N. D.

Kaushika
Anuradha Mishra
Anil K. Rai

Solar
Photovoltaics
Technology, System Design, Reliability
and Viability
Solar Photovoltaics
N. D. Kaushika • Anuradha Mishra • Anil K. Rai

Solar Photovoltaics
Technology, System Design, Reliability and
Viability
N. D. Kaushika Anuradha Mishra
Centre for Energy Studies National Institute of Technology Calicut
Indian Institute of Technology Delhi Calicut, India
Delhi, India

Anil K. Rai
Department of Electrical & Electronical
Engineering
Ajay Kumar Garg Engineering College
Ghaziabad, Uttar Pradesh, India

Co-published by Springer International Publishing, Cham, Switzerland, with Capital Publishing

Company, New Delhi, India.

Sold and distributed in North, Central and South America by Springer, 233 Spring Street,
New York 10013, USA.

In all other countries, except SAARC countries—Afghanistan, Bangladesh, Bhutan, India,

Maldives, Nepal, Pakistan and Sri Lanka—sold and distributed by Springer, Tiergartenstr.
15, 69121 Heidelberg, Germany.

In SAARC countries—Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan

and Sri Lanka—printed book sold and distributed by Capital Publishing Company, 7/28,
Mahaveer Street, Ansari Road, Daryaganj, New Delhi 110 002, India.

ISBN 978-3-319-72403-4 ISBN 978-3-319-72404-1 (eBook)

https://doi.org/10.1007/978-3-319-72404-1

Library of Congress Control Number: 2017964703

© Capital Publishing Company, New Delhi, India 2018

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by the registered company Springer International Publishing AG part of
Springer Nature.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

Decentralized electricity-generating units are effective means of energy conserva-

tion. However, being based on diesel generation, these offer only a short-term
solution as the world’s fossil fuel production has already started declining. In this
regard, mature and sustainable decentralized electricity generation technologies such
as wind power and solar photovoltaic, solar thermal electric and hybrid renewable
power systems are being investigated around the world. Solar conversion technology
based on photovoltaics has several positive attributes and is on track to become a
leading competitor for next-generation power if system efficiency and cost-effective
approaches could be addressed.
In recent years, extensive research based on a multitude of solar cell technologies
and balance of system (BOS) components has exhibited the potential for significant
cost reduction and efficiency as well as functionality amelioration to make solar
photovoltaics a viable proposition. In this monograph, an overview of global
research activities on solar photovoltaics is presented. In particular, Chap. 1 encom-
passes the introduction to solar photovoltaic power and power generation from other
renewable resources; Chap. 2 deals with solar radiation measuring instruments and
an estimation of average solar radiation on a tilted surface; Chap. 3 presents a review
on the fundamentals of photovoltaic generation and formation of p-n junction in
semiconductor materials; Chap. 4 deals with different solar cell materials and wafer-
based fabrication technologies, while Chap. 5 exposes mathematical models of
transport processes in solar cells; Chaps. 6 and 7 introduce the electrical character-
istics of solar cells and power loss in array networks due to mismatch in solar cell
characteristics; Chap. 8 displays BOS other than solar PV array and different types of
electronic regulations; Chap. 9 presents the repertoires of space, terrestrial and cost-
effective applications of solar PV systems, while Chap. 10 discusses a knowledge-
based approach for the design of solar PV system and web access to solar PV design
aid expert systems for mass media adaptation. The reliability considerations of solar
PV systems and discussion on high-performance solar cells starting from first-
generation to third-generation solar cells are presented in Chaps. 11 and 12, respec-
tively, whereas Chap. 13 exposes the cost-benefit analysis of engineering systems/
projects.
v
vi Preface

This monograph covers some of the materials required for solar photovoltaic
courses that are being taught at engineering colleges and departments of some
universities. It is intended to serve research students, scientists, engineers, policy
experts and programme managers. Sincere thanks are due to Dr. R.B. Gangadhar, for
the useful discussions during the preparation of the monograph.

Delhi, India N. D. Kaushika

Calicut, India Anuradha Mishra
Uttar Pradesh, India Anil K. Rai
Contents

1 Introduction to Solar Photovoltaic Power . . . . . . . . . . . . . . . . . . . . 1

1.1 Conventional Power Supplies . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 Gap in Supply and Demand . . . . . . . . . . . . . . . . . . . . 1
1.1.2 Fuel Security Concerns . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.3 Environmental Concerns . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.4 Rising Prices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Power Generation from Renewable Resources . . . . . . . . . . . . . 3
1.2.1 Small Hydro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2.2 Biomass and Waste . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2.3 Ocean Gradient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2.4 Wind Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2.5 Solar Power Conversion Technologies . . . . . . . . . . . . . 8
1.2.6 Comparative Study . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.3 Scope of the Monograph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2 Solar Radiation Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1 The Solar Constant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2 Solar Spectral Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3 Atmospheric Effects on Solar Radiation . . . . . . . . . . . . . . . . . . 16
2.4 Solar Radiation Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.5 Solar Radiation Measuring Instruments . . . . . . . . . . . . . . . . . . . 20
2.6 Solar Radiation on a Tilted Surface . . . . . . . . . . . . . . . . . . . . . 21
2.7 Estimation of Average Solar Radiation . . . . . . . . . . . . . . . . . . . 23
3 Fundamentals of Photovoltaic Generation: A Review . . . . . . . . . . . 27
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.2 Atomic Bonding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.2.1 Ionic Bonding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.2.2 Metallic Bonding . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.2.3 Covalent Bonding . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.2.4 Van der Waals Bonding . . . . . . . . . . . . . . . . . . . . . . . 28

vii
viii Contents

3.3 Crystal Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

3.4 Band Gap Structure of Metal, Semiconductors and Insulators . . . 30
3.4.1 Commonly Used Semiconductor Materials . . . . . . . . . . 32
3.4.2 Types of Semiconductors: Doping . . . . . . . . . . . . . . . . 33
3.4.3 Drift, Diffusion and Recombination Currents . . . . . . . . 35
3.5 P-N Junction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.6 P-N Junction Solar Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4 Wafer-Based Solar Cells: Materials and Fabrication
Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.1 Photoionic Processes in Solar Cells . . . . . . . . . . . . . . . . . . . . . 43
4.2 Solar Cell Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.3 Silicon Solar Cell Wafer Technology . . . . . . . . . . . . . . . . . . . . 47
4.3.1 Sand to Metallurgical Grade Silicon . . . . . . . . . . . . . . 48
4.3.2 Metallurgical to Semiconductor Grade Silicon . . . . . . . 48
4.3.3 Polycrystalline Silicon to Single
Crystal Silicon Wafers . . . . . . . . . . . . . . . . . . . . . . . . 49
4.3.4 Fabrication of Silicon Solar Cells . . . . . . . . . . . . . . . . 49
4.3.5 Electrical Contacts and Encapsulation . . . . . . . . . . . . . 50
4.3.6 Improved Solar Cell Structures and Techniques . . . . . . 52
4.4 Indian Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5 Mathematical Model of Transport Processes . . . . . . . . . . . . . . . . . . 55
5.1 Physical Model of Solar Cells . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.1.1 Excess Carrier Generation: Photo Generation . . . . . . . . 55
5.1.2 Built-in Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
5.2 The Continuity Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
5.3 Solution of Continuity Equation . . . . . . . . . . . . . . . . . . . . . . . . 63
5.4 Piecewise Exponential Computational Scheme . . . . . . . . . . . . . 67
5.5 Illustrative Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
6 Electrical Characteristics of Solar Cells . . . . . . . . . . . . . . . . . . . . . . 73
6.1 Solar Cell: A Power Source . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
6.2 Equivalent Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6.3 Current-Voltage Characteristics . . . . . . . . . . . . . . . . . . . . . . . . 75
6.4 Performance Measurement Procedures . . . . . . . . . . . . . . . . . . . 77
6.5 Temperature and Solar Irradiance Effects . . . . . . . . . . . . . . . . . 80
7 Solar PV Module and Array Network . . . . . . . . . . . . . . . . . . . . . . . 81
7.1 Solar PV Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
7.2 Mismatch Losses in Array Network . . . . . . . . . . . . . . . . . . . . . 83
7.3 Formulations of Fractional Power Loss in Solar PV
Module and Array . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
7.3.1 Solar Cell Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
7.3.2 Fractional Power Loss in a Module . . . . . . . . . . . . . . . 87
7.4 Computational Results and Discussion . . . . . . . . . . . . . . . . . . . 89
Contents ix

8 BOS and Electronic Regulations . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

8.1 Balance of System (BOS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
8.2 Classification of BOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.2.1 Mechanical BOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.2.2 Electrical BOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.2.3 Electronic BOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
8.3 Electronic Regulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
9 Repertoires of Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
9.1 Space Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
9.2 Terrestrial Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
9.3 Cost Effective Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
9.3.1 Solar Photovoltaics for Buildings . . . . . . . . . . . . . . . . 107
9.3.2 Concentrating Applications . . . . . . . . . . . . . . . . . . . . . 112
9.4 System Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
10 Solar Photovoltaic System Design . . . . . . . . . . . . . . . . . . . . . . . . . . 115
10.1 Solar PV System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
10.2 System Design Considerations for Particular Location . . . . . . . . 118
10.2.1 Mathematical Formulations . . . . . . . . . . . . . . . . . . . . . 119
10.3 Solar PV Design Aid Expert System . . . . . . . . . . . . . . . . . . . . 121
10.3.1 Development of Knowledge Base for Indian Region . . . 121
10.4 Web Access to Solar PV Design Aid Expert System . . . . . . . . . 125
11 System Reliability Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . 127
11.1 The Reliability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
11.2 Site Dependence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
11.3 System Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
11.4 No Sun Days and Storage Autonomy . . . . . . . . . . . . . . . . . . . . 129
11.5 Power Loss and Hot Spot Effects . . . . . . . . . . . . . . . . . . . . . . . 129
11.5.1 Manufacturer’s Tolerances in Cell Characteristics . . . . . 130
11.5.2 Environmental Stresses . . . . . . . . . . . . . . . . . . . . . . . . 130
11.5.3 Shadow Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
11.5.4 Array Operation at Low Voltage . . . . . . . . . . . . . . . . . 131
11.6 System Reliability Improvements . . . . . . . . . . . . . . . . . . . . . . . 131
12 High Performance Solar Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
12.1 Solar Cell Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
12.2 First Generation Solar Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
12.3 Second Generation Solar Cells . . . . . . . . . . . . . . . . . . . . . . . . . 140
12.4 Third Generation Solar Cells . . . . . . . . . . . . . . . . . . . . . . . . . . 141
12.5 Limited Demonstration of Concentrating Solar Photovoltaic . . . 142
13 Solar PV System Economics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
13.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
13.2 System Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
13.3 Time Value of the Money . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
x Contents

13.4 Formulations for Evaluation of Money Value

Over Time Span . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
13.4.1 Notations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
13.4.2 Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
13.5 Cost-Benefit Analysis in Engineering Systems/Projects . . . . . . . 147
13.5.1 Category I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
13.5.2 Category II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
About the Authors

N. D. Kaushika Ph.D. formerly professor at the Centre for Energy Studies, Indian
Institute of Technology Delhi, and subsequently director of reputed engineering
institutions in Delhi and the National Capital Region, is a specialist in renewable
energy and environment. He has been a recipient of S.S. Bhatnagar Research
Endowment Award (Hariom Ashram Prerit) for research in energy conservation in
1987. He has promulgated the establishment of several research centres in engineer-
ing colleges, authored four books and contributed to articles in several reputed
journals as well as authored chapters in books by international publishers as Aca-
demic Press, USA and Elsevier.

Anuradha Mishra Ph.D. joined the Centre for Energy Studies, IIT Delhi, and
obtained her Ph.D. in 2004 specializing in “modelling and simulation of solar power
systems”. Presently, she teaches both graduate and postgraduate courses at the
National Institute of Technology Calicut (NITC) where she is also an independent
researcher.

Anil K. Rai Ph.D. was awarded his Ph.D. from IIT Delhi in 2008. He was the
coinvestigator in the Department of Science and Technology, Government of India
that sponsored research projects on solar photovoltaics. Presently, Prof. Rai works at
the Department of Electrical and Electronics Engineering at the Ajay Kumar Garg
Engineering College affiliated to Dr. A.P.J. Abdul Kalam Technical University,
Lucknow. He has published several articles in international journals. His areas of
interest are modelling and simulation of renewable energy systems, system engi-
neering and electrical power quality.

xi
Chapter 1
Introduction to Solar Photovoltaic Power

1.1 Conventional Power Supplies

Electricity is the most convenient form of energy for industrial, agricultural, com-
mercial and domestic activities; it is the key input for the economic and industrial
development of a country. In recent years, countries all over the world have taken up
industrialization, transport, communication as well as automation projects on large
scale and to maintain reasonable growth they must have a sustainable electricity
supply to achieve their developmental goals. However, the current state of affairs is
quite dismal due to the following challenges.

1.1.1 Gap in Supply and Demand

The developmental activities have led to an ever increasing escalation in the gap
between supply and demand for electricity. This is more pronounced in developing
countries where there is a chronic shortage of grid electricity caused by large amount
of generation and distribution losses. It may be illustrated from the facts and figures
for Indian electricity supplies as: power deficit 2.6%, peak power deficit 3.2%,
transmission and distribution losses 23% and plant load factor of 65%.

1.1.2 Fuel Security Concerns

The per capita consumption of electrical energy is a reliable index to connote the
progress made by a nation (for example, for 2011 per capita consumption of
electrical energy was 753 kWh for India, 4332.5 kWh for UK and 9538.8 kWh for

© Capital Publishing Company, New Delhi, India 2018 1

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_1
2 1 Introduction to Solar Photovoltaic Power

USA). So if the standards of living in the developing world are to improve, many-
fold increase in the electricity supplies is needed.
Presently major means of electrical generation include: hydro generation using
falling water resources, fossil fuel generation using coal, oil gas and nuclear gener-
ation using fissionable nuclear materials. The Earth has fixed amount of
non-replenishable resources of fossil fuels and nuclear materials and it has been
estimated that the proven reserves (that is recoverable with today’s technology) will
be exhausted in about 50–200 years. Hydro power generation is restricted to
geographically and geologically suited locations and availability of potential energy
of water. The accelerated generation from fossil fuels is also not encouraging due to
limited reserves of coal, oil and gas, more so, in view of oil crisis in recent years. All
these resources are able to cater to world’s increasing power requirements partially.

1.1.3 Environmental Concerns

The excessive use of fossil fuels (coal, oil and gas), nuclear fuel and hydro resources
is not benign to the environment and ecological balance on earth. For example, USA
has already developed around 50% of its hydro potential and further expansion is
suffering a set back because of ecological considerations. The adverse effect of the
excessive mining and use of fossil fuels are well documented. For example, coal
mining generates:
1. Ridges and valleys which result in erosion of top soil.
2. Waste material forms huge refuge banks occupying hundreds and thousands of
acres of land.
The combustion of coal at the power plant site adds fly ash, smoke, soot and
gaseous oxides of sulphur, nitrogen and carbon to the atmospheric air. For each kW
of electricity installation about 2.5 tonnes of coal is burnt annually and a plant of
200 MW capacity annually emits 16,000 tonnes of toxins and residues. Furthermore,
the typical conversion efficiency of power plant is 40%; consequently, 60% of heat
of combustion of coal goes to air and water environment.
The carbon dioxide and heat added to atmosphere have resulted in the increase of
earth’s planetary temperature which is often referred to as ‘Global Warming’; it is
threatening the world in terms of reduction of polar snow cover and raising of the
equatorial ocean level to flood the low lying islands and coastal regions; oxides of
nitrogen and sulphur are posing a threat to plants and vegetations kingdom through
‘acid rains’.
Nuclear power generation is to be carried out under stringent safety regulations
which involve cost. The limitations due to hazards of nuclear waste material and its
disposal impose further restrictions on nuclear power generation.
1.2 Power Generation from Renewable Resources 3

1.1.4 Rising Prices

Hydroelectric, thermal and nuclear power generating stations have following

prerequisites:
1. detailed site investigations for the design and construction work of generators,
distribution centres etc.
2. survey of load requirement and
3. laying of transmission cables, towers etc.
The above activities continue for several years and the project is often fraught
with several socio-economic problems before the generating stations start their
operations and deliver power to the user. In cases, where the generating point is
situated far away from the location where the power is to be used, the cost of
transmission lines and their maintenance makes the cost of power prohibitive and
remote areas suffer a setback in electrification.
In view of above challenges, non-conventional methods of power generations
have received steadily increasing attention around the world to control
1. dependence on fuels with escalating costs; and
2. pollution resulting from the use of nuclear and fossil fuels.
Non-conventional methods of power generation include solar photovoltaics, solar
thermal, bioconversion, small hydro, wind and ocean temperature gradients and
currents, which are also known as renewable resources of energy. These resources
are replenished almost as fast as they are used and they do not lead to environmental
pollution. They utilize sunlight, solar heat, biological waste, winds and ocean
currents which are otherwise unutilized.

1.2 Power Generation from Renewable Resources

In recent years, power generation from such renewable resources as solar, wind,
biomass, small hydro (including minimicro hydel), and ocean gradient have been
seriously counted upon to bridge the gap between the global demand and supply of
power. Extensive efforts over the past several decades aimed at research, develop-
ment, testing and demonstration of renewable power technologies have been made
throughout the world. Total worldwide R&D and demonstration investment on
emerging renewable power technologies has exceeded hundreds of billion dollars.
Substantial base of field experience and knowledge concerning these technologies
has been established. In what follows we present a brief outline of these
technologies.
4 1 Introduction to Solar Photovoltaic Power

1.2.1 Small Hydro

Hydro power is obtained from the potential and kinetic energy of water flowing from
a height. A hydro power project of capacity up to 25 MW is referred to as small
hydro power plant. The hydro power potential of a site is dependent on the discharge
and head of water as follows:

P¼ηQgH

where P ¼ power in kW, η ¼ overall power generating system efficiency, Q ¼ dis-

charge (rate of flow) in m3/s, g ¼ acceleration due to gravity (9.8 m/s) and H ¼ head
(height) in metres.
Small hydro power projects are classified on capacity as follows:
Micro hydro: up to 100 kW
Mini hydro: 101–2000 kW (i.e. 2 MW)
Small hydro: above 2 MW up to 25 MW
Small hydro power projects are also classified on the basis of head as follows:
Ultra low head: below 3 m
Low head: above 3 and up to 40 m
Medium/high head: above 40 m
Development of the small hydro power plant involves:
1. civil work making reservoir; and
2. choice of a suitable turbine and generator. In the simplest system a suitable
centrifugal pump may be selected to work as a turbine by matching its reverse
characteristic and coupling it to an alternator. The centrifugal pumps offer several
such advantages as low cost, easy installation and operation but can provide only
one discharge for a given head and speed whereas hydraulic turbine has flow and
therefore load control.
The cost of mini/micro hydel power plants is a sensitive function of the magni-
tude of head; higher the value of head lower the cost. Illustrative cost of small hydro
power plant in India is Rs. 7–8.5 crores per MW.

1.2.2 Biomass and Waste

Biomass includes agro-waste and residues such as bagasse, rice husk, straw, cotton
stalk, coconut shells, soya husk, de-oiled cakes, coffee waste, jute wastes, ground nut
shells, saw dust etc. These are also new and renewable resource for power genera-
tion. Municipal and industrial wastes can also be useful for electricity generation, but
are basically different forms of biomass.
1.2 Power Generation from Renewable Resources 5

The processes for conversion of agro waste and residues into useful form of fuels/
energy include:
1. thermo chemical processes (gasification, pyrolysis and liquefaction);
2. biochemical processes (anaerobic digestion and fermentation); and
3. direct conversion.
Over the years all these processes have been tried and tested and shown techni-
cally feasible, but they differ considerably in their efficiency of conversion and
degree of commercialization.
For electricity generation, gasification and direct combustion are the serious
alternatives. Gasification is a century-old technology. In this process the biomass
(agro-waste or residue) is reacted with a limited amount of air which is insufficient
for combustion. An inflammable gas known as producer gas or fuel gas is generated.
The fuel gas is a mixture of carbon monoxide, methane, carbon dioxide and
hydrogen. The gas, after cleaning and cooling, can be used as gaseous fuel in diesel
engines with 70–80% cut in diesel consumption. During the World War II many
producer gas powered vehicles were used in Europe and Australia. Gasified-based
power units are now commercially available; however, there are very few working
systems around the world though there have been several reports of the success of
experimental systems. This appears to be due to the fact that gasifier units are too
complex and messy to be viable for small scale (10–100 kW) utilization. Further-
more, the producer gas is a fuel of relatively low energy density; consequently the
combustion engine exhibits excessive wear and tear and the system maintenance is
expensive. Similarly the product of the biochemical process is the biogas which also
is a relatively low calorific value fuel (5000 k calorie/m3) as against the common
fuels like natural gas having calorific value (8600 k calorie/m3). In contradiction to
this is the process of direct combustion to raise steam. It is the simplest process and
has been used for about 200 years. All the biomass is combusted directly so the
efficiency of conversion is superior to gasifiers. Well-designed agro-waste combus-
tion systems are commercially available and they exhibit good reliability and
operational characteristics; the resultant ash is recycled back to the ecosystem. The
system is used for electricity generation and its waste heat may be utilized for such
applications as crop drying.

1.2.3 Ocean Gradient

Oceans are the largest natural collectors of solar energy. The upper layers of tropical
ocean are heated by the absorption of solar radiation; the water at depth of about
100–500 m is colder due to flow from polar regions. The absorption of solar energy
takes place according to Lambert’s law of absorption. It states that the absorption of
6 1 Introduction to Solar Photovoltaic Power

light decays exponentially with distance into the material. The reduction in the
intensity of a ray of original intensity along a length x is

I ðxÞ ¼ I 0 eKx

where I0 ¼ Intensity of radiation at the surface x ¼ 0, I ¼ Intensity of radiation at

distance x below the surface, and K ¼ absorption coefficient (unit L1). Its value is
0.05–0.5 depending upon type of water.
A typical variation of temperature with ocean depth is illustrated in Fig. 1.1
(Sukhatme 1984).
There are two types of energy which can be harnessed from ocean: thermal
energy from ocean surface warm water and mechanical energy from the tides and
waves. Ocean thermal energy conversion technology uses this temperature differ-
ence for obtaining power from sea. The largest reported temperature difference is
about 20  C and about 10  C is lost across heat exchanger walls. The resultant
temperature difference corresponds to 3.4% Carnot efficiency. Typical OTEC turbo
generators may have an efficiency of about 60%. So the net obtainable efficiency
will be about 2%. The operation of system auxiliaries and pumping of large
quantities of water will also consume power; this may well reduce the net efficiency
to a negative value. In fact it is for GW range that OTEC efficiency becomes
positive. The poor thermodynamic efficiency of OTEC system is obviously due to
the small temperature difference between source and sink.
A dam is used for allowing the water to pass through turbines to convert tidal
energy into electricity. Tides are driven primarily by the gravitational pull of sun and
moon on the earth, whereas the waves are driven primarily by the wind. Tide induced
currents are the underwater currents utilized to drive the tidal turbines in a similar
way as wind turbines are driven by the wind.

Fig. 1.1 A typical variation of temperature with ocean depth

1.2 Power Generation from Renewable Resources 7

1.2.4 Wind Power

Wind energy utilization is an ancient craft dating back thousands of years; it has been
in vogue for various such applications as ship propulsion, milling grain, water
pumping and more recently for generation of electricity. The basic principle
involved in wind (air in motion) energy utilization is that the wind exerts pressure
at an obstacle and is capable to do work; the rate of this work is called wind power.
The global potential of wind power for electricity generation was estimated to be
432,419 MW; for India it was about 25,088 MW by the end of 2015.
Wind machines are classified on the basis of the direction of their axis of rotation.
Two broad categories are: wind axis (or horizontal axis) machines and cross wind
(or vertical axis) machines. The pumping wind mills have multi bladed rotors in a
single wheel form generally known as fan-mill. The minimum wind speed required
for a wind mill is 2 m/s and with a wind speed of 10 m/s it could develop power up to
30 kW; the cut out wind speed is 17 m/s. Wind electricity generators are propeller
type with small number of blades (2 or 3) so that starting torque is low and they are
fast running. Reliable wind generators ranging from 100 W–3 MW rated output are
now commercially available. Some of the technological incorporations of 1980s in
machines are fibre glass reinforced polyester blades, microprocessor controller and
pitch regulations. The pitch regulation is rather a high tech feature; it allows the
blades to pivot along their own longitudinal axis. At low wind velocity the pitch
regulated blade offers large surface area against the wind and as the wind velocity
increases the surface area decreases gradually. The resultant system enables opti-
mum energy capture and has prolonged life span. It is cost effective in suitable wind
regimes. Total estimated capacity of wind power generation in the world at the end
of 2014 is 50 GW. The share held by Germany is 44.9%, USA has 4674 MW. Indian
setup in Gujarat, Tamilnadu and Andhra Pradesh has total capacity of 1702 MW.
In the international market the capital costs of wind power system varies in the
range of $1500–3000 /kW rated capacity; corresponding power costs are about 6–10
cent/kWh for grid connected systems. Generation costs of 5 cents/kWh are achiev-
able in windy regions. Wind farms can be built for a total investment as low as US
$1000 per installed kW. The cost of operation and maintenance for a wind power
system is estimated as 1 cent/kWh. The cost of maintenance, operation and fuel for
nuclear and coal based power plants are 1.9 and 2.3 cents per kWh respectively. In
India grid connected power systems are being promulgated to an investment cost of
Rs. 3.5  107/MW of rated output. The unit rating of individual generators range
between 55 kW and 400 kW corresponding to rotor diameter of 15 m to 34 m. Each
55 kW unit saves 400  103 kWh of electricity per year and avoids the addition of
3–5 tonnes of carbon dioxide, 16–28 tonnes of slag and fly ash and 160–280 kg of
particulate matter in atmosphere.
8 1 Introduction to Solar Photovoltaic Power

1.2.5 Solar Power Conversion Technologies

Solar radiant energy may be converted directly into electricity through processes like
photovoltaic, thermoelectric and thermionic or indirectly through thermal conver-
sion often referred to as thermodynamic conversion. The solar photovoltaics use a
semiconductor device called solar cell for the conversion. It is in most advance stage
of development as compared to other direct conversion methods. In the indirect
method solar heat is produced in a collector or concentrator and imparted to a
working fluid to increase its thermal energy which is subsequently converted into
mechanical power through Rankine, Brayton or Stirling cycles or through specially
designed devices.

1.2.5.1 Solar Thermal Electric

Solar radiant energy may be converted into electricity through thermal route often
referred to as thermodynamic process. Solar thermal electric is the subject of
intensive research and development activities in several developed countries, viz.
the USA, Europe, Japan and Australia and in some developing countries such as
India, Spain and Senegal etc. In thermodynamic process heat is collected from the
sun rays in solar collector and is transferred in a heat exchanger to a working fluid
which in turn activates a prime mover and alternator.
Several options in solar collectors, working fluid, prime movers, generators and
storage units are now available. Based on these possibilities wide range of systems
have been proposed and tested. Collector is the costliest component in a solar
thermal electric system. The variations in collector technologies include solar chim-
ney, solar ponds, flat plate collectors, evacuated tubular collectors, parabolic trough
collectors, hemispheric bowl, parabolic dish and central receiver. The temperature
range of heat generated by these collectors is illustrated in Table 1.1. Most advanced
solar thermal power systems include parabolic dishes with high efficiency engines,
central receiver power plants and solar boosted parabolic trough or solar ponds
conventional thermal generators.

