OPTIMIZATION OF MULTIPLE DISTRIBUTION

GENERATING UNIT FOR VOLTAGE STABILITY

INDEX USING ANALYTICAL APPROACH
A THESIS
Submitted
In partial fulfillment for the award of the Degree of

MASTER OF TECHNOLOGY
IN
POWER SYSTEMS ENGINEERING
By
MALLADI VENKATA SEETA RAMA SASTRY
(Roll No: 148W1D5609)
Under the guidance of

Dr. GUMMADI SRINIVASA RAO

AssociateProfessor

DEPARTMENT OF

ELECTRICAL & ELECTRONICS ENGINEERING

VELAGAPUDI RAMAKRISHNASIDDHARTHA
ENGINEERING COLLEGE

(AUTONOMOUS)

(AFFILIATED TO JNTUK KAKINADA)

VIJAYAWADA-520007
OCTOBER - 2016

CERTIFICATE

This is to certify that the thesis / dissertation entitled

Optimization of Multiple Distribution Generating Unit for Voltage
Stability Index using Analytical Approach that is being submitted
byMr.

MALLADI

VENKATA

SEETA

RAMA

SASTRY

bearing

Roll

No:148W1D5609 in partial fulfillment for the award of the degree of

Master of Technology in Power Systems Engineering to the Electrical &
Electronics Engineering Department, Velagapudi Ramakrishna Siddhartha
Engineering College is a record of bonafide work carried out by him under
my guidance and supervision.
The results embodied in this thesis have not been submitted to
any other University or Institute for the award of any degree or diploma.

Signature of Guide

Signature of HOD

Dr. GUMMADI SRINIVASA RAO

NARASIMHAM
Associate Professor
EEE

Signature of the Examiner

Dr. P.V.R.L.
HOD of

ACKNOWLEDGEMENTS
I am extremely gratified to my guide Dr. GUMMADI SRINIVASA RAO
Associate Professor, for his insurmountable guidance and constant encouragement
given throughout this project. I am very much indebted for his valuable suggestions
and inspiration he has proffered throughout the course of the project.
It is my sincere obligation to thank our head of Electrical and Electronics
Engineering department Dr. P.V.R.L. NARASIMHAM, Principal Dr. A.V. RATNA
PRASAD and PG program coordinator of EEE Dr. B. SRINIVASA RAO for their
immeasurable beneficent help, with timely suggestions and ideas which wrapped me
in an inconceivable compliance.
I would like to thank all the project review panel members Dr. M.S.K.
RAYALU, Dr. A. RAMADEVI for their consistent encouragement and suggestions
to improve this project work.
I acknowledge my gratitude to the staff, classmates and my senior Mr. A
Balu for their continuous help and support.
It is my sincere commitment to thank to Mr. Mahesh Kumar. Putti, Mr. Ch
Sai Ravi Chandra for extending their support towards project and giving financial
support.
Last but not least, I would like to thank my Parents and family members for
their support in every part of my life and this work.

Malladi Venkata Seeta Rama Sastry

(148W1D5609)

ABSTRACT
Voltage stability has become an important issue of power system stability.
This work is concentrated on finding weak busses with voltage stability index
analysis for Radial Distribution System (RDS). In this workOptimization of Multiple
Distribution Generating Unit using Voltage Stability Index and Analytical Approach
with Forward, Backward Sweep method (FBS).
In this work a noval approach has been implemented to reduce the losses in
the radial distribution system. This work has been done in two phases.From the first
phase of work identification of weak nodes with voltage stability index, which are
most sensitive to voltage collapse using the Branch Injection and Branch Current
(BIBC) and Branch Current and Branch Voltage (BCBV) matrix have been analyzed.
This feature enables us to set an index threshold to monitor and envisage system test
case stability, so that a proper action can be taken to prevent the system from
collapse.
From second phase, optimal allocation of Multiple DG with the objective as
minimizing the distribution line losses with cost attributes. It is vitally essential to
characterize the size and location of multiple DGs to be placed and the cost of
DGs,. Thecost of line losses are calculated when the DGsare placed at optimal bus.
By virtue of some intrinsically characteristics of distribution systems, as the structure
is radial in nature, immensely colossal number of nodes, with an extensive range of
R/X ratios, the conservative techniques divergesinthe distributionsystem forthe
determination of optimum size and location ofthese multipleDistributed Generations
(DG). The projectedapproach and results are validated on IEEE 15, IEEE 33, and
IEEE 69 bus standard test systems.

Table of Contents
Chapter 1.....................................................................................1
INTRODUCTION...........................................................................1
1.1 Transmission & Distribution of Electricity......................2
1.2 Distributed generation (DG)...........................................2
1.3 Technologies for distributed generation..........................4
1.4 Background of the project..............................................5
Chapter 2.....................................................................................8
POWER FLOW ANALYSIS.............................................................8
2.1 Power flow analysis NR method......................................8
2.2 Forward, Backward Sweeping Method..........................12
2.3 Voltage stability index..................................................15
2.4 Mathematical modeling for Distribution Generation.....17
2.5 DG Impact on Voltage Drop..........................................20
2.6 Maximum DG allowed capacity....................................21
2.7 Cost analysis................................................................22
Chapter 3...................................................................................23
ALGORITHM IMPLEMENTATION................................................23
3.1 Constraints of Allocating DGs......................................23
3.2 Algorithm for proposed method....................................23
3.3 Algorithm for Optimal Size and Allocation of DG..........24
Chapter 4...................................................................................26
SIMULATION RESULTS..............................................................26

4.1 Test Systems................................................................26

4.2 Optimal Location and Size............................................26
4.3 Without DG..................................................................27
4.3.1 15 bus Radial Distribution Network(RDN)...............27
4.2.2 33 bus Radial Distribution Network......................29
4.2.3 69 bus Radial Distribution Network......................31
4.4 With DG.......................................................................33
4.4.1 15-Bus Radial Distribution Network.......................34
4.4.2 33-Bus Radial Distribution Network.......................36
4.4.3 69-Bus Test System................................................38
4.5 Multipe Distibution generation (DG).............................41
4.5.1 15 bus distribution network.....................................41
4.5.2 33 bus distribution network.....................................43
4.5.3 69 bus distribution network.....................................44
4.6 Result of Voltage...........................................................46
FUTURE SCOPE.........................................................................47
CONCLUSIONS...........................................................................48
REFERENCES............................................................................49
Appendix....................................................................................51

List of Table
Table 1.1 Different types of DG.....................................................4
Table 4.1 Voltage and VSI without DG for 15 bus test system.....28
Table 4.2 Voltage and VSI without DG for 33 bus system............30
Table 4.3 Voltage and VSI without DG for 69 bus system............32
Table 4.4 Review of result in various test system without DG.....33
Table 4.5 Voltage and VSI with DG for 15 bus............................35
Table 4.6 Voltage and VSI with DG for 33 bus system.................37
Table 4.7 Voltage and VSI with DG for 69 bus system.................39
Table 4.8 Review of results for One Distribution generators........41
Table 4.9 Review of results for multiple DG for 15 bus...............42
Table 4.10 Review of results for multiple DG for 33 bus system. .44
Table 4.11 Review of results for multiple DG for 69 bus system. .46
Table A.1 Load Data for 15 bus system......................................51
Table A.2 Line Data for 15 bus..................................................51
Table A.3 Line Data for 33 bus system.......................................52
Table A.4 Load Data for 33 bus system......................................53
Table A.5 Line Data for 69 bus system.......................................54

10 | P a g e

List of figure
Figure 1.1 Energy Consumed Worldwide

Figure 1.2 Comparison of T & D Systems with DG

Figure 2.1 Two bus system

Figure 2.2 Single line layout of Distibution system

12

Figure 2.3 15 Bus Radial Distribution System (RDS)

13

Figure 2.4 Single Line Layout of System

15

Figure 2.5 One line diagram to illustrate the Voltage drop

20

Figure 4.1 Bus Radial Distribution Network

27

Figure 4.2 Voltage and Voltage stability index for 15 bus RDS

28

Figure 4.3 33 Bus Radial Distribution Network

29

Figure 4.4 Voltage and Voltage Stability Index for 33 bus RDS

29

Figure 4.5 69 Bus Radial Distribution Network

31

Figure 4.6 Voltage and Voltage Stability Index for 69 bus RDS

32

Figure 4.7 Voltage and Voltage Stability Index for 15 bus RDS

34

Figure 4.8 Real power loss and DG size

35

Figure 4.9 Voltage and Voltage stability index for 33 bus RDS

36

11 | P a g e

Figure 4.10 Real power loss and DG size

37

Figure 4.11 Voltage and Voltage stability index for 69 bus RDS 38
Figure 4.12 Real power loss and one DG size

40

Figure 4.13 Two DG allocation with Real power loss

43

Figure 4.14 Two DG allocation with Real power loss

43

Figure 4.15 Two DG allocation with Real power loss

45

Subscripts
N=No of Busses
PL=Real power loss of considered test case
QL=Reactive power loss of considered test case
Vm=bus voltage of mth bus
Sm=injected apparent power of mth bus
SI=stability index
Vs=sending bus voltage
rnn=Branch resistance .
xnn=Branch reactance.
Pm2=Total real power load fed towards node m2,
Qm2=Total reactive power load fed towards node m2

12 | P a g e

jj=branch number
PDgm=OptimalRealGenarationatmthbus
Pm=Injection real power at mth bus
Qm=Injection reactive power at mth bus
QDgm=OptimalReactiveGenerationatmthbus
EC=Energyrate
T=Timedurationinhours
m=Voltage angle on mth bus.

