OPTIMAL LOCATION FOR LOSS REDUCTION ON

A 7-BUS BAR POWER GRID

SYSTEM BY CAPACITOR PLACEMENT

 Adam Misbawu Department of
Electrical/Electronic Engineering
Kumasi Technical University
Kumasi, Ghana
Adam.misbawu@kstu.edu.gh
Solomon Nchor A.
Department of Physics
Kwame Nkrumah University of Science and
Technology Kumasi, Ghana
sanchor@knust.edu.gh

Margaret Richardson Ansah

Department of Computer Engineering
University of Ghana
Accra, Ghana
marichardson@ug.edu.gh
Abstract
Reduction of power loss in transmission and distribution
system is key to improve the efficiency of power system.
This paper presents a method of minimizing the loss
associated with the reactive component of the branch
current by optimal placement of shunt capacitor on an
IEEE seven Bus-bar system by Power-World Simulator.
The project started by drawing the one-line diagram of
the 7-bus power grid system with rated values suggested
by the simulator tool.
Upon completion, all the data regarding the Voltage per unit, mega
volt-ampere reactive losses, megawatt and mega volt-ampere at each
buss were recorded. Analysis was done on the system using the
power world simulator.
From the simulation results, the original system recorded a low power
factor of 0.79 and a mega volt-ampere reactive loss of eighteen
percent (18%).
After analyzing the system and installing the capacitor bank at the
optimal location, the power factor improved to 0.98 and the system’s
mega volt-ampere reactive reduced to 6.4%.
The optimal location for the capacitor bank was identified as the bus
that injects more reactive power to the system. The bus also has a low
voltage per unit value and high conductance as compared to all other
buses.
INTRODUCTION
Transmission and distribution losses are a major
concern in Ghana.
An article published by Lawrence Quartey (2015) in the
Africa report on Ghana energy crisis, in May 2015
placed Ghana on top of the top 30 countries in the
world that loose most of its electrical energy transmitted
and distributed.
Ghana loses 26% of its electrical energy transmitted and
distributed.
The level of reported losses in a year is influenced by, both
technical and operational losses.
Studies have indicated that as much as 13% of total power
generated is consumed as loss due to the heating of conductors
at the distribution level [1].
[2] were able to reduce losses on power lines using
Compensation devices. In their work, it was noted that, if surge
impedance loading of the line corresponds to the line loading,
then a flat voltage profile can be achieved.
The compensation device was chosen bearing in mind, the
above-mentioned idea. However, if the appropriate
compensation device is installed in line, the effective surge
impedance will yield a virtual natural load canceling the
actual load of the line.
But in the practical system, the demand curve is not smooth, it
varies with time. To have surge impedance loading equal to
the natural loading of the line, compensation device should be
operated according to the variable load condition.
 [3] implemented shunt capacitive in his work. In his work,
the capacitor was connected at the substation or at the
midpoint of the line. It was not uniformly distributed
throughout the line and automatic control of the shunt
capacitive compensation was used to maintain the voltage at a
constant level. The effect of shunt capacitive compensation is
continuously distributed throughout the line.
[4] commented on the shunt capacitive compensation that it is
required when the alternator active power output is greater
than the surge impedance loading of the line. Huge amount of
power can be transmitted by doing shunt compensation. But in
the practical system, it requires large number of capacitors.
Capacitor Allocation and Network Reconfiguration
