OPEN AFRICAN POWER 2022

ENEL FOUNDATION

Improving the Efficiency and Power Quality of the Grid and Off-grid Power Supply
System for Rural Customers using Optimal Allocation of Distributed Generation

Capstone Project

Demsew Mitiku Teferra

Addis Ababa Science and Technology University

Addis Ababa

Ethiopia

July 2022
Abstract

One of the UN millennium development goal is the target set for the energy sector, SDG7. Energy
is becoming more sustainable, affordable, widely available, reliable, and efficient to be accessed
by energy customers. Access to electricity in poorer and rural customers should be accelerated,
energy efficiency and power quality need to be improved, and special attention is also given to the
renewable energy mix in electricity generation to attain the sustainable development goal of the
energy sector. Electricity access can be achieved by either increasing the generation capacity or
improving the efficiency of the power supply in the generation, transmission, distribution, and
end-user devices. System efficiency and power quality are some of the criteria for choosing
electricity supply competitors. Long-distance power transmission and distribution networks in the
grid-connected system and elongating the distribution line of the off-grid power supply system are
the cause of low power quality and inefficiency of overall system operation. Optimal sizing and
allocation of the distributed generation (DG) system in a grid-connected power supply system or
an island microgrid network is supposed to be a good solution for power quality and efficiency
improvement of the existing electricity supply system in accordance with the SDG7. A new
optimization algorithm based on the hybrid of particle swarm optimization and harmony search
algorithm will be applied for the implementation of DG in a grid-connected and island mode
microgrid system for maximizing the efficiency and power quality of the overall electricity supply
system.

1. Introduction

The difficulties in bringing electricity to rural areas are formidable: Low population densities in
the rural communities result in high capital and operating costs of electricity. Consumers are often
poor, and their electricity consumption is too low which results in very low attraction of electricity
producers [1]. On the other hand, people are using electricity than ever before, with the proportion
of the global population having access to this service rising from 83 percent in 2010 to 87 percent
in 2015, then to 89 percent in 2017. Still, 840 million people were without this essential service
in 2017, mostly in sub-Saharan Africa. In that region, only 44 percent of the population had access,
and an estimated 573 million people still lacked electricity [2]-[4]. Therefore, the efficiency of the
electric power supply system, the quality of service, and the quality of power supply should be
improved to make the power supply system more affordable and sustainable.

2
Long-distance power transmission and distribution networks in the grid-connected system and
elongating the distribution line of the off-grid power supply system are the cause of low power
quality and inefficiency of overall system operation [5], [6]. Optimal sizing and allocation of the
distributed generation system in a grid-connected power supply system or an island microgrid
network is supposed to be a good solution for power quality and efficiency improvement of the
existing electricity supply system in accordance with the SDG7 [7]. Meta-heuristic optimization
algorithms like genetic algorithm, artificial bee colony algorithm, particle swarm optimization
algorithm, ant colony algorithm, simulated annealing, differential evolution, firefly algorithm,
harmony search algorithm, cuckoo search algorithm, etc. are used to implement allocation and
sizing of distributed generation for power quality, reliability and efficiency improvement of an
existing power supply system [8]-[11].

This paper will investigate how distributed generation (DG), particularly when it contained solar
PV, wind, and battery energy storage systems, can become a good power quality and efficiency
improvement solution in a grid-connected distribution and an islanded microgrid system designed
for rural electrification using particle swarm optimization and harmony search algorithms. The rest
of the paper is organized as follows. Sections 2 and 3 present the research objective and scope of
the research. Section 4 is about research methods. In Sections 5 and 6 description of the research
and expected deliverables are discussed; Section 7 is about the research implementation plan.
References and acronyms are presented at the end of the paper.

2. Research Objective

The objective of this research is to improve the efficiency and power quality of the grid and off-
grid power supply system for rural customers using the optimal allocation of distributed
generation.

3. Scope of the Research

The scope of this research comprises the model of the grid-connected and islanded microgrid
system in Simulink, defining the objective function and constraints of the optimization problem
using Matlab, and validation of the newly developed model using existing mathematical and
software models. The newly developed model is tested in the existing grid-connected and islanded

3
microgrid power supply system for future system development and its adaptation in an existing
power supply system. The research is planned to be completed in the seven-month time frame.