Table 1.1 Solar thermal technology and their temperature range

S. No. Solar thermal power technologies Temp. range ( C)
1. Solar chimney 30–70
2. Solar pond 50–80
3. Flat plate collector 50–120
4. Evacuated tubular collectors 90–200
5. Parabolic trough 150–400
6. Hemisphere bowl 200–500
7. Paraboloidal dish 300–600
8. Central Receiver 300–600
1.2 Power Generation from Renewable Resources 9

The parabolic trough has been explored by LUZ and subsequently by SOLEL.
They have used solar parabolic trough collectors to preheat water prior to generating
superheated steam in a conventional natural gas system. Given sufficient solar input
the plant can operate at full rated power using solar energy alone. The parabolic
trough solar technology is the most proven and lower cost large scale solar power
technology available today. Field experiments on solar power tower concept have
been carried out in Spain, Italy, Japan, France and USA. World’s largest solar power
plant of 10 MW capacity was constructed near Barstow (CA). It used Rankine cycle
steam turbine operating at 500  C. More recently a solar thermal test plant, based on
central receiver concept, in Newcastle, Australia has generated super critical steam at
a pressure of 23.5 MPa and 570  C (Source: Dr. Alex Wonhas, CSIRO Energy
Director).
The parabolic dish systems have the highest operating temperatures and hence
higher conversion efficiencies. During 1980 to 2000 several investigations on design
and materials of dish technology ware reported. Most notable development is that by
reducing the weight of the structure, reduction in the system cost becomes possible
(Kaneff 1983). A rationale with respect to economic design of a dish concentrator is
to use the inherent strength of a paraboloidal shell involving a deep dish wherein no
frame would be required which in turn would reduce the weight and cost. A deep
dish will have a shorter focal length in comparison to a shallower dish (Kaneff 1983).
The surface receiver can be used as a focal absorber when the rim angle is greater
than 90 . A tradeoff between focal length, rim angle, nature of absorber and its life
properties have been carried out by Kaushika (1986) to evolve a viable design
involving deep dish of imperfect optics. Yet another viability approach is the dish-
sterling system wherein heat to electricity conversion is carried by sterling engine
mounted in the focal region of the dish. It is proved by gaseous hydrogen or helium,
usual working fluids of sterling system.
The parabolic dish systems have an advantage over power tower of being
modular. The estimated electricity generation costs of the three leading solar thermal
technologies are as follows (Solar Directory 1989):

Technology Installed costsU.S. $/kW Overall generation U.S. cents/kWh

Parabolic dish 1400 6
Power tower 3000–2000 15–6
Parabolic trough 4000–2000 20–10

The facts and figures change over a period of time; the intention here is to
give illustrative figures for comparative studies. The cost corresponds to systems
larger than 5 MW and lower figure applies above 50 MW. The coal-based
electricity costs are 4 cents/kWh. Thus in case of parabolic dishes the economics
of scale is possible; lower power level and scaling up to larger size plants can be
achieved in steps.
10 1 Introduction to Solar Photovoltaic Power

1.2.5.2 Solar Photovoltaics

Solar photovoltaic (PV) technology offers a convenient method for conversion of

sunlight, available in abundance during a major part of the year, directly into
electricity. The technique of conversion of sunlight directly into electricity by
solar cells employing photovoltaic effect is environment friendly, easy to maintain
and operate. The solar photovoltaic effect was first observed by Becquerel in 1839,
when he directed sunlight on one of the electrode of an electrolytic cell. Later,
Adams and Day in 1877 observed the effect in selenium. Amongst others the
pioneering work of Lange (1930), Schottky (1930) and Grondhal (1933) led to the
development of the photographic exposure meter and many other useful devices.
Subsequently in 1954, Chaplin, Fuller and Pearson of Bell Telephone Laboratory
fabricated diffused junction crystalline silicon cell with an efficiency of 6% and
Reynold et al. (1954) made a similar device with cadmium sulphide. At the outset,
single crystal silicon homojunction solar cell attracted considerable research and
development for outer space as with the advent of the space age in 1950s, the need
for a light and reliable power source for satellites was realized. The first solar
powered satellite, the American Vanguard I, was launched early in 1958. Since
then hundreds of satellites launched for scientific, military, meteorological, commu-
nications and other purposes have been powered by solar cells. Progressive improve-
ments have raised the conversion efficiency of space solar cells to nearly 20%. At the
outset, cost of solar cells was prohibitively high (US dollar 10 million per kW). Up to
1975, production of crystalline silicon solar cells, mainly for space programmes,
averaged about 100 kW per year and the cost reduced to only 1 million US dollars
per peak kW. Subsequently terrestrial applications of solar PV received attention;
cost reduction has been the main objective. In this regard following aspects of solar
photovoltaics have been investigated:

Solar Cells and Modules

The most common material used for solar cell is silicon which is available in
abundance on earth. The technology of production of semiconductor grade crystal
silicon is very well understood as this is used in fabrication of electronic components
including diodes, transistors, integrated circuits and microchips. The manufacturing
process of silicon PV modules comprises following steps:
1. Polysilicon production;
2. Single crystal ingot production;
3. Cutting ingots into wafers;
4. Cell production; and
5. Module assembly.
Based on the silicon wafer production technology, crystalline silicon cells are
classified into three main categories:
1.2 Power Generation from Renewable Resources 11

• Monocrystalline referred to as single crystalline: these cells are circular in shape;

• Polycrystalline, referred to as multi-crystalline; and
• EFG ribbon silicon and silicon sheet-defined film growth.
These multicrystalline and EFG ribbon silicon cells are non-circular and do not
involve waste of space in module formation.
Yet another classification based on the state and quality of the silicon material
involves three types of solar cells that have been produced and marketed:
1. Mono-crystalline silicon solar cell;
2. Polycrystalline silicon solar cells; and
3. Amorphous silicon solar cell/Thin film cell.
Furthermore, dozens of PV materials and process technologies have been pro-
posed and tested for the amelioration of efficiency and reduction of cost of solar
cells.
Thin film solar cells based on amorphous Si (a-Si), Cu (InGa), Se2 (CIGS) and
CdTe and nano/micro/poly-Si have received serious attention for their prospective
high potential for cost reduction and quality improvement and are now receiving an
increasing weight in the market. The laboratory cell efficiency is 25.6% for mono-
crystalline and 20.8% for multicrystalline wafer-based technology. The highest
laboratory efficiency in thin film technology is 21% CdTe and 20.5% for CIGS
solar cells (Fraunhofer ISE: 2016). The most common cells used in terrestrial
applications as mentioned above are silicon cells. Silicon wafer-based PV technol-
ogy is about 93% and thin film technologies are about 7% of the total annual
production in 2015. The share of multi-crystalline technology is now 69% of total
annual production.
Silicon might not be so able to meet long-term cost goals, since the thermody-
namics fundamentally limit their conversion efficiency up to 31% or 41% depending
on the concentration of incoming sunlight. This is known as Schockley-Quetsser
efficiency limit.
Dye-cells and organic photovoltaics are also advancing and hold promise for
affordable low cost technologies. Next-generation photovoltaics and nano-
architectured solar cells may enable several advanced solar cells design concepts
to be exploited (e.g. intermediate band and multiple exiton devices, hot carrier and
up/down conversion devices, etc.) and predicted to be very high-efficiency
approaches to solar energy conversion. However, the substantial theoretical and
practical barriers in their implementation in useful devices for a commercial per-
spective are still to be removed. Theory and experiments directed at gaining a deeper
understanding of the phenomena at atomic/molecular/nano- and meso-scale
involved in the development of improved, emerging and new generation high
efficiency/low cost solar cells are being currently researched.
12 1 Introduction to Solar Photovoltaic Power

Systems and Applications

The cells are connected in series/parallel to form modules which are used as a
photovoltaic generator. Depending on the load requirement, these modules are
connected to form arrays. Photovoltaic generators are modular in nature and can
be set up quickly and easily as small arrays and increased in size as demand for
power grows. So, solar PV generation is uniquely suited to remote areas and hilly
regions as the sunshine including diffused radiation is a distributed source and it
involves onsite generation cum utilization. For such areas, the transportation of fuel
and laying of transmission lines becomes an uneconomical and unreliable proposi-
tion. Such a situation is prevalent in remote areas and hilly regions with scattered
population where sunshine is abundant all round the year. Solar photovoltaic
generators also provide power for several other terrestrial applications like telecom-
munications, water pumping, cathode protection of pipelines and other metallic
structures, rural electrification, navigational aids, alarm systems, traffic warning
lights and other consumer products like pocket calculators. The best applications
correspond to distribution end of utility system or beyond the distribution system.
The global cumulative solar PV installations reached to 242 GWp at the end of 2015
(Fraunhofer ISE: 2016) with a region wise status as follows:

S. No. Country Capacity of installations (%)

1 China 21
2 Germany 16
3 Italy 8
4 Rest of Europe 16
5 Japan 14
6 North America 13
7 Rest of World 12

Three configurations of SPV power plant system have been proposed and tested:
1. Non-tracking
2. Tracking
3. Tracking-concentrating systems
In the laboratory, high concentration multi-junction solar cells achieve an effi-
ciency of up to 46% and with concentration technology module efficiency of up to
38.9% have been achieved (Fraunhofer ISE: 2016).

1.2.6 Comparative Study

Based upon our review of the technological activities, field experiences, cost trends
and resource magnitudes, it may be concluded that wind power is closest to use in
suitable windy regions. Solar appears to be a universal remedy for power generation.
1.3 Scope of the Monograph 13

The resource is best known, has appropriate magnitude, freely available and the
technologies are widely applicable, non-polluting, environment friendly and deliver
power directly or via thermal route. However the thermal route is cumbersome and
troublesome. The direct conversion technology based on solar photovoltaics seems
most promising. Its positive attributes may be summarized as follows:
1. Well Developed and Proven Technology
Initially developed for space applications, the SPV technology has undergone
exhaustive R&D, and has passed many stringent laboratory and field tests over a
period of time spanning half a century.
2. Decentralized Generation of Power
SPV solar plants can be setup and maintained in remote areas which are
otherwise costly to connect with electrical transmission lines and roads.
3. Simple and Reliable
Presently available solar PV power plants involve reliable electronic and
electrical components that have been time tested.
4. Modular Configuration
SPV power plants of all sizes can be developed using basic building blocks
and can be easily installed in succession.
5. User and Eco-Friendly
Solar power plants can be installed without safety problems and do not have
any adverse effect on the environment and ecological balance known to this date.
6. Low Operation and Maintenance Costs
Their capability for unattended operation and lack of need for water or other
fluid for cooling confers in them low operation and maintenance costs and high
degree of reliability.

1.3 Scope of the Monograph

In conclusion, solar conversion technology based on photovoltaics has several

positive attributes and is on track to become a leading competitor for next generation
green power if system efficiency and cost effective approaches could be addressed to
make solar photovoltaic a commercial prospective. In recent years, extensive
research based on micro and nano-solar cell technology has exhibited the potential
for significant cost reduction and efficiency as well as functionality amelioration.
This monograph presents an overview of global research activities in this direction
which encompasses wafer-based solar cells materials and fabrication technologies,
mathematical model of transport processes, mismatch losses in array network, BOS
and electronic regulations, repertoires of applications, solar PV system design and
reliability and prospective high performance solar cells. It considers mono-
crystalline and multi-crystalline silicon-based solar cells and discusses the simula-
tion model efficiencies as well as the practical efficiencies of thin film micro
crystalline, nano crystalline, hybridized silicon-based solar cells, multi junction
14 1 Introduction to Solar Photovoltaic Power

solar cells and thin film cadmium telluride (CdTe) cells. Furthermore, it points out
that high efficiency of solar cell device reached so far is not followed by high system
efficiency due to several drawbacks. In this regard it presents work on fault tolerant
circuitry to combat mismatch losses in cell network and balance of the system. It also
considers cost reduction options like building integrated photovoltaics and the
concentrating systems.
Chapter 2
Solar Radiation Characteristics

2.1 The Solar Constant

Sun is a star, a vast spherical mass of hydrogen and helium in proportion 3:1 by
weight; the proportion of other elements is less than 1% of total mass (1.9
1030
kg). The temperature in its central interior region is estimated to be very high
~2106 K; consequently huge amount of thermonuclear energy is generated in the
central core by the process of nuclear fusion. Some of the basic characteristics of
sun-earth system are illustrated in Fig. 2.1.
The radiant energy flowing out of the sun is 3.7
1026 Watts. This energy flow
corresponds to a rate of depletion of solar mass as 4.1
106 tonnes per second or 1%
of the total mass in billion years. The solar energy intercepted by the earth (SE) may
be calculated as follows:

3:7
1026 2
SE ¼ πr
4πR2
where R is the sun-earth distance ¼ 1.495
1011 m and r is the earth’s
radius ¼ 6.35
106 m.
So we have SE ¼ 173
1015m.
The solar energy available outside the earth’s atmosphere is called extra-
terrestrial solar radiation. The amount of this energy that crosses in unit time through
unit area placed perpendicular to the direction of energy flow is referred to as solar
energy flux. At sun-earth distance this flux is essentially constant and is called the
solar constant. Owing to the variation of sun-earth distance solar constant varies
slightly with day of year as follows:

© Capital Publishing Company, New Delhi, India 2018 15

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_2
16 2 Solar Radiation Characteristics

Fig. 2.1 Sun-earth system

  
360n
I α ¼ I sc 1 þ 0:033 cos ð2:1Þ
365

where Iα ¼ extra-terrestrial radiation, Isc ¼ solar constant and n ¼ day number of

the year.
The expected value of solar constant is 1367 W/m2 (Duffie and Beckman 1980).

2.2 Solar Spectral Distribution

Every hot body emits radiant energy which is characteristics of body temperature,
size and nature of its surfaces. The spectrum of solar radiation too is characteristic of
sun’s temperature. However, since sun’s energy is generated at its centre, which has
a temperature of millions of degree, the central hot core terminates in a relatively
colder surface called photosphere. For the purpose of solar spectrum and the
corresponding temperature of sun, we always consider the temperature of photo-
sphere. As viewed outside the earth’s atmosphere, the spectrum of solar radiation is
characteristic of that emitted from a black surface at a temperature of 5762 K. This
implies that the solar spectral irradiance ranges between the wavelength region of
0.2–2.0 μm and has maximum invisible light region at 0.48 μm (Fig. 2.2). The
brightness of solar disc also varies from its centre to edges. However, for engineering
calculations, the disc is often assumed to be of uniform density.

2.3 Atmospheric Effects on Solar Radiation

Sun is a sphere of diameter 1.39
106 km situated at a distance of 1.50
108 km. It
subtends an angle of 32 min at earth’s surface. So sun’s radiation beam outside the
earth’s atmosphere is almost parallel. During its passage through atmosphere, it is
affected by the process of reflection, absorption (selective as well as continuous) and
scattering due to clouds, particulate matter and gas molecules. All the processes are
wavelength dependent and the magnitudes of effects depend on the thickness of the
atmosphere penetrated. A measure of atmospheric thickness is air mass which is
2.4 Solar Radiation Geometry 17

Fig. 2.2 Spectral distribution of Extra-terrestrial solar radiation

expressed by the ratio of the optical thickness of the atmosphere through which
direct sun’s rays pass to the optical thickness if the sun was at zenith (vertical
incidence) and is given as follows:

m ¼ ð cos θz Þ1

where θz is zenith angle.

The net atmospheric effect is that radiation received at earth’s surface consists of
two components: (i) beam radiation—the radiation received from sun without
change of direction and (ii) diffuse radiation—the radiation received from all
directions. The sum of beam and diffuse radiation is referred to as total radiation
or global radiation.

2.4 Solar Radiation Geometry

For the simulation and design of solar energy systems one is interested in the value
of solar intensity on a surface having an arbitrary orientation; this in turn depends on
the angle which the sun’s rays make with the normal to the surface. This angle is
referred to as incidence angle, θ for the planar surface. Solar incidence angle depends
on the latitude of location (φ), day of the year characterized by the declination angle
(δ), solar hour angle (ω), surface azimuthal angle (rp), and slope of planar surface to
horizontal (β). Several texts have dealt with the sun-earth geometry related to solar
incidence angle. In what follows we present a discussion closely based on Boes
(1981) and Sukhatme (1984).
Earth revolves round the sun in an elliptical orbit and this motion may be
represented by sun’s virtual motion in celestial sphere whose centre is centre of
earth. The plane of the elliptical orbit is called the ‘ecliptic plane’.
18 2 Solar Radiation Characteristics

The angle between the sun-earth vector (which lies in elliptic plane) and earth’s
equatorial plane is called the solar declination angle, δ. It is maximum (23.45 ) on
June 21 and minimum (23.45 ) on December 21. It may be expressed as
 
360
δ ðin degreesÞ ¼ 23:45 sin ðn þ 284Þ
365:24

where n is day number of the year (1st January onwards).

The rotation of earth about its axis causes the day-night cycle and is the basis of
solar hour angle which is an angular measure of time and is equivalent to 15 per
hour. It is measured from solar noon (i.e. when the sun is on the local meridian) and
is positive (negative) before (after) solar noon.
The latitude of site, φ, is the angle made by the projection of the line on the
equatorial plane with the radial line joining the site to the centre of earth. It varies
from 90 to +90 and measured as positive for northern hemisphere
conventionally.
The sun’s position on celestial sphere may be represented by solar azimuth angle,
rs (sun’s angular distance from the south) and solar altitude angle, α (sun’s angular
distance from horizon). Similarly the orientation of fixed planar surface is
represented by its azimuth angle, rp (angle in the horizontal plane between line
due south and the projection of normal to the surface on the horizontal plane) and its
tilt angle with the horizontal, β. The azimuth angle is measured from the south,
positive towards east. As an illustration rp ¼ 0 for south wall, +90 for east wall,
90 for west wall and 180 for north wall of a building. β is taken positive for south
facing surfaces. All the above angles are illustrated in Fig. 2.3.
Following Boes (1981), solar incidence angle θ may be expressed as:


cos θ ¼ cos α cos r s  r p sin β þ sin α cos β ð2:2Þ

where solar altitude angle α is given by

sin α ¼ cos ϕ cos δ cos ω þ sin ϕ sin δ

and the solar azimuth angle, rs by:

 
1 cos δ sin ω tan δ
r s ¼ sin if cos ω >
cos α tan ϕ

and
 
1 cos δ sin ω tan δ
r s ¼ 180  sin if cos ω <
cos α tan ϕ

Following simple cases are often of interest:

1. Vertical surface (viz. walls)
2.4 Solar Radiation Geometry 19

Fig. 2.3 Sun-earth angles Solar Radiation

θ
C

β
B A

Zenith

Normal to
Sun horizontal surface

qz
θ

b
β
N
aW

rs
rp
S F


β ¼ 90

cos θ ¼ cos α cos r s  r p

2. Horizontal surface


β ¼0
cos θ ¼ sin α

3. Surface facing due south


rp ¼ 0
cos θ ¼ cos α cos r s sin β þ sin α cos β

4. Vertical surface facing due south

20 2 Solar Radiation Characteristics

 
β ¼ 90 , r p ¼ 0
cos θ ¼ cos α cos r s

Some authors (Duffie and Beckman 1980; Sukhatme 1984) have used a simple
formulation as an alternative to Eq. (2.2), which is as below:

cos θ ¼ sin ϕ sin δ cos β  sin δ cos ϕ cos r p sin β

þ cos ϕ cos δ cos ω cos β þ cos δ sin ϕ sin β cos r p cos ω
þ cos δ sin β sin r p sin ω

2.5 Solar Radiation Measuring Instruments

(i) Pyranometer
Pyranometer works on the principle of thermoelectric effect. It is used to measure the
global radiation and when shaded from beam radiation by using shading ring it
measures the diffused radiation. A precision pyranometer is designed to respond to
radiation of all wavelengths to accurately measure the total power in the incident
spectrum. It contains a thermopile whose sensitive surface consists of circular,
blackened, hot junctions, exposed to sun, the cold junctions being completely
shaded. The temperature difference between the hot and cold junctions is the
function of radiation falling on the sensitive surface. The sensing element (thermo-
pile) is covered by two concentric hemispherical glass domes to shield it from wind
and rain.

This also reduces the convection currents. A radiation shield surrounding the outer
dome and coplanar with the sensing element, prevents direct solar radiation from
heating the base of the instruments.
2.6 Solar Radiation on a Tilted Surface 21

(ii) Pyrheliometer
It measures beam radiation from the sun at normal incidence and works on the
principle of thermoelectric effect using thermopile having sensitivity 8–10 μV/W/m2.
It uses a long collimator tube to collect beam radiation whose field of view is limited to
a solid angle of 5.5 . It often uses two axes tracker for continuous recordings.

(iii) Sunshine recorder

This instrument measures the duration in hours of bright sunshine during the course
of a day. It essentially consists of a glass sphere mounted on its axis parallel to that of
earth within a spherical section. The bowl and glass sphere are arranged in such a
way that the sun’s rays are focused sharply at a spot on a card held in a groove in the
bowl. The card is prepared from a special paper bearing time scale. As the sun’s
elevation changes, the focused bright sunshine burns a path along this paper. The
length of the trace thus obtained on the paper is the measure of the duration of the
bright sunshine.

2.6 Solar Radiation on a Tilted Surface

The total radiation incident on a surface consists of

(i) Beam or direct radiation
(ii) Diffuse or scattered radiation
(iii) Solar radiation reflected from the ground or surroundings.
The solar radiation that reaches the earth’s surface directly without being
absorbed or scattered is called beam or direct radiation. The direct radiation produces
a shadow when interrupted by an opaque object.
22 2 Solar Radiation Characteristics

Diffuse radiation is that radiation received from sun after its direction has been
changed by reflection and scattered by the atmospheric molecules and by dust
particles. The diffuse radiation is received by earth surface from all directions.
The practical distinction between the two components is that only the beam
radiation can be focused. Even on a clear day there is some diffuse radiation.
Normally the beam radiation and diffuse radiations on horizontal surface are
recorded. After knowing beam and diffuse radiations on horizontal surface, Liu and
Jordan (1962) have given a formula to evaluate total radiation on tilted surface.

I T ¼ I bh Rb þ I dh Rd þ ρRr ðI bh þ I dh Þ

where Rb, Rd and Rr are known as conversion factors for beam, diffuse and reflected
components respectively. Ibh and Idh are the beam and diffuse components respec-
tively on horizontal surface.

ρ ¼ reflection coefficient of ground

¼ 0:2 for concrete or glass surface
¼ 0:7 for snowcovered surface

Sun Rays
Normal
Normal
Sun Rays

θz
β
Horizontal surface Tilted surface

For horizontal surface

The beam radiation of horizontal surface Ibh ¼ IN cos θz.
where IN is Sun rays and given by

I bh
IN ¼
cos θz
For tilted surface

I bt ¼ I N cos θt
I bh
¼
cos θt
cos θz

Or,

cos θt
I bt ¼ I bh
cos θZ
But
2.7 Estimation of Average Solar Radiation 23

cos θz ¼ cos ϕ cos δ cos ω þ sin ϕ sin δ

cos θt ¼ cos ðϕ  βÞ cos δ cos ω þ sin ðϕ  βÞ sin δ

where
ϕ ¼ Latitude of the site
β ¼ Optimum tilt angle
ω ¼ Sun hour angle
δ ¼ Solar declination angle
θt ¼ Zenith angle on tilted surface
Thus

cos θt
Rb ¼
cos θz

It is defined as the ratio of flux of beam radiation incident on tilted surface to that
on a horizontal surface.
Rd: It is defined as the ratio of flux of diffuse radiation incident on tilted surface to
that on a horizontal surface. It is given by

1 þ cos β
Rd ¼
2
Rr: The reflected components come from the ground and other surrounding
objects

1  cos β
Rr ¼
2
Thus

cos θt ð1 þ cos βÞ ð1  cos βÞ

I T ¼ I bh þ I dh þ ρðI bh þ I dh Þ
cos θz 2 2

2.7 Estimation of Average Solar Radiation

Measured solar radiation data are the best source of information for estimating
average radiation on a surface relevant to solar energy utilization. In absence of
such data one uses empirical relationship based on data on average hours of sunshine
or average percent of possible sunshine hours; the sunshine data is widely available
and is collected with Campbell Stories Instrument. Cloud cover data may also be
used in this connection, but it generally embraces uncertainty resulting from visual
estimates.
24 2 Solar Radiation Characteristics

Solar radiation data are the basic requirement in the modelling and simulation of
various energy and environment systems. For example, the global, diffuse and direct
solar radiation data at horizontal and tilted surfaces is an essential element in
modelling and performance evaluation of various solar systems which include:
1. Evaluation of monthly mean hourly performance during the year-round cycle on
tilted flat plate collector and photovoltaic arrays.
2. The energy needs of a building wherein a large number of factors related to
climate and structural configurations control the indoor temperature. The effec-
tiveness of these factors is highly intricate to be assessed independently in a
physical model.
3. Determination of solar fraction of hybrid systems and design of control logic for
the operation of solar power plants.
In practice the solar radiation data is generally variable and enormously incon-
sistent (Sfetsos and Coonick 2000; Dorvlo et al. 2002). These data are often not
available for the desired locations and are not of the required accuracy. Several
empirical models have therefore been developed for the estimation of solar radiation
at different geographical and meteorological conditions (Reddy 1971; Hottel 1976;
Sabbagh et al. 1977; Barbaro et al. 1978; Goh 1979; El-Nashar 1981; Ogeman et al.
1984; Supit and Van Kappel 1998). These empirical relations rely on easily available
meteorological parameters such as atmospheric temperature and mean duration
sunshine per hour, relative humidity and total rainfall.
In recent years several authors have proposed and tested the efficacy of artificial
neural network modelling for estimation of solar radiation. For example, Kaushika
et al. (2014) have presented an artificial neural network model based on interrela-
tionship of direct, diffuse and global solar radiations. In this paper, an explicit
computational model based on multi output frame of neural network is developed.
The approach involves the estimation of solar transmission through clear atmosphere
(simple clear day model) and its modification by the allowance of weather phenom-
ena through ANN analysis. The model is rigorous as it simultaneously considers the
global, diffuse and direct components of solar radiation as target variables in the
analysis. Computational algorithm is as follows:
As a first approximation, solar beam radiation on earth under clear sky condition
can be estimated from the consideration of atmospheric transmittance, sun-earth
geometry and climate type following the simple model due to Hottel. The solar beam
radiation relation is


Sbn ¼ I a an þ a1 ek sec θz

The term within the brackets is all wavelength transmittance of solar radiation
through the atmosphere. It is a function of time of the day, day of the year and
location on earth. Ia is the extra-terrestrial radiation and can be expressed as
2.7 Estimation of Average Solar Radiation 25

  
360 n
I a ¼ I sc 1 þ 0:033 cos
365

where Isc is the solar constant 1367 W/m2, and n is number of day (n ¼ 31 for
January 31) and θz is solar zenith angle or solar incident angle on a horizontal plane.
The beam radiation falling on a horizontal surface is therefore given by

Sbh ¼ Sbn cos θz

For the estimation of diffuse component of solar radiation falling on a horizontal

surface (Sdh), Liu and Jordan (1960) have given an expression in terms of atmo-
spheric transmittance for beam radiation


Sdh ¼ I a cos θz 0:2710  0:2939 a0 þ a1 ek sec θz

This expression is rather approximate and may be used as guide in search of

experimental relationship. So the global radiation may be expressed as

Sgh ¼ Sbh þ Sdh

The above equations may be used to estimate the values of direct, diffuse and
global radiations. The deviations of these estimates from observations may be
considered due to climate and weather phenomena and are characterized by the
atmospheric clearness indices as follows:

Observed value of solar radiation

Atmospheric clearness index ¼
Calculated value of solar radiation
A value compatible with observed values may therefore be obtained by multi-
plying the atmospheric clearness index with the value calculated from simple clear
day model. The clearness index value for direct, diffuse and global radiation have
been determined by Kaushika et al. (2014) by using Artificial Neural Network based
on long-term solar radiation data in Indian region. A contour map of these determi-
nations for typical hours of a day is shown in Fig. 2.4.
26 2 Solar Radiation Characteristics

Fig. 2.4 Contour maps of clearness indices of global solar radiation for typical hours of a day for
Indian region
Chapter 3
Fundamentals of Photovoltaic Generation:
A Review

3.1 Introduction

Photovoltaic generation of electricity is caused by the radiation separating positive and

negative charges in absorbing material. If the electric field is present these charges can
generate an electromotive field and produce a current for use in an external circuit. The
fundamentals of photovoltaic generation, therefore, involve the knowledge of science of
materials and the photovoltaic effect therein which in turn involve such basic concepts as
atomic bonding, metals, semiconductors and insulators, crystal structure (silicon), band
gap energy, built-in electrostatic field, PN junction, interaction of matter and radiation. In
this chapter we present a review of these concepts.

