Science and Technology Publishing (SCI & TECH)

ISSN: 2632-1017
Vol. 8 Issue 2, February - 2024

Power Loss Minimization In IEEE 33 Bus

Network Through Optimal Location And Sizing
Of Distributed Generation Using Particle Swarm
Optimization Technic
Ezuruike Okafor S. F. 1
Department of Electrical/Electronic Engineering
Imo State Polytechnic,
Omuma , Imo State , Nigeria

Agbaji Adai Samuel 2

Department of Computer Engineering
University of Uyo,
Uyo,Akwa Ibom State, Ngeria
[email protected]

Maduka N.C. 3
Department of Physics
Federal University Gusau,
Zamfara State Nigeria
[email protected]

Abstract—In this study, power loss minimization 1. INTRODUCTION

in IEEE 33 bus network through optimal location Distributed Generation (DG) is as small-scale power
and sizing of distributed generation using Particle generation technologies installed close to the utilisation
Swarm Optimization technic is presented. Load
point, and this includes solar photovoltaics, wind turbines,
flow analysis was conducted using the Newton
Raphson method for the baseline scenario where biomass systems and fuel cells, with each having its own
there was no distributed generation (DG) in the strengths and applications [1,2,3]. Again, [4] and [5] stated
network. The case study IEEE 33 bus has a total that DG units can be installed as island grids, or they could
real power demand of 3715 kW and a total reactive also be linked with the existing grid, helping to increase
power demand of 2300 kVar. 201.891 the total energy security while facilitating the process of
reactive power loss of 134.641kW which amounts electrification and energy transition (towards renewable
to 5.853956522 % of the total real power is
energy sources).
observed. The DG placement was done using the
Particle Swarm Optimization (PSO) technic for 1 Importantly, DG plays a key role in modern power
DG, 2DG and 3 DG. The PSO optimization at 30% systems [7]. Firstly, DG improves energy security by
penetration shows that the introduction of DGs reducing the dependency on centralised power plants for
considerably reduced the power losses and also electricity generation [8,9]. Again, [10] went further to
improved on the voltage stability in the power elaborate that DG reduces transmission and distribution
system. The scenario with 2 DGs provides the losses because the generated power is always near their
most substantial improvements, balancing power
consumption points. Secondly, [11] contributed that DG
loss reduction and voltage profile enhancement
effectively. Specifically, the scenario with 2 DGs enables more renewables integration into the energy
gave the highest (over 65 %) reduction in both system, which is vital for energy transition and shrinking
reactive and reactive power with the lowest VDI of the carbon footprint. Also, [12], highlighted that
2.61. The first DG of 567 kW was located at bus economically, adding DG defers the investments on
number while 14, while the second DG of 567 kW building new large-scale power plants with high-capital
was located at bus number 32. In all, the PSO load
outlays, creates jobs in local economies and stimulate
placement and sizing showed remarkable
reduction in the power losses and the best technological innovation for smart grids and energy
performance occurred with two DG scenario. storage. Distributed generation can also be used to
minimize power loss [13,4].
Keywords— Loss Minimization, IEEE 33 Bus However, in order to realize the aforementioned
Network, Optimal Location and Sizing, Distributed
benefits of DG in the power system, the DG must be
Generation, Particle Swarm Optimization Technic

www.scitechpub.org
SCITECHP420332 1852
 Science and Technology Publishing (SCI & TECH)
ISSN: 2632-1017
Vol. 8 Issue 2, February - 2024

