B.

Venkatesh Reddy 13

Sizing of DG Units Using Exact Loss Formula to

Improve Efficiency of Radial Distribution System
B. Venkatesh Reddy

Abstract: Loss reduction in distribution systems has been a systems, making power marketable to sell to utilities,
subject of great concern since the evolution of the generating environmental friendly power and ancillary
interconnected power system. In the recent past, with increasing services like voltage control and loss reduction [1].
interest in climate change and energy security, renewable
energy integration and energy efficiency, including loss
II. DG TECHNOLOGIES
reduction, have been considered as twin-pillars of sustainable
energy solutions. When renewable energy is integrated by Distributed generation takes place on two-levels: the local
considering loss reduction as an additional goal, it would lead to level and the end-point level. Local level power generation
multi-fold benefits. This paper presents the application of plants often include renewable energy technologies that are
distributed generation for loss reduction. The two key issues of site specific, such as wind turbines, geothermal energy
the most suitable location and appropriate size of distributed production, solar systems (photovoltaic and combustion), and
generation for loss reduction have been discussed. Analytical
some hydro-thermal plants. These plants tend to be smaller
expressions have been developed for finding the appropriate size
of different types of distributed generations. Methodologies are and less centralized than the traditional model plants. They
presented for locating the DG in primary distribution network, also are frequently more energy and cost efficient and more
assuming primary energy resources are evenly distributed along reliable. Since these local level DG producers often take into
the network. The analytical expressions and placement account the local context, the usually produce less
methodologies have been tested on IEEE 33 bus system. environmentally damaging or disrupting energy than the
larger central model plants.
Keywords: Distributed Generation (DG), Exact Loss The technologies used for distribution generation are
Formula, Optimum Location, Power Factor, Power Loss reciprocating engines, gas turbines, micro turbines, fuel cells,
photovoltaic systems, wind Energy, hydro electric resources,
bio mass etc., DG technologies can meet the needs of a wide
I. INTRODUCTION range of users, with applications in the residential,
Distributed generation is an approach that employs commercial, and industrial sectors. Decision makers at all
small-scale technologies to produce electricity close to the levels need to be aware of the potential benefits that DG can
end users of power. DG technologies often consist of modular offer. In some instances, DG technologies can be more cost
(and sometimes renewable-energy) generators, and they offer effective than conventional solutions. Among other things,
a number of potential benefits. In many cases, distributed DG can be used by utilities to enhance existing systems and to
generators can provide lower-cost electricity and higher delay the purchase of transmission and distribution
power reliability and security with fewer environmental equipment. In addition, DG units can help to meet the
consequences than can traditional power generators. changing demand of end users for premium, reliable or
With large amount of DG penetration it is critical that the “green” power.
power system impacts be assessed accurately so that DG can Depending on the operating scheme and relative
be applied in a manner that avoids causing degradation of performance of the DG system and the power plants
power quality, reliability and control of utility system. The supplying the grid, fuel consumption, carbon and other
penetration of DG like wind generators, solar panels and pollutant emissions and noise pollution can all increase or
small combined heat and power units causes a change in the decrease with DG adoption [2]. For these reasons, DG policy
classic definition of the electricity grid. In the new scenario needs to encourage applications that benefit the public, while
power flow will not only vary in amount but also in direction. discouraging those from which the public incurs a net cost. By
For DG implementation the consumers’ interest would be considering the capital DG costs, maintenance cost, size
reliability, Power quality and safety where as network range, start-up capability, efficiency, influence on public etc.,
operator interests would be customer interests, usage of the the technology that will be suitable for particular application
grid components, grid losses, safety, stability of the grid and should be adapted.
profitability.
Distributed generation (DG) devices can be strategically III. LOAD FLOW APPROACH
placed in power systems for grid reinforcement, reducing Four variables associated with each node are Bus voltage
power losses and on-peak operating costs, improving voltage magnitude (V), Voltage angle (δ), Real power (P) and
profiles and load factors, differing or eliminating for system Reactive power (Q). Each node introduces two equations,
upgrades and improving system integrity, reliability and namely the real and reactive power balance equations.
efficiency. The minimum data required to fully specify system
Distributed Generation has many benefits, including conditions are impedances (usually in per-unit) for all series
increasing the reliability of electric power, reducing electric and shunt branches of the distribution network. Network
utility bills, improving the payback of required generation elements are represented as lumped complex impedances at