4. Research Method

In this research, an optimal allocation of distributed generation (DG) in the grid-connected radial
distribution network and an islanded radial microgrid system will be designed to maximize the
power quality and efficiency of the overall power system performance. The performance of the
newly developed optimal allocation methodology based on the hybrid algorithm of particle swarm
and harmony search optimization methods will be tested in IEEE_33, IEEE_15 radial distribution
systems, and an islanded radial microgrid system which is designed for a rural community power
supply option.

The hybrid of PSO and HSA optimization algorithms is chosen for optimal allocation and sizing
of DG in a radial distribution network and an islanded microgrid network to improve overall
system performance. The hybrid algorithm improves the weakness of the PSO and HSA
optimization algorithms alone. This hybrid algorithm enhances the cognitive and social searching
performance of the objective function to a certain level than the individual algorithms.

Constraints
_ The max. and min. capacity limit of the DG
_ The voltage limit constraints
_ The thermal limit of the lines

Input Output
_ Initialize the Optimization Method _ Optimal DG size
algorithm Hybrid [PSO-HSA] _ Optimal DG
_ Input variables location

Single Objective Function

-Goal Programming
-Scalarization

Multi-Objective Function
- Maximizing System Efficiency
- Maximizing Power Quality (voltage)

Fig. 1: Development Process of a hybrid PSO_HSA algorithm for sizing and siting of DG.

4
Fig. 1 and Fig. 2 demonstrate how the hybrid PSO_HSA algorithm is implemented for the optimal
allocation of DG in a radial distribution network. The proposed optimization method combines the
PSO and HSA algorithms to optimize the size and site of the distributed generation system which
is integrated into a radial distribution network. The harmony search rate (HSR), harmony memory
considering rate (HMCR), and pitch adjustment rates (PAR) are the parameters of HSA.

Start

Formulate the (MOOP) Optimization Problem

Initialize the particle with random position and velocity and

the harmony memory with HSR, HMCR and PAR.

Evaluate the fitness of Particles, fp Evaluate the fitness of HSA, fh

Update PSO Parameters Update HSA Parameters

No Yes
based on the improvised fp < fh? based on the best value of the
value of HM. particles

No No
Termination criteria?

Yes

End

Fig. 2: The proposed hybrid PSO-HSA method for optimal allocation of DG.

5
5. Description of the Research
The system cost, reliability, and overall efficiency are the most important parameters to be
considered in the sizing and siting of the distributed generation (DG) in the radial distribution
network. An optimally sized and sited DG in the radial distribution system effectively mitigates
the power network efficiency and power quality challenges. The IEEE 15, IEEE 33 bus radial
distribution systems, and the islanded microgrid system designed for rural communities are used
for testing the new model's performance.

5.1. Problem Formulation

The Power flows in a distribution system and microgrid system are computed by the following set
of simplified recursive equations derived from the single-line diagram shown in Fig. 3. The power
flow analysis can be used to obtain the power quality and efficiency of the IEEE 15 and IEEE 33
bus radial distribution systems. The objective function is to find the total power loss in the network
and the power quality index of the network.

Vo Vk Vk+1 Vn
Pk+jQk Pk+1+jQk+1
Rk+jXk PL,k+1+jQL,k+1
Yk1 Yk2

Fig. 3. Single line diagram for load flow problem formulation.

𝑃𝑘+1 = 𝑃𝑘 − 𝑃𝑙𝑜𝑠𝑠,𝑘 − 𝑃𝐿,𝑘+1 (1)

𝑄𝑘+1 = 𝑄𝑘 − 𝑄𝑙𝑜𝑠𝑠,𝑘 − 𝑄𝐿,𝑘+1 (2)

where Pk and Qk are the real and reactive power flowing out of bus k, 𝑃𝑙𝑜𝑠𝑠,𝑘 and 𝑄𝑙𝑜𝑠𝑠,𝑘 are the
real and reactive line losses between bus k and k+1, 𝑃𝐿,𝑘+1 and 𝑄𝐿,𝑘+1 are the real and reactive
loads at bus k+1.