This micro level objective can be approached by doing capacitor
allocation at the sensitive load buses with optimum value [5]. Optimal
capacitor placement of the network balances the conglomerated load.
If there is a lump of load at a certain bus, then by shifting the load
efficiently to another light load bus can reduce the active power loss. It
also stabilizes the system and maintains the nominal voltage at the
buses. Though with the prevailing condition of Power System,
capacitor allocation and network reconfiguration is very much tough
and cumbersome approach to reduce the line loss, but it is a useful and
less hazardous way to minimize the line loss economically. However
physically it is a well understood fact that capacitor allocation and
network reconfiguration is necessary for loss minimization. [6].
 Power Factor Improvement
The importance of installing the capacitor is to improve the
power factor. In the solution to improve the power factor by
[7], power factor correction capacitors were added to the plant
power distribution system to act as reactive power generators
and provide the needed reactive power to accomplish kW of
work. In his work, KVAR was not considered. However, [3]
argued that One should consider the average value of the
active power as well as the average power factor in the
system. Using these two values, one can calculate the
capacitor bank for the average operating condition needed to
improve the overall power factor.
Optimization Strategies for minimizing Power Losses
Analytical approach, genetic algorithm approach, ant colony
approach, particle swamp and heuristic approach were adopted for
power loss minimization by capacitor placement and network
reconfiguration in [8].
Loss reduction by installing capacitors on primary feeders was
studied in [9]. In his work, capacitor installation at the load end was
beneficial rather than placing it in the substation. However, a method
proposed by [8] differ from [9]. he argued that for optimum size and
location of shunt capacitors for reduction of losses in distribution
feeder, the capacitor should be placed on the distribution feeder by
doing power factor correction. By this approach, the power quality
was also improved.
Particle Swarm Optimization (PSO) has been implemented for
optimal capacitor placement by [10]. in their work, AC load
flow calculation and harmonic load flow calculations were
done. The type of capacitor is tested in every compensation
and if the iteration number reaches the maximum number of
iterations, then the computation procedure is stopped, and the
best solution or particle is considered as the optimal solution.
Otherwise, computation procedure continues until the iteration
number reaches the maximum iteration number. PSO has not
proven to solve the problem of power loss due to its
deficiency in a number of iterations needed to be performed.
Heuristics Search Strategies for capacitor placement has been studied
in the distribution system by [11]. In their approach, cost saving was
given more importance than loss reduction. The computation time and
the complexity of the simulation procedure were significantly reduced
by this heuristic approach. The test system which is considered for
simulation is a balanced three phase network with balanced three
phase voltages though the effect of harmonics was not considered in
their work, but they realized that if the peak power loss for the test
system is computed and it is checked for each bus to find out the
largest effect of reactive power compensation, then the sensitive bus
or node is selected on the basis of highest impact or change in loss and
the capacitor should be placed at that particular node to reduce the real
power loss.
 SYSTEM DESIGN
 