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569) Vol. 10, Issue. 1 ,Nov. 2014
B. Venkatesh Reddy 14

rated frequency (e.g. distribution lines, in-phase transformers, losses in the system .The cost of power transmission is
series and shunt reactors and capacitors). Distribution lines reduced by minimizing power losses. This can be partially
with non-negligible charging capacitance are represented by accomplished by DG unit placement in the network.
their simple equivalent π networks and active-power and By considering the above points into consideration, we
reactive-power generations and loads at each node. must consider another important point, which is power
Distribution load flow solution typically provides voltage reversal in the network. Because if DG size at a bus exceeds
magnitude and angle at each bus bar, real and reactive power certain value, power loss at that bus becomes negative. This
loss at each bus bar, power flows and MVA loadings at both situation must be avoided. The exact location and size of DG
ends of each branch of the distribution system, power is calculated by using exact loss formula. Using the exact loss
generation or consumption of each static shunt- compensating formula sensitivity factors is calculated and bus with highest
device and total system losses. sensitivity factor is selected as bus with highest power loss.
There are many solution techniques for load flow analysis.
A. Problem Formulation
The solution procedure and formulations can be precise with
values adjusted or unadjusted. The acceptable load flow In order to place the DG unit in the network, we must
analysis method should meet the following requirements: calculate the loss coefficients by using Exact Transmission
 They should have high speed and low storage Loss Formula and it is derived below. The bus power Si
requirements, especially for real time large system injected into bus can be represented as generated power minus
applications, as well as multiple case and interactive the bus load. By adding all n bus powers, we therefore obtain
applications. the total generated power minus the total load. i.e., we obtain
 They should he highly reliable, especially for the network losses.
n n
ill-conditioned problems, outage studies and real-time
applications. PL  jQ L   S i   Vi I i* (4)
i 1 i 1
 They should have acceptable versatility and simplicity.
Where
The approach utilized forward and backward sweep
PL  Total transmission loss
algorithm based on Kirchoff’s Current Law (KCL) and
QL  Total reactive power loss
Kirchoff’s Voltage Law (KVL) for evaluating the node
Si  Bus power injected into jth bus
voltages iteratively. In this approach, computation of branch
Vi  Bus voltage at ith bus
current depends only on the current injected at the
Ii  Bus current vector of ith bus
neighboring node and the current in the adjacent branch. This
The total real power loss in power systems is represented by
approach starts from the end nodes of sub lateral line, lateral
(5), popularly known as “exact loss formula” [4]
line and main line and moves towards the root node during n
branch current computation. The node voltage evaluation
begins from the root node and moves towards the nodes
PL   [
i 1, j 1
ij ( Pi Pj  Qi Q j )   ij (Qi Pj  Pi Q j )] (5)
located at the far end of the main, lateral and sub lateral lines
[3]. rij
Voltage at any node is given by
where  ij  Cos ( i   j )
ViV j
Vn  Vn 1  I b Z b (1)
rij
Where Vn-1 = voltage at (n-1)th node. and  ij  Sin( i   j )
b = (n-1) ViV j
Ib = Current in the branch b i and j in the suffix indicates values at ith and jth nodes
Zb = Impedance of the branch b respectively.
The real and reactive power loss in the network is given by,
Nb B. Types of DG
Real power loss, P   | I b | Rb 2
(2) DG can be classified into four major types based on their
b 1 terminal characteristics in terms of real and reactive power
Nb delivering capability as follows:
Reactive power loss, Q   | I b |2 X b (3) 1. DG capable of injecting P only.
b 1 2. DG capable of injecting Q only.
3. DG capable of injecting both P and Q
IV. DG SIZING AND PLACEMENT 4. DG capable of injecting P but consuming Q.
Even though power contribution to the transmission system Photovoltaic, micro turbines, fuel cells, which are
by DG is small, it is necessary in the present power scenario as integrated to the main grid with the help of
many power companies are focusing on reduction of losses in converters/inverters, are good examples of Type 1. Type 2
the system and maintaining good voltage profile. In a large could be synchronous compensators such as gas turbines. DG
transmission system network with high power losses, it is units that are based on synchronous machine (cogeneration,
difficult to select a particular bus from many buses so as to gas turbine, etc) fall in Type 3. Type 4 is mainly induction
place a DG unit for loss reduction. Power losses are present at generators that are used in wind farms. Type 3 DG is studied
every bus and identification of bus with highest power loss is in this paper.
important because losses at that bus includes majority of total