13 | P a g e

Chapter 1
INTRODUCTION
Electricity demand is growing the fastest of all the energy
consumed worldwide. According to global energy consumption 2000
fromFig 1, the electricity accounted for 9%.This rise in consumption
is

more

connectedwithtechnologicallydeveloped

countries

and

continues in the future.It is predicted that the worlds net electricity

consumption almost will increase 36.704% from 20,567TWh in
2016[1]to 28, 1402 billion kWh in 2030. As because of this increase,
proper planning and operation increases its margin of research [2].In
developing countries,the most insistent challenge of modernization,
growth, and global competitiveness is production ofreliable and lowcost electricityand delivering.They are still 1.6 billion without
electricity in work.

Figure 1.1Energy Consumed Worldwide

1.1

Transmission& Distribution of Electricity

A centralized power generation is followedby a typical power

system.The transmission and distribution system acts like a barrier

betweenthis
continuity

centrally
of

generated

delivering

to

power

the

load

and
is

the

more

consumer.
important

A
for

transmission and distribution companies to maintain their reliability

in the system.Reliability according to consumer can be defined as
thecontinuous supply to the consumer end or theability of the power
system components to deliver electricity to all points of consumption.
Reliability is often measured by outage indices defined by the
Institute of Electrical and Electronics Engineers (IEEE) Standard
1366.The only option for the seller is to provide continuous supply
99.90% to act as reliable one and as a result ofthis arduous task, the
electric power system should work hard in all aspects where he wants
to achieve with this extremely complex network. New challenges are
facing day to day in this complex power system network constantly.
This thesis deals with oneof most important new challengein front of
system operators - Distributed generation (DG) with cost attributes
and voltage profile improvement.Th cost plays one of the major role
inselection of DG in the system.
1.2

Distributed generation (DG)

According to supremacy in load increment, these concentrated

generations willnot satisfy the needs of consusmer, so that this DG

has

come

into

picture

with

new

difficulties

and

advantages.Depending upon this increase in load the generation

should increases instantly, but concentrated generation did not

satisfy the load demand in optimal concern. So the concept of
Distribution generation (DG) has a space to curtail this gap between
generation and demand in a distributed way.Concerning to the
liberalization of Electricity Systems by Electric Power Research
Institute (EPRI), new methodologies for planning and operation of
power system are required, dealing with issues ranging from system
dynamics

to

longer

term

considerations

[5].

Introducingthe

Distribution Generation (DG) in system changes the scenarioof

distributionnetwork across the world the way it works.The other way
they

are

Generation

called
(EG)

asDispersed
but

for

this

Generation
thesis,

(DG)and

the

term

Embedded
Distributed

Generation(DG) isused. Figure 2 shows the difference between

transmission and distribution with DG.

Figure 1.2 Comparison of T& D Systems with DG

The distribution system is a highly passive network and solely

for the distribution and delivery of loads. So thatthere is increasingits
penetration of DG is changing the role of distribution systems. The
increased explosionand bidirectional active and reactive power flows
of these Distributed Generatorsin thesystem (DG) has shown the way
to

change

the

variationbychanging

uniqueness
the

of

the

technical

network,

with

characteristics.

more
These

(DG)Distribution Generators are on the networks, and just about the

operate closer to theirlimits for reliable operation. In developed
countriesfor

any

distribution

network,the

payment

to

the

generatorstrengthening required forconnection and can present a

considerable cost to the generator. This may be a barricadeto more
DG penetration.

1.3 Technologies for distributed generation

Typical available size per module according to progression of
technology in recent years.
Table 1.1 Different types of DG
Battery storage
Biomass
Combined cycle gas
Combustion turbine
Fuel cells, molten carbonate
Fuel cells, phosacid
Fuel cells, proton exchange
Fuel cells, solid oxide
Geothermal
Internal combustion engines
Micro hydro

500 kW5 MW
100 kW20 MW
35400 MW
1250 MW
250 kW2 MW
200 kW2 MW
1 kW250 kW
250 kW5 MW
5100 MW
5 kW10 MW
35 kW1 MW

Micro-Turbines
Micro-Turbines
Ocean energy
Photovoltaic arrays
Small hydro
Solar thermal, central receiver
Solar thermal, Lutz system
Sterling engine
Wind turbine
1.4

35 kW1 MW
35 kW1 MW
100 kW1 MW
20 Watt100 kW
1100 MW
110 MW
1080 MW
210 kW
200 Watt3 MW

Background of the project

Traditional power flow analysis techniquesuch as the Newton-

Raphson(N-R) methods are frequently used in powertransmission

systems. However, for special cases which are usually calledillconditioned power systems, these methods might be unstable and
diverge. Therefore, systemsLoad flow analysis of distribution systems
has received much attention because of its uniqueness as R\X ratio,
with this distribution network are ill-conditioned and normal
conventional Newton Raphson and fast decoupled methods are
inefficient at solving such networks as the range of test case
increases in bus number [4]. The main aim in this thesis is to
develop a new load flow technique for radial distribution test system
and this proposed method involves a straight forward arithmetical
expression of voltage magnitudes without trigonometric equations.
This method involves KVL and KCL matrixes and analysis is worked
out [5].
Concerning to the liberalization of Electricity Systems by
Electric Power Research Institute (EPRI), new methodologies for
planning and operation of power system are required, dealing with

issues ranging from system dynamics to longer term considerations

[6]. One of the main changes being seen in power systems across the
world is the increased proliferation of what is known as Distributed
Generation (DG) ranges from a few kilowatts 5KW up to 50 MW with
respect to its penetration level [6]. According to International Energy
Agency (IEA), Distributed generator (DG) as a plant serving a
consumer on-site or providing support to a distribution network,
connected to the grid at distributed level voltages. Some objectives
are produced by IEA and EPRI on distribution generator (DG) in
distribution systems, reduction of fossil energy consumption in
electric power generation and mitigate power losses without emission
of harmful carbon particles, development in nonconventional energy
based Distibuted generation DG all over the world in recent years[3].
With multi-objective index-based approach, size and allocation of
multiple DG for different load models is determine and performed
with PSO [9,12]. For distribution network, a goal programming
technique is developed on GA & PSO which is presented in [10] to
calculate the DG location and size in [11] with multi objective
function. An enhanced analytical method with Loss Sensitivity
Factor(LSF) as objective is proposed in reference [13]. Voltage stability
index as root, exact loss formula is used [14] to find an optimum
sizing for DG in each bus for voltage stability improvement. The
standard equations for formation of distribution loss is referred form
reference [15] .

The mounting penetration of Distribution Generator (DG) is

altering the role of distribution network, which are originally passive
networks, purely for the delivering of electricity to the consumer. The
increased number of this installation of distributed generators (DG)
has directed to change in the uniqueness of the network. For the
operator, to operate the system within safe limits and reliable, these
generators are altering their technical characteristics of the networks.
In analytical approach method, a new work has been proposed to
quickly estimated losses for identifying the best location [16] the load
flow is required to be performed with respect to how many number of
DG installed(according to CIGRE max number of DG installed in RDS
are 3). This method requires less computation, as this we are
adopting multiple DG for getting stability with its range of constraints
with cost analysis [13, 15, 16].
The proposed method is classic since its input data is the same
as the conventional bus branch oriented data used by most utilities.
The proposed method is also novel since it takes advantage of the
topological characteristics of distribution systems and solves the
distribution load flow efficiently. The proposed method is based on
the Newton Raphson formulation and utilizes the branch voltage as
state variables. Byusing those ideas, a constant Jacobian matrix can
be developed, and the traditional Newton Raphson technique can be
utilized, equations are derived to find the solution.

