OPTIMAL ALLOCATION OF HYBRID RENEWABLE ENERGY SOURCES FOR

AJAYI CROWTHER UNIVERSITY POWER SYSTEM FLEXIBILITY

BY

OLADAPO, OMOLADE GIDEON

(19EG02/015)

A PROJECT REPORT SUBMITTED TO

THE DEPARTMENT OF ELECTRICAL AND ELECTRONICS ENGINEERING,

FACULTY OF ENGINEERING,

AJAYI CROWTHER UNIVERSITY, OYO,

OYO STATE.

IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AWARD OF

BACHELOR OF ENGINEERING DEGREE (B.ENG) IN ELECTRICAL AND

ELECTRONICS ENGINEERING.

JUNE, 2024
 CERTIFICATION

This is to certify that this research work titled “Optimal Allocation of Hybrid Renewable

Energy Resources for Ajayi Crowther University Power System Flexibility”, was carried

out by OLADAPO, Omolade Gideon with Matriculation Number 19EG02/015 of the

Department of Electrical and Electronics Engineering, Ajayi Crowther University, Oyo, Oyo

State, Nigeria for the award of Bachelor of Engineering in Electrical and Electronics

Engineering under my supervision.

Dr. O.A. Ajeigbe DATE

Supervisor

Dr. O.A. Ajeigbe DATE

Head of Department

Prof. Z.K Adeyemo DATE

Dean, Faculty of Engineering
 DEDICATION

This project report is dedicated to God Almighty and in memory of my late father,

Mr. Oladapo Adesanya Israel.

ii
 ACKNOWLEDEGMENT

I appreciate God Almighty, for the grace he gave me and the strength to pull through during

the course of this research.

I appreciate my mother and my brothers, Prof. (Mrs) Adejoke Rasidat Oladapo, Oladapo

Omololu and Oladapo Omolayo for their immense support, love and provision throughout my

undergraduate studies. Only God can reward you abundantly.

I thank my supervisor wholeheartedly, Dr. O.A Ajeigbe for training and guidance he gave me

on this research topic as he did not only supervise me but also taught me the fundamentals that

I will need in the labour market.

I am forever grateful to the lecturers from Computer Engineering, Civil Engineering,

Electrical and Electronics Engineering and Mechanical Engineering for the knowledge that

have been imparted to me thus far.

I acknowledge the Dean, Professor Z.K Adeyemo for giving us, the student a vast knowledge

of engineering that can never be taken for granted.

I mostly thank all my friends and roommates: Buoye Eniola, Oladimeji Samuel, Enahoro

Andrew, Jiya Joshua, Aworinde Oreoluwa, and my wonderful mates for their love, support

and care. I also thank my level mates, though I cannot mention your names. I am glad to have

met you all.

iii
 TABLE OF CONTENTS

CERTIFICATION.....................................................................................................................i

DEDICATION...........................................................................................................................ii

ACKNOWLEDEGMENT.......................................................................................................iii

LIST OF FIGURES..................................................................................................................vi

ABSTRACT.............................................................................................................................vii

CHAPTER ONE........................................................................................................................1

INTRODUCTION.....................................................................................................................1

1.1 Background to the study............................................................................................1

1.2 Statement of research problem..................................................................................3

1.3 Sub-problems...............................................................................................................4

1.3.1 Sub-problem 1: Evaluation of existing power system infrastructure.............4

1.3.2 Sub-problem 2: Development of mathematical optimisation planning model.

1.3.3 Sub-problem 3: Determination of optimal number, timing and locations of

HRESDG units in the network..........................................................................................5

1.3.4 Sub-problem 4: Generation and planning for flexibility in Ajayi Crowther

University power system....................................................................................................6

1.4 Hypothesis....................................................................................................................6

1.5 Aims and objectives....................................................................................................7

1.6 Contribution of the study...........................................................................................8

1.7 Delimitations................................................................................................................8

iv
 1.8 Outline of the chapters...............................................................................................9

CHAPTER TWO.....................................................................................................................11

LITERATURE REVIEW.......................................................................................................11

2.1 Introduction...............................................................................................................11

2.2 Hybrid renewable sources distributed generation (HRESDG)............................13

2.2.2 Battery energy storage system..........................................................................16

2.3 Some recent application of HRESDGs and their key result.................................18

2.4 Formulation of HRESDG allocation planning problem........................................22

2.4.1 Objectives of HRESDG allocation problem....................................................23

2.5 Decision variables and constraints of HRESDG allocation..................................26

2.6 Modeling of uncertainties in HRESDGs allocation...............................................28

2.6.1 Methods for modeling/accounting the uncertainties of renewable energy

resources............................................................................................................................29

2.7 Load profile model....................................................................................................31

2.8 Optimisation algorithms for HRESDGs allocation problem................................32

2.8.1 Conventional Method........................................................................................33

2.8.2 Intelligent search (IS) methods.........................................................................40

2.8.3 Potential intelligent search methods................................................................43

2.8.3 Probable hybrid optimisation methods...........................................................45

2.9 Summary....................................................................................................................45

CHAPTER THREE.................................................................................................................47

v
OPTIMAL ALLOCATION OF HYBRID RENEWABLE ENERGY SOURCES

DISTRIBUTED GENERATION IN AJAYI CROWTHER POWER SYSTEM’S

FLEXIBILITY.........................................................................................................................47

3.1 Site description for HRESDGs allocation...............................................................48

3.2 Modeling of renewable energy resources of site under study...............................49

3.2.1 Modeling of solar irradiance............................................................................49

3.3 Modeling of solar PV and battery energy storage system power (MW)..............50

3.3.1 Modeling of solar power output.......................................................................50

3.3.2 Modeling of battery energy storage system power output.............................51

3.4 Mathematical formulation of the planning problem.............................................52

3.4.1 The objective function.............................................................................................52

3.4.2 HRESDG allocation constraints............................................................................55

3.5 Modeling of load demands............................................................................................59

3.6 Capacity Utilization Factor......................................................................................59

3.7 Summary....................................................................................................................59

CHAPTER FOUR...................................................................................................................61

RESULT AND DISCUSSION................................................................................................61

4.1 The ACU-10 bus location.........................................................................................62

4.1.1 Results of the optimal REHDGS’ allocation Problem.......................................62

4.2 Summary.........................................................................................................................66

CHAPTER FIVE.....................................................................................................................68

CONCLUSION........................................................................................................................68

vi
 5.1 Summary of work.....................................................................................................68

5.2 Significant contributions..........................................................................................69

5.3 Future scope..............................................................................................................71

REFERENCES........................................................................................................................73

vii
 LIST OF FIGURES

Figure 2.1: Main sources of RES systems.................................................................................13

Figure 2.2: Working principle of Solar PV...............................................................................15

Figure 2.3: Working principle of Battery energy storage system (deye.com)..........................17

Figure 4.1: Aerial view of the Bus location...............................................................................62

Figure 4.2: Optimal locations and sizes of PV power installed throughout the planning

horizon……………………………………………………………………………………64

Figure 4.3: Optimal capacities of BESS installed throughout the planning horizon:………..65

viii
 ABSTRACT

The need for energy and environmental sustainability drives the adoption of renewable energy

systems to reduce greenhouse gas emissions. Efficient planning of energy sources is crucial

for a reliable distribution network. This study presents the optimal allocation of renewable

energy resources into the power system of Ajayi Crowther University (ACU) to enhance

power supply flexibility and sustainability. The integration of hybrid renewable energy

sources distributed generation (HRESDG) such as solar, and battery bank storage can

significantly reduce dependency on fossil fuels power generation and their environmental
ACU.
impacts in However, the intermittent nature of renewable energy sources poses

challenges to the stability and reliability of the power system. By considering power system

flexibility, these challenges have been carefully examined in this study and transformed into

opportunities, facilitating a seamless transition to a more sustainable energy mix.

In this study, HRESDG optimisation algorithm is employed to plan and design the strategic

allocation (size, location and timing) of hybrid renewable sources distributed generation

(HRESDG) with the ultimate goal of maximising the renewable power generated and

absorbed into the ACU campus network at a minimum net present value of total cost. The

analysis of uncertainty and modeling of the hybrid renewable system parameters is crucial as

it helps in managing risks, designing reliable and flexible systems. This study aims to provide

a review on uncertainty modeling approaches for distribution system studies and application.

The scope of this study addresses the optimal allocation of non-dispatchable solar, and battery

bank storage on a smart network considering the flexibility of power system. The allocation of

hybrid renewable energy resources into ACU power system provide flexibility in

accommodating the intermittent nature of solar energy. Several tools have been developed

with focus ranging from resource identification to parameter characterization and model

formulations. Historical meteorological data for solar irradiance have been analyzed to

ix
provide suitable hybrid power models. Optimization models for solar, and battery bank

storage reserves for power system simulation studies have been developed. Long term power

system flexibility has been evaluated through ACU system studies and analysis.

The results of the study demonstrate the feasibility and benefits of optimal RES allocation into

the ACU power system. The proposed framework provides valuable insights for policymakers

and energy planners to develop strategies for a more sustainable and flexible power system in

the university campuses.

x
 CHAPTER ONE

INTRODUCTION

This chapter introduces the study's context and rationale, and outlines the issues under

investigation. It then presents the research hypothesis, aim and objectives, along with a list of

contributions stemming from the study. Finally, the chapter concludes with an overview of the

research's subsequent chapters.

1.1 Background to the study

The strategic allocation of hybrid renewable energy sources like solar PV and battery energy

storage within Ajayi Crowther University's power system represents a crucial research area.

This focus is driven by advancing technologies, economic viability, environmental

advantages, resource synergies, resource availability, and the growing capacity of these

sources to meet the Ajayi Crowther university's high energy demands. Building upon prior

studies on developing power models for renewable energy integration and the concept of

flexibility in ACU's power system, this research tackles the significant challenges posed by the

determination of sizes and locations of these hybrid renewables in the Campus network. These

challenges, compounded by the predominantly passive nature of current distribution networks,

hinder renewable energy integration. They particularly impact system operation and the

integration capacity of renewable power into the power system.

Leveraging distributed generations (DG) emerges a potent strategy for tackling the technical,

economic, and ecological challenges in the power system. Moreover, it caters to the rising

energy needs effectively (Santos et al., 2017). DGs, as generating units, are strategically

positioned within distribution systems, close to load centers, to meet immediate power needs.

Their deployment aims to decrease on-peak operating costs, delay network upgrades,

minimize losses, enhance power quality and reliability, lower transmission and distribution

1
(T&D) expenses, alleviate T&D loading, diversify energy sources, and fortify power system

and enhance power stability (Zubo et al., 2017; Ajeigbe et al., 2019 ; Georgilakis &

Hatziargyriou, 2015, ; Abdmouleh et al., 2017). Moreover, DGs, in contrast to centralized

generations, are modular units that occupy small landmass or area using smaller generators,

lower capital costs and shorter construction times. The conventional DG can be grid-

connected or standalone (Hosenuzzaman et al., 2015; Ajeigbe et al., 2018). DG systems

exhibit a broad spectrum of sizes and power levels tailored to specific applications and

demands. They range from single units as small as 1 kW to multi-unit installations capable of

generating up to 300 MW. These systems are categorized based on their power output,

comprising micro-scale DGs (1 kW - 5 kW), small-scale DGs (≥5 kW - 5 MW), medium-scale

DGs (≥5 MW - 50 MW), and large-scale DGs (≥50 MW - 300 MW). (Ackermann et al., 2001;

Zubo et al., 2017). DGs are segmented into four distinct types based on their active and

reactive power behaviours. P-type DGs, classified as Type 1, solely generate active power

without interacting with the distribution network for reactive power. Q-type DGs, or Type 2,

either generate or absorb reactive power from the distribution network but do not contribute

active power. Type 3 DGs, or P Q+ DGs, contribute both real and reactive power to the

distribution network. Lastly, Type 4 DGs, or P Q− DGs, generate real power while absorbing

reactive power from the system (Zubo et al., 2017; Jordehi, 2016). DG may also be renewable

energy based or non-renewable energy sources or a hybrid of both.

The hybrid nature of renewable energy-based DGs presents an attractive option, given their

renewable and inexhaustible characteristics, synergy with other energy sources, environmental

sustainability, technological readiness, and economic feasibility (Ajeigbe et al., 2018). Yet,

the power output of HRESDGs like photovoltaic (PV) (Santos et al., 2017; Jordehi, 2016).

The variability arises due to factors such as weather conditions (solar). A system’s ability to

dynamically adjust its operating parameters in response to changing condition is its flexibility.

2
Power system flexibility enables the power system to quickly respond to changes in generation

output of HRESDGs, helping to balance the supply-demand equation and reduce the need for

conventional backup power plants (katiraci and Iravani, 2018). It allows for efficient

utilization of renewable energy resources that are inherently intermittent and dependent on

variability factors.

In summary, the optimal locations, capacities, and operational efficiencies of HRESDGs

hinges on network topology and variables. However, integrating these elements presents

numerous challenges to Ajayi Crowther power system operations and planning. In this study,

there is a need to revamp the Renewable Energy Hybrid Distributed Generations-Allocation

Expansion Planning (HRESDGs-AEP) network model to effectively capture the variability

and intermittency of HRESDGs in the network. This improved model leads to more accurate

assessment of the system’s flexibility requirements, thereby allowing proper designing and

implementation of strategies to enhance power system of ACU. The significant contribution

lies in the optimization of HRESDG allocation, and power system flexibility variables to

attain sufficient energy/power utilisation in the campus.

1.2 Statement of research problem

The inadequate power supply from the Ibadan Electricity Distribution Company (IBEDC) has

hindered certain day-to-day activities at Ajayi Crowther University Campus, including water

pumping, cyber learning, and the creation of an efficient learning environment. The research

problem addressed in this study is the need to improve the flexibility and reliability of the

power system at Ajayi Crowther University through the optimal allocation of HRESDG into

the distribution system.

Existing research on DG allocation optimization has largely overlooked the integration of

renewable energy resources and the intermittent nature of solar PV/battery energy storage

3
hybrid resources. These studies often fail to optimize the absorption of power from

intermittent renewable energy hybrid DGs (HRESDGs) into the distribution network while

considering the flexibility of the network at minimal cost. While most models prioritize cost

minimization, they neglect how well a network can quickly adjust its operating conditions to

retain its maximum power performance which is very important in addressing the challenges

caused by the intermittency of HRESDG resources such as solar PV and battery energy

storage.

This research aims to investigate the allocation of hybrid renewable energy sources distributed

generations (HRESDG) into ACU power system, with a focus on enhancing the system’s

flexibility.

1.3 Sub-problems

These are the following sub-problems that are identified by the study;

1.3.1 Sub-problem 1: Evaluation of existing power system infrastructure

Prior to the optimal allocation of predetermined renewable energy sources at Ajayi Crowther

University, it is crucial to assess the current power system infrastructure. This evaluation

involves identifying the infrastructure's capacity and configuration, evaluating its age and

condition, and assessing the overall reliability and efficiency of the system. Therefore, the

following questions are put into consideration:

1. What is the current capacity and configuration of the power system infrastructure and

key components of the power system infrastructure, such as power plants, substations,

transformers, and distribution lines?

