DESIGN AND EVALUATION OF A GRID CONNECTED PHOTO VOLTAIC

(PV) SYSTEM: A CASE STUDY OF MOI UNIVERSITY

ADMINISTRATION BLOCK

BY

KOECH, VINCENT KOSGEI

A Thesis Submitted to the School of Engineering, Department of Mechanical and

Production Engineering in Partial Fulfillment of the Requirements of the Award of the

Degree Master of Science in Energy Studies

(Renewable Energy Resources Option)

`
MOI UNIVERSITY

2021
 ii

DECLARATION

Declaration by Candidate

I hereby declare that this thesis is my original work has not been presented to any other

institution for the award of any certificate, and therefore may not be reproduced in

whole or part unless with the consent of my approval or that of Moi University.

Signature: ________________________________ Date:___________________

Koech, Vincent Kosgei

Reg No: TEC/PGMP/03/15

Declaration by the Supervisors

This research thesis has been submitted for examination with our approval as University

supervisors.

Signature: ________________________________ Date:___________________

Dr. Stephen Talai

Lect. Dept. of Mechanical and Production Engineering

Signature: ________________________________ Date:___________________

Dr. Lawrence Letting

Snr. Lecturer, Dept. of Electrical and Telecommunication Engineering

 iii

DEDICATION

I dedicate this work to my family for unwavering support, encouragement they gave

me as I researched, their patience and understanding during the entire study period.

May all Glory be to God.

 iv

ACKNOWLEDGMENT

I would like to acknowledge the invaluable guidance, encouragement and assistance

provided by my supervisors; Dr. S. Talai and Dr. L. Letting during the research process

and development of this thesis.

 v

ABSTRACT

Moi University relies on utility grid whose major sources are centralized generation
with petroleum, hydro and geothermal resources. With increasing students enrolment
every year, the grid power is not adequate, hence frequent outages are experienced.
During the occurrence, diesel generators are utilized as backup. However, fuel is
expensive for its operation. Fortunately, Moi University administration block is at
geographical location of 0.286°N latitude and 35.294°E longitude, where availability
of solar resource is throughout the year. The average has been indicated to be 5.56
kWh/m2/day. Therefore, the main aim of this research was to design, simulate and
evaluate a grid-connected system for Moi University administration block. The specific
objectives were: to evaluate the optimal values of solar resource parameters with
respect to tilt and orientation; to map and collect power rating data of power consumer
in the administration block; to design and simulate a grid connected PV system based
on the outcome of first and second objectives and finally to evaluate the economic
impact of incorporating battery bank to the grid connected system. To design an ideal
grid connected PV system this study had to get ideal solar resource parameters through
experimentation by varying solar panel tilt and orientation. Administration block’s
power demand was obtained through physical load audit and utility company bills
(KPLC bills) to obtain load demand as well as identifying the amount critical load
power consumption. Lastly, a grid connected system was designed, simulated at this
ideal orientation and its performance evaluated by PVsyst software. The economical
evaluation was done with respect to current market prices. The orientation of the
building was found to be desirable for having most of its roof surface area on its
longitudinal length sloping downward midway along North and South axis to forming
solar plane Azimuth 159° and −21° . The obtained load was 19745 kWh per Month,
with 66.24 kWh being daily critical load. The energy balance between load and
available solar resource done by PVsyst yielded a feasible and economical system of
100 Kw grid-tie system with a battery bank of 2145 Ah at 48 V. This battery bank was
sized to specifically support the critical load sub panel when this system intentionally
island during outage. The study obtained an ideal panel orientation as −21° tilt angle
and 159° azimuth angle. The designed system simulated yearly production of 265.21
mWh leading to which a savings of KES 4.3 million per year. This translated to
breakeven of 2.7 years for a 100 kW grid-connected system without battery bank while
3.4 years for the same system with batteries. In conclusion, such a system with these
savings on electric bills, improved power reliability, and short time of return on
investment can be adopted to take advantage of abundant solar energy resources.
Therefore, this study recommends a grid connected solar system not only for
administration block but to the whole university to take advantage of adequate solar
resource within the geographical location. Future work is needed on ways of enhancing
energy efficiency and its impact on sizing solar system as alternative power supply.
 vi

TABLE OF CONTENTS

DECLARATION ...........................................................................................................ii
DEDICATION ............................................................................................................. iii
ACKNOWLEDGMENT............................................................................................... iv
ABSTRACT ................................................................................................................... v
TABLE OF CONTENTS .............................................................................................. vi
LIST OF TABLES ....................................................... Error! Bookmark not defined.
LIST OF FIGURES ...................................................................................................... xi
ABBREVIATIONS/ACRONYMS............................................................................ xiii
OPERATIONAL DEFINITION OF TERMS ............................................................ xiv
CHAPTER ONE .......................................................................................................... 1
INTRODUCTION........................................................................................................ 1
1.1 Background of the Study ......................................................................................... 1
1.2 Utility Grid Supplying Moi University .................................................................... 1
1.3 Moi University Administration Building Electricity Bills....................................... 2
1.4 Kenya Government Power Planning ........................................................................ 3
1.5 Power Line Losses State .......................................................................................... 4
1.6 Distributed Generator Islanding ............................................................................... 5
1.7 Problem Statement ................................................................................................... 6
1.8 Justification of the Study ......................................................................................... 7
1.9 Objectives of Study .................................................................................................. 8
1.9.1General Objective ............................................................................................... 8
1.9.2 Specific Objectives ............................................................................................ 8
1.10 Scope of the Study ................................................................................................. 8
CHAPTER TWO ....................................................................................................... 10
LITERATURE REVIEW ......................................................................................... 10
2.1 Solar Photovoltaic (PV) Technology ..................................................................... 10
2.2 Types of Solar Systems .......................................................................................... 10
2.2.1 Off-grid solar systems ..................................................................................... 10
2.2.2 Hybrid solar systems ....................................................................................... 11
2.2.3 Battery-based grid-tie inverter ........................................................................ 12
2.2.4 Grid-tied solar systems .................................................................................... 12
2.3 Type of Feed in Grid Connected Inverters ............................................................ 14
 vii

2.4 Categories of Grid Connected System ................................................................... 14

2.4.1 Central Grid-Connected PV Systems .............................................................. 14
2.4.2 Distributed Grid-Connected PV Systems........................................................ 15
2.5 Equipment for Grid-Tied Solar Systems................................................................ 15
2.5.1 Grid-Tie Inverter (GTI) ................................................................................... 15
2.5.1.1 Micro-Inverters ......................................................................................... 16
2.5.1.2 Central inverters ....................................................................................... 16
2.5.1.3 String inverter ........................................................................................... 16
2.5.2 Power meter..................................................................................................... 16
2.6 Solar Insolation ...................................................................................................... 17
2.7 The Solar Resource - Kenya .................................................................................. 17
2.8 Components of Solar Irradiance ............................................................................ 18
2.8.1 Global horizontal irradiance ............................................................................ 18
2.9 Policy and Regulation Framework in the Solar PV Sector .................................... 20
2.10 Status of the Kenyan Solar Market ...................................................................... 22
2.11 Moi University Energy Situation ......................................................................... 22
2.12 Grid Connected PV Systems in Kenya ................................................................ 23
2.13 Site Assessment ................................................................................................... 24
2.13.1 Roof specifications ........................................................................................ 24
2.13.2 Fixed angles scheme...................................................................................... 25
2.13.3 Effects of Shading on Arrays ........................................................................ 26
2.14 Effects of Temperature and Irradiance ................................................................ 28
2.15 Other Derating Factors ......................................................................................... 29
2.16 Ways to Optimize a Commercial PV System ...................................................... 29
2.16.1 Module spacing and tilt ................................................................................. 30
2.16.2 Azimuth Optimization ................................................................................... 31
2.16.3 Inverter design and shade tolerance .............................................................. 31
2.16.4 A new era of system design........................................................................... 32
2.17 Simulation of Solar PV Systems .......................................................................... 32
2.18 Intentional Islanding ............................................................................................ 33
2.19 Inverters interconnection requirement ................................................................. 36
2.20 Intentional Islanding Configuration Through AC-coupling ................................ 38
2.21 Guidelines for Sizing and Interacting Two Types of Inverters............................ 40
2.22 Battery sizing ....................................................................................................... 42
 viii

2.23 Solar PV System Economic Analysis .................................................................. 43

CHAPTER THREE ................................................................................................... 45
METHODOLOGY .................................................................................................... 45
3.1 Introduction ............................................................................................................ 45
3.2 Data Collection ...................................................................................................... 45
3.3 Geographical Location ........................................................................................... 45
3.4 Administration Block Grid Supply ........................................................................ 46
3.5 Electricity Outages ................................................................................................. 46
3.6 Shading Analysis ................................................................................................... 47
3.7 Calculation of Solar Irradiation ............................................................................. 48
3.8 Experiment to Obtain Optimal Tilt Angle ............................................................. 51
3.9 Investigation of Offices Load ................................................................................ 51
3.10 Roof Top Tilt and Orientation ............................................................................. 52
3.11 Dimensioning of the PV System .......................................................................... 53
3.11.1 Module dimensions ....................................................................................... 53
3.12 Battery Capacity Determination .......................................................................... 56
3.13 Equipment’s ......................................................................................................... 58
3.13.1 Solar modules ................................................................................................ 58
3.13.2 Inverter .......................................................................................................... 59
3.13.3 PVGIS Website ............................................................................................. 60
3.14 STC Standard Test Conditions of Panels ............................................................. 61
3.15 AC Connection to Sub-panel ............................................................................... 61
3.16 Sensitivity Analysis of the Investment ................................................................ 61
CHAPTER FOUR ...................................................................................................... 63
RESULTS AND DISCUSSION ................................................................................ 63
4.1 Introduction ............................................................................................................ 63
4.2 Site Sun Path .......................................................................................................... 63
4.3 Average Sun hours ................................................................................................. 63
4.4 Administration Block Roof Analysis ..................................................................... 64
4.5 Sizing the Solar Plant ............................................................................................. 66
4.6 Electrical Load Mapping........................................................................................ 68
4.7 Critical Load Battery Backup ................................................................................ 71
4.8 Plant Layout ........................................................................................................... 73
4.9 PVSYST System Simulation ................................................................................. 75
 ix

4.10 Cost of the System without Batteries ................................................................... 86

4.11 Cost of Electricity ................................................................................................ 87
4.12 Return On Investment (RoI) of the system without batteries........................... 89
4.13 Sensitivity Analysis ............................................................................................. 89
CHAPTER FIVE ....................................................................................................... 90
CONCLUSION AND RECOMMENDATION ....................................................... 90
5.1 Conclusion ............................................................................................................. 90
5.2 Recommendation ................................................................................................... 91
REFERENCES ............................................................................................................ 92
APPENDICES ............................................................................................................. 94
Appendix I: July 2018 Electric Bill.......................................................................... 94
Appendix II: August, 2019 Electric Bill .................................................................. 94
Appendix III: Array Setup........................................................................................ 95
Appendix IV: PVGIS Curves of Admin GIS Location at 00, 30 & Optimum Tilt. .. 95
 x

LIST OF TABLES

Table 1.1: Administration block electricity bill for 2 nd July – 1st August 2018............. 3
Table 1.2: Failure curve of sampled countries between 1960 and 2014 ........................ 5
Table 2.1: Policies supporting solar (Source: GIZ Kenya, 2014) ................................ 21
Table 2.2: Array tilt vs site latitude ............................................................................. 25
Table 2.3: Derating factors (Rodgers CRC press London, 2010) ................................ 29
Table 3.1: Characteristic of panels............................................................................... 58
Table 3.2: Electrical Characteristics ............................................................................ 58
Table 4.1: Moi administration block irradiation table ................................................. 64
Table 4.2: KPLC Bill 02/07/2018-01/08/2018 ............................................................ 67
Table 4.3: KPLC Bill 02/07/2019-01/08/2019 ............................................................ 67
Table 4.4: Mapped Appliance Table ............................................................................ 68
Table 4.5: Mapped lighting points table ...................................................................... 69
Table 4.6: Critical load table ........................................................................................ 71
Table 4.7: Cost of Incorporating Battery Back up ....................................................... 72
Table 4.8: Inverter Dc Input parameters verses designed system parameters ............. 75
Table 4.9: Experiment results table ............................................................................. 76
Table 4.10: Summary table of various simulation ....................................................... 81
Table 4.11: Simulation data of 100KW plant with tilt 3° and azimuth 159° ............... 82
Table 4.12: Prices based on Chloride Exide – major solar distributor in Kenya ......... 86
Table 4.13: KPLC Cost of electricity .......................................................................... 87
Table 4.14: System Production .................................................................................... 88
Table 4.15: Sensitivity of ROL to array orientation .................................................... 89
 xi

LIST OF FIGURES

Figure 1.1: Lessos 232/33 KV, Substation 23MVA TX ................................................ 2

Figure 1.2: Average global power line losses between 1960 and 2014. ........................ 5
Figure 1.3: Energy balance facets structure ................................................................... 9
Figure 2.1: Off-grid system (Energy Informative,2014.)............................................. 11
Figure 2.2: Hybrid system structure (Energy Informative,2014.)................................ 11
Figure 2.3: Grid tie system structure (Energy informative, 2014.) .............................. 12
Figure 2.4: Ground mounted grid tie PV ..................................................................... 14
Figure 2.4: Horizontal irradiance ................................................................................. 19
Figure 2.5: Diffuse normal irradiance. ......................................................................... 20
Figure 2.6: Partially shaded cell/cells (Altestore, 2014) .............................................. 26
Figure 2.7: Fully shaded cell/s (GIZ 2014) .................................................................. 27
Figure 2.8: Panel I-V Curve of a module at different temperature. ............................. 28
Figure 2.9: Anti-islanding of grid tie inverters ............................................................ 39
Figure 2.10: Current path when grid power is present (Source: Outback USA, 2019).
.................................................................................................................. 39
Figure 2.11: Flow of power uninterrupted to critical loads during outage .................. 40
Figure 2.12: Predesigned consumer board from outback ............................................ 42
Figure 3.1: Location of Moi University Administration in GPS ................................. 46
Figure 3.2: Horizontal shading device, overhang, side view ....................................... 47
Figure 3.3: Vertical shading device, vertical fin, top view .......................................... 47
Figure 3.4: Incline surface and resource influencing parameters ................................ 49
Figure 3.5: Tilt angle experiment................................................................................. 51
Figure 3.6: Orientation of north facing roof as captured ............................................. 52
Figure 3.7: Moi Administration admin coordinate ...................................................... 53
Figure 3.8: Trojan battery ............................................................................................ 57
Figure 3.9: JKMS295M-60V Maxim Dimension ........................................................ 59
Figure 3.10: Inverter .................................................................................................... 60
Figure 3.11: Efficiency Curve Graph (Source: SMA solar, 2015) .............................. 60
Figure 4.1: Solar path at Kesses -Moi University ........................................................ 63
Figure 4.2: Global irradiance, temperature and wind velocity curves ......................... 64
Figure 4.3: Roof Analysis Administration block ......................................................... 64
Figure 4.4: Administration Block Moi University ....................................................... 65
 xii