Chapter 2
POWER FLOW ANALYSIS
2.1 Power flow analysis NR method
The current flowing from the mth bus in system at kth iteration

S
I km
V m

k
m

As we see that the current injects into bus m

(2.0)

I m YmnVn

(2.0)

n 1

Where, again, we emphasize that the Y mn terms are admittance

matrix elements. Substitute Eq. (2.2) into Eq. (2.1)

Sm Vm

Y
n 1

* *
mnVn Vm Ymn Vn
n 1

(2.0)

Recall that Vm is a phasor, having magnitude and angle, so that

Vm | Vm | m

.Also, Ymn, being a function of admittances, so that

Ymn Gmn jBmn

Then we may rewrite Eq. (2.3) as
N

Sm Vm Ymn *Vn * Vm m (Gmn jBmn ) * Vn n Vn n (Gmn jBmn ) Vn n

j 1

j 1

j 1

j 1

j 1

Vm m Vn n (Gmn jBmn ) Vm Vn ( m n ) (Gmn jBmn )

(2.0)

The Eq. 2.3 & 2.4 are called the power flow equations. For the
complex function of sinusoids, the Euler relation is used,

V | V | m | V |{cos j sin }
we may rewrite Eq. (2.4) as

S m Vm Vn ( m n ) (Gmn jBmn )
n 1
N

Vm Vn cos( m n ) j sin( m n ) (Gmn jBmn )

j 1

(2.0)

Inside parentheses after performing multiplications the terms in the

of Eq. (2.5) and recall that

S m Pm jQm

, we can express Eq. (2.5) as

two equations, one for the real part, P m, and one for the imaginary
part, Qm
N

Pm Vm Vn Gmn cos( m n ) Bmn sin( m n )

n 1
N

Qm Vm Vn Gmn sin( m n ) Bmn cos( m n )

(2.0)

n 1

It is interesting to consider the case of Eq. 6 if bus m, relabeled

as bus n, is only connected to one other bus, lets say bus n. Then
the bus m injection is the same as the flow into the line mn. The
situation is illustrated in Fig. 2.1

Figure 2.3 Two bus system

For the situation illustrated in Fig. 3, Eq. (2.6) becomes
2

Pm Vm Gmm Vm Vn Gmn cos( m n ) Vm Vn Bmn sin( m n )

2

Qm Vm Bmm Vm Vn Gmn sin( m m ) Vm Vn Bmn cos( m n )

(2.0)

If the line mn admittance is Y=G-jB, as shown in Fig2.1, then G mn=-G

and Bmn=B (fromEq. 2.5). If there is no bus (m) shunt reactance or line
charging, then Gmm=G and Bmm=B. Under these conditions,Eq. (2.7)
becomes

Pm Vm G Vm Vn G cos( m n ) Vm Vn B sin( m n )

(2.

Qm Vm B Vm Vn G sin( m n ) Vm Vn B cos( m n )

0)

If we simply rearrange the order of the terms in the reactive equation,

then we have
2

Pm Vm G Vm Vn G cos( m n ) Vm Vn B sin( m n )

(2.0)

Qm Vm B Vm Vn B cos( m n ) Vm Vn G sin( m n )

As

f1 ( x)

P2 ( x ) P2

f N 1 ( x )

P2

PN ( x ) PN

PN

f ( x)
Q 0

QN 1 ( x ) QN 1

QN 1
f N ( x)
0

( x )

f2 N N

QN ( x ) QN

QN

(2.0)

The solution update formula is given by Eq. (2.10), repeated here for
convenient andclearly, an essential step in applying NR to the power
flow problem is to enable calculation of the Jacobian elements

( i 1)

(i )

x x

(i )

(i )

x J

(i )

f (x )

(2.0)

And the Jacobian is at

J P
Q
J

J PV

J QV

P

|V| Q

(2.0)

He resulting expressions is given below

J nm P

Pn ( x)
Vn Vm Gmn sin( n m ) Bmn cos( n m )
m

(2.13)

J nn P

Pn ( x)
2
Qn ( x ) Bnn Vn
n

(2.14)

J nm Q

Qn ( x )
Vn Vm Gmn cos( n m ) Bmn sin( n m )
m

(2.15)

J nn Q

Qn ( x )
Pn ( x ) Gnn Vn
m

(2.16)

J nm PV

Pn ( x)
Vn Gnm cos( n m ) Bmn sin( n m )
Vm

(2.17)

J nn PV

Pn ( x ) Pn ( x)

Gnn Vn
Vn
Vn

(2.18)

J nm QV

Qn ( x)
Vn Gmn sin( n m ) Bnm cos( n m )
Vm

(2.19)

J nnQV

Qn ( x ) Qn ( x )

Bnn Vn
Vn
Vn

(2.20)

We are now in a position to provide the algorithm for using NR to

solve the power flow problem. Before doing so, it is helpful to more
explicitly define the mismatch vector, from Eq. (2.1) or (2.20) as
f1 ( x )

P2 ( x) P2

PN ( x ) PN

f N 1 ( x )

f ( x )
QN 1 ( x ) QN 1

f N ( x)

(
x
)
2 N N 1
G

P2

QN ( x ) QN

PN

0
(2.21)

QN 1

QN

The Newton Raphson power flow solution techniques and a host of

their derivatives have efficiently solved well behaved power systems.
However the shortcomings have been encountered when their
algorithms are generally implemented and applied to ill-conditioned

and poorly initialized power system. The distribution network is radial

innature, having a high R/X ratio, whereas the transmission systemis
loop in nature having a high X/R ratio.

2.2 Forward, BackwardSweepingMethod

This approach is a direct application of Kirchhoffs Current Law
(KCL) and Kirchhoffs Voltage Law (KVL). Given the voltage of the root
bus as in Fig. 2.2and an initial voltage guess of other buses, the
algorithm takes three steps for each iteration.
Step 1: Nodal current calculation:The current injection at each node
mis calculated using Eq. (2.21)
I mk (

Sm *
P jQ
) ( m k m )* ym (V k 1m )
k
V m
V m

(2.22)

WhereSm is the power injection at node m, Vm is the voltage of node

m, calculated from iteration k, ym is the shunt element of node m.

Figure 2.4 Single line layout of Distibution system

Step 2: Backward Sweep: Starting from the last ordered branch,
current flowJl in branch l is calculatedusing Eq. (2.22)

J m ( k ) I mr J mr , l b, b 1
Where

J mr

I mr

(2.23a)

is the current injection of node mrcalculated from step 1, is

the currents in branches emanating from node mr.

Step 3: Forward Sweep: Starting from the root bus, the node voltages
are updated using Eq.2.23

Vmr( k ) Vmr( k ) Z m J m( k ) , m 1, 2,......, N

(2.23b)

Where msand mrdenote the sending and receiving end of branch l,Z l
is the series impedance of branch l.A simple distribution system
shown in Fig. 2.3 is used as an example. The power injections can be
converted to theequivalent current injections by (2.22), and the
relationshipbetween the busmcurrent injections and branch currents
can beobtained by applying Kirchhoffs Current Law (KCL) to
thedistribution network. The branch currents can then beformulated
as functions of equivalent current injections. Forexample, the branch
currents B1, B2 and B15, can be expressedby equivalent current
injections as

Figure 2.5 15 Bus Radial Distribution System (RDS)

Branch current for 15 bus system from Fig.2.3
B1=I2+I3+I4+I5+I6+I7+I8+I9+I10+I11+I12+I13+I14+I15;
B2=I3+I4+I5+I11+I12+I13+I14+I15;
B3=I4+I5+I14+I15;
B4=I5;
B5=I9+I10;
B6=I10;
B7=I6+I7+I8;
B8=I7;
B9=I8;
B10=I11+I12+I13;
B11=I12+I13;
B12=I13;
B13=I14;
B14=I15;
Therefore, the relationship between the bus current injections and
branch currents can be expressed as

[ B ] [ BIBC ][ I ]

(2.24)

Where BIBC is the bus-injection to branch-current matrix.The

relationship between branch currents and bus voltages as shown in

Figure 4 can be obtained by below equations. For example, the

voltages of bus and impedances are considered in the following steps
Step 1: Read BIBC matrix, Zbbranch impedance vector.
Step 2: Convert Zb vector to a diagonal matrix Z by setting off
diagonal elements to zero.

Z diag ( Z b )

(2.25)

Step 3: Multiply transpose of BIBC matrix with Z matrix.

BCBV [ BIBC ]T * Z

V1
V2
Z12

V1 V3 Z12
V1 V4 Z12

V1 V5 Z12
V1 V6 Z12

V1 V7 Z12
V V Z
1 8 12
V1 V9 Z12
V V Z
1 10 12
V1 V11 Z12

V1 V12 Z12
V1 V13 Z12

V1 V14 Z12
V V Z
1
15
12

Z 23

Z 23

Z 34

Z 23
0

Z 34
0

Z 45
0

0
Z 26

0
0

0
0

0
0

0
0

0
0

0
0

0
0

0
0

Z 26

Z 67

Z 26

Z 67

Z 78

Z 29

Z 29

Z910

Z 23

Z311

Z 23

Z311

Z1112

Z 23

Z311

Z1112

Z1213

Z 23
Z 23

Z 34
Z 34

0
0

0
0

0
0

0
0

0
0

0
0

0
0

0
0

0
0

Z 414
0

0
0

0
0

0
0

0
0

0
0

0
Z 415

B1

B2
B3

B4
B5

B6
B
* 7
B8
B
9
B10

B11
B12

B13
B
14

2.3 Voltage stability index

Voltage instability results from the inability of the case to
provide the power requested by loads. The driving force for voltage
instability is increased load. The voltage stability margin is a
parameter that identifies the near collapse nodes. The node with

small stability indices are called weak nodes and then should be
reinforced by injecting reactive power. In the present analysis, voltage
stability margin is calculated for time variant realistic ZIP load model.
The impact of DG on voltage stability improvement has also
determined.

