer capability of the existing distribution network to
stability enhancement.
maintain reliability and continuity of supply. Otherwise the
It can be observed that some of the buses repeatedly con-
new construction of distribution lines is required, which is
tribute to critical modes which represent the non-localized
not a practical solution always. Distributed generation can
behavior of the modes and determine the weak region in the
only provide the alternatives to strengthen the existing distri-
system. The existence of such a weak region clearly indicates
bution network without addition of new lines. The power
that the reactive power support to this region is exhausted
transfer capacity of distribution network with the placement
and attention must be given to ensure availability of reactive
of DG units can be assessed as follows [22]. The active and
power for maintaining voltage stable operation of the system.
reactive power transfer capacities are given by (21) and (22),
The comparison between Voltage Sensitive Index (VSI)
respectively:
obtained with CPF and the bus participation factors corre-
PBase  PDG þ PLoss  PDG;Loss sponding to the least stable modes (i.e. Point C) is represented
APTC ¼  100 ð21Þ
PBase in Table 5 for the base case test system. The results of CPF and
Modal analysis follow the same order of the voltage sensitivity;
QBase  QDG þ QLoss  QDG;Loss thus, the ranking of the buses suggested by both analyses
RPTC ¼  100 ð22Þ
QBase together can be confirmed to identify the buses which are most
prone to voltage instability. Hence, Bus 18 and Bus 118 are
where PBase , QBase are active and reactive power of the slack bus
identified as the best choice for DG unit placement to enhance
in base case, respectively; PDG , QDG are active and reactive
the voltage stability for 33- and 136-node RDN, respectively.
power of the slack bus with DG unit, respectively; PLoss ,
Further, the aim of this paper is to determine the optimal loca-
QLoss are active and reactive power losses in the distribution
tion of DG units that can improve the overall performance of
lines in base case, respectively; PDG;Loss , QDG;Loss are active
the system in terms of indices given in Table 1. Hence, place-
and reactive power losses in distribution lines with DG units,
ment of DG units is also analysed for the subsequent buses
respectively.
in the ranking with comparative lower voltage sensitivity in
order to evaluate overall system performance.
6. Results and discussions
6.2. Determination of type of DG units
In this work, placements of different types of DG discussed in
Section 2 are considered for the analysis. Power flow, continu- The characteristic of three types of DG units for reactive
ation power flow and Modal analysis are performed to select power compensation in order to enhance the voltage stability
the best type among different DG units and to ensure its has been discussed in Section 2. Type 1 and Type 3 DG units
optimal location that can enhance the voltage stability. The comprised of synchronous generators and synchronous com-
penetration level of 26% is considered for both test systems. pensators, respectively, are modeled to deliver reactive power
i.e. 33-nodes and 136-nodes RDNs. in normal and emergency condition. They will operate in volt-
age regulation mode and maintain their voltage to nominal
6.1. Determination of optimal location for DG placement value. Type 2 DG units comprise of PV cell which will neither
supply nor absorb any reactive power.
The CPF is carried out by assuming base case loading equal to Any one type of DG unit is placed at the candidate buses
zero. The active and reactive power loads are increased with determined by CPF and Modal analysis at a time and its influ-
constant power factor until the voltage collapse. Fig. 3 shows ence on the variation of critical modes have been analysed.
the bus voltage profile of the 33- and 136-node RDN obtained Figs. 4 and 5 show the comparison between eigenvalue
with CPF, where all bus voltages correspond to voltage obtained for both the systems without and with the placement
collapse point. Bus 18 is found to be the most critical bus in of different types of DG units on the candidate buses. The
33-nodes RDN with a voltage of 0.35954 pu corresponding results clearly show that the placement of any type of DG unit
Optimal selection of generating units and its placement 193

1 1

0.9 33-Nodes RDN

0.9
0.8
at Critical Loading

at Critical Loading
Bus Voltages (pu)

Bus Voltages (pu)

0.8
0.7 136-Nodes RDN

0.6
0.7

0.5
0.6
0.4

0.5
0 5 10 15 20 25 30 35 0 20 40 60 80 100 120 140
Bus Numbers Bus Numbers
(a ) 33-Nodes RDN (b) 136-Nodes RDN
Figure 3 Bus voltages profile of test systems corresponding to critical loading without placement of DG.

Table 2 Eigenvalues for most critical buses at different operating points.