3.2 Atomic Bonding

The basis of atomic bonding is the structure of atoms which states that atoms are
formed with nucleus (positive charge) in the centre and electrons (negative charge)
orbiting around the nucleus at different energy levels or shells. In solids the atoms
are closely packed and they create interatomic bond which is often referred to as a
chemical bond. In this circumstance the atoms try to obtain a full outer shell. There is
an important principle regarding electrons filling up shells: the Pauli’s exclusion
principle which states that two electrons cannot occupy the same energy level. The
mode of filling up of shells determines the chemical, electrical and optical behaviour
of the matter. In chemical bonds different or identical atoms come together for
mutual minimization of electron energies and shell filling. There are four main
types of bonding as follows:

© Capital Publishing Company, New Delhi, India 2018 27

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_3
28 3 Fundamentals of Photovoltaic Generation: A Review

3.2.1 Ionic Bonding

This bonding occurs when two atoms of different species come together and the
outer most electron of one atomic species is donated to other to fill its outer orbit.
Consequently, the species involved in this bond are positively and negatively
charged ions and they attract each other. An example of ionic bonding is NaCl
(Sodium chloride).

3.2.2 Metallic Bonding

Metallic bonding occurs in metals. In this bonding outer-most electrons of atom are
not accepted by any other atoms and move freely in the solid. The electron cloud
around the nuclei actually acts as glue that holds the lattice of atoms together. The
mobile electrons are responsible for the good conduction of electricity in metals such
as silver which has conductivity about almost 6
107 mho/m.

3.2.3 Covalent Bonding

In this bonding none of the outer-most electrons are able to move freely but are
shared amongst atoms involved in the bond. Diamonds are an example of carbon
atoms in a covalent bond. This is responsible for the electrical insulation character of
diamonds and fused quartz (thermal conductivity less than 2
1017 mho/m). In
case of silicon and germanium, the covalent bonding is rather weak and some
electrons may be released to move freely when sufficient thermal energy is provided.
It gives the covalent bond atoms the character of semiconductor. Silicon and
germanium have crystalline structure. Their atoms are arranged in an ordered array
known as crystal lattice. These materials have four valence electrons in their outer
shell. The neighbouring atoms form covalent bonds by sharing four electrons with
each other so as to achieve inert gas structure.
In the case of intrinsic semiconductor, the covalent bonds have to be broken
thermally to provide electrons for conduction.

3.2.4 Van der Waals Bonding

It occurs in atomic species where the atoms behave like an instantaneous dipole and it
inducts a very small but opposite dipole moment on another atom. Consequently a very
weak bond of attraction is created. It is found in noble gases such as argon, crypton etc.
3.3 Crystal Structure 29

3.3 Crystal Structure

Materials can be broadly classified into two categories: crystalline and

non-crystalline solids.
A solid in general is said to be a crystal if the constituent particles (atoms, ions or
molecules) are arranged in a three dimensional periodic manner. In a crystalline
material the arrangement of atoms is in periodically repeating pattern whereas no
such regularity of arrangement is found in a non-crystalline material. A crystalline
solid can be either a single crystal, where the entire solid consists of only one crystal.
It has well-defined crystal structure, higher density due to closed packing of atoms
and exhibits sharp diffraction pattern upon irradiation.

Non-crystalline solid often possesses entangled chain without periodicity. It has

lower density as the packing of atoms is zigzag, does not show sharp diffraction
pattern and melts over a range of temperatures. However in both forms, the coordi-
nation number is almost the same. Formation of non-crystalline structure is charac-
terized by several factors as listed below:
(i) Formation of one-dimensional chain molecule.
(ii) Formation of two-dimensional sheet molecules.
(iii) Non-formation of three-dimensional primary bond.
(iv) Absence of primary bonds in all the directions.
(v) Non-parallel, entangled chain configuration.
(vi) Weak secondary bond.
(vii) Open network of the atomic packing.
Typical crystalline and amorphous structures are illustrated in Fig. 3.1.
30 3 Fundamentals of Photovoltaic Generation: A Review

Fig. 3.1 Crystalline and amorphous structures of solid

3.4 Band Gap Structure of Metal, Semiconductors

and Insulators

The characteristics of a solid can be understood in terms of band energy diagram as

shown in Fig. 3.2. Its energy band structure and the manner in which electrons
occupy the energy-band are central to the electrical properties of a solid. The energy
levels split into N levels of energy where N is the number of atoms in the crystals.
The electrons in the outer-most orbit of an atom are known as valence electrons.
Under normal condition of the atom, valence band corresponds to the electrons of
highest energy. This band may be filled completely or partially but never empty. The
energy band which corresponds to the valence electrons is called valence band.
The next higher permitted energy band is called the conduction band. It refers to
allowed energy of bound electrons. It may be either empty or partially filled with
electrons. The gap between the valence band and conduction band is called forbid-
den energy gap. It is denoted by Eg and may be defined as the energy difference
between the most energetic bound electrons and minimum allowed energy of
conduction electrons governed by quantum mechanical effect. So forbidden energy
gap is a region in which no electron can stay as there is no allowed energy state. The
width of the forbidden energy gap represents the bondage of valence electrons to the
atom. The greater the forbidden energy gap more tightly the valence electrons are
3.4 Band Gap Structure of Metal, Semiconductors and Insulators 31

Empty or
Conduction band
partially filled
Band Energy (eV) Eg
Fully or
Valence band
partially filled

Completely filled
inner bands

Fig. 3.2 Energy band diagram of a solid

Fig. 3.3 Energy-band overlap

diagram for three types of
material: Metals, Conduction
Electron energy

Semiconductors and band

Insulators
Fermi level Bandgap

Valence
band

metal Semiconductor insulator

bound to the nucleus. When valence electron absorbs enough energy, it jumps across
the forbidden energy gap and enters the conduction band.
Fermi Level refers to the highest occupied molecular energy orbital at absolute
zero. It is usually found at the centre between the valence and conduction bands. The
particles in this state each have their own quantum states and generally do not
interact with each other. When the temperature begins to rise above absolute zero,
these particles will begin to occupy states above the Fermi level and states below the
Fermi level become unoccupied.
Depending upon the band theory, all solids can be classified as metals, semi-
conductors and insulators (Fig. 3.3).
In metals (conductors) there is no forbidden energy gap between the valence
bands and conduction bands. They overlap each other. The valence–band energy is
same as conduction band energy. It is very easy for a valence electron to become a
conduction electron. Therefore, without supplying any additional energy, a metal
already contains a large number of free electrons and works as a good conductor.
Insulator has a very wide forbidden-energy gap; because of this, it is impossible
for an electron in the valence band to enter into the conduction band. At room
temperature, an insulator does not conduct because there are no conduction electrons
in it. However, it conducts only if its temperature is very high or a very high voltage
is applied across it.
32 3 Fundamentals of Photovoltaic Generation: A Review

Semiconductor material is one whose electrical properties lie in between those of

insulators and good conductors. The forbidden energy gap is of the order of 1 eV.
The energy provided by the heat at room temperature is sufficient to lift electrons
from the valence band to conduction band.

3.4.1 Commonly Used Semiconductor Materials

A large number of tetravalent (group IV of Periodic Table) materials such as carbon

in diamond state, silicon, germanium and gray tin are semiconductors. The forma-
tion of allowed energy bands is related to the overlapping of atomic orbitals in the
crystal. The concept of band gap is important in understanding the behaviour of these
semiconductors. It needs quantum mechanical considerations. However, there is a
simple intuitive explanation based on the fact that all of valence electrons are
involved in strong covalent bonds that require some energy to break. The minimum
energy required for breaking the covalent bond in these materials is 7 eV, 1.12 eV,
0.75 eV and 0.1 eV respectively. Carbon in diamond state, having forbidden energy
gap of 7 eV, behaves more or less as an insulator. Silicon and germanium need
1.12 eV and 0.72 eV respectively to break a covalent bond; therefore, silicon and
germanium are considered to be most suitable semiconductor materials.
Silicon is most common and widespread element on the earth. The word silicon
has origin in the Latin word silex which means flint. It does not occur free in nature.
It is always found as a compound and almost always as an oxide. The simplest
compound of silicon and oxygen is silicon dioxide which is commonly known as
silica or sand. Silicon is obtained from silicon dioxide. Silicon is separated from
oxygen by the use of a reducing agent like coke which is one of the cheapest
reducing agent. Silicon dioxide is heated with coke in an electric furnace

SiO2 þ 2C ! Si þ 2CO "

Silicon also occurs in nature as quartz, flint and precious stones like opal,
agale etc.
Silicon belongs to group IV of the Periodic Table and placed below carbon in the
table. The atomic number and atomic mass of silicon are 14 and 28 respectively and
having 14 protons, 14 neutrons and 14 electrons. The electronic configuration of
silicon is 2, 8 and 4. Like carbon it has four electrons in its outermost shell or valence
shell. Hence it is tetravalent.
Properties of Silicon
• Silicon is a grey, hard and shining substance with density 2.33 gm/cm3 at room
temperature. Band gap energy 1.12 eV at 300 K.
• It melts at 1410  C and boils at 365  C.
3.4 Band Gap Structure of Metal, Semiconductors and Insulators 33

• It is a crystalline solid and its atomic structure is similar to that of diamond. Each
silicon atom is surrounded by four other silicon atoms. All the silicon atoms are
held together by a network of covalent bonds.

3.4.2 Types of Semiconductors: Doping

The classification of semiconductors can be done as shown in Fig. 3.4.

An intrinsic semiconductor is one which is made of the semiconductor material
in its extremely pure form. Intrinsic semiconductor materials are: pure germanium
and silicon which have forbidden energy gap of 0.72 eV and 1.1 eV respectively.
The energy gap is so small that even at ordinary room temperature, electrons jump
across the small energy gap between the valence and conduction bands creating
positively-charged holes in the valence band i.e. an intrinsic semiconductor has
equal number of conduction electrons and holes. Although the holes exist in the
valence band, they are similar to conduction electrons as they serve as mobile charge
carriers. The method of movement is little different. Essentially a hole moves
through the crystal when electrons from neighbouring silicon atom move into fill
the hole and replace the missing electron. When this happens, the neighbouring
silicon atom then inherits the broken bond and the process advances. Both holes and
electrons move with quite similar dynamics although they have opposite charges and
exist in different electronic bands.
In an intrinsic semiconductor, Fermi level lies exactly in the middle of the
forbidden energy gap i.e. midway between the conduction and valence bands
(Fig. 3.5).

Fig. 3.4 Types of Semiconductor

semiconductor materials

Extrinsic or doped Intrinsic or pure

semiconductor semiconductor

N-type P-type

Fig. 3.5 Fermi level in Eg Conduction band

intrinsic semiconductor
EF Eg

Eo
Valence band
34 3 Fundamentals of Photovoltaic Generation: A Review

The Fermi level may be defined as the energy which corresponds to the centre of
gravity of conduction electrons and holes weighed according to their densities. The
number of electrons in conduction band is nc and given by


nc ¼ NP E g

where N is total number of electrons in both conduction and valence bands and
P (Eg) represents the probability of an electron having energy Eg.
The value of P(Eg) can be found from Fermi Dirac probability distribution
function and is given by (e.g. Theraja and Theraja 1988)

1
PðE Þ ¼ EE
exp kT
F
þ1

where k is Boltzmann’s constant, T is the absolute temperature, P(E) is the proba-

bility of finding an electron having any particular value of energy E and EF is Fermi
level.

1 N
∴P E g ¼   ∴nc ¼  
ðEg EF Þ ðEg EF Þ
exp kT þ1 exp kT þ1

The number of electrons in valence band nv is given by

N v ¼ NPð0Þ

where N is total number of electrons in both conduction and valence bands and P
(0) represents the probability of an electron having zero energy in valence band.
The value of P(0) can be found by putting E equal to zero in Fermi Dirac
probability distribution function and is

1 1 N
Pð0Þ ¼ 0E ¼ E : Hence nv ¼ E
exp kT
F
þ1 exp F
kT þ1 exp F
k T þ1

The total number of electrons N ¼ nc + nv.

N N
Or, N ¼   þ E
ðEg EF Þ exp F
þ1
exp kT þ1 kT

E
which on simplification gives E F ¼ 2g .
Extrinsic or doped semiconductors are those intrinsic semiconductors to which
some suitable impurity or doping agent dopant has been added in extremely small
amount—about 1 part in 108. The usual doping agents are pentavalent atoms having
five valence electrons (antimony, arsenic, and phosphorus) or trivalent atoms having
three valence electrons (indium, gallium, aluminium, boron). Depending upon type
of doping material used, extrinsic semiconductors can be divided into two cases:
3.4 Band Gap Structure of Metal, Semiconductors and Insulators 35

BAND ENERGY
Fig. 3.6 Fermi level in N-
type extrinsic Conduction band Fermi level
EF
semiconductor ED
Donor level
Valence band

Fig. 3.7 Fermi level in P-

BAND ENERGY
Conduction band
type extrinsic
Acceptor level
semiconductor
EA
EF
Valence band Fermi level

• N-type extrinsic semiconductor

• P-type extrinsic semiconductor
In N-type extrinsic semiconductor, a pentavalent material like antimony is added
to pure silicon crystal. Each antimony atom forms covalent bonds with the surround-
ing four silicon atoms. The fifth electron can be easily excited from the valence bond
to the conduction band by the application of electric field or increase in thermal
energy. Thus, antimony is called donor impurity and makes the pure silicon an N-
type extrinsic semiconductor. The addition of antimony increases the number of
conduction electrons. Hence, Fermi level shifts towards the conduction band
because of increased concentration of electrons in the conduction band (Fig. 3.6).
In N-type semiconductors electrons are the majority charge carriers while holes
are the minority charge carriers.
In P-type extrinsic semiconductor, a trivalent material like boron is added to a pure
silicon crystal. The three valence electrons of boron atom form covalent bonds with four
surrounding silicon atoms. The fourth silicon atom is unable to form a covalent bond with
boron atom because boron atom does not have the fourth electron in its valence orbit and
gives rise to a hole. Thus boron causes as many positive holes in a silicon crystal as there
are boron atoms thereby producing a P-type extrinsic semiconductor. In this type of
semiconductor, conduction is by the movement of holes in the valence band; hence holes
are the majority charge carriers whereas electrons are minority charge carriers. Since
concentration of holes in the valence band is more than the concentration of electrons in
the conduction band, Fermi level shifts nearer to the valence band and acceptor level lies
immediately above the Fermi level (Fig. 3.7).
Conduction is by means of hole movement at the top of valence band, the
acceptor level readily accepting electrons from the valence band.

3.4.3 Drift, Diffusion and Recombination Currents

Movement of charge carriers occur in semiconductor materials through following

mechanisms:
36 3 Fundamentals of Photovoltaic Generation: A Review

• Drifting of charge carriers under the influence of applied electric field and
• Diffusion of charge carriers from a high charge density region to low charge
density region.
Above 0 K when no electric field to the semiconductor crystal is applied, the
motion of the charge carriers is random and they collide with each other and fixed
ions. Because of this the net average velocity of these charge carriers in any given
direction is zero and hence no current exists in the crystal.
When electric field is applied to the semiconductor crystal, the motion of charge
carriers is directed. Due to this the net average velocity, known as drift velocity, is in
the direction of applied electric field. Since the holes and electrons move in opposite
direction because of their opposite charges but produce a current in the same
direction. The drift velocity is proportional to electric field strength and is given as

v ¼ μE

where v ¼ drift velocity, E ¼ applied electric field strength and μ ¼ proportionality

constant, known as mobility.
The current density due to electron drift is e μe n E where μe is electron mobility
and n electron density.
The current density due to hole drift is e μh p E where p is hole density. The total
current density due to electron and hole drift is

J ¼ e μe n E þ e μh p E ¼ eðn μe þ p μh ÞE

where μe ¼ electron mobility, n ¼ current density, μh ¼ hole mobility, h ¼ hole

density and J ¼ total current density.
Diffusion is gradual flow of charge carriers from a region of high density to a
region of low density due to non-uniform distribution of charge carriers in a semi-
conductor crystal. It is a force-free process and leads to an electric current without
the benefit of an applied electric field. The diffusion of charge carriers is proportional
to the carrier density gradient, and the constant of proportionality is diffusion
constant or diffusion coefficient. The current density due to electron diffusion is

dn
J e ¼ e De and due to hole diffusion
dx
dp
J h ¼ e Dh
dx
2
where De, Dh ¼ electron and hole diffusion constants respectively ofunit, ms , dndx ¼
electron density gradient and dp
dx hole density gradient.
In semiconductor materials, drift and diffusion processes occur simultaneously
and, therefore, the expression for total electron and hole current densities become
3.5 P-N Junction 37

dn Amp
J e ¼ e μe n E þ e De
dx m2
dp Amp
and J h ¼ e μh pE  e Dh
dx m2
Apart from drift and diffusion, a third phenomenon called recombination occurs
in semiconductor crystals. The recombination is a result of collision of an electron
with a hole and the recombination rate is directly proportional to the carrier concen-
tration. It also depends on the carrier life time that is defined as time for which, on an
average, a charge carrier will exist before recombination. Its value varies from
nanoseconds to microseconds. The total carrier flow in a semiconductor is sum of
the three flows as discussed above. The current in the semiconductor is the sum of
the electrons and hole currents.

3.5 P-N Junction

When n-type material (characterized by electrons as majority carriers and holes as

minority carriers) is in metallurgical contact with a piece of p-type material (char-
acterized by holes as majority carriers and electrons as minority carriers) a p-n
junction is formed (Fig. 3.8). This type of junction is commonly encountered in
many semiconductor devices such as transistors, diodes and photodiodes. The solar
cell is a large area photodiode and understanding of this junction is important to
understand the operation of solar cell.
The density of donor and acceptor atoms is equal in the junction plane. The
conduction electrons in N-type region and holes in P-type region are highly mobile
and they possess thermal energy and exhibit kinetic motion that causes them to drift
randomly through the crystal. Consequently, at the junction, conduction electrons
from donor atoms in the n-region diffuse into the p-region and combine with
acceptor atoms, producing a negatively charged impurity atoms. The opposite action
also takes place: conduction holes from acceptor atoms in the p-region diffuse into
the n-region and combining with electrons, producing a positively charged impurity
atoms. The net result of this movement is the disappearance of conduction electrons
and holes from the vicinity of the junction and establishment of reverse electric field
which is positive on N-type side and negative on P-type side. The area in which this
field is set up is called depletion layer (Fig. 3.9).
If a majority carrier, either an electron or a hole, tries to cross into depletion layer
either it can be trapped by the row of fixed ions of opposite sign or it may succeed in
entering the depletion layer where it will be repelled by the row of similar-charged
impurity ions guarding the other region. At the end, an equilibrium condition is
reached when depletion layer has widened to such an extent that no electrons or
holes can cross the P-N junction. Because of this charge separation, an electric
38 3 Fundamentals of Photovoltaic Generation: A Review

Fig. 3.8 P-N junction N - type P - type

Junction

Fig. 3.9 Depletion layer in p-n junction

potential difference is established across the junction known as junction or barrier

potential (Fig. 3.10) which depends on doping density, electronic charge and
temperature. This junction potential, at room temperature, is about 0.3 V and
0.7 V for germanium and silicon respectively (e.g. Theraja and Theraja 1988). For
any junction, the doping density and electronic charge are constant, thus making
barrier potential dependent on temperature. As temperature increases, more minority
charge carriers are generated leading to an increase in drifting across the junction. It
has been found that the barrier potential is decreased by about 2 mV per degree
centigrade for germanium and silicon.
Figure 3.11 shows the energy bands of the trivalent impurity atoms in the P-
region at a slightly higher level than those of the pentavalent impurity atoms in N-
region at the instant of junction formation. This is because of higher core attraction
for valence electrons in pentavalent atoms. Therefore, trivalent valence electrons are
at a higher energy level. However, there is some overlap between respective bands of
the two regions. Due to this reason, some high energy electrons near the top of N-
region conduction band diffuse into the lower part of the P-region conduction band.
As diffusion continues, depletion layer begins to form and energy bands in N-region
shift downward due to loss of high energy electrons.
When the top of the conduction band in N-region is at the same level of bottom of
conduction band in P-region an equilibrium condition is reached with establishment
of barrier potential. For an electron to diffuse across the junction, it must receive
energy from the external source.
When the P-N function is forward biased, junction offers a low resistance and
permits easy flow of electrons across the junction. In the case of reverse biased, P-N
junction offers high resistance and no current flows across the junction.
3.6 P-N Junction Solar Cells 39

Fig. 3.10 Junction

potential in p-n junction

P - side N -side
Conduction band
Eg
Conduction band
Fermi level
Energy

Fermi level
Valence band

Valence level

Depletion
Region

Distance

Fig. 3.11 Energy band diagram in P-N junction

3.6 P-N Junction Solar Cells

A solar cell is a large area junction device made from a thin (250-400 μm) wafer cut
from a pure silicon crystalline block. In the cell fabrication process a p-type doped
silicon wafer 300 μm thick is taken and the phosphorous (having five valence
electrons) is diffused into the top region of slice such that about 0.5 μm of silicon
wafer is converted into n-type which leads to the formation of p-n junction
(Fig. 3.12).
A current collected grid is connected to the top and bottom ends of the solar cell to
enable the external characteristics. The front metal contact is in the form of a narrow-
fingered grid, while the back contact usually covering the entire back surface. The
front contact surface has an antireflective coating to keep the photons going into the
crystal instead of bouncing back. Titanium oxide and tantalum pentoxide are the
materials being used as an antireflective coating. It can be square, square with
cropped corners, circular or semicircular. Nowadays it is usually 100 mm square
or 100 mm diameter (Fig. 3.13).
40 3 Fundamentals of Photovoltaic Generation: A Review

Negative
contact grid

Antireflective
coating

n-type
silicon
Back p-n junction
contact P-type
silicon

Fig. 3.12 P-N junction solar cell

Fig. 3.13 100 mm circular and square solar cells

When light falls on front surface, the photon energy in the excess of band gap
interact with valence electron and lift them to conduction band. This movement
leaves behind holes; so each photon is said to generate an electron-holes pair.
In crystalline silicon, electron-hole generation takes place throughout the thick-
ness of the cell in concentration depending upon the irradiance and spectral compo-
sition of the light. The electrons and holes diffuse through the crystal in an effort to
make an even distribution. Some recombine after a lifetime of the order of 1 ms
neutralizing their charges and giving up energy in the form of heat. Others reach the
junction before their life time has expired where they are separated by the reverse
field, the electrons being accelerated towards the negative contact and holes towards
positive contact. If the cell is connected to a load (Fig. 3.14), electrons will be pushed
from the negative contact through the load to the positive contact, where they will
recombine with holes. This constitutes an electric current.
3.6 P-N Junction Solar Cells 41

Fig. 3.14. p-n junction

solar cell connected to a load
Chapter 4
Wafer-Based Solar Cells: Materials
and Fabrication Technologies

4.1 Photoionic Processes in Solar Cells

Solar cells convert solar energy into electrical energy by photoionic process referred to as
photoelectric effect. The solar radiation may be considered as discrete bundles of energy
called photons. The energy (E) of photons associated with radiation of frequency (ν) and
wave length (λ) may be expressed as E ¼ hν ¼ h c\λ where h is Planck’s constant and c is
the velocity of light. The radiation intensity is proportional to photon number density. As
discussed in Chapter 2 the solar spectral irradiance ranges between the wave length
region of 0.2–2.0 μm and has maximum invisible light region at 0.48 μm.
With the solar radiation, photoelectric effect is possible in a number of materials but
the materials that have shown best performance in sunlight are the semiconductors. In a
direct absorber semiconductor material, the incoming photons and electrons interact
directly. The photons of energy (Ep) greater than the forbidden energy gap (Eg) are
completely absorbed and the electrons are transferred to conduction band. The excess
energy (Ep  Eg) is given up to lattice as thermal energy. For photons with energies less
than Eg, the material will appear transparent (Backus 1977). Such absorption is not
readily attained in indirect absorber semiconductors wherein the photon interacts with the
electron as well as phonons from the lattice to provide momentum change. Silicon and
germanium are indirect absorbers. Doped crystals of these materials are, therefore, used.
They are of two types: (a) n type which is doped with an impurity such as phosphorous of
5 valence electrons and (b) p type, which is doped with an impurity such as boron of
3 valence electrons. After the photons are absorbed and electron hole pairs are created, the
charges are to be separated. In solar cells the necessary electric field is provided by a
junction formed by alloying or diffusion processes. The junction can be made inside the
same semiconductor and is then referred to as homojunction in contrast to a
heterojunction made between two different semiconductors such as CdS and
Cu2S. The junction can also be made between a metal and semiconductor, e.g. Si and
Ag. Yet another approach due to Gover and Stella (1974) suggest a new geometry of
multi junctions for solar cells.

© Capital Publishing Company, New Delhi, India 2018 43

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_4
44 4 Wafer-Based Solar Cells: Materials and Fabrication Technologies

In the junction region free electrons of n type semiconductor neutralize holes of

p type region. Consequently n region interface is left with excess positive charges
and p region with negative charges. These charges give rise to an electric field
(depletion layer) at the p-n interface. If the photon-produced minority carriers (holes
in n type and electron in p type) are able to diffuse to depletion layer region before
they recombine, they are swept across the junction and separated giving rise to a
photo voltage.

4.2 Solar Cell Materials

In the pioneering work at Bell Telephone laboratory on solar cells, Chaplin et al.
(1954) used crystalline silicon to fabricate a diffused junction solar cell and Reyn-
olds et al. (1954) made a similar breakthrough with a cadmium sulphide device.
Following this, several semiconductors have been tried for the fabrication of solar
cells. These include silicon (crystalline, polycrystalliane, amorphous), germanium,
gallium phosphide, gallium arsenide, gallium antimonide, indium phosphide, indium
arsenide, indium antimonide, zinc oxide, zinc sulphide, cadmium sulphide, cadmium
selenide, lead sulphide, lead selenide and cadmium telluride. The PV materials of
commercial potential are given in Table 4.1 (Kazmerski and Emery 1992; Kaushika
1996). The efforts in cell fabrication technology are summarized in Fig. 4.1.
These semiconductor materials have intermediate conductivities (1–10 mho/m)
between insulators (e.g. fused quartz has conductivity less than 2  1017 mho/m)
and good conductors (e.g. silver has conductivity about 2  107 mho/m). This
property results from the covalent bonding structure of their crystal lattice. Silicon
has a chemical valence of 4 which implies that out of its 14 electrons, four are
available to interact with other atoms. The regular structure of silicon crystal is such
that each atom is firmly held in the crystal lattice by sharing its valence electrons
with each of surrounding four neighbours.

Table 4.1 The PV materials with their efficiencies

Typical cell voltage at Typical cell current at short Efficiency
Material open circuit (volts) circuit (mA/cm2) (%)
Mono-crystalline 0.6 35 14–15
Poly crystalline 0.6 35 12–13
Gallium arsenide 1.0 27 24
Tandem amorphous sil- 1.8 10 25
icon (Two cells)
Amorphous silicon 0.9 15 9–10
Copper indium 0.4 35 10
diselenide (CIS)
Cadmium sulphide, 0.7 25 10
Cadmium telluride
4.2 Solar Cell Materials 45

For a III-IV semiconductor (e.g. gallium arsenide GaAs) the electron pair is
provided by atoms of 1:1 mixture of III and IV column elements which behaves
like an average lattice atom that contributes four valence electrons. The same
argument applies to II and VI semiconductor e.g. cadmium sulphide. This situation
is similar to that of an insulator and is in contradiction to that of metals where many
electrons are free to move about to constitute an electric current upon the application
of electromotive force. In terms of energy band concept or the band theory of
materials, the energy of electrons can be represented by a one dimensional energy
diagram consisting of allowed and forbidden bands. Conduction bands are empty in
insulators and have a sea of electrons in good conductors. In semiconductors the
forbidden gap (the gap between conduction and valence bands) is much narrower.
Energy gaps of some common semiconductors are shown in Table 4.2.