I mk (

Sm *
P jQm *
) ( m k
)
k
V m
V m

(2.26)

For two bus system the current flow from mth bus to nthfrom Fig 2.4

Figure 2.6 Single Line Layout of System

Im

| Vm | m | Vn | n
Rm jX m

(2.27)

Pm jQm * | Vm | m | Vn | n
)
V km
Rm jX m

( Pm jQm ) *( Rm jX m ) V k m (| Vm | m | Vn | n )
( Pn Rm Qn X m ) j ( Pn X m Qn Rm ) | Vm || Vn | cos( n m ) | Vn |2 j | Vm || Vn | sin( n m )
Separating real and imaginary parts of the above equations

( Pn Rm Qn X m ) | Vm || Vn | cos( n m ) | Vn |2

( Pn X m Qn Rm ) | Vm || Vn | sin( n m )
From the above equations

(2.28)
(2.29)

{| Vm |2 2( Pm 2 Rmn Qm 2 X mn )}2 4.0{( Pm 2 2 Qm 2 2 )( Rmn 2 X mn 2 )} 0

(2.30)

Above equation has a homogenous solution and does not depend on

the phase angle of the system, for problem calculation. In a

distribution system, the V is not so important because the

variation of (V) from the substation to the rootend of a distribution
feeder is only a few degrees. The square root of the |V (n)| solution of
the quadratic equation gives a realistic value. From the above
derivations.

( Rmm 2 X mm 2 )( Pm 2 2 Qn 2 2 ) Amm
| Vm |2 4( Pm 2 Rmn Qn 2 X mn ) Bmm

0.707[ Bmm [ Bmm 2 4 * Amm 2 ]1/ 2 ]1/ 2

0.707[ Bmm [ Bmm 2 4* Amm 2 ]1/ 2 ]1/2
0.707[ Bmm [ Bmm 2 4* Amm 2 ]1/ 2 ]1/2
0.707[ Bmm [ Bmm 2 4* Amm 2 ]1/2 ]1/2

From the above equations, fourth is a feasible solution as first two are
nearly equal to zero and not feasible solutions. The third solution is
negative and not feasible.

0.707[ Bmm [ Bmm 2 4* Amm 2 ]1/2 ]1/2 Vn

Bmm 2 4 * Amm 2 0
and the final equation is

| Vm |4 4( Pm 2 Rmn Qm 2 X mn ) | Vm |2 4.0{( Pm 2 X mn Qm 2 Rmn )}2 SI

(2.31)

Voltage stability margin is determined for each bus using Eq.

(2.31) and the bus with minimum VSM is determined. VSM of each
bus is a number between 0 and 1. The method is based on the BIBC
matrix and helps us to develop a computerized algorithm for the
calculation of each bus index value of any distribution system.
2.4 Mathematical modeling for DistributionGeneration
DG can be classified into four major types based on their
terminal characteristics in terms of real and reactive power delivering
capability as follows:
Assuming

(sign)* tan(cos 1 ( DG pf )
Sign = +1: DG injects reactive power;
Sign = 1: DG is consuming reactive power;
The active and reactive power injected at bus (m), where the DG
located, are given by Eq. (2.32) and (2.33), respectively,

Pm PDGm Pdm

(2.32)

Qm QDGm Qdm Pm * PDGm Qdm

(2.33)

As loss is derived below

N

m 1
n 1

m 1
n 1

Pl jQl S m Vm I m*

Vm

Z
m 1
n 1

mn

In

m 1
n 1

m 1
n 1

Sm Z mn I n I m*
I n | I n | cos n j || I n | sin n
I m | I m | cos m j || I m | sin m
Where

m m m
m Tan(

Qm
)
Pm

Pl jQl Z mn (| I n | cos n j || I n | sin n )* (| I m | cos m j || I m | sin m )

m 1
n 1

m 1
n 1

m 1
n 1

Pl jQl Z mn (| I m || I n | (cos m cos n sin m sin n )) j ( Z mn | I m || I n | (cos m sin n cos n sin m ))

Considering real term form the above Eq.

N

Pl Z mn | I m | | I n | (cos m cos n sin m sin n )

m 1
n 1

Pl Rmn | I m | | I n | cos( m n )
m 1
n 1
N

Ql X mn | I m | | I n | sin( m n )
m 1
n 1

Im

Qm
Pm
Im
| Vm cos m |
| Vm sin m |

Pl Rmn (
m 1
n 1

Pm Pn cos( m n )
)
| Vm | cos m * | Vn | cos n

Pl Rmn (
m 1
n 1

Pm Pn cos[( m n ) (m n )]
)
| Vm || Vn | cos m cos n

From the expansion of cosine function

N

Pl [ mn ( Pm Pn QmQn ) mn (Qm Pn PmQn )

m 1
n 1

(2.34)

Ql [ mn ( Pm Pn Qm Qn ) mn (Qm Pn Pm Qn )
m 1
n 1

(2.35)

Where

mn Rmn (

cos( m n )
)
| Vm || Vn |

mn Rmn (

sin( m n )
)
| Vm || Vn |

mn X mn (

cos( m n )
)
| Vm || Vn |

mn X mn (

sin( m n )
)
| Vm || Vn |

The total active power loss of the system from Eq.2.34

N
dPl
dPl

( [
dPDGm
dPDGm m 1

mn

( Pm Pn QmQn ) mn (Qm Pn Pm Qn ))

n 1

dPl
dPDGm

dPl
dPDGm

( [ mn (( PDGm PDm ) Pn ( * PDGm QDm )Qn ) mn (( * PDGm QDm ) Pn ( PDGm PDm )Qn ))
m 1
n 1

mn ( Pm * Qm ) mm ( Pm Qm )

( mn * Pn mnQn )

m 1
nm

( mn * Qn mn Pn ) 0
m 1
n m

As
N

X m ( mn * Pn mnQn )
m 1
nm

Ym ( mn * Qn mn Pn )
m 1
nm

mm ( PDGm PDm 2 PDGm QDm ) mm (QDm PDm ) X m * Ym 0

After minimizing loss the optimal size of DG at each m th bus can be
written as

PDGm

mm ( PDm * QDm ) mm ( * PDm QDm ) X m Ym

2 * mm mm

(2.36)

Type 1 DG: Only active power injection by DG with PF as Unity, so

DG pf 1; 0

Eg:Photovoltaic, Micro turbines.

PDGm

mm ( PDm ) mm (QDm ) X m
mm

Type 2 DG:Type 2 DG: Only reactive power injection by DG with PF

as Zero. So

DG pf 0; .

Eg:Synchronous Condenser such as Gas Turbines

QDGm

mm (QDm ) mm (QDm ) Ym
mm

Type 3: Active (P) and Reactive (Q) power injection mutually and the

PF ranges between

0 DG pf 1;

th

size at m bus is given by

PDGm

constant sign=+1, the DG optimal

Eg:Synchronous Machine.

PDGm

mm ( PDm * QDm ) mm ( * PDm QDm ) X m Ym

2 * mm mm

Type 4: Mutual Active (P) and Reactive (Q) power injection and the PF

ranges between

th

0 DG pf 1;

at m bus is given by

PDGm

constant sign

, the DG optimal size

Eg: Induction Generators (Wind Generators)

PDGm

mm ( PDm * QDm ) mm ( * PDm QDm ) X m Ym

2 * mm mm

2.5 DG Impact on Voltage Drop

By using power electronic devices, DGsare connectedto the grid
directly. Various modes of operation for power flow are performed
bySynchronous generators or power electronic interfaced DG.

Figure 2.7 One line diagram to illustrate the Voltage drop

For maintaining voltage levelreliable and constant, reactive
power is varied within constraints as the load is varied. Another way
of justifying the above statement isDG should provide both activeand
reactive power for the improvement of voltage profile.The exchange of
reactive power is varies with the types of DG installed in the system.
As they are four types of DG classification stated above reactive power
will flow. A simple layout is shown in Fig.2.5 where the load is
operated with both grid and Distribution Generator (DG). The system
will vary its characteristics when DG is allocated andthe regulation
between two nodes approximated by

V V1 V2

RLN ( PL PDG ) X LN (QL ( QDG ))

V2

(2.37)

Considering the assumption that, if the active and reactive power

generated is more than the load allocated, then the DG will inject
power to load aswellas to the grid, then the voltage will rise at
substation/grid. According to Eq. 2.37, if the DG is a sink to reactive
power,then there is rise/down in the voltage drop as this is the
unconditional

case,i.e.;

the

voltage

may

increase

or

decrease

depending upon the active and reactive power relation between load
and line R/X ratio.