optimally sized and optimally located within the power Generally, the optimal placement and sizing of
distribution network. Consequently, in this paper, loss distributed generation on distribution network will lead to
minimization in IEEE 33 bus network through optimal reduction in power. The actual percentage of reduction
location and sizing of DG using particle swarm achieved depends on some factors among which is the
optimization (PSO) technic is presented. The study approach used. In this study, the total power demand on the
considered the baseline low flow analysis where there is no distribution network is determined for the baseline scenario
DG. The Newton Raphson approach is used for the load where there is no distributed generation (DG) in the power
flow. Further simulation are then conducted for different network. The load flow analysis is conducted using the
scenarios where 1 DG, 2 DG and 3 DGs are installed using Newton Raphson method presented by [15] and the
the PSO to determine the optimal size and location of each flowchart is shown in Figure 1.
of the DGs on the power distribution network. The results The DG placement is done using the Particle
are presented and discussed; detailing the effectiveness of Swarm Optimization (PSO) technic. Three cases were
PSO to minimize the power loss and enhance voltage considered in this study, namely, a case of 1 DG, another
stability through optimal DG installation. case of 2 DG and finally 3 DG. Each of the three cases
were implemented using the PSO algorithm captured in the
2. METHODOLOGY pseudo code given in section 2.2. The control parameters of
the PSO algorithm used to conduct the study are presented
2.1 The load flow analysis for the baseline scenario in Table 1. The line data of the IEEE 33 bus are given in
Table 2 while the load data are given in Table 3

www.scitechpub.org
SCITECHP420332 1853
 Science and Technology Publishing (SCI & TECH)
ISSN: 2632-1017
Vol. 8 Issue 2, February - 2024

Start

Read load data:

Primitive Matrix, Slack Us Voltage,
Real and Reactive Bus Power, etc.

Formulate Y-matrix using load flow data

Make initial assumptions, Vi and θi for I = 1,2,…,n

Set iteration count i = 0

Compute Pi and Qi for i= 1,2,…,n sing power flow

equations

Evaluate power mismatch; ΔPi and ΔQi

NO
Max ΔPi >ε
Max ΔQi >ε

Calculate line flows and line

YES losses

Evaluate Jacobian Matrix Elements

ΔP 𝐽1 𝐽2 Δθ
Output Load Flow Solution;
ΔQ 𝐽3 𝐽4 ΔV Voltage Magnitude, Phase Angle,
etc

Evaluate increments and update variables

End

i =i+1

Figure 1 The flowchart for the Newton Raphson Algorithm-based load flow analysis for the baseline scenario
without DG (Source: [1])
2.2 The Particle Swarm Optimization (PSO) vii. End for
Pseudocode for DG Placement and Sizing viii. Evaluate the fitness of each solution (i.e., power
loss and voltage stability for the given DG
i. Initialize necessary parameters and candidate placement and sizing);
locations for DG; ix. Update the best personal and global positions;
ii. while not termination do: x. Update particle positions and velocities based
iii. Generate particle swarm; on the best found positions;
iv. for each particle do: xi. End while
v. Select a candidate location for DG based on xii. Return the best solution found;
particle's position; xiii. End procedure.
vi. Determine the size of DG at the selected
location;

www.scitechpub.org
SCITECHP420332 1854
 Science and Technology Publishing (SCI & TECH)
ISSN: 2632-1017
Vol. 8 Issue 2, February - 2024

Table 1: Control parameters of the PSO algorithm used to conduct the study
Parameters Values Explanation
Defines the number of search agents (particles) exploring the solution
Population size 50
space simultaneously.
Sets the maximum number of times the PSO loop (particle movement
Maximum iterations 100
and update) will be executed.
Minimum weight of Controls the exploration-exploitation balance of particles. Higher
0.4
inertia (𝜔 ) 𝜔 allows wider exploration initially.
Maximum weight of This value influences the exploitation behavior in later iterations,
0.9
inertia (𝜔 ) guiding particles towards promising regions of the search space.
Social acceleration Influences exploitation in later iterations, guiding particles towards
1.5
coefficient (𝑐 ) promising regions.
Controls the influence of a particle's personal best (Pbest) on its
Cognitive acceleration
2.0 movement. Allows particles to explore promising regions based on
coefficient (𝑐 )
their experiences.