2. How well does the current infrastructure meet the energy demand of the system, both

in terms of capacity and reliability?

4
 3. What are the environmental impacts of the existing power system infrastructure,

including emissions and land use?

1.3.2 Sub-problem 2: Development of mathematical optimisation planning model.

Modeling HRESDGs requires dealing with a complex mathematical model that can be quite

challenging to formulate accurately (Jordehi, 2016; Ajeigbe et al., 2019; Abdmouleh et al.,

2017; Georgilakis & Hatziargyriou, 2013). To optimize the allocation of intermittent hybrid

DG, it's crucial to accurately model the uncertainty associated with HRESDG units,

particularly the intermittent nature of solar resources. This involves quantifying their

potentials and predicting their stochastic operation. Quantifying the potentials of solar

resources is challenging due to their site-specific characteristics, making a universal model for

different locations impractical. Furthermore, aligning the power outputs of these variable

resources with varying weather conditions to meet the load demand adds complexity. Despite

these obstacles, some studies have effectively integrated these uncertainties into their models.

The following questions are pertinent:

1. How can accurate optimization and uncertainty models be developed to simulate the

integration of various renewable energy sources?

2. What parameters and data are essential for creating reliable uncertainty models for the

campus?

3. What are the potential cost savings and environmental benefits that can be achieved

through the implementation of these optimization models?

1.3.3 Sub-problem 3: Determination of optimal number, timing and locations of

HRESDG units in the network

Optimizing the planning and design of HRESDGs in a distribution system involves assigning

renewable active and reactive power outputs while minimizing total cost over the planning

5
horizon. However, this task is complex due to its non-linear, multi-objective nature with many

constraints. One major challenge is identifying the most optimal locations for these generators

(Jordehi, 2016; Zubo et al., 2017; Ajeigbe et al., 2018; Theo et al., 2017). The variable power

output of HRESDGs resources can significantly be impacted if their allocation is sub-optimal.

Numerous researchers have explored various approaches to determine the optimal locations

for HRESDGs to mitigate this issue (Teng et al., 2002; Keane & O’Malley, 2005; Porkar et

al., 2011; Munoz-Delgado et al., 2015; Sfikas et al., 2015). The optimal sizing and placement

of HRESDGs can reduce costs.

1.3.4 Sub-problem 4: Generation and planning for flexibility in Ajayi Crowther

University power system

The Ajayi Crowther University (ACU) power system, like many modern energy systems,

faces challenges related to generation and planning for power system flexibility. With the

integration of renewable energy sources (RES) and the need for a reliable and sustainable

electricity supply, it is crucial to enhance the system's flexibility to accommodate the

inadequacies of the utility source. The ACU power system currently relies on conventional

sources such as grid power and diesel generators. However these sources are irregular,

inadequate and costly leading to challenges in matching supply with demand. To generate

flexibility in the integration of HRESDGs into ACU power system, the following question are

considered:

1. What are the key considerations in evaluating the flexibility requirement of ACU

power system?

2. How can energy storage systems and demand side management be used to enhance

ACU power system’s flexibility?

3. What are the challenges in monitoring and adjusting flexibility enhancement strategies

in ACU power system?

6
1.4 Hypothesis

Optimal allocation of renewable energy sources into Ajayi Crowther University power system

postulates increased system flexibility, reduced reliance on non-renewable energy sources, and

improved sustainability. This can be achieved by strategically placing renewable energy

sources such as solar PV systems, and battery storage systems in locations with high

renewable energy potential, coupled with efficient energy storage systems and smart grid

technologies. Power system flexibility is proposed as the primary method to annihilate

intermittency and variability impact of HRESDGs in Ajayi Crowther University distribution

system and achieve the optimal allocation of Renewable Energy Hybrid Distributed

Generators (HRESDGs) at minimum cost. An approach involves the increase in the share of

renewable energy sources and investing in energy storage systems to store excess energy for

use during periods of low generation. This can help balance supply and demand and reduce

reliance on fossil fuels. This method aims to improve the operational characteristics of Ajayi

Crowther University proposed network systems.

1.5 Aims and objectives

The study aims to develop and design models integrating solar PV and BESS into the ACU

distribution system to enhance power system flexibility. The specific objectives are stated

below:

1. To develop a new multi-stage mathematical optimization planning models that

considers the intermittency and variability of these renewable energy sources.

2. To concurrently ascertain the optimal number, timing, and locations of HRESDG units

within Ajayi Crowther University distribution system. The overarching objective is to

maximize the integration of renewable DG power into the system at the lowest cost.

3. To incorporate flexibility considerations, such as demand-side management and energy

storage, into the optimization model.

7
1.6 Contribution of the study

The study on optimal allocation of renewable energy sources in Ajayi Crowther University

Power System Flexibility has the following contributions:

1. It will contribute to the body of knowledge on renewable energy integration and power

system flexibility, particularly in the context of educational institutions like Ajayi

Crowther University.

2. It will explore the use of various modeling and simulation techniques for assessing the

performance of renewable energy resources in the power system and for evaluating the

impact of their integration on power system flexibility.

3. Implementing optimal allocation strategies will lead to cost savings in the long run as

renewable energy sources generally have lower operating costs compared to non-

renewable sources and also contribute to reducing carbon emissions.

4. The findings of the research can inform policy decisions related to renewable energy

integration and sustainability in educational institutions and beyond.

Overall, the research on optimal allocation of renewable energy resources has the potential

to make significant contributions towards a more sustainable, resilient, and efficient

energy system for Ajayi Crowther University and beyond.

1.7 Delimitations

The following are excluded from the study:

1. The research only focused on the optimal allocation of solar photovoltaic and

battery bank storage and does not consider other renewable resources.

8
 2. The study only focused specifically on the power system of Ajayi Crowther

University, limiting the generalizability of the findings to other institutions or

regions but may be scalable to other institutions and regions.

3. The research was conducted within a specific timeframe and may not capture long-

term trends or developments in renewable energy technologies.

4. The study did not take into account financial limitations, including budget

constraints, funding availability, and cost-effectiveness of renewable energy

projects.

5. There was no prototype constructed for the optimal allocation of HRESDG units

developed in this research.

1.8 Outline of the chapters

A brief outline of the chapter layout of this study is as follows:

Chapter 1 presents the introduction and justification of the research. This Chapter highlights

background information, problem statement, sub-problems, research hypotheses, and

delimitations of the study.

Chapter 2 reviews the existing works that are relevant to the areas of developing optimization

model for this study. For clarity of presentation, the existing studies are grouped into 8

significant constituents that are investigated in the course of this research. The chapter serves

as the knowledge warehouse and theoretical foundation for the systematic conduct of this

research.

Chapter 3 describes the research methodology to optimally allocate HRESDGs into Ajayi

Crowther power system. It presents the models that calculate the output power from solar PV

modules, battery energy source system and the formulation of planning problem which helps

to attain the research objectives.

9
Chapter 4 presents how the optimisation model is implemented to generate the best possible

location and size of solar pv and battery energy storage system to be integrated into Ajayi

Crowther university at a minimum NPV cost. It summarizes the analysis and implementation

of the methodology.

Chapter 5 present summary of key findings, its contribution and potential areas for future

research

10
 CHAPTER TWO

LITERATURE REVIEW

2.1 Introduction

The integration of renewable energy hybrid distributed generation (HRESDG) such as solar

photovoltaic (PV) and battery bank storage into distribution system is expected to increase in

the future years. This is due to favourable technological advancement, economic profitability,

and environmental benefits (Ajeigbe et al., 2018). The power system at Ajayi Crowther

University, like many others, faces the challenge of integrating renewable energy resources

while maintaining system flexibility and reliability. Therefore, exploring solutions and

strategies for integrating HRESDG units is crucial to ensure steady power supply and

maximize renewable energy usage at the university.

Renewable Energy Hybrid Distributed Generators (HRESDGs) like solar PV, and battery

bank storage offer a wide array of advantages beyond their environmental benefits. They

provide low fuel costs, reduce greenhouse gas emissions, and contribute to energy security and

independence. Additionally, they create new job opportunities and stimulate economic growth

in regions where they are implemented (Ajeigbe et al., 2018). The integration of renewable

energy DGs into the power system at Ajayi Crowther University faces challenges due to the

variability and uncertainty of HRESDG resources and the passive setup of the network

system. These challenges are further propelled by the lack of reactive power compensation in

most renewable DG units. Additionally, the variability of large-scale HRESDG power can

lead to persistent system oscillations, requiring careful management for successful integration.

(Slootweg et al., 2003; Ajeigbe et al., 2020). Suboptimal allocation or ineffective optimization

of HRESDG units can escalate system oscillations and magnify the impact of intermittencies

on HRESDG units within distribution network systems (Li et al., 2017). Therefore,

11
strategically planning and designing the optimal allocation and timing of HRESDGs in the

Ajayi Crowther University power system emerges as a viable methodology to address the

intermittency issues. Notably, a significant portion of existing research in this domain has

leveraged mixed-integer linear programming (MILP) for its diverse advantages (Ajeigbe et al.,

2019). The research delved into the intricate optimization of where and how to place

renewable DGs, factoring in voltage stability constraints for distributed generation planning.

Its core objective was to reduce the net present value (NPV) of overall costs, which includes

investment, production, maintenance, energy losses, and unserved energy. Reference (Santos

et al., 2017) focused on determining the ideal size and location for HRESDGs, considering

transient stability constraints. Their objective was to effectively distribute solar PV and wind

DGs across distribution networks to minimize the net present value (NPV) of total cost.

(Mishra et al., 2019) and (Alturki et al., 2018) extensively employed power transfer capability

constraints to optimize the placement and sizing of renewable energy resource DGs in the

planning and design phases. The primary aim was to maximize the total DG capacity

integrated into the networks while minimizing costs.

In this study, an innovative joint multi-stage mathematical optimization formulation is

extensively detailed for the expansion of distribution networks. This formulation not only

models the sizing, placement, and timing of HRESDG units and battery banks but also

rigorously constrains the small signal stability variables.

Furthermore, the comprehensive literature review aims to extensively synthesize and analyze

the existing body of knowledge on enhancing the integration of HRESDGs for power system

flexibility, with a specific and thorough focus on Ajayi Crowther University. Through

meticulous examination of relevant research studies, industry reports, and academic

publications, this review extensively provides insights into the challenges, strategies, and best

12
practices for effectively integrating renewable energy resources into the university's power

system

2.2 Hybrid renewable sources distributed generation (HRESDG)

Worldwide, there is a rapid increase in energy usage and demand, surpassing the capabilities

of current energy resources to meet these escalating needs (Ajeigbe et al., 2018,

Hosenuzzaman et al., 2015). This makes RES like Solar PV an essential part of the global

shift toward more sustainable and environmentally-friendly energy systems. Renewable DGs

are RES units located in distribution networks close to load centres in order to meet immediate

power demand, reduce on-peak operating cost, reduces losses and enhance power stability

(Jordehi, 2016, Zubo et al, 2017, Georgilakis et al, 2015). The literature review will explore

recent research studies on optimal allocation of HRESDGs in distribution networks and

provide references that reflect current trends in the studies.

Renewable
Energy Sources

Solar Energy Wind Energy Biomass Energy Hydro-power

Figure 2.1: Main sources of RES systems

13
2.2.1 Solar photovoltaic

Solar PV technology is a pivotal component of renewable energy systems, offering a

sustainable method to generate electricity. Its popularity has surged due to cost reductions and

environmental considerations. Solar PV operates by converting sunlight into electricity

through semiconducting materials. When sunlight strikes the PV cells, it excites electrons,

generating an electric current. This process, known as the photovoltaic effect, forms the core

of solar PV technology. Factors such as solar irradiance and temperature significantly affect

the performance of solar PV systems. Arrays of interconnected solar panels can be scaled up

to meet varying energy demands.

In just a year, from 2018 to 2019, the world witnessed a remarkable 21% relative growth in

the cumulative installed capacity of photovoltaic (PV) systems, soaring from 483.1 GW to

580.2 GW. This surge underscores the accelerating global transition towards solar energy as a

key pillar of sustainable power generation (Irena, 2020). The highest contribution of currently

installed PV systems is identified in Asia, followed by China (175 GW), Japan (55.5 GW),

and India (26.8 GW). Europe comes in the second rank in terms of PV installed capacity with

considerable shares in Germany (45.9 GW), Italy (20.12 GW), and the UK (13.4 GW). The

efficiency of current photovoltaic (PV) technologies falls below 23%, emphasizing the

imperative for further advancements to enhance competitiveness. Various parameters affect

the efficiency of PV systems, with meteorological conditions playing a crucial role. Factor

such as solar radiation can influence the performance of PV plants. While desert regions with

high solar exposure are ideal for PV energy output, the accumulation of dust poses a

challenge. Research extensively explores the impact of dust deposition on PV system

efficiency, highlighting the need for effective mitigation strategies.

14
Cell temperature is a critical factor influencing the performance of solar PV systems. Elevated

temperatures can reduce the power output of PV panels, primarily affecting the open-circuit

voltage and ultimately leading to lower efficiencies. Managing heat and implementing

effective cooling techniques are crucial to improving PV system performance. Researchers

globally have developed various approaches to cool PV modules, conducting extensive

experimental and simulation studies to evaluate their potential in enhancing PV conversion

efficiencies. These studies underscore the importance of addressing temperature effects to

optimize the performance of solar PV systems (Hasanuzzaman et al., 2016) and (Reddy et al.,

2015).

Overall, solar photovoltaic technology harnesses the energy of sunlight to generate electricity

cleanly and sustainably, making it a key technology in the transition to renewable energy.

Figure 2.2: Working principle of Solar PV

15
2.2.2 Battery energy storage system

The hybrid configuration of the energy system in energy technologies is proving to gain

increasing success, incorporating numerous hybrid systems. (Lee and Kim., 2019) conducted

an in-depth analysis to determine the pulse power capability of batteries and SCs about

the power usage time and experimental verification. A Lithium-ion battery with 44V,

11Ah, and an SC with 36V, 30F is employed. Compared to conventional methods, the

proposed technique can provide a robust improvement in the performance and stability

of storage systems, particularly when combined with the estimation of the battery’s state of

charge.

Presently, there is a necessity to supply reliable, clean and sufficient energy to loads critical in

nature. As such, a secondary reserve power system has become increasingly important in

meeting this demand (Wang et al., 2017). Ordinarily, support systems depend largely on

batteries as a storage medium, but a higher system may utilize a generator set. Nevertheless,

each technology has its peculiar flaw. Specifically, batteries as a storage system present

discharge proficiency of high value, but at increased load current, the proficiency decays

rapidly with their rate capacity outcome (Okundamiya et al., 2022). The design and

development of a rechargeable storage system with high power and energy densities coupled

with a longer life span are of boundless significance. A novel technology super-battery is

ideally suited to the market for reserve power systems. Employed as combined or hybrid

energy sources such as super-battery and battery systems are depicting a new wave of backup

energy of higher dependability and increased reliability.