33 node RDN 136 node RDN
Bus no. Eigenvalues at different operating points Bus no. Eigenvalues at different operating points
Point A Point B Point C Point A Point B Point C
Bus 18 0.21498 0.1807 0.08446 Bus 118 0.02264 0.01904 0.00753
Bus 33 0.69787 0.645 0.4572 Bus 38 0.04461 0.04255 0.03896
Bus 18 2.25493 2.06237 1.53569 Bus 62 0.04546 0.04364 0.04050
Bus 22 2.74458 2.71166 2.65266 Bus 97 0.05750 0.05517 0.05124
Bus 25 4.17262 3.95781 3.38813 Bus 15 0.06110 0.05809 0.05265

Table 3 Bus participation factors for most critical buses obtained with modal analysis for 33-nodes RDN.
Bus participation factors for most critical buses obtained with modal analysis
k1 ¼ 0:08446 k2 ¼ 0:4572 k3 ¼ 1:53569 k4 ¼ 2:65266 k5 ¼ 3:38813
Bus 18 Bus 33 Bus 18 Bus 22 Bus 25
Bus 18 0.1279 Bus 33 0.1824 Bus 18 0.1954 Bus 22 0.4504 Bus 25 0.1668
Bus 17 0.1246 Bus 32 0.1737 Bus 17 0.1575 Bus 21 0.3157 Bus 24 0.1106
Bus 16 0.1100 Bus 31 0.1608 Bus 10 0.1030 Bus 20 0.2112 Bus 26 0.0633
Bus 15 0.1027 Bus 30 0.1142 Bus 11 0.1018 Bus 25 0.0041 Bus 27 0.0622
Bus 14 0.0946 Bus 29 0.0983 Bus 09 0.0966 Bus 19 0.0040 Bus 06 0.0606
Bus 13 0.0827 Bus 28 0.0598 Bus 12 0.0957 Bus 24 0.0030 Bus 33 0.0545
Bus 12 0.0603 Bus 18 0.0386 Bus 08 0.0722 Bus 23 0.0014 Bus 07 0.0531
Bus 11 0.0569 Bus 17 0.0359 Bus 33 0.0264 Bus 05 0.0009 Bus 18 0.0489
Bus 10 0.0551 Bus 16 0.0231 Bus 07 0.0239 Bus 04 0.0008 Bus 23 0.0385
Bus 09 0.0404 Bus 27 0.0215 Bus 13 0.0233 Bus 03 0.0007 Bus 32 0.0381

on any of the candidate buses significantly increases the different DGs at the selected candidate buses is shown in
magnitude of the critical modes in positive direction which Fig. 6. It is noticed that the placement of any type of DG at
helps the system to keep away from the voltage instability. any of the candidate bus can successfully improve the VSM.
The ranking for the most associated buses in critical modes But improvements in VSM are quite noticeable only when
in Figs. 4 and 5 with the placement of different DGs can be DGs are placed at first two candidate buses as determined
identified as given in Table 6. by CPF and Modal analysis. On another end, the improve-
To determine the best types of DG for voltage stability ments in VSM are marginal when DGs are placed at lower
enhancement, different types of DGs are placed at the critical ranked buses. Hence, it is proposed to place the DGs unit at
buses and the improvement in VSM has been observed. The the bus with the least mode shape in the base case to gain
comparison between VSM values after the placement of the maximum benefit of the voltage stability enhancement.
194 P. Mehta et al.

Table 4 Bus participation factors for most critical buses obtained with modal analysis for 136-nodes RDN.
Bus participation factors for most critical buses obtained with modal analysis
k1 ¼ 0:00753 k2 ¼ 0:04050 k3 ¼ 0:03896 k4 ¼ 0:05124 k5 ¼ 0:05265
Bus 118 Bus 38 Bus 62 Bus 97 Bus 15
Bus 118 0.07245 Bus 38 0.05929 Bus 62 0.06596 Bus 97 0.11639 Bus 15 0.08417
Bus 117 0.07222 Bus 35 0.05875 Bus 61 0.06562 Bus 96 0.11429 Bus 17 0.08392
Bus 116 0.0691 Bus 37 0.05871 Bus 60 0.06444 Bus 95 0.11075 Bus 16 0.08287
Bus 110 0.06882 Bus 34 0.05845 Bus 59 0.06154 Bus 99 0.1093 Bus 14 0.08217
Bus 115 0.06861 Bus 39 0.05735 Bus 58 0.05839 Bus 98 0.10717 Bus 13 0.08156
Bus 109 0.06695 Bus 36 0.05718 Bus 57 0.05614 Bus 94 0.10406 Bus 12 0.08038
Bus 114 0.06607 Bus 33 0.05677 Bus 56 0.05336 Bus 93 0.09194 Bus 11 0.07916
Bus 113 0.06582 Bus 32 0.05624 Bus 55 0.05324 Bus 92 0.08359 Bus 10 0.07613
Bus 112 0.0648 Bus 31 0.05606 Bus 54 0.05288 Bus 91 0.05155 Bus 9 0.07462
Bus 111 0.06384 Bus 30 0.05546 Bus 53 0.05176 Bus 90 0.04476 Bus 8 0.07312