The real and reactive line loss between bus k and k+1 is computed using the following equations.

𝑃𝑘2 +𝑄𝑘2
𝑃𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1) = 𝑅𝑘 𝑉𝑘2
(3)

𝑃𝑘2 +𝑄𝑘2
𝑄𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1) = 𝑋𝑘 𝑉𝑘2
(4)

6
 5.2. The Objective Functions

The problem is a dual objective function of maximizing efficiency (minimizing the total power
loss) and minimizing the average voltage deviation index (AVDI).

The total active (𝑃𝑇,𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1)) and reactive (𝑄𝑇,𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1)) power loss of the whole
feeder in the distribution network is computed using:

𝑃𝑇,𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1) = ∑𝑛𝑘=1 𝑃𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1) (5)

𝑄𝑇,𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1) = ∑𝑛𝑘=1 𝑄𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1) (6)

The efficiency, 𝑓1 is the first objective function to be maximized which is given in Eq. 7.

𝑓1 = (𝑃𝑇,𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1)𝑤𝑖𝑡ℎ𝐷𝐺 − 𝑃𝑇,𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1)𝑏𝑎𝑠𝑒𝑐𝑎𝑠𝑒 )/𝑃𝑇,𝑙𝑜𝑠𝑠 (𝑘, 𝑘 + 1)𝑏𝑎𝑠𝑒𝑐𝑎𝑠𝑒 (7)

The average voltage deviation index (AVDI) is the measure of the power quality of the power
supply system and it is given in Eq. 8. AVDI should be minimized.
1
𝑓2 = AVDI = 𝑏𝑟 ∑𝑏𝑟
𝑘=1|1 − 𝑉𝑘 |
2
(8)

Therefore, the objective is to maximize 𝑓1 and to minimize 𝑓2, the multi-objective optimization problem
is formulated in Eq. 9.
𝑤
𝑓(𝑘) = minimize⁡( 𝑓1 + 𝑤2 𝑓2 ) (9)
1

Where, br is the number of branches in the network, w1 and w2 are the weight factor.

5.3. The Set of Constraints

The thermal limit of the lines, the capacity limit of the DG, and the bus voltage minimum and
maximum limit values, are given in Eq. 10-Eq. 12 respectively, are the constraints considered to
evaluate the objective function given in Eq. 9.
𝑚𝑎𝑥
𝑆(𝑘, 𝑘 + 1) ≤ 𝑆𝑙𝑖𝑛𝑒 (10)

𝑆𝑚𝑖𝑛 ≤ 𝑆𝐷𝐺 ≤ 𝑆𝑚𝑎𝑥 (11)

𝑉𝑚𝑖𝑛 ≤ 𝑉𝑘 ≤ 𝑉𝑚𝑎𝑥 (12)

The forward-backward sweep load flow algorithm will be implemented for load flow analysis.

7
6. Expected Deliverables
In this research, the distributed generator (DG) is placed in an existed radial distribution network
or an islanded microgrid system in order to maximize the efficiency of the network and improving
the power quality of the power supplied to the customer. The size (capacity) of the DG and where
it should be placed are the determinant variables of the DG for the overall system performance
improvement. The optimal values of these decision variables are determined using an optimization
algorism based on the hybrid of PSO and HSA. The newly developed algorithm will be applied
and implemented in the standard IEEE radial distribution system and an islanded microgrid
network which will be designed for electrification of rural community. An improved overall
system efficiency and supply power quality results are expected from the incorporation of the new
modeling approach and also validation of the model will be done for verification of the new
methodology in a bigger system.

8
7. Research Implementation Plan
The research implementation plan and activity time frame required to complete the research
work are presented in Table 1.

Table 1. Research activity and timetable.