(Li et al., 2017) hinted that the crucial challenges impacting the power systems are the

variability attributes of renewables and the inability of generated energy profiles to meet the

load requirement. To this end, the authors proposed a super-battery/battery hybrid model

16
based on a micro-grid employed as a case study to examine the working compatibility of the

scheme. Furthermore, the evaluation depicts the capability of super-batteries in reducing the

stress resulting from the discharge/charge profile of the system.

The optimal sizing of a PV system incorporating a super-battery for improved self-

consumption and reduced grid fluctuation is proposed in (Jaszczur and Hassan, 2020). A

simple controller and inverter topology supply energy into the power system. The technical

feasibility of the model is simulated, time stepped analysed with the effect of the electrical

load taken into consideration. The results of power flow and energy self-utilization for a single

household, with the system increasing (from 53% to 100%) in initial self-consumption as

super-batteries increases by 16, while partly cloudy days (36%-80%) for increment in 10

super-batteries.

the battery’s state of charge.

17
 Figure 2.3: Working principle of Battery energy storage system (deye.com)

2.3 Some recent application of HRESDGs and their key result

In the field of electrical power systems, Renewable Energy-based Hybrid Distributed

Generation (HRESDG) systems have emerged as a significant revolution. These systems offer

the capability to provide cost-effective and carbon-free electricity to consumers (Singh et al,

2020). For instance, a study using various types of biomass gasifiers in an off-grid renewable

hybrid system composed of photovoltaic energy, hydrokinetic turbines and batteries, shows

that the change of the cost in the apparatuses impacts the total cost of the system. However,

the photovoltaic system have the highest sensitivity among other systems (Cano et al, 2020).

HRESDG systems applications summary shows high dependency on PV cells. Solar systems

play a significant role in most of the HRESDGs system. However, these systems are more

applicable on large industrial sectors. This combination can happen by adding wind power,

biomass, bio generators, hydro systems and many other systems. Optimization of HRESDGs

attempts on reducing intermittency, reducing the size of PV, reducing CO2 emissions,

increasing RES fraction and lessening generators working hours. Table 1 shows a summary of

some recent HRES applications, technologies used and corresponding key results.

References Main Technologies Key results Improvement

Objectives used & savings
Khawaja et al, Sizing and Solar System −80 kW reduction Energy &
2019 energy in size of PV Economy
management of −50% reduction in
hybrid energy the size of
systems using Electrolyze and the
finite automata hydrogen tank

18
 Designing a Solar, wind -HRES with Energy
Li Q et al, 2019 desalination and reverse optimized
with hybrid osmosis installation
renewable systems. contains: 111
energy systems photovoltaic panels
using persistent and 5 wind turbines.
neural systems -The designed
system reduced
potential loss of
power supply
probability by
18.3% compared
with the base case.
Optimum sizing Solar, wind -40% Drop in CO2 Environment
Das Bk et al., of a hybrid turbine, emissions
2021 system for Micro Gas −33% rise in
electric and Turbine and renewable energy
thermal loads Li-ion system fraction
using excess battery
energy and
waste heat
Salisu et al., Analysis of the Solar and -Cost of energy: Environment
2019 effectiveness of Wind 0.110$/kWh; & Economy
a hybrid PV– systems -Present cost: 1.01$;
wind–diesel– -Operating cost:
battery energy 4723$;
system applied -Renewable
in a remote fraction:98.3%;
community. -Greenhouse gas
emission: 2889.36
kg/year;
Manrique et al, Studying of a Solar– −80% reduction in Economy
2020 hybrid solar– biomass operation cost.
biomass coffee

19
 bean drying system.
system.
Suresh et Proposing and Solar, wind -The minimum Energy
al. ,2020 enhancing a turbine and storage capacity of
Hybrid PV– Pumped the system:
Wind Turbine- Hydro 3930615 kWh;
Pumped Hydro Energy -Levelized Cost of
Storage Energy Storage Energy 0.27 $/kWh;
System in systems.
Coastline
Communitie
Murugaperumal Proposing and Solar, wind -Percentages of Not specified
et al., 2019 estimating turbine and generated power:
different bio -solar power: 86.8%
working generators -wind power: 12.7%
approaches for system. -bio generators
rural power: 0.5%.
electrification -Losses: 6.8%
using a hybrid battery bank losses
renewable and 1.78%
energy system converter losses
Taghizadeh et Optimizing Solar, fuel −5.52% lessening in Economy
al., 2019 hybrid energy cell and total increase of
system to battery operation cost with
provide thermal system. demand response
and electrical program.
operation. −13.28% lessening
in Cost deviation
with demand
response program.
Kim et al., 2019 Study of hybrid Solar -Payback period: 6- Economy
renewable system, year. -Benefit-cost
energy system Thermal ratio: 1.7.
with solar energy and

20
 heating for net heat pump
zero energy
community
Akpinar and Optimal sizing Solar PV, Lower cost Economy
AKpinar, 2020 and techno wind compared to diesel-
economic turbines, only system
analysis of battery
hybrid energy
renewable storage
energy system
for a small
island
Hu et al., 2020 Optimal Solar PV, Reduction in energy Energy and
planning and wind cost and improve Economy
operation of a turbines, a system efficiency
renewable battery
energy based energy
microgrid storage
considering system
demand
response and
battery energy
storage system
Zhang et al., Optimal design Solar PV, -Reliable electricity Energy and
2021 of a hybrid wind access environment
renewable turbines and -Reduced
energy system biomass greenhouse gas
for a rural area energy emission
in china

21
 Al-Saadi et al., Optimal design Solar PV, -Cost saving Economy and
2021 and operation wind turbine -Environmental environment
of a hybrid and diesel benefit
renewable generator as
energy system backup
for a residential
building in
Oman
Sharma and Optimal sizing Solar PV, -Reduction in diesel Economy
kumar of a standalone wind turbine consumption and
hybrid and battery operating cost
renewable energy
energy system storage
for an off-grid
telecom tower
in india

Table 1: Summary of some recent Hybrid RES applications

2.4 Formulation of HRESDG allocation planning problem

The optimization problem associated with Renewable Energy-based Hybrid Distributed

Generation (HRESDG) systems can be approached from two perspectives: planning and

design, or real-time operations. In the context of expansion planning, the focus is on

identifying the most efficient type, size, location, and timing for installing HRESDGs within a

distribution network. This process considers various constraints, including those related to

electrical network operations, investment, and HRESDG specifications. The overarching goal

is to maximize the power of HRESDGs while minimizing costs over a defined planning

horizon. Operating constraints in an electrical network are crucial, encompassing factors like

line flow capacity, network radiality, safety, and reliability. To achieve minimum cost,

22
aggressive optimization of economic targets is essential. This includes minimizing investment

and operating costs, reducing energy losses, cutting emissions from conventional grid power

purchases, and decreasing reliability costs. Planning the allocation of Renewable Energy-

based Hybrid Distributed Generation (HRESDG) systems is an exceedingly intricate and

challenging task. It involves tackling a non-linear, complex, mixed-integer, multi-objective

optimization problem, laden with stringent constraints. The quest for a global optimum

solution is tedious, given the sheer complexity and interdependencies inherent in such a

problem. The optimal allocation planning of intermittent hybrid Distributed Generation (DG)

systems is an exceptionally challenging task. It involves navigating through conflicting

objective functions, such as maximizing DG capacity while minimizing network instability.

Additionally, a variety of constraints must be carefully managed, including DG voltage

magnitude limits and line flow constraints. Complex decision variables, such as DG type,

number, size, location, and power factor, further exacerbate the complexity of the problem.

Furthermore, accurately modeling uncertainty, particularly in terms of intermittency, adds

another layer of intricacy to an already complex optimization process (Jordehi et al., 2016;

zubo et al., 2017; Theo et al., 2015).

2.4.1 Objectives of HRESDG allocation problem

The objective functions of a Hybrid Renewable Energy Sources Distributed Generation

(HRESDG) system can be expressed mathematically in two ways: linear or non-linear. A

linear formulation of the allocation problem leads to linear objective functions, while a non-

linear mathematical representation results in non-linear objective functions. The optimization

of Renewable Energy Sources Hybrid Distributed Generation (HRESDG) allocation can be

framed in either single-objective or multi-objective terms. In the context of single-objective

optimization, viewpoints such as those of a distribution energy resources developer or a

distribution system operator (DSO) are considered etc. (Jordehi et al., 2016; Georgilakis et al.,

23
2013; abdmouleh et al., 2017). In the existing research works, the most common single-

objective functions usually used in descending commonality are the following:

 System losses minimization (copper, power, and energy).

 Voltage deviations and drops minimization, or voltage profile and stability

enhancement.

 Costs minimization (total, energy/operation, investment, maintenance, and emissions),

coupled with penalties for compensating losses.

 Total harmonic distortions (THD) levels (current and voltage) minimization.

 Social welfare and profit maximization.

 Benefit/cost ratio maximization.

 Renewable DG penetration maximization.

 DG capacity maximization.

 Network system loadability limits maximization.

 System reliability metrics maximization (e.g., system average interruption duration

index [SAIDI]).

 Short-circuit level minimization.

 Distribution system revenue maximization.

 Uniform power flow maximization among feeders.

 Renewable power maximization.

A single objective optimisation problem optimises one objective function subject to some

inequality and equality constraints as can be written in mathematical form as the following:

Min or max F s ( x ) =f s (x) (2.1)

With the constraint,

R ( x )≤ 0 , (2.2)

24
 S ( x )=0 , (2.3)

Where F s ( x ) is the objective function with one objective and expression (2.2) and (2.3) are

inequality and equality constraints, respectively.

In a Distributed Generation (DG) allocation scenario, the multi-objective function requires

combining or adding multiple single objectives, each with conflicting goals. It's important to

note that no single solution can satisfy all objectives, so these multi-objective functions must

be concurrently minimized or maximized through a single-objective formulation (Syahputra et

al, 2018, Mohandas et al, 2015, Dixit et al, 2019). The multi- objective formulation are

grouped into (Zubo et al., 2017; Ajeigbe et al., 2019; Georgilakis et al, 2015).

 In the context of multi-objective optimization, the weighted sum formulation assigns

weights to each objective and combines them into a single objective function. While

this approach is conceptually simple, its application in non-convex optimization

problems poses significant challenges.

 In a multi-objective formulation with numerous contrasting objectives, a wide array of

conflicting goals are considered, and the ultimate solution involves striking the best

possible compromise among the available feasible solutions.

 In the context of optimization, the goal multi-objective index employs the goal

programming method to streamline a multi-objective formulation into a single, unified

objective function.

Each of these multi-objective formulations has its merits and demerits that makes it

suitable for a particular allocation problem. A multi-objective optimisation problem

simultaneously optimises n objective functions subject to some inequality and equality

constraints as in (4) to (6).

Min or Max F m ( x )=[f m 1 ( x )+ f m 2 ( x )+ .... f mn ( x ) ] (4)

25
 With the constraints

R ( x )≤ 0 (5)

S ( x )=0 (6)

Where F m ( x ) is the objective function that consists of n objectives and expressions (5) and

(6) are inequality and equality constraints, respectively

2.5 Decision variables and constraints of HRESDG allocation

In the allocation of Renewable Energy-based Hybrid Distributed Generation (HRESDG)

systems, decision variables are the parameters that need to be determined. These variables

include the type of DG system, its size and location, the number of units, power factors,

installation timing, the amount of active and reactive power it generates, storage capacity,

slack bus power (a reference point in power system analysis), bus voltage magnitude, and

phase angle. These variables are crucial in optimizing the design and operation of HRESDG

systems to meet various objectives such as minimizing costs, improving system reliability, and

maximizing renewable energy penetration. (Jordehi et al., 2016; Ehsan et al., 2018). Bus

voltage magnitude and phase angle are critical parameters in power systems that play a crucial

role in ensuring network stability. The voltage magnitude at a bus is vital for maintaining

voltage stability and preventing voltage collapse, which can lead to system failures. On the

other hand, the phase angle between different buses is essential for determining the small-

signal stability of the system, indicating its ability to withstand small disturbances. Monitoring

and controlling these parameters are key aspects of maintaining a stable and reliable power

system.

Constraints are applied on the DG allocation problem to exert restrictions over the

optimisation of the objective function(s) in respect of some decision variables. The most

26
common constraints used in the optimal HRESDG allocation problem formulations are

grouped into six.

 System reliability constraints: are essential for maintaining uninterrupted power supply

to end-users. They include constraints such as maximum SAIDI (System Average

Interruption Duration Index), limits on short-circuit levels, and radiality constraints.

These constraints ensure that the distribution system remains reliable and stable, even

under varying operating conditions.

 Technical constraints in DG allocation include power balance equality constraints

based on Kirchhoff’s current law and inequality constraints like transformer or line

overloading, capacity limits, transmission supply limits, and constraints on the number

of buses available for DG installation. These constraints are essential for maintaining

the reliability and stability of the distribution network.

 DG capacity constraints: These constraints related to the generation limits of

distributed generation units include various inequality constraints such as DG

penetration limits, capacity bounds, constant power factor requirements, discrete size

limitations for DG units, and restrictions on the maximum number of DGs allowed.

These constraints are crucial for ensuring that the DG allocation plan adheres to

technical specifications and system constraints.

 Power quality constraints: These constraints are essential for maintaining high-quality

power in distribution systems. These constraints are represented as inequality

constraints and include limits on current and voltage total harmonic distortion (THD),

as well as restrictions on voltage rise and sag. Compliance with these constraints helps

to ensure a stable and reliable power supply, minimizing the risk of equipment damage

and disruptions in power delivery

27
  Network stability constraints: Network stability constraints, including voltage drop or

deviation, bus voltage magnitude limits, and voltage angle constraints, are critical

parameters for ensuring power system stability. By monitoring and controlling these

parameters, the risk of voltage instability and system collapse can be minimized,

contributing to a more reliable power supply.

 Safety constraints are crucial for ensuring the safety of both the network and the

public. These constraints are typically represented as inequality constraints and include

considerations such as right-of-way constraints. Ensuring compliance with these

constraints is essential for preventing accidents and maintaining the safe operation of

the distribution network.

2.6 Modeling of uncertainties in HRESDGs allocation

Planning of distributed generation into the Ajayi Crowther power distribution system network

involves addressing various uncertainties and variability, especially with the integration of PV

and BESS. These uncertainties arise from the random and variable nature of operational

conditions over time. (Georgilakis et al, 2015). These uncertain parameter are grouped into

two based on either technical or economic effect (Zubo et al, 2017). In distributed generation

planning, technical uncertainties encompass variables such as generation values (impacted by

the intermittent nature of HRESDGs), demand values, generation and line outages, and

instrumentation or device failures. Economic uncertainties include factors like the cost of

production (covering fuel, maintenance, operation, labor, etc.), business taxes, market prices,

inflation rates, unemployment rates, and gross domestic product. These uncertainties present

challenges in accurately predicting and planning for the integration of distributed generation

into the power distribution system.