Table 2: The line data of the IEEE 33 bus test system (Source :[16])

Line Data
S/N From Bus To Bus Line Resistance (Ω) Line Reactance (Ω)
1 1 2 0.0922 0.047
2 2 3 0.493 0.2511
3 3 4 0.366 0.1864
4 4 5 0.3811 0.1941
4 5 6 0.819 0.707
6 6 7 0.1872 0.6188
7 7 8 0.7114 0.2351
8 8 9 1.03 0.74
9 9 10 1.044 0.74
10 10 11 0.1966 0.065
11 11 12 0.3744 0.1238
12 12 13 1.468 1.155
13 13 14 0.5416 0.7129
14 14 15 0.591 0.526
15 15 16 0.7463 0.545
16 16 17 1.289 1.721
17 17 18 0.732 0.574
18 2 19 0.164 0.1565
19 19 20 1.5042 1.3554
20 20 21 0.4095 0.4784
21 21 22 0.7089 0.9373
22 3 23 0.4512 0.3083
23 23 24 0.898 0.7091
24 24 25 0.896 0.7011
25 6 26 0.203 0.1034
26 26 27 0.2842 0.1447
27 27 28 1.059 0.9337
28 28 29 0.8042 0.7006
29 29 30 0.5075 0.2585
30 30 31 0.9744 0.963
31 31 32 0.3105 0.3619
32 32 33 0.341 0.5302

www.scitechpub.org
SCITECHP420332 1855
 Science and Tecchnology Publishing (SCI & TE ECH)
ISSN: 2632-11017
ue 2, February - 2024
Vol. 8 Issu 2

Taable 3: The lo
oad data of th
he IEEE 33 bus
b test system
m (Source :[116])

Bus Real Powerr Demand Reactive

R Poweer
Bu
us Number Reeal Power Demand (kW)Re
eactive Powerr Demand (kVaar)
Number (kWW) Demand
D (kVar)
1 0 0 18 900 40
2 100 660 19 900 40
3 90 440 20 900 40
4 120 880 21 900 40
5 60 330 22 900 40
6 60 220 23 900 50
7 200 1000 24 4220 200
8 200 1000 25 4220 200
9 60 220 26 600 25
10 60 220 27 600 25
11 45 330 28 600 20
12 60 335 29 1220 70
13 60 335 30 2000 600
14 120 880 31 1550 70
15 60 110 32 2110 100
16 60 220 33 600 40
17 60 220 17 600 20
Tottal 37115 2300

3. RESULTS
R A
AND DISCUSSIONS Raaphson load flowf analysiss are shown in Table 4. The
ressults in Table 4 shows a totaal real power demand of 3715
3.1
1 The results on the baselin
ne case witho
out DG kW
W and a total reactive poower demand d of 2300 kV Var.
201.891the tota al reactive poower loss of 134.641kW wh hich
The Newton R Raphson loaad flow anaalysis The key mounts to 5.8
am 853956522 % of the tota al real poweer is
parrameters of thhe bus network obtained from
f the New
wton observed.
Table 4 The key paarameters of the
t bus netwoork obtained
d from the Newton Raphsoon load flow analysis
a

commpared acrosss four scenarioos: the base caase (no DGs) and
3.2
2 The resultss on the DG sizing and placement
p ussing
casses with 1, 2, and 3 DG ununits. The PSO O optimizationn at
PSO
30%% penetratio on shows thaat the introd duction of DGs
D
connsiderably red
duced the pow wer losses annd also improoved
The results in Table 5 presennts the outcom
mes of optimizzing on the voltage stability
s in thee power systeem. The scenario
Disstributed Genneration (DG) placement inn a power systtem witth 2 DGs pro ovides the mo
most substantiaal improvemeents,
usiing Particle Swarm Opttimization (P PSO) which is ballancing pow wer loss redduction and voltage proofile
imp plemented at a 30% level of
o penetration
n. The results are

ww
ww.scitechpub
b.org
SC
CITECHP4203332 1856
 Science and Tecchnology Publishing (SCI & TE ECH)
ISSN: 2632-11017
ue 2, February - 2024
Vol. 8 Issu 2