Table 5 Comparison of results obtained with CPF and modal analysis for base case system.
33-nodes RDN 136-nodes RDN
CPF analysis Modal analysis CPF analysis Modal analysis
Bus no. VSI Bus no. BPF Bus no. VSI Bus no. BPF
Bus 18 0.05920 Bus 18 0.12793 Bus 118 0.0301 Bus 118 0.0725
Bus 17 0.05870 Bus 17 0.12458 Bus 117 0.0301 Bus 117 0.0722
Bus 16 0.05701 Bus 16 0.11002 Bus 116 0.0293 Bus 116 0.0691
Bus 15 0.05589 Bus 15 0.10268 Bus 110 0.0293 Bus 110 0.0688
Bus 14 0.05475 Bus 14 0.09461 Bus 115 0.0293 Bus 115 0.0686
Bus 13 0.05296 Bus 13 0.08266 Bus 109 0.0289 Bus 109 0.0670
Bus 12 0.04823 Bus 12 0.06029 Bus 114 0.0285 Bus 114 0.0661
Bus 11 0.04712 Bus 11 0.05694 Bus 113 0.0285 Bus 113 0.0658
Bus 10 0.04650 Bus 10 0.05511 Bus 112 0.0284 Bus 112 0.0648
Bus 33 0.04432 Bus 09 0.04036 Bus 111 0.0283 Bus 111 0.0638

Also, Fig. 6 shows that VSM has been enhanced greatly placement of DGs at buses with lower ranking of voltage sen-
with the placement of synchronous generator on the most volt- sitivity. The placement of synchronous generator depicts the
age sensitive buses. This indicates that the selection of syn- best performance for improvement in voltage profile among
chronous generator as DG units on most voltage sensitive all DG types.
buses can carry more load before becoming voltage unstable. For 136-nodes RDN, the improvement in bus voltage pro-
The capability of synchronous generator of delivering both file almost equally affected for the nodes in lateral feeder where
active and reactive power makes it more promising for enhanc- DG has been placed. For the main and other lateral feeders,
ing the voltage stability margin. On the other end, the VSM the placement of DGs cannot alter the voltage profile much.
with synchronous compensator is less due to its inability to It is observed from Fig. 7b that the type of DG units brings
supply active power. The load relies only on the source gener- different variations in bus voltage profile. The placements of
ation (i.e. sub-station); hence, larger value of the current fed DGs at location other than Bus 118 are not capable to prevent
right from the source has been responsible for greater real the drastic voltage drop in valley near buses 100–120. The per-
and reactive power losses in the distribution feeders. Thus, formances of fuel cell and synchronous compensators have
the system becomes voltage unstable earlier as compared to caused more variations in bus voltages as compared to syn-
that of synchronous generator. chronous generator. The DGs with reactive power generation
Figs. 7a and 7b and Figs. 8a and 8b show the comparative capability are modeled in voltage regulation mode; hence, they
bus voltage profile with the placement of all types of DGs on try to maintain nominal voltage of 1.0 pu at its node by sup-
each candidate bus for the normal as well as stressed loading plying/absorbing reactive power. Therefore, the reactive power
condition, respectively. For the stressed loading condition in generation by DG units which affect the performance of sys-
Figs. 8a and 8b, the bus voltage profile are plotted nearer to tem is noted in Table 7. For 136-nodes RDN, the synchronous
kcritical ¼ 3:4005 and kcritical ¼ 3:6863, respectively, for 33- generator at Bus 118 absorbs (1348) kvar to maintain its ter-
nodes and 136-nodes RDNs. These given values of k corre- minal voltage to nominal value of 1.0 pu; hence, its result is
spond to kcritical for the base case test systems without any inferior as compared to synchronous compensator and fuel
DGs. For 33-nodes RDN, the placement of DGs on the first cell. The synchronous compensator provides reactive power
two candidate buses can significantly improve the bus voltages support of 2059 kvar to maintain the terminal voltage of 1.0
profile in comparison with base case which can be seen in pu. Even if the fuel cell neither generates nor absorbs any reac-
Fig. 7a whereas improvement is not much noticeable with tive power, its placement raises the voltage up to 1.0478 which
Optimal selection of generating units and its placement 195

5 5

4 Without DG 4
With Syn Gen
Eigenvalue

Eigenvalue
With Syn Com
3 With Fuel Cell
3

2 2

1 1

0 0
1 2 3 4 5 0 1 2 3 4 5 6
Critical Modes with Placement of DGs at Bus 18 Critical Modes with Placement of DGs at Bus 33

5 4

4
3
Eigenvalue

Eigenvalue
3
2
2
1
1

0 0
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Critical Modes with Placement of DGs at Bus 22 Critical Modes with Placement of DGs at Bus 25

Figure 4 Eigenvalues obtained with the placement of different DG units at the candidate buses for 33-nodes RDN.