28
2.6.1 Methods for modeling/accounting the uncertainties of renewable energy resources

Modeling and accounting for the uncertainties of renewable energy resources are essential for

ensuring the accuracy and reliability of power system operation and planning. These

uncertainties stem from various factors, including weather conditions, resource availability,

and forecasting errors. To incorporate these uncertainties into power system models

effectively, several advanced methods and techniques have been developed:

 Probabilistic Modeling: In distributed generation planning, it's common to use

probabilistic models that consider the stochastic nature of renewable energy

generation. These models factor in elements like weather patterns and seasonal

variations, which impact the output of sources such as solar. By employing these

models, planners can better understand the variability of renewable energy generation,

making more informed decisions about integrating distributed generation into the

power distribution system (Zhang et al., 2019). You can estimate the parameters of

probability distributions based on observed data using statistical techniques like Monte

Carlo simulation or historical data analysis. (Hong et al., 2020). The models offer a

spectrum of potential generation outcomes, complete with their corresponding

probabilities, by taking into account the probability distributions.

 Ensemble Forecasts: Ensemble forecasting is used to account for uncertainties in

weather forecasting, which directly impacts renewable energy generation forecasts.

This technique involves generating multiple forecasts using different models or initial

conditions to capture a range of possible outcomes (Yang et al., 2021). It utilizes a

mix of weather forecasting models, numerical weather prediction models, and varying

input parameters to produce a range of scenarios that illustrate the uncertainty in

renewable energy generation. Ensemble forecasts offer a probabilistic outlook on

29
 future renewable energy generation, enabling the assessment of different possible

outcomes (Ding et al., 2020).

 Scenario-based Modeling: Scenario analysis helps planners consider a variety of

future conditions, such as technology advancements and policy changes, to assess

their impact on renewable energy integration. This approach enables the evaluation of

different scenarios and their implications for decision-making. These scenarios cover

the spectrum of uncertainties related to renewable energy generation. They serve as

inputs in power system simulation models to assess system performance and make

well-informed decisions considering the array of possible outcomes.

 Robust Optimization: Robust optimization seeks solutions that are resilient to

uncertainties. It involves developing models that can perform well without relying on

specific probability distributions, making it suitable for situations with significant

uncertainty. (Ramezanpour et al., 2019). These methods optimize system operation or

planning for different scenarios of renewable energy generation. The objective is to

minimize the worst-case performance across all scenarios, considering the uncertainty

in renewable energy resources. Robust optimization provides reliable strategies that

perform well under a wide range of uncertainties. (Neshat et al., 2021).

 Real-Time Monitoring and Forecasting: By continuously updating predictions based

on the latest data, real-time monitoring and forecasting techniques ensure that

estimates of renewable energy generation remain accurate and responsive (Wang et

al., 2020). By integrating real-time measurements, historical data, and advanced

forecasting algorithms, these techniques enhance prediction accuracy. Continuous

updates to the forecasts enable system operators to adjust power system operations in

response to changing conditions and uncertainties (Aadil et al., 2021).

30
  Data Assimilation: Data assimilation updates models and forecasts based on real-time

data, improving the accuracy of renewable energy forecasts and system operation.

This approach is particularly useful for short-term decision-making (Jiang et al.,

2020). These methods reduce uncertainties in renewable energy generation forecasts

by updating model states and parameters with observed data. Techniques like Kalman

filtering or particle filtering, known as data assimilation, improve the accuracy and

timeliness of predictions by integrating real-time measurements into forecasting

models (Lu et al, 2019).

2.7 Load profile model

The load profile model for a distribution network can be represented in various ways, such as

time-varying, single load level, multi-load level, fuzzy, or probabilistic models. However, the

time-varying load model, which considers different numbers and types of users and uses

hourly load values, provides a more detailed representation. This dynamic load model is

particularly useful for evaluating both long-term performance and analysis of the distribution

system. Furthermore, the load connected to the distribution network can be either concentrated

on the network buses or distributed along the network lines (Georgilakis et al, 2015). The

concentrated static load is modeled in optimal HRESDG allocation problem depending on the

power relationship with voltage as either (a) constant current, (b) constant power or variable

power, and (c) constant impedance.

 Constant power load model: Constant power model is a load representation where

power has a proportionality relationship with phase angle (ie ∝θ ), but independent of

voltage magnitude changes. This model can be efficient in the evaluation of long-term

dynamic stability of the distribution system.

 Constant current load model: Constant current or variable power model depends on

bus voltage magnitude. This presents a static model and represents power as an

31
 exponential function of voltage magnitude (9). This load model can achieve static

voltage stability of the system.

α
P=P n V ( 7 )

β
Q=Q n V (8)

Where P,Q , Pn, Qn are the actual values of real and reactive nominal voltages values of

real and reactive powers and nominal voltages values of real and reactive powers

respectively, α and β are the real and reactive power exponent values and these values

are common representative of the values used in most research works

 Constant Impedance load model:

The power voltage magnitude relation of a constant impedance model

2.8 Optimisation algorithms for HRESDGs allocation problem

The optimization of HRESDG allocation involves determining the optimal type, size, location,

and operating parameters of DG units to achieve specific objectives, such as minimizing

energy costs, reducing greenhouse gas emissions, improving system reliability, and enhancing

overall system performance. This optimization process is complex and challenging due to the

stochastic nature of renewable energy resources, uncertainties in demand, and the dynamic

nature of power system operation. To address these challenges, various optimization

algorithms and techniques were introduced. These optimisation approaches are classified into

conventional method, intelligent search method (IS), probable hybrid method and potential

intelligent search method. (Ajeigbe et al., 2020) The conventional method is sub divided into

analytical, numerical, and power flow methods, the intelligent search method is sub divided

into: genetic algorithm (GA), fuzzy logic (FL), simulated annealing (SA), tabu search (TS),

differential evolution algorithms (DEA), particle swarm optimisation (PSO), ant colony

optimisation (ACO), and harmony search (HS) method. Potential IS is subdivided into :

32
artificial bee colony (ABC), shuffle frog leaping algorithms (SFLA), cuckoo search (CSO),

plant growth simulation (PGSA), shuffled bat (SBA), biogeography-based optimisation

(BBO), imperial competitive (ICA), firefly algorithms (FA), intelligent water drop (IWD),

bacteria foraging (BFO), artificial immune system (AIS), big-bang big-crunch (BBBC), and

ant lion optimisation (ALO) method. Probable hybrid method is subdivided into:

analytical/analytical, analytical/IS, numerical/IS, IS/IS, and PF/IS.

2.8.1 Conventional Method

Recent years have seen a surge in interest in using analytical and traditional methods,

particularly those based on linear programming, for solving Distributed Generation (DG)

allocation problems, as evidenced by a review of conventional approaches. They are;

2.8.1.1 Numerical methods

Linear and non-linear programming are classical optimization methods that have seen some

use in addressing distributed generation (DG) allocation challenges. However, these

approaches, including linear programming (LP) and mixed-integer linear programming

(MILP), are frequently noted for their inflexibility. They typically necessitate that objective

functions exhibit convexity, linearity, and continuity, criteria that can be challenging to fulfil

in real-world scenarios (Abdmouleh et al., 2017).

 Linear programming (LP): LP is used for solving mathematical models whose

formulations exhibited linear relationships in order to minimise or maximise the

objective function. LP is mostly applied to power system optimisation problems to

determine optimal DG sizes, as it gives exact solutions (Zubo et al., 2017; Abdmouleh

et al., 2017). Reference (Hamam et al., 2010), (Dicorato et al. 2008), and (Alturki et

al., 2018) use LP to achieve high penetration and maximum energy harvesting of DGs,

33
 respectively, while maintaining voltage magnitude limits. Hamam et al., describes a

partitioning algorithm based on LP method to solve excessive memory requirement

and computation time of long-term generation plant mix problems. Dicorato et al.,

presented an LP-based Energy Flow Optimization Model (EFOM) for DG exploitation

and optimal energy-efficient technology diffusion, aiming to minimize environmental

impact and operational costs. While these LP algorithms have shown efficiency in

computation time and optimization performance, none of them considered network

stability in their evaluations. Future research could focus on integrating network

stability considerations into LP-based optimization models for a more comprehensive

analysis.

 Mixed integer linear programming (MILP): MILP, is a mathematical approach used

to solve problems by optimizing a linear objective function while adhering to linear

constraints, with some variables restricted to integer values. Implementing MILP can

be challenging for large-scale problems and can require considerable computing

resources. MILP is utilized in the strategic planning of smart grid technologies to

determine the best size and location for integrating renewable Distributed Generation

(DG) (Santos et al., 2017). (Muñoz-Delgado et al., 2015) developed a MILP solution

for the multi-stage expansion planning of distribution systems, considering total costs

of distributed generation (especially wind power) and the distribution network.

(Alturki et al., 2018) proposed a MILP approach to increase Distributed Generation

(DG) capacity in radial distribution networks by reconfiguring ties and switches. All

the research works considered here aimed to maximize total DG capacity but didn't

assess the impact of renewable power variability or report global optimality.

 Non-linear programming (NLP): NLP is applied to tackle mathematical models

characterized by non-linear objective functions and constraints, all involving

34
 continuous variables. The computation in NLP hinges on the derivatives of these

functions. When solving a non-linear problem, an iterative approach is taken, selecting

a search path. This is often done by specifying initial partial derivatives or reduced

gradients, which is known as a first-order method. Included in this category are

techniques like the reduced gradient method and other search methods (Wu et al.,

1979). In DG allocation planning, second-order NLP methods like the Newton-

Raphson method and successive quadratic programming utilize derivatives of

constraints and power flow functions (Cutsem et al., 1991 ; Kermamshahi et al.,1998)

 Mixed integer non-linear programming (MINLP): MINLP is used to solve

mathematical models with both continuous and discrete variables, as well as non-linear

objective functions and constraints. It has been applied extensively to find the best

locations and sizes for Distributed Generators (DGs) in power systems, using power

loss sensitivity indices to locate DGs economically and operationally. However,

MINLP faces challenges due to its large number of design variables and the time-

consuming nature of its computations (Zubo et al., 2017; Abdoumouleh et al 2017).

MINLP is employed in several studies to optimize the allocation of different types of

Distributed Generators (DGs) in power systems. (Porkar et al., 2011) and (Al Abri et

al., 2013) use MINLP to consider factors like electricity prices and electronically

interfaced DGs for voltage stability improvement, respectively. (Salyani et al., 2018)

apply MINLP for the simultaneous expansion planning of HV/MV substations,

multiple DG units, and robust MV feeder routing, using an adaptive Genetic Algorithm

(GA) to handle uncertainties in renewable generations, demand, and electricity and

fuel prices.

 Decomposition Method: The decomposition method simplifies large-scale allocation

problems by removing some constraints. In DG allocation, it divides the problem into

35
 two parts: one with simpler constraints and another with more complex constraints. In

their research, (Wang et al., 2014) and (Mena and García., 2015) utilize Benders

decomposition (BD) to optimize the selection, sizing, and siting of DG units

leveraging PV and wind resources. Their optimization goals include managing

investment, maintenance, and operational costs, as well as emissions, fuel costs, and

micro-grid profits. Their findings underscore the effectiveness of their methodologies,

demonstrating the robustness of their solutions across diverse conditions

 Dynamic Programming (DP): DP is a strategic optimization method that breaks

down intricate problems into more manageable components. Its defining trait is its

multi-stage approach, offering a versatile framework with various optimization

techniques to address specific aspects of a broader problem formulation. (Zubo et al,

2017). (Khalesi et al., 2017) employ Dynamic Programming (DP) to optimize

Distributed Generator (DG) placement in distribution networks, aiming to reduce

power loss, enhance voltage profile, and boost system reliability. Their study examines

various load conditions to maximize network operator profits. (Saif et al., 2013)

propose a DP algorithm for optimal allocation of distributed energy resources (DERs),

finding that well-sized and positioned DERs enhance system reliability. Other

researchers, such as (Popovi and Popovi. 2010), as well as (Martin et al., 2017) use DP

algorithms to solve multi-period planning problems for distributed generations, also

evaluating system voltage stability during DG integration.

 Quadratic Programming: Quadratic Programming (QP) focuses on optimizing

quadratic objective functions within linear constraints. (Bhowmik et al., 2000)

introduce a two-stage iterative solution technique that combines QP and Mixed Integer

Quadratic Programming (MIQP) for planning radial distribution systems.

In the initial stage, the focus is on pinpointing optimal substation locations, followed

36
 by fine-tuning the network configuration in the subsequent stage. (Sfikas et al., 2015)

implement a Sequential Quadratic Programming (SQP) algorithm to optimize the sizes

of photovoltaic, wind, and mass-based DG units, along with battery devices, in both

standalone and grid-connected microgrids. Their optimization objectives revolve

around minimizing energy costs and annual energy losses. The study highlights the

superior cost-efficiency of mass units compared to other DG types and underscores the

advantages of integrating battery devices into standalone microgrid operations. They

conclude that expanding the range of DG types and batteries leads to reduced energy

costs from DG units. (Vovos et al., 2005) and (AlHajri et al., 2010) leverage SQP and

Optimal Power Flow (OPF) algorithms to optimize DG allocation problems,

addressing fault level constraints in the former while excluding them in the latter.

 Gradient search (GS): The Gradient Search (GS) method, also called gradient

descent or ascent, was pioneered by Newton for solving minimization or maximization

problems. It's an iterative algorithm that finds local minima or maxima by adjusting the

current point based on the negative or positive gradient of the objective function. (Rau

and Wan, 1994) advocate for the GS method in optimally allocating DGs in meshed

networks, while (Vovos and Bialek, 2005) propose a similar strategy that accounts for

fault level constraints.

2.8.1.2 Analytical Method

Analytical methods are adept at formulating numerical equations to address optimization

problems, relying on calculations, mathematical, and theoretical analysis. Their accuracy is

contingent upon the quality of the developed model and can be strengthened by combining

them with other models, utilizing simulation results from the initial model. These methods

offer advantages such as rapid computation and straightforward implementation to achieve

problem convergence. However, their accuracy in handling complex problems may be

37
compromised by simplifying assumptions. Some of the methods under analytical methods are

stated below.

 Precise Loss calculation method: (Archaya et al., 2006) introduced two analytical

methods based on precise loss formulas for determining the optimal locations and sizes

of Distributed Generation (DG) units. These methods consider four types of DG units

and aim to minimize total power losses in distribution systems. Their approach builds

upon the work of (wang et al., 2004) which was limited to a single type of DG. On the

other hand, (Tawfeek et al., 2018) proposed a method that combines exact loss

formulas with Particle Swarm Optimization (PSO) to optimally allocate four types of

DG units in radial distribution networks. This method aims to minimize total active

power losses.