enhhancement effffectively. Speecifically, the scenario witth 2 lossses, with the most pronouunced benefits observed when w
DGGs gave the highest (oveer 65 %) reduction in bboth two o DGs are opttimally placedd.
reaactive and reaactive power with the loweest VDI of 2 .61. Thhe comparison n of the voltagge profiles acrross different bus
The first DG off 567 kW wass located at bus number whhile num mbers for variious scenarioss of DG penettration, optimiized
14,, while the seecond DG of 567 kW waas located at bus usiing PSO is presented in Figure 4. The ressults
num mber 32. demmonstrates th hat increasinng DG peneetration throough
opttimal placement using PSSO significanttly enhances the
The comparisonn of real (activve) and reactive power lossses volltage profile of the systtem. The most m consideraable
acrross different scenarios off DG placem ment in a pow wer im
mprovement is observed witth the integrattion of two DGs, D
sysstem is illustraated in Figuree 3. The scennarios include the whhich providess the best vvoltage regulaation across the
basse case (no DGs) and caases with 1, 2, and 3 D DGs, nettwork. Addin ng a third DG G still impro oves the volttage
opttimized usingg PSO. The chartc effectiveely demonstraates proofile but does not offer as ssignificant an advantage as the
thaat the integratiion of DGs ussing PSO optim mization leadds to seccond DG.
a significant
s reeduction in both
b real andd reactive pow wer
Table 5: PSO-Optimiz
P zed DG Placeement Resultss Implementeed at 30% Peenetration)

PParameter Base C
Case 1 DG 2 DGss 3 DGs
Total reaal power loss (kW) 201.8991 86 70 74
Total reactiive power loss (kVAR) 134.6441 59 46 48
Real poweer loss reductioon (%) ‐ 57.4 65.3 64.3
% Reactive power loss red duction ‐ 56.1 65.8 64.3
Minimmum voltage (pu) 0.91334 0.936067 0.9596444 0.9543035
Minimum voltage bus nu umber 18 18 30 31
Maximmum voltage (p pu) 0.98334 0.997931 0.9979442 0.9979402
Maximum m voltage bus nu umber 2 2 2 2
Voltage devviation index, VDI
V (%) 11.288 4.32 2.61 2.62
DG Loocation (Bus No o.) ‐ 30 14, 32 32, 11, 16
DG Reaal Power Size (kkW) ‐ 1115 567, 5677 372, 372, 372
DG Reactivve Power Size (kVAR)
( ‐ 540 270, 2700 180, 180, 180

Fiigure 3:Real and

a reactive power losses for all scenarios using PSSO.

ww
ww.scitechpub
b.org
SC
CITECHP4203332 1857
 Science and Tecchnology Publishing (SCI & TE ECH)
ISSN: 2632-11017
ue 2, February - 2024
Vol. 8 Issu 2