0.06 0.06
Without DG
0.05 0.05 Syn Gen
Syn Com
Eigenvalue

Eigenvalue

0.04 0.04 Fuel Cell

0.03 0.03

0.02 0.02

0.01 0.01

0 0
1 2 3 4 5 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5
Critical Modes with Placement of DGs at Bus 118 Critical Modes with Placement of DGs at Bus 38

0.06 0.06

0.05 0.05
Eigenvalue

Eigenvalue

0.04 0.04

0.03 0.03

0.02 0.02

0.01 0.01

0 0
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5
Critical Modes with Placement of DGs at Bus 62 Critical Modes with Placement of DGs at Bus 97

Figure 5 Eigenvalues obtained with the placement of different DG units at the candidate buses for 136-nodes RDN.

is also not a desirable condition. This comprehensive analysis respectively, for 33- and 136-nodes RDNs on voltage stability
suggests that while selecting the DG units to achieve the par- enhancement are clearly depicted in Figs. 8a and 8b where the
ticular objectives, the constraints must be designed very care- reductions in bus voltages are limited even for the stressed
fully and more powerful solution techniques should be loading conditions.
adopted. The size of the distribution network, number of lat-
eral branches, loading on different nodes, load models and 6.3. Determination of multiple location and type of DG
active and reactive power generation capability of DG units
significantly affect the performance of the system. The benefi- The placement of first DG unit on the most voltage sensitive
cial impact of placement of DGs on Bus 18 and Bus 118, bus determined by CPF and Modal analysis results in greater
196 P. Mehta et al.

Table 6 Identification of critical modes for most associated buses with the placement of DG.
33 RDN 136 RDN 33 RDN 136 RDN 33 RDN 136 RDN 33 RDN 136 RDN 33 RDN 136 RDN
Without Without DG at Bus DG at Bus DG at Bus DG at Bus DG at Bus DG at Bus DG at Bus DG at Bus
DG DG 18 118 33 38 22 62 25 97
18 118 33 38 18 118 18 118 18 118
33 38 12 35 18 62 33 38 33 38
18 62 22 37 22 97 18 97 18 97
22 97 25 34 28 15 25 15 22 15
25 15 25 39 25 73 25 73 27 73

7 12
Without DG
Without DG
6
Voltage Stability Margin (VSM)

Voltage Stability Margin (VSM)

With Syn Gen 10 With Syn Gen
With Syn Com
With Syn Com
5 With Fuel Cell With Fuel Cell
8
4
6
3
4
2

2
1

0 0
Bus 18 Bus 33 Bus 22 Bus 25 Bus 118 Bus 38 Bus 62 Bus 97

Figure 6 Comparative voltage stability margin (VSM) without and with DG units for 33- and 136-nodes RDNs.

DG Placed at 18 DG Placed at 33
1 1

0.98 0.98
Bus Voltages (pu)

Bus Voltages (pu)

0.96 0.96

0.94 0.94

0.92 0.92

0.9 0.9
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35
Bus Numbers Bus Numbers

DG Placed at 22 DG Placed at 25
1 1

Without DG
0.98 0.98
Bus Voltages (pu)

Bus Voltages (pu)

Syn Gen
Syn Com
0.96 Fuel Cell
0.96

0.94 0.94

0.92 0.92

0.9 0.9
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35
Bus Numbers Bus Numbers

Figure 7a Comparative bus voltage profile of 33-nodes RDN without and with DGs for normal loading condition.
Optimal selection of generating units and its placement 197

DG Placed at Bus 118 DG Placed at Bus 38

1.04
1.02 1
Bus Voltage (pu)

Bus Voltage (pu)

1
0.98
0.96 0.95

0.94
0.92
0.9
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
Bus Numbers Bus Numbers

DG Placed at Bus 62 DG Placed at Bus 97

1.02
Bus Voltage (pu)

Bus Voltage (pu)

1
1
0.98

0.96
0.95 Without DG
With Syn Gen 0.94
With Syn Com 0.92
With Fuel Cell
0.9
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
Bus Numbers Bus Numbers

Figure 7b Comparative bus voltage profile of 136-nodes RDN without and with DGs for normal loading condition.

DG Placed at Bus 18 DG Placed at Bus 33

1
1
Stressed Loading (pu)

Stressed Loading (pu)

Bus Voltages with

Bus Voltages with

0.8
0.8

0.6 0.6
Without DG
0.4 With Syn Gen
0.4
With Syn Com
With Fuel Cell
0.2 0.2
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35
Bus Numbers Bus Numbers

DG Placed at Bus 22 DG Placed at Bus 25

1 1
Stressed Loading (pu)

Stressed Loading (pu)

Bus Voltages with

Bus Voltages with

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35
Bus Numbers Bus Numbers

Figure 8a Comparative bus voltage profile of 33-nodes RDN without and with DGs for stressed loading condition.
198 P. Mehta et al.