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569) Vol. 10, Issue. 1 ,Nov. 2014
B. Venkatesh Reddy 15

C. Sizing at various locations V. COMPUTATIONAL PROCEDURE

-1
Assuming a = (sign) tan (cos (PFDG)), the reactive power When power factor of DG is set to be equal to that of
output of DG is combined total loads, computational procedure to find
QDGi = a PDGi (6) optimal size and location of one of four types of DGs is
in which described in the following.
sign = +1: DG injecting reactive power; Step l: Run load flow for the base case.
sign = -1: DG consuming reactive power: Step 2: Find the base case loss using (5).
PFDG is the Power factor of DG. Step 3: Calculate power factor of DG using (11).
The active and reactive power injected at bus i, where the Step 4: Find the optimal size of DG for each bus using (9)
DG located are and (8).
Pi = PDGi – PDi (7) Step 5: Place DG with the optin1al size obtained in step 4 at
Qi = QDGi – QDi = aPDGi – QDi (8) each bus, one at a time. Calculate the approximate
loss for each case using (5) with the values α and β of
The active power loss can be rewritten and the total active the base case
power loss of the system is minimum if the partial derivative Step 6: Locate the optimal bus at which the total loss is
of with respect to the active power injection from DG at bus i minimum corresponding with the optimal size at that
becomes zero. After simplification and rearrangement we get bus.
the optimal size of DG at each bus i [5] for minimizing loss Step 7: Run load flow with the optimal size at the optimal,
can be written as location obtained in step 6. Calculate the exact loss
using (5) and the values α and β after DG placement.
 ii ( PDi  QDi )   ii (aPDi  QDi )  X i  aYi It is noted that when the type and power factor of DG is
PDGi 
a 2 ii   ii given; the computational procedure to find the optimal size
where (9) and location of DG is as described earlier apart from step 3.
n At this step, the power factor of DG is entered rather than
Xi   (P 
j i , j  i
j ij  Q j  ij )] using (11). In Exhaustive Load Flow (ELF) method, optimal
location and power factor are obtained with a number of load
n flow solutions.
Yi   (Q 
j i , j  i
j ij  Pj  ij )]
VI. RESULTS AND DISCUSSION
In the above expression PDGI gives the optimal capacity of This chapter provides the case of IEEE 33 Bus radial
DG unit to be placed in transmission system. The power distribution system. The single line diagram of IEEE 33 Bus
factor of DG depends on operating conditions and type of radial distribution system is shown in figure 5.1.
DG. When the power factor of DG is known, the optimal size
of type 3 DG at each bus i for minimizing losses can be found
by assuming 0 < PFDG < 1, sign = +l and "a" is a constant, the
optimal size of DC at each bus i for the minimum loss is given
by (9) and (6) respectively.
D. Optimal Location
For optimal location, first the optimal sizes at various
locations have been calculated for different types of DG and
the losses were calculated with optimal sizes for each case.
The case with minimum losses is selected as the optimal
location for each type of DG. Based on this method, one can
avoid exhaustive computation and save time, especially for
large-scale distribution systems as trend of loss reduction can
be captured with α and β coefficients from the bases case.
Fig. 1 Single Line Diagram of IEEE 33 Bus System
E. Optimal Power Factor
In the distribution system root node is bus 1, main line has
The power factor of the single load [6] is given by eighteen buses including bus 1 and lateral line emanating
P from three buses i.e., bus 2, bus 3 and bus 6. There are no sub
PF  (10) lateral lines and minor lines present in the system.
2 2
P Q The load flow study is conducted on the bus system and the
It can be proved that at the minimum loss occur when power results are carefully noted down. The losses and voltage
factor of DG is equal to the power factor of load. profile of each bus is noted down, by using the computational
procedure mentioned above the optimum power factor and
PDG the optimum values of DG to be placed at each bus is
PF  PFDG  (11)
2
PDG 2
 QDG obtained. The analysis by placing DG is done and the results
are given in this section.