Figure 4.5: Front view of Administration Block ......................................................... 65

Figure 4.6: Aerial view of Administration Block ........................................................ 65
Figure 4.7: Unused equipment stilled hooked to power .............................................. 70
Figure 4.8: DOD vs Cycle Life Curve Graph .............................................................. 72
Figure 4.9: Plant layout ................................................................................................ 74
Figure 4.10: Average power vs Angle curve ............................................................... 76
Figure 4.11: Diagram of the south facing roof surfaces .............................................. 77
Figure 4.12: Normalized production Bar graph ........................................................... 77
Figure 4.13: Optimized tilt of the south facing roof .................................................... 78
Figure 4.14: Normalized productions bar graph .......................................................... 78
Figure 4.15: The North facing roof surfaces ................................................................ 79
Figure 4.16: Normalized production Bar graph 3 ........................................................ 79
Figure 4.17: Optimized tilt of the south facing roof .................................................... 80
Figure 4.18: Normalized Production Bar graph 4 ........................................................ 81
Figure 4.19: Yearly production bar graph.................................................................... 83
Figure 4.20: Performance Ratio PR Bar Graph ........................................................... 83
Figure 4.21: Monthly Global Ho &Glob Inc kWh/m2 ................................................. 84
Figure 4.22: Loss Diagram .......................................................................................... 85
 xiii

ABBREVIATIONS/ACRONYMS

BoP Base of pyramid

BOS Balance of System

DG Distributed generation

EIA U.S Energy informative administration

FiAH Feed-in Approved Holder

FiT Feed in Tariff

GDC Geothermal Development Cooperation

IPP Independent power producers

IPP Independent Power Producers

IRR Internal Rate of Return

Ken Gen Kenya Electricity Generating Company

KESI Kenya Electricity Supply Industry

KPLC Kenya Power and Lighting Company

LPG Liquefied Petroleum Gas

MoE Ministry of Energy

MoEP Ministry of Energy and Petroleum

MPPT Maximum Power Point Tracking

MWh Mega Watt Hours

MWp Megawatts peak

NEC National electrical code

NGO Non-Governmental Organizations

REA Rural Electrification authority

SLP Solar lighting products

 xiv

OPERATIONAL DEFINITION OF TERMS

Connection point The point on the Distribution System, electrically closest

to the FiAH’s plant, at which FiAH generated energy is

exported.

Harmonic A sinusoidal component of a periodic wave or quantity

having a frequency that is an integral multiple of the

fundamental frequency.

Intentional Islanding Permitting one or a number of distributed sources to

continue operating autonomously and provide

uninterrupted service to local customers and revenue to

the mini-grid operator during outages on the main grid

Islanding A condition in which a portion of the utility system that

contains both load and distributed resources remains

energized while isolated from the remainder of the utility

system

Performance Ratio Is a measure of the quality of a PV plant that is

independent of location and it therefore often described

as quality factor. The performance ratio (PR), stated as

percent describes the relationship between the actual and

theoretical energy outputs of the PV plant. It therefore

indicates the proportion of the energy that is actually

available for export to the grid after deduction of energy

loss as thermal losses and conduction losses and of

energy consumption for operation. The closer the PR

 xv

value determined for a PV plant approaches 100%, the

more efficiently the respective PV plant is operation.

Specific energy production It is computed by applying a transposition model (Hay or

Perez) to the horizontal hourly values. The result depends

namely on the diffuse irradiance.

Expressed in [kWh/kWp/year], is an indicator that allows

the comparison of the system quality between

installations in different locations and orientations.

Traditional networks Are conceptualized as systems with large generators

connected away from the load side, so power transfer

occurs from the higher voltage level to the point of

consumption at lower voltage levels.

Transpose Factor The Transposition Factor is the ratio of the incident

irradiation (GlobInc) on the plane, to the horizontal

irradiation (GlobHor). I.e. what you gain (or loose) when

tilting the collector plane. It may be defined in hourly,

daily, monthly or yearly values.

 1

1 CHAPTER ONE

INTRODUCTION

1.1 Background of the Study

According to Kenyan Electricity Supply Industry (KESI), Kenya’s installed capacity is

at 2.3 GW with electricity generation mix of consists of hydro, geothermal, thermal and

wind (energypedia.info, 2020.). The base load is supplied by the cheaper hydro and

geothermal power generating plants while the peak power is supplied by the generators

like diesel, thermal and gas plants. Significant of this electrical energy is consumed by

industries in the urban centers.

Moi University is situated at an ideal location for solar resource throughout the year for

the utilization of photovoltaic for the generation of electricity. It is closer to the equator

where the sun is sun is overhead in most part of the year. The administration block is at

coordinate 0.282˚N, 35.295˚ E (Figure 3.6). At the moment, the university heavily relies

on utility grid whose major sources are centralized generation with petroleum, hydro,

and geothermal resources therefore marred with line losses due to long line

transmission and distributions. Diesel generator is used as a backup during power

outages. The whole administration block consumes roughly 31756.5KWh per month

with each unit going for Ksh. 16.17 making a bill of Ksh. 513502.60 per month while

the whole Kesses Main campus monthly bill is Ksh. 14,895,744 according to KPLC bill

of July 2018.

1.2 Utility Grid Supplying Moi University

Moi university electricity is fed by installed Lessos 2.5 MVA 33/11 substation. The

substation load has increased between the year 2017 and 2018 from 11MW to 18MW

presently with five evacuation feeders namely;

 2

 Eldoret - Lessos interconnector

 Chemelil Feeder

 Fluorspar Feeder

 Kapsabet Feeder

 Lessos Feeder

Figure 1.1: Lessos 232/33 KV, Substation 23MVA TX

(KPLC, 2020.)

The Eldoret-Lessos interconnector serves Moi University, Ngeria Prisons, Kabiyet and

Sangalu Dairies and others. Lessos power supply by distanced centralized generations

as far as Tororo through long range transmission as Bujagali-Tororo-Lessos High

Voltage power line ((REA, 2009).

1.3 Moi University Administration Building Electricity Bills

Due to numerous academic activities taking place at Moi university administration

block, the bill is significantly huge (over Kshs. 760,000 per month). The administration

has a significant power consumption during the day as shown by KPLC electricity bill

in Table 1.1. for the period of 2nd July, 2018 – 1st August, 2019.
 3

Table 1.1: Administration block electricity bill for 2nd July – 1st August 2018

Meter No. Previous Current Reading Consumption Consumption

reading reading Type type
040016113631 0 96 Real 96 Demand KVA
040016113631 0 94 Real 94 Demand KW
040016113631 563873 582889 Real 19016 High rate
040016113631 368516 378147 Real 9631 Low rate
Source: (KPLC bill)

1.4 Kenya Government Power Planning

The second medium Plan 2013–2017 of Vision 2030, it identifies energy as one of the

infrastructure enablers for Kenya’s transformation into “a newly-industrializing,

middle-income country providing a high quality of life to all its citizens in a clean and

secure environment”. Access to competitively priced, reliable, quality, safe and

sustainable energy is essential for achievement of the Vision. Government have put up

a 20-year feed-in-tariff for renewable resources established, as well as a zero rating of

export duty and a removal of VAT on renewable equipment, the Kenyan government

is actively facilitating renewable energy growth at utility scale, commercial and

industrial (C&I) scale, as on-grid solution (Global Energy insights,2018).

Furthermore, parliament came up with Energy Bill 2015 with intensions to unbundle

electricity transmission and distribution while liberalizing licensing of electricity

generation, transmission and distribution. This reflects the changing environment of

energy regulation in Kenya since it recognizes different sources of renewable energy

and creates the corresponding licensing and regulatory agencies. These include the

introduction of an energy and petroleum institute (which shall be responsible for the

nuclear energy program) and regulation of mid-stream petroleum, areas which did not

exist under the previous regulatory regime - Kenya gazette, August 2015.
 4

1.5 Power Line Losses State

Centralized generation exposes large regions on constant outages and huge

transmission and distribution losses. This leads to 600hrs/year of outages compared to

120hrs/year in South Africa. To guard themselves, firms resort to expensive backups

like diesel generators which lead to pollution (IEA, 2018). Moi University is one of the

institutions with standby diesel generator as an energy solution during times of outage.

Transmission losses at about 19.45% and power theft costs Kenya power 17.5 billion

annually. With such trend, KPLC was push spent Sh10 billion in a countrywide network

upgrade project – Operation boresha stima – launched in August 2015. In 2014, it also

had spent Sh1 billion in a programme – Boresha meme viwandani – aimed at boosting

power supply in manufacturing industries (Ben Chuma, 2016).

The status quo as is for now is that government investment on renewable energy

resource so far has been mostly by centralized generation systems – Garissa solar plant

55 MW and Lake Turkana wind plant 300MW will be connected to the national grid.

This centralized generation comes with long range transmission and distribution which

is characterized with high line losses and large area blackouts every time when a minor

grid fault occurs – 25 days/year of blackout in Kenya compared to 5 days/year in South

Africa (IEA Statistics). These have forced end users to invest in diesel-powered that are

costly and not environmentally friendly.

 5

Table 1.2: Failure curve of sampled countries between 1960 and 2014
Country 1960 2014 Curve
Kenya 18
South Africa 8
South Africa 6
China 5
United Kingdom 9 8
USA 9 6
Australia 13 5
Austria 8 5
Azerbaijan 14

The power failures, averaging two days a month, rank Kenya eighth on the list of

African countries that suffer the longest duration of electricity outages and power

losses.

Figure 1.2: Average global power line losses between 1960 and 2014.
(Source: IEA statistics, OECD/IEA 2018)

1.6 Distributed Generator Islanding

This is a condition in which a portion of the electric utility system that contains both

load and DG resources - which the utility has no control over remains energized while

isolated from the remainder of the utility system. In the case of this study, Islanding
 6

will allow the designed system to power administration bock even with loss of power

in the grid. This is unusual characteristics of Grid tie inverters which go off during grid

failure.

The battle for electricity customers in an increasingly competitive and deregulated

market environment is one of the challenges facing the electric power utilities of today.

Customers expect a reliable and efficient supply of power from their utilities. One of

the advantages that a DG can provide to the electric utility and to customers is the

possibility of improving the continuity of supply by implementing safe intentional

islands in the event of upstream utility supply outage. Implementing intentional

islanding of DG in a deregulated era will have an impact on electricity market prices.

This problem is considered in this thesis by solving the optimal power flow problem

while accounting for islanded operation (H.H. Zeineldin).

1.7 Problem Statement

Moi university electricity is supplied by fault-prone traditional grid which gradually

becoming insufficient due to rapidly expanding annual electricity demand at a rate of

13.5% as a result of rise in population and expanding economy (Megorden, 2017).

Kenya population rose by 2.3% in 2018 according to Kenya national bureau of

statistics. Therefore, the university suffer an average of 600 hours of blackout in a year

as the rest part of the country supplied by the same grid with centralized generations.

Just like other firms, Moi University resort to diesel generators to fill the gap that come

with these outages. Running diesel generators is costly. Also, its operation is not

environmentally friendly and this couple up with tedious routine maintenance of the

generator.
 7

Therefore, this research aimed at designing an economical and feasible Grid-connected

solar system with sized battery bank for critical loads to harness adequate solar resource

of 5.56 𝑘𝑊ℎ ⁄ (𝑚^2/𝑑𝑎𝑦) available to administration block against 19000 𝑘𝑊ℎ ⁄

𝑚𝑜𝑛𝑡ℎ as daytime consumption. a solution to these problems which is worsening with

time to decouple part of demand from the grid through decentralized generation.

1.8 Justification of the Study

Moi University administration block has electric demand of 374.5 MWh yearly (KPLC

Bill, 2018). The solar resource of this geographical resource is 5.56 sun-hours a day

(Table 4.1). This solar resource can be harnessed to generate and substitute of

265.2 𝑀𝑊ℎ ⁄ 𝑌𝑒𝑎𝑟 electricity, hence, grid savings Ksh. 4.3M per year. Grid-

connected system without batteries makes the system more economical since the

batteries are the most replaceable component in the PV system making the system

expensive in the long run. Furthermore, the system will raise awareness of renewable

resources among the students, thus, this will attract new apprentices to engage in the

PV industry and stimulate solar energy research. .

Administration offices being operational during daytime, hence, the peak load is during

the day which coincides with day time irradiation of sun resource. 5.564 sun-hours of

administration bock is abundant enough resource in tropical locations like in Kenya to

substitute most of administration demand-19016Kwh/month during the day and

9631KWh/month during the night in July, 2018.

Furthermore, solar PV systems will improve the voltage profile in the distribution lines,

thereby reducing transmission and distribution losses. This will encourage the

introduction of Building Integrated Photovoltaic (BIPV) legislation that would

accelerate solar PV penetration and reduce reliance on fossil fuel generators.

 8

1.9 Objectives of Study

1.9.1General Objective

To design, simulate and evaluate a grid connected 𝑃𝑉 system for Moi university

administration block.

1.9.2 Specific Objectives

i. To evaluate the optimal values of solar resource parameters with respect to tilt

and orientation.

ii. To map and collect power rating data of power consumer in the administration

block.

iii. To design and simulate a grid connected PV system for Moi university

administration block based on (i) and (ii) above.

1.10 Scope of the Study

The scope of this study is limited to collection and analysis of solar insolation, loads

demand data for Moi University administration block. The analyzed trends enables the

sizing of a typical grid connected solar photovoltaic electricity generation system. This

study does not cover energy management.

The flow plan of this thesis is as shown in Figure 1.3. The two main values that was

used to obtain energy balance of the power system sized and designed are;

i. The solar resource assessment of the site is a theoretical study based on

meteorological data provided by PVGIS website.

ii. To find the load power demand, a field survey of the administration block to

Moi University will be conducted.