DG Placed at Bus 118 DG Placed at Bus 38

with Stressed Loading (pu)

with Stressed Loading (pu)

1 1
Bus Voltages

Bus Voltages
0.8 0.8

Without DG
0.6 With Syn Gen 0.6
With Syn Com
With Fuel Cell

0.4 0.4
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
Bus Numbers Bus Numbers

DG Placed at Bus 62 DG Placed at Bus 97

with Stressed Loading (pu)

with Stressed Loading (pu)

1 1
Bus Voltages

Bus Voltages
0.8 0.8

0.6 0.6

0.4 0.4
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
Bus Numbers Bus Numbers

Figure 8b Comparative bus voltage profile of 136-nodes RDN without and with DGs for stressed loading condition.

Table 7 Active and reactive power generation without and with DGs for 33- and 136-nodes RDNs.
Synchronous generator Synchronous compensator Fuel cell
P (kW) Q (kvar) V (pu) P (kW) Q (kvar) V (pu) P (kW) Q (kvar) V (pu)
DG placed at Bus 18 (33-nodes RDN)
1115 139 1 0 1592 1 1115 0 0.99091
DG Placed at Bus 118 (136-nodes RDN)
4761 1348 1 0 2059 1 4761 0 1.0478

Table 8 VSI with placement of different types of DG units on candidate buses for 33-nodes RDN.
VSI with DG placed at Bus 18 VSI with DG placed at Bus 33
Bus no. Syn gen Syn com Fuel cell Bus no. Syn gen Syn com Fuel cell
Bus 33 0.0841 0.0632 0.0639 Bus 18 0.0822 0.0663 0.0656
Bus 32 0.0837 0.063 0.0636 Bus 17 0.0813 0.0658 0.065
Bus 31 0.0823 0.0623 0.0627 Bus 16 0.0782 0.0641 0.0629
Bus 30 0.0762 0.0591 0.0584 Bus 15 0.0762 0.063 0.0616
Bus 29 0.0712 0.0563 0.055 Bus 14 0.0741 0.0619 0.0603
VSI with DG placed at Bus 22 VSI with DG placed at Bus 25
Bus no. Syn gen Syn com Fuel cell Bus no. Syn gen Syn com Fuel cell
Bus 18 0.0585 0.0596 0.0592 Bus 18 0.062 0.0599 0.0591
Bus 17 0.0581 0.0591 0.0587 Bus 17 0.0615 0.0594 0.0586
Bus 16 0.0565 0.0574 0.0571 Bus 16 0.0596 0.0578 0.057
Bus 15 0.0554 0.0562 0.056 Bus 15 0.0584 0.0567 0.0559
Bus 14 0.0543 0.0551 0.0549 Bus 14 0.0572 0.0556 0.0548
Optimal selection of generating units and its placement 199

Table 9 BPF with placement of different types of DGs for 33-nodes RDN.
BPF with DG placed at Bus 18 BPF with DG placed at Bus 33
Bus no. Syn gen Syn com Fuel cell Bus no. Syn gen Syn com Fuel cell
Bus 33 0.1518 0.1606 0.148 Bus 18 0.1459 0.14311 0.1459
Bus 32 0.1459 0.1541 0.1424 Bus 17 0.1422 0.13938 0.1422
Bus 31 0.1372 0.1446 0.1339 Bus 16 0.1239 0.12205 0.1239
Bus 30 0.1048 0.1095 0.1025 Bus 15 0.1153 0.11363 0.1153
Bus 29 0.0933 0.0973 0.0913 Bus 14 0.1054 0.10416 0.1054
Bus 28 0.0641 0.0658 0.0629 Bus 13 0.0905 0.09 0.0905
BPF with DG placed at Bus 22 BPF with DG placed at Bus 25
Bus no. Syn gen Syn com Fuel cell Bus no. Syn gen Syn com Fuel cell
Bus 18 0.1278 0.1279 0.1278 Bus 18 0.1283 0.12826 0.1283
Bus 17 0.1245 0.1246 0.1245 Bus 17 0.1251 0.12499 0.1251
Bus 16 0.11 0.11 0.11 Bus 16 0.1105 0.11039 0.1105
Bus 15 0.1027 0.1027 0.1027 Bus 15 0.1032 0.10311 0.1032
Bus 14 0.0946 0.0946 0.0946 Bus 14 0.0952 0.09503 0.0952
Bus 13 0.0827 0.0827 0.0827 Bus 13 0.0831 0.083 0.0831