 Two-third (2/3) rule: The 2/3 rule is an analytical method that uses an intuitive

approach to place DG units or batterys in distribution systems, often with the aid of

power flow graphical displays. It suggests sizing the DG unit at 2/3 of the load and

installing it at a location 2/3 along the length of the feeder line. This method aims to

minimize the total VAR-miles of flow, losses, and voltage impacts. (Willis, 2000)

conducted an analysis on a uniform radial distribution network to study how voltage

and feeder losses in distribution systems affect the optimal placement of DG units.

This analysis utilized the 2/3 rule with zero points to determine the placement of

distributed generation units. The study found that placing the DG unit at the end of the

most loaded branch of the feeder was optimal. However, this method may not be

effective when loads are distributed unevenly on radial feeders. The author concluded

that the algorithm used in the study is approximate, and the optimality of the solution

obtained may be questionable.

38
  Perturbation analysis: These methods are used to simplify complex equations by

linearizing them around initial operating points. These methods involve adjusting

parameters to observe their impact on results, reducing computational time for large-

scale systems. They help locate DG units based on sensitivity indices and assess

uncertainties from renewable energy sources. Various studies, such as those by

(GÃuzel and Hocaoglu, 2009) and (Murty and Kumar, 201), propose methods to

optimize DG unit size and location to minimize power losses and improve voltage

profiles. However, while these methods excel at finding optimal locations, their ability

to determine solution optimality is limited.

2.8.8.4 Power flow (PF) methods

 Optimal power flow (OPF): The main goal of Optimal Power Flow (OPF) is to

determine the most cost-effective operating strategy for a power system, taking into

account both transmission and distribution considerations. OPF is instrumental in

optimizing the placement and sizing of Distributed Generators (DGs) while

considering economic factors. Researchers, such as (Karatepe et al., 2015) and (Ochoa

et al.,2011) have utilized OPF to assess various DG integration strategies, considering

factors like DG output variability, network losses, line capacity, and voltage profiles.

Additionally, (Vovos et al,2005) and (Vovos and Bialek, 2005) have proposed OPF

algorithms to determine optimal DG sizes, incorporating fault level constraints and

considerations for protection equipment and switchgear in distribution network system

(DNS) planning. (Meng et al.2019) have extended an OPF algorithm to include

distribution energy resources (DERs) and solid-state transformers (SST), aiming to

minimize generation costs in DNS integration.

 Continious power flow (CPF): CPF is a valuable tool in the allocation of renewable

energy resources. (Hemdan and Kurrat, 2011) utilized the Continuation Power Flow

39
 (CPF) method to integrate DGs into distribution systems efficiently, meeting the rising

load demand. Their study showed substantial benefits from DG integration, including

reduced power losses and improved voltage profiles. In a similar vein, (Hedayati et al.

2008) applied CPF to identify optimal locations for DG placement in a distribution

system. This approach resulted in reduced power losses, enhanced voltage profiles and

system stability, as well as increased power loading and transfer capacity

2.8.2 Intelligent search (IS) methods

Artificial intelligence (AI) involves implementing intelligent behavior in machines. In the

context of optimizing the location and sizing of Distributed Generators (DGs) in power

systems, AI plays a crucial role. Heuristic methods, a subset of AI, are algorithms that

expedite the process of finding near-optimal solutions. While heuristic methods offer

simplicity and robustness, their accuracy and precision can be questionable, and they often

require significant computational resources.

2.8.2.1 Genetic algorithm (GA)

Genetic Algorithms (GAs) are heuristic optimization tools introduced by John Holland in

1975. They simulate natural selection processes, including mutation, selection, crossover, and

inheritance, to search for optimal solutions. GAs evolve a population towards maximum

fitness under specified selection rules, with each member represented as chromosomes. The

algorithm starts with a randomly generated population, and through generations, elements are

modified to form a new population. This process continues until a satisfactory solution is

found or a maximum number of iterations is reached. GAs are suitable for both continuous

and discrete parameters and excel in finding global optimal solutions for various functions

without requiring derivatives (Zubo et al., 2017; Abdmouleh et al, 2017) They are particularly

useful for complex and ill-defined problems. However, GAs can be time-intensive for

complex large problems due to the need for repeated fitness function evaluations In the

40
literature, GAs are widely used for solving DG allocation problems. For example, (Silvestri et

al., 1999) and (Kashyap et al., 2017) proposed GAs to minimize distribution power losses by

optimizing the allocation of distributed generation units. Kashyap et al.'s optimal solution

showed a maximum percentage reduction in active power loss compared to other methods.

Their approach also considered voltage constraints and evaluated the voltage profile within

limits, but did not model renewable DG sources or power system uncertainties. Various

configurations have been suggested to improve the Genetic Algorithm (GA) method for

solving DG location and sizing problems. These include Non-Dominant Sorting Genetic

Algorithms (NSGA II), Quantum GA (QGA), Adaptive Genetic Algorithm (AGA), and

Genetic Algorithm and Multi-Attribute Decision Making (GA-MADM). Previous studies have

demonstrated that these methods can enhance the optimality of solutions compared to using

genetic algorithms alone for determining the optimal locations and sizes of DGs in distribution

networks.

2.8.2.2 Simulated annealing (SA)

Simulated Annealing (SA) was developed in 1983 by Kirkpatrick et al. and further defined by

Viado Cerny in 1985. It models optimization problems as annealing processes to find global

optimal solutions. SA is iterative and applicable to combinatorial optimization problems,

mimicking crystallization processes in physical systems at discrete search spaces (zubo et al.,

2017; Abdmouleh et al., 2017; Ha et al., 2017). The key element of SA is the cooling

criterion, utilizing initial and final temperatures (t_initial and t_final, respectively) and a

cooling rate (β). The process starts with a probable solution, followed by system disturbance,

and determines a new likely solution using a probabilistic acceptance criterion. SA algorithms

are known for their simple implementation and ability to provide good solutions to various

combinatorial problems, although they can be computationally intensive and lack an upper

limit on computation time. They may also face challenges with local minimum termination

41
and lack of information on the deviation of local minimum from the global minimum. Efforts

have been made to adapt SA to DG planning problems, including modifications to the creation

of random solutions and control of the annealing schedule. In the literature, SA algorithms are

used to locate and determine the capacities and sizes of DGs by converting the original model

to an equivalent model using either 𝜖-constrained or weighted average method. While SA can

reduce computation time compared to Genetic Algorithms (GA) and Tabu Search (TS), it may

sacrifice global optimality of solutions. In some studies, SA is employed to obtain optimal

planning solutions for radial networks to minimize total cost or to find optimal locations and

sizes of DGs to minimize power losses and enhance voltage profiles. Feasibility-preserving

SA algorithms have been proposed for solving DN reconfiguration problems, and an improved

SA-PSO algorithm is introduced to enhance location selection and capacity finding of DGs.

2.8.2.3 Particle swarm optimisation (PSO) method

The Particle Swarm Optimization (PSO) method, developed by Eberhart and Kennedy in

1995, draws inspiration from the social behavior of bird flocking and fish schooling. PSO

operates with particles moving randomly in a multidimensional search space, evaluating

fitness based on their own and neighboring particles' positions. They adjust their positions to

enhance their solutions, aiming for global optimal solutions. PSO is known for its simplicity,

robustness, and parallel computation capability with short runtimes. It requires minimal

parameter tuning and converges swiftly. PSO has found extensive use in solving Distributed

Generation (DG) placement and sizing challenges. It has been instrumental in optimizing the

allocation of multiple DG units to minimize power losses, operational costs, improve voltage

profiles, stability indices, or enhance power quality. Notably, PSO surpasses Genetic

Algorithms (GA) in solution quality and iteration efficiency. It effectively addresses DG

allocation problems even with inaccurate mathematical models or when uncertainties are

42
present, Numerous advanced PSO variants have been proposed, including PSO with inertial

weight (PSO-IW), PSO with constriction factor (PSO-CF), dynamic weighted aggregation

PSO (DWAPSO), binary PSO (BPSO), improved PSO (IPSO), adaptive PSO (APSO),

decimal-coded quantum PSO (DQPSO), hybrid PSO, and social learning PSO (SLPSO).

However, initializing design parameters can be challenging with PSO, and it may converge

prematurely to local minima in complex scenarios. Some studies did not consider renewable

DGs, and uncertainties of intermittent renewable DGs were not consistently modeled.

2.8.3 Potential intelligent search methods

These are some other optimisation algorithms that are recently developed and implemented to

efficiently solve the DGs allocation problems.

2.8.3.1 Artificial bee colony algorithm

Dervis Karaboga introduced the Artificial Bee Colony Algorithm (ABCA) in 2005 as a novel

optimization approach inspired by honey bees' foraging behaviour (Karaboga et al., 2005)

ABCA has been applied in various studies to solve optimal DG placement and sizing

problems, aiming to minimize power losses and enhance voltage stability in distribution

systems. For instance, Mohandas et al. and Dixit et al. utilized ABCA for this purpose. In

(Abu-Mouti and El-Hawary, 2011), optimal DG allocation was achieved using ABCA by

adjusting control inputs, colony size, and iteration number. (El-Zonkoly, 2013) and (Kefayat

et al., 2015) applied ABCA to solve distribution network planning problems, determining

optimal values for reinforcements and suitable schedules for installing new DG units. In

(Padma Lalitha et al., 2010) ABC algorithms were proposed and compared with the PSO

method. The results showed that ABC outperformed PSO in terms of providing superior

solutions and faster convergence. However, these studies did not include indices for assessing

the long-term dynamic stability of the system.

43
2.8.3.2 Cuckoo search algorithm (CSA)

The Cuckoo Search Algorithm (CSA), devised in 2009 by Yang and Deb, is an optimization

technique inspired by the brooding parasitic behavior of certain cuckoo species. This

algorithm has been employed in various studies, such as those by (Moravej and Akhlaghi,

2013), and (Aranizadeh et al., 2019) where it was used to enhance the allocation of biomass

and solar-thermal based DG units, aiming to reduce power losses and improve voltage

profiles. Furthermore, (Nguyen et al., 2016) proposed the utilization of CSA to determine the

optimal placement and sizes of distributed generation units, with the objective of minimizing

network losses and enhancing the voltage stability index.

2.8.3.3 Firefly algorithm

The Firefly Algorithm (FA), introduced by Yang in 2009, is designed to solve complex non-

linear multidimensional optimization problems efficiently. Inspired by the behavior of

fireflies, where brighter individuals attract others, FA is used to optimize the allocation of

Distributed Generation (DG) units (yang, 2010). This optimization aims to minimize both

active and reactive power losses while enhancing line loading. Researchers, including Nadhir

et al., have applied FA to determine optimal sizes and locations for multiple DGs on balanced

radial networks, with a focus on reducing power loss. Furthermore, Othman et al. have

adapted the traditional FA to better handle constrained optimization problems. This modified

FA offers advantages such as easier implementation, increased stability, and simpler concepts

compared to the original algorithm.

2.8.3.4 Shuffled bat algorithm (SBA)

Yang introduced the Shuffled Bat Algorithm (SBA) in 2010, inspired by micro-bats'

echolocation behavior, to tackle optimization problems. (Yammani et al., 2016) investigated

SBA's performance in optimizing distributed generation (DG) placement and sizing for load

enhancement in radial distribution systems. Meanwhile, (Candelo-Becerra and Hernández-

44
Riaño, 2015) used the bat algorithm to determine optimal DG sizes, numbers, and locations in

radial distribution networks. They also proposed a hybrid SBA-LSF approach for optimizing

battery bank sizes and locations.

2.8.3 Probable hybrid optimisation methods

Efforts to enhance optimization methods have led to the development of hybrid methods,

which combine existing techniques to improve solution quality, computation efficiency, and

implementation simplicity. These hybrids merge heuristic and conventional methods, creating

novel approaches to problem-solving. In the context of allocating and sizing Distributed

Generators (DGs), researchers have proposed various advanced hybrid algorithms (Zubo et

al., 2017). These hybrids integrate metaheuristic methods like Genetic Algorithms (GAs) and

Particle Swarm Optimization (PSO) with traditional optimization algorithms such as Linear

Programming (LP) or Quadratic Programming (QP). This integration aims to address the

complexities associated with DG allocation optimization.

2.9 Summary

The literature review on the optimal allocation of Renewable Energy-Based Hybrid

Distributed Generators (HRESDGs) in the Ajayi Crowther University power system flexibility

highlights various methods and approaches used to enhance the integration of renewable

energy sources into the power system.

The literature review starts by examining different renewable energy sources such as solar

photovoltaic (PV), and battery storage bank. It underscores the recent advancements in

renewable energy technologies, emphasizing their potential environmental benefits and

contribution to sustainability. However, it also acknowledges the challenges posed by the

variable and intermittent nature of these energy sources when integrating them into the current

power infrastructure. The review discusses the challenges associated with integrating

45
renewable energy, such as the intermittency and variability of renewable sources, which can

impact the stability and reliability of the power system. It reviews the Integration of decision

variables and constraints into HRESDGs allocation.

To address these challenges, decision variables and constraints were introduced into

HRESDGs allocation. The literature review explores different methods for modeling and

evaluating uncertainties in renewable energy resources. They also proposed several

optimization techniques to address these challenges, including Genetic Algorithms (GAs),

Particle Swarm Optimization (PSO), and other metaheuristic methods. These optimization

techniques are used to determine the optimal locations and sizes of HRESDGs in the power

system, taking into account factors such as power losses, voltage stability, and system

reliability.

In conclusion, it stresses the ongoing requirement for research and innovative approaches to

effectively manage the integration of these resources. This is seen as a critical step towards

achieving a more sustainable and dependable energy system.

46
 CHAPTER THREE

OPTIMAL ALLOCATION OF HYBRID RENEWABLE ENERGY SOURCES

DISTRIBUTED GENERATION IN AJAYI CROWTHER POWER SYSTEM’S

FLEXIBILITY

To optimally allocate Hybrid Renewable Energy Sources Distributed Generation (HRESDG)

within the Ajayi Crowther University Power System Network, a comprehensive and effective

framework is essential. This process aims to enhance the power grid's adaptability, reliability,

and environmental sustainability. Building on prior renewable energy studies at Ajayi

Crowther University, which focused on developing power models for renewable energy

integration within the ACU network, this study seeks to advance that research by optimizing

the sizing, timing, and placement of renewable DG units. The goal is to minimize costs and

enhance flexibility during integration into the distribution network. The strategic allocation

framework involves two levels. Firstly, the modeling of renewable energy resources and land

is required to describe the HRESDG potential at specific locations. For the site, a typical

HRESDG setup will include solar photovoltaic and battery energy system models that utilize

solar irradiance resources. Effective optimization of a hybrid generation plant hinges on

precise modeling of each resource. The planning problem for HRESDGs is addressed using a

multi-stage optimization model. The objective is to maximize the power generated and

absorbed by the Ajayi Crowther University distribution network while minimizing the net

present value (NPV) cost. This model employs stochastic mixed integer linear programming

(MILP), known for accurately representing the problem's physics and mechanics and

achieving a global optimal solution with minimal computational demand. Constraints are

47
applied to guide the optimization of the objective function(s) in relation to the decision

variables. The optimization model's performance is assessed using MATLAB syntaxes.