 9

These two portions were the foundation that built a suitable PV system. To find the

optimal solution, the computer software PVsyst and Sunny Design will be used, which

will result in a numerical model that will end with economical evaluation of the system.

Daily solar rad data Module tilt Space measurement Electric consumer identification

Potential solar energy Load demand compilation

ENERGY BALANCE

Array Inverters

No of modules in series & no of batteries Backup system

Optimization results

Sensitivity Analysis

Figure 1.3: Energy balance facets structure

 10

CHAPTER TWO

LITERATURE REVIEW

2.1 Solar Photovoltaic (PV) Technology

Historically, PV deployment has been slowed by real or perceived challenges such as

high capital costs, lack of scale in manufacturing, shortage of raw materials, and

balance of system (BOS) performance limitations. Recent reductions in cell prices,

increased manufacturing and more aggressive policies have driven a rapid growth in

installed global solar PV capacity. Annual growth in installed PV has averaged over

58% since 2006. With an installed capacity greater than 137 GWs worldwide, solar

photovoltaic (PV) technology has become an increasingly relevant energy supply

resource (Rethinking Energy, IRENA 2014).

2.2 Types of Solar Systems

The Energy Informative of August 14, 2013. Categorized solar photovoltaic systems

depending on:

i. Connection to grid – off grid and grid tied

ii. Connection to other energy sources- hybrid mostly supplemented with

genset or grid

iii. Design use- stand alone and micro grid

iv. Size- Pico solar homes system as m-kopa delights.

These are discussed as follows:

2.2.1 Off-grid solar systems

An off-grid solar system (off-the-grid, standalone) is the alternative to one that is grid-

tied. For homeowners that have access to the grid, off-grid solar systems are usually

out of question. It ensures access to electricity at all times, off-grid solar systems require
 11

battery storage and a backup generator (if you live off-the-grid). On top of this, a battery

bank typically needs to be replaced after 10 years. Batteries are complicated, expensive

and decrease overall system efficiency.

Figure 2.1 : Off-grid system (Energy Informative,2014.)

2.2.2 Hybrid solar systems

Hybrid solar systems combine the best from grid-tied and off-grid solar systems. They

can either be referred as off-grid solar with utility backup power, or grid-tied solar with

extra battery storage.

Figure 2.2: Hybrid system structure (Energy Informative,2014.)

 12

2.2.3 Battery-based grid-tie inverter

Hybrid solar systems utilize batter-based grid-tie inverters. These devices combine can

draw electrical power to and from battery banks, as well as synchronize with the utility

grid. The bottom line is that, currently for the vast majority of homeowners, tapping the

utility grid for electricity and energy storage is significantly cheaper and more practical

than using battery banks and/or backup generators.

2.2.4 Grid-tied solar systems

Grid-tied, on-grid, utility-interactive, grid intertie and grid back-feeding are all terms

used to describe the same concept – a solar system that is connected to the utility power

grid.

Figure 2.3: Grid tie system structure (Energy informative, 2014.)

Though Grid Tied Systems is used only used when the sun is available it has numerous
advantages (Outback USA, 2019);

i. Save more money with net metering

A grid-connection will allow you to save more money with solar panels through better

efficiency rates, net metering, plus lower equipment and installation costs: - Batteries,

and other stand-alone equipment, are required for a fully functional off-grid solar
 13

system and add to costs as well as maintenance. Grid-tied solar systems are therefore

generally cheaper and simpler to install.

The solar panels will often generate more electricity than what you are capable of

consuming. With net metering, homeowners can put this excess electricity onto the

utility grid instead of storing it themselves with batteries.

Net metering plays an important role in how solar power is incentivized. Without it,

residential solar systems would be much less feasible from a financial point of view.

Most utility companies guarantee to buy electricity from homeowners at the same rate

as they sell it themselves.

ii. The utility grid is a virtual battery

Electricity has to be spent in real time. However, it can be temporarily stored as other

forms of energy (e.g. chemical energy in batteries). Energy storage typically comes

with significant losses. The electric power grid is in many ways also a battery, without

the need for maintenance or replacements, and with much better efficiency rates. In

other words, more electricity (and more money) goes to waste with conventional battery

systems.

According to EIA data, national, annual electricity transmission and distribution losses

average about 7% of the electricity that is transmitted in the United States. Lead-acid

batteries, which are commonly used with solar panels, are only 80-90 efficient at storing

energy, and their performance degrades with time.

Additional perks of being grid-tied include access to backup power from the utility grid

when your solar systems stop generating electricity for one reason or another. Further,

more it helps to mitigate the utility company`s peak load and as a result, the efficiency

of our electrical system as a whole goes up.

 14

2.3 Type of Feed in Grid Connected Inverters

The feeding method can be sub categorized as:

a) Direct Feed – Connection point at grid (MV to HV) and

b) Indirect Feed - Connection point at customer (only applicable for LV)

For MV, connection is to be made to a bus bar at a substation. No connection is to be

made directly to an overhead line or cable.

2.4 Categories of Grid Connected System

There are two major types of grids connected solar systems. These are:

2.4.1 Central Grid-Connected PV Systems

The existing electricity system typically consists of central power stations using a

variety of fuel sources such as coal, gas, hydro or diesel that provide power to end-users

via transmission lines and a distribution system (Vijay K. Sood, Haytham Abedelgawad

Distributed Energy Resource in Micro grid, 2019). The power station connects to the

transmission lines. The power produced is consumed by end-users at their actual

location such as factories, businesses and homes as in the figure 2.4.

Figure 2.4: Ground mounted grid tie PV

(Source: SMA Solar Technology AG, 2017)
 15

2.4.2 Distributed Grid-Connected PV Systems

Grid-connected PV systems are distributed throughout the electricity grid. This is the

most common type of PV system. Hence, this thesis focused on this type of PV System.

Distributed grid-connected PV systems are further categorized into two kinds:

i. Commercial and

ii. Residential

Commercial systems are generally greater than 10kWp and are located on buildings

such as factories, commercial businesses, office blocks and shopping centers (Vijay k.

Sood, Haytham Abedelgawad Distributed Energy Resource in Microgrid, 2019). The

power generated by these systems is typically consumed by the loads within the

building, so no excess power is pumped to the electricity grid.

Residential systems refer to those installed on homes and are generally smaller than

commercial systems, typically between 1 and 5kWp (Vijay k. Sood, Haytham

Abedelgawad Distributed Energy Resource in Microgrid, 2019). The power generated

by these systems is first consumed by any loads operating in the house during the day;

excess power is fed into the grid providing electricity to nearby buildings.

2.5 Equipment for Grid-Tied Solar Systems

There exist differences between the equipment needed for grid-tied, off-grid and hybrid

solar systems. Standard grid-tied solar systems rely on the following components:

 Grid-Tie Inverter (GTI) or Micro-Inverters

 Power Meter

2.5.1 Grid-Tie Inverter (GTI)

Role of a solar inverter is to regulate the voltage and current received from the solar

panels. Direct current (DC) generated is converted into alternating current (AC), which
 16

is the type of current that is utilized by the majority of electrical appliances. In addition

to this, grid-tie inverters- also known as grid-interactive or synchronous inverters,

synchronize the phase and frequency of the current to fit the utility grid (nominally

50Hz). The output voltage is also adjusted slightly higher than the grid voltage in order

for excess electricity to flow outwards to the grid (SMA, 2019).

2.5.1.1 Micro-Inverters

Micro-inverters go on the back of each solar panel, as opposed to one central inverter

that typically takes on the entire solar array.

Micro-inverters are certainly more expensive, but in many cases yield higher efficiency

rates (SMA, 2019). Home owners who are suspect to shading issues should definitely

look into if micro-inverters are better in their situation.

2.5.1.2 Central inverters

These inverters are connected to a series of strings of solar panels rather to just a single

string as in the String inverter. In this study, the generated DC power from a series of

solar panels is connected in parallel with the output from other series of solar panels at

a combiner box. The DC output from a combiner box is then connected to a single

central inverter.

2.5.1.3 String inverter

This is connected to a series of solar panels rather than to just a single panel as in the

case of micro inverters. AC current from a string inverter can then be combined with

output from other String inverters. A “string” here means series of solar panels.

2.5.2 Power meter

Most home owners will need to replace their current power meter with one that is

compatible with net metering. This device, often called a net meter or a two-way meter,
 17

is capable of measuring power going in both directions, from the grid to your house and

vice versa.

2.6 Solar Insolation

The surface of the sun, with a temperature of about 5 800 Kelvin, is emitting

electromagnetic radiation. The energy is spread out in the universe and when the

radiation reaches the outside of the earth’s atmosphere the mean energy content is 1367

W/m2, named the solar constant. About 40% of this radiation reaches the surface of

earth; the rest is reflected or absorbed by the atmosphere. The maximum radiation is

about 1 100 W/m2, this is in places near the equator, like Kenya. The energy density

per unit area, W (λ), as a function of wavelength, λ, is given by

𝒉𝒄
𝒘(𝝀)𝑨 = 𝟐𝝅𝒉𝑪𝒓𝟐 𝝀−𝟓 (𝒆𝝀𝑲𝑻 − 𝟏) ( 0. 1)

Where h is Planks constant (h= 6.63𝑥10^(−34) 𝐽𝑠 c is speed of light in vacuum (C=

3𝑋108 m/s), k is Boltzmann’s constant (K= 1.38𝑋10−23J/K), t is temperature of

blackbody in degrees Kelvin, λ is Wavelength.

2.7 The Solar Resource - Kenya

Kenya is located near the equator and has a great potential for solar power. The average

irradiance is between 4 and 6 𝐾𝑊ℎ/(𝑚2 /𝑑𝑎𝑦. Depending on the conversion

efficiency of solar modules, 10-18% of this energy can be converted to electric energy.

There exist regional and periodic differences in the solar resources of the country.

Seasonally, a place like Nairobi experiences high level of solar isolation at December

solstice. Similarly, sun radiation decreases between June solstice and September

equinox.

On the other hand, Kisumu has a good solar radiation throughout the year- both in

solstices and Equinoxes (timeanddate.com). Therefore, it is important that solar

 18

consumers to be conversant with solar resources in their area and on the long-term

implications of these resources on the system they are putting up.

2.8 Components of Solar Irradiance

Radiation reaching the earth’s surface is represented in a number of ways. These are

explained as follows:

2.8.1 Global horizontal irradiance

This is the total amount of shortwave radiation received from above by a surface

horizontal to the ground. It includes both Direct Normal irradiance (DNI)- as in Figure

2.5 and the diffuse horizontal irradiance (DHI). Where, Direct Normal irradiance is

solar radiation that comes in a straight line from the direction of the sun at its current

position in the sky. Figure 2.4 shows diffuse horizontal irradiance which is solar

radiation that does not arrive on a direct path from the sun, but has been scattered by

molecules and particles in the atmosphere and comes equally from all direction.
 19

Figure 2.4: Horizontal irradiance

 20

Figure 2.5: Diffuse normal irradiance.

2.9 Policy and Regulation Framework in the Solar PV Sector

The key government institution with direct influence in the solar energy sector focuses

on key policy, regulation, standardization and taxation issues in the solar energy sector

that have positive and negative impacts on solar PV market structure as in the Table 2.1
 21

Table 2.1: Policies supporting solar (Source: GIZ Kenya, 2014)

Key Policy/regulation Positive Effect in the Negative Effect in
Legal document
statement value chain the value chain
National Energy Tax and other concessions are
and petroleum planned to encourage
This may shift the
policy, draft 2015 investment in oil and gas,
focus to industrial
exploration of coal and Lower prices of solar
energy needs
geothermal, development of products
neglecting micro-
hydroelectric power as well as
level
other forms of renewable energy
such as wind, solar and biomass
Uncoordinated approach in
Drag rapid
policy implementation and
diffusion of solar
promotion of solar energy
PV products
projects.
Review of FiT to
The existing FiT structure for accommodate net
each technology metering shall promote
SHS
Energy (Solar The regulations shall apply to a
photo voltaics) solar PV system manufacturer,
• Improve quality of solar
regulation, 2012 importer, vendor, technician,
product in the market
contractor, system owner, a solar
• Consumer satisfaction
PV system installation and
consumer devices
• Improve quality of solar • Ensure traceability and
product in the market reduced quacks in the
• Consumer satisfaction market
Any person by himself, servant,
or agent undertakes or carries
out any solar PV system
manufacture, import, vending or Reduced existing
installation work without being actors creating
Ensure quality of
the holder of a license then in shock on the
imported products
force appropriate to the work upward adoption
undertaken or carried out or trend of solar PV
without being under the
direction of such a license-
holder;
Sessional paper No rigorous attempts have been
No.4 on energy, made to project cost effective
Dragging on rapid
2004 demand for the other renewable
diffusion of solar
energy sources including solar,
PV product
wind, biogas and municipal
wastes.
lack of awareness on the
• PAYG credit models
potential opportunities and
like MKOPA have
economic benefits offered by
transformed the sector
solar technologies; and,
Lack of appropriate credit and
financing mechanisms to
facilitate acquisition of solar
technology by the rural
population and urban poor.
 22

2.10 Status of the Kenyan Solar Market

Utilization of solar resources in Kenya begun in the 1870s, due to government’s use of

solar photo voltaic systems to operate broadcast installations in remote regions

(ESMAP, 2016). Around 1980s, international donors and NGOs started a key role in

the development of solar energy sector in Kenya to power social services, like school

lighting, vaccine refrigeration and water pumping (GVEP, 2014). Later, donor role had

gradually reduced over the years; paving way for private sector as from 1990s.

Today SLP entrepreneurs are leading the solar lighting industry who often relying

purely on market-based models, utilizing the latest technology, and designing based on

consumer tastes (Wright, 2015). Technology is improving at a rapid rate, business

models such as mobile phone enabled PAYG are maturing, and the focus by industry

players and market facilitators on addressing key market failures means that the SLP

market is ready for a substantial inflow of private sector investment and exponential

growth. The impact of heightened campaign and awareness creation by lighting Global

has seen the growth of SLP to close to a million by 2014.

Most of these modern solar lighting products incorporate features such as mobile

charging and consumer credit systems such as the pay-as-you-go (PAYG) (Yarime

et.al, 2015).