20 August 2022
Task Duration

20 June 2022

20 Sept 2022
20 July 2022

20 Nov 2022
20 Oct 2022
Description

20-Dec-22
Task ID

(Day)
Task

1. Literature Review 210

2. Proposal development 30

3. ▪ Mathematical model of load 30

flow analysis, the forward-

backward algorithm
4. ▪ Simulink Model of Microgrid, 30

IEEE-15, and IEEE-33 radial

distribution system
5. ▪ Preparation of the MATLAB 30

code for PSO, HAS, and hybrid

PSO-HAS algorithm
6. ▪ Integrating the forward- 30

backward load flow algorithm

with the optimization algorithm
7. ▪ Simulation of the model and 30

interpreting the result of the

model, discussing its output.
8. ▪ Preparing the report, writeup 30

and presenting the findings of

the report, and also the
publication of findings.

9
8. References
[1]. Douglas F. Barnes. “The Challenge of Rural Electrification Strategies for Developing
Countries.” Published July 11, 2007, Routledge, 368 Pages, ISBN 9781933115436.
[2]. Fabio Riva, Helene Ahlborg, Elias Hartvigsson, Shonali Pachauri, Emanuela Colombo.
“Electricity access and rural development: Review of complex socio-economic dynamics
and causal diagrams for more appropriate energy modeling, Energy for Sustainable
Development”, Volume 43, 2018, Pages 203-223, ISSN 0973-0826,
https://doi.org/10.1016/j.esd.2018.02.003.
[3]. Jimenez Mori, Raul Alberto. “Rural Electricity Access Penalty in Latin America: Income
and Location”, Jun 2016.
[4]. United nation department of Economics and social affairs, statistics division.
https://unstats.un.org/sdgs/report/2019/goal07/#:~:text=Globally%2C%20electrification%
20of%20rural%20areas,electricity%20live%20in%20rural%20areas.
[5]. Vikash Anand and S. K. Srivastava. “Causes, Effects, and Solutions of Poor-Quality
Problems in the Power Systems”, Journal of Engineering Research and Applications
www.ijera.com ISSN: 2248-9622, Vol. 4, Issue 5(Version 7), May 2014, pp.67-74.
[6]. A. Moreno-Munoz, J.J.G. de-la-Rosa, M.A. Lopez-Rodriguez, J.M. Flores-Arias, F.J.
Bellido-Outerino, M. Ruiz-de-Adana (2010). “Improvement of power quality using
distributed generation”. 32(10), 1069–1076. doi: 10.1016/j.ijepes.2010.06.002.
[7]. M Salmani, J Pasupuleti, V K Ramachandaramurthy. “Optimal Placement and Sizing of
Multiple DG in Microgrid Systems”. International Journal of Recent Technology and
Engineering (IJRTE) ISSN: 2277-3878, Volume-8, Issue-4, November 2019.
[8]. Xin-she Yang, “Metaheuristic Optimization”. (2011), Scholarpedia, 6(8):11472.
doi:10.4249/Scholarpedia.11472.
[9]. H. E. A. Talaat and E. Al-Ammar, "Optimal allocation and sizing of Distributed Generation
in distribution networks using Genetic Algorithms," 11th International Conference on
Electrical Power Quality and Utilisation, 2011, pp. 1-6, doi: 10.1109/EPQU.2011.6128840.
[10]. I. J. Hasan, M. R. A. Ghani and C. K. Gan, "Optimum distributed generation allocation
using PSO in order to reduce losses and voltage improvement," 3rd IET International
Conference on Clean Energy and Technology (CEAT) 2014, 2014, pp. 1-6, doi:
10.1049/cp.2014.1476.
[11]. M. T. A. Y. Mohammadi and M. Faramarzi, "PSO algorithm for sitting and sizing of
distributed generation to improve voltage profile and decreasing power losses," 2012
Proceedings of 17th Conference on Electrical Power Distribution, 2012, pp. 1-5.

10
9. Acronyms

AVDI
Average Voltage Deviation Index, 7

DG
Distributed Generation, 2, 3, 4, 5, 6, 7, 9

HMCR
Harmony Memory Consideration Rate, 5
HSA
Harmony Search Algorithm, 4, 5, 8
HSR
Harmony Searching Rate, 5

IEEE
Institute of Electrical and Electronics Engineering, 4, 6, 8

MOOP
Multi-objective optimization problem, 5

PSO
Particle Swarm Optimization Algorithm, 4, 5, 8, 9
PV
Photovoltaice, 3

SDG
Sustainable Development Goal, 2

UN
United Nation, 2

11