The methodology discusses the mathematical modeling of renewable resources and load, the

calculation of the output power from HRESDGs during the planning period, and the

formulation of the optimization model for solving the renewable allocation planning problem.

3.1 Site description for HRESDGs allocation

The first milestone towards enhancing power supply at Ajayi Crowther University Campus is

identifying suitable sites for integrating solar, and battery bank system. A comprehensive

identification methodology has been developed to assess the availability of these resources,

confirming complementary locations within the campus that meet the minimum energy

production requirements. This task was facilitated by the extensive work of the National

Centre for Environmental Information (NCEI) and SolarGIS (Solar Geographic Information

System) global solar and wind model. Their valuable criteria played a crucial role in

determining the optimal integration of these renewable resources at Ajayi Crowther University

Campus. They are:

 Solar resources exceed specific thresholds. For instance, solar power is at least 1740.8

kWh/m²/year.

 Effective planning and integration of the distribution network rely on understanding

key aspects such as the layout (including both current and proposed generation from

diverse resources), load pockets, available space along distribution lines, and potential

future generation sources. This crucial network information is vital for ensuring the

efficient and seamless incorporation of various power generation methods.

 Data availability for the site is crucial for planning. Ajayi Crowther University

Campus, situated in the Western part of Oyo Township, Nigeria (7.8496° N, 3.9480°

48
 E), spans an area of 0.51 km² and hosts a population of approximately 5,000 people.

The campus has an average annual electricity demand of around 2.983 MWh.

3.2 Modeling of renewable energy resources of site under study

Five years of historical data on renewable resources at Ajayi Crowther University were

analyzed using Beta, Weibull, Gamma, and Log-normal probability distribution functions to

estimate hourly solar irradiance. The data were categorized into four seasons—summer,

autumn, winter, and spring—with each month assumed to have thirty days, resulting in 120

data points per season (5 years x 3 months x 30 days). Researchers employed MATLAB's

"dfittool" to analyze the seasonal data, determining the shaping (a and k) and scaling (b and c)

parameters for the Beta and Weibull distribution functions. These parameters were then used

to transform the raw irradiance to match the Beta and Weibull distributions. A one-year study

period was chosen and divided into seasons, each represented by a typical day. The typical day

for each season was selected based on the lowest standard deviation from the season's 24-hour

average values. This day was then divided into 24 hourly segments. The calculated parameters

were utilized to generate frequency distributions for irradiance for each hour of the typical day

in each season.

Consequently, beta and Weibull distribution functions for each hour were produced using

(3.1) resulting in a total of 96 hourly periods across the year (24 hours per day, across four

seasons).

3.2.1 Modeling of solar irradiance

Solar irradiance data for each hour of a typical day in every season generally exhibited two

distinct distribution patterns. By analyzing these seasonal segments, we identified the

parameters for the Beta probability density function (PDF). These parameters were

49
subsequently used to characterize the hourly behavior of solar irradiance data for a typical day,

as illustrated in (3.1)

{
Γ ( s +q ) s−1 q−1
∗Si ∗( 1−Si ) 0 ≤ Si≤ 1 , s ≥ 0 , q ≥ 0
f B ( Si ) = Γ ( s ) Γ ( q ) (3.1)
0 ; otherwise

Where:

f B ( Si ) = the beta distribution function of Si

Si = solar irradiance

s = the scaling parameter of the beta distribution function

q = the shaping parameter of the beta distribution function

Γ = the Gamma function

3.3 Modeling of solar PV and battery energy storage system power (MW)

3.3.1 Modeling of solar power output

The output power of the PV module in each state was influenced by the solar irradiance,

temperature at the site, and the module's characteristics. After generating the Beta PDF for a

24-hour period of a typical day, the PV module's output power during different states was

estimated using its power performance curve for each hourly segment, as outlined in equation

(3.2). The average solar irradiance value for a specific state ( I ap ¿ was used to derive the state

output power from the module's power curve

T cp=T A + I ap ( NOT0.8−20 )
pv

I p=I ap [I sc + K c ( T cp −25 ) ]

V p=V oc −K c∗T cp

50
 V mppt∗I mppt
FF=
V oc∗I sc

P pv ,h=N∗FF∗V p∗I p

(3.2)

where T cp is the temperature of the PV cells (◦C) at state p; T A is the ambient temperature of

the site (◦C); K c is the current temperature coefficient (A/◦C); K c is the voltage temperature

coefficient (V/◦C); NOT pv is the nominal operating temperature of PV cell (◦C); Isc is the

short circuit current (A); V oc is the open circuit voltage (V); P pv ,h ( I ap) is the output power of

the PV module at state p; I apis the average solar irradiance of state p; FF is the fill factor (-);

I mppt is the current at the maximum power point (A); V mppt is the voltage at the maximum

power point (V); and prs is the solar panel rated power (MW). Availability factor P pv ,h (l ap )/ p rs

(parameter AFp p .h) is the output power of all PV modules at the ith bus. The parameter AFp p .h

was computed using the hourly irradiance data of each state. The calculation of the AFp p .h

parameter was global, as the parameter was time dependent and affected all the PV modules

equally.

The number (n p . i) of rated PV generators ( prs) per bus is one of the decision variables for the

problem. Multiplying the number of PV generators ( n p . i) and the rated PV generator ( prs ) with

the parameter AFp p .h in the power balance equation gave the actual power output of the solar

based DG units.

3.3.2 Modeling of battery energy storage system power output

The BESS unit’s output is dependent on both the solar-PV output and the system’s demand.

By modeling both the solar-PV and BESS units, it is possible to determine their actual output

for the specified time. A battery is the most important part of a BESS because it retains

51
electrical energy as chemical energy. BESS is operating as a load at charging state and a

generator in discharging state.

The BESS energy E BESS ( t ) at time “t” can be calculated using the followings equation:

{ }
EBESS ( t−1 ) , Pdiff ( t )=0
E (t−1 )+ P diff ( t )∗∆ t∗η ch, P ( t )<0
E BESS ( t )= BESS diff

P (t)
EBESS ( t−1 )+ diff ∗∆ t , P diff ( t )> 0
ηdisch

(3.3)

Pdiff (t) = P Load (t) − P RDG (t) (3.4)

The power difference Pdiff (t) at time “t” can be calculated as the diference between the load

power P Load (t) and the RDGs generated power P RDG (t) at time “t.” The time span of t is

denoted by∆ t . The charge and discharge efficiencies are denoted as η ch andη disch, respectively

3.4 Mathematical formulation of the planning problem

This section introduces a multi-stage optimization model designed to determine the optimal

sizes, quantities, timings, and locations of HRESDGs, which include solar PV, wind, micro

hydro and biomass DGs. The primary objective of this model is to maximize the power

generated and integrated from HRESDGs into Ajayi Crowther power Distribution Network

System (DNS) while minimizing costs. This model is formulated as a stochastic mixed-integer

linear programming (MILP) optimization problem implemented in a MATLAB environment.

Additionally, a linearized alternating current (AC) network model utilizing fast decoupled

power flow (FDPF) was employed to effectively capture the network system's characteristics.

3.4.1 The objective function

The goal of this planning formulation is to maximize the integration of renewable energy into

ACU Distribution Network System (DNS) from the Distribution System Operator’s (DSO)

52
perspective. This involves optimally allocating solar PV distributed generator and BESSs

while minimizing costs. The objective function of the formulated Mixed-Integer Linear

Programming (MILP) optimization problem is to minimize the Net Present Value (NPV) of

the total cost, while complying with the given linear constraints.

( 1+d )−t I (1+ d )−t

C T NPV = ∑ C t + ∑ ( 1+ d )−t (¿ ¿Ct M +C t E +C t X )+ ∑ (¿ C T M +C T E +C T X )¿ ¿ ¿
t ∈Ω
t d t ∈Ω
t
t ∈Ω
t d
(3.5)

The first term in equation (3.5), the cost term C t I represents the total investment cost,

amortized in annual installments over the lifespan of the installed components. This approach

assumes reinvestment in identical components upon the expiration of their lifespan, as

described by (Santos et al., 2017) The cost valuation follows the principle of an infinite or

perpetual planning horizon, as outlined in (Blank, 2017). In this study, the overall investment

cost is calculated as the sum of the investment costs for both existing and new distributed

generators (DGs) and battery banks, as specified in equation (3.6).

¿
( )
The capital recovery factor d 1+ d was used to amortize all investment costs, ensuring a
¿¿

return on capital for each component. The specific capital recovery factors for different

for all g ∈ DG; and for

LT
components were calculated as follows: for generators, d (1+ d )
g

¿¿

for all cb ∈ cb. Here, LT represents the lifetime of each component,

LT cb
battery banks, d (1+ d )
¿¿

and d is the interest rate on the investment. The formulation of the investment variables (

x−, i ,t −x−,i ,t −1) included in equation (3.6) ensures that the investment cost of a component is

accounted for only once. This stipulates that components can only be utilized at the beginning

of a planning year, regardless of whether they are acquired mid-year, thus preventing mid-year

utilization.

53
 i
C t =RF
g
∑ ∑ I C g ,i ( x g ,i , t−x g ,i ,t −1 ) + RFcb ∑ ∑ I C cb ,i ( x cb , i−x cb ,i , t−1 )
DG i cb i
g ∈Ω i ∈Ω cb ∈Ω i∈Ω

(3.6)

The second term in (3.5) corresponds to the operation and welfare costs throughout the time

stages. This term consists of three cost terms vis-à-vis: total maintenance cost ( C t M ), total

energy cost (C t Et), and total emission cost (C t X ). Equation (3.7) models the maintenance cost,

M
Ct , which is the sum of the respective maintenance costs of existing and new DGs (

E N E N
MC g ∧MC g and ) and battery banks ( MC cb ) and MC cb at each planning stage under the

principle of a perpetual planning horizon. The cost term C t E denotes the total energy cost in

the system based also on a perpetual planning horizon. C t E is the sum of the costs of power

generated by the existing and new DGs and the purchase from the utility for each stage as

modeled in (3.8). The costs of power generated by the existing and new DGs are the

multiplication of the cost of unit energy produced (i.e., expected cost of operation) and the

amount of power produced. The cost term C t X sums up the emission costs associated with

existing and new DGs and the power from the utility feeders, as characterized in (3.9). The

emission cost is the expected costs of emissions based on the power produced from the

existing and new DGs and purchased from the utility, respectively. Here, the cost function of

emission was assumed to be linear for the sake of simplicity. The cost function, in reality, was

highly non-linear and non-convex.

Ct =
M
∑ ∑ MC g E∗u g ,i , t + ∑ ∑ MC cbE∗u cb ,i , t + ∑ ∑ MC g N∗u g ,i , t + ∑ ∑ MC cbN∗u cb ,i , t
DG i cb i DG i cb i
g ∈Ω i∈Ω cb ∈Ω i ∈Ω g ∈Ω i∈Ω cb∈Ω i∈Ω

(3.7)

Ct = ∑ ∑ ∑ ∑ (OC ¿¿ g , i , s , h , t ¿¿ E P g ,i , s ,h ,t +¿¿OC g ,i , s ,h , t N P g ,i , s ,h , t ¿¿❑)∗90+ ∑ ∑ ∑ EP ς , s ,h ,t ss P

E
s h DG i s h ς
s ∈ Ω h∈ Ω g ∈ Ω i∈ Ω s ∈Ω h ∈Ω ς∈Ω

(3.8)

54
 CtX= ∑ ∑ ∑ ∑ γ s ,tCO E ER g P g ,i , s ,h ,t E∗90+ ∑ ∑ ∑ ∑ γ s , tCO E ERg Pg , i ,s , h ,t N + ∑ ∑∑
E N
2 2

s h DG i s h DG i
s ∈Ω h ∈Ω g ∈Ω i ∈Ω s ∈Ω h ∈Ω g ∈Ω i ∈Ω s∈ Ωs h∈Ω i ς ∈Ω

(3.9)

CO 2 E
whereγ s , t , ER g, and OC g ,i , s ,h , t are the penalty for emissions ($/tCO2 e ), emission rate (tCO2 e

/MWh), and operation costs ($) for the respective components.

The third term in (3.5) presents the net present value of the costs incurred for production

(maintenance and energy costs) and emissions (welfare) after the last planning stage. This cost

is also referred to as the end effect, accounting for the residue values of the invested

components. It should be noted that this term is also estimated based on the principle of a

perpetual planning horizon and depends on the operation and emission costs of the last time

stage.

3.4.2 HRESDG allocation constraints

Constraints were applied on the HRESDG allocation problem to exert restrictions on the

optimization of the objective function(s) with respect to the decision variables. The constraints

used in the optimal REHDG allocation problem formulations were as below:

1. Complex power flow: The network AC power flow commonly known as Kirchhoff’s law of

voltage is highly non-linear and non-convex. The equations representing it are given below

2
Pk =V i Gk −V i V j (Gk cosθk + Bk sinθ k ) (3.10)

2
Qk =V i Gk +V i V j (G k cosθ k −Bk sinθ k ) (3.11)

The principle of the fast decoupled power flow (FDPF) model was applied to linearize (3.10)

and (3.11) to yield (3.12) and (3.13). The expression for voltage magnitude at bus i is given as

in (3.14)

55
Pk =[B ij∗θi ] (3.12)

Qk =[B ij∗V i ] (3.13)

V i=[V i+ ∆ V i ] (3.14)

Where V imin ≤ ∆ V i ≤ ∆ V imax

2. Voltage magnitude and angle constraints: The voltage magnitude and angle must be

constrained to maintain the voltage of the system

min max
V ≤ V ≤V (3.15)

min max
θ ≤ V ≤θ (3.16)

3. Power flow limits: On any line k, the power flow must be within the specified limits for that

line as in (3.17) and (3.18).

max
0 ≤ Pk ≤ Pk (3.17)

min max
Qk ≤ Qk ≤Q k (3.18)

4. Active and reactive power limits of power from transmission feeders: The power supply

from the feeder should have minimum and maximum limits for technical reasons. Inequalities

(3.19) and (3.20) enforce the limits.

SSmin SS SSmax
Pς ,s , h ,t ≤ Pς ,s , h ,t ≤ Pς , s , h ,t (3.19)

SSmin SS SSmax
Qς ,s , h ,t ≤ Qς ,s , h ,t ≤ Qς ,s , h ,t (3.20)

Where Pς ,s , h ,t =( 1−e )∗∑ P D i and Qς ,s , h ,t =( 1−e )∗∑ Q D i

SSmax SSmax
i i
i∈Ω i∈ Ω

5. Active and reactive power limits of REHDGs: The output of a unit generator should not

exceed its capacity multiplied by the generation binary variables. This ensures that the power

56
generation variable of a generator is zero when it is either unused or un-invested. The capacity

at a given year y and state s is considered to be constrained between maximum and minimum

values. The active and reactive capacity limits of exiting REHDGs are given in (3.21) and

(3.22), respectively. Equations (3.23) and (3.24) give the corresponding limits for the new

REHDGs.