2.11 Moi University Energy Situation

Since inception of Moi University in 1984, electricity demand has been growing up

over time. A sharp rise in the number of students’ enrolment - 21.3% occurred in

2011/12 due to double intake adopted by ministry of education to avoid eighteen-

month break between students completing high school and joining university

(commission for university education, 2014). In the academic year 2018/19 Moi
 23

university reacted to this rise in demand by stopping students from cooking in the

hostels and considering putting up solar water heating solutions to all hostels to avoid

use of electricity to warm water (Student notice, 2017).

2.12 Grid Connected PV Systems in Kenya

Currently, there are four grid-connected solar power plants in operation in Kenya.

These include:

i. 60𝐾𝑊𝑝 Plant at a Tambuzi ltd (AHK, 2013; Hansen et al., 2015);

ii. 72𝑘𝑊𝑝 Plant installed Uhuru flower farm (Earley, 2015)

iii. 575𝐾𝑊𝑝 Plant installed at the UN compound in Nairobi (Pedersen, 2016);

iv. Plant at the SOS Children’s village in Nairobi (60𝐾𝑊𝑝) (AHK, 2013);

The first two plants financed by international donors while the other two were financed

by the industrial plant owners. The existing plants were delivered on a turnkey basis by

total system suppliers from abroad in cooperation with local consultancy companies

and installation contractors (Dinnewell, 2014). The second plant was constructed by the

UK-based company Arun Construction Services in cooperation with the local company

Azimuth Power (modules from Centro solar AG and inverters from SMA Solar Systems

(Hille and Franz, 2011). In the third plant, the tea farm owner commissioned the UK-

based company, Solar Century, to deliver the plant, including import of key

components, in cooperation with the Kenyan-based companies East African Solar Ltd.

and Azimuth Power (Solar Century, 2014). An additional plant at the Strathmore

University (𝑂. 6 𝑀𝑊𝑝), which in 2015 signed a PPA with KenGen, seems to be close

to starting operation. Kenyan companies Quest works and ReSol were contracted as the

total system provider and installation contractor respectively, and major components

will be sourced from European and Chinese suppliers - panels from Jinko Solar while

the inverters are from Solar edge. Various projects of significant scale are under
 24

development in Kenya as part of the feed-in tariff system. Currently, it offers a tariff of

US$ 0.12/kWh for project developers (ERC, 2015). This includes the Garissa project

(55 MW), the Samburu project (40 MW), the Green millenia Energy project (40 MW),

the Alten Kenya Solar farm project (40 MW), the Nakuru project (50 MW), the Witu

Solar Power project (40 MW), and the Kopere Solar Park project (17 MW), (Hansen et

al., 2015; IREK, 2015).

Project planning preparation for these projects started in 2012, (ERC, 2015; Eberhard

et al., 2016). Most of the projects are supported by a number of donors and development

banks, like the German development agency and the World Bank.

2.13 Site Assessment

A site assessment aims to determine the location of the PV array, the roof specifications,

and amount of shading, available area and other considerations.

During the site assessment the installer should collect roof tilt and orientation, roof

space and strength, site coordinates, shading on location and other site specific required

to optimize system design. In most urban areas, the array is located on the roof of a

building, or in cases where there is a large, clear area of ground that will not be shaded.

2.13.1 Roof specifications

Orientation: Ideal orientation is where a module receives maximum sunlight (this is

true south for the northern hemisphere or true north for the southern hemisphere).

Unfortunately, when a PV array is installed on a roof its orientation is governed by the

direction of the roof. Using a compass and magnetic declination data, installers should

determine the orientation that the roof is facing and its bearing from the ideal

orientation. The orientation of the roof will be the same as the orientation of the modules

and will be required for energy yield calculations.

 25

Tilt angle: Tilt of the modules will follow the tilt angle (or pitch) of the roof. The tilt

angle should be measured using an inclinometer or an angle finder; it may also be

available on architectural drawings of the building. The optimum tilt for a system is

equal to the latitude of that location. In cases where the pitch of the roof is not equal to

the optimum tilt angle the PV array’s energy yield will be affected.

Once the tilt and orientation angles of the PV array have been determined, the designer

may want to calculate their effect on energy yield. This is normally done using data

tables rather than by hand; with current development, data tables are available.

Space: This is an available area to which the panel are laid. It plays a vital role as if not

optimized; it limits the amount of energy that can be generated.

2.13.2 Fixed angles scheme

National Renewable Energy Laboratories (NREL) research affirmed that fixed angles

for different latitudes can produce very good results as indicated in Table 2.2. However,

for locations very near to the equator, the purpose of tilting the array is to enable its

natural cleaning during rainy seasons to avoid dust accumulation. Furthermore, tilting

the array also provides free air space under the array which enhances the cooling of the

array (Solar century, 2014).

Table 2.2: Array tilt vs site latitude

Latitude Recommended tilt angle

Latitude < 15 degrees 15 degrees

15 degrees < Latitude < 20 degrees Latitude

20 degrees < Latitude < 35 degrees Latitude + 10 degrees

35 degrees < Latitude Latitude +15 degrees

(Source: National Renewable Energy Laboratories, 2009)

 26

2.13.3 Effects of Shading on Arrays

A solar PV module reacts by generating electricity well when it receives direct radiation

from the sun (Nielsen, R. 2005). When the sun is obstructed from the module, the output

of the module is reduced proportional to the amount of shading (Ramaprabha, 2013).

Module cell shading is either caused by trees, flying objects and even clouds. Short

lasting shading does not affect the daily charging current but persistent shading reduces

the daily output current of the modules.

As seen in Figure 2.6, a “soft” source, like a distant tree branch or cloud causing a

diffused shade that can significantly reduce the amount of light reaching a solar panel’s

cells. “Hard” sources block light from reaching solar cells, such as debri or bird

dropping sitting on top of the panel. Even if one full cell is hard shaded, the voltage of

a solar panel drops to half in order to protect itself. When enough cells are hard shaded,

the module will not convert any energy, and eventually become a significant drain of

energy on the entire system over time (Boxwell, 2012).

Figure 2.6 Partially shaded cell/cells (Altestore, 2014)

Partial shading on cell of a 36-cell solar panel will reduce its power output. Due to all

cells are connected in a series string, the weakest cell will bring the others down to its

reduced power level. Therefore, whether half of one cell is shaded, or half a row of cells

is shaded, the power decrease will be the same and proportional to the percentage of

area shaded that is 50 percent.

 27

Figure 2.7: Fully shaded cell/s (GIZ 2014)

When a full cell is shaded, it uses energy produced by the remaining cells, and trigger

the solar panel to protect itself. The solar panel will route the power around that series

string. If one full cell in a series string is shaded, it causes the module to reduce its

power level to half of its full available value (GIZ 2014). If a row of cells at the bottom

of a solar panel is fully shaded, the power output may drop to zero.

In a centralized inverter system, where panels are connected in series, if one of the solar

panels is shaded in an array, the rest of the panels’ output reduces. So, it is important

when choosing a grid connected solar power system that you often prefer the tested and

true technology of a centralized inverter system with pretty good prices. But with

consideration of the effects of shading, however, it’s easy to understand how micro

inverter and Solar Edge systems have become so popular (SolarEdge, 2016).

The use of both Solar Edge and Micro inverter systems allows each solar panel in an

array to maximize power output independently, thereby maximizing a system’s power

generation. If one solar panel is shaded in either of these systems, the rest of the array’s

panels can still operate at full capacity. Solar Edge provides DC to AC power

optimization for each solar panel, while micro inverters provide DC to AC optimization

at the module level. Both of these systems allow solar panels to be facing different

orientations giving you more design flexibility if part of your installation site is in the
 28

shade. A centralized inverter system requires panels to facing the same direction (SMA,

2019).

2.14 Effects of Temperature and Irradiance

The performance of the solar PV array normally depends on the amount of irradiance

reaching the array surface (Jinko Solar, 2018.). Unfortunately, this also causes heating

on the solar cells of the array as the day progresses and affects the output voltage of the

module.

When irradiance reaching the module is reduced, the output current and hence the

power is reduced. Where, current is greatly affected while the voltage is affected

slightly. Extreme high temperatures also reduce the lifespan of the modules by fast wear

and tear. When effects of temperature are not clearly accommodated in the design of a

solar PV array then a temperature derating factor is also applied.

Figure 2.8: Panel I-V Curve of a module at different temperature.

(Source: Jinko Solar, 2018.)
 29

2.15 Other Derating Factors

Table 2.3 gives the impacts on performance of these already discussed and other

derating factors which are controllable. These are aging, that can be controlled by

timely replacement of modules, dust can be reduced by tilting the modules, wiring can

be reduced by using the right size of conductors and mismatch, can be eradicated by

using modules from the same manufacturing standards.

Table 2.3: Derating factors (Rodgers CRC press London, 2010)

Derating Due to Factor percentage

Mismatch 0.98

Efficiency 0.95

Dust 0.97

Shading 0.99

Wiring 0.98

Temperature 0.96

Aging 0.98

2.16 Ways to Optimize a Commercial PV System

Designers of commercial solar are increasingly using system optimization to drive

performance and decrease system costs. As module and inverter prices drop, the other

major opportunity to bring down system cost is by soft-cost reduction through design

optimization. Design optimization improves system performance without increasing

cost, which can actually impact decision on either the project will be economical or not

(Michael La Marca. project engineer, 2015). Because of this, many designing firms are

now focused on developing optimization techniques that can be applied to find the most

cost-effective designs on a site-by-site basis. This is particularly common in

 30

commercial systems given that the system size is large enough to merit optimization -

project price tags reach into the millions of dollars, and yet each location and customer

profile is sufficiently unique to require a customized analysis. With system engineers’

quest for the highest leverage design parameters, these optimization drivers consistently

rise to the top (Paul Grana of Folsom Labs, 2015). The following are the various ways

of optimization:

2.16.1 Module spacing and tilt

Every system designer should have a tradeoff. In the process of maximize the sunlight

on each module, the designer has to tilt for maximum yield and space the modules far

apart. This will reduce the number of modules that can fit in the array

(Rheinlände,2004). Alternatively, the designer can place the modules closer together,

and reduce the tilt in order to minimize shading. This is a tradeoff between energy

density, which is maximum yield per module and power density- maximum kilowatts

per square foot. On this, the industry is moving toward designs with lower tilt and

narrower spacing, sacrificing energy density for improved power density (Horn, 2014).

This move has been driven primarily by the advent of lower-cost modules, well

explained that as hardware costs fall, it becomes efficient to use more modules and

maximize the total generation from a rooftop. With ground mounted system this meant

large land that could be used for agriculture is used (Ferrer-Martí et al., 2012). Also,

new technologies are capitalizing on east-west racking which enables even greater

power density by alternating the module tilt to fit the modules even closer on the rooftop

and maximize the energy yield of the array.

 31

A cascade of simulations can be run and an optimal configuration for each project

selected. For a 2 percent drop in kWh/kWp from reducing tilt and row spacing can

improve the power density by more than 20 percent (Green, 2015).

2.16.2 Azimuth Optimization

Most rooftops are not perfectly south facing. This leaves the system designer with a

choice that must make- either to align the modules to the south to maximize the energy

yield or face them in the direction of the building to maximize the number of modules

that can fit on the roof.

Orientation choice ties in with the biggest themes facing design engineers, this is a

tradeoff between energy productivity versus array power, and the time-of-day profile

of energy production, which can benefit in places with (NREL, 2018). Mostly, this is

not the case involved with the existing centralized solar plants system in Kenya as the

main design concern amount of power being feed to the grid and not the daytime load

profile. Utilities increasingly need later-afternoon production, and changing the array’s

azimuth can be the most cost-effective way to achieve that while with a commercial

building, you can fit a bigger system by going with the building, which almost always

offsets the slight reduction in energy yield that is associated with arrays that do not

perfectly face south (Mayfield 2014).

2.16.3 Inverter design and shade tolerance

The new inverter topologies such as three-phase string inverters, micro-inverters and

optimizers, gives commercial system engineers various options beyond standard central

inverters (Ferrer-Martí et al., 2012). Micro-inverters and three-phase string inverters

change the whole array design, leading to fewer DC wires and more AC wires.

Furthermore, it caused change in the labor mix- these new technologies require more
 32

labor during installation but offer more modularity and easier replacement (SMA,

2015).

Also, with module-level optimization, systems can be designed closer to shade.

Historically System designers eliminated any modules that are shaded. But with module

prices getting down, it currently makes more sense to selectively add some of those

modules back in the system, especially if they are only shaded in the small part of the

day when productivity is low. Also, only recently have new software that enable

detailed shade and mismatch calculations, enabling system designers to rigorously

analyze the losses from shade and assess the system cost-benefit of adding modules.

This is facilitated with switching to string inverters, which does a good job of

minimizing the shading losses on the project (La Marca, 2015).

There is better economics for solar projects designed to enhance a system array in a

finite area vs. designing solely to optimize system efficiency (Schiemann,2014). Even

with a resulting increase in shading, the overall increase in energy production from the

project makes for better economics for customers (IEA and World Bank, 2015).

2.16.4 A new era of system design

Sophisticated solar developers find it easy to embrace optimization techniques. They

are able to deploy lower-cost solar projects without sacrificing quality- with lower

hardware costs. Experienced developers, actually do the analysis and realizes that a

lower tilt and tighter spacing can improve system economics significantly (Mayfield,

2015).

2.17 Simulation of Solar PV Systems

The world market has various solar database and software programs for analyzing solar

photovoltaic (PV) systems, either for commercially, personal or study use (IRENA, 2018).
 33

Solar resource information is required in all stages of the development of a PV project.

Reliable solar radiation statistics is required for system siting, design, and for financing

(IEA, 2015). Mostly, monthly averages, probability statistics of typical meteorological years

(TMY) is sufficient. This information is sufficient also for the manufacturing industry and

for policy makers defining support programs (IRENA, 2018).

Given that factors as solar irradiance, ambient temperature, shading and other derating factors

are never constant, it implies that the output of the solar PV array will be varying as frequently

as the parameter changes (Malerba, 2005). This means that long hand calculation cannot

adequately give an accurate design where all conditions are matched. There are solar PV

design software’s to simplify the solar PV systems model design process like sunny design,

PVSyst and HOMER which was developed by the National Renewable Energy Laboratories

(NREL) of USA (NREL, 2010).

2.18 Intentional Islanding

There are instances under which islanding operation may be required, especially with

the case of a mini-grid being integrated with a central grid that has history of reliability

problems. The mini-grid interconnection may be made in a way that permits the mini-

grid to continue operating autonomously and provide uninterrupted service to local

customers and revenue to the mini-grid operator during outages on the main grid. This

capability is known as intentional islanding.