Table 10 VSI with placement of different types of DG units on candidate buses for 136-nodes RDN.
VSI with DG placed at Bus 118 VSI with DG placed at Bus 38
Syn gen Syn com Fuel cell Bus no. Syn gen Syn com Fuel cell
#
0.01794 (106) 0.01120 (15) 0.0222 (113) Bus 117 0.03186 0.03339 0.03190
0.01733 (107) 0.01104 (17) 0.0222 (114) Bus 118 0.03186 0.03339 0.03190
0.01669 (113) 0.01102 (13) 0.0221 (107) Bus 110 0.03104 0.03252 0.03108
0.01669 (114) 0.01097 (85) 0.0221 (112) Bus 115 0.03099 0.03247 0.03103
0.01654 (112) 0.01090 (16) 0.0221 (106) Bus 116 0.03099 0.03247 0.03103
VSI with DG placed at Bus 62 VSI with DG placed at Bus 97
Bus no. Syn gen syn com Fuel cell Bus no. Syn gen Syn com Fuel cell
Bus 117 0.03015 0.03279 0.03029 Bus 117 0.0334 0.0308 0.0338
Bus 118 0.03015 0.03279 0.03029 Bus 118 0.0334 0.0308 0.0338
Bus 110 0.02937 0.03195 0.02951 Bus 110 0.0325 0.0300 0.0330
Bus 115 0.02933 0.03189 0.02946 Bus 115 0.0325 0.0300 0.0329
Bus 116 0.02933 0.03189 0.02946 Bus 116 0.0325 0.0300 0.0329
#
Numbers in bracket indicate bus number.

Table 11 Evaluation of indices for the performance measurement of 33-nodes RDN.

Indices kw losses kvar losses APLI RPLI AVDI VSI % APTC % RPTC VSM
Without DG 211 143 – – 1.80 9.19 – – 3.40
DG at Bus 18 Syn gen 140 101 33.59 29.53 0.88 5.77 1.92 1.74 4.70
Syn comp 286 220 35.41 53.67 1.01 2.22 1.91 2.99 3.57
Fuel cell 151 108 28.49 24.19 0.96 7.83 1.64 2.83 4.69
DG at Bus 33 Syn gen 92 72 39.11 33.23 0.95 6.55 3.15 2.99 4.67
Syn comp 262 209 24.34 46.29 1.06 3.50 1.31 2.46 3.74
Fuel cell 117 90 44.66 37.01 1.27 7.40 2.95 1.77 3.42
DG at Bus 22 Syn gen 230 165 9.05 15.24 1.77 8.43 0.37 0.95 3.42
Syn comp 212 145 0.40 1.14 1.78 9.12 5.37 5.86 3.40
Fuel cell 219 155 3.95 8.19 1.78 8.49 1.98 1.77 3.42
DG at Bus 25 Syn gen 170 121 19.39 15.24 1.59 8.85 1.15 0.97 3.58
Syn comp 255 182 21.06 26.99 1.63 8.12 0.74 1.33 3.50
Fuel cell 180 126 14.68 11.98 1.65 8.21 0.90 0.703 3.56
200 P. Mehta et al.

Table 12 Evaluation of indices for the performance measurement of 136-nodes RDN.

Indices kw losses kvar losses APLI RPLI AVDI VSI % APTC % RPTC VSM
Without DG 320 703 3.408 0.030 3.686
DG at Bus 118 Syn gen 675 928 1.11 0.32 3.047 0.018 21.74 20.83 8.170
Syn comp 364 705 0.14 0.00 2.483 0.011 0.47 23.81 4.230
Fuel cell 545 736 0.70 0.05 2.566 0.022 23.14 0.88 4.548
DG at Bus 38 Syn gen 442 830 0.38 0.18 0.111 0.032 24.24 16.80 3.687
Syn comp 328 702 0.02 0.00 0.107 0.033 0.08 21.44 3.686
Fuel cell 407 771 0.27 0.10 0.108 0.032 24.62 1.59 3.687
DG at Bus 62 Syn gen 616 983 0.92 0.40 0.120 0.030 22.38 31.35 3.688
Syn comp 334 706 0.04 0.00 0.109 0.033 0.14 17.05 3.686
Fuel cell 501 825 0.56 0.17 0.112 0.030 23.61 2.84 3.688
DG at Bus 97 Syn gen 503 1000 0.57 0.42 0.114 0.033 23.59 20.68 3.687
Syn comp 323 704 0.01 0.00 0.113 0.031 0.03 13.70 3.686
Fuel cell 460 917 0.44 0.31 0.114 0.034 24.05 4.98 3.687