Emin E Emax
Q g ,i U g , i ,t ≤ Qg , i , (t ) ≤Q g ,i U g , i ,t (3.21)

Emin E Emax
Q g ,i U g , i ,t ≤ Qg , i ,t ≤ Qg ,i U g ,i ,t (3.22)

Nmin N Nmax
P g ,i x g , i ,t ≤ P g , i ,t ≤ P g ,i x g ,i ,t (3.23)

Nmin N Nmax
Q g ,i x g , i ,t ≤ Qg , i ,t ≤ Qg , i x g ,i , t (3.24)

6. Reactive power limits of the battery bank: The reactive power supply from battery banks

should be bounded. Inequalities (3.25) and (3.26) limit the reactive power generated to

between zero and maximum capacity.

0≤ QEcb, i ,s , h ,t ≤ QEmax
cb ∗U cb ,i ,t (3.25)

0≤ QNcb, i ,s , h ,t ≤ QNmax
cb ∗x cb ,i ,t (3.26)

7. REHDGs and battery banks’ penetration limits in the DNS: The total penetration limits in

the system are determined by the government policy on the average penetration of the

renewable DG units and reactive compensators.

∑ AF g ,h ,t∗P Eg ,i , s ,h , t + ∑ AF g , h ,t∗P Ng ,i , s ,h ,t ≤ ϵ∗∑ PDi , s ,h , t ; ∀ ∈ 2 : N i

t t t
(3.27)
t ∈Ω t ∈Ω t ∈Ω

∑ cb
Q cb, ,s , h ,t + ∑ Q g ,i , s ,h , t ≤ ϵ ∑ QD i ,s , h ,t ; ∀ ∈2 : N i
E
t
N
t
(3.28)
cb∈Ω t ∈Ω t ∈Ω

57
where ϵ is the maximum penetration limit as a percentage of the load profile and AF g ,h ,t is the

capacity utilization factor of the generators.

8. Active and reactive power balance (Kirchhoff current law): When the distribution network

expansion planning includes new renewable generations with supply from the utility

substation, then the new network is modeled as (3.29) and (3.30). At each node, power

balance must be observed. This is achieved with the constraints:

ss
Pς ,sh , t + ∑ DG
E N
(P g ,i , s ,h ,t ¿+ Pg , i ,s , h ,t )+ ∑ i
Pk , s , h, t − ∑ i
P k ,s , h ,t =¿ PD i, sh ,t + P Lk ,s , h ,t ¿ ¿
g ∈Ω ¿ ,k ∈Ω out ,k ∈Ω

(3.29)

ss
Q ς ,sh , t + ∑ DG
E N
(Q g ,i , s ,h ,t ¿+Q g ,i ,s , h ,t )+ ∑ cb
E N
(Q cb ,i , s ,h ,t ¿+Q cb , i ,s , h ,t )+ ∑ i
Q k , s , h ,t − ∑ i
P k ,s , h ,t =¿ QDi , s , h ,t +Q Lk , s
g∈Ω cb ∈ Ω ¿ ,k ∈ Ω out ,k ∈ Ω

(3.30)

3.5 Modeling of load demands

The time varying load demand (hourly load demand) of the test systems which in this case is

Nigerian load profile was scaled with the Ajayi Crowther University Campus 10 bus network

load demand profile (Ezeh and Ezeh, 2021).

3.6 Capacity Utilization Factor

The Capacity Utilization Factor (CUF) is a key performance metric used to evaluate the

efficiency and productivity of power generation systems, particularly in HRESDGs like solar,

wind, hydro, and biomass. It is defined as the ratio of the actual output of a power plant over a

58
period of time to its potential maximum output if it had operated at full capacity during that

same period. CUF provides a clear measure of how effectively a power plant is being utilized.

Higher CUF indicates better utilization and efficiency of the installed capacity.

The CUF is typically expressed as a percentage and is calculated using the following formula:

Actual Energy Output (Mwh)

CUF= ∗100
Maximum possible Energy Output ( Mwh)

The natural variability of resources like sunlight and wind greatly affects the CUF. For

instance, solar power plants have a CUF ranging from 15% to 25%, depending on the location,

technology used, and other factors, Wind farms generally have a CUF ranging from 20% to

40%. Offshore wind farms tend to have higher CUFs compared to onshore due to more

consistent wind speeds, Hydro plants can have a CUF ranging from 30% to 70%, heavily

influenced by the availability of water flow and Biomass plants can achieve a CUF of around

60% to 80%, as they can operate more continuously compared to intermittent renewable

sources.

3.7 Summary

Ajayi Crowther University aims to evaluate the potential for generating renewable energy at

its campus by optimizing the allocation of various renewable resources. This will enable the

university to assess and maximize the amount of power generated from renewable sources

such as solar PV, and BESS. The optimal allocation of renewable energy sources involves

understanding and incorporating the power output characteristics influenced by various

factors.

For solar power, the solar irradiance and temperature of the site, along with the characteristics

of the photovoltaic (PV) modules, play a significant role. The main influencing parameter for

solar power is solar irradiance.

59
The main objective of this optimization model is to maximize the HRESDGs power generated

and absorbed into the distribution networks at a least possible NPV of the total cost. The

model was formulated as a stochastic mixed integer linear programming (MILP) problem,

while the non-linear AC network was made linear with the principle of fast decoupled power

flow in order to characterize the network without the loss of generality, maintain accuracy,

and reduce the computational complexity. Constraints are applied to the Hybrid Renewable

Energy System Distributed Generation (HRESDG) allocation problem to impose restrictions

on the optimization of the objective function(s) concerning the decision variables.

In summary, the optimal allocation of renewable energy resources at Ajayi Crowther

University involves creating and utilizing models that assess the potential power generation

from various renewable energy sources. These models consider power output characteristics,

capacity utilization factors, and the integration of both intermittent and continuous power

generation. By optimizing the allocation of renewable energy sources such as solar, the

university can ensure that it maximizes the power generation while adhering to constraints

such as power balance, capacity limits, environmental impact, budget, and technical

specifications. This approach allows the university to make informed decisions regarding the

implementation of renewable energy systems to effectively meet its power requirements in the

most efficient and sustainable manner.

CHAPTER FOUR

RESULT AND DISCUSSION

The optimisation was run considering the data provided in the month of June 2024. However,

the code could be also applied when dealing with other months or other dataset related to

different extension. In this study, a simulation on the average daily load demand was chosen to

better depict the obtained results, such as the behavior of the generators. A 10 bus system was

60
studied over a year planning period to test the developed model. The following data and

assumptions was made:

 For the sake of simplicity, 2% of investment costs was considered for the maintenance

costs of the corresponding components.

 The slack bus (Bus 1) was considered as the substation node whose voltage magnitude

and angle were taken to be 1.0 pu and 0.95 pu, respectively.

 Using the capacity utilization factor, solar DG of 600kWp installed capacity were

considered to be deployed as the potential DG power.

 The HRESDGs’ penetration limit, e, of 30% was considered, which was well above

the target level for Nigeria’s 2030 renewable projections.

 The battery bank storage system with 200kWh of installed capacity was considered to

be the minimum deployable reactive compensator in the system. The unit investment

cost for the battery bank was N37500/kWh.

 The price of electricity supplied by the substation (utility) was taken to be N212/kWh

for Ajayi Crowther University.

 The cost of power produced by solar energy DGs was N127/kWh.

 The power factor, pf, of renewable DGs was set to 0.95 lagging, meaning that the DGs

was Type II and always absorbed reactive power.

The optimizations, in this study, were implemented on an Intel(R) Core™ i3-10110U CPU

at 2.10 GHz and 8 GB of RAM using MATLAB R2019a. An optimality gap of 0.1% was

considered as the stopping criterion.

4.1 The ACU-10 bus location

The drone view of containing the network data of the ACU-10 bus system in Ajayi Crowther

is shown below in figure 4.1. This system has 10 load buses, which are located within specific

61
locations at the university. The time varying load demand (hourly load demand) of the test

system as scaled with the actual Ajayi Crowther load profile is used in this study.

Figure 4.1: Aerial view of the Bus location

4.1.1 Results of the optimal REHDGS’ allocation Problem

Intermittent renewable generations such as PV based DGs usually operate with fixed lagging

power factors, meaning such DGs always absorb reactive power from the network. Hence, the

power factor of PV in this study was set to 0.95 lagging. However, recently, there have been

photovoltaic generator systems emerging that provide reactive power support. Yet, the

optimization of renewable energy DGs’ integration into DNS presented here was lacking

reactive power support from these generations, and the optimization results are discussed

below.

Based on the stopping criterion considered, the computation time required to obtain optimal

solutions was 18s. The optimal solutions solar PV and the battery banks are depicted in Tables

1 and 2, respectively.

62
Table 1. Optimal investment solution of renewable energy hybrid distributed

generations (HRESDGs) for distribution system planning.

Type Bus T1 Type Bus T1

Solar 1 200kWp Solar 6 0

Solar 2 0 Solar 7 0

Solar 3 200kWp Solar 8 0

Solar 4 0 Solar 9 0

Solar 5 200kWp Solar 10 0

Table 2. Optimal investment solution of Battery bank storage system for distribution system

planning

Type Bus T1 Type Bus T1

BESS 1 200KWh BESS 6 0

BESS 2 0 BESS 7 0

BESS 3 200kWh BESS 8 0

BESS 4 0 BESS 9 0

BESS 5 200kWh BESS 10 0

Table 1 shows that equal Solar DG units were installed on bus 1, 3 and 5 and no solar unit

were installed on any other bus. This was because of the roof orientation and tilt thereby

increasing the capacity utilization factor of solar PV than any other bus location.

Furthermore, the power from battery energy storage system provided supportive services

such as spinning reserves to make up the inadequate supply from the intermittent

PVDGs. The totals of 600kWp of renewable power were installed in the network.

63
From Table 2, it can be deduced that the optimal locations of battery banks were

mainly on the loaded buses. This is a normal power system operation where battery

banks are used to compensate for the active power deficit in the system, thus helping to

enhance the power system flexibility and sustainability of network. It also shows the

total capacity of the battery banks invested in and installed to be 600kWh. The results

from Tables 1 and 2 are captured in Figure 4.2 demonstrate the complementarity of

intermittent solar PV generations and battery bank storage generation. Consequent upon

these results, the hybrid renewable generators were optimally allocated close to one

another and achieved a high optimality gap of 0.1%. The inclusion of reactive

compensators greatly increased the capacity of renewable DG units that were integrated

into the system by helping to maintain active and reactive power balance when reactive

power consuming generators were installed.

250

200
Power (KWp)

150

100

50

0
1 2 3 4 5 6 7 8 9 10
Bus number

Figure 4.2: Optimal locations and sizes of PV power installed throughout the planning

horizon

64
 250

200

150
Power (KWh)

100

50

0
1 2 3 4 5 6 7 8 9 10
Bus number

Figure 4.3: Optimal capacities of BESS installed throughout the planning horizon

For this particular study, ACU has an annual electricity demand of around 2.949 MWh,

averaging 8,191.2kWh per day. Cost of electricity per kWh is 212 naira. Net cost in a year is

N633.835 M.

Adding the 600 kWp PV parks and 600 kWh BESS reduced existing electricity bill by

73.6%.

amounting to the total NPV investment costs of N714.295 M for the three planning stages.

The cost of 600kWh BESS account for about 30% of the total cost of investment. The NPV

costs of maintenance, energy, and emission throughout the planning horizon were

correspondingly equal to N14.28 M, N67.410 M, and N574.221 M, respectively. The overall

total NPV cost for the whole planning stages year was N1.370 B. With the integration of

600kWP and 600kWh storage of HRESDG’ power, a total sum of N318.292 M would be

saved if the whole power to satisfy the load demand had come only from coal fired

conventional generations.

65
4.2 Summary

In the study of optimal allocation of hybrid renewable energy at Ajayi Crowther University

(ACU), various types of data were collected. This included electrical load data showing the

university's hourly energy demand, historical consumption patterns, and peak and off-peak

periods. Data on renewable energy resources such as solar radiation and potential biomass

resources were also gathered. Additionally, information about the existing power system

infrastructure, including transmission lines, substation capacities, and current power

generation capacities, was collected. Economic data on the costs of different renewable

energy technologies, energy prices, potential savings, and available incentives were

considered, along with environmental data to assess the potential impact of renewable

installations and carbon footprint reductions.

Using a Mixed-Integer Linear Programming (MILP) model, the optimal locations and sizes

for various renewable energy sources were determined. The goal was to minimize power

losses and maximize system flexibility while considering constraints like power balance,

grid capacity, renewable resource availability, and economic feasibility.

The results indicated that photovoltaic (PV) systems should be installed on rooftops and

open spaces with high solar exposure. Battery Energy Storage Systems (BESS) were

integrated to store excess energy generated during sunny periods for use during cloudy

periods or at night. BESS were best located near solar PV locations to ensure efficiency. PV

systems were recommended to cover about 60% of the energy demand, and the remaining

40% from battery storage to ensure continuous power supply.

The hybrid system was found to significantly enhance the flexibility of the power system,

better handling the variability in renewable energy generation and demand fluctuations. This

integration led to a more stable and reliable power supply, reducing dependence on any

66
single energy source. Economically, the optimal allocation resulted in substantial cost

savings of 73.8% of the existing electricity bill. This is due to reduced fossil fuel

consumption and lower operational costs. Environmentally, there was a significant reduction

in greenhouse gas emissions, aligning with sustainability goals and promoting a greener

campus.

CHAPTER FIVE

CONCLUSION

5.1 Summary of work

The optimal allocation of renewable energy sources involves determining the most efficient

67
and cost-effective placement and sizing of renewable energy units within a ACU power

system to achieve objectives such as minimizing power losses, reducing costs, enhancing

reliability, and increasing renewable energy utilization. The process typically involves using

advanced optimization algorithms and mathematical models. The testing methodology

includes data collection, mathematical modeling, optimization algorithms, simulation, and

validation. Key results showed significant reductions in power losses and cost savings,

increased renewable energy penetration, enhanced system flexibility, and environmental

benefits.

This study successfully developed mathematical optimization planning model that accounts

for the intermittency and variability of renewable energy sources and investment cost. By

concurrently determining the optimal size, timing, and locations of Hybrid Renewable

Energy Source Distributed Generation (HRESDG) units within the Ajayi Crowther

University distribution system, the model enhances system reliability and cost-effectiveness.

Furthermore, incorporating flexibility considerations, such as demand-side management and

electrical appliances and gadgets retrofitting, significantly improves the adaptability of the

power system.

In this study, optimal allocation of solar and battery storage system led to a 73.8% reduction

in Ajayi Crowther university yearly electricity bill and a 100% increase in the power

efficiency. Hybrid systems demonstrated improved reliability and cost-effectiveness, with up

to a 65% increase in renewable energy utilization and a 30% reduction in greenhouse gas

emissions.