The recent adoption IEEE standard 1547.4-2011 specifically looked on power systems

that include intentional islanding. Implementing intentional islanding requires that the

system perform various steps, reliably, in correct sequence and timing:

 34

i. The distributed generator must identify an abnormal condition on the utility grid

and disconnect a circuit breaker to separate the generator and islanded mini grid

load from the main grid (Outback, 2019).

ii. On disconnecting, the distributed generator must immediately switch from

“synchronized mode” to “autonomous mode” engaging controls to regulate

frequency (Nelson, 2011).

In addition to the generator configuration, the settings of various protective relays are

needed to be different in islanded mode. Since small generators particularly inverters

typically produce less fault current than large generators on the main grid. Also, voltage

or frequency tolerances need to be broader in island mode. More so, inverter-based

generation may need low-voltage ride-through (LVRT) and frequency ride-through

(FRT) capability to continue operating during voltage disturbances due to faults or

sudden load changes, especially while the grid is transitioning to islanded mode (IEEE,

2015).

The system should continue to sense line voltage on the main grid, and when main grid

power returns to stable conditions, initiate reconnection, and return to control regimes

appropriate for grid-connection (Pilo, 2013).

The protective relay settings controlling intentional islanding must be selected based

on the local grid operating conditions and coordinated with the utility’s protective

relays (Outback,2019). Separating from the main grid due to minor disturbances lead

to lost opportunities for revenue generation from selling power to the utility. On the

other hand remaining on for quite long and only disconnecting at more extreme

disturbances on the grid can lead to cases of excessive voltage and frequency sag hence

brownouts and blackouts with possible equipment damage.

 35

In the same way, settings for condition that constitute to stable conditions on the main

grid for reconnection must consider the timing and effects of re-closers, if they exist on

the main grid’s feeder to which the distributed generator reconnects. When a line is

disconnected after a fault, a re-closer will automatically re-energize the line after a short

delay (Xantrex, 2017). The distributed generator’s controls must:

i. Disconnect the generator fast enough to avoid being on when a recloser

energizes the circuit; and

ii. Wait to resynchronize until grid voltage and frequency are stable (typically a

few cycles).

An intentionally islanded mini-grid faces, all the operational challenges of a standalone

mini-grid with no connection to the grid. Specifically, the mini-grid must meet the

demand and balance local load with local generating resources at all times may

experience significant load growth after connecting to a central grid. This means, the

original generation capacity of the mini grid may no longer be sufficient to supply all

of the load in islanded mode.

This makes it important to have demand-side management strategies ready to

implement on short notice. Normally, automatically opening circuit breakers to curtail

large, non-critical loads or by using smart load-limiting devices on household circuits

that can detect the islanded condition and respond automatically are used. Currently

there is no intentionally islanded mini-grids that face this condition and have

implemented these measures. In principle the same relay that opens the grid intertied

breaker to initiate islanding could simultaneously send a signal to open breakers to large

or non-priority loads (SMA,2015).

 36

2.19 Inverters interconnection requirement

Solar modules produce DC and inverters are an important component of these systems.

For Grid-interactive inverters, the grid controls both frequency and voltage (Victron,

2013). These inverters are designed to export power to the utility grid and incorporate

many of the functions traditionally performed by protective relays, including

synchronization, over/under voltage protection, and frequency protection (Malerba,

2011).

Pumping power to the utility grid requires a grid-interactive inverter. However, most

grid-tie inverters cannot operate without a grid connection and will go off if one is not

available- that is, they do not support intentional islanding (GIZ,2016).

A standalone inverter regulates its frequency and voltage and can operate without a grid

connection. Some standalone inverters allow the grid to be used as a backup for the

renewable generation while others allow the renewable generation to serve as a backup

when the grid is down. However, the inverter cannot be paralleled with the grid such

that either the inverter is providing power to the AC loads, or the grid is providing

power as the inverter is off. These inverters have separate terminals for the grid

connection and for the AC loads (Outback USA, 2019).

Some inverters can operate in both standalone and grid-interactive mode, allowing both

grid export and off-grid (intentionally islanded) operation. These inverters offer the

most flexibility, allowing both intentional islanding and power export. In addition to

the main grid connection, some of these inverters like the SMA Sunny Island allow

other inverters or small induction generators to be connected to the AC load. This

feature allows the construction of AC-coupled mini-grids using a variety of energy

 37

sources. As with standalone inverters, these dual-function inverters generally have

separate terminals for the grid connection and the AC loads (SMA 2012).

Multiple inverters can also be added in parallel on the AC bus to accommodate higher

capacities (SMA 2012). To accommodate multiple-phase loads in larger systems, a

three-phase inverter can be used, or three single-phase inverters can be networked

together with one on each phase (SMA 2012). Also, large, non-critical loads can be

connected separately to the utility grid and not through the inverter.

As of mid-2012, the following inverters sized for the residential market in

industrialized countries were capable of both grid-tie and standalone operation (SMA

2012): Out Back Power GS, GFX, GTFX, and GVFX series; SMA Solar Technology

Sunny Island (SI) series; Schneider Electric XW series; and Princeton Power Grid tied

Inverter and Battery Controller (GTIB) 480-100. Rotating machines (synchronous and

asynchronous generators), which generate relatively pure sine wave output using

rotating magnetic fields, inverters synthesize a sine wave using solid-state electronic

components. If the synthesized waveform is not exactly a true sine wave, current is

introduced at frequencies that are multiples (harmonics) of 50 or 60 Hz, depending on

the local utility frequency (harmonic distortion) (SMA, 2012).

Transformer less inverters may accidentally introduce DC current into the grid, a

process referred to as DC injection. These abnormal currents may damage utility

transformers and other system components and can cause problems for other utility

customers (SMA, 2012). Therefore, grid tie inverters must meet strict requirements for

power quality, including limits on total harmonic distortion and DC injection.

Some small mini grids are entirely DC, with no inverters at all; for example, Mera Gao

Power operates all-DC solar micro grids in 155 villages in Uttar Pradesh, India. These
 38

systems, the renewable energy source produces DC electricity, which is used to charge

a battery bank to which the loads are directly connected (Sungrow, 2018). Customers

must use low power LED lights and appliances specifically designed for low-voltage

DC, with 12 volts the most common system voltage. These DC micro grids are very

small and dissimilar to warrant interconnection with utility AC power. The low voltage

distribution DC system would be difficult to adapt safely to AC power (KPLC, 2018).

In any case, concerns about such systems quickly becoming obsolete may be

unwarranted; considering the high fees paid by households to connect to the main grid,

electrification policies that makes a community electrified even though only a small

fraction of the households connected to the grid, and the intermittency of national grid

power in rural areas, operators of micro grids like Mera Gao Power may find they retain

a sufficient customer base even when the main grid arrives (GIZ, 2015)

2.20 Intentional Islanding Configuration Through AC-coupling

The grid-as-battery is good concept until the grid is no longer there – the grid connected

inverter requires a powered grid to stay per the UL1741 requirement for safety. Without

grid power to keep the GT inverter operating, the available solar power just sits on the

roof unused (Altestore, 2013). During an outage, a home or business with PV electricity

potentially available remains in the dark just like everyone else unconnected to grid as

indicated by Figure 2.9.

 39

Figure 2.9: Anti-islanding of grid tie inverters

(Source: Altestore, 2013.)

Incorporating a battery-based off-grid technology to get a smarter type of grid/hybrid

inverter technology is able to use PV and other renewable DC energy sources to keep

the batteries charged while selling excess power to the grid. Batteries come in as link

to facilitate islanding (Outback USA, 2019). It typically calls for additional load center

with circuit breakers and electrical connections for the building’s critical loads to create

a point at which the grid-tied (GT) inverter and the battery-based (BB) inverter to

couple and share their energy to the loads-power flow shown by Figure 2.10. In the

normal mode of operation with grid power available, energy from the PV array flows

through the GT inverter to the critical load panel, with any excess energy flowing

through the load panel to the BB inverter, and then out to the grid.

Figure 2.10: Current path when grid power is present (Source: Outback USA,
2019).
 40

During outage time, the BB inverter activates an internal transfer switch that opens its

connection to the grid. This avoids the inverter from trying to power other homes on

the grid, and also, keeping energy off the power lines so that utility workers don’t get

electrocuted (Outback USA, 2019). The BB inverter also provides a power source to

the GT inverter that keeps it online and inverting the DC power to AC power for the

critical loads as shown by Figure 2.11.

Figure 2.11: Flow of power uninterrupted to critical loads during outage

(Source: Altestore, 2017)

2.21 Guidelines for Sizing and Interacting Two Types of Inverters

There are two norms that are encouraged for a successful interaction of the two different

inverters;

a) The daily critical load watt-hours shouldn’t exceed 80% of the Battery bank’s watt-

hours (Häberlin, 2012). This assumes that the backup system will only be used a few

days or perhaps a week or two per year, so discharging batteries to an 80% depth will

not significantly reduce their life below what is considered the normal life cycle for a

battery. The rate or speed of which the batteries are both charged and discharged will
 41

affect their overall capacity (Altestore, 2017). The slower the rate of charge or

discharge, the more capacity in the battery.

b) Inverter power rating should be 1.25% of the GT inverter power rating (Altestore,

2017).

This ensures that the GT inverter does not overwhelm the charging circuitry in the

Outback inverter if the load demand goes to zero and all available GT inverter power

is flowing to the Outback inverter. While admittedly an unlikely scenario, for safety

and equipment protection it’s best to follow this guideline. Daily load or battery

charging power must be lower than PV power. This prevents either the daily load

demand or battery charging from exceeds the power from the PV array, or adds an

optional generator to the backup system. In a situation where, available PV power

exceeds load demand, requiring that the GT inverter is disconnected with an Outback

remote operated circuit breaker (Figure 2.12) if the excess power begins to overcharge

the batteries. In reality, the condition in a backup system whereby the available PV

power is out-producing the load and battery charging demand doesn’t occur. Critical

loads will hardly turn off completely and many conditions, especially on cloudy days,

will call for other source of power to meet load and battery charging demand (Outback

USA, 2019).
 42

Figure 2.12: Predesigned consumer board from outback

2.22 Battery sizing

Without a well sized battery backup, the grid connected system will only supply power

to loads when the grid is available. Therefore, it will not solve the fundamental problem

that this research seeks to solve, therefore the battery bank is sized to supply power to

critical load all times - during the day, grid failure as well as in the night.

To ease the battery specification since most batteries are rated at their 10hr, 20hr and

100hr discharge rate; a discharge rate of 20 hours will be most appropriately taken as

the typical discharge rate.

The total energy that must be supplied by the battery bank is determined by the

following equation 2.2;

𝑬_𝑻 = 𝑬_ ⅆ𝒍/𝜼_𝒊𝒏𝒗 (0.1)

Where;
 43

𝐸𝑇 = Total energy in watt hours to be supplied by battery bank during grid

failure.

𝐸ⅆ𝑙 = Total AC energy to be supplied by grid connected PV system. This is

determined from estimated daily critical load as prioritized from load

mapping list and allowance for future load growth.

2.23 Solar PV System Economic Analysis

There are few studies on Solar systems in Kenya considering the economical aspect of

the installation. Therefore, this study was anchored on Kenyan 2030 vision on

sustainable and reliable energy research by the virtue of being one of its a pillar of the

vision.

In a research by the Solar Energy Foundation, the payback time of a small off-grid

system was calculated as a guideline for the investment required installing the PV

system (Breyer et al. 2009). It shows that the payback time of a 10 W PV module was

2 to 4 years depending on the energy usage of the household. Batteries account for 10-

40% of the total capital cost for off-grid PV systems (Häberlin, 2012).

The demand for more efficient and cost-effective storage systems was high and

companies around the world were doing research on solar energy storage solutions.

According to Tesla Motors (2015) announced the release of a newly developed lithium-

ion battery especially adapted for solar power named Power Wall. The batteries come

in two models, 10 kWh weekly cycle and 7 kWh daily cycle, with a price of $ 3 500

and $ 3 000 per unit. The peak power was 3.3 kW. This low price of the new innovation

creates huge opportunities for the PV industry to engage more customers (Tesa, 2015).

Much research concerning evaluation of the use of renewable resources has been

conducted by using life cycle evaluations, such as energy payback time and greenhouse-
 44

gas emission rate (GIZ, 2018). The energy payback is the ratio of energy input

compared to energy output rate. A study on strengthening the case for recycling

photovoltaics, an energy payback analysis showed that the aluminum frame and

mounting rails of the PV system was responsible for the majority of energy payback

time reduction, as they have robust recycling technologies (Goe and Gaustad, 2014).

Materials in the PV cell such as indium, gallium and silicon do not. Although indicators

of sustainability and environmental impact was important, the economic consequences

should also be highlighted since it was of major importance when it comes to

influencing policy- and decision makers to develop new markets.

 45

CHAPTER THREE

METHODOLOGY

3.1 Introduction

The research was carried out through literature review and experimentations with an

ultimate aim of coming up with better system design that harness more solar resource

to generate enough power that satisfy Moi University administration block

economically.

This involved finding of optimal module tilt angle and load determination. It further

carried out numerical calculations and inputs. The PVsyst software was used to

determine optimal system solution.

3.2 Data Collection

The specific data required were site solar insolation, geographical position, and

electrical load. Solar insolation data were obtained from PVGIS and PVsyst by

inputting optimal parameters of design obtained as indicated by equations (3.1) to

(3.15). The information on electrical loads, and rooftop specifications such as tilt and

wall orientation wall were measured.

3.3 Geographical Location

GPS Location of Moi university administration block in GPS is Latitude 0.286°N,

longitude 35.294°E and altitude of 2203M as seen in Figure 3.7. The construction was

such that its longitudinal length is at 21° northern drift from true East (refer Fig 3.6).
 46

Figure 3.1: Location of Moi University Administration in GPS

3.4 Administration Block Grid Supply

From the incoming utility electric supply line, a service ring circuit of transformers if

formed to serve different sections of the university. Administration block is supplied

by 315 KVA TX Pole mounted at the utility transformer at Upper hill hostel.

3.5 Electricity Outages

The challenge of Kenya grid electricity system is the regular occurrence of grid failure

(IRENA, 2016). This regular grid failure affects the work in the university since it

depends on electricity for its production. However, no records on occurrences and

duration of blackouts are kept.