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569) Vol. 10, Issue. 1 ,Nov. 2014
B. Venkatesh Reddy 16

decreased for the other buses as the branch impedances comes

1
into action.
0.9 1
Without DG
0.8 With DG
0.98
0.7
0.96
) 0.6
A
V .)
M( 0.5 .u
(p 0.94
e e
iz d
itu
S
0.4 n
g
G a
D M
0.92
ge
0.3 at
l
o
V

0.2 0.9

0.1
0.88

0
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
Bus Number 0.86
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
Fig. 2 Optimal DG Sizes at all buses for IEEE 33 Bus System Bus Number

Figure 2 shows the bar representation of optimal DG Sizes Fig. 5 Voltage at all buses with DG at bus 8 and without DG
at all buses for IEEE 33 Bus System. From the figure we can Figure 5 shows voltage at all buses with DG at bus 8 and
observe that the DG size do not follow a regular manner and without DG. The voltage profile was improved and it can be
the size is independent of location of bus (main line or lateral seen as difference between the lines blue and red. The bus
line). with more voltage improvement is 18 and bus with less
300
voltage improvement is bus 2.
280 1
Without DG
With DG
260 0.98

) 240
W 0.96
(K .)u
s 220 .p
s (
oL 0.94
r ed
e
w
200 uti
n
o ga
P
la M 0.92
e 180 eg
R
ltao
160 V 0.9

140
0.88

120
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 0.86
Bus Number 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
Bus Number
Fig. 3 System Total Real Power Loss with optimum DGs at all buses
Fig. 6 Voltage at all buses with DG at bus 7 and without DG
Figure 3 shows the graphical representation of system total Figure 6 shows voltage at all buses with DG at bus 7 and
real power loss with optimum DGs at all buses with optimal without DG. The voltage profile was improved and it can be
DGs placed at each bus. The lowest value of loss is around seen as difference between the lines blue and red. The buses
140 KW for bus 8. It is clearly seen from graph the optimal with more voltage improvement are 17 and 18 and buses 19,
location is bus 8 as total losses of system is less. DG of size 20 and 21 are with less voltage improvement.
0.932 MVA is placed at bus 8 and the graphs are plotted using 1
Without DG
MATLAB. With DG
80 0.98
Without DG
With DG
70 0.96

).
60 .u
p( 0.94
e
d
uit
n
g
) 50 a
W M 0.92
e
(K g
s tal
os o
L V
r
e
40
w 0.9
o
P
e
v 30
ti
c
A
0.88

20

0.86
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
10 Bus Number

Fig. 7 Voltage at all buses with DG at bus 9 and without DG

Figure 7 shows voltage at all buses with DG at bus 9 and
0
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
Branch Number

Fig. 4 Line Active Power Loss with DG at bus 8 and without DG without DG. The voltage profile was improved and it can be
Figure 4 shows line active power loss with DG at bus 8 and seen as difference between the lines blue and red. The bus
without DG. The losses are reduced and it can be seen as with more voltage improvement is 18 and bus with less
difference between the lines blue and red, loss is considerably voltage improvement is bus 2.
reduced for buses one to eight excluding 6. Losses are not

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569) Vol. 10, Issue. 1 ,Nov. 2014
B. Venkatesh Reddy 17