2.6 Maximum DG allowed capacity

The maximum number of DG and allowable capacity of DG to be
placed in the system will depend upon three factors. Those are voltage
limits of system, fault levels, thermal limits of considered lines, lines
and substation. The above three technical factors affect the size and
place of DG. For minimizing losses in the distribution systems,
distribution companies (in India DISCOMs) operate these DGs locally
to handle at load side within above technical factors. Before
introducing the DG in steady state / practical system the (DISCOM)
distribution companies will make sure that if any violation in these
three technical factors for secures operation. Some of the worst cases
the system faces when distribution system

Maximum load and no generation

Minimum load and maximum generation
Minimum voltage at the load when DG size exceeds its
range.

2.7 Cost analysis

The economics attributes of energy loss, active and reactive power of
DGcalculated based on the mathematical model represented as:
(i) Energy losses (EL): The cost of energy loss on an annual basis is
given by

EL cost =(Totalreal power loss)*(E c *T)

R's

Where
EC=4. 63

Rs/kWh

T = 8760

hours/year

(ii) Cost characteristic of DG is selected as per the data available[17]

2
C ( PDG ) a * PDG
b * PDG c

(2.38)

Rs/kWh

Where a = 0, b = 20 c = 0.25.
Cost of reactive power supplied by DG as
2
2
C(Q DG )=[Cost(Smax )-Cost( (Smax
-Q dg
) )*K Rs/hr

(2.39)

Where
K=0. 05 to 0.1

Chapter 3
ALGORITHM IMPLEMENTATION
3.1 Constraints of Allocating DGs
CIGRE

engineers

designed

and

given

some

important

constraints regarding to optimally allocate of Distribution Generation

(DG) into the system.Below steps shows the constraints for allocating
DGin a constant system.
The DG can only be installed at customer locations (loads).
5 MW capacity will be the maximumvalue of DGto
be installedin the system.

 The Power balance condition should be followed for

installation of DG.
The operator should follow the voltage limits as
10.05within a tolerance of 0.001
For isolated networks considered as a special case,the total DG power
cannot begreater than the total light load power.

3.2 Algorithmfor proposed method

The objective is to find out minimum voltage and voltage
stability index for identifying weak nodes in the test case with
considered IEEE standard test case data. Several steps had been
evaluated in this algorithm for finding out the above objectives.
Step 1: Standard test case data are considered for running the load
flow.
Step 2: Select central bus at which the substation is placed.
Step 3: Make element ordering for ordering the root bus with respect
to from the bus (sort row operation with respect to from bus).
Step 4: Formulate BIBC and BCBV matrix.
Step 5: Initialize all the voltages by 1.0pu and considers tolerance of
0.0001.
Step6: Calculate load current by Eq. (2.22) andEq. (2.23a) for every
iteration.
Step 7: Calculate voltages with obtaining currents and matrices up to
which the test case has converged with specified tolerances and find
out minimum voltage index and value.

Step 8: Calculate the voltage stability index by Eq. (2.31) from

obtaining voltages by injection of total real power load and reactive
power load fed through the node for and find out minimum voltage
stability index and value.

3.3 Algorithm for Optimal Size and Allocation of DG

Minimizing the Real Power loss (Ploss) is the main objective with
the injection of power fromDistribution Generation (DG) at optimal
node.With the proposed method determine above the optimal size
and allocation of DG with optimum power factorgives step by step as
follows,
Step 1: Run the base power flow with BIBC matrix and the BCBV
matrix for finding the bus voltages and voltage stability index
Step 2: Finding the optimum size of added DG for each bus except
the reference bus using (2.36).
Step 3: Real power losses are calculated from (2.34) for each bus
individual for allocating optimum size of DG to the bus.
Step 4: Minimum power loss node is choosen as the optimal node for
allocating DG.
Step 5: Check voltage and power balance constraintsaccording to
CIGRE.
The same algorithm is to be followed for all the classifications of DG.

Chapter 4
SIMULATION RESULTS
4.1 Test Systems
3 test cases had been considered for this methodology. They
are 15 bus, 33bus and 69 bus test systems. These test cases, data
had considered from an IEEE standard test case from the power

system. Figure 8 shows 15 bus test case, Figure 8 shows 33 bus test
case and Figure 10 show 69 bus test cases. By this voltage stability
index also increase at that particular node as because of the reactive
power injection.
4.2 Optimal Location and Size
Technically speaking, distribution and transmission networks
are designed for somewhat different purposes. The main difference is
that distribution systems are usually not designed for the connection
of active power generators. Furthermore, distribution networks
usually have a radial or loop design, rather than a mesheddesign
typical for transmission networks. Therefore, the power flow in
distributionnetworks usually is unidirectional and little or no
redundancy exists.From the analytical expressions proposed above,
the optimal sizes at diverse locations the losses are calculated. The
Position with minimum losses is considered as the optimal location at
which the Distributed generation (DG) should be added. As loss
reduction is improved with and coefficients from the base case for
DG allocation. The bus voltages will increase and voltage security will
enhance with the optimal allocation of DG.
4.3

Without DG
The algorithm followed by Forward, Backward Substitution

Method (FBSM), a multi port compensation technique and forms

basic formulation for Kirchhoffs laws. From the above proposed

algorithm the outputs are simulated in MATLAB 2013 and shown

below for different test systems.
4.3.1 15 bus Radial Distribution Network(RDN)

Figure 4.8 Bus Radial Distribution Network

The results of the considered 15 bus test system shown in Fig. 4.1
has been tabulated in Table 4.1From Table. 4.1, it is noticed that a
minimum voltage of 0.9445pu appears at 13 th bus and the
corresponding voltage stability index (VSI)is 0.800464pu at 13 th bus.
The bus number verses voltage mangnitudes and voltage stability
index profile is drawn and is shown in Fig. 4.2. From Fig. 4.2, it is
noticed that the bus number 13 is weak busaccording to stability
index. From Table 4.4 the real power loss in the 15 bus system is
61.795kW. So that the percentage of real power loss in totalsystem is

identified that 2.038% of the total load in the system for the
operating power factor of 0.72 lag.

Figure 4.9 Voltage and Voltage stability index for 15 bus RDS
Table 4.1 shows the Voltage and Voltage Stability Index(pu) without
allocating DG using above proposed method.
Table 4.2 Voltage and VSI without DG for 15 bus test system
Bus
1
2
3
4
5
6
7
8

Voltage
in p.u
1
0.971231
0.956618
0.950854
0.949867
0.958199
0.955979
0.956924

VSI in
p.u
1
0.993369
0.888237
0.837232
0.817435
0.845511
0.839001
0.865304

9
10
11
12
13
14
15

0.967920
0.966847
0.949909
0.945791
0.944481
0.948554
0.948390

0.847343
0.878410
0.830201
0.814112
0.800165
0.819422
0.824790

4.2.2 33 busRadial Distribution Network

Figure 4.1033 Bus Radial Distribution Network

The results of the considered 33 bus test system shown in Fig. 4.3
has been tabulated in Table 4.2From Table 4.2, it is noticed that a
minimum voltage of 0.90365pu appears at 18th bus and the
corresponding voltage stability index (VSI) is 0.6675pu at 18 th bus.
The bus number verses voltage mangnitudes and voltage stability
index profile is drawn and is shown in Fig. 4.4. From Fig. 4.4, it is
noticed that the bus number 18 is weak bus according to stability
index.

Figure 4.11 Voltage and Voltage Stability Index for 33 bus RDS
From Table 4.4 the real power loss in the 33 bus system is
211.264kW. So that the percentage of real power loss in total system
is identified that 5.6686% of the total load in the system for the
operating power factor of 0.8225 lag.Table 4.2 shows the Voltage and
Voltage Stability Index(pu) without allocating DG using above prposed
method.