Designing the One-Line Diagram of the Power flow
System
The one-line diagram used in this project work was
created with the Power World Simulator software tool, the
software provides all the necessary quantities required by a
typical power grid system for the purpose of analyzing the
power system.
The quantities include generators, buses, circuit breakers,
transmission lines, filters, and three phase loads. The
modeled power system for this work is as shown in Figure1.
The system consists of seven 138 kV buses, five generators, eleven
transmission lines with circuit breakers, 6 static loads and one power
factor correction capacitor (PFC). The software allows changes to be
made on the quantities to display actual electrical quantities on the
one-line diagram. For instance, the per unit (PU) voltages and phase
angles in degrees are provided for each voltage bus. The actual
power outputs in real and imaginary are displayed for each of the
generators. The actual power flows in real and imaginary are
provided for all transmission lines. The ratings of the connected load
and PFC are included on the one-line as well.
The power system one-line diagram created using the software is
then used as the main work for the project. Figure 1 shows the one-
line diagram of the power flow system.
Schematic diagram for the One-line 7
Buss Bars Power Grid System.
 The Load Flow Analysis for the One-Line Diagram
After completing the one-line diagram of the power system, the
simulation menu was run and the Newton-Raphson type of load flow
was selected to be performed on the network.
This is because there are little of no errors in the Newton-Raphson
iterations, it also takes less iterative cycle to reach the solution and it
can be used to solve high voltages and gives accurate values. The
magnitudes and angles of the voltages are determined for each bus
using the Power World Simulator. The resulting power flows are
shown as arrows as triangles. Green represents real power and blue
represents imaginary power. The movement of the arrows along the
transmission lines represents actual direction of real and imaginary
power flow. The arrows are scaled to show the magnitude of the
powers flowing on each transmission line.
Table 1 Power flow values for bus-bar two
Fig. 2 Configuration of Bus-Bar Two.
 Load Flow Analysis Results
The Newton-Raphson load flow analysis was performed on the
power system described in Figure 1. The animated results of this
load flow are shown in Figure 2. For every transmission and
distribution network, one of the bus bars are chosen as the reference
or the slack bus, in this project work bus 7 was set up to be the
reference or the slack bus with 1 per unit output voltage at 0 ° phase
angle. The connected load was 201 MW and 62 MVAr (inductive).
The resulting bus voltages and phase angles are shown on the one-
line diagram. The resulting power flows for each transmission line
feeding the connected load are also displayed. The pie charts
associated with each transmission line provides an indication of
transmission line loading.
For this work, the resistance of each transmission line was not
included, only the reactive component. The power flows (real and
imaginary) for the purpose of this work are in the same direction.
Original System Settings for the One-Line Diagram
Table 2 shows the performance of the various buses for the original
settings of the power grid system.
Bus 1 uses 93% of its MVA value which translates to a high
performance of the bus. This indicates little loss on the bus. Bus 1 has
a 103 MW, 19 MVAr generator connected to it. Bus 1 has no load
connected to it but supplies a total of 110.2 MVA to both bus 2 and
bus 3. The per unit voltage on the bus is 1.05 at an angle of 6.17. The
power factor for the bus is 0.93.
Table 2 Improved data after installing the
capacitor on busbar two
Bus 2 has a high MVAr loss of about 84.41MVAr and it uses
only 86% of its MVA value. Bus 2 has a total active power
of 165 MW and an apparent power of 190 MVA. The per
unit value of the bus is 1.04 at an angle of 4.22. Bus 2 has a
load of 40 MW, 20 MVAr connected to it. Bus 2 is connected
to bus 3, bus 4 and bus 6.
The slack bus chosen for the system is bus 7. It has a power
factor of about 0.93 and a per unit value of 1.04 at 0 degrees.
It has an active power of 201 MW, a reactive power of 62.24
MVAr and an apparent power of 215.5 MVA.
From Table 2 bus 6 performs better with a power factor of
0.97 and as low as 6.24 MVAr losses better than the rest of the
buses, whiles bus 2 performs poorly with as low as 0.86
power factor and a significant 84.41MVAr loss among the rest
of the buses. On the other hand, as bus 2 is performing poorly,
the losses on bus 7 is also high making the entire system
unhealthy.
Figure 3 illustrates the variations of the MVA, MW and MVAr
values of the original system. To be totally convinced that bus
2 has the greatest loss on the system, the load MVAr on bus 2
is increased.
Fig 3. Schematic Diagram for the One-Line
Power Grid System during Simulation
PWS was used to design the one-line power flow system and
considering one of the seven busses as the slack or reference bus.
The system parameters are the standard IEEE 7-bus bar values.
The system was run and the data on each bus was recorded. The
system initial settings recorded a low power factor of 0.79. The
system was then disturbed and showed more loss especially on bus
2. A capacitor bank rated 60 MVAr was installed on bus 2, and the
system experiences an increase in power factor of 0.93. A capacitor
bank rated 80 MVAr was installed on bus 4 and showed a
phenomenon increase in power factor of 0.98.
Finally, 5% of the total power injected to the system was lost due to
the impedance of the transmission lines, when the estimated reactive
power of 100 MVAr of the capacitor bank installed on bus 4.
Fig 4. Schematic diagram for the 7 bus-bars power grids
after installation of the capacitor bank on each bus bar.