In conclusion, the optimal allocation of hybrid renewable energy sources at ACU showed the

potential for notable improvements in power system flexibility and sustainability. By

strategically integrating PV and battery storage systems, the university can achieve a more

reliable, cost-effective, and environmentally friendly energy system. Continued research and

68
integration of other readily available renewable energy resources are essential to further

enhance the system's performance sustainability and adaptability.

5.2 Significant contributions

The research on the optimal allocation of hybrid renewable energy at Ajayi Crowther

University (ACU) makes several specific contributions that advance the understanding and

implementation of renewable energy systems. One of the main contributions is the

development of a Mixed-Integer Linear Programming (MILP) model tailored to determine the

optimal locations and sizes of hybrid renewable energy sources, including photovoltaic (PV)

systems and Battery Energy Storage Systems (BESS). This model is crucial for minimizing

power losses and maximizing system flexibility, making it a significant tool in the field of

energy optimization.

The MILP model integrates multiple renewable energy sources and storage systems into a

cohesive and efficient network. By carefully selecting the optimal locations for PV systems,

the model takes advantage of areas with the highest solar exposure, ensuring maximum energy

generation. The inclusion of BESS is particularly important as it addresses the intermittent

nature of renewable energy sources. By storing excess energy generated during peak

production times. BESS ensures a continuous power supply even during periods of low

generation, such as at night or during cloudy days. This integration of PV systems and BESS

significantly enhances the power system's ability to handle fluctuations in energy generation

and demand, leading to a more reliable and stable power supply.

The research also provides a detailed economic analysis, showing that the optimal allocation

of renewable energy sources can lead to substantial cost savings. These savings primarily arise

from reduced fossil fuel consumption and lower operational costs associated with renewable

energy systems. Fossil fuels are not only expensive but also subject to price volatility and

supply chain disruptions. By reducing reliance on these conventional energy sources, the

69
university can achieve long-term financial stability and predictability in energy costs.

Additionally, renewable energy systems typically have lower operational and maintenance

costs compared to traditional power plants. For instance, PV systems have no moving parts

and thus require minimal maintenance.

This economic analysis makes a strong financial case for investing in renewable energy

technologies. The initial capital investment in renewable energy infrastructure may be

significant, but the long-term benefits outweigh these costs. The reduction in operational

expenses and the potential for government incentives and subsidies for renewable energy

projects further enhance the financial attractiveness of such investments. Moreover, the

research highlights the potential for generating additional revenue through the sale of excess

energy back to the grid, where applicable, providing an additional income stream for the

university.

Environmentally, the study demonstrates significant reductions in greenhouse gas emissions

through the integration of renewable energy sources. This aligns with global sustainability

goals and highlights the potential for academic institutions to contribute to environmental

protection and climate change mitigation. Renewable energy sources, such as solar pv produce

no greenhouse gas emissions during operation, in stark contrast to fossil fuel-based power

plants. By adopting renewable energy technologies, Ajayi Crowther University can

significantly reduce its carbon footprint, contributing to the fight against climate change and

setting an example for other institutions to follow.

The methodology and findings of this research are scalable and replicable for other academic

institutions and similar facilities. The MILP model provides a valuable reference framework

for optimizing renewable energy integration in different contexts, taking into account local

resource availability, infrastructure constraints, and economic factors. This scalability makes

70
the research applicable beyond the confines of Ajayi Crowther University, offering a blueprint

for other universities and organizations aiming to enhance their energy systems and achieve

sustainability goals

5.3 Future scope

The research on the optimal allocation of hybrid renewable energy at Ajayi Crowther

University (ACU) lays a strong foundation for future exploration and advancements in

sustainable energy systems. The future scope of this research encompasses several promising

areas that can enhance the robustness, efficiency, and applicability of renewable energy

systems, addressing both technical and socio-economic challenges. One critical area for future

research is the integration of additional renewable energy sources, particularly wind energy

and biomass. Wind power can complement the existing photovoltaic (PV) systems and

biomass generators by providing a diversified energy mix that enhances overall system

reliability. Comprehensive studies on wind patterns, optimal turbine placements, and the

integration of wind energy into the existing hybrid system can significantly bolster the energy

supply's resilience. By integrating local wind resources and developing optimized strategies

for wind turbine deployment, researchers can identify how best to incorporate wind energy

into the university’s energy portfolio.

Advancements in energy storage technologies also present a crucial avenue for future

research. While Battery Energy Storage Systems (BESS) are integral to the current system,

emerging storage solutions such as super capacitors, flywheels, and hydrogen storage offer

promising alternatives that could improve energy management. Future studies should

investigate the feasibility, cost-effectiveness, and operational benefits of these advanced

storage technologies. Integrating a variety of storage options can enhance the system’s

capacity to manage fluctuations in energy generation and demand, ensuring a more stable and

reliable power supply.

71
The application of advanced optimization techniques is another vital area for future research.

The current study utilizes a Mixed-Integer Linear Programming (MILP) model, but exploring

more sophisticated methods, such as machine learning algorithms and artificial intelligence,

can further improve optimization accuracy and efficiency. These advanced techniques can

process larger datasets and consider more complex variables, providing deeper insights into

optimal energy allocation strategies. Machine learning models, for instance, can predict

energy generation and consumption patterns with greater precision, enabling more effective

resource management and enhancing the overall performance of the energy system.

Finally, fostering community engagement and education initiatives can amplify the impact of

renewable energy research. Developing educational programs, workshops, and outreach

activities can raise awareness about the benefits and functioning of renewable energy systems

among students, staff, and the broader community. Encouraging participation in energy-saving

practices and sustainability initiatives can create a culture of environmental stewardship and

inspire future generations to champion renewable energy solutions

72
 REFERENCES

Abdmouleh Z, Gastli A, Ben-Brahim L, Haouari M, Al-Emadi N. A. Review of optimization

techniques applied for the integration of distributed generation from renewable energy

sources. Renew Energy. 2017;113:266-280

Ackermann T, Andersson G and Söder L. (2001). Distributed generation: a definition. Electr

power syst res. 2001;57(3):195-204.

Adaramola, M. S., Oyewola, O. M., & Ogunjuyigbe, A. S. O. (2020). Optimal sizing and

techno-economic analysis of a hybrid renewable energy system for a rural community in

Nigeria. Energy report, 6, 739-750.

Ajeigbe O. A, Chowdhury S. P and Olwal T. O, Abu-Mahfouz AM. Harmonic control

strategies of utility-scale photovoltaic inverters. Int J Renew Energy Res. 2018;8(3):1354-

13

Ajeigbe, O. A., Munda, J. L. and Hamam, Y. (2019).“Towards maximising the integration of

renewable energy hybrid distributed generations for small signal stability enhancement: A

review,” International Journal of Energy Research, pp. 1–42.

Akpinar, E. K., & Akpinar, S. (2020). Optimal sizing and techno-economic analysis of a

hybrid renewable energy system for a small island. Renewable Energy 145, 1885-1900

Al-Saadi, H. M., Al-Farsi, M. A., & Kazem, H. A. (2021). Optimal design and operation of a

hybrid renewable energy system for a residential building in Oman. Renewable energy,

171, 528-543.

Alturki M, and Khodaei A. (2018). Increasing distribution grid hosting capacity through

optimal network reconfiguration. In: 2018 North American Power Symposium (NAPS)

IEEE; 2018: 1-6.

73
Cano A, Arévalo P, and Jurado F. (2020). Energy analysis and techno-economic assessment of

a hybrid PV/HKT/bat system using biomass gasifier: Cuenca-ecuador case study. Energy

2020;202:117727. http://dx.doi.org/10.1016/j.energy.2020.117727.

Das B. K, and Hasan M. (2021). Optimal sizing of a stand-alone hybrid system for electric and

thermal loads using excess energy and waste heat. Energy 2021;214:119036.

http://dx.doi.org/10.1016/j.energy.2020.119036

Dubey, S., Tiwari, P., & Sahoo, A. K. (2020). Optimal sizing of a standalone hybrid

renewable energy system for a residential building in India. Energy, 204, 117976.

Ehsan A, Yang Q. Optimal integration and planning of renewable distributed generation in the

power distribution networks: a review of analytical techniques. Appl Energy. 2018;210:44-

59.

Ezeh, C.I. and Ezeh, O.O. (2021). “Electricity Load Demand and Forecast for Nigeria’s

Residential Sector,” Energy Reports, 2021.

Georgilakis P. and Hatziargyriou, N. (2015). “A review of power distribution planning in the

modern power systems era: Models, methods and future research.” Electric Power Systems

Research, vol. 121, pp. 89–100, 5

Hosenuzzaman M, Rahim N, Selvaraj J, Hasanuzzaman M, Malek A and Nahar A. (2015).

Global prospects, progress, policies, and environmental impact of solar photovoltaic power

generation. Renew Sustain Energy Rev. 2015;41:284-297

Hu, S., Liu, X., & Wu, Y. (2020). Optimal planning and operation of a renewable energy-

based microgrid considering demand response and battery energy storage system. Applied

energy 267, 114896

74
International Renewable Energy Agency. Power system flexibility for the energy transition

part Overview for policy makers [Internet]. 2018, Available from: https://irena.org/-

/media/Files/IRENA/Agency/Publication/2018/Nov/IRENA_Power_system_flexibility_1

_2018pd

Jordehi A.R. (2016). Allocation of distributed generation units in electric power systems: a

review. Renew Sustain Energy Rev. 2016; 58:893-905

Katiraei F., & Iravani R. (2018). Power system flexibility for enhanced grid integration of

renewable energy sources. Proceedings of the IEEE

Khawaja Y, Allahham A, Giaouris D, Patsios C, Walker S, and Qiqieh I. (2019). An

integrated framework for sizing and energy management of hybrid energy systems using

finite automata. Appl Energy 2019;250 (2018):257–72.

http://dx.doi.org/10.1016/j.apenergy.2019.04.185

Kim MH, Kim D, Heo J, Lee D.W. (2019). Techno-economic analysis of hybrid renewable

energy system with solar district heating for net zero energy community. Energy

2019;187:115916. http://dx.doi.org/10.1016/j.energy.2019.115916.

Li P, Qi J, Wang J, Wei H, Bai X, and Qiu F. (2017). An SQP method combined with gradient

sampling for small-signal stability constrained OPF. IEEE Trans Power Syst. 2017;32:

2372-2381

Li Q, Loy-Benitez J, Nam KJ, Hwangbo S, Rashidi J, and Yoo C.K. (2019). Sustainable and

reliable design of reverse osmosis desalination with hybrid renewable energy systems

through supply chain forecasting using recurrent neural networks. Energy 2019;178:277–

92. http://dx.doi.org/10.1016/j.energy.2019.04.114.

75
Li, F., Yang, L., Chen, C., & Li, P. (2019). Power System Flexibility for High Renewable

Energy Integration: A Review. Renewable and Sustainable Energy Reviews.

Manrique R, Vásquez D, Chejne F, Pinzón A. Energy analysis of a proposed hybrid solar–

biomass coffee bean drying system. Energy 2020; 202:117720.

http://dx.doi.org/10.1016/j.energy.2020.117720.

Mohandas N, Balamurugan R, Lakshminarasimman L. Optimal location and sizing of real

power DG units to improve the voltage stability in the distribution system using ABC

algorithm united with chaos. Int J Electr Power Energy Syst. 2015;66:41-52.

Munoz-Delgado G, Contreras J, and Arroyo J. M. (2015). Joint expansion planning of

distributed generation and distribution networks. IEEE trans Power Syst.;30:2579-2590

Murugaperumal K, Srinivasn S and Satya Prasad GRKD (2020). Optimum design of hybrid

renewable energy system through load forecasting and different operating strategies for

rural electrification. Sustain Energy Technol Assess 2020;37(2019):100613.

http://dx.doi.org/10. 1016/j.seta.2019.100613

Porkar S, Poure P, Abbaspour-Tehrani-fard A., and Saadate S. (2011). Optimal allocation of

distributed generation using a two-stage multi-objective mixed-integer-nonlinear

programming. Eur Trans Electr Power. 2011;21(1):1072-1087

Salisu S, and Mustafa M.W, Olatomiwa L, and Mohammed O. O. (2019). Assessment of

technical and economic feasibility for a hybrid PVwind-diesel-battery energy system in a

remote community of north central Nigeria. Alexandria Eng J 2019;58(4):1103–18.

http: //dx.doi.org/10.1016/j.aej.2019.09.013

76
Santos, S. F., Fitiwi, D. Z., Shafie-Khah, M., Bizuayehu, A., Catalao, J. and Gabbar, H.

(2017). “Optimal sizing and placement of smart grid enabling technologies for maximising

renewable integration.” in Smart Energy Grid Engineering, pp. 47–81

Sfikas E, Katsigiannis Y, and Georgilakis P. (2015). Simultaneous capacity optimization of

distributed generation and storage in medium voltage microgrids. Int J Electr Power

Energy Syst. 2015;67:101-113

Sharma, R., Singh, B., & Kumar, A. (2021). Optimal sizing of a standalone hybrid renewable

energy system for an off-grid telecom tower in India. Renewable Energy, 173, 1298-1312.

Singh S, Chauhan P, and Singh, N. J. (2020). Capacity optimization of grid connected

solar/fuel cell energy system using hybrid ABC-PSO algorithm. Int J Hydrogen Energy

2020;45(16):10070–88. http://dx.doi.org/10.1016/j.ijhydene.2020.02.018.

Suresh V, Muralidhar M, and Kiranmayi R. (2020). Modeling and optimization of an off-grid

hybrid renewable energy system for electrification in a rural areas. Energy Rep 2020;

6:594–604. http://dx.doi.org/10.1016/j.egyr.2020.01.013.

Syahputra R, Wiyagi RO, Suripto S, Jamal A, Putra KT, & Mujaahid F. A. (2018). novel

fuzzy approach for multi-objective optimization of distribution network configuration in

complex system. Int J Appl Eng Res. 2018;13(2):1120-1127.

Taghizadeh M, Bahramara S, Adabi F, and Nojavan S. (2019). Optimal thermal and electrical

operation of the hybrid energy system using interval optimization approach. Appl Therm

Eng 2020;169(2019):114993. http://dx.doi.org/10.1016/j.applthermaleng.2020.114993.

Teng, J. H., Luor, T. S and Liu Y. H., (2001). Strategic distributed generator placements for

service reliability improvements. IEEE Power Engineering Society Summer Meeting, Vol.

2: IEEE; 2002:719-724

77
Zhang, Q., Zhang, L., & Yuan, Y. (2021). Optimal design of a hybrid renewable energy

system for a rural area in China. Energy conversion and management, 232, 113951.

Zhang, W., Sun, H., Sun, T., & Xu, Y. (2020). Modeling and forecasting of renewable energy

generation: A comprehensive review. Renewable and Sustainable Energy Reviews, 127,

10986

Zubo, R. H. A., Mokryani, G., Rajamani, H. S., Aghaei, J., Niknam, T. and Pillai, P. “(2017).

Operation and planning of distribution networks with integration of renewable distributed

generators considering uncertainties: A review.” Renewable and Sustainable Energy

Reviews, vol. 72, pp. 1177–1198, 2017

78