Bus
1
2
3
4
5
6
7
8
9

Table 4.3 Voltage and VSI without DG for 33 bus system

Voltage
VSI
in pu
in pu
1
1
0.997013 0.999351
0.98283 0.984634
0.975301 0.931404
0.967858 0.903331
0.949340 0.873939
0.945849 0.811926
0.93215 0.798706
0.925803 0.754540

10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

0.919931
0.919055
0.917527
0.911347
0.909073
0.907652
0.906271
0.90424
0.90363
0.99648
0.99288
0.99217
0.99153
0.979217
0.972509
0.96916
0.947414
0.944855
0.933371
0.925125
0.921575
0.917366
0.916439
0.916152

0.734331
0.716141
0.713386
0.708532
0.689855
0.682981
0.678712
0.674648
0.668573
0.938349
0.985920
0.971842
0.969076
0.914943
0.919136
0.894430
0.817349
0.805467
0.795874
0.758530
0.732482
0.721143
0.708203
0.705361

4.2.3 69 bus Radial Distribution Network

Figure 4.1269 Bus Radial Distribution Network

The results of the considered 69 bus test system shown in Fig. 4.5
has been tabulated in Table 4.3From Table 4.2, it is noticed that a
minimum voltage of 0.91125pu appears at 65th bus and the
corresponding voltage stability index (VSI) is 0.6910pu at 65 th bus.
The bus number verses voltage mangnitudes and voltage stability
index profile is drawn and is shown in Fig. 4.6. From Fig. 4.6, it is
noticed that the bus number 65 is weak bus according to stability
index.From Table 4.4 the real power loss in the 33 bus system is
219.54kW. So that the percentage of real power loss in total system is
identified that 5.6026% of the total load in the system for the
operating power factor of 0.82 lag.Table 4.3 shows the Voltage and
Voltage Stability Index(pu) without allocating DG using above
prposed method.

Figure 4.13Voltage and Voltage Stability Index for 69 bus RDS

Table 4.4 Voltage and VSI without DG for 69 bus system
Bus
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

Voltage in
pu
1
0.999962
0.999924
0.999818
0.998900
0.989058
0.978823
0.976383
0.975136
0.969597
0.968377
0.964876
0.961638
0.958428
0.955251
0.954660
0.953686
0.953676

VSI in
pu
1
0.999991
0.99984
0.99967
0.999004
0.993033
0.95471
0.917546
0.90863
0.903898
0.883795
0.879272
0.866634
0.855098
0.843778
0.832671
0.83062
0.827222

19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54

0.953162
0.952832
0.952299
0.952291
0.952212
0.952039
0.951852
0.951775
0.951753
0.999917
0.999841
0.999735
0.999716
0.999622
0.999396
0.999142
0.999067
0.999909
0.999718
0.999542
0.999492
0.999489
0.998715
0.998391
0.998348
0.998339
0.99823
0.998229
0.999763
0.998386
0.994139
0.993537
0.976344
0.976333
0.972081
0.968527

0.827196
0.825412
0.824275
0.822421
0.822397
0.822124
0.821527
0.820879
0.820611
0.999285
0.999668
0.999368
0.998941
0.998866
0.998489
0.997586
0.996574
0.999295
0.999636
0.998874
0.998172
0.99797
0.997955
0.994872
0.993581
0.99341
0.993374
0.992939
0.99669
0.999031
0.993488
0.976773
0.90427
0.908682
0.929005
0.89224

55
56
57
58
59
60
61
62
63
64
65
66
67
68
69

0.963625
0.960202
0.941053
0.931678
0.927778
0.921737
0.914278
0.913968
0.913629
0.911717
0.911258
0.968281
0.968248
0.964481
0.964429

0.879047
0.862013
0.848398
0.784707
0.753792
0.740891
0.722079
0.698752
0.697833
0.696887
0.691022
0.867553
0.879033
0.856983
0.865313

Table 4.3 shows the Voltage and Voltage Stability Index(pu) without
allocating DG using above proposed method.
Table5.4Review of resultin various test system without DG
Test

Power

case

factor

15
bus
33
bus
69
bus

0.98
0.85
0.82

Voltage
Stability
index

Minimum
Voltage

Load demand

Activepower

with real

loss, cost

power loss

Rs/year

0.8004@

0.9445

1.226M@61.795

13
0.6675@

@13
0.90367

kW
3.715MW@211.

18
0.69102

@18
0.9113

26kW
3.971MW@219.

@65

@65

54kW

25.30lakh
85.69lakh
89.210lakh

4.4 With DG
The algorithm considered above is formulated for DG in the
system. Belowresults shows that optimal sizes and total losses with

the optimal placement of DG at a time, for 15, 33 and 69 bus test

systems respectively.
Notice in all the cases, optimal sizes of DG obtained from the
proposed method capable of voltage profile improvement. Similarly,
total losses obtained from the method without updating and for
the purpose of identifying best location changes when they are
updated with DGas the losses are decreased.
4.4.1 15-BusRadial Distribution Network
With analytical approach, Table 4.8 shows the simulation
results for proposed network. The table contains location at which the
DG allocatedand the optimal size of DG in MVA, power factor of DG
unit and cost attributes for different busses. For better reduction of
losses in the system, installation of the DG numberwithin their
ranges according to CIGRE .For proposed 15 bus network Table 4.5
shows the voltage and stability index in tabular form. Form that bus 7
has minimum voltage of 0.9794 pu having the voltage stability index
of 0.9241pu at 7th bus only. The voltage and index had improved after
placing DG in the optimal position.

Figure 4.14 Voltage and Voltage Stability Index for 15 bus RDS
The voltage and stability index is plotted in Fig. 4.7 where minimum
voltage point and index point is highlighted. The test case needs both
real and reactive power injection to reduce losses.All the results above
discussed wrt to the 15 bus radial distribution system are tabulated
in Table 4.8.
Table 4.6 Voltage and VSI with DG for 15 bus

Bus
1
2
3
4
5
6
7
8
9

Voltage in
pu
1
0.9946678
1.0002883
0.9945240
0.993537
0.9816263
0.9794047
0.9803501
0.9913563

VSI in
pu
1
1.0874743
1.0566929
0.9999285
0.9781005
1.0099772
0.924110
0.951719
1.011980

10
11
12
13
14
15

0.9902833
0.9935754
0.9894544
0.9881440
0.9922236
0.9920596

0.966490
0.992236
0.973742
0.958250
0.980273
0.986147

From Fig. 4.8 it is noticed that minimum power loss appears at bus 3.
The optimal power factor for this considered system is 0.72 lagging
and corresponding optimal size is 1.4206 MVA has a cost for real
power is 20,457/hr (1022.9 kW) Rs.The total losses with optimal
location, optimal size, having best power factor are 15.7881 kW with a
loss cost 6.40lakhRs/year.

Figure 4.15 Real power loss and DG size

4.4.2 33-Bus Radial Distribution Network
For proposed 33 bus network Table 4.6 shows the voltage and
stability index in tabular form. Form that bus 18 has minimum
voltage of 0.9545 pu having the voltage stability index of 0.8323pu at
18

th

bus only. The voltage and index had improved after placing DG in

the optimal position. The voltage and stability index is plotted in Fig.
4.9 where minimum voltage point and index point is highlighted.

Figure 4.16 Voltage and Voltage stability index for 33 bus RDS
In 33-bus test system, from Fig. 4.10 bus 6 found to be the best
location with minimum power loss. The optimal power factor for this
system obtained is 0.8225 lagging and corresponding optimal size is
3.0135 MVA has a cost for real power is 49,774/hr

(2488.78 kW)

according to cost analysis. The test case needs both real and reactive
power injection to reduce losses. The total losses with optimal
location, optimal size, having best power factor is 68.495 kW with real
power loss cost as 30.382 lakh Rs/year.

Bus

Table 4.7 Voltage and VSI with DG for 33 bus system

Voltage
VSI in
in pu
pu

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

1
0.999061
0.995824
0.996421
0.997437
1.000408
0.996894
0.983175
0.976814
0.970931
0.970056
0.968529
0.962332
0.96004
0.958615
0.957229
0.955185
0.954570
0.998529
0.994931
0.994223
0.993582
0.992217
0.985506
0.982161
0.998474
0.995904
0.984399
0.976136
0.972566
0.968370
0.967446
0.967160

1
1.007541
1.036341
1.013525
1.017225
1.070142
1.000482
0.983005
0.932588
0.908919
0.888491
0.885134
0.878495
0.857191
0.849232
0.844195
0.839228
0.832328
0.988825
0.994015
0.979871
0.977086
0.996343
0.968583
0.943045
1.006503
0.993118
0.980108
0.936899
0.907139
0.893753
0.879169
0.87595

All the results above discussed wrt to the 33 bus radial distribution
system are tabulated in Table 4.8.

Figure 4.17 Real power loss and DG size

4.4.3 69-Bus Test System
For proposed 69 bus network Table 4.7 shows the voltage and
stability index in tabular form. Form that bus 27 th has minimum
voltage of 0.9694 pu having the voltage stability index of 0.8834pu at
27th bus only. The voltage and index had improved after placing DG in
the optimal position.

Figure 4.18 Voltage and Voltage stability index for 69 bus RDS
From Fig. 4.8 it is noticed that minimum power losss appears at bus
61. The test case needs both real and reactive power injection to
reduce losses. The optimal power factor for this system obtained is
0.82 lagging and corresponding optimal size is 2.2228MVA has a cost
for real power is 36,232 RS/hr (1811.51 kW). The total losses with
optimal location, optimal size, having best power factor is 23.67 kW
with real power loss cost as 9.60lakh Rs.