Fig. 4.1 Cell fabrication technologies (Kazmerski and Emery 1992; Kaushik 1996)
46 4 Wafer-Based Solar Cells: Materials and Fabrication Technologies

Table 4.2 Energy gaps of Material Energy gap (eV)

some semiconductors
Crystalline silicon 1.12
Amorphous silicon 1.65
Germanium 0.66
Gallium phosphide 2.26
Gallium arsenide 1.42
Indium phosphide 0.92/1.35
Indium arsenide 0.36
Cadmium sulphide 2.6
Cadmium telluride 1.4

Some electrons may receive enough energy from sunlight or thermal agitations to
jump across the influence of an electric field. The material in which the conduction is
due to electrons excited from valence band to conduction band is called intrinsic
semiconductor. In this case only a limited portion of the solar energy can be used to
excite electrons, so there is a theoretical limit on the efficiency of photovoltaic
conversion. The efficiency of photovoltaic conversion is illustrated as follows:
The energy flux Eλ represents energy per unit area per unit time per unit
wavelength interval at wavelength λ. So if ν is frequency of a spectral component
of radiation and h is plank’s constant (6.6260755  1034 Ws2), the energy of each
photon is hν and the number density of photon (Nν) in terms of solar spectral density
Eν of electromagnetic radiation is given by

Eν
Nν ¼
hν
So, the number of photons in the spectral range (dν) is represented by

dN ¼ N ν dν
Eν dν
or, dN ¼
hν
The solar ionization process in solar cell requires minimum photon energy
(Eg ¼ h νg) where Eg is band gap energy of semiconductor material. The solar
Rνg R1
spectral part of E ν dνor E λ dλ is not being used in photon ionization and gets
0 λg
converted into heat.
Only the fraction of solar energy input used for carrier generation is in the
frequency range greater than νg. So the number of photons participating in carrier
generation is Np, which is the integral of dN from νg to 1 i.e.
4.3 Silicon Solar Cell Wafer Technology 47

Z1
Eν dν
NP ¼
hν
νg

The fraction of solar energy input (Ep) actually used for charge carrier generation
is EgNp corresponding to number of photons with energy >Eg. So we have

Z1
E ν dν
Eg N p ¼ Eg
hν
νg

The fraction of solar energy used for photon ionization is, therefore, given by

R1 E ν dν
Eg hν
νg
ηc ¼ R1
E ν dν
0

This efficiency depends on the band gap of semiconductor material and the temper-
ature. Silicon has an energy gap of 1.08 eV; hence, sunlight with wavelength shorter than
1.15 m can only excite electrons to conduction band. This portion of sunlight contains
about 78% of total energy. Of this, only a fraction equal to 1.08 eV divided by the energy
of incident photons is used to overcome the gap energy while the rest is dissipated as heat.
This yields about 42% as maximum efficiency of silicon solar cells (McDaniels 1979).
Based on the above consideration, a calculated curve of maximum efficiency as a
function of energy gap of semiconductors is shown in Fig. 4.2.

4.3 Silicon Solar Cell Wafer Technology

The considerations of input solar spectrum, operating temperature and availability of

semiconductor grade material and its processing like cutting, polishing, alloying or
diffusion and contact making etc. have often favoured silicon for making solar cells.
Consequently, silicon solar cells have reached the state of commercial realization.
The technology of silicon solar cell wafer fabrication comprises the following steps
(Green 1979):
(i) Sand to metallurgical grade silicon,
(ii) Metallurgical to semiconductor grade silicon,
(iii) Semiconductor grade silicon to single crystal silicon wafers,
(iv) Silicon wafers to solar cells, and
(v) Electrical contacts and encapsulation.
48 4 Wafer-Based Solar Cells: Materials and Fabrication Technologies

Fig. 4.2 Efficiency vs. band gap energy

4.3.1 Sand to Metallurgical Grade Silicon

In the commercial process, quartzite, an ore of silicon dioxide obtained from earth’s
crust is reduced in large arc furnaces by carbon (in the form of wood chips, coke and
coal) resulting in silicon.

SiO2 þ 2C ! Si þ 2CO ð4:1Þ

Silicon is poured into shallow troughs from the furnace and blown with oxygen or
oxygen/chlorine mixture for further purification.
Its purity is around 98–99%, iron and aluminum being the major impurities.

4.3.2 Metallurgical to Semiconductor Grade Silicon

Metallurgical grade silicon is converted into a gaseous compound which is con-

densed and refined by fractional distillation. Ultra pure silicon is extracted from the
refined product.
A bed of fine metallurgical grade silicon particles is fluidized with HCl in the
presence of a Cu catalyst to promote the reaction.

Si þ 3HCl ! SiHCl3 þ H2 ð4:2Þ

The gases are passed through a condenser and the resulting liquid is subjected to
multiple fractional distillation to produce semiconductor grade SiHCl3
4.3 Silicon Solar Cell Wafer Technology 49

(trichlorosilane). The SiHCl3 is reduced by hydrogen by heating mixtures of the

gases. This results in deposition of silicon in polycrystalline form on an electrically
heated rod.

SiHCl3 þ H2 ! Si þ 3HCl ð4:3Þ

This step of producing polycrystalline silicon increases the production cost and
needs improvement in technology for cost reduction.

4.3.3 Polycrystalline Silicon to Single Crystal Silicon Wafers

For fabrication of semiconductor devices, silicon must be both pure and be in a

crystalline form with minimum number of defects in crystal structure. This is
achieved by a commercial process often referred to as Czochralski process.
The polysilicon is melted in a crucible after addition of trace levels of one of the
dopants. This is generally boron, a p type dopant, for silicon solar cells. With good
temperature control, using a seed crystal, a large single crystal of silicon is pulled out
from the melt in cylindrical form called ingots. The n or p type single crystal ingots
are now commercially available.
The crystal ingot is sliced into thin wafers of about 300–500 micron thickness.
This process results in more than 50% waste as cutting loss.

4.3.4 Fabrication of Silicon Solar Cells

Silicon slices, cut parallel to (1,1,1) plane from 1 ohm-cm n or p type single crystal
ingots, are used in the fabrication of solar cells, for example Gangadhar et al. (1967),
on the laboratory scale fabrication of silicon solar cells in India. These cells are now
commercially manufactured. In what follows is a discussion based on Gangadhar
et al. (1967).
The thickness of the wafers is 0.5 mm which is adequate for absorption of
majority of sunlight useful in photovoltaic energy conversion. The shorter wave-
lengths of solar radiation are absorbed in the surface region of the wafer whereas the
longer wavelengths penetrate deeper. The lifetime of minority carriers is then
measured. Silicon wafers in which minority carrier lifetime is more than five
microseconds are only used. The minority carrier lifetime is measured by the well
known method of conductivity modulation in a point contact. If the minority carrier
lifetime is low the charges generated by photons away from the junction are lost by
recombination before reaching the junction.
The slices are lapped and polished to mirror finish on both the surfaces. This is
followed by a 4–5 min etching in a mixture containing six parts of fuming nitric acid
and one part of hydrofluoric acid. By keeping etchant temperature at about 15  C and
50 4 Wafer-Based Solar Cells: Materials and Fabrication Technologies

employing mechanical agitation, localized etching is minimized and a reproducible

and excellent surface finish is achieved. The slices are rinsed in deionized water and
degreased in a trilyne vapour degreaser designed for the purpose. The slices are
preserved in acetone till diffusion. Fabrication of solar cells is carried out either by
diffusing phosphorous on p type wafer or by diffusing Boron on n type wafer.

4.3.4.1 Diffusion of Phosphorus

Phosphorous is diffused in p type silicon slices using P2O5 as the source, using dry
oxygen as carrier gas. Three specimens were used in each run. It consists of a silica
tube passing through two controlled temperature zones with such temperature
gradient between them as not to let the impurity condense in it. P2O5 was placed
in the lower temperature zone, at a temperature of 220–250  C determined by the
volatility of the impurity. The vapours of P2O5 were carried with the help of dry
oxygen over silicon slices maintained at some fixed temperature between 950 and
1020  C (with the variation in temperature not exceeding 2  C). The slices were
kept vertical in the boat to allow P2O5 vapours to react uniformly with both the
surfaces. After about 5 min the boat carrying P2O5 was withdrawn and the oxygen
flow continued till no more vapours were seen at the exhaust. Phosphorus was
diffused in silicon slices for 45–60 min. The slices were then shifted to a temperature
zone of 700  C and were allowed to cool gradually.

4.3.4.2 Diffusion of Boron

In this process, boric acid (H3BO3) was used as an impurity source in n type silicon
material. Boric acid was placed in the vicinity of the slices, maintained at a fixed
temperature between 1000 and 1150  C (with variation in temperature not exceeding
2  C). Nitrogen was used as the carrier gas and the quantity of moisture in air was
critically controlled. It was essential to protect the surface from damage and to
control the rate of diffusion of boron in silicon. The source material was removed
from the diffusion zone after 5 min. Impurity was allowed to diffuse for 45–60 min
which was followed by cooling of slices in a way similar to that during phosphorus
diffusion.

4.3.5 Electrical Contacts and Encapsulation

In order to remove any oxide layer formed on the surface, slices obtained by either of
the above mentioned processes were etched for 5 min in 48% hydrofluoric acid. This
was followed by thorough rinsing in deionized water. The junction at the rim and the
4.3 Silicon Solar Cell Wafer Technology 51

back of these slices was removed by lapping. The active surface was coated with
some photo resistant paint leaving out gridded structure of optimum size; rim was
fully covered with the paint. In the literature, the use of Kodak Photo Resist (KPR) is
extensively recommended for the above purpose. However, KPR is an imported
material. In India, the use of an alternate material known as Haunsa Resist/SS is
recommended. This paint is not affected by acids and plating solutions and is readily
soluble in acetone. The constituents of nickel plating solution are (gms/litre): nickel
chloride, 30; sodium hypophosphite, 10 and ammonium chloride, 50.
After nickel deposition the slices were washed and Haunsa Resist/SS was cleaned
off the surface. Wire connections were made using a special composition Wood’s
metal and tin with zinc chloride as flux. The cell was then encapsulated such that the
top surface of the container helped in concentrating the sun’s energy on the cell, thus
increasing the effective output power.
The rectangular solar cells may have a line contact parallel to the longer edge.
Such contact geometry results in increased probability of recombination for the
charge carriers at larger distances from the contact. To improve collection of carriers,
more interconnected contacts (grids) are provided on the active surface of the cell.
Such a grid pattern leads to a lower contact resistance. The efficiency of such a cell
depends on the width of the grid lines and their separation. While the grid lines
reduce series resistance, they also reduce the active area available for incident light,
thereby reducing the output power. Narrower grid lines reduce useful output power.
If the separation of grid lines is too much, charge carriers would be lost by
recombination, reducing the output power again. In practice, a compromise is
made between these factors and an optimum spacing S0 and width T00 of the grid
lines are calculated. Knowing the terminal voltage V near the maximum power point
and the other data of the solar cell without grid structure, the values of grid structure
can be calculated from the equations for the grid structure (Wolf 1960)

ρT  1=2
3=4
T ¼ 25=4 1=2
BCj0 eBV W 3=2 ð4:4Þ
ρp
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
3 2T 2T
S¼  ð4:5Þ
BCρp j0 eBV 3
 
where C ¼ 1  I L ðe
I0 BV
 1Þ, B ¼ AKT
q
, ρT ¼ the sheet resistance of the contact
strip, ρp¼the sheet resistance of the diffused p layer, j0 ¼ the dark saturation current
density, IL ¼ the light generator current and I0 ¼ the saturation current.
It is apparent from Eqs 4.4 and 4.5 that for higher light intensities the width of
the grid lines should increase and their separation should decrease. Solar cells
have highly polished surfaces and so the contacts made by electrodeless depo-
sition of nickel or vacuum evaporation of gold do not adhere properly to the cell.
The adherence of contacts is improved by sintering to form an alloy with silicon.
52 4 Wafer-Based Solar Cells: Materials and Fabrication Technologies

During sintering, the junction is penetrated, seriously degrading the cell char-
acteristics. Silver-titanium contacts are used to avoid junction penetration
(Smith et al. 1963). Silver-cerium contacts have also been developed for super
blue cells.
About 30% of the incident light is reflected from the surface of silicon solar cells
which have highly polished surfaces. Antireflection coatings are an obvious solution
to reduce the reflection losses of useful light from the cell surface. Silicon monoxide
has the desired properties for use as antireflection coating and it serves excellently
well on bare silicon cells. An antireflection film of SiO is vacuum evaporated on to
the surface of the solar cell. The quality of the film is checked by looking at the
interference colours in diffused light after suspending these cells for 30 min in steam
over boiling water. Adherence of the film is tested by the scotch tape test (Smith et al.
1963).
Because of the negative temperature coefficient for power output of the silicon
solar cells, it would be desirable to keep the temperature of the cell as low as
possible. Bare cells have an emittance of about 0.32 at room temperature,
whereas for borosilicate glass it is 0.85. Thus, with a borosilicate glass slide
kept in thermal contact with the cell the effective emittance of the cell would
increase appreciably and the rise of cell temperature would not be as high. The
use of a cover glass on uncoated cell leads to a slight improvement in cell
performance.

4.3.6 Improved Solar Cell Structures and Techniques

Several solar cell materials and fabrication steps have been researched for improve-
ment in conversion efficiency. The conversion efficiency of laboratory solar cells has
increased from 17.1% (Milstein and Tsno 1984) to 24% (Wang et al. 1990). The
production cells have always been approximately 5% less efficient (in absolute
efficiency) than laboratory cells. The efficiency difference is mainly due to the
difference in area of wafers used in production (50 cm2) and laboratory cells (4 cm
2
). Corresponding metallization coverages are 10–15% and 2–3% and hence the
difference in efficiencies. Most of the cell efficiency improvements have originated
from improved cell structures and processing techniques. Initial improvement was
obtained by reducing the recombination along the top cell surface using thermal
oxide passivation; the improvements were augmented by the reduction in bulk
recombination resulting from an alloyed aluminum rear contact. The structure may
be referred to as Passivated Emitter Solar Cell (PESC) (Blakers and Green 1986).
Further improvements in cell performance were obtained in the new cell structure,
the Passivated Emitter and Rear Cell (PERC) (Blakers et al. 1989; Green et al. 1990).
The main structural difference arises at the rear contact wherein a thermally grown
oxide is used to electronically passivate most of the rear cell surface. The rear ohmic
4.4 Indian Status 53

connection is made through the holes in passivating oxide. However, the resultant
aluminum-silicon contact not only has a high contact resistance but also exhibits
rectifying properties. The deficiency has been overcome by locally diffusing boron
into contact areas at the rear of the cells. The resultant structure is referred to as
Passivated Emitter Rear Locally (PERL) diffused cell (Wang et al. 1990). It exhibits
24% efficiency. The structure incorporates reasonably effective light trapping
schemes. There seems to be some scope of improvement in this area.
The technology of fabrication, comprising large number of steps starting from
quartzite to encapsulated silicon solar cells, had also undergone several develop-
ments over years. Spacing and width of contact grid lines are optimized to
(i) increase active area for incident light,
(ii) reduce contact resistance and
(iii) improve collection of charge carriers.
Antireflection coatings are employed to reduce reflection losses of incident light
from the cell surface. Single crystal silicon solar cells have a higher conversion
efficiency compared to the less expensive polycrystalline and amorphous cells.
Other technological improvements include better cell structures and processing
techniques. Passivated emitter solar cell, passivated emitter and rear cell, and
passivated emitter rear locally diffused cell have resulted from these improvements
exhibiting a conversion efficiency of around 24%.

4.4 Indian Status

In India, mono-crystalline silicon solar cell was fabricated on laboratory scale at

Indian Institute of Technology Delhi during 1967 (Jain et al. 1968). The cell active
area was 2.4 cm2 and the illumination level was 80 mW/cm2. The conversion
efficiency was around 10%. Solar cell fabrication has since reached a commercial
status and several manufacturers are now marketing single crystal solar cells. Some
of the physical and material characteristics of solar cell manufactured by Central
Electronic Ltd. are as follows:
A. Material: Single crystal silicon
Resistivity range: 0.5–3.0 ohm-cm
Orientation: h1, 0, 0i
Type: p-type boron doped
Thickness: 400 μm
Cell area: 78.54 cm2
B. Process
Surface preparation: Chemical etching 20–30 μm silicon
Texturisation: Low conc., alkaline solution by preferential etching
54 4 Wafer-Based Solar Cells: Materials and Fabrication Technologies

C. Diffusion
Gaseous diffusion: Phosphorus (POCl3)
Junction depth: 0.5 μm
Surface concentration: 1020 atoms/cm2 graded
Temperature of 900  C
diffusion:
D. Front screen printing: Silver paste + drying
E. Back screen printing: Silver + Aluminum paste + drying + firing at
700  C
Chapter 5
Mathematical Model of Transport Processes

5.1 Physical Model of Solar Cells

The physical processes that determine the role of a solar cell as an energy converter
include absorption of light leading to generation of electron-hole pairs, separation of
charges in depletion region, surface and bulk recombination and transport of charge
carriers by diffusion and drift. Over the years efforts have been made to understand
these processes (e.g. Fahrenbruch and Bube 1983; Green 1982) and ameliorate the
efficiency of solar cells.

5.1.1 Excess Carrier Generation: Photo Generation

It is possible to generate excess carriers in the p- and n-regions by illuminating them

with light. A flux of monochromatic light If (Watts/cm2) incident on the surface of
the p-region is considered. If the wavelength of light is λ (cm), the number of
photons F 00 (photons/cm2 sec) striking the surface will be given by

If λI f
F 00 ¼ ¼ ð5:1Þ
hν hc
where h is Planck’s constant, and ν and c are frequency and velocity of light,
respectively.
Part of this light is reflected from the surface and the remaining is transmitted. If
the reflection coefficient of the surface is R, the transmitted flux is

F 0 ¼ F 00 ð1  RÞ ð5:2Þ

© Capital Publishing Company, New Delhi, India 2018 55

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_5
56 5 Mathematical Model of Transport Processes

These photons are absorbed in accordance with Lambert’s law which states that
the absorption of light decays exponentially with distance into the material. The flux
at a distance x from the surface is given by,

F ¼ F 0 eðαxÞ photons ð5:3Þ

where α is the absorption coefficient of light of wavelength λ in the semiconductor.

The plot of α vs λ for silicon obtained from the available data is given in Fig. 5.1.
A volume of the plane bounded by x and x + Δx (Wolf 1969) is considered.
Assuming the quantum yield to be unity (i.e., an electron-hole pair is created for each
photon absorbed), the number of electron-hole pairs per cm3 per sec, injected
between the planes at x and x + Δx is given by


F 0 eðαxÞ  eðαðxþΔxÞÞ
ð5:4Þ
Δx
Making Δx approach zero, the rate of generation of electron-hole pairs at a
distance x from the surface is given by

Eph [eV]

3.5 3.0 2.5 2.0 1.5 1.3 1.2

106

105

104
α ( cm-1)

25°C

103 −200°C

Conversion
102 1.237
l (cm) =
Eph (in eV)

10
0 0.2 0.4 0.6 0.8 1.0 1.2
λ (μm)

Fig. 5.1 Absorption coefficient (α) vs wavelength (λ) in single crystalline silicon (Wolf 1969)
5.1 Physical Model of Solar Cells 57

 
g ¼ F 0 α eðαxÞ pairs=cm3 sec ð5:5Þ

The photons incident on the cell can be subdivided into five categories as
described below:
(i) A portion of the incident light is reflected away. Some of the photons absorbed
near the surface generate electron-hole pairs which recombine and do not reach
the junction.
(ii) Photons absorbed near the junction, in both p- and n-regions, generating
electron-hole pairs which are separated by the built-in field near the junction
give rise to useful power, contributing to the conversion efficiency of the cell.
(iii) Photons absorbed deep in the collector region generate electron-hole pairs, so
far from the junction that they recombine and, hence, are not useful.
(iv) Photons with insufficient energies to generate electron–hole pairs are absorbed
and generate heat.
(v) Finally, some of the photons having longer wavelengths are not absorbed in
either base or collector.
Furthermore, if the cell thickness is less, some of the photons which can generate
electron-hole pairs are not absorbed, causing a lower output current. It is found that
0.4–0.5 mm thickness of silicon is adequate to absorb almost all the incident solar
photons that generate electron-hole pairs which contribute to the power output of the
cell. Fabrication of cells of greater thickness will only increase the number of
photons absorbed away from the junction, which would not contribute to the
power output of the cell. Also, longer wavelengths which have insufficient energy
to generate electron-hole pairs are absorbed and generate heat, thereby increasing the
cell temperature which in turn leads to a drop in cell power output (0.45% per degree
Kelvin).

5.1.2 Built-in Field

The diffused p-n junction solar cell (Fig. 5.2) is characterized by a

non-homogeneously doped (assume p-type base) over a constant impurity back-
ground of opposite conductivity referred to as collector. The non-homogeneous
doping in the base region is the result of the fabrication of solar cell by impurity
diffusion. The p-n junction solar cell under consideration is formed by the diffusion
of p-type (acceptor) impurities into an n-type wafer containing n-type (donor)
impurities. The impurity profile in the p-region is described by (i) a Gaussian
function if the diffusion is carried out from an infinitesimally thin layer of dopant
and (ii) a complementary error function, when the diffusing species is made avail-
able to the surface to maintain the concentration at a constant value (Jain and
Al-Rifai 1967). It may be mentioned that the surface concentration of impurity at
58 5 Mathematical Model of Transport Processes

N0
17
10

Na (x)

1016

NB
1015

p
n

0 w

Fig. 5.2 Diffused p-n junction solar cell

the end of diffusion is determined by the time and temperature of diffusion, in the
former cases while in the latter case it is limited by the solubility of the diffusion
impurity at the temperature at which diffusion is carried out.
The resultant impurity profile results in a built-in electric field. This field influ-
ences the carrier transport in the diffused base region and hence it is to be considered
in carrier transport process in this region. It is a common practice to approximate the
impurity profile in the diffused region by a negative exponential. Such an approx-
imation leads to a constant built-in field in place of a spatially varying built-in field
resulting from a realistic profile described by complementary error function or
Gaussian function. The exponential approximation of a constant built-in field in
base region results in considerable simplification in the mathematical analysis of the
physical processes (Gangadhar and Bhattacharya 1968; Fahrenbruch and Bube
1983) for photo diodes and solar cells. The expressions for built-in field for various
impurity profiles can be derived as described below. The concentration in the
diffused layer of the solar cell, arise from N0 at the surface to NB at the junction at
x ¼ w as shown in Fig. 5.3.
The width of the p-region, over which there is a grading of the impurities, is
designated by w (cm). The implications of the non-homogeneous impurity distribu-
tion will be considered presently. For the following discussions it is assumed that
5.1 Physical Model of Solar Cells 59

Fig. 5.3 Impurity profile N0 Exp. Profile

and piecewise exponential 17 CERF Profile
approximation 10
Piecewise
exponential
approximation
Na (x) (l = 5)

1016

NB
1015

p n

0 W

(i) The material is non-degenerate and the Einstein relationship

kT
D¼ μ ð5:6Þ
q

is valid. D and μ are the diffusion coefficient and mobility of charge carriers
respectively.
(ii) The impurities are completely ionized at room temperature.
Hence

pð x Þ ¼ N a ð x Þ

where p and Na are position dependent concentrations of holes and the acceptor
impurities.
Due to the non-homogeneous impurity distribution and the consequent majority
carrier distribution, by Fick’s law, there is a natural tendency for diffusion. The
diffusion current is given by

dpðxÞ
qDp ¼ ð5:7Þ
d ð xÞ
60 5 Mathematical Model of Transport Processes

where Dp and dp(x)/dx represent the diffusion constant and the concentration
gradient of holes, respectively. This diffusion current is, at equilibrium, balanced
by the drift current which is given by

qμp EpðxÞ ð5:8Þ

when μP is the hole mobility and E (V/cm) is the built-in field, assumed positive from
surface to the junction, in the p-region. Equating (5.7) and (5.8), we have

DP 1 dpðxÞ
E ð xÞ ¼ ð5:9Þ
μP pðxÞ dx

and in view of Eq. (5.6).

kT 1 dpðxÞ
E ð xÞ ¼ ð5:10Þ
q pðxÞ dx

Equation (5.10) shows that the built-in field is determined by the impurity
distribution in the p-region. It is of practical importance to find the built-in field
for different impurity distributions encountered in practice.
(i) Gaussian impurity profile
For the impurity profile of the form:
x2
N a ðxÞ ¼ N 0 e4D∗ t ð5:11Þ

where D* is the diffusion constant (cm2 sec1) determined by the impurity and the
temperature of diffusion which is carried out for a period of t seconds. N0 and Na(x)
represent the impurity concentrations (cm3) at the surface and at any point x in the
diffused p-region.
The built-in field, in view of Eqs (5.10) and (5.11), is given by

kT x
E ð xÞ ¼  ð5:12Þ
q 2D∗ t

(ii) Complementary error function profile

For the impurity profile of the form:
 
x
N a ðxÞ ¼ N 0 erfc pffiffiffiffiffiffiffiffi ð5:13Þ
2 D∗ t
The built-in field, in view of Eqs (5.10) and (5.13), is given by
5.2 The Continuity Equation 61

x2
kT 1 e4D∗ t
E ð xÞ ¼ pffiffiffiffiffiffiffiffiffiffiffi
ð5:14Þ
q πD∗ t erfc pxffiffiffiffiffiffi
∗
2 D t

(iii) Exponential profile

For an impurity profile of the form:

pxffiffiffiffiffi
N a ð xÞ ¼ N 0 e 2 D∗ t ð5:15Þ

The built-in field, in view of (5.10) and (5.15), is given by

kT 1 dN a ðxÞ kT 1
E¼ ¼ pffiffiffiffiffiffiffiffi ð5:16Þ
q N a ðxÞ dx q 2 D∗ t

The negative sign in Eqs (5.12), (5.14) and (5.16) shows that the field (assumed
positive) is directed from junction to the surface, i.e., in a direction favourable for the
minority carrier flow towards the junction. It is noted that the built-in field given by
Eqs (5.12) and (5.14) is a function of distance into the diffused p-region. For an
assumed exponential distribution, the built-in field given by Eq. (5.16) is constant for
the whole of the p-region. Assuming NB ¼ 1016 cm3, N0 ¼ 1018 cm3 and
w ¼ 0.5 μm, the built-in fields given by Eqs (5.12), (5.14) and (5.16) are calculated.
Figure 5.4 shows the field and its variation in the diffused region for the above
distributions. It is seen that for a Gaussian profile, the field is found to have a linear
variation, while for the complementary error function (CERF) distribution, the field
could be linearized incurring an error within 5%. It is also seen that (i) built-in field
due to a Gaussian impurity profile has a zero value at the surface, (ii) the Gaussian
profile has a larger field gradient compared to the corresponding CERF case, (iii) the
field due to Gaussian profile being lower near the surface, increases rapidly towards
the junction and exceeds the field due to the later at and near the junction than that
due to the CERF distribution, and (iv) the constant field due to an assumed
exponential profile has, in the diffused layer, field values greater near the surface,
nearly equal midway and less than, near the junction, compared to the corresponding
values resulting from the other two impurity profiles considered.