 47

3.6 Shading Analysis

In PV, shading analyzes the shading caused by surrounding objects and vegetation. In

cases like analysis or design of BIPV systems, exact analysis of "shadow-voltaic"

systems as overhangs, vertical shading fins and awnings are also very important in

showing the overall system efficiency in harnessing the sun. Calculations were done

using equations 3.1 to equation 3.3 with dimensions as in Figure 3.2 and figure 3.3

Figure 3.2: Horizontal shading device, overhang, side view

Figure 3.3: Vertical shading device, vertical fin, top view

 48

(𝒉 = (𝑫 ⋅ 𝐭𝐚 𝐧 𝒂 ))/(𝐜𝐨 𝐬(𝝓 − 𝝍) (0. 1)

𝝎 = 𝑫 ⋅ 𝐭𝐚 𝐧(𝝓 − 𝝍) (0. 2)

𝐭𝐚 𝐧 𝒂
𝜸 = 𝒕𝒂𝒏−𝟏 (𝐜𝐨 𝐬(∅−𝝍)) (0. 3)

Where

D - geometry of horizontal shading device (overhang dimension)

α - sun height, Φ - solar azimuth, Ψ - plane azimuth

ꭣ - geometry of vertical shading device (vertical fin)

γ - vertical shadow angle (VSA)

3.7 Calculation of Solar Irradiation

Solar radiation data of various forms; Monthly average values of the radiation incident

on a horizontal surface 𝐺ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 were obtained from PVGIS Website. Simulation

was done using PVSyst software.

To increase the amount of radiation intercepted, modules are often installed at an

inclination angle facing south for locations in northern hemisphere, as seen in Figure

3.4. Tilting the solar array at an angle 𝛽 to the incoming light increases the module

output.

This study utilized PVGIS and PVSyst to simulate monthly and yearly energy output

of designed system in two scenarios, namely:

i. Array with roof’s tilt and orientation performance

ii. Optimized array with optimal site parameters performance

 49

Figure 3.4: Incline surface and resource influencing parameters

(Source: PV Education, 2013).

The amount of solar radiation incident on a tilted module surface 𝐺𝑚𝑜𝑑 was calculated

with help of equations 3.4 to 3.7 (Wenham, 2011)

𝑮𝒉𝒐𝒓𝒊𝒛𝒊𝒏𝒕𝒂𝒍 = 𝑮𝒊𝒏𝒄𝒊ⅆ𝒆𝒏𝒕. 𝑺𝒊𝒏 𝒂 (0. 1)

𝑮𝒎𝒐ⅆ = 𝑮𝒊𝒏𝒄𝒊ⅆ𝒆𝒏𝒕. 𝑺𝒊𝒏 (𝒂 + 𝜷) (0. 2)

Where

𝛼: The elevation angle of the module [°],

𝛽: The tilt angle of the module from the horizontal ground [°],

𝐺ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 : Solar radiation measured on horizontal surface [(𝑘𝑊 ⁄ 𝑚^2 ) ⁄

𝑑𝑎𝑦] and

𝐺𝑖𝑛𝑐𝑖ⅆ𝑒𝑛𝑡 : Solar radiation measured perpendicular to the sun [(𝑘𝑊 ⁄ 𝑚^2 ) ⁄

𝑑𝑎𝑦].

The elevation angle of the site is specified as:

𝛼 = 90 − φ + δ (0.3 )

Where φ is the latitude and δ is the declination angle given as:

𝛿 = 23.45° ⋅ sin [360/365 * (284+d)] (0.4)

 50

where 𝑑 is the day number of the year, 23.45° is the angle of earth axis of rotation from

the ecliptic axis and 284 + 𝑑 is equivalent to 𝑑 − 81 which are the days of the year when

the declination of the sun equals zero.

By using numerical methods in calculating the incoming radiation depending on the

degree of tilt; the optimum angle 𝛽opt can be estimated. This algorithm is based on

finding the angle that gives the optimum solar radiation of each specific month. The

average of all monthly optimum angles is found as the most efficient inclination for a

fixed PV array.

The incident radiation varies during the year since the module is set at a fixed angle.

The average annual solar radiation incident on the tilted angle 𝛽opt was calculated as in

equation 3.8
𝒏
𝑮𝒎𝒐ⅆ,𝒊
𝑮𝒐𝒑𝒕 = ∑ (0. 5)
𝒊=𝟏 𝒏

Where:

𝐺opt: the average solar radiation at the specific month [kW/m2/day] and

𝑛: the number of the month where 𝑛 = 1 is January and 𝑛 = 12 is December.

The number of daylight hours 𝑁 is given by (Duffie and Beckman, 2006).

𝟐
𝑵= 𝒄𝒐𝒔−𝟏 (−𝒕𝒂𝒏∅𝒕𝒂𝒏𝖉) (0. 6)
𝟏𝟓

This is an important correlation for estimation of the sunlight hours. The above

equations will give all underlying data needed to calculate the potential solar energy

given at the site.

 51

3.8 Experiment to Obtain Optimal Tilt Angle

The experiment was carried out with four setup each comprising of 50 W rated solar

panel powering one 7W bulb as a load and two multi meters-(one measuring voltage

while the other one current). Each solar panel on each setup was mounted at Azimuth

angle corresponding to south fazing roof of block but different tilt angle i.e. 5°, 15°,

25°, 35° and last one set at block roof angle. Timer was used to facilitate simultaneous

Voltages and current readings at the same irradiance while recording. This was

conducted on a sunny day of 15th August 2019. The experiment results were not

dependent on date and time but on sensibility of power with respect to tilt.as seen in

Figure 3.5.

Figure 3.5: Tilt angle experiment

3.9 Investigation of Offices Load

For a sustainable and economical design of the grid connected PV system, data from

two sources were collected. The first being electrical bill data from the administration

block. Secondly, mapping and collection of daily watt-hour compilation of all

 52

individual electrical load as per manufacturers specifications. This facilitated the size

determination of the battery-based inverter as well as battery bank size.

3.10 Roof Top Tilt and Orientation

The orientation and tilted angle of the roof was determined with the help of scan the

earth software installed in a phone., The tilt and orientation were recorded when the

phone was kept on parallel to the roof.

Figure 3.6: Orientation of north facing roof as captured

With the orientation of the North facing roof, the south facing roof’s orientation was

obvious given that it is parallel and opposite to northern facing roof.

 53

Figure 3.7: Moi Administration admin coordinate

3.11 Dimensioning of the PV System

The dimensioning involved the following:

3.11.1 Module dimensions

The PV modules were incorporated to match the load power demand. Not only the

irradiated energy of the site needed to be known, but also the efficiency of the module

selected for the installation. Once the load and the absorbed radiation of the site was

established, the PV module area to meet the load was determined through an energy

balance. A PV module generates the energy, 𝐸pv, obtained by equation 3.10

𝑬𝒑𝒗 = 𝑨𝑪 . 𝜼𝑺𝒀𝑺 . 𝑮𝒎𝒐ⅆ (0. 1)

Where:

𝐴c: the surface area of the module [m2],

𝜂sys: the total system efficiency [%] and

𝐺mod: the incident solar radiation on the tilted array [kW/m 2/day].
 54

This energy can either be used directly to supply a primary load, as can be seen in Figure

2.11 or be accumulated in a backup battery for deferred utilization.

The total system efficiency 𝜂sys considers all losses from converting solar energy into

direct current electricity and then transform it into alternating current electricity, thus,

it involves all the deficiencies of the components in a PV system. This ratio is given by

equation 3.10 (Antony et al, 2007).

𝜼𝒔𝒚𝒔 = 𝜼𝒑𝒗 . 𝜼𝒑𝒗−𝒃𝒂𝒕𝒕 . 𝜼𝒄𝒄 . 𝜼𝒃𝒂𝒕𝒕. 𝜼ⅆ𝒊𝒔𝒕 . 𝜼𝒊𝒏𝒗 (0. 2)

Where:

𝜂pv: efficiency of the PV modules,

𝜂pv-batt: efficiency due to voltage drop in cables,

𝜂cc: efficiency of the charge controller,

𝜂batt: efficiency of the batteries,

𝜂cable: efficiency of the distribution cables from PV battery to loads and

𝜂inv: efficiency of the inverter.

The universal index used for comparing the efficiency of solar cells from different

vendors is the peak power 𝑊p, which is estimated under Standard Test Conditions

(STC). STC is equivalent with an irradiance 𝐺STC of 1000 W/m2 and a temperature 𝑇STC

of 25°C. The peak power of a solar cell has a direct correlation to its efficiency 𝜂PV,

which is defined as the ratio of generated power to incident solar energy (Häberlin,

2012)
𝒘𝒑
𝜼𝒑𝑽 = 𝑮 (0. 3)
𝒔𝑻𝑪 ⋅𝑨𝑪

An inverter was required to upturn DC power into AC power with a desired output

voltage for utilization of conventional electrical appliances that demanded AC power.

 55

A converter can be used instead since it can convert AC to DC and vice versa depending

on the direction of power flow.

The inverter provides for low battery cut-out and cut-in operation (Messenger and

Ventre, 2010). The selected inverter was needed to meet the requirements of the load

as well as monitor the battery voltage.

After choosing inverter, the string size (number of modules in series) for the PV

modules has to be determined. The modules can be connected either in series and/or in

parallel before they are connected to the inverter.

The size of the string determines the amperage and the voltage that will be going into

the inverter. The maximum number of modules allowed to be connected in series as

given in equation 3.13 (Messenger and Ventre, 2010)

𝐕𝐢 𝐦 𝒂𝒙
𝑴𝒎𝒂𝒙 = 𝜟𝑽 𝒄 (0. 4)
𝑽𝒐𝒄 +(𝑻 𝒎𝒊𝒏−𝑻 𝑺𝑻𝑪)⋅ 𝟎 ⁄𝜟𝑻

And the minimum number of modules tolerable to be connected in series is

𝐕 𝐢 𝒎𝒊𝒏
𝑴𝒎𝒊𝒏 = 𝜟𝑽 (0. 5)
𝑽𝒎𝒑 +(𝑻 𝒎𝒂𝒙−𝑻 𝑺𝑻𝑪)⋅ 𝒐𝒄⁄𝜟𝑻

Where:

𝑉i, max: maximum inverter voltage [V],

𝑉i, min: minimum inverter voltage [V],

𝑉m p: maximum power voltage [V],

𝑉o c: open circuit voltage [V],

Δ𝑉o c/Δ𝑇: temperature coefficient of 𝑉oc [V/K] – Technical data for 295 W Jinko

PV module,

𝑇max: maximum monthly averaged temperature of Kenya (23.0°C) and

 56

𝑇min: minimum monthly averaged temperature of Kenya (12°C) according to

world weather online.com.

The total number of modules has to be equal or exceed the wattage of the inverter as

well as match the number of modules needed to meet the system load. The final

combination also has to match the measured location area for the placement of the PV

modules.

Equations (3.10) to (3.14) gives the number of modules required to meet the load and

was simulated in PVSyst. Such a system can only supply power at the same time when

the solar resource is present, meaning that it can function to supply critical offices

during power outages.

The administration block load from the KPLC bills was accurate since previous meter

reading was subtracted from current meter reading, thus, actual. The consumption and

separation of high rate and low rate depicts time and night utilization respectively. With

just two bills of July, 2018 and June, 2019-meter reading gave actual consumption.

The number of PV modules was found by determining the number of components

needed to equal the same load during the day and battery backup that can supply night

lighting with a few the critical loads of offices during power outages.

3.12 Battery Capacity Determination

The assessment of the energy required during grid failure is based on its nature and how

the system operates during failure. Given that there was no data of grid failure at the

university and the fact that the system is expected to supply power to some of the loads

during failure, the energy needed during failure was typically fraction of total energy.
 57

The crucial equipment that allows minimal operation during outage were considered

for backup as in Figure 3.8. Every office will be backed with one 4 feet fluorescent and

a desktop. Also, ethernet network is need to be available all the times. Due to the quality

of Trojan batteries from USA this research used them in the design. The Trojan battery

in Figure 3.8, is manufactured in USA was used. Its specifications are given in Figure

4.8.

Figure 3.8: Trojan battery

Batteries used in all solar systems are sized in Ampere hours under standard test

conditions (Temp: 250C). The depth of discharge is a measure of how much of the total

battery capacity has been consumed. For most batteries’ manufacturers recommend the

maximum allowable depth of discharge is 0.5 (50%) - 0.7 (70%) in order to get more

cycles which, translate to lifespan.

The battery bank capacity was given by equation 3.15

C x = G ft / V dc × E tot / DOD max (0.1)

Where

Cx = battery capacity, for a specified discharge rate in ampere hours.

 58

Etot = total energy in watt hours to be supplied by battery bank during grid

failure

Gft = the number of days the battery bank needs supply during grid failure.

DODmax = design maximum depth of discharge

3.13 Equipment’s

The descriptions of equipment used were as follows:

3.13.1 Solar modules

Jinko panels model JKMS295M-60V Maxim were used for the design and their features

are elaborated in table 3.1

Table 3.1: Characteristic of panels

Description Specifications
Weight 41.90 lbs
Solar Cells Monocrystalline PERC
3.2mm, Anti-Reflection Coating, High
Front Glass
Transmission, Low Iron, Tempered Glass
Frame Anodized Aluminum Alloy (Black)
J-Box IP67 Rated
Output Cables TUV 1x4.0mm², Length: 900mm or Customized

Table 3.2: Electrical Characteristics

Maximum Power (Pmax) 295 Watts
Maximum Power Voltage (Vmp) 32.40 Volts
Maximum Power Current (Imp) 9.10 Amps
Open-Circuit Voltage (Voc 39.70 Volts
Short-Circuit Current (Isc) 9.61 Amps
Module Efficiency STC 0.1802
Operating Temperature -40°C to +85°C
Maximum System Voltage 1000VDC (IEC)
Maximum Series Fuse Rating 15 Amps
Power Tolerance 0~+3%
 59

Figure 3.9: JKMS295M-60V Maxim Dimension

3.13.2 Inverter

SMA inverter from Germany- Model Sunny Tripower 25000TL-JP-30 was used as it has

a well distribution- function that provide voltage regulation by controlling reactive

power flow in the circuit. This facilitated renewable penetration even with their

intermittent nature characterized with rapid, large and random fluctuation in supply

(Ieeexpore, 2019)
 60

Figure 3.10: Inverter

Figure 3.11: Efficiency Curve Graph (Source: SMA solar, 2015)

3.13.3 PVGIS Website

The Photovoltaic Geographical Information System (PVGIS) provides web access to:

 solar radiation and temperature data

 PV performance assessment tools

 61

3.14 STC Standard Test Conditions of Panels

Usually, when the performance of the solar module is tested under standard conditions

as;

 Irradiance 1000 W/𝑚2

 Module temperature 250 C

 Spectrum AM 1.5

3.15 AC Connection to Sub-panel

When a site service contains more than one panel board, the panels fed from the main

service panel are referred to as subpanels. The NEC, in 705.12(D) and NEC 690.64(B)

in the 2008 NEC, allows the inverter OCPD to be connected at any location in the

premises wiring system, provided that the 120% of busbar and conductor ampacity

limitation was observed.