Bus2
1
2

Table 4.8 Voltage and VSI with DG for 69 bus system

Voltage
VSI in
in pu
pu
1
1
0.999980
1.000064

3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38

0.999961
0.999910
0.999618
0.996619
0.993506
0.992789
0.992461
0.987013
0.985813
0.982369
0.979184
0.976028
0.972903
0.972322
0.971363
0.971354
0.970848
0.970523
0.969999
0.969992
0.969914
0.969743
0.969559
0.969483
0.969462
0.999953
0.999878
0.999771
0.999752
0.999658
0.999432
0.999179
0.999104
0.99994
0.999755
0.999579

0.999986
1.000036
1.001875
1.023185
1.012152
0.980480
0.974781
0.969859
0.949021
0.944318
0.931201
0.919224
0.907466
0.895942
0.893807
0.890282
0.890252
0.888398
0.887215
0.885291
0.885265
0.884981
0.884360
0.883687
0.883410
0.999650
0.999814
0.999514
0.999087
0.999011
0.9986353
0.9977326
0.9967201
0.9996612
0.9997821
0.9990205

39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69

0.999528
0.999526
0.998752
0.998428
0.998385
0.998375
0.998266
0.998266
0.999854
0.998478
0.994311
0.993629
0.992750
0.992740
0.992630
0.992836
0.993196
0.993632
0.996102
0.997319
0.997788
0.998726
0.999940
0.999652
0.999330
0.997542
0.997100
0.985719
0.985687
0.981981
0.981930

0.9983184
0.9981158
0.9981015
0.9950181
0.9937268
0.9935562
0.9935196
0.9930846
0.9995579
0.9993965
0.9938506
0.9771325
0.9702658
0.9713156
0.9958784
0.9845891
0.9909036
0.9875707
1.0564498
1.0251409
1.0060607
1.0173064
1.0276842
0.9997618
0.9986123
0.9973232
0.9901407
0.9321698
0.9440876
0.9211955
0.9298473

The voltage and stability index is plotted in Fig. 4.11 where minimum
voltage point and index point is highlighted. The test case needs both
real and reactive power injection to reduce losses. All the results
above discussed wrt to the 69 bus radial distribution system are
tabulated in Table 4.8.

Figure 4.19 Real power loss and one DG size

From this analysis an interesting point had been seen that optimal
power factor of DG for minimizing losses is in close agreement with
the load power factor, so that a guidance to select the appropriate
technological option for minimizing losses in the system, is given
choice.The total results with single DG above discussed are given in
Table 4.8 with cost attributes and with this real power loss plays
important role in the planning of DG at various busses in the system.

Table 4.9Review of results for One Distribution generators

Powe
Test

case

facto
r

15
bus

33
bus

69
bus

0.72

Load
DG

Voltage

Min

demand

value

Stabilit

Voltag

withreal

in MVA

y index

power

0.825

0.82

0.8323@

0.9546

18

@18

68.495k

MVA@6
1

@27

@27

/hr

/year

kW)

@7

0.9694

costRs

81kW
3.715MW

0. 8831

costRs

1022.9

MVA@3

2.2228

loss

W@15.78

0.9791

@6

DG

20,457(

0.9241@

MVA

ActiveP

loss
1.2264M

1.4205

3.0135

Active

49,774(
2488.7
8kW)

W
3.9718M

36,232

W@23.67

(1811.5

kW

1kW)

6.40
lakh

30.4
lakh

9.60
lakh

4.5 MultipeDistibution generation(DG)

4.5.1 15 bus distribution network
In 15-bus test case, from Fig. 4.8 it is noticed that bus 3 found
to be the best location with minimum power loss.Extending the
annalysis for multiple for further reduction of real power loss. For this
once again run the load flow with previously obtained values for
finding another optimal loacation continued to previous one.

Table 4.10 Review of results for multiple DG for 15 bus

%
Real
DGco
Test

Installed DG

Voltag

Ploss

chang

Power

st

case

MVA

kW

e in

injecti

Rs

loss

on

/hr

1022.9

20,45

kW
1035.5

0.9445

No DG
1DG

2DG

size

1.42

@13
0.9794

bus

@7

size

1.418

bus

0.464
4
6

0.9875
@11

61.79
15.82

74.38

7.47

87.90

kW+33
9.03kW

27,49
0

From Fig. 4.13 it is noticed that

bus 6 had the low value having

minimum power loss at 8.302kW. For the optimal power factor 0.72
lagg for this system,the optimal size is 1.8824MVA has a cost for real
power is 27,490 (1035.5kW+339.0316kW) Rs/hr. The test case needs
both real and reactive power injection to reduce losses. Table 4.9
shows the simulated results for 15 bus distribution systems.Further
increase in number of DGs, will decrese the real power loss in the
system and voltage profile will improve.

Figure 4.20 Two DG allocation with Real power loss

4.5.2 33 bus distribution network
In 33-bus test case, from Fig. 4.10 it is noticed that bus 6 found
to be the best location with minimum power loss. Extending the
annalysis for multiple for further reduction of real power loss. For this
once again run the load flow with previously obtained values for
finding another optimal loacation continued to previous one.

Figure 4.21 Two DG allocation with Real power loss

From Fig 4.14 it is noticed that bus 30 had the low value having
minimum power loss at 56.387kW.

For the optimal power factor

0.8225 lagg for this system,the optimal size is 3.22164MVA has a

cost for real power is 61,095 (2420.18kW + 630.385kW) Rs/hr. The
test case needs both real and reactive power injection to reduce
losses. Table 4.10 shows the simulated results for 33 bus distribution
systems.Further increase in number of DGs, will decrese the real
power loss in the system and voltage profile will improve.

Table 4.11Review of results for multiple DG for 33 bus system

Tes

Installed DG

Voltag

Ploss

Chang

cas

MVA

kW

in loss

Real
power
injection

DG
cost
Rs
/hr

No

0.9036

211.2

DG

@18

1D

size

3.038

0.9546

68.85

bus

@18

2D

size

2.548

0.9765

56.38

bus

@18

.
663
30

67.40

2488.78

49,77

kW

2420.18k
73.30

W+630.3
8kW

61,05
8

4.5.3 69 bus distribution network

In 69-bus test case, from Fig. 4.11 it is noticed that bus 17
found to be the best location with minimum power loss. Extending
the annalysis for multiple for further reduction of real power loss. For
this once again run the load flow with previously obtained values for
finding another optimal loacation continued to previous one.

Figure 4.22 Two DG allocation with Real power loss

From Fig. 4.15 it is noticed that

bus 17 had the low value

having minimum power loss at 7.4647kW.

For the optimal power

factor 0.8225 lagging and corresponding optimal size is 2.844MVA

has a cost for real power is 45,931 (1776.834kW+519.884kW) Rs/hr.
The test case needs both real and reactive power injection to reduce
losses. Table 4.11 shows the simulated results for 33 bus distribution
systems. Further increase in number of DGs, will decrese the real
power loss in the system and voltage profile will improve.

Table 4.12Review of results for multiple DG for 69 bus system

Test

Installed DG

case

1DG

MVA

Ploss

chang

power

kW

e in

injectio

loss

89.21

1811.51

kW

0.9113

DG

@61

size

2.222

0.9694

bus

61

@27

2DG
bus

0.649
0

2.195

17

61

Real

Voltag

No

size

DG
cost
Rs/hr

219.54
23.67

36,232

1776.83
0.9765
@65

7.4647

96.59

4kW

+519.88

45,931

4kW

4.6 Result of Voltage

Tables4.2, 4.3, 4.4, 4.6, 4.7 and 8 and figures, Fig. 4.2 to Fig.
4.13 indicate the maximum voltages for 15, 33, and 69-bus test
systems, respectively. In all of these cases, after single and
multipleDG units are added, although satisfying all the power and
voltage constraints, the total losses can reduce considerably. This was
checked with this complete power flow and improved voltages are
obtained.

FUTURE SCOPE
The DG should participate according real time whole sale
electrical market. Deregulation of electrical market is derived by two
objectives

namely

optimal

size

and

allocation

throughprofit

maximization and social welfare maximization.Social welfare is

defined as the difference between total benefit to loads (consumers)
minus total cost of production (players).To calculate the profit
maximization which is evaluated as revenue minus cost for the
particular DG.For getting maximum profit for the player (DG owner)
the optimization process will identify the node and corresponding
optimal DG size.Locational Marginal Price (LMP) identifies the
Distribution Generation (DG) placement.
Hence this work may be extended to find optimal size and
optimal location using LMP ranking and consumer payment (CP)
based ranking in deregulated environment

CONCLUSIONS
A

Forward,

Backward

Sweeping

method

(FBSM)

has

been

implemented to find out the bus voltages, voltage stability index

and system losses.
The optmal size and optimal location with Type 3 DG have been
found using analytical approach to reduce the system losses and to
improve the voltage profile.
The same analytical approach has been implemented to multiple
DGs (Two) to reduce the system .
From this proposed approach, the system losses are reduced from
61.795kW to 15.78kW, 211.26kWto 68.975kWand219.54kW to
23.674kW for 15, 33 and 69 bus systems respectivelyfor single DG
alloaction.
The cost of DG for proposed test systems are, 20,457(1022.9kW)
Rs/hr, 49,774 (2488.78kW) Rs/hr and 36,232 (1811.51kW) Rs/hr
for 15, 33 and 69 bus systems respectively..
From this proposed approach system for multiple DG alloaction,
the system losses are reduced from 61.795kW to 7.47kW,

211.26kW to 56.387kW and 219.54kW to 7.4647kW 15, 33 and 69

bus systems respectively.
The
cost
of
DG

for

proposed

are,27,490(1034.4kW+339.031kW)

test

Rs/hr,

(2420.18kW+630.38kW)

systems
49,774

Rs/hr

and

36,232(1776.834kW+519.884kW) Rs/hrfor 15, 33 and 69

bus

systems respectively.
The main theme of this thesis voltage profile improvement rather
than voltages stability improvement by injecting real power into the
system.