5.2 The Continuity Equation

The one dimensional continuity equation for electrons in the p-region of the p-n
junction solar cell can be written as (Gangadhar et al. 1967).
62 5 Mathematical Model of Transport Processes

Fig. 5.4 Field and its 5000 N0

variation in the diffused = 102
NB
region
w = 0.5 mm

4000

E (V/cm) 3000

2000

1000
Gaussian
CERF
EXPN

0
0 0.1 0.2 0.3 0.4 0.5
X (μm)

 
d2 n dn n  npx
Dn 2 þ μn E  þ F 0 α eðαxÞ ¼ 0 ð5:17Þ
dx dx τn
where Dn, μn and tn are the average values of the diffusion coefficient, mobility and
lifetime of electrons; npx ¼ ni2/N(x), and ni, n and npx are the intrinsic, total and
equilibrium concentration of electrons (cm3), respectively, in the p-region. Due to
non-homogeneous doping, the mobility which depends on the value of impurity
concentration becomes position dependent in the diffused p-region. The variation of
mobility with impurity concentration (doping) is shown in Fig. 5.5 (Wolf 1969).
Also, E, the built-in field has a spatial variation, as seen from Eqs (5.7) and (5.9). A
general analytical solution of Eq. (5.17) for an arbitrary impurity profile taking both
mobility and built-in field variations into account is difficult.
The methods described in the literature consist of those which consider, for the sake of
mathematical simplicity, the impurity distribution in the diffused region to the exponential
resulting in a single built-in field value and also employ a suitably averaged value of
mobility (Jordan and Milnes 1960; Wolf 1963; Gangadhar and Kaushika 1992a). In what
follows, the different methods for solving the continuity equation are reviewed.
5.3 Solution of Continuity Equation 63

Fig. 5.5 Electron and hole mobility vs impurity concentration in silicon (Wolf 1969)

5.3 Solution of Continuity Equation

In constant field method/single exponential approximation method, the impurity

distribution in the diffused region is assumed to be exponential, giving rise to a
constant built-in field in it and its value is determined by the surface to background
impurity concentration ratio and the width of the diffused region. It is also assumed
that the mobility of minority carriers can be represented by a suitably averaged value
for the whole of the non-homogeneously doped p-region. The p-region is divided
into a number of sections and the mobility is averaged in each section. The average
mobility for the whole of the p-region is obtained by taking the arithmetic mean of
these averaged values. It is noted that the division of the p-region into about 20 parts
gives a reasonably accurate estimate of the average mobility. The single exponential
model is only an approximate one. This is because the practical impurity profiles
described by a complementary error function of a Gaussian function give rise to a
spatially varying built-in field. Also, the use of a single averaged mobility value for
charge carriers does not adequately represent the practical situation wherein the
mobility is position dependant.
64 5 Mathematical Model of Transport Processes

The p-n junction solar cell is considered to have a thin diffused p-type base with a
exponential acceptor impurity profile and a long n-type collector having uniform
donor concentration. The governing one-dimensional continuity equation for the
minority carriers in the p-type region, with a photon flux F 00 of monochromatic light
incident on the front surface of the cell is given by Gangadhar and Bhattacharya
(1968):

d2 n dn n npx
Dn þ μn E  ¼ F 0 α eðαxÞ  ð5:18Þ
dx2 dx τn τn
where Dn, μn and tn are the average values of the diffusion constant, mobility and
lifetime of electrons, respectively; n and npx are total and equilibrium concentration
of electrons; α, the absorption coefficient of the semiconductor and E built-in electric
field. Also, the acceptor profile (Figs. 5.2 and 5.3) will be taken as an exponential
decaying from the surface, i.e.

N a ðxÞ ¼ N 0 eβx

where N0 is the impurity concentration at the surface, and β is a constant(cm1).

From Eq. 5.11 we have,

1 dN a
E¼ ð5:19Þ
N a dx
kT
E¼ β ð5:20Þ
q

and from Eqs (5.5) and (5.20).

1
β ¼ pffiffiffiffiffi
2 Dt ð5:21Þ
npx ¼ N p0 eβx
F 0 ¼ F 00 ð1  RÞ ð5:22Þ

where npx and np0 are the minority carrier concentration at any point in the bulk and
at the surface respectively.
Na is the concentration of the acceptor impurity and R is the reflection coefficient.
A low illumination level is assumed wherein number of minority carriers injected by
light is much smaller than the concentration of majority carriers. Thus, conductivity
modulation effects are neglected.
The appropriate boundary conditions at the surface x ¼ 0 and at the junction
x ¼ w are obtained as described below.
If the contact to the base is a perfect ohmic contact, i.e., no minority carriers are
injected or extracted from the base, the only minority carriers lost are those which
recombine at the surface of the p-layer i.e.,
5.3 Solution of Continuity Equation 65

  dn
σ n  npx jx¼0 ¼ Dn jx¼0 þ μn E n jx¼0 ð5:23Þ
dx
and

njx¼w ¼ npw ð5:24Þ

where σ is the surface recombination velocity.

The solution of Eq. (5.18) is given by:

αF 0 eαx
n ¼ Aem1 x þ Bem2 x 
þ npx ð5:25Þ
Dn α2 þ βα  L12
n

where.
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ffi
β β þ L2
2 4
n
m¼ ð5:26Þ
2
αF 0
A¼    ½ðσ þ Dn α þ Dn βÞem2 w  ðσ þ Dn m2 Þeαw 
1
Dn α2 þ βα  2
Ln ð5:27Þ
1

½ðσ þ Dn m2 Þe m2 w  ðσ þ Dn m1 Þem1 w 

and

αF 0
B¼    ½ðσ þ Dn α þ Dn βÞem2 w  ðσ þ Dn m2 Þeαw 
1
Dn α þ βα  2
2
Ln ð5:28Þ
1

½ðσ þ Dn m1 Þe m1 w  ðσ þ Dn m2 Þem2 w 

where Ln is the average diffusion length of electrons. The electronic component of

current is given by:

dn
J n ¼ qDn jx¼w þ qμn E n jx¼w ð5:29Þ
dx
Evaluating

dn
jx¼w and njx¼w
dx
66 5 Mathematical Model of Transport Processes

it can be shown that:

qF 0 α
jn ¼  
1
Dn α2 þ βα  2
Ln
 
m1 ½ðσ þ Dn α þ Dn βÞem2 w  ðσ þ Dn m1 Þeαw em1 w ð5:30Þ
ðσ þ Dn m2 Þem2 w  ðσ þ Dn m1 Þem1 w
 
m2 ½ðσ þ Dn α þ Dn βÞem1 w  ðσ þ Dn m2 Þeαw em2 w
 þ αeαw
ðσ þ Dn m2 Þem2 w  ðσ þ Dn m1 Þem1 w

Obviously jn can be written as:

jn ¼ jnphoto þ jsndark ð5:31Þ

where jnphoto represents the electron component of the photocurrent which is a

function of the incident photon flux. jsndark stands for electron component of the
dark reverse saturation current of the cell.
The above expressions are useful for the investigation of the overall performance
of the p-n junction solar cell as a solar energy converter. In the present book
calculations have been carried out for the case of silicon cells.
The total photocurrent jphoto due to creation of electron-hole pairs by the incident light
is given by

jphoto ¼ jnphoto þ jpphoto

where jpphoto is due to light generated minority carriers of the n-region (collector).
For short wavelengths, i.e. for larger values of α, the electron hole pair generation is
limited to the p-type base region. In such a case, the photocurrent is due to electrons
of the p-type base region only, and is given by.

jphoto ¼ jnphoto

Figure 5.6 shows the dependence of the collection efficiency, defined as jphoto/qF0
on the surface recombination parameter σw/Dn for built-in field parameter βw ¼ 0.1
and with w/Ln ¼ 0.2. It may be mentioned that βw ¼ 0 represents the case of a solar
cell having a uniform concentration of p-type impurities in the base region.
It may be noted that for higher built-in field the rate of decrease of collection
hhefficiency with respect to surface recombination σw/Dn is reduced. It means that
under such a condition the performance of the solar cell is less sensitive to the changes
of surface properties.
5.4 Piecewise Exponential Computational Scheme 67

1.0
βw = 5

0.8

0.6 =1
JPhoto
0.4 =0
qF0

0.2

0
0.1 1.0 10.0
sw
Dn

Fig. 5.6 Collection efficiency jphoto/qF0 vs dimensionless surface recombination σw/Dn with
dimensionless field βw as parameter

5.4 Piecewise Exponential Computational Scheme

This method utilizes the simplicity of exponential model and yet effectively takes
into account the field and mobility gradients. The method had been developed earlier
(Gangadhar and Bhattacharya 1968) for the study of admittance characteristics of the
p-n junction diode with a narrow diffused region; subsequently for the study of
spectral response of solar cells by Gangadhar and Kaushika (1992b). In this method,
the diffused p- region is divided into a number of sections, say l, and in each section,
the impurity profile N ¼ N(x) is approximated by a separate exponential such that

kT
Ei ¼ β ¼ a constant ð5:32Þ
q i
1 N i1
with βi ¼ ln , i ¼ 1, 2, 3 . . . l ð5:33Þ
wi Ni
where wi is the width of the ith section and Ni is the actual impurity concentration
at the interface of the ith and (i + l)th sections. This procedure, where the three
types of impurity profiles are subjected to a piecewise exponential approximation, is
shown in Figs. 5.2 and 5.3. For the three types of profiles N0 ¼ 1017 cm3 and
NB ¼ 1015 cm3, are assumed. The impurity concentration due to an assumed
exponential distribution significantly differs from the Gaussian or complementary
error function distribution.
It is found that
(i) for a large number of sections, i.e., for a large value of l, Ei as obtained from Eqs
(5.7), (5.9) and (5.11) is close to the actual average field value in each section
and
68 5 Mathematical Model of Transport Processes

Table 5.1 Average field values in the diffused region

Impurity Average built-in field, Ei, V cm1, in different sections

distribution (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Gaussian 120 350 590 820 1050 1290 1590 1760 1990 2210
CERF 582 730 870 1020 1160 1310 1450 1600 1745 1890
Exponential 1190 1190 1190 1190 1190 1190 1190 1190 1190 1190
N0 ¼ 1017 cm3, NB ¼ 1015 cm3, w ¼ 1 μm

(ii) the division of the p-region into 10 sections, i.e., 1 ¼ 10 serves the purpose with
Ei, values approaching the average field in the ith section.
The Ei values with w ¼ 1 μm, N0 ¼ 1017 cm3 and NB ¼ 1015 cm3 have been
calculated for Gaussian, complementary error function and exponential impurity
distributions and are shown in Table 5.1 (Gangadhar and Kaushika 1992b). Simi-
larly the minority carrier mobility values μn, averaged in each of the l-sections
denoted by μni for the three impurity profiles can be calculated (Gangadhar and
Kaushika, 1992b).
The method is applicable to all types of impurity profiles and using a few sections
a reasonably accurate characterization of the transport property of an homoge-
neously doped device is possible.
The above approach of ‘piecewise exponential approximation’ had been made
use of in the analysis of spectral response of solar cells in the book. This method has
the following features:
(i) The diffused region is divided into a number of sections (Figs. 5.2 and 5.3) and
the impurity distribution is approximated by an exponential in each of the
sections. As can be seen from Fig. 5.3, this amounts to taking the actual
impurity concentration unlike the single exponential approximation. The prox-
imity to the actual value is more whenever the number of sections is increased.
(ii) The mobility of minority carriers is the averaged values in each of the sections,
whereas a single averaged value is used in the single exponential approxima-
tion. Thus, the mobility value is nearer to the actual value especially for a large
number of sections. This in a way takes care of mobility variation.
(iii) The diffusion coefficient approaches the actual value and its variation is taken
into account as discussed above, for mobility.
In view of above, it is seen that the method of piecewise exponential approxima-
tion is a realistic approach as it takes into account the actual impurity concentration,
its variation as well as the value and the variation of built-in field; values of mobility,
its variation; and value of diffusion coefficient, and its variation; to a considerable
extent. Also, the method is independent of whether the resultant impurity concen-
tration in the diffused region, depending on the technique of fabrication, is described
by an exponential, Gaussian, complementary error function or any other function.
5.4 Piecewise Exponential Computational Scheme 69

Formulation
Let us consider a p-n junction solar cell having a thin diffused p-type base and long
homogeneously doped n-type collector. The impurity concentration and its distribu-
tion in the base region gives rise to spatially varying built-in field and mobility
values. The variations may be taken into account by a method known as integration
by piecewise exponential approximation developed earlier for characterization of a
narrow diffused diode (Bhattacharya and Gangadhar 1971). In this approach the
diffused region is divided into a number of sections, say l, and in each section, the
built-in field Ei and βi may be expressed as follows:
(i) Uniform distribution: βi ¼ 0
(i ¼ 1, 2,. . ., l)
(ii) Exponential distribution: βi ¼ constant
(i ¼ 1, 2, 3. . ., l)
(iii) Gaussian and CERF distribution: βi is variable with i.
A typical p-type Gaussian impurity profile in the base region and its piecewise
exponential approximation are illustrated in Fig. 5.2. Figure 5.3 shows a comple-
mentary error function profile in the diffused p-region subjected to piecewise
exponential approximation. The built-in field in each of the section is given by
Eq. (5.32). For a large number of sections the built-in field Ei is close to average field
in each section. The one dimensional continuity equation for electrons in the ith
section of the diffused p-region, with a photon flux F 00 of monochromatic light
incident on the front surface of the cell is given by
2
∂ ni ∂ni ni  npxi
Dni þ μni Ei  þ F 0 αeαx ¼ 0, i ¼ 1, 2, 3:: . . . , l ð5:34Þ
∂x 2 ∂x τn
where Dni and μni are the diffusion coefficient and mobility suitably averaged in the i
th
section, τn is the lifetime of electrons, ni and npxi are the total and equilibrium
concentration of electrons and α is the absorption coefficient of the incident radiation
in the semiconductor. By making a large number of sections, the diffusion coeffi-
cient and mobility values approach actual values and their spatial variation.

n2i
npix ¼ ð5:35Þ
ni j x

where ni is the intrinsic carrier concentration, and.

F 0 ¼ F 00 ð1  RÞ ð5:36Þ

where R is the reflection coefficient.

The solution of Eq. (5.34) is given by:

ni ¼ Ai emi x þ Bi eli x  Gi F 0 eαx þ npxi ð5:37Þ

where.
70 5 Mathematical Model of Transport Processes

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
β β2i 1
m i , li ¼ i  þ
2 4 Dni τn

and
 1
1
Gi ¼ Dni ¼ α þ βi  ð5:38Þ
αDni τn

The constants Ai (i ¼ 1, 2,. . ., l ) and Bi (i ¼ 1, 2,. . ., l ) are evaluated using the

boundary conditions corresponding to each section. The assumptions made to obtain
the boundary conditions are:
1. The minority carrier density, at the interface of any two sections, is continuous;
which gives:

Ai emi xi þ Bi eli xi  Aiþ1 emiþ1 xi  Biþ1 eliþ1 xi  F 0 eαxi ðGi  Giþ1 Þ ¼ 0

ð5:39Þ
i ¼ 1, 2, . . . ðl  1Þ

2. The minority carrier current density (due to diffusion and drift) is also continuous
at the interface of any two adjacent sections. As a result,

Ai li emi xi þ Bi mi elxi  Aiþ1 liþ1 emiþ1 xi  Biþ1 miþ1
eliþ1 xi 

þDni Gi F 0 eαxi ½α þ βi   Dnðiþ1Þ Giþ1 F 0 eαxiþ1 α þ βiþ1 ¼ 0 ð5:40Þ
i ¼ 1, 2, . . . ðl  1Þ

Equations (5.39) and (5.40) give (2l – 2) boundary conditions and two more are
obtained by considering the conditions at the surface (x ¼ 0) and at the junction
x¼w

∂n1
σðnÞ jx¼0 ¼ Dn1 jx¼0 þ μn1E 1 n1 jx¼0 ð5:41Þ
∂x
and

n1 ¼ npx jx¼w ð5:42Þ

where σ is the recombination velocity at the p-type surface and w is the depth of the
p-n junction. The constants Ai and Bi can be calculated using Eqs (5.39, 5.40, 5.41
and 5.42), through the Determinant method of solving simultaneous equations. The
electron component of the photo current density is given by:

∂n1
jnphoto ¼ qDn1 jx¼w þ qμn1 E1 n1 jx¼w ð5:43Þ
∂x
5.5 Illustrative Results 71

Finally, the quantum efficiency, ηQ and the collection efficiency, ηC of the

diffused base region, are given by

jnphoto
ηQ ¼ ð5:44Þ
qF 0
jnphoto
ηC ¼ ð5:45Þ
qF 0 ½1  eαw 

5.5 Illustrative Results

The efficiencies of solar cell having a Gaussian distribution in the base region have
been computed corresponding to solar radiation wavelength range of 0.4–0.8 μm.
Data used are NB ¼ 1015 cm3, N0 ¼ 1017 cm3, w ¼ 1 μm, τn ¼ 109 sec and
l ¼ 10.
Table 5.2 (Gangadhar and Kaushika 1992b) shows values of η0 and ηC obtained
for different wavelengths and their variation with recombination velocity at the
surface.
With a view to see the variation of the spectral response with base region acceptor
impurity profile and associated built-in field and mobility values, calculation similar
to the above have been made for the four distributions namely complementary error
function, Gaussian, exponential and epitaxial, and compared in Table 5.3
(Gangadhar and Kaushika 1992b).
It is seen that the efficiencies are different for different distributions. Hence, the
realistic distribution must only be considered and must not always be approximated
by a single exponential distribution. The present method enables the realistic distri-
bution to be taken into account and is adaptable on a microcomputer.

Table 5.2 Spectral response (with σ as variable)

o' (cm/s)
10 105 106 107
λ (μm) ηQ ηC ηQ ηC ηQ ηC ηQ ηC
0.40 0.916 0.916 0.779 0.779 0.429 0.429 0.263 0.263
0.45 0.862 0.906 0.791 0.831 0.608 0.640 0.522 0.549
0.50 0.690 0.945 0.645 0.884 0.533 0.730 0.479 0.657
0.55 0.524 0.954 0.495 0.900 0.420 0.764 0.384 0.694
0.60 0.376 0.963 0.357 0.914 0.307 0.787 0.283 0.727
0.65 0.248 0.954 0.236 0.907 0.205 0.790 0.191 0.733
0.70 0.189 0.945 0.180 0.901 0.157 0.786 0.146 0.732
0.75 0.150 0.961 0.143 0.915 0.125 0.801 0.117 0.747
0.80 0.091 0.959 0.087 0.916 0.077 0.806 0.071 0.751
72 5 Mathematical Model of Transport Processes

Table 5.3 Spectral response (with impurity profile as variable)

CERF Gaussian Exponential Epitaxial
λ (μm) ηQ ηC ηQ ηC ηQ ηC ηQ ηC
0.40 0.931 0.931 0.916 0.916 0.927 0.927 0.826 0.826
0.45 0.899 0.946 0.862 0.906 0.890 0.936 0.804 0.846
0.50 0.695 0.953 0.690 0.945 0.685 0.941 0.628 0.863
0.55 0.528 0.961 0.524 0.954 0.517 0.939 0.479 0.870
0.60 0.378 0.970 0.376 0.963 0.367 0.933 0.344 0.874
0.65 0.250 0.957 0.248 0.954 0.237 0.917 0.227 0.876
0.70 0.190 0.951 0.189 0.945 0.176 0.893 0.173 0.878
0.75 0.151 0.965 0.150 0.961 0.134 0.854 0.137 0.879
0.80 0.092 0.965 0.091 0.959 0.022 0.235 0.084 0.879
Chapter 6
Electrical Characteristics of Solar Cells

6.1 Solar Cell: A Power Source

To understand the behaviour of a solar cell as an electric power source, let us review
the familiar characteristics of p-n junction diode and its behaviour under forward and
reverse bias when dark. When n-type silicon semiconductor (characterized by
electrons as majority carriers and holes as minority carriers) is in metallurgical
contact with a piece of p-type semiconductor (characterized by holes as majority
carriers and electrons as minority carriers) the boundary formed is p-n junction
(Fig. 6.1). Owing to the process of natural diffusion electrons would flow from
regions of high concentration (n-type) to the regions of low concentration ( p-type)
and the similar diffusion for holes. The electrons leaving the n-type side will expose
positive charges in n-type side. The holes leaving the p-type side will expose the
negative charges. This will result in a charge imbalance on both sides; it will set up
an electric field that will oppose the natural diffusion tendency of electrons and holes
and an equilibrium situation will be obtained. The region of transition is known as
depletion layer, because the region is depleted of holes and electrons. The diode
behaviour of such a p-n junction in dark conditions for forward bias and reverse bias
conditions is well known and is shown in Fig. 6.2.
A forward bias voltage ( p connected to positive of the battery and n to the negative)
reduces the field in depletion region allowing a diffusion of majority carriers across the
junction to the region where they are in minority resulting the I-V curve of first quadrant
in Fig. 6.2. In reverse bias condition, electric field in depletion region is increased which
strongly opposes the diffusion of majority carriers to zero. However, the field direction
and magnitude is such that it attracts minority carriers from the neutral regions and moves
them by drift across the depletion region resulting in a negative or reverse current through
the structure. The concentration of minority carriers is order of magnitude lower than the
majority carriers for each of the two regions. Consequently the reverse current is orders of
magnitude lower than the forward current. Thus a solar cell under dark conditions acts as
a rectifier.

© Capital Publishing Company, New Delhi, India 2018 73

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_6
74 6 Electrical Characteristics of Solar Cells

0.5 μm N - region

300 μm P - region

dark current
photo current

Fixed Ions
Majority
Majority Charge carriers
Charge carriers

Minority Minority
Charge carriers Charge carriers

Depletion layer

Fig. 6.1 p-n junction diode behaviour of solar cell

Fig. 6.2 Junction diode

behaviour in dark condition
for forward bias and reverse
bias conditions Current
-200 V
Voltage

0.7 V

Under illumination semiconductor plasma is generated in the solar cell and it is

separated by the drift of minority carriers across the depletion regions resulting in
large reverse current. The light generated is superimposed onto the normal rectifying
current-voltage characteristics of the diode. This results in the I-V curve covering the
fourth quadrant where electrical power (IνV) is negative which means that the power
can be extracted from the device.
6.3 Current-Voltage Characteristics 75

Fig. 6.3 Equivalent circuit I

of a solar cell
ISH RS
IPh ID RP
V Load

6.2 Equivalent Circuit

A solar cell may be regarded as a current source shunted by a diode and a resistance
Rp, in series with Rs (lumped internal series resistance) to produce a terminal voltage
V and current, I through the load resistance. This equivalent circuit is shown in
Fig. 6.3.
It consists of a constant current generator shunted by the junction which acts like
a positively biased diode. Rs is the lumped internal series resistance and RL is the
load resistance. The series resistance arises mainly due to
(i) the thin diffused layer on a silicon solar cell and
(ii) ohmic contact on the two opposite surfaces and bulk silicon resistivity.
The shunt resistance Rp is the result of the leakage current across the junction and
the edge of the solar cell. Shunt resistance and capacitance can normally be
neglected. Furthermore, in view of the fact that solar cell is a power source in
contrast to an ordinary diode which absorbs power, in practice it is customary to
regard the direction of photo-generated current to be positive.

6.3 Current-Voltage Characteristics

Following the above solar cell equivalent circuit I-V characteristics may be
expressed as follows:

I ¼ IG  Ij
h i V  IR
¼ I G  I 0 e AKT Þ  1 
ðqðV IRs Þ
s
Rp

Normally we take Rp ! 1 and Rs ! 0, which gives

76 6 Electrical Characteristics of Solar Cells

Fig. 6.4 I-V characteristics

of crystalline silicon Isc
solar cell
Imax Pmax

Current

Vmax Voc
Voltage

h qV i
I ¼ I G  I 0 eðAKT Þ  1 ð6:1Þ

where IG ¼ photo-generated current, I0 ¼ dark saturation current of the diode; it is

proportional to e(Eg/KT), q ¼ electronic charge (1.6022  1019 C), K ¼ Boltzmann
constant (1.38  1023Joule/ K), T ¼ absolute temperature of cell and V ¼ terminal
voltage.
In accordance with Eq. 6.1 current-voltage curve of a crystalline silicon solar cell will
be represented as in Fig. 6.4. The parameters indicated therein correspond to Standard
Test Conditions (STC); cell temperature of 25  C and solar irradiance of 1000 W/m2.
The short circuit condition corresponds to zero voltage across the terminals of the
cell; Ij is very small and practically all the generated current passes through the
external circuit. The short circuit current is therefore a useful measure of photo-
generated current and the two terms mean the same. The open circuit voltage (Voc)
corresponds to I ¼ 0, V ¼ Voc which from Eq. (6.1) is given as follows:
 
AKT IG
V oc ¼ loga þ1 ð6:2Þ
q I0

Furthermore, solar cell current-voltage (I-V) relationship is such that the power
output is zero at short circuit and open circuit and the output is maximum (Pmax) at
intermediate point (Vmax, Imax). For maximum power output of cell, its load operation
corresponds to Vmax, Imax conditions. Maximum power, Pmax corresponds to the area
of largest rectangle that can be fitted in I-V curve. The conversion efficiency is the
maximum power output expressed as a percentage of the input power. The cell
conversion efficiency (η) is given by

Pmax
η¼  100 ð6:3Þ
Irradiance  Area
Obviously, nearer the I-V curve to rectangular shape, higher is the maximum
power and conversion efficiency. The rectilinearity of I-V curve is represented by a
parameter referred to as fill-factor which is defined as
6.4 Performance Measurement Procedures 77

Pmax
ff ¼ ð6:4Þ
I sc  V oc

6.4 Performance Measurement Procedures

The performance of a crystalline solar cell is a sensitive function of

(i) Solar irradiance,
(ii) Cell temperature and
(iii) Spectral irradiance distribution.
Consequently, the experimental arrangements for determining the performance of
a solar PV device—be it a cell, a module, a sub array or an array—must involve the
device exposure to known temperature and stable sunlight (natural or simulated).
The measurement of I-V characteristics must involve the simultaneous recording
of the solar irradiance and cell temperature. For uniform interpretation it is often
desired that the measured characteristics be transposed to Standard Test Conditions
(STC) which are as follows:
1. Irradiance: 1000 W/m2
2. Cell temperature: 25  C
3. Reference spectral irradiance distribution due to Bird et al. (1983).
The power output at STC is often referred to as rated or peak power.
In natural sunlight the spectral content varies with location, weather, time and day
of the year. Similarly in solar simulator it depends on light source and the optical
system that is affected by aging. In practice it has been found that spectral changes
can affect outdoor measurements on PV device as much as 15%; if the irradiance is
measured with an instrument that is not spectrally selective such as pyranometer or
other thermopile type radiometer. In such a situation the measurement on PV device
cannot be related to a reference spectral irradiance distribution for rating purposes.
This problem may be bypassed by measuring the irradiance with a reference device
that has essentially the same spectral response as the test (PV) specimen and has been
calibrated in terms of short circuit current per unit of irradiance of reference spectral
distribution. The solar device measures the irradiance in terms of reference distribu-
tion and the specimen measurements give the response corresponding to same
reference distribution. In this way, location, time and weather conditions in outdoor
conditions and aging of simulator in indoors is automatically taken care. In exper-
imental measurements we have studied I-V characteristics with known cell temper-
ature (or else have affected temperature correction) and in stable solar irradiance
which has been measured by a solarimeter (Surya Mapi) of same spectral response as
PV (test) specimen.
78 6 Electrical Characteristics of Solar Cells

+ A –
+
0-2A
V 50 watts
50 ohms
– 0 - 20 V I

V
+ –
A

0 - 20 V
– 12 V
0-2A
V
I
+
50 watts
50 ohms

– +
A
+
0-2A
V
0 - 30 V
– 0 - 20 V 3A I

V
50 watts
50 ohms

Fig. 6.5 Schematic diagram of measurement procedure of I-V characteristics

The general schematic of measurement procedure is illustrated in Fig. 6.5. The

I-V curve can be traced by varying load through range. The variable load should
preferably be an electronic circuit which sweeps the load from short circuit to open
circuit condition. The accuracy of measurements is in conformity with international
standard procedures as indicated below:
Parameter Accuracy
Temperature 1  C
Current and voltage 0.5%
Furthermore, the open circuit voltage Voc can be obtained by isolating the test
specimen by means of a switch. The short circuit current in ideal situation is measured
at zero voltage and electronic load (negative) is used to offset the voltage drop across the
external series resistance. Yet another approach could be to measure the short circuit
current at a voltage up to 3% of open circuit voltage and extrapolate it to zero voltage.
Our experiments have also involved the measurements of I-V characteristics in
natural sunlight. To ensure stable irradiance, measurements were carried out on clear
days and during 30 minutes of local noon hours.
6.4 Performance Measurement Procedures 79

Y- channel Range for Y-channel

Variation in A: 0 to Isc
A 0 − 0.5V 0 − 5V 0 − 20V
0 − 2.5A 0 − 2.5A 0 − 2.5A
V X- channel Potentiometer

Protentiometer could be electric or electronic 50 Ω

A or else 0- Isc

An Electronic I-V Plotter

Fig. 6.6 Electronic measurement of I-V characteristics

For small power units (e.g. a solar cell or a small string/module of solar cells) manually
adjusted rheostat may also be used. Typical values of variable load and precision resistors
for such cases were 50 Ω (50 W) and (0.5–1) Ω respectively. In case of simple manual
observational approach, the two circuits as shown in Fig. 6.6 may be used.
The power produced in each of the above cases is very small and there must be an
intermediate value of RL for which power is maximum, Pmax. The corresponding
output voltage and current are referred to as optimum voltage, Vm and optimum
current, Im, respectively.
Pmax ¼ I m V m ð6:5Þ

The electrical quality of a solar cell is measured in terms of fill factor, ff.