VSM for any type of DG unit which can be observed from locations have been presented. It is observed that the DG units
Fig. 6 and Tables 11 and 12. The ranking of the weakest buses operating in voltage control mode and having reactive power
suggested by VSI and BPF can provide better insight to deter- regulation capability effectively enhance the voltage stability
mine the next location of DG placement as well as its type. For margin irrespective of the system size. The DG units with only
33-nodes RDN, Bus 33 can be identified as the next candidate active power generation capability give better results as com-
bus for the placement of DG in order to enhance voltage sta- pared to DG units with only reactive power generation capa-
bility as indicated by Tables 8 and 9. For this system, the syn- bility. The placement of any type of DG unit results in
chronous generators will provide the better results as it gives voltage profile improvement as compared to the system with-
highest VSI. The careful observation of VSI for 136-nodes out DG in normal as well as stressed loading condition.
given in Table 10 reveals the fact that the placements of differ-
ent types of DG leave the system with different degrees of volt- References
age sensitivity. Hence, the problem of multiple DG placement
becomes much more complex in order to have system wide [1] Working Group on Distributed Generation Integration. Summary
benefits. of distributed resources impact on power delivery systems. IEEE
The different performance indices given in Table 1 are eval- Trans Power Del 2008;23(3) [July].
uated and given in Tables 11 and 12 for 33-nodes and 136- [2] Georgilakis Pavlos S, Hatziargyriou Nikos D. Optimal distributed
nodes RDNs, respectively. The impacts of DG units placement generation placement in power distribution networks: models,
at different candidate buses on real and reactive losses and methods, and future research. IEEE Trans Power Syst 2013;28(3)
[August].
their indices, aggregate voltage deviation index, power transfer
[3] U.S.–Canada Power System Outage Task Force. Final report on
capabilities are calculated. It is observed that losses have been the August 14, 2003 blackout in the United States and Canada:
reduced to greater extent with placement of synchronous gen- Causes and recommendations. Technical Report; April 2004.
erator on Bus 33 as compared to Bus 18. But more emphasis [4] Federal Energy Reg. Comm. Principles for efficient and reliable
has been given to higher value of VSM and lower value of reactive power supply and consumption; February 2005.
AVDI while placing DGs. For 136-nodes RDN, placement [5] Kundur P. Power system stability and control. New York: Mc-
of synchronous generator greatly increases the voltage stability Graw-Hill; 1994.
margin. But it has detrimental impact on active and reactive [6] Donnelly MK, Dagle JE, Trudnowski DJ, Rogers GJ. Impacts of
power losses. Hence, weightage should be decided to strike the distributed utility on transmission system stability. IEEE
the balance between enhancement of VSM and minimization Trans Power Syst May 1996;11(2):741–6.
[7] El-Fergany AA, Abdelaziz AY. Capacitor placement for net
of losses.
saving maximization and system stability enhancement in distri-
bution networks using artificial bee colony-based approach. Int J
7. Conclusion Electr Power Energy Syst 2014;54:235–43 [January].
[8] El-Fergany AA, Abdelaziz AY. Multi-objective capacitor alloca-
The paper has proposed a methodology to find the optimal tions in distribution networks using artificial bee colony algo-
location and type of DG unit among different types which rithm. J Electr Eng Technol 2014;9(2):441–51. http://dx.doi.org/
10.5370/JEET.2014.9.2.441.
can enhance the voltage stability. The comprehensive analysis
[9] El-Fergany AA, Abdelaziz AY. Capacitor allocations in radial
with the use of CPF and Modal analysis are carried out on distribution networks using cuckoo search algorithm. IET Gener
well-known 33-nodes and 136-nodes radial distribution net- Transm Distrib 2014;8(2):223–32.
work for assessment of voltage stability characteristic with [10] Rueda-Medina AC, Franco JF, Rider MJ, Padilha-Feltrin A,
the placement of different DG units. The ranking of the volt- Romero R. A mixed integer linear programming approach for
age sensitive buses are determined with extensive studies. The optimal type, size and allocation of distributed generation in
changes in mode shapes and bus participation factors after radial distribution systems. Int J Electr Power Syst Res
the placement of different types of DG units at different 2013;97:133–43.
Optimal selection of generating units and its placement 201