REFERENCES
[1]

Yearbook.Enerdata.Net/Electricity-Domestic-Consumption

Data -By-Region
[2] World Final Electricity Consumption between 1980 and
2030Twh
[3] Tsikalakis, A. G., and N. D. Hatziargyriou. "Environmental
benefits of distributed generation with and without emissions
trading." Energy Policy35.6 (2007): 3395-3409.
[4] Das D, Kothari DP, Kalam A. A simple and efficient method
for

load

flow

solution

of

radial

distribution

networks.

International Journal of Electrical Power and Energy Systems

1995; 17 (5): 335346.
[5] A. Alsaadi, and B. Gholami An Effective Approach for
Distribution

System

Power

Flow

Solution

International

Scientific Index, Electrical and Computer Engineering, Vol: 3,

No: 1, 2009 waset.org/Publication/14357
[6]SN Singh, Jacob Ostergaard and Naveen Jain Distributed
Generation in Power Systems: An Overview and Key Issues
2009.
[7] See Electric Power Research Institute
[8] IEEE Standard for Interconnecting Distributed Resources
with Electric Power Systems, IEEE STD 15472003
[9]ElZonkoly AM. Optimal placement of multi-distributed
generation units, including different load models using particle
swarm optimization. Swarm Evolutionary Computation 2011;
1: page 5059.
[10]Moradi MH, Abedini M.
A combination of genetic algorithm and particle swarm
optimization for optimal DG location and sizing in distribution
systems. Electric Power Energy System 2012; 34: page 6674.
[11]Vinothkumar

K,

Selvan

MP.

Distributed

generation

planning a new approach based on goal programming. Electric

Power Compon Syst 2012; 40: page 497512
[12]Hung Duong Quoc, Mithulananthan Nadarajah. Multiple
distributed

generator

placement

in

primary

distribution

networks for loss reduction. IEEE Trans Industrial Electronics

2013; Vol 60: No4: pg17001708.
[13]Parizad A, Khazali A, Kalantar M. Optimal placement of
distributed generation with sensitivity factors considering

voltage stability and losses indices. Proceedings of ICEE May

1113; 2010, page. 848855.
[14]

Power

System

Optimization

D.

P.

Kothari, J.

S.

Dhillon - Electric power systems 2004

[15]Acharya N, Mahat P, Mithulananthan N. An analytical
approach for DG allocation in primary distribution network.
Electric Power Energy System 2006; 28:66978.
[16]V.V.S.N. Murty, Ashwani Kumar Optimal placement of DG
in radial distribution systems based on new voltage stability
index under load growth Electrical Power and Energy Systems,
Volume 69, July 2015, page: 246256
[17]Gautam Durga, Mithulananthan Nadarajah. Optimal DG
placement in deregulated electricity market. Electric Power
System Res 2007; 77:16271636.

Appendix
Table A.13 Load Data for 15 bus system
Bus

Bus

1
2
3
4
5

Type
1
3
3
3
3

Voltage

Del

Pg

Qg

Pload

Qload

1
1
1
1
1

angle
0
0
0
0
0

0
0
0
0
0

0
0
0
0
0

0
44.1
70
140
44.1

0
44.9883
71.41
142.8
44.9883

6
7
8
9
10
11
12
13
14
15

3
3
3
3
3
3
3
3
3
3

1
1
1
1
1
1
1
1
1
1

0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0

140
140
70
70
44.1
140
70
44.1
70
140

142.82
142.82
71.41
71.41
44.9883
142.82
71.41
44.9883
71.41
142.82

Table A.14 Line Data for 15 bus

Bus

Form

To

Resistanc

Reactanc

1
2
3
4
5
6
7
8
9
10
11
12
13
14

bus
1
2
3
4
2
9
2
6
6
3
11
12
4
4

bus
2
3
4
5
9
10
6
7
8
11
12
13
14
15

e
1.35309
1.17024
0.84111
1.52348
2.01317
1.68671
2.55727
1.0882
1.25143
1.79553
2.4485
2.01317
2.23801
1.19702

e
1.32849
1.14464
0.82271
1.0276
1.3579
1.1377
1.7249
0.734
0.8441
1.2111
1.65185
1.3579
1.5047
0.8074

Shunt

Tap

4.668
0
0
0
0
0
0
0
0
0
0
0
0
0

ratio
1
1
1
1
1
1
1
1
1
1
1
1
1
1

Table A.15 Line Data for 33 bus system

Bus

Form

To

Resistanc

Reactanc

bus
1

bus
2

e
0.0922

e
0.047

Shunt

Tap

5.234

ratio
1

2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
2
19
20
21
3
23
24
6
26
27
28
29
30
31
32

3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

0.493
0.366
0.3811
0.819
0.1872
1.7114
1.03
1.04
0.1966
0.3744
1.468
0.5416
0.591
0.7463
1.289
0.732
0.16
1.5042
0.4095
0.7089
0.4512
0.898
0.896
0.203
0.2842
1.059
0.8042
0.5075
0.9744
0.3105
0.341

0.251
0.1864
0.1941
0.707
0.6188
1.2351
0.74
0.74
0.065
0.1238
1.155
0.7129
0.526
0.545
1.721
0.574
0.1565
1.3554
0.4784
0.9373
0.3083
0.7091
0.7011
0.1034
0.1447
0.9337
0.7006
0.2585
0.963
0.3619
0.5302

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

Table A.16 Load Data for 33 bus system

Bus

Bus
Type

Voltage

Del
angle

Pg

Qg

Pload

Qload

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

1
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

0
100
90
120
60
60
200
200
60
60
45
60
60
120
60
0
0
90
90
90
90
90
90
420
420
60
60
60
120
200
150
210
60

Table A.17Line Data for 69 bus system

0
60
40
80
30
20
100
100
20
20
30
35
35
80
10
0
20
40
40
40
40
40
50
200
200
25
20
20
70
600
70
100
40

Bus

Form

To

Resistanc

Reactanc

Shun

Tap

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

bus
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
3
28
29
30
31
32
33
34

bus
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35

e
0.0005
0.0005
0.0015
0.0251
0.366
0.3811
0.0922
0.0493
0.819
0.1872
0.7114
1.03
1.04
1.058
0.1966
0.3744
0.0047
0.3276
0.2106
0.3416
0.014
0.1591
0.3463
0.7488
0.3089
0.1732
0.0044
0.064
0.3978
0.0702
0.351
0.839
1.708
1.474

e
0.0012
0.0012
0.0036
0.0294
0.1864
0.1941
0.047
0.0251
0.2707
0.0619
0.2351
0.34
0.345
0.3496
0.065
0.1238
0.0016
0.1083
0.069
0.1129
0.0046
0.0526
0.1145
0.2475
0.1021
0.0572
0.0108
0.1565
0.1315
0.0232
0.116
0.2816
0.5646
0.4873

t
-4.68
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

ratio
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68

3
36
37
38
39
40
41
42
43
44
45
4
47
48
49
8
51
9
53
54
55
56
57
58
59
60
61
62
63
64
11
66
12
68

36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69

0.0044
0.064
0.1053
0.0304
0.0018
0.7283
0.31
0.041
0.0092
0.1089
0.0009
0.0034
0.0851
0.2898
0.0822
0.0928
0.3319
0.174
0.203
0.2842
0.2813
1.59
0.7837
0.3042
0.3861
0.5075
0.0974
0.145
0.7105
1.041
0.2012
0.0047
0.7394
0.0047

0.0108
0.1565
0.123
0.0355
0.0021
0.8509
0.3623
0.0478
0.0116
0.1373
0.0012
0.0084
0.2083
0.7091
0.2011
0.0473
0.1114
0.0886
0.1034
0.1447
0.0012
0.0012
0.0036
0.0294
0.1864
0.1941
0.047
0.0251
0.2707
0.0619
0.2351
0.34
0.345
0.3496

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1