V m Im
ff ¼ ð6:6Þ
V oc I sc
Thus Voc, Isc, Vm, Im, Pmax and ff are the basic electrical parameters of solar cells
which are of interest in their practical utilization.
To determine the I–V characteristics of a solar cell, the voltage across the cell is
measured using a voltmeter and the current flowing through it is measured using an
ammeter. The load resistance is varied and the corresponding voltage and current are
measured at a particular solar insolation, measured using a solarimeter. A typical
silicon solar cell was fabricated at the Indian Institute of Technology, Delhi during
1967. The cell active area was 2.4 cm2 and the illumination level was 80 mW/cm2.
The conversion efficiency was around 10%.
80 6 Electrical Characteristics of Solar Cells

6.5 Temperature and Solar Irradiance Effects

The I-V characteristics for a silicon solar cell fabricated at the Indian Institute of
Technology, New Delhi (Jain et al. 1968) was measured at different temperatures as
early as in 1968. The solar cell had been placed in a thermostat whose temperature is
varied over a wide range of temperatures. The cell is kept under constant illumina-
tion by a lamp. The cell temperature had been measured with a thermocouple and a
temperature indicator. The objective of the study was to measure the I-V character-
istics of the solar cell, the short circuit current, open circuit voltage and power output
of the cell at different temperatures. The study enables assessment of their temper-
ature dependence and concludes that
(i) The short circuit current increases slightly with increase of temperature.
(ii) The open circuit voltage decreases rapidly at the rate of 2.3 mV  C1.
(iii) Power output of the cell decreases at the rate of about 0.6%  C1.
Subsequently various authors have investigated the effects of changes in irradi-
ance and temperatures. The typical temperature coefficients (Treble 1991) are:
dI SC 
¼ 0:1 Am2 C1
dT
dV OC 
¼ 2:2 mV C1
dT
dPMAX 
¼ 0:5% C1
dT
With the increase in the intensity of illumination, the short circuit current
increases linearly up to a reasonably high level of illumination. The open circuit
voltage increases slightly with increase in the intensity of illumination. Also, the
open circuit voltage is less sensitive to increase of illumination as well as variations
in illumination level.
At higher levels of illumination, it is desirable to provide cooling to the cells. This
is owing to the fact that at higher levels of illumination, the power output of the cell
decreases with the increase of temperature.
Chapter 7
Solar PV Module and Array Network

7.1 Solar PV Module

The smallest operational unit of all SPV systems is the solar cell. Its size is chosen
according to planned application. To obtain the desired voltage and current ratings
the individual solar cells are connected in series and/or parallel. A string of cells
connected in a pure series arrangement is referred to as series string or sub string. A
group of series string or sub strings connected in parallel forms a series blocks. A
branch circuit is composed of a series blocks between the positive and negative
terminals of a power unit. An array is a collection of branch circuits (Fig. 7.1). The
array is constructed in the field and transported from factory to field in small units
called modules. The modules can be used to construct mW to MW size SPV arrays.
Module is a panel for structural purposes having solar cells connected in series or
parallel encapsulated in weather resistant laminate to protect the cells and enclosed
in the sturdy corrosion resistant frame.
It may be noted that present-day modules are made from non-circular solar cells.
Based upon the state and quality of silicon material three types of solar cells/modules
have been produced and marketed. Mono-crystalline silicon solar cells have almost
no defects and impurities. They have solar conversion efficiency of 15–17%. They
are expensive to produce. Polycrystalline silicon cells/modules are slightly poorer
grade monocrystalline silicon or semiconductor grade silicon. They exhibit white
speckles on surface due to impurities. They are less expensive since simpler pro-
cesses are involved in their production. They have solar conversion efficiency of
11–13%. Amorphous silicon cells/modules are made from amorphous silicon rather
than silicon of crystal structure. They absorb light more effectively and are relatively
thinner. They have lower conversion efficiency of 5–7%. However the reduced cost
often overweighs reduced efficiency leading to a net increase in ratio of performance
to cost.
Crystalline silicon modules are presently the dominant commercial product and
deliver approximately 100–130 W/m2 at STC. Amorphous silicon (a-Si) thin-film

© Capital Publishing Company, New Delhi, India 2018 81

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_7
82 7 Solar PV Module and Array Network

Fig. 7.1 Arrangement of solar cells in string, series block, branch circuit and array

modules, which are beginning to enter the market, require less material to produce
than the thick crystalline products and so can be made less expensively. Today’s
commercial a-Si modules deliver 40–50 W/m2 under full sun at STC. The modules
are connected electrically in series and parallel to build arrays.
7.2 Mismatch Losses in Array Network 83

7.2 Mismatch Losses in Array Network

Solar cell fabrication procedure involve several controls; despite all automations the
resultant cells possess significant manufacturer’s tolerances. Performance-wise they
exhibit variations in current-voltage characteristics. In particular, current (Imax) and
voltage (Vmax) of individual cells operating at maximum power (Pmax) exhibit
statistical distribution. So when mass-produced solar cells are connected in series
and parallel within a network, the statistical distribution of cell parameter leads to
electrical mismatch in the network; the mismatch limits the short circuit current of
the combination to a value closer to the lower end of the distribution than to average.
Consequently the maximum power output of the network is often less than the sum
of the maximum power of individual cell; this is referred to as mismatch loss.
Similar results would occur with elements combined in parallel; open circuit
voltage distribution will generate inter string power dissipation. This mismatch in
fill-factor is not significant since the mismatch averages out producing no
power loss.
In the above context a question of practical importance is: how much variations in
cell characteristics can be allowed to meet the intended module and array perfor-
mance specifications.
Several authors (e.g. Bucciarelli 1979; Gonzales and Weaver 1980; Chamberlin
et al. 1995; Kaushika and Gautam 2001) have investigated the lowering of available
power in solar PV arrays and found it to be due to small variations in cell charac-
teristics. The mismatch losses in solar PV arrays can be due to a variety of reasons as
follows (Sayed and Partin 1975; Treble 1991; Kaushika 1996).
• Manufacturer’s tolerances in cell characteristics
• Environmental stresses
• Shadow problem
Physical differences between cells or cell processing material may arise during
normal processing and different cells may have slightly different characteristic
parameters. Consequently in an array the output power of the combination is less
than the sum of maximum output power of constituents. This power loss may tend to
enhance with time due to degradations resulting from aging.
The environmental stresses in solar PV array include the partial or full opening of
a string due to cell cracking (provoked by hail impact) or otherwise. This not only
gives rise to mismatch loss but also could result in excessive heating of the regions of
power dissipation, which in turn can lead to solder melting and damage the
encapsulant. The regions of excessive heating are referred to as hot spots (Sayed
and Partin 1975; Feldman et al. 1981).
In the field, solar cells arrays are subject to shadows from both predictable sources
as well as from such unpredictable sources as bird droppings or fallen leaves. The
percentage power loss is much greater than the percentage of array area, which is
shadowed; for smaller arrays with few or no parallel connections, one leaf could
cause the system output to drop to a fraction of rated power, eventually resulting in
84 7 Solar PV Module and Array Network

system failure. Shadowed cells in series with illuminated cells block the current flow
in entire series connection and tend to become reverse biased. Consequently, hot
spots are developed in the localized regions of shadowed cell, which can damage the
encapsulation or crack the cell causing module failure.
In this chapter, we investigate the effect of aging of solar cells in a module or
array and hence reduction in available power. This investigation is based on real life
data of ensemble of fresh and aging solar cells and is intended to enable the sizing of
solar PV system in such a way that it can deliver power to the given load without
failure for longer duration (Arab et al. 2004; Hussein et al. 2004; Kaushika et al.
2005).

7.3 Formulations of Fractional Power Loss in Solar PV

Module and Array

7.3.1 Solar Cell Model

The solar PV module is formed by an appropriate series-parallel combination of

solar cells that provides the required rated output voltage and current under normal
conditions. Following Chapter 6 the electrical equivalent circuit of the solar cell is
shown in Fig. 7.2.
Applying Kirchhoff current law, the solar cell terminal current, I can be expressed
as

I ¼ I ph  I D  I SH ð7:1Þ

where Iph is photo-generated current (Amp), ID ¼ diode current (Amp), and ISH is
shunt current (Amp).
The photo-generated current (Iph) depends on both irradiance and temperature. It
is measured at some reference conditions such as reference temperature, Tc,ref,
reference radiation Gref and reference photocurrent Iph,ref and is related as follows
(Chenni et al. 2007; Rai et al. 2011):

ISH Rs
IPh ID RP
V Load

Fig. 7.2 Equivalent circuit of solar cell model

7.3 Formulations of Fractional Power Loss in Solar PV Module and Array 85

G


I ph ¼ I ph, ref þ I SCC T c  T c, ref
Gref

where G is the actual solar radiation (W/m2), Tc, the actual cell operating temperature
of cell (K) and Iscc is the manufactured supplied temperature coefficient of the short
circuit current (A/K).
The diode current is given by Shockley equation
   
eð V c Þ
I D ¼ I 0 exp 1 ð7:2Þ
η KT c

where Vc is voltage across diode (V); I0, reverse saturation current (A); η, diode
ideality factor; Rs, series resistance (Ω); e, electron charge 1.602  1019 C and K is
Boltzmann constant, 1.38  1023 (J/K) whereas the shunt current, ISH is given by

Vc
I SH ¼ ð7:3Þ
Rp

Rp is the shunt resistance (Ω).

The Eq. (7.1) can be written as
  
eðV þ I Rs Þ V þ I Rs
I ¼ I ph  I 0 exp 1  ð7:4Þ
η KT c Rp
0 0
Let V ¼ V + I Rs and I 0 ¼ I þ RVp . Equation (7.4) becomes as
   
eV 0 V 
I 0 ¼ I ph þ I 0  I 0 exp max
ð7:5Þ
η KT c V 
max

Divide the above equation by I 

max
   
I0 I ph þ I 0 I0 eV 0 V 
¼    exp max
I
max I max I max η K Tc V
max
I ph þ I 0 I0 eV 
Let α ¼  , β¼  and C ¼ max
I max I max AKT c

The above equation can be written as

 
I0 V0
¼ α  βexp C  ð7:6Þ
I
max V max

where V  
max is average voltage at maximum power, I max ¼ average current at
maximum power and C is parameter related to fill-factor of solar cell.
86 7 Solar PV Module and Array Network

This expression is similar to that of Bucciarelli’s model (1979) with parameter α,

β and C. It is assumed to provide an adequate expression of I-V curve near maximum
power point. The parameter C can be matched with the curvature of I-V curve in the
neighbourhood of maximum power point. It can also be expressed in terms of cell
fill-factor. Thus in Bucciarelli’s model series and parallel resistances end up encap-
sulated in the parameter C. Furthermore, by definition we know that the fill-factor
(Treble 1991) is given by

Pmax V max Imax

Fill factor ¼ ¼ ð7:7Þ
I sc V oc I sc V oc
where Isc ¼ short circuit current of solar cell (Amps) and Voc ¼ open circuit voltage
of the cell (Volts).
The maximum power point current Imax and maximum power point voltage Vmax
will satisfy Eq. (7.6) and can be written as follows:
 
Imax V max
¼ α  βexp C ð7:8Þ
I
max Vmax
d ðIV Þ
For maximum power, ¼0
dV
dI
or, I þ V ¼0
dV
dI max
and at maximum power, I max þ V max ¼0 ð7:9Þ
dV max
From Eqs. (7.8) and (7.9) we have
h
i
I max
I exp C VV max

β¼ max max
ð7:10Þ
C VV max

max

Substitution of the value of β in Eq. (7.9) gives

h
i
 þ 1
Imax
I exp C VV max
α¼ max max
ð7:11Þ
C VV max

max

From Eqs (7.10) and (7.11) we see that in general I-V curve near the maximum
power point may be represented by Eq. (7.6) wherein the α and β are
non-dimensional quantities which may be defined in terms of Imax and Vmax.
The fill-factor ( ff) in terms of C can be written as (Bucciarelli 1979)

Pmax V max Imax C2

ff ¼ ¼ ¼ ð7:12Þ
I sc V oc I sc V oc ð1 þ C Þ½C þ ln ð1 þ CÞ
7.3 Formulations of Fractional Power Loss in Solar PV Module and Array 87

Table 7.1 Fill-factor and cell S. No. Fill-factor C

characteristics
1 0.3 1
2 0.4 1.72
3 0.5 2.85
4 0.58 4.22
5 0.67 6.83
6 0.7 8.13
7 0.75 11.8

Fig. 7.3 Fill-factor vs. cell characteristics

The Eq. (7.12) is an implicit relation; it may be used to evaluate C as a function of

ff. Some typical values are given in Table 7.1.
Figure 7.3 portrays the variation of cell characteristic parameter C with the fill
factor; it shows that C lies in the range 8–11 for the ensemble of fresh cells.

7.3.2 Fractional Power Loss in a Module

The fractional power loss due to mismatch for L solar cells connected in series
operating at common current is given by (Bucciarelli 1979)
 
Cþ2 2 1
E½ΔP ¼ ση 1  ð7:13Þ
2 L

The fractional power loss due to mismatch for M solar cells connected in parallel
operating at common voltage is given by (Bucciarelli 1979)
88 7 Solar PV Module and Array Network

 
Cþ2 2 1
E ½ΔP ¼ σξ 1  ð7:14Þ
2 M

Now consider a network of Nb series blocks of M parallel sub strings and each sub
string having L cells in series as shown in Fig. 7.4. The fractional power loss due to
mismatch is given by (Bucciarelli 1979)
   
Cþ2 1
2 
2 Nb
E ½ΔP ¼ ση 1 
2
 ση  σξ ðM  1Þ ð7:15Þ
2 T T

where T ¼ LMNb i.e. total number of cells in the network.

From these formulations it may be noted that mismatch losses in both the series
and parallel arrays are a sensitive function of the fill factor, variance in distribution of

Fig. 7.4 Network of Nb

series blocks of M parallel
sub strings and each sub
string having L cells in
series (Kaushika and Rai
2007)
7.4 Computational Results and Discussion 89

maximum power parameters of solar cell ensemble and total number of cells
connected in the network. Gautam and Kaushika (2002a) have developed an effi-
cient algorithm to simulate the performance of solar PV arrays.

7.4 Computational Results and Discussion

The formulations of mismatch losses are valid if the deviations, in cell max-power
current and voltages, from the mean values for the ensemble of cells used in the
network are small (limited to a few percent of their mean values). In the present
investigation we have considered the mono-crystalline solar cells manufactured by
Central Electronics Limited (CEL), New Delhi that has been manufacturing silicon
solar cell for about three decades. Upon our request they supplied us the data for
open circuit voltage (Voc), short circuit current (Isc), fill factor and maximum power
parameters of constituent cells of two modules each made from an ensemble of
36 cells manufactured by them. CEL, New Delhi had also supplied to us yet another
ensemble of cells that at the time of this investigation are representative of aging
cells with impaired performance resulting from erosion at the top layer of the cell.
These cells have been synonymously referred to aging or soiled cells in the present
work. The variance and coefficient of variation (CV) of maximum power parameters
of the ensembles are given in Table 7.2.
The standard deviation is found to be 2.5% of mean value for Imax for the fresh
cells. The values of CV are derived from the following relationships:
σI
ση ¼ ð7:16Þ
I
max

Similarly σξ of the distribution of maximum power voltage is given by

σV
σξ ¼ ð7:17Þ
V
max

The fill factor, in general, varies from 0.5 to 0.75 for the mono-crystalline silicon
solar cells. For example, for the cells manufactured by Central Electronics Limited
(CEL) in India, it ranges from 0.70 to 0.75. The fill-factor was found to deteriorate to
a value as low as 0.30 as a result of aging of cells.
It may be noted from Table 7.2 that the coefficient of variation (CV) is larger for
Imax (fresh as well as aging cells) than the CV for Vmax by a factor of 2.12 and 3.22
respectively for fresh and aging cells. Computational results of fractional power loss
as a function of the number of cells and the variance of the distribution of maximum
power parameters are illustrated in Figs. 7.5a and 7.5b). Both series and parallel

Table 7.2 Variance of cell max-power current and max-power voltage of fresh and aged cells
Category ση 2 σ ξ2 ση σξ
Fresh cell 2.7  105 6  106 5.2  103 2.45  103
Aged cell 3.36  103 3.24  104 0.058 0.018
90 7 Solar PV Module and Array Network

Fig. 7.5a Fractional power loss vs. number of cells connected in series for different variance of cell
max-power current (Kaushika and Rai 2007)

Fig. 7.5b Fractional power loss vs. number of cells connected in parallel for different variance of
cell max-power voltage (Kaushika and Rai 2007)
7.4 Computational Results and Discussion 91

Fig. 7.6 Fractional power loss vs. number of series blocks for different parallel sub strings
(Kaushika and Rai 2007)

interconnection schemes have been considered. Variance in Imax, ση2 is relevant in

series connection and the variance in Vmax, σξ2 is relevant in parallel combination;
hence considered accordingly. The CV for Imax has greater impact on power loss in
series string than CV for Vmax has for parallel string. However, comparing Figs. 7.5a
and 7.5b for the same values of CV, the fractional power loss is observed to be the
same. For example in Fig. 7.5b for a CV of 0.0048 the fraction power loss is about
0.025–0.0275. From Fig. 7.5a, for CV of 0.0048 the corresponding fractional power
losses lie in range of 0.02–0.04. The effect of series paralleling on the fractional
power loss of network is investigated in Fig. 7.6. We have considered an ensemble of
144 (T ¼ 144) fresh cells for the computation of fractional power loss for Nb series
blocks of M parallel strings each having L cells given by L ¼ T/MNb. Results show
that the fractional power loss tends to be lower for larger number of series blocks and
higher values of M (number of parallel strings). For example, for M ¼ 1, there is no
variation with increase in Nb. Furthermore, for Nb ¼ 1, the fractional power loss is
due to paralleling and hence due to CV of Vmax whose range of variation is very
small. To illustrate the point let us consider a module made up of a series string of
36 cells. It will have T ¼ 36, Nb ¼ 1, M ¼ 1 and L ¼ 36; the fractional power loss
(ΔP) is 0.00016. Now if it is series-paralleled with Nb ¼ 4, M ¼ 3, L ¼ 3, the
fractional power losses will reduce to 0.00011 (ΔP). In series string the power loss
due to mismatch increases significantly with aging of solar cells. However, it can be
reduced by an appropriate series paralleling arrangement. For example, for the
ensemble of solar cells used in present work the fractional power loss of 12%
resulting from the aging (soiling) can be counterbalanced by series paralleling in
an array oversized by 2.4%. The variation of fractional power loss as a function of
cell’s lumped characteristic parameter, C is investigated in Fig. 7.7. It shows that
aging cells are more sensitive to C and hence to fill-factor.
92 7 Solar PV Module and Array Network

Fig. 7.7 Fractional power loss vs. cell characteristics (Kaushika and Rai 2007)
Chapter 8
BOS and Electronic Regulations

8.1 Balance of System (BOS)

The Balance of System (BOS) encompasses all components of solar PV system other
than array. The array is the energy generating subsystem. It consists of PV modules,
panel structures, interconnecting wires, junction boxes along with lightning arresters
and associated accessories.
It generates variable DC only during sunshine hours which must be stored if the
power is to be used during night or on a cloudy day. Solar generated electricity can
be stored in battery bank. The battery banks can usually store sufficient electricity to
power the load through two to five completely cloudy days and nights. Charge
controller is yet another important component in connection with electrical storage
subsystem like battery. These devices protect batteries from excessive charge when
the module produces more electricity than the batteries can store. Also, they prevent
deep discharge of batteries in no sun conditions. In absence of charge controllers,
battery may suffer extreme wear and tear and become less effective, last for shorter
time and possibly even short circuit. Sometimes during over charging the gassing
from the acid of batteries may take place which in turn may become a fire hazard.
The output energy delivered from solar PV array to battery (load) may be
maximized by using a maximum power point tracker (MPPT). The maximum
power point tracker is used to correct the variations in the current voltage charac-
teristics of array resulting from the variability of incident solar radiation and ambient
temperature. It is basically a dc-to-dc converter that converts the input voltage of the
load to a value coinciding with load line, without any change in the delivered power,
but with a corresponding change in current.
The battery bank serves the load either directly or after conversion into AC. The
battery power produced tends to be at a low voltage (12 V) and household applica-
tion requires 240 V. An inverter is needed to convert direct current (dc) into
alternating current (ac). Inverters are electronic devices employed in both stand-
alone and utility-interactive PV systems to convert the direct current (dc) electricity

© Capital Publishing Company, New Delhi, India 2018 93

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_8
94 8 BOS and Electronic Regulations

produced by the PV array to alternating current (ac) common to the utility grid and
required by most appliances and electronic devices. In grid connected systems, the
dc power of the solar PV array is to be converted into AC power of fixed AC voltage.
The AC should be sinusoidal of 50 Hz with no harmonics. Inverters are specified in
terms of their capacity (in watts or kilowatts), their output voltage (e.g., 120 Vac) and
their power quality. All utility-interactive PV systems require an inverter, as do
stand-alone systems supplying ac loads. Stand-alone systems supplying only dc
loads do not need inverters.
A meter is required to ensure that the system owner is credited for any excess
power that is generated by the PV system and fed into the mains grid.
SPV power systems have been operated in various modes which can be classified
into following three categories:
(a) Stand Alone: The entire power is generated by SPV array and stored in a battery
to provide it in response to demand.
(b) Hybrid: In addition to SPV array such other means as wind generator and diesel
generators are also used to produce power.
(c) Grid Connected Systems: In such systems the output of SPV plant is connected
to the grid line and there is no storage battery.
The schematic diagram of a standalone PV power supply is shown in Fig. 8.1.
In this system the solar array needs the use of an optimum load into which it can
deliver its maximum power. But the requirement varies with insolation and
temperature. The battery in turn offers a load to the array that varies with its
state of charge and temperature. The mismatches of different system components
offer problem in optimum sizing of the system and results in loss of energy and
degradation of components. A matched interconnection of these power system
components under variable conditions may be realized with the help of Maxi-
mum Power Point Tracker (MPPT). So in practice solar photovoltaic system
requires several components in addition to PV modules. These are called Balance
of System (BOS).

Fig. 8.1 Schematic of standalone photovoltaic power supply system

8.2 Classification of BOS 95

8.2 Classification of BOS

In view of the above, BOS includes dc-dc converter, inverter, charge controller,
maximum power point tracker, solar irradiance sensor, electronics for energy man-
agement, electrical switches, circuit breakers, fuses, wires & cables, protection
devices etc. depending on the technical specifications and requirements of the PV
system. All the components of the balance of systems may be classified in three
categories: Mechanical, Electrical and Electronics BOS.

8.2.1 Mechanical BOS

Mechanical BOS includes PV panel structures, battery racks, poles and stays for
carrying electrical wires and cables etc. to withstand high wind speed and hail storms
for stability.

8.2.2 Electrical BOS

Since a PV system is an electrical energy conversion unit, the electrical BOS sub-
systems are similar to those used in conventional electrical power plants. This
includes lightning arresters in the field to control panels consisting of required
cable and wiring, circuit breakers, switchgears and so on.
Batteries are needed in stand-alone systems to store electricity for use at night and
during periods of low sun. Liquid-electrolyte, lead-acid storage batteries are, by far,
the most widely used. Deep-cycle batteries are required for long-term service.
Batteries are connected in series to provide the necessary system voltage, and in
parallel to provide the required amount of storage rated in amp-hours at the system’s
voltage.
Captive-electrolyte, or “gel-cell”, batteries are available for applications where
liquid electrolyte is undesirable, but their initial cost is typically greater than for
those with liquid electrolyte. Many new battery technologies are being investigated
for use in electric vehicles. When commercialized, these new batteries will benefit
PV systems as well.
String combiners are prefabricated junction boxes that are useful to facilitate the
parallel connection of the series strings of PV modules, called source circuits, which
make up the PV array. String combiners often have integral lightning protection
devices, such as silicon-oxide varistors, which shunt damaging high-voltage tran-
sients to ground. Article 690 of the National Electrical Code (NEC), which governs
96 8 BOS and Electronic Regulations

PV systems, requires over current protection and a disconnect means for PV array
source circuits, and most string combiners include over current protection in the
form of pull-out fuses or dc-rated circuit breakers, which also can serve as string
disconnects for array testing and service.
Disconnect switches are needed for system servicing and personnel safety, as set
forth in NEC Article 690. They are typically installed on the inverter’s input and
output, at the array output and on the battery bank’s output. Most disconnect
switches also include over current protection, either as fuses or circuit breakers. A
fuse or other means of “catastrophic” over current protection is also necessary at the
battery bank and, large battery banks even have fuses internal to their string wiring to
isolate portions of the bank in the event of a fault. Be certain to employ dc-rated
components in all dc circuits.
There are some main components that are key to any PV system, assuming the PV
array is mounted and the necessary cables and switches are in place. This equipment
is often referred to as balance of system equipment required to complete the system.
This includes some of the equipment explained below, and wires, conduit, a ground-
ing circuit, fuses, safety disconnects, outlets, metal structures for supporting the
module and any additional materials that are part of the PV system.
Batteries are sometimes necessary to store electricity for use during times when it
is needed. Or to help meet loads when PV modules are not generating sufficient
power to meet load requirements. Electricity storage is a critical component of many
PV systems. If the power is needed at night or on a cloudy day, solar generated
electricity can be stored in batteries. These battery banks usually store sufficient
electricity to power the home through two to five completely cloudy days and nights.

8.2.3 Electronic BOS

The electronic BOS is required to monitor, convert, manage and control the PV
generated energy to the loads. Commonly used electronic BOS in solar PV system
may be divided into following categories.

8.2.3.1 Energy Conversion Subsystem

Energy conversion subsystem may have dc/dc converters and MPPTs; or dc/ac
inverters.