Now consider the current flow at the main service panel, the 2011 NEC requires

installers to calculate the sum of the supply OCPDs at the main service panel based on

the rating of inverter OCPD, like in the case of 200A service board a 40-amps and not

the 200-amp feeder breaker that feeds the subpanel

3.16 Sensitivity Analysis of the Investment

The research analyzed the system cost and benefit of conventional on-grid power

system components, off-grid solar systems and hybrid of solar systems. For

quantification and analysis, power production and consumption were kept equivalent

for all the three energy systems. The model enabled end customers to make a decision

based on system parameters and its corresponding results in terms of power savings.

This model is a conglomeration of solar system studies and its direct impact on end

customers on an annual basis.

 62

For more precision cost benefit model, it used real time data such as current energy

prices for residential systems, solar system costs from different agencies, costs of power

components, construction and labor costs from solar and roofing contractors, carbon

emission reduction data through different conventional sources.

PVSyst calculates the cost of the investment and will find the optimal system to meet

the load. The two most important setups were evaluation of a sensitivity analysis by

varying important parameters such as PV capital cost, and grid electricity price.

Once the optimal system was found, its’ payback period was calculated with respect to

the accumulative development of the national grid electricity price in Kenya.

 63

CHAPTER FOUR

RESULTS AND DISCUSSION

4.1 Introduction

This chapter presents results and discussions of various findings of the study. It starts

from sizing the system that meets administration block load and simulation of this

system on existing site parameters.

4.2 Site Sun Path

Figure 4.1: Solar path at Kesses -Moi University

4.3 Average Sun hours

The sun-hour varies throughout the year. For Grid connected system with no net

metering and with high FiT values, average monthly sun-hours are used in sizing unlike

in off-grid system where the lowest monthly average value is used.

 64

Table 4.1: Moi administration block irradiation table

Jan Feb Mar. Apr May Jun Jul Aug Sept Oct Nov Dec Year
Hor. Global 6.29 6.39 6.18 5.55 5.2 5 4.4 4.54 5.77 5.75 5.54 6.2 5.564
Amb.
17.5 18.5 18.5 17.8 17 16 16 15.8 16.4 17.3 17.1 17 17.17
Temperature
Wind
4 4.1 3.9 3.3 2.8 2.6 2.5 2.5 2.8 3.5 3.9 3.9 3.317
Velocity

Worth noting is that wind velocity is directly proportional with Global irradiation as

shown in Figure 4.2.

Figure 4.1: Global irradiance, temperature and wind velocity curves

4.4 Administration Block Roof Analysis

Figure 4.1: Roof Analysis Administration block

The main width was found to be 18𝑚 while the roof forming slopes of was 170 from

the opposite walls, resulting to two faces of the roof; one facing south at an azimuth of

210 and the other facing North with an azimuth angle of 1590 (Hypotenuse) as captured

by Google campus (Figure 3.6). Width of the roof =W roof = 9/cosine 17°= 9.41m
 65

Figure 4.2: Administration Block Moi University

Figure 4.3: Front view of Administration Block

Figure 4.4: Aerial view of Administration Block

The available longitudinal length of the North/South facing roof to lay the panels

without considering the shading effects of the protruding sections of roof was

= 136𝑀 – (18 + 2 𝑋 16) = 86𝑚

 66

Total area on either side of the roof (North/South) was

𝐴 = 86 𝑥 9.41 = 809.26𝑚2

4.5 Sizing the Solar Plant

Taking high-rate meter reading as captured in monthly electric bill of 1st July– 2nd

August, 2018 (Table 4.2) - previous meter reading of July 2018 bill and also previous

meter reading in 1st July- 2nd August 2019 (Table 4.3) gave a twelve months actual

consumption.
 67

Table 4.1: KPLC Bill 02/07/2018-01/08/2018

Previous Current Reading Consumption
Meter No. Consumption
reading reading Type type

040016113631 0 96 Real 96 Demand KVA

040016113631 0 94 Real 94 Demand KW

040016113631 563873 582889 Real 19016 High rate

040016113631 368516 378147 Real 9631 Low rate

Table 4.2: KPLC Bill 02/07/2019-01/08/2019

Meter No. Previous Current Reading Consumption Consumption
reading reading Type type

040016113631 0 99 Real 99 Demand kVA

040016113631 0 97 Real 97 Demand kW

040016113631 800803 823052 Real 22249 High rate

040016113631 506052 519062 Real 13010 Low rate

Actual consumption was

800803 − 563873 = 236930 𝑘𝑊ℎ/𝑦𝑒𝑎𝑟

236930/12 = 19744.17𝑘𝑊ℎ/𝑚𝑜𝑛𝑡ℎ

19744.17/30 = 658.14𝑘𝑊ℎ/𝑑𝑎𝑦

658.14/5.6 = 117.53𝑘𝑊
 68

4.6 Electrical Load Mapping

Electrical power rating of the equipment, was presented as in Table 4.4.

Table 4.1: Mapped Appliance Table

Item Description Qty Wattage Total watts Time of Watt-Hrs
Usage (Hrs)
Switches 1 480 480 2 960
Shredder 1 240 240 2 480
Scanner 11 15 165 2 330
Printers 1 696 696 2 1392
Coil 1 2400 2400 1 2400
Conditioning fan 1 134.4 134.4 2 268.8
Cooker 5 1200 6000 2 12000
Decoder 2 10 20 8 160
Desktop 218 200 43600 8 348800
Dvd player 1 8 8 8 64
Electric coil 3 1200 3600 2 7200
Kettle 12 2200 26400 2 52800
Fan 3 134.4 403.2 2 806.4
Fax machine 3 19.2 57.6 5 288
Fridge 4 100 400 15 6000
Internet switch 5 66 330 24 7920
Laptop 37 50 1850 6 11100
Microwave 9 1200 10800 2 21600
Photocopier 15 1224 18360 2 36720
Printers 106 480 50880 2 101760
Projector 2 312 624 2 1248
Tv 2 94 188 8 1504
Ups 20 528 10560 8 84480
Dispenser 7 550 3850 6 23100
Water heater 3 220 660 1 660
182706.2 724041
Grand Total
 69

Table 4.2: Mapped lighting points table

Item Total Total use hrs Total Watt

Item Qty
Load watts per day Hours

2 FT Tube lights 18 97 1,746 8 13,968

4 FT Tube lights 36 1,353 48,708 8 389,664

5 FT Tube lights 58 19 1,102 8 8,816

CFL LIGHTS 10 13 130 8 1,040

Total 55,230 413,488

From the physical mapping of electrical equipment and their consumption. The sum

watt-hours of a day are divided by average daily sun-hours

Daily consumption = Appliances + lighting

724041.2 + 413488.0 = 1137529.2 𝑊ℎ/𝑑𝑎𝑦

Then, the system size will be

1137529.2/5.6 = 203.13𝑘𝑊

The value of 203.13 kW was ignored since it was far from the accurate. The reason was

that during audit, the operation time of lights and appliances are purely guessed work.

Some equipment may have ceased being used and yet to be faced out as seen in Figure

4.7.
 70

Figure 4.1: Unused equipment stilled hooked to power

This method is only good for small housed hold system where exaggeration will not

impact so much on the cost of the system (Hawkins, 2010).

For the case of this study, the mapping data obtained was used to size battery bank for

critical loads - 66240Wh. Its accuracy was found to be high since the end user usually

dictate backup time depending their experiences on duration of discomfort caused by

outage.

From the KPLC bill the highest peak power was 99Kw. It was found that, the

economical and sustainable size system for administration block was 100kW. This is

because the peak load of administration block is 99kVa which is definitely during the

day when most activities is on. Also, analysis of KPLC bill resulted to 117kWh per day

and that net metering bill is not enacted, the system being small compared to set

threshold FiT value of 500KWp and export of excess power meant extra bill to the

university.
 71

4.7 Critical Load Battery Backup

From the physical mapped load list, this research picked crucial loads that was in the

offices taking into consideration of existing 235 offices in administration block. These

were internet server and network switches, a desktop and two 4ft tube florescent tubes

in a third the number of offices in administration block.

Table 4.1: Critical load table

Item Total Total use hrs Total Watt

Item Qty
Load watts per day Hours
Desktop 120 75 9,000 4 36,000
4 FT Tube lights 36 150 5,400 2 10,800
switches 66 5 330 24 7,920
Server 480 1 480 24 11,520
Total 15,210 66,240

Size of Battery bank at nominal DC voltage of 48V and 65% depth of discharge

66240 Wh
= 2,123 Ah
48 × 0.65

Battery base inverter of 15KVA at 48V was enough to serve critical load service board.

This research used 4V Trojan 2145Ah batteries. For a 48V nominal voltage system, it

took 12 Trojan batteries all connected in series.

On off-grid system of such size of critical load as indicated by Table 4.6, 65% Depth

of discharge would have at least seven years lifespan with 2600 life cycle for daily

battery use- Figure 4.8. But, for this case where batteries are used as backup, 2600 life

cycle represent quite long-life span given that cycles are utilized only when utility grid

is down.
 72

Figure 4.1: DOD vs Cycle Life Curve Graph

Table 4.2: Cost of Incorporating Battery Back up
ITEM DISCRIPTION EXPECTED
QTY Unit price DISC VAT
PRICE
4V 2145 AH Trojan
12 167,500.00 15% 0% 1,708,500.00
BATTERY

6KVA INV/CHG 3 227,000.00 15% 0% 578,850.00

AVS & change over 2 14,500.00 0% 16% 33,640.00

Installation Materials 1 32,200.00 0% 16% 37,352.00

Labour 1 34,700.00 0% 16% 40,252.00

GRAND TOTAL VAT 2,398,594.00

Table 4.7 is the current market price of doing battery backup systems of 12 pieces of
4V 2145Ah batteries
 73

4.8 Plant Layout

The total rating of the system was100KW taking 360 pieces of Jinko panels model

JKMS295M-60V Maxim shall cover a total rooftop area of 589 m2. With administration

block’s longitudinal length of 135m, larger roof surfaces facing north and south was

used for simulation. The plant uses a total of 360 solar panels and 4 three phase inverters

with rating of 25 KW - model Tri power 25000TL-JP-30 as in Figure 4.9. These

inverters were used for DC to AC conversion and the output was fed to the 11 KV grid.
 74

Figure 4.1: Plant layout

The array dimensions and space required was computed as follows:

Length = 0.996 × 18 = 17.928 m

𝑊𝑖𝑑𝑡ℎ = 1.657 ∗ 5 = 8.285𝑚

 75

The inverter manufacturers specification that should not be exceeded by pant

parameters is shown in table 4.8.

Table 4.1: Inverter Dc Input parameters verses designed system parameters

Inverter dc input Parameter Layout/Array parameters

Mpp voltage Range/Rated input Voltage: No. of panels in series X Voc

390V to 800V/600V
18 x 39.70 = 714.60V

Maximum input current: 33A No. of strings in an array Y I sc

5 x 9.61 = 48.05A

No. of independent MPPT inputs 2

Min input Voltage/Start input Voltage: 150V/188V

150V/188V

4.9 PVSYST System Simulation

This research looked into the tilt and orientation of the two main longitudinal faces of
the roof – south and north. It simulated energy production from each independently and
further investigated the possibility of energy enhancement. Results and Analysis

PVsyst allows the user to import the metrological data from various sources and analyze

grid connected solar systems depending on the specifications of the system and

characteristics of its components such as PV module, inverters and others components.

The experimented tilt angles at existing roof orientations helped to point out the existing

relationship between the latitude of a location and the optimal tilt angle to harness the

sun energy.
 76

Table 4.1: Experiment results table

Tilt angle° 1st 2nd 3rd 1st 2nd 3rd AVERAGE

Current Current Current Voltage Voltage Voltage POWER

reading reading reading reading reading reading

5 0.452 0.442 0.436 20.178 20.071 20.037 8.909

15 0.596 0.575 0.593 20.525 20.359 20.552 12.042

25 0.460 0.450 0.456 20.345 20.112 20.248 9.214

35 0.447 0.446 0.448 20.085 19.870 19.933 8.923

Plotting these values in Table 4.9 against tilt angles resulted to curve as seen in Figure

4.10.

AVG Power
14.000
12.000
10.000
8.000
6.000
4.000
2.000
0.000
0 5 10 15 20 25 30 35 40

Figure 4.1: Average power vs Angle curve

The results obtained from simulations of varying scenarios are given in Figure 4.11 to

figure 4.18.
 77

Figure 4.2: Diagram of the south facing roof surfaces

Figure 4.3: Normalized production Bar graph

For orientation tilt 17, azimuth -21 in figure 4.11, transposition factor is 0.99 and loss

with respect to the optimum is -0.7. This mean ratio of incident to horizontal is close to

one but there are minimal losses incurred in harnessing the sun on the panel due to its

orientation. The system’s production is at 259.2 mWh annually.

 78

Figure 4.4: Optimized tilt of the south facing roof

Figure 4.5: Normalized productions bar graph

For orientation tilt 100, azimuth -210 in figure 4.13, transposition factor is 1 and loss

with respect to the optimum is 0.0%. This mean ratio of incident to horizontal is one

and zero gain/losses incurred in harnessing the sun on the panel due to its orientation.

The system’s production of 263.9 mWh annually.

 79

Figure 4.6: The North facing roof surfaces

Figure 4.7: Normalized production Bar graph 3

For orientation tilt 170, azimuth 1590 in figure 4.15, transposition factor is 0.98 and -

2.4% loss with respect to the optimum. This mean ratio of incident to horizontal is less

than one and significant losses incurred in harnessing the sun on the panel due to its

orientation. The system’s production of 258.6 mWh annually.