1 determine the optimal location among the buses present in the

Without DG
With DG
network. The reliability of electric power systems can be
0.98
enhanced by distributed generation, as the system is less
0.96 dependent on centralized facilities. The use of distributed
)
u.
p(
generators at selected locations can also help distributors
e
d 0.94 overcome local bottlenecks.
tui
n
g
a
M 0.92 A. Future Scope
e
g
at
lo The proposed method can be extended for a network with
V 0.9
large number of buses. An ideal power system, which has the
0.88 benefits of DG as well as grid, can be formulated by
accommodating these DGs into the grid along with central
0.86
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
power units so that surplus power generated by DGs can
Bus Number easily sent to the regions as backup power and as peak load
Fig. 8 Voltage at all buses with DG at bus 31 and without DG sharing.
Figure 8 shows voltage at all buses with DG at bus 31 and Since India is a developing country, with availability of
without DG. The voltage profile was improved and it can be huge amount of renewable energy sources like solar, biomass,
seen as difference between the lines blue and red. The bus wind and wave energy. So some amount of energy is
with more voltage improvement is 33 and bus with less contributed to meet the power demands. Therefore it is
voltage improvement is bus 2. necessary to locate more DGs and they must be connected to
Table : Summary of Results for IEEE 33 Bus System (PFDG=0.83)
the distribution system.

Bus TPL (KW) TQL (KW) Voltage (p.u)

No
ACKNOWLEDGMENT
* # % * # % * # &
I express my deep sense of gratitude to Dr. A. Lakshmi
7 292.4 173.7 40.5 196 113.3 42.2 0.925 0.958 3.57 Devi, Associate Professor, Department of Electrical and
8 292.4 146.2 50.0 196 93.8 52.1 0.903 0.958 6.11 Electronics Engineering, Sri Venkateswara University
9 292.4 167.8 42.8 196 107.9 44.9 0.893 0.949 6.24 College of Engineering. I am greatly indebted to her for her
valuable guidance, constant encouragement and timely
31 292.4 169.0 42.1 196 117.7 39.9 0.894 0.924 3.24
directions for bringing out this paper.
* without DG # with DG
% Percentage Loss reduction & Voltage increment
REFERENCES
Table shows the summary of results for IEEE 33 bus
system. The voltage increment is more at bus 9 compared to [1] Manu Avinash. G and Krishna Kumar, “Distributed
bus 8, but the loss reduction is very less so choice of optimal Generation and its Social Impact,” Energy Saving Now, Vol.7,
bus is justified. DG placement improves voltage profiles of September 2010.
the adjacent buses more compared to the buses far away from [2] Gregory W. Massey, “Essentials of Distributed Generation
the installation; this is also seen from the figures. DG at 7 or 8 Systems,” Jones & Barlett Publishers, Canada, pp. 7-13, 2010.
or 9 improves voltage of bus 18 but DG at bus 31 improved [3] K. Vinoth Kumar and M. P. Selvan, “A Simplified Approach
for Load Flow Analysis of Radial Distribution Network,”
the voltage of bus 33.
International Journal of Computewr and Information
Engineering, vol. 2, April 2008.
VII. CONCLUSION [4] D. P. Kothari and J. S. Dhillion, “Power System Optimisation”,
DG has the potential to play a major role as a compliment or Prentice Hall of India Private Ltd., New Delhi, 2010.
alternative to the electric power grid under certain conditions. [5] Duong Quoc Hung, Nadarajah Mithulanathan and R. C.
Bansal, “Analytical Expressions for DG Allocation in Primary
DG is fundamentally distinct from the traditional central plant
Distribution Network,” IEEE Trans. on Energy Conservation,
model for power generation and delivery in that it can deliver vol. 25, no. 3, September 2010.
energy close to loads within the power distribution network. [6] V. K. Mehta and Rohit Mehta, “Principles of Power System”,
The need for new capacity and DG technology advancements, S. Chand & Company Ltd., New Delhi, 2010.
are collectively laying the groundwork for the possible
widespread introduction of DG.
The operating conditions of a power system after B. Venkatesh Reddy has completed his B. Tech
connecting DG sources can change drastically as compared to from S.V.U College of Engineering. He is
the base case. The planning of DG installations should pursuing his M. Tech, Power Electronics from
ASRA, Affiliated to JNTUH, Hyderabad,
therefore, consider several factors: the best technology to be
Telangana. His areas of interest are Control
used, the number of units of DG and capacities, where should Systems, Power Electronics and Power Systems.
they be installed and type of connection should be used etc., E-mail id: [email protected]
The problem of DG allocation and sizing should be
approached carefully. If DG units are connected at non
optimal locations, the system losses can increase, resulting in
increased costs. This shows that all the sites are not
practicable for DG implementation and it is necessary to

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569) Vol. 10, Issue. 1 ,Nov. 2014