Dc/Dc Converters and MPPTs

Since the output of a solar PV array is variable dc, a dc-to-dc converter is required to
convert the dc output from the array to a voltage suited to the requirement of the
battery or load. A dc-to-dc converter may be located between the solar PV array and
8.2 Classification of BOS 97

battery bank to ensure optimum power. Among the dc-to-dc converters, the boost
converter is preferred because it can draw current continuously at the input side. The
voltage for optimum power transfer varies with insolation and array operating
temperature. A feedback control is required in order to ensure that the power
converter is operating at optimum voltage transformation ratio.
A maximum power point tracker (MPPT) is required to minimize the mismatch
between the solar PV array output and load. In stand-alone solar PV system without
battery storage, the module or array generates variable dc output during the day and
delivers it to the dc load. In such a simple system, the amount of energy delivered to
the load depends not only on the incident solar radiation and ambient temperature
but also on the operating load on the array current-voltage characteristics. If the load
is so chosen that the array operating point is at its peak power point during the
mid-day the array then delivers maximum possible amount of energy at that time.
But since both the incident solar radiation and the ambient temperature vary during
the course of a day, the position of the array peak power point varies with time.
Therefore, it is quite unlikely that the operating point will be coinciding with the
array peak power point at all times. There is some loss in the delivered energy due to
mismatch between the locus of load line and locus of peak power points. The power
loss due to load line mismatch can be overcome by using a maximum power point
tracker in the system.
A maximum power point tracker is an essentially dc-dc converter, which contin-
uously monitors the output current and voltage from the module or array as long as
there is enough sunshine and module is generating electricity. At the same time it
computes the peak power available, and then it converts the input dc voltage to the
load to a value coinciding with the load line without any change in the delivered
power, but with a corresponding change in the current. Thus, the peak power at any
instant is fed to the load at a voltage appropriate to coincide with its load line. The
possible power loss due to mismatch is partially eliminated.
It has built-in control logic, usually operated by a microprocessor that senses the
array voltage and current at frequent interval (typically every 30 ms), computes the
power output and compares it with the previous value. If the power output has
increased, the array voltage is stepped in the same direction as the last step.
If the power has decreased, the array voltage is stepped in the opposite direction.
Eventually, the array voltage reaches near the maximum power point and remains
constant irrespective of irradiance, temperature and load. The power consumption of
MPPT is typically 3–7% of the array output. Many techniques for maximum power
point tracking of SPV arrays are now known which include:
(i) Hill climbing/Perturb and Observe method
(ii) Incremental Conductance method
(iii) Neural network
(iv) Fuzzy Logic Control
(v) Fractional Open Circuit Voltage method
(vi) Fractional Short Circuit Current method
(vii) Ripple Correlation Control (RCC) method
98 8 BOS and Electronic Regulations

(viii) Current Sweep method

(ix) DC-Link Capacitor Droop Control method
(x) Load Current or Load voltage Maximization method
(xi) dP/dV or dP/dI Feedback control method
(xii) State-based MPPT method
(xiii) One Cycle Control method
(xiv) Slide Control method
The Perturb and Observe method as well as the Incremental Conductance algorithms
are the most applied techniques. Esram and Chapman (2007) have reviewed all the above
mentioned techniques, which can be broadly classified in two categories: PV array
dependent MPPT techniques and PV array independent MPPT techniques.
A summary of PV array dependent and independent MPPT techniques is shown
in Tables 8.1a and 8.1b respectively.

Table 8.1a Characteristics of PV array dependent MPPT techniques

MPPT True Analog or Periodic Convergence Sensed
technique MPPT? Digital? tuning? speed parameter
Neural network Yes Digital Yes Fast Varies
Fuzzy Logic Yes Digital Yes Fast Varies
control
Fractional VOC No Both Yes Medium Voltage
Fractional ISC No Both Yes Medium Current
Current Sweep Yes Digital Yes Slow Voltage,
Current
State-based Yes Both Yes Fast Voltage,
MPPT Current
One Cycle No Both Yes Fast Current
control

Table 8.1b Characteristics of PV array independent MPPT techniques

True Analog or Periodic Convergence Sensed
MPPT technique MPPT? Digital? tuning? speed parameter
Hill-climbing/P & O Yes Both No Varies Voltage,
Current
Incremental conductance Yes Digital No Varies Voltage,
Current
RCC Yes Analog No Fast Voltage,
Current
DC Link Capacitor Droop No Both No Medium Voltage
Control
Load Current or Load Volt- No Analog No Fast Voltage,
age maximization Current
dP/dV or dP/dI Feedback Yes Digital No Fast Voltage,
control Current
Slide control Yes Digital No Fast Voltage,
Current
8.2 Classification of BOS 99

Fig. 8.2 Schematic diagram of ANN controlled SPV system

Rai et al. (2011) have discussed an artificial neural network (ANN) based
maximum power point tracking controller. The schematic block diagram of such a
controller is shown in Fig. 8.2. The controller consists of ANN tracker, optimal
control unit and dc-to-dc converter. The ANN tracker estimates the voltages and
currents corresponding to maximum power delivered by solar PV array for variable
cell temperature and solar radiation. The cell temperature is considered as a function
of ambient air temperature, wind speed and solar radiation. The ANN model consists
of three layers consisting one input layer of two neurons, one hidden layer of nine
neurons and one output layer of two neurons. The tracker is trained employing a set
of 124 patterns using back propagation algorithm in MATLAB. The mean square
error of tracker output and target values is set to be of the order of 10 5 and the
successful convergent of learning process takes 1281 epochs (Fig. 8.3). The accu-
racy of ANN tracker has been validated employing different test data sets. The
output of the control unit is a combination of voltage and current estimated by ANN
tracker and load resistance.
An ANN model of the maximum power point tracking controller is simulated
with SIMULINK (Fig. 8.4). The temperature and radiation block shows the input
layer. The individual inputs are weighted by corresponding elements of the weight
matrix (input_weights). The resultant product is added with bias matrix (input_bias)
that is transformed by hyperbolic tangent sigmoid transfer function (tansig), which is
the input to the hidden layer. The neurons of hidden layer are applied to the output
layer, which is approximated by a linear transfer function (purelin). Figure 8.5 shows
the actual data and the ANN-estimated maximum power current and voltage varia-
tions with the temperature and radiation. One can notice the good agreement
between the two sets of results. The relative error between experiment data and
simulated values on maximum power point is less than 5%.

Dc/Ac Inverters

A dc-to-ac inversion is required for ac loads. Inverters are broadly classified into two
categories: stand-alone inverter and grid tied inverter. In stand-alone system, the
inverter controls the voltage and frequency of the AC; sometimes called off grid
inverter. The grid tied inverters are intended to feed power into AC supply. The
100 8 BOS and Electronic Regulations

100

10-1
Training-Blue Goal-Black

10-2

10-3

10-4

10-5

10-6
0 200 400 600 800 1000 1200
1281 Epochs

Fig. 8.3 Mean square error vs. number of iteration

Output_weights
hidden_layer_weights
Input_weights
-C- Matrix
-C- Matrix
-C- Multiple + 1
Matrix + +
+ Multiple +
Multiple
1 + Purelin Peak Power
Product2 i&v
tansig Product1 tansig1
Radiation & Temp Product
-C-
-C-
-C- Output_layer Biases
hidden_layer Biases
input_bias

Fig. 8.4 ANN model of the maximum power point tracker implemented with Simulink

difference between the two categories is because of control algorithm which is

implemented using microprocessor/microcontroller.
An inverter converts the dc from the array or battery bank to single-phase or
three-phase ac as per load requirement. Quality inverters provide sine-wave electri-
cal output that conforms to utility power quality standards in terms of harmonic
content and voltage and frequency stability established for the power grid. All
utility-interactive inverters must meet minimum power quality standards set by the
American National Standards Institute (ANSI).
In modern solid-state inverters two techniques are mainly used to get a sinusoidal
output:
(i) Pulse-width modulation (PWM): In this technique, sinusoidal waveform is
generated by switching square-wave inverter stages at a rate higher than the
8.2 Classification of BOS 101

9
Actual
Actual
8
Actual
ANN 10°C
7
ANN 25°C
ANN 40°C
6
Current (A)

0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Voltage (V)

Fig. 8.5 Maximum power current vs. maximum power voltage for different values of temperature
and radiation

fundamental frequency. The variation in the conduction time of the power

switches (pulse width) controls the output voltage. The total harmonic distortion
(THD) in the inverter output is inversely proportional to the switching rate.
(ii) Waveform synthesis: In this technique, a stepped output waveform approximat-
ing to a sine wave is constructed by combining the phased outputs of several
inverter stages at the fundamental frequency switching. Higher level inverters
are used to lower the harmonic distortion and better approximation to sine wave.

8.2.3.2 Energy Control and Management Subsystem

This subsystem may have power condition and synchronizer, charge controller to
protect the battery bank from overcharged or deep discharge, timer, load switching,
monitoring, data logging etc. for control and management.
The charge controller regulates the flow of energy between the solar PV array and
the battery bank. These electronic devices prevent damage to the batteries by
protecting against both overcharging by the PV array and deep discharging by the
load. In the absence of charge controllers, the battery may suffer extreme wear and
tear and become inefficient. The life of the battery will be short and in worst
condition a short circuit may occur. In the case of standalone solar PV system
with battery storage, the floating condition occurs when battery voltage fluctuates
102 8 BOS and Electronic Regulations

Table 8.2 Summary of regulation process

S. No. Constraints Command Remarks
1 If Vb > Vmax Disconnect battery from PV
array
2 If Vb > Vmax and Reconnect the battery to the Vmax – Vt1 gives the hysteresis
Vb < Vt1 PV array voltage
3 If Vb < Vmin Disconnect the battery from
load
4 If Vb < Vmin and Reconnect the battery to load Vmin – Vt2 gives hysteresis
Vb > Vt2 voltage

between maximum (100%) and minimum (50–60%) state of charge (SOC) of the
battery. This type of system is auto regulated during the day and at night the
regulation makes use of blocking diode which blocks the reversal of power from
battery to PV array.
In the case of overcharge and deep discharge conditions the battery voltage
reaches at some critical values. In that case the electronic regulation is characterized
by two functions:
(i) Knowing the constraints (Vmin and Vmax).
(ii) Collecting information on the state of charge of the battery (battery voltage, Vb).
The summary of the regulation process are given in Table 8.2. A number of
different models are available, from basic to sophisticated one. Desirable features
include low-voltage load disconnect, field adjustability and provisions for a finishing
charge. Because battery charging characteristics change with temperature, battery-
temperature compensation is recommended when selecting a charge regulator in all
but the most-basic systems.

8.3 Electronic Regulators

The solar PV array has an optimum load into which it can deliver its maximum
power which varies with solar insolation and temperature. The battery offers a load
to the array by providing a bias to it that varies with battery state of charge and
temperature. The load derives power from the battery and changes its state of charge
which varies with load pattern. The interconnection of these power system compo-
nents must be designed to have them matched under all conditions. For small or
intermediate applications, stand-alone PV systems are self-sufficient in generating,
storing and supplying electrical power to local electrical loads.
8.3 Electronic Regulators 103

Fig. 8.6a Series regulator

Fig. 8.6b Shunt regulator

The electronic regulator may be either a stand-alone device or a control circuitry

integrated within a battery-powered device, or battery recharger. In PV array,
basically there are two methods by which the voltage and current can be controlled
either using series regulator or shunt regulator. In series regulator, a resistance is
introduced in series with array to reduce the load current and allowing the array
voltage to rise up to its open circuit value whereas a shunt regulator dumps currents
from pulling down both the load and array voltage (Figs. 8.6a and 8.6b). In both the
regulators a blocking diode is required to prevent the battery bank being discharged
either through the array or through the shunt regulator. These regulators may be
implemented by using transistors as variable resistor whose value is a function of the
state of charge of the battery. This approach requires the transistor to dissipate a large
amount of power and is generally not used.
The regulation can also be achieved if the transistor is used as switch which turns
on and off in such a way that the mean current flowing to the battery is reduced
(Figs. 8.7a and 8.7b). In the series switch regulator, the switch connects and
disconnects the current to battery bank. The mean current depends on the duration
for which the switch is on. If the switching is fast the voltage across the battery will
be constant. The excess power from the PV array is dissipated in cells of the PV
array.
The shunt switching regulator also operates in similar manner. The blocking
diode prevents the battery bank being discharged through the switch and also acts as
a series switch which is off when the controlled switch is on. The excess power is
dissipated in the cells of the PV array.
104 8 BOS and Electronic Regulations

Fig. 8.7a Series switch

regulator.

Fig. 8.7b Shunt switch

regulator.
Chapter 9
Repertoires of Applications

9.1 Space Applications

The solar conversion efficiency of first silicon solar cell was reported to be 6%. The
estimated cost of production of solar cell was USD 600 per Wp. The applications
that could be envisioned at that time were of power generation in remote locations
where fuel could not be easily delivered. The obvious application was satellites
where the requirement of reliability and light weight made the cost of solar PV
unimportant. The first ever application of solar photovoltaics in space was in 1958 in
Vanguard-I satellite. A small solar PV power unit of 6 mW, developed by space
engineers at United States Army Signal Corps (USASC), was used as a power source
in a back up radio transmitter. It was a rather successful trial and it sped the
laboratory test and evaluation research in area of solar PV. Simultaneously (1958
onwards) in USSR solar PV units were used to power the entire electronics of
satellites. These units operated successfully for 2 years. I960 onwards American
satellites also started using solar PV units to power the electronic systems. The
maximum capacity was of 150 Wp which paved the way for further capacity increase
of solar PV power units in satellites as follows:
Name of mission Year Solar PV power unit (Wp)
Nimbus 1964 500
Orbiting astronomical observatory 1966 1000
Extensive space applications were followed by the growth of market for space
programme and research inputs which in turn resulted in solar PV power cost
reduction to about USD 100–200 Wp. This price also was high to support mass
scale terrestrial applications.

© Capital Publishing Company, New Delhi, India 2018 105

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_9
106 9 Repertoires of Applications

9.2 Terrestrial Applications

Large scale solar PV application was triggered in early 1970 due to rapid escalation
of the cost of fossil fuels and increasing environmental concerns about their use.
In the initial stages solar PV was used for such assorted applications as solar
lantern in hilly areas and street lights in remote villages. In 1980 some commercial
companies came in the field of solar PV application such as in India Tata-BP Solar.
They identified several need-based clients such as:
1. Illuminating hamlets in scorching deserts;
2. Hospitals in far-flung areas using solar refrigeration for polio drops preservation;
3. Agricultural water pumping in villages which were not connected with grid; and
4. Distant towns entrepreneurs using solar PV power for furtherance of their busi-
ness opportunities (Figs. 9.1a and 9.1b)

Fig. 9.1a Solar PV power for illuminating hamlets in scorching deserts (Adapted from Tata BP
Solar promotion pamphlet)

Fig. 9.1b Solar PV power for remote refrigeration (Adapted from Tata BP Solar promotion
pamphlet)
9.3 Cost Effective Applications 107

This awareness of PV technology is rising and there are an increasing number of

systems beginning to appear as the concern for the need of ‘clean’ electricity is
gaining importance. The global SPV module in 1989 was 45 MW. It has now
increased to over 200 GWp. Some typical applications for photovoltaics today are
as follows:
Lighting and security and other consumer products
Village power
Remote refrigeration
Electricity for individual buildings or entire communities
Water pumping and irrigation
Supplementing the utility grid for distribution support
Communication
Remote monitoring equipment
Aids to navigation
Cathodic protection
PV/Diesel hybrid generators
Military applications

9.3 Cost Effective Applications

The reduction in cost of solar PV power has been the aim of some applications
during the last decades. These applications are described hereunder.

9.3.1 Solar Photovoltaics for Buildings

The distributed PV systems on buildings may be considered as an attractive appli-

cation to attain commercialization of solar PV power on the utility grid. It could be a
preferred power source for buildings in remote areas not served by the utility grid.
The PV systems on buildings would also avoid the cost of land required for ground-
mounted systems, as well as the cost of site development, foundations, structural
support systems, underground electrical distribution and the utility connection. In
practice the PV modules would become an integral part of the building, often serving
as the exterior weathering skin. The building would provide the PV aperture area and
support structure, and the building’s utility service would become the consumer of
PV electricity. In addition, PV systems on buildings would become a reliable power
source in areas plagued with frequent power cuts. The design and installation of
effective and efficient solar PV depends on several parameters as follows:
(i) Location: The amount of sunlight available will affect the amount of electricity
generated.
108 9 Repertoires of Applications

(ii) Orientation: The orientation of a building and positioning of solar arrays are
important factors to maximize energy production. In the northern hemisphere, a
south facing façade will collect maximum solar energy throughout the year and
produce the maximum electricity.
(iii) Tilt angle: The solar incident angle (θ) on a surface varies with declination
(time of the day and day of the year) and latitude of the location. At noon time
(time of maximum irradiance) it is expressed as

θ ¼ϕδ

where ϕ ¼ latitude of the site and δ ¼ declination angle.

So, several options are possible for the mounting of modules/array. It may be
placed at fixed tilt angle to maximize the generation of yearly average energy. The
tilt may be seasonally adjusted in accordance with the energy requirement in
summer, equinoxes and winter respectively. Owing to cost considerations, fixed
tilt arrays are often used.
The above parameters can easily be optimized in the design of a building. The
inclined building roof or the appropriate building wall may be chosen to maximize
the electricity production. The resultant systems are known as Building-integrated
PV systems (BIPV). The BIPV systems, therefore, have the potential to be architec-
turally attractive, which adds to the commercial acceptance of the technology on the
part of design professionals, building owners, and the general public.
Furthermore, BIPV components replace conventional building materials and
labour, thus further reducing the installed cost of the PV system.
With reduced installation costs, improved aesthetics and all the benefits of
distributed generation, building-integrated PV systems have become an attractive
option around the world for both residential and commercial buildings.
The BIPV power systems have been operated in various modes which can be
classified into following three categories:
(a) Stand-alone: The entire power is generated by SPV array and stored in a battery
to provide it in response to demand.
(b) Hybrid: In addition to SPV array such other means as wind generator and diesel
generators are also used to produce power.
(c) Grid-connected systems: In such systems the output of SPV plant is connected to
the grid line and there is no storage battery.
Photovoltaic systems for buildings can be either stand-alone, hybrid or grid-
connected. In a stand-alone system, the building has no connection to the utility
grid and often relies on a bank of batteries to store power for use at night and during
periods of limited sun. In India the hybrid systems have been considered to avoid
excessive cost of solar panels and the battery bank. For example, a 10 kWp solar PV
and Biomass Gasifier Hybrid Power Plant have been set up by Tata Energy Research
Institute to serve the hostel and 10 numbers (PL-11) standalone dusk-to-dawn street
lights. In a grid-connected or utility-interactive (UI) system, the building receives
electricity from both the PV array and the utility grid. Some PV systems are capable
9.3 Cost Effective Applications 109

Fig. 9.2 Indira Parayavaran Bhawan, New Delhi

of both stand-alone, utility-interactive modes of operation and or hybrid mode. In

India an innovative application of building integrated PV system has been demon-
strated in office building of Ministry of Environment, Forest and Climate Changes.
The building is referred to as Indira Parayavaran Bhawan (Fig. 9.2) and it is located
at Jorbagh Road, Aliganj, New Delhi (28.58 N). This is net zero energy multi-
storeyed building with 100% on-site power generation.
The details of the BIPV system are as follows:
Capacity of power generation: 930 kWp
Annual energy requirement: 14 lakh units (kWh)
Net energy consumption: Zero
Total area of solar panels: 460 m2
Photovoltaic panel: Mono-crystalline (20% efficiency)
Nature of power generation: Grid-Interactive
Owing to the advantages and promise of continued cost reductions, the potential
for PV-powered buildings has escalated. In a best-case scenario, Arthur D. Little,
Inc. (1995), has projected that the annual U.S. market for PV on buildings could
exceed $2.5 billion in the next 10 years. Japanese and European markets, already
under development, hold similar potential.

9.3.1.1 Photovoltaics as Glazing Material

Mounting PV panels on buildings have been carried out since the solar photovoltaics
panels became commercially available. In the initial stages building clading with
solar photovoltaic panels was carried out in retrofit manner. The traditional PV
110 9 Repertoires of Applications

modules were manufactured with aluminium frames. These modules were designed
for grouping on racks to be placed on buildings or on the ground for the reduction of
building electrical load. The installation was often carried out in the post construc-
tion phase of the building. Installing the modules meant drilling holes in a south
facing roof and installing them on expensive racks. In the event of roof repair the PV
installation had to be removed. In recent years the demand for solar PV module as a
building element has been recognized; the modules could then mimic windows or
skylights or they can become part of the actual roofing material. This application has
been referred to as building integrated photovoltaics (BIPV). Hundreds of projects in
Europe, USA and Asia have been completed and several others are in the pipeline. In
what follows we present case studies of some of these projects.
BIPV means that the solar features become part of the weather-proof skin of a
building. The array takes over the normal function of roof and is not just placed on
top of an existing structure. It serves both as a roof covering and as an electrical
production device. Based on this concept, several BIPV R&D demonstration units
have come up around the world. The photovoltaics websites give an introduction to
the construction and performance of these BIPV roofs and walls. Some of these units
that we have studied through website may be listed as follows:
(i) BIPV at the School of Construction Management of University of Reading,
UK
The University of Reading was refurbished a few years ago. In this process
south elevation of the roof was considered an excellent site for BIPV. It is
unshaded and captures maximum solar energy in winter. The roof generates
electric power directly and provides part of the electric supply to the building.
The solar photovoltaic roof is also being used as an R&D demonstration facility
for advance training and research in the school of construction and management
and engineering.

(ii) Folk Centre has conducted several projects in the field of building inte-
grated photovoltaics
The centre has good theoretical and practical know-how in the area. It offers
general consultancy regarding production and marketing of building integrated
photovoltaics, including installation, operation & maintenance and supervision
of local production. The characteristics of some of these projects are as follows.
(A) The Building integrated photovoltaics in “Skibsted Fjord” training
centre
The solar cells are integrated in the window panes applying the princi-
ples of security window panes. The PV cells, in the installation in the
facades of Skibstedfjord, are placed in various patterns, and demonstrate
the possibilities to use the cells as an architectural effect. The patterns can
be used for shading as well as for decoration purposes.
9.3 Cost Effective Applications 111

(B) Folk Centre main building

Various organs of the main building have been integrated with solar
photovoltaic panel wherein the panels have been architecturally integrated
to serve some blind filters of sunlight in summer and generate electricity
for use in the building.
Some of these projects have been funded by Danish energy authority.
The intended objective of these projects is to demonstrate the use of
photovoltaics as roofing which adds a new architectural dimension to
solar cell. Until now solar photovoltaics have been retrofitted on top of
existing structures.
(iii) BIPV Projects in India
In India also several BIPV projects have been commissioned. The notable of
these projects are:
(A) 10 kWp solar PV Gasifier Hybrid Power Plant for Tata Energy Research
Institute, Delhi hostel and 10 numbers (PL-11) standalone dusk-to-dawn
street lights.
(B) Building integrated solar PV system plant at Mamata Machinery,
Hyderabad.
(C) 36.3 kWp, hybrid PV power plant Matrimandir, Auroville, Pondicherry.

9.3.1.2 Solar PV Metal Roofing

Solar PV metal roofing means that solar features become part of weather-proof roof
or wall of the building. Presently research and development activities in the field of
building integrated photovoltaic systems are concentrating on the development of
constituent products. It includes three general areas: integral roof modules, roofing
tiles and shingles, and integral modules for vertical facades and sloped glazing.
Many BIPV systems have been integrated into roofs because these have the highest
solar exposure and so provide the highest power output. Recent innovations have
also been made in regard to their use as vertical curtain walls and awnings because
integrated PV modules can replace facade materials, such as granite and architectural
glass. New semi-transparent thin-film modules are also being developed, so PV can
even function as a skylight or a window. In solar PV metal roofing the factors that are
often considered are as follows: The nature of PV technology, type of PV module,
type of battery technology and characteristics of charge controller and inverter.
In India, BIPV systems have used three types of modules: solar PV modules,
solar PV metal roofing modules and dummy modules. The salient features of dummy
modules are without solar cells. Instead of cells the back tedler is screen printed with
solar cell pictures which look similar to actual solar cells. The other related materials
used shall remain same except solar cells. Whatever the method of integration, BIPV
systems are often grid-connected.
This means that the solar power is fed into the existing mains electricity supply of
the building.
112 9 Repertoires of Applications

Fig. 9.3 Solar concentrator (Refraction type)

9.3.2 Concentrating Applications

For cost reduction, concentrating module is an important approach to overcome the

problem of high cost of solar cell. The low-cost optical concentrator focuses solar
radiation onto the cell array of area much smaller than the concentrator aperture
(Fig. 9.3). The concentration of radiation also leads to an increase in photon
generation current and hence cell efficiencies. The operation of solar cells under
high illumination requires the removal of heat from the cells to combat the drop
(0.45% of power per degree Kelvin) in efficiencies of cells. The outstanding
problems of this technology are the development of
• Efficient low cost concentrator
• Cost-effective tracking systems
• Concentrator cells
• Efficient low cost heat sink

9.4 System Classification

The solar PV power systems have been proposed and tested for generation of electricity
in outer space as well as in terrestrial environment. The concept of solar power satellite
was proposed in 1964 for the generation of PV electricity in outer space, conversion to
microwave of laser beams to be captured and reconverted into ac electricity for
9.4 System Classification 113

Fig. 9.4 Terrestrial solar PV system grouping with their capacities

distribution to terrestrial load centre. It now appears unattainable even by the end of
2016. However, the use of PV electricity in satellite electronic systems and spacecrafts
seems plausible. In terrestrial area the PV electricity has been used in capacity range of
10 Wp–500 Wp. The solar PV system grouping and their capacities range has been
shown in Fig. 9.4.
Chapter 10
Solar Photovoltaic System Design

10.1 Solar PV System

The Solar Photovoltaic (SPV) system consists of energy generating subsystem modules,
energy storage subsystem battery bank and power conditioning units to supply power,
either stand-alone or in conjunction with grid (grid back-up). It is gaining increased
importance as a renewable source due to its advantage like absence of fuel cost, little
maintenance and no noise and wear due to the absence of the moving parts, etc. In
particular, energy conversion from solar cell arrays received considerable attention in the
last four decades. Its size is chosen according to planned application. Several demonstra-
tion units of large size solar PV power plants have been constructed, operated and
monitored around the world. Most of these systems use silicon crystal solar cells. The
module may be flat in geometry and use unconcentrated solar irradiance or else it may
have concentrating optics (refracting or reflecting) as parts of its structure. Several options
are possible in module mounting as follows:
• PV modules may be placed at fixed tilt angle at which the average yearly energy
generated is maximized.
• The tilt may be seasonally adjusted; for example at latitude
15, 0 and latitude
+15 in summer, Equinoxes and winter respectively.
• The modules may be mounted on a tracking unit that follows the sun. Single angle
axis trackers follow the sun daily from east to west and two axis trackers further
include elevation control to correct for seasonal north-south sun movement.
This is accomplished by a Power Conditioning Unit (PCU), which consists of
following sub-systems:
• DC interface
• Charge controller
• Inverter
• Protection circuit
• AC interface

© Capital Publishing Company, New Delhi, India 2018 115

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_10
116 10 Solar Photovoltaic System Design

From practical point of view the solar PV system may be divided into following
types:
(i) Stand-alone solar PV system without storage
(ii) Stand-alone PV system with storage
(iii) Solar PV-hybrid system
(iv) Grid-Interactive solar PV system
In stand-alone solar PV system without any storage subsystem, PV modules or
array supply power to an electrical load (Fig. 10.1). Since the solar PV modules
generate variable dc during day time, the load should be chosen in such a way that it
should have capability to be operated by variable dc output. The internal resistance
of the solar cell depends on incident solar radiation and ambient temperature; the
maximum energy will be delivered if internal resistance of the solar cell is equal to
the load resistance. Since both the incident solar radiation and ambient temperature
vary during the course of a day, it is quite unlikely that internal resistance of solar
cell will coincide with the load resistance.
Some power loss occurs due to this mismatch. The power loss due to this
mismatch can be overcome by using a maximum power point tracker, which is
essentially a dc-to-dc converter (Fig. 10.2a). The converter converts the input dc
voltage available from the module to the load voltage without any change in the
delivered power, but with a corresponding change in the current. This type of system
can only operate as long as enough solar radiation is incident on the module or array.
Such a system cannot operate either during night time or whenever there is no solar
radiation.
The system can operate an ac load if a dc-to-ac inverter is incorporated between
array and the ac load (Fig. 10.2b).
The stand-alone solar PV system with storage battery becomes capable of sup-
plying power during any time of the day and it also provides stable load voltage and
power except for water pumping and battery charging applications (Fig. 10.3). A
blocking diode is used in order to prevent the reversal of power from battery bank to
PV module or array during night time or whenever there is no solar radiation.
In modern sophisticated large capacity stand-alone system, an electronic energy
management, control and monitoring subsystems are being used.
A PV-hybrid system is a combination of a stand-alone PV system and either
renewable energy sources such as wind, solar thermal or biomass generator or diesel
generator (Fig. 10.4). In such a system both the PV and non-PV sources meet the

Fig. 10.1 Schematic diagram of solar PV system without battery storage

10.1 Solar PV System 117

Fig. 10.2a Schematic diagram of solar PV system with MPPT controller for dc load

Fig. 10.2b Schematic diagram of solar PV system with MPPT controller for ac load

Fig. 10.3 Schematic

diagram of solar PV system
with battery storage

Fig. 10.4 Schematic diagram of solar PV system with battery and back-up generator
118 10 Solar Photovoltaic System Design

Fig. 10.5 Grid-interactive solar PV system

demand together. In PV-wind hybrid system appear to be a natural mix because both
are complementary to each other. The PV energy has been also successfully inte-
grated in hybrid mode with micro-hydel energy sources.
A grid-interactive PV system, solar PV acts only as a supplementary peak-load
sharing electrical energy source, whose load becomes the utility grid itself. In grid-
interactive PV system an appropriate power condition subsystem, which converts the
input dc energy to stable ac energy synchronized to the grid energy, is used for the
connection and synchronization of PV array to the utility grid (Fig. 10.5).
The surplus power is fed to the grid through a second inverter, which is usually a
line commuted.