 80

Figure 4.8: Optimized tilt of the south facing roof

 81

Figure 4.9: Normalized Production Bar graph 4

For orientation tilt 30, azimuth 1590 in figure 4.17, transposition factor is 1 and 0.0%

loss with respect to the optimum. This mean ratio of incident to horizontal is one and

no losses incurred in harnessing the sun on the panel due to its orientation. The system’s

production is 265.2 mWh annually.

Table 4.2: Summary table of various simulation

Glob System Normalized pv loss

Tilt° Azimuth° on Prod Specific Prod Performance Prod kWh/kWp
Coll mWh/yr kWh/kWp/yr Ratio kWh/kWp/day /day
3 159 2026 265.2 2497 0.835 6.84 1.20

10 -21 2036 263.9 2484 0.837 6.81 1.18

17 -21 2013 259.2 2441 0.838 6.69 1.14

17 159 2099 258.6 2435 0.834 6.67 1.16

On analysis of the values in Table 4.10 it is certain that the normalized energy from the

optimized 3-degree tilt angle (6.84 kWh / kWp / day) is more than the rest which get
 82

lower to the lowest 6.67 of 17-degree tilt angle. PV- array losses is also higher for this

tilt angle just as yearly system production and specific production.

Therefore, the suitable orientation for solar energy harnessing at Moi university

administration was obtained as Tilt angle of 3° and Azimuth angle of 159°.

Table 4.3: Simulation data of 100KW plant with tilt 3° and azimuth 159°

GLOBLE GLOBLE
PRODUCTION
HORIZONTAL INCIDENCE E-GRID MWH
RATIO
kWh/m2 kWh/ m2
JAN 254.5 248.2 21.96 0.833

FEB 237.6 234.0 20.52 0.826

MAR 266.6 266.0 23.26 0.823

APR 250.9 253.8 22.36 0.830

MAY 245.4 251.1 22.33 0.838

JUN 228.9 235.4 21.18 0.847

JUL 239.3 245.4 22.09 0.847

AUG 251.6 255.6 22.86 0.842

SEP 254.1 254.9 22.57 0.834

OCT 263.7 260.8 22.99 0.830

NOV 248.2 242.5 21.50 0.835

DEC 250.5 243.3 21.59 0.836

TOTALS 2991.3 2991 265.21 0.835

 83

Syst Prod
23.5
23
22.5
22
21.5
21
20.5
20
19.5
19
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Figure 4.10: Yearly production bar graph

Figure 4.11: Performance Ratio PR Bar Graph

 84

Monthly GlobHor & GlobInc

600

500

400

300

200

100

0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

GLOBHOR KWH/M2 GLOBINC KWH/M2

Figure 4.12: Monthly Global Ho &Glob Inc kWh/m2

 85

Figure 4.13: Loss Diagram

From the loss in Figure 4.23 the effective irradiation on the collector plane was

2903𝑘𝑊ℎ/𝑚^2 and the global horizontal irradiance was 2991𝑘𝑊ℎ/𝑚^2 . The solar

radiation incident on the solar panels will convert into electrical energy. On PV

conversion, the nominal array energy was 308 mWh, while the PV array efficiency was

found to be 18.02% at standard test condition (STC) (Figure 4.23). Array virtual energy

obtained was 271.1 mWh. After the inverter losses the available energy obtained at the

inverter output was 265.2 mWh. The importance of doing simulation is to have optimal
 86

orientation where the modules are well exposed to the sun and rays are harness with

minimal losses so as to obtain more solar energy conversion to electrical energy.

4.10 Cost of the System without Batteries

Taking the current prevailing market prices, the total system cost without batteries was

as in the table 4.12. The total cost formed the basis of economic analysis. The costing

assumed there is a dedicated room within administration bock to house the inverters

and other balance of system- without batteries.

Table 4.1: Prices based on Chloride Exide – major solar distributor in Kenya
PRICE PRICE
Unit price
ITEM DESCRIPTION QTY VAT EXCL VAT INCL VAT
(Kshs.)
VAT(Kshs.) (Kshs.)
SOLAR MODULE C-SI
360 16,994.20 0% 6,117,910.24 0.00 6,117,910.24
295 W

25kW GRID-TIE
4 470,159.76 0% 1,880,639.02 0.00 1,880,639.02
INVERTER

AC DISTRIBUTION BOX 1 89,204.55 16% 89,204.55 14,272.73 103,477.27

COMMUNICATION
1 73,568.18 16% 73,568.18 11,770.91 85,339.09
EQUIPMENT

DC CABLE (4 mm2) 540 127.50 16% 68,850.00 11,016.00 79,866.00

AC CABLE (10 mm2) 270 293.25 16% 79,177.50 12,668.40 91,845.90

PANEL MOUNT - ROOF

1 997,159.09 16% 997,159.09 159,545.45 1,156,704.55
MOUNT

REVERSE CURRENT
1 850,777.78 16% 850,777.78 136,124.44 986,902.22
PROTECTION

ELECTRICAL
1 560,713.79 16% 560,713.79 89,714.21 650,428.00
INTEGRATION GEAR

INSTALLATION
1 713,438.63 16% 713,438.63 114,150.18 827,588.81
LABOUR

GRAND TOTAL 11,431,438.78 549,262.32 11,980,701.10

 87

4.11 Cost of Electricity

According to Moi university electric bill for the month of July, 2019; the total monthly

cost was Ksh. 755 960.09. This was after a consumption of total units of 35259Kwh in

both high and low rating. The main part of the bill that solar system will substitute was

the high rating of the bill, A unit kwh cost was obtained as shown in Table 4.13:

Table 4.1: KPLC Cost of electricity

Energy

High-rate cons 22249KWh X 2 266,988.00

low-rate cons 11399kwh x 12 136,788.00

low-rate cons 1611kwh x 6 9,666.00

max Demand KVA 99kva x 6 79,200.00

Fuel Energy Cost 35259kwh x 3.67 129,400.53

Total Energy 622,042.53

Levies and adjustments

Forex exchange 35259kwh x 0.0452 1,593.71

Inflation 35259kwh x 0.25 8,814.75

ERC
35259kwh x 0.03 1,057.77
LEVY

REP
413442 x 5 20,672.10
Levy

Warma Levy 35259kwh x 0.0169 595.88

Total levies and Adjustments 32,734.21

Rounding Adjustment 0.10

VAT 101,192.16

Total monthly bill 755,969.00

KWh Price = 1(12+3.67+0.0452+0.25+0.03+0.0169+.16) Kenya shillings

= Ksh. 16.17/kwh
 88

Therefore, using this rate, the yearly saving obtained from the simulated yearly system

production before introducing batteries to enhance self-consumption was as shown in

the table 4.14.

Table 4.2: System Production

Month Syst Prod MWh Unit Cost Saving (Kshs)

Jan 21.96 16.17 355,093.20

Feb 20.52 16.17 331,808.40

Mar 23.26 16.17 376,114.20

Apr 22.36 16.17 361,561.20

May 22.33 16.17 361,076.10

Jun 21.18 16.17 342,480.60

Jul 22.09 16.17 357,195.30

Aug 22.86 16.17 369,646.20

Sep 22.57 16.17 364,956.90

Oct 22.99 16.17 371,748.30

Nov 21.5 16.17 347,655.00

Dec 21.59 16.17 349,110.30

Totals 265.21 16.17 4,288,445.70

The total cost of the system after incorporating battery backup was obtained as Kshs.

14,379,295.10. The break-even point was

11980701.10 + 2398595.00 = 14379295.10 shillings

14379295.10/4288445.70 = 3.35 years

 89

4.12 Return On Investment (RoI) of the system without batteries

ROI= 11,980,701.10/ 4288445.70

= 2.793 roughly 3 Years

This calculation assumes the university has already existing estates and maintenance

team capable of doing maintenance without any extra cost. The reduction in ROI time

compared to the case with batteries is small because the battery backup load is just

Critical loads that need to run during outage times.

Supposed the whole administration load were backed up by batteries, the cost of the

system would be significantly large to give economic justification. Therefore, backing

up of only critical loads improve reliability of grid connected system as well as

controlling system cost to economic viability region.

4.13 Sensitivity Analysis

The study further studied the relation of array orientation to system production and

return on investment time as shown in table 4.15

Table 4.1: Sensitivity of ROL to array orientation

Tilt System System production market Scenario’s ROL- With

Azimuth°
(°) Prod mWh/yr Price (Ksh) batteries(Yrs)

3 159 265.2 4,288,284 3.353

10 -21 263.9 4,267,263 3.370

17 -21 259.2 4,191,264 3.431

17 159 258.6 4,181,562 3.439

 90

CHAPTER FIVE

CONCLUSION AND RECOMMENDATION

5.1 Conclusion

This study has ascertained adequate solar resource in Moi University. If harnessed with

well-designed systems, they will meet the demand of intended facility, hence; reduce

dependence on the grid electricity. The ROI difference is small between systems with

and without batteries. This favors battery incorporation and the fact that batteries will

last longer when used as backup than when used on daily basis as in off-grid scenario.

Moi university administration block being operational for 8 hours and mainly on

daytime duration of the day, have the potential of leveraging on this adequate solar

resource. This will minimize its reliance on fault prone grid and exploitation of

nonrenewable resources such as fossils fuels.

With grid connected PV system, batteries are not mandatory. They only serve as backup

for major critical load during outage. This means a considerable small battery bank

whose cost is significantly small compared to backing up whole building loads.

Moreover, in this case batteries are only used in the event of power outage. Therefore,

the useful lifecycles of batteries are reserved for outage period- the batteries last longer

than in the case of off grid system.

The four scenarios of simulation yielded different values with highest being 265.2Mw

while the lowest at 258.6 Mw per year, with a range of 6.6 MW in a year for a 100kw

system. This shows that with proper installation at optimal site parameter, one can

enhance production with a significant value than can positively impact on economic

feasibility of the system as well as shortening the breakeven duration.

 91

Current technologies have enhanced production of solar system. The likes of

centralized inverter with a number of MPPT capabilities have allowed utilization of

space that even get shaded in part some part of the day. Some OEM recognizes

developing countries without Net metering and have developed functionalities in

inverter that limits export of power to grid or gadget (limiter) that facilitate grid

connected inverters with normal one directional meter.

5.2 Recommendation

This study recommends a grid connected solar system not just for administration block

but to whole university to leverage on adequate solar resource in order save on electric

bills, for university to serve as a forerunner in encouraging adoption of solar energy and

attract a clique of solar professions.

Thoroughly analysis of site and optimal parameters is paramount for decision making

otherwise wrong decision would be made if inadequate research is done on a site as

well has impacting on the grid substitute value as well as pay-back period of the system.

The cost addition on grid connected system due to precaution being put in place to avoid

export of power to the grid in term of power limiter can be done away if the net metering

bill is enacted. Also, the threshold capability for FiT should be lowered to increase

target market as well as participation.

 92

REFERENCES

AHK, (2013). Target Market Study Kenya Solar PV & Wind Power, Delegation of
German Industry and Commerce in Kenya.
Alternative Energy Tutorials, Home of Alternative and Renewable Energy Tutorials.
www.alternative- energy- tutorials.com/energy-Articles/solar-cell-i-v-
characteristic.html
Eberhard A & K Gratwick (2010) IPPs in Sub-Saharan Africa: Determinants of
Success. MIR Working Paper, Energyinformative.org/grid-tied-0ff-grid-
andhybrid-solar-systems
ERB (Electricity Regulatory Board) (2005) Retail Electricity Tariffs Review Policy.
ERC (Energy Regulatory Commission) (2008). Kenya Electricity Grid Code.
Government of Kenya (2004) session Paper No. 4 of 2004 on Energy. Nairobi.
Government of Kenya (2007). https://www.elsevier.com/books/distributed-
energy-resources-in-microgrids/chauhan/978-0-12-817774-7
http://re.jrc.ec.europa.eu/esti/ http://www.erc.go.ke/retail-tariff.pdf.
https://energypedia.info/wiki/Publication_-_Renewables_2020_Global_Status_Report
https://greenminigrid.afdb.org/gvep-international-presentation-financing-mini-grids-
east-africa-2014
https://www.esmap.org/node/57633
https://www.pv-magazine.com/2020/12/29/solarcentury-to-install-11-2-mw-of-solar-
8-5-mw-battery-at-mozambique-mine/
https://www.researchgate.net/scientific-contributions/Hatem-H-Zeineldin-
2113245701
https://www.tesla.com/solarpanels
IMF (International Monetary Fund) (2010) Kenya: Staff Report for the 2009 Article IV
Consultation. Washington DC: IMF.
Jayanna Kanchikere, K. Kalyan kumar, “Proposal for 1 kW Roof-Top Solar PV Plant”,
International Research Journal of Engineering and Technology (IRJET), Vol. 4,
Issue 7, July2017
Kefa V.O Rabah, Integrated Solar Energy Systems for Rural Electrification in
Kenya Vision 2030, from http://safaricom.co.ke/foundation/file admin/template/main/
downloads
Kenya_VISION_2030-final_report-October_2007.pdf.
Manu Kumar D. M., Ganesha T., Mallikarjunayya C. Math, “Performance and
Evolution of Grid Connected to 5MW Solar Photovoltaic Plant in
shivanasamudra”, International Journal of Research in Advent Technology,
Vol. 3, No. 1, June 2015
Mark Hankins: “Solar Energy systems for Africa” November 2010
 93

NREL, “PVWatts Calculator,” [Online]. Available:

https://pvwatts.nrel.gov/pvwatts.php. [Accessed 2 February 2019]
Ondraczek, J., 2014. Are we there yet? Improving solar PV economics and power
planning in developing countries: The case of kenya. Renewable and
Sustainable Energy Reviews, 30, pp.604–615. Available at:
http://dx.doi.org/10.1016/j.rser.2013.10.010.
SE4All, 2016. Sustainable Energy for All ( SE4All ). Kenya’s Investment Prospectus.
, pp.1–63.
SolarCentury, Williamson Tea 1MWp solar farm, Available at:
http://www.solarcentury.com/za/wpcontent/uploads/sites/6/2014/05/Williamso
n-Tea-Changoi-CS-web.pdf.
www.globallegalinsights.com/practical-areas/energy-laws-and-regulations/kenya
www.irena.org
 94

APPENDICES

Appendix I: July 2018 Electric Bill

Appendix II: August, 2019 Electric Bill

 95

Appendix III: Array Setup

Appendix IV: PVGIS Curves of Admin GIS Location at 00, 30 & Optimum Tilt.