Fig 4. Schematic diagram for the 7 bus-bars power grids after

installation of the capacitor bank on each bus bar.
RESULT AND DISCUSSION
From the simulation results, it was found out that.
The original system recorded a low power factor of 0.79 which
translates to seventy-nine percent (79%) usage of the total power to
the system and an MVAr loss of eighteen percent (18%).
After analyzing the system and installing the capacitor bank at the
optimal location, the power factor improved to 0.98 which translates
to ninety-eight percent (98%) usage of the total power to the system,
and the system’s MVAr reduced to 6.4%. The optimal location for the
capacitor bank was identified as the bus that injects more reactive
power to the system.
The bus also has a low voltage per unit value and high conductance as
compared to all other buses.
 
When the same capacitor was installed on bus bar
4, the MW and the MVAr lines were in phase
indicating tremendous improvement than when it
was installed on bus 2. As indicated in the Fig 4.14,
the value of the capacitor has been raised from 60
MVAr to 80 MVAr resulting into optimal loss
reduction in the system.
CONCLUSION
The project started with the drawing of the one-line diagram of 7
buses power grid system with rated values of components needed as
suggested by the simulator tool. The project was further analyzed by
categorizing the analysis by cases. For case 1, the value of the load
MVAr demand at the selected bus resulted from the previous findings
was doubled the original size. This was to show more convincing
result of MVAr losses at the bus.
For case 2, the capacitor bank was installed on each buss one at a
time to indicate the optimal location of the bus that can improve the
whole system the most. All the MVAr losses for each installation of
the capacitor bank were recorded. For case 3, the capacitor values
were varied at the selected bus. This was to estimate the capacitor
range needed for power loss minimization.
Finally, the analysis required finding the optimal capacitor location
to minimize power losses, the appropriate size and rating for the
capacitor, the percentage value of power losses reduction for the
power system and the kVAr rating for the desired Power Factor was
calculated.
It can be concluded that, for an IEEE 7-Bus bar system loss
minimization, the bus that will improve the entire system the most
when the shunt capacitor is installed on it should be the bus that
inject more reactive power to the system.
The bus should have a low voltage per unit value as compared to all
other busses in the system. Again, the bus should have low resistance
and high conductance.
RECOMMENDATIONS
Implementation of optimal placement of multiple capacitor banks on
transmission and distribution network can be a good approach for
loss reduction.
Transformer tap changing option with capacitor allocation and
distributed generation may be done for the same reduction of real
power loss. A suitable and optimal combination for both schemes
can be studied for the extension of the work. The findings from this
project work were based on the IEEE seven bus bar system, the
findings can be extended to other IEEE buses.
By placing the shunt capacitor to the original 7-bus system and
improving the systems power factor, it can be recommended for
future work to be performed on the analysis to determine the cost
when the system power factor is improved.
THANK
YOU
REFERENCS
[1] H.N.Ng.M.M.A.Salama (1995). Fuzzy Optimal Capacitor Sizing and Placement. Research, 87, Elsevier.
[2] Neagle, N.M., and Samson, D.R. (1956) ‘Loss reduction from capacitors installed on primary feeders’, AIEE Trans.
Vol.75,PartIII.
[3] Khodr, H.M., Olsina, F.G., Olivera-DeJesus, P.M.D., and Yusta, J.M. (2008) ‘Maximum savings approach for location
and sizing of capacitors in distribution systems’, Electric Power Systems Research, 78, Elsevier.
[4] Songur, M and Ercan B. POWERHU- (1997) A PC Based Electric Power System Software Package for Electric Power
System Courses, NASAR: IEEE.
[5] A.M.Sharaf and S.T. Ibrahim (1996). Optimal capacitor placement in distribution networks. Electric Power System
Research 37, Elsevier
[6] Chis, M., Salam, M.M.A., and Jayaram, S (1997), ‘Capacitor placement in distribution systems using Heuristic Search
strategies’, IEEE Transactions on Elec. Eng. Vol. 144.
[7] M.H. Shwehdi and M.R. Sultan. (2000), Power Factor Correction Capacitors; IEEE Power Engineering Society
Summer Meeting, Vol.3,
[8] M.V.Vitornio, R., M.M.Jorge, H., and M.P.Neves, Luis. (2009) ‘Network reconfiguration using a Genetic Approach
for loss and reliability optimization in distribution systems’, IEEE Power Eng., Lisbon, Portugal.
[9] Chang, C-F. (2009) ‘Optimal reconfiguration and capacitor placement by robust searching. http://dx.doi.org
[10] Yu, X-m., Xiong, X-y., and Wu, Y-w. (2004) ‘A PSO based approach to optimal capacitor placement with harmonic
distortion consideration’, Electric Power Systems Research 71
[11] Mekhamer, S.F., El-Hawary, M.E., Soliman, S.A., Mustafa, M.A., and M.M Mansour. (2002) ‘New Heuristic
Strategies for reactive power compensation of radial distribution feeder’, IEEE Transactions on strategies of power loss
reduction, Vol.13, No. 1