Figure 4: Voltage proofile compari son with incrreasing DG penetration ussing PSO.
4. CONCLUSIO
C ON applicaations, and policies. En nergy Strate egy
Review ws, 48, 1010 96.
A study
s of poweer reduction in IEEE 33 bu us network baased 4. Alawassa, K. (2022). Distributed Energy
E Resourrces
on distributed generation sizings and placement
p ussing Electriccal Systems: C
Current status and Future
Parrticle Swarm m Optimizatioon (PSO). Newton
N Raphhson prospecctive. Jordan JJournal of En
nergy, 1(1): 377–
meethod was usedd in the load flow analysis to determine the 50.
pow wer losses aand voltage profile and other relevvant 5. Almam mari, M. and A Albadi, M. (20022). Impacts of
parrameters of thhe bus netwoork under the condition off no distributed generration on power systtem
disstributed generration on the bus. The PSO O load placemment protecttion. Renew wable Energ gy and Pow wer
andd sizing showeed remarkablee reduction in the power lossses Qualityy Journal, 20: 413–418.
andd the best perfformance occuurred with two o DG scenarioo. 6. Rudressha, S. J., Ankalliki, S. G.,
REEFERENCES S Ananth hapadmanabhha, T. and Girish, V. (202 23).
1. Sackeyy, D. M., Komlla, D. S., Möller, B., Kagbettor, Effects of integrattion of multi-type DGs on
P. K., O
Owusu-Manu,, D.-G. and Am mevi, A. (20222). voltagee stability annd loss minimization in the
Impactt of distributted power generation
g onn a distribution system m. Internatio onal Journall of
distribuution networrk: A case study
s of miccro‐ Scientiffic Researcch in Eng gineering and a
hydro power plantt in Ghana. 2022 IEEE/ IET Management (IJSREEM), 7(2):1-6
Internaational Utility
y Conference e and Exposittion 7. Mehiga an, L., Deanee, J. P., Galllachóir, B. Ó., &
(IUCE),, 5p. Bertsch, V. (20188). A review of the role e of
2. Saxena, V., Kumar,, N., and Nangia, U. (20222). distribu
uted generattion (DG) in future
f electricity
Recent trends in thhe optimizatio on of renewaable systemms. Energy, 1163, 822-8836.
distribuuted generatioon: A revieww. Ingenieríaa E Abushamah, H. A A. S., Haghifam, M. R., &
InvestiggacióN/Ingeniiería E Invvestigación,422(3): Boland di, T. G. (20021). A nove el approach for
97702. distribu
uted generaation expan nsion plann ning
3. Nadee em, T. B., Siddiqui,
S M., Khalid, M. , & considering its addded value compared with w
Asif, M
M. (2023). Disstributed ene ergy systemss: A centralized generaation expansion. Sustaina able
review of classification, technologiies, Energy y, Grids and Networks, 25, 2 100417.

ww
ww.scitechpub
b.org
SC
CITECHP4203332 1858
 Science and Technology Publishing (SCI & TECH)
ISSN: 2632-1017
Vol. 8 Issue 2, February - 2024

8. Boruah, D., & Chandel, S. S. (2023). A 12. Samala, R., & Kotapuri, M. (2018). Power loss
comprehensive analysis of eight rooftop grid- reduction using distributed
connected solar photovoltaic power plants generation. Modelling, Measurement and
with battery energy storage for enhanced Control A, 91(104), 10-18280.
energy security and grid resiliency. Solar 13. Liu, C., Lv, X., Guo, L., Cai, L., & Su, K.
Energy, 266, 112154. (2017). Control Strategy for Power Loss
9. Chillab, R. K., Jaber, A. S., Ben Smida, M., and Reduction considering Load Variation with
Sakly, A. (2023). Optimal DG location and sizing Large Penetration of Distributed
to minimize losses and improve voltage profile Generation. Advances in Electrical
using Garra Rufa optimization. Sustainability, Engineering, 2017(1), 5764054.
15(2): 1156-1169. 14. Mander, D. K., & Virdi, G. S. (2017). Result
10. Ali, A., Huma, Z., Zahoor, N., Yuanqing, X., Ullah, analysis on load flow by using Newton
I., and Hussain, A. (2022). Research on Raphson method. International Journal of
conventional power distribution systems with Advanced Research in Electrical, Electronics
integration of modern smart distributed and Instrumentation Engineering, 6(7), 5835-
generation for power quality in renewable 5844.
generation. 2022 International Conference on 15. Santoso, D. B., Sarjiya, S. and Sakti, F. P. (2018).
Electrical Engineering and Sustainable Optimal sizing and placement of Wind-Based
Technologies (ICEEST), Lahore, Pakistan, 7p. distributed generation to minimize losses using
11. Bhargava V. and Sinha, S. K. (2022). DG flower pollination algorithm. JTERA (Jurnal
placement in distribution system for voltage Teknologi Rekayasa), 3(2):167-176.
profile enhancement and loss reduction by
reactive power loss minimization. 2022
International Conference on Computing,
Communication, and Intelligent Systems (ICCCI
S), 6p.

www.scitechpub.org
SCITECHP420332 1859