[11] Gözel T, Eminoglu U, Hocaoglu MH. A tool for voltage stability [29] Blaabjerg F, Teodorescu R, Liserre M, Timbus AV. Overview of
and optimization (VS&OP) in radial distribution systems using control and grid synchronization for distributed power generation
MATLAB graphical user interface (GUI). Simulink Model systems. IEEE Trans Ind Electron October 2006;53(5):1398–409.
Practice Theory 2008;16(5):505–18. [30] Gao B, Morison GK, Kundur P. Voltage stability evaluation
[12] Hung Duong Quoc, Mithulananthan N, Bansal RC. Analytical using modal analysis. IEEE Trans Power Syst 1992;7(4)
expressions for DG allocation in primary distribution networks. [November].
IEEE Trans Energy Convers 2010;25(3) [September]. [31] C. A. Cãnizares, ‘‘Voltage stability assessment procedures and
[13] Hung Duong Quoc, Mithulananthan N. Multiple distributed guides”. Available: <http://www.power.uwaterloo.ca>.
generator placement in primary distribution networks for loss [32] Cãnizares CA (Ed.), ‘‘Voltage stability assessment: concepts,
reduction. IEEE Trans Industr Electron 2013;60(4). practices and tools”; August, 2002. Available: <http://www.
[14] Ochoa LF, Dent CJ, Harrison GP. ‘‘Network distributed gener- power.uwaterloo.ca>.
ation capacity analysis using OPF with voltage step constraints. [33] Ajjarapu Venkataramana, Christy Colin. The continuation power
IEEE Trans Power Syst January 2010;25(1):296–304. flow a tool for steady state voltage stability analysis. IEEE Trans
[15] Karimyan Peyman, Gharehpetian GB, Abedi M, Gavili A. Long Power Syst 1992;7(1), February.
term scheduling for optimal allocation and sizing of DG unit [34] Milano F. Power system analysis toolbox (PSAT), Version 1.3.4,
considering load variations and DG type. Int J Electr Power User Manual; 2005, Available <http://www.power.uwaterloo.ca/
Energy Syst 2014;54:277–87. fmilano>.
[16] Ghosh S, Ghoshal SP, Ghosh S. Optimal sizing and placement of [35] Mantovani JRS, Casari F, Romero RA. Reconfiguracßão de
distributed generation in a network system. Int J Electr Power sistemas dedistribuicß ão radiais utilizando o critério de queda de
Energy Syst 2010;32:849–56. tensão. SBA Controle Autom 2000;11(2):150–9.
[17] El-Fergany Attia. Study impact of various load models on DG
placement and sizing using backtracking search algorithm. Appl
Soft Comput J 2015;30:803–11 [May]. Pankita Mehta completed her BE degree in
[18] El-Fergany Attia. Optimal allocation of multi-type distributed Electrical Engineering in 2003, from
generators using backtracking search optimization algorithm. Int Saurashtra University, Rajkot, India, and her
J Electr Power Energy Syst Jan 2015;64:1197–205. M.E. degree in Electrical Power System in
[19] Mistry Khyati D, Roy Ranjit. Enhancement of loading capacity 2006 from Sardar Patel University of Vallabh
of distribution system through distributed generator placement Vidhyanagar, India. Currently she is working
considering techno-economic benefits with load growth. Int J as Assistant Professor in Charotar University
Electr Power Energy Syst 2014;54:505–15. of Science and Technology and pursuing PhD.
[20] Sajjadi Sayyid Mohssen, Haghifam Mahmoud-Reza, Salehi Her main research interests include Electrical
Javad. Simultaneous placement of distributed generation and machines and power system.
capacitors in distribution networks considering voltage stability
index. Int J Electr Power Energy Syst 2013;46:366–75.
[21] Chakravorty M, Das D. Voltage stability analysis of radial
distribution networks. Int J Electr Power Energy Syst 2001;23 Dr. Praghnesh Bhatt did his PhD from S V
(2):129–35. National Institute of Technology, Surat, India
[22] Hedayati H, Nabaviniaki SA, Akbarimajd A. A method for in 2012. He is working as Professor in
placement of DG units in distribution networks. IEEE Trans Department of Electrical Engineering at C S
Power Del 2008;23(3):1620–8 [July]. Patel Institute of Technology, Charusat,
[23] Hill DJ, Hiskens IA. Load recovery in voltage stability analysis India. His research areas of interests are
and control, presented at the National Science Foundation/ECC Power System Analysis, Power system
Workshop Bulk Power System. Davos, Switzerland: Voltage Dynamics and Stability, Grid integration of
Phenomena III; 1994 [August]. wind power generation, Power System Pro-
[24] Arya LD, Koshti Atul, Choube SC. Distributed generation tection and Distributed Generation.
planning using differential evolution accounting voltage stability
consideration. Int J Electr Power Energy Syst 2012;42:196–207.
[25] Ettehadi M, Ghasemi H, Vaez-Zadeh S. Voltage stability-based
DG placement in distribution networks. IEEE Trans Power Del Dr. Vivek J. Pandya has completed his B.E.
January 2013;28(1). (Electrical Engineering, Sardar Patel Univer-
[26] Esmaili Masoud, Firozjaee Esmail Chaktan, Shayanfar Heidar sity), 1995, M.E. (Electrical Power Systems,
Ali. Optimal placement of distributed generations considering Sardar Patel University), 2003, and Ph.D.
voltage stability and power losses with observing voltage-related (Power System Protection (Electrical Engi-
constraints. Int J Appl Energy 2014;113:1252–60. neering), from Maharaja Sayajirao Univer-
[27] Freitas Walmir, Vieira Jose CM, Morelato Andre. Influence of sity, Vadodara, India. His areas of interests
excitation system control modes on the allowable penetration level are Electrical Machines, Electrical power sys-
of distributed synchronous generators. IEEE Trans Energy tem and Power system protection. He had
Convers 2005;20(2):474–80. published several papers in reputed journals
[28] Puttgen HB, MacGregor PR, Lambert FC. Distributed genera- and conferences. Currently he is working as Associate Professor at
tion: semantic hype or the dawn of a new era? IEEE Trans Power PDPU, School of Technology, Raysan, India.
Energy Manage 2003;1(1):22–9.