10.2 System Design Considerations for Particular Location

Conventional power plants engineering design is based on installed capacity,

whereas solar PV system is highly modular in nature; so system capacity is often
not a consideration in its design. However, modular capacity addition can be
accomplished with due consideration of mismatch losses (Kaushika and Rai
2007). The solar PV system performance depends upon site parameters
(i.e. climatological data), system configuration and load parameters. Therefore, it
would be desirable to size each PV system individually for each set of site and load.
The solar PV systems can be operated in several modes as mentioned in Section 10.1
such as stand-alone with and without battery storage, hybrid and grid connected in
accordance with their several applications. The design methods of stand-alone
systems only have reached a stage of maturity; therefore, to start, we consider
stand-alone solar PV system for a particular location.
10.2 System Design Considerations for Particular Location 119

10.2.1 Mathematical Formulations

Generally, the PV array is kept inclined at an angle to horizontal, whereas in general

the data for incident solar radiation are available corresponding to flat horizontal
surface. Following Sukhatme (1984) and Bhattacharya (1998), G(β) the factor by
which global radiation on a horizontal plane is to be multiplied to obtain that on a
surface kept at tilt angle β facing South (or North in Southern hemisphere) is given
by:

r A
GðβÞ ¼ Rð1
r Þ þ ð1 þ cos βÞ þ ð1
cos βÞ ð10:1Þ
2 2
where


cos ðϕ
βÞ cos δ sin ωst þ ωst sin ðϕ þ ωst Þ sin δ
R¼
cos ϕ cos δ sin ω þ ω sin ϕ sin δ
ωst ¼ cos
1 ð
tan ðϕ
βÞ tan δÞ ð10:2Þ
ω ¼ cos
1 ð
tan ϕ tan δÞ ð10:3Þ


360ð284 þ N Þ
δ ¼ 23:45 sin ð10:4Þ
365

Furthermore, solar array output depends on solar radiation, wind speed and
ambient temperature which in turn governs the solar cell temperature. The operating
temperature of solar cell can be obtained as follows (Marwali et al. 1997):
  
T c ¼ T amb : þ T diff
T gradvw
T J
T A ¼ NOCT
20 ð10:5Þ
T J ¼ NOCT þ T A
20

For Indian climate NOCT is taken as 45  C (Treble 1991; Kaushika et al. 2005)

So T J ¼ ðT A þ 25Þ ð10:6Þ

The simple and convenient way to express the incident solar radiation on a unit
surface area is Equivalent Hours of Full Sunlight (EHFS). It is defined as number of
hours of incident sunlight at a place, when the sunlight intensity is constant value of
1000 W/m2.
To incorporate the effect of cell operating temperature the EHFS is converted into
another parameter called UAO, which is expressed in units of Watt hour per peak
Watt per day as follows (Bhattcharya 1998)
" #
Temp:Coefficient in%
UAO ¼ EHFS  1
100
 ð10:7Þ
ðMonthly max:amb:air temp:Þ
120 10 Solar Photovoltaic System Design

PV Array Output Energy ¼ (Peak Power Array capacity at STC)  (UAO) So the
energy balance may be expressed as.
PV Array Output Energy ¼ WL For system without battery storage.
WL ¼ daily mean load, in Wh (including all losses). For example if WL0 is actual
load, then WL ¼ WL0 /( f1  f2  f3) where f1 ¼ field wiring and mismatch loss factor,
f2 ¼ wiring and switch gear loss factor and f3 ¼ inverter efficiency system with
battery storage

CDB
WL ¼
n
n WL
or C¼ ð10:8Þ
DB
where C ¼ capacity in Wh, n ¼ no. of autonomy days, D ¼ maximum depth of
discharge and B ¼ battery efficiency.
The first approximation of the array size for a particular location is calculated as
follows:

Daily load in Watt hour

Minimum Array Peak Wattage ¼
Daily mean UAO for the year

Monthly array output is calculated and matched with the load to get the SOC of
the battery of the month. If SOC (1-D) is less than the specified value then the array
size is increased by 0.1% and the simulation is repeated. Finally the array peak
wattage required to maintain the given minimum value of SOC is obtained.
For every location, array is tilted at a fixed angle called optimum array tilt, which
is defined as the angle at which the array is tilted from the horizontal plane to receive
maximum solar radiation averaged over the year. The optimum array tilt angles will,
therefore, be the angle for which the total annual value of Unit Array Output is
maximum. Kaushika and Rai (2006) have calculated the optimum tilt angle for
14 stations in Indian region and its variation with latitude is shown in Fig. 10.6 which
shows the variation of optimum tilt angle of the array with latitude and fit of second

Fig. 10.6 Variation of 35

optimum tilt angle of the
30
array with latitude
Tilt angle (degree)

25

20

15
Tilt angle
10 Poly. (Tilt angle)

0
0 5 10 15 20 25 30

Latitude (degree)
10.3 Solar PV Design Aid Expert System 121

degree polynomial therein. The corresponding value of R2 is also shown; in regres-

sion analysis R2 value is an indicator of the accuracy of best fit. The relationship
between optimum tilt angle of the array and latitude of the site under consideration is
given by Eq. (10.9).

Optimum tilt angle ¼
0:0342 ðlatitudeÞ2

þ 2:265 ðlatitudeÞ
6:8898 ð10:9Þ
R2 ¼ 0:95

10.3 Solar PV Design Aid Expert System

A solar PV design expert system for the design of solar PV unit for various locations
spread over wide region like Indian continent may also be developed. The develop-
ment of knowledge base is an essential constituent of such an effort.

10.3.1 Development of Knowledge Base for Indian Region

A knowledge base for the design of stand-alone solar PV system corresponding to

Indian region for autonomy day of 3 is developed. It considers two modes of system
OPERATION: with battery storage and without battery storage. The knowledge base is
based on the year round performance of the system corresponding to climatological
data (Mani 1980) of 14 stations in Indian region. The knowledge base approach
involves combining both site and array characteristics in a single parameter referred
to as EUAO and expressing the composite parameter as a function of geographical
co-ordinates. A simulation model has been developed for the determination of
EUAO for a given load and battery capacity with the main idea that the SOC of
the battery does not fall below the specified depth of the discharge during the poor
insolation period. The different steps of the mathematical procedure are as follows:
1. Determination of load parameter such as average daily load in Watt-hour (Wh).
2. The battery capacity C is calculated from eq. (10.8).
3. A parameter referred to as EUAO is obtained as follows:

Daily load in Watt hour

EUAO ¼
Required array peak wattage

Kaushika and Rai (2006) have written a program in C language using the above
algorithm. The starting and final results of simulation for New Delhi, India for
autonomy days 3 are shown in Tables 10.1a and 10.1b respectively. The variation
122 10 Solar Photovoltaic System Design

Table 10.1a EUAO and array capacity corresponding to climatological data of New Delhi, India
for autonomy days 3 and daily mean load 1 kWh in first iteration [Starting: UAO ¼ 5.23 (Wh/Wp/
day), Array capacity ¼ 191.05 (Wp)]
Waste
At the end of Monthly array Monthly Surplus/ Battery energy
month output (Wh) load (Wh) Deficit (Wh) SOC (Wh) (Wh)
January 35528.14 31,000 4528.14 4411.76 4528.14
February 33600.61 28,000 5600.61 4411.76 5600.61
March 35981.54 31,000 4981.54 4411.76 4981.54
April 32402.35 30,000 2402.35 4411.76 2402.35
May 31823.15 31,000 823.15 4411.76 823.15
June 25811.09 30,000
4188.91 222.85 0.00
July 21943.93 31,000
9056.07
8833.22 0.00
August 22656.28 31,000
8343.72
17176.95 0.00
September 26441.98 30,000
3558.02
20734.97 0.00
October 32455.31 31,000 1455.31
19279.65 0.00
November 32975.26 30,000 2975.26
16304.40 0.00
December 34024.04 31,000 3024.04
13280.36 0.00
After first iteration: EUAO ¼ 5.20 (Wh/Wp/day), Array capacity ¼ 192.20 (Wp)

Table 10.1b EUAO and array capacity corresponding to climatological data of New Delhi, India
for autonomy days 3 and daily mean load 1 kWh after last iteration
At the end of Monthly array Monthly Surplus/ Battery Waste
month output (Wh) load (Wh) Deficit (Wh) SOC (Wh) energy (Wh)
January 37627.16 31,000 6627.16 4411.76 6627.16
February 37167.15 28,000 9167.15 4411.76 9167.15
March 42866.29 31,000 11866.29 4411.76 11866.29
April 39734.81 30,000 9734.81 4411.76 9734.81
May 38864.02 31,000 7864.02 4411.76 7864.02
June 33224.18 30,000 3224.18 4411.76 3224.18
July 29211.83 31,000
1788.17 2623.60 0.00
August 29259.14 31,000
1740.86 882.74 0.00
September 34772.85 30,000 4772.85 4411.76 1243.82
October 40668.31 31,000 9668.31 4411.76 9668.31
November 39635.66 30,000 9635.66 4411.76 9635.66
December 37861.94 31,000 6861.94 4411.76 6861.94
After last iteration: EUAO ¼ 4.33 (Wh/Wp/day), Array capacity 230.78 (Wp)

of EUAO with latitude and longitude for autonomy days 3 is depicted in Figs. 10.7a and
10.7b respectively. Figure 10.7a shows the variation of EUAO as a function of latitude
and fit of polynomial therein. It may be noted that the variation of EUAO with latitude
exhibits larger scatter (R2 ¼ 0.24) than the variation with longitude (R2 ¼ 0.42). This
10.3 Solar PV Design Aid Expert System 123

Fig. 10.7a Variation of 5

EUAO with latitude for
different longitudes (arid 4

EUAO (Wh/Wp/day)
and non-arid zones)
3

1 Non-Arid Linear (Arid)

Arid Linear (Non-Arid)
0
0 10 20 30
Latitude (degree)

Fig. 10.7b Variation of 5

EUAO with longitude
(cumulative station) for 4.5
autonomy days 3 4

3.5
EUAO (Wh/Wp/day)

2.5

1.5

1 EUAO
Poly (EUAO)
0.5

0
70 75 80 85 90 95
Longitude (degree)

could be due to specific geometry of Indian Peninsula where the cloudiness index and
ambient temperature is controlled by the monsoon. We have, therefore, considered the
variation of EUAO with longitude in the composition of our knowledge base. Further-
more, statistically significant polynomial fits are obtained when arid (Latitude 24 –30
and Longitude 68 –75 ) and non-arid zones are considered separately (Fig. 10.8).

EUAO ¼
0:0053ðlongitudeÞ2
þ0:8447ðlongitudeÞ
29:344 ð10:10Þ
R2 ¼ 0:42

For arid zone

124 10 Solar Photovoltaic System Design

Fig. 10.8 Variation of 5

EUAO with longitude for 4.5
autonomy days 3 when arid

EUAO (Wh/Wp/day)
4
and non-arid zones are 3.5
considered separately
3
2.5
2
1.5
1 Non-Arid Poly (Non-arid)
0.5 Arid Poly (Arid)
0
70 75 80 85 90 95
Longitude (degree)

EUAO ¼
0:0036ðlongitudeÞ2
þ0:5386ðlongitudeÞ
15:783 ð10:11Þ
R2 ¼ 0:79

For non-arid zone

EUAO ¼
0:0064ðlongitudeÞ2
þ1:034 ðlongitudeÞ
37:413 ð10:12Þ
R2 ¼ 0:56

Similarly the variation of EUAO with latitude and longitude for autonomy days
4–7 and the variation of EUAO as a function of latitude and fit of polynomial therein
can be obtained.
Kaushika and Rai (2006) developed a software tool referred to as an expert
system (Fig. 10.9) for the design of solar PV systems in Indian region. It uses the
values of latitude and as well as longitude of location and the number of autonomy
days as input and calculates the Effective Unit Array Output (EUAO) from the
corresponding polynomial equations. The EUAO is independent of load. For differ-
ent autonomy EUAO will be different. It then calculates other sizing parameters as
follows:
(i) Required array capacity (Wp): It is given by

Daily
mean load in kWh

Array capacity ¼
EUAO

(ii) Number of modules in series (Ns): It is given by

10.4 Web Access to Solar PV Design Aid Expert System 125

Fig. 10.9 Schematic diagram of an expert system

Load operating voltage

Ns ¼
Effective operating voltage of each module

(iii) Number of series string in parallel (Np): It is given by

Required array capacity

Np ¼
N S  Effective peak power rating of each module

Where effective peak power rating of each module is effective operating voltage of
each module  Imax.
The flow chart of calculations is given in Fig. 10.10.

10.4 Web Access to Solar PV Design Aid Expert System

The website corresponding to the expert system has been prepared and placed on
Internet and outline of the uploading procedure is as follows:
(a) We make self-extracting zipped file. The self-extracting zipped file does not
require win zip software for extracting the zipped file. The file can be
downloaded from anywhere on any machine. The necessity of zip in general
126 10 Solar Photovoltaic System Design

Fig. 10.10 Flow chart for the expert system

arises when there is more than one file accomplishing the required task. In the
present case there are more than one file required to run the knowledge-base
(e.g. the executive/compiled file of knowledge base program and the executive
file of turbo C compiler). The zipping saves space as well as download time.
(b) The executive files of knowledge base and Turbo C are zipped in a single file
named setup.exe with the help of WinZip software. Afterwards we make this
setup.exe file self-extracting.
(c) Then this setup.exe file is on the web server for downloading.
The address of the websites is http://www.nd-kaushika.tripod.com
Chapter 11
System Reliability Considerations

11.1 The Reliability

The cost and reliability are two outstanding barriers in the promulgation of solar PV
technologies on mass scale. In solar PV power system context reliability involves
two factors: (i) the solar array need the use of an optimum load to deliver maximum
power. This requirement varies with solar irradiance, temperature and battery bank
characteristics and often difficult to be realized in practice. (ii) The reliability also
means the ability of the system to continue functioning without failure for the period
of time intended and under the given operating conditions; for example, during the
array operation in field, the partial shadowing, soiling, cracking (provoked by hail
impact) or opening of a string may occur due to solder melting or damage to
encapsulate and in extreme cases it can lead to system failure. In both cases, the
system reliability can only be guaranteed on a probabilistic basis. It essentially
involves probability of success/failure, adequate performance, on/off time and
operating conditions. Thus the probability provides the numerical input for the
assessment of reliability and also the first index of system adequacy. For mission-
oriented project this definition of reliability can be regarded as suitable measure.
However, it becomes rather unsuitable measure for those continuously operated
systems that can accommodate failure. The suitable measure used for these systems
is availability. For example, in power system, availability is the percentage of time
the system will deliver power to its load (Stember 1981). The required availability is
different for different types of systems. Availability may be obtained by dividing
“up-time” by “up-time plus down time” or it can be expressed as

MTBF
A¼
MTTR þ MTBF
where A ¼ availability, MTTR ¼ mean time to repair and MTBF ¼ mean time
between failure.

© Capital Publishing Company, New Delhi, India 2018 127

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_11
128 11 System Reliability Considerations

Thus availability can be increased by increasing reliability i.e. making MTBF

longer or by decreasing MTTR.

11.2 Site Dependence

The performance of a solar PV system is site dependent since PV array output is a

function of cell temperature and solar irradiance which depends on the geographical,
climatological and environmental factors like the latitude, global, diffuse and
reflected solar radiation, ambient air temperature and wind speed. The array perfor-
mance also depends on the uniformity and homogeneity of solar irradiance on the
PV array surface.
Furthermore at particular location, the angle at which the sunlight is incident on
the PV collector surface influences the array output. The performance of the array
will, therefore, depend on whether it is made to track the sun during the day, month
and year, or it is installed at a fixed tilt angle. A fixed tilt array should face due
geographical South in the northern hemisphere and due North in southern hemi-
sphere. The deviation in the azimuth angle from due south (or due north) introduces
randomness in system performance. The average daily performance is also depen-
dent on the tilt angle at which the array is installed with respect to the horizontal. The
optimum array tilt angle for a fixed-tilt array depends on the local geographical and
climatological factors. For example from geographical considerations for maximum
yearly energy gain the tilt angle is taken to be 0.9 L where L is latitude angle. For
seasonal adjustment the angle to be adjusted L  15 . The load as well as system
losses do not remain constant over the year. Thus solar PV systems are also highly
load-specific.

11.3 System Availability

The system availability means its availability to serve the load; it is an important
design parameter for solar PV systems. It is the percentage of time in a whole year,
during which the system is able to satisfactorily meet the load. The availability is
100% if the system is able to meet satisfactorily all designated loads during that
period. Conventional reliability parameter used in this connection is Loss-of-Load-
Probability (LOLP) expressed as the total number of days, during the period of one
year, when the system is incapable of meeting the load. The system availability may
be expressed in terms of the Loss-of-Load-Probability (LOLP) as follows:

365  LOLP
System Availability ¼  100%
365
11.5 Power Loss and Hot Spot Effects 129

System sizing exercises often attempt to ensure a high value of annual availability
or a low value of LOLP. In case of solar PV systems this approach may lead to an
expensive system design, because that may demand high system array capacity. On
the other hand, a low value of the system availability will normally not be desirable
from the viewpoint of the user. Therefore, the system designer must use a pragmatic
and judicious approach to ensure reasonably high system availability without mak-
ing it unduly expensive. An annual availability value between 90% and 99% is
generally considered acceptable for most stand-alone systems.

11.4 No Sun Days and Storage Autonomy

The measure of ‘sun shine days’ has been used as a measure of reliability in solar PV
systems. Larger the number of sun shine days higher the reliability of the solar PV
system. The number of consecutive off-sun shine days in a month decides the period
during which the storage battery is required to meet the load completely. The storage
battery provides storage autonomy to counter balance the possibility of no PV
generation during off-sun shine (e.g. rainy or cloudy) days. The storage autonomy
is referred to as NO-sun days. Larger value of storage autonomy means a larger
battery capacity and higher cost.
The capacity C of a storage battery, having storage autonomy of n days, may be
expressed as follows:

nL
C¼
DB
where L is the daily mean value of load, expressed in the same units as the battery
capacity C, D is the permitted Maximum Depth of Discharge (MDOD) expressed as
fraction and B is the battery efficiency.

11.5 Power Loss and Hot Spot Effects

The undesirable feature of solar PV arrays delivering power lower than the array
rating in field conditions could be fundamentally due to mismatch in the output
of the individual cells. This power loss may be due to variety of reasons as
follows.
130 11 System Reliability Considerations

11.5.1 Manufacturer’s Tolerances in Cell Characteristics

Solar cell fabrication procedure involves several controls despite all automations; the
resultant cells possess significant manufacturer’s tolerances. Performancewise they
exhibit variations in current voltage characteristics. In particular, current (Imax) and
voltage (Vmax) of individual cells operating at maximum power (Pmax) exhibit
statistical distribution. So when mass produced cells are connected in series and/or
in parallel within a network, the statistical distribution of cell parameter leads to
electrical mismatch in the network; the mismatch limits the short circuit current of
the combination to a value closer to the lower end of the distribution than to average.
Consequently the maximum power output of the network is often less than the sum
of the maximum power of individual cell; this is referred to as mismatch loss. It may
tend to enhance with time due to degradations resulting from aging. The detailed
analysis of power loss due to mismatch in cell characteristics has been carried out in
Chapter 7.
Similar results would occur with strings combined in parallel; open circuit voltage
distribution will generate inter-string power dissipation.

11.5.2 Environmental Stresses

It includes the partial or full opening of a string due to cell cracking (provoked by
hail impact) or otherwise. This not only gives rise to a mismatch loss but also could
result in excessive heating of the regions of power dissipation which in turn can lead
to solder melting and damage the encapsulant. The regions of excessive heating are
referred to as hot spots.

11.5.3 Shadow Problem

In the field solar cells arrays are subject to shadows from both predictable sources as
well as from such unpredictable sources as bird droppings or fallen leaves. The
percentage power loss is much greater than the percentage of array area which is
shadowed; for smaller arrays with few or no parallel connections, one leaf could
cause the system output to drop to a fraction of rated power, eventually resulting in
system failure. Shadowed cells in series with illuminated cells block the current flow
in entire series connection and tend to become reverse biased. Consequently, hot
spots are developed in the localized regions of shadowed cell, which can damage the
encapsulation or crack the cell causing module failure.
11.6 System Reliability Improvements 131

11.5.4 Array Operation at Low Voltage

The operation of the array at low output voltage occurs in some array voltage
regulation schemes and is also likely to occur when maintenance is carried out and
some sections of the array are sorted out for the safety of the maintenance crew. A
low output cell in such cases gets reversed biased and exhibits hot spots. For this
reason some manufacturers now specify ‘safe operating regions’ for their modules.

11.6 System Reliability Improvements

Quality and redundancy are the two main attributes by which the reliability can be
affected. Quality concerns not only the physical materials and components used in
the system but also the quality of manufacture, testing, calibration, transport and
operation. This in turn depends on the quality and experience of personnel involved
and the ergonomics and environment of work place. This measure is rather cost-
intensive for solar PV system and has been adopted only in such mission-oriented
systems as space satellites. The redundancy is based on the basic premise that
component will fail from time to time and that there should be sufficient back up
in the system so that the function of failed component is absorbed by another.
Redundancy is of two types: active and standby. Active redundancy corresponds
to the situation that the system components share the normal function amongst
themselves; when one or more component fail, the activity is automatically absorbed
by the remaining components. This is also referred to as parallel redundancy.
In contradistinction to this is the standby redundancy wherein the redundant compo-
nent or subsystem remain in a standby state and are switched into the system only when
main operating component fails. Reliability improvement requires additional investment
and raises a question where or on what the money should be invested in the system to
achieve maximum reliability benefit. Reliability and economics, therefore, act as
conflicting variables in the decision process. In what follows, therefore, we examine
the efficacy of redundancy in terrestrial solar PV system.
The function block diagram of a simple PV system is shown in Fig. 11.1a; it
exhibits the subsystem level interconnections. Following Stember (1981) and
Billinton and Allan (1992), the above system may be analysed to the subsystem
level using two reliability techniques:

Fig. 11.1a Functional block diagram of a solar PV system.

132 11 System Reliability Considerations

Fig. 11.1b Fault Tree for

solar PV system

Table 11.1 Results of fault State of subsystem

modes and effects analysis
1 2 3 System failure SFNOS
(FMEA) for solar PV system
0 0 0 0
0 0 1 1
0 1 0 1
1 0 0 1
0 ¼ Operating, 1 ¼ Failed, SFNOS ¼ System failed no output sun
available

• Failure Modes and Effects Analysis (FMEA)

• Fault Tree Model
Fault Tree (Fig. 11.1b) and results of FMEA (Table 11.1) are shown in terms of
the consequences of the failure of each subsystem. The system is series connected
and an interruption of any of the subsystems will prevent the output being delivered
to the load. Consequently, the system is operational only if all the subsystems are
operational. Each of the subsystem has a similar effect on the system and its failure
causes the failure of the system.
Let us further consider the above system with redundant PV array (Fig. 11.2a).
The corresponding FMEA (Table 11.2) and Fault Tree (Fig. 11.2b) show that a
failure of either of the PV array does not cause system failure; failure of both only
causes the failure. So due to redundant PV array the failure mode is significantly
reduced. However, PV array is a costliest component of the system, so its back up
redundancy at subsystem level is not an attractive option.
System reliability can also be improved by affecting fault tolerance at component
level. In what follows we examine the effectiveness of this approach.
The fault tolerant circuitry can be considered as an arrangement of solar cell
interconnections, which is capable of producing acceptable power output even if the
11.6 System Reliability Improvements 133

Fig. 11.2a Solar PV system with redundant PV array

Fig. 11.2b Fault tree for

solar PV system with
redundant PV array

module/array itself suffers from the effects of electrical mismatches. Rauschenbach

(1971) proposed and showed that the electrical output of shadowed solar cell arrays
could be considerably improved if each series block of parallel string was shunted by
a diode.
Subsequently Sayed and Partin (1975) and Swalesh and Green (1982) analysed
the effect of shading on terrestrial solar cell arrays and investigated the reduction in
power loss and hot spot effects by judicious application of shunt diode. Gonzalez
and Weaver (1980) have shown that the efficiency of an array can also be enhanced
by redundant circuit design. Green et al. (1981) reported a new technique for
integrating the bypass diodes into solar cell structure. Bishop (1988) and Shephard
and Sugimura (1984) have studied that the effect of electrical mismatch can be
reduced by series-paralleling solar cell interconnection scheme. The effect of elec-
trical mismatches can be substantially reduced and hence the performance of the
134 11 System Reliability Considerations

Table 11.2 Results of fault State of subsystem

modes and effects analysis
1A 1B 2 3 System failure SFNOS
(FMEA) for solar PV system
with redundant PV array 0 0 0 0 0
0 0 0 1 1
0 1 0 0 0
0 0 1 0 1
1 0 0 0 0
0 0 1 1 1
0 1 0 1 1
1 0 0 1 1
0 1 1 0 1
1 0 1 0 1
1 1 0 0 1
0 ¼ Operating, 1 ¼ Failed, SFNOS ¼ System failed no output sun
available

array can be improved by a proper arrangement of cross connections in the module.

These authors presented experimental results verifying proposed cell processing
sequence and system performance of the cells. The concept has the proof but has
not yet reached the stage of commercial exploitation. However, most of the present
day manufacturers of solar modules use an external bypass diode to take care of
partial shadow and opening of a string due to crack in cell. Simultaneously with this
practice it has also been realized that one may obtain substantially more output from
shadowed array with proper arrangement of cell interconnections in the modules.
Several interconnection schemes have been proposed (e.g. Brandstetter 1984; Singer
et al. 1984; Gautam and Kaushika 2001; Gautam and Kaushika 2002b). It is of great
practical importance to find the configurations less susceptible to hot spots and
excessive power loss. In what follows we investigate four configurations of array:
• Series-paralleling (SP) configuration
• Series parallel with bypass diode
• Bridge-linked (BL) configuration
• Partially cross-ties configuration
The schematic of series-paralleling concept is illustrated in Fig. 11.3a. To ascer-
tain the effectiveness of various degree of series paralleling, the power output of an
array under specific failure condition must be determined for variety of series-
parallel configuration. The power loss varies inversely with degree of series
paralleling, but limit imposed by physical constraints, as well as cost and thermal
problem, make it impossible to achieve negligible power loss by means of unlimited
series paralleling.
Figure 11.3b exhibits the use of bypass diode. A bypass diode is connected in
parallel with a given number of series blocks. The bypass diode does not conduct
unless the series block is reversed from normal polarity by a small amount, usually
say 0.5 V. Then the diode passes unlimited current at its operating voltage. The
11.6 System Reliability Improvements 135

Fig. 11.3 Different interconnection schemes for solar PV array

purpose of using a diode is to prevent the series block from operating at high reverse
voltage. In this state it absorbs power from the system and cells operating with
reversed polarity becomes hot, leading to physical degradation of cell and module.
This condition is known as back biasing which results in hot spot problem.
Figure 11.3c illustrates bridge-linked configuration, which is accomplished by
making cross connections between parallel strings of cells of series-parallel config-
uration. The cross connections are in bridge rectifier fashion so called as bridge-
136 11 System Reliability Considerations

linked configuration. The power loss varies inversely proportional to square root of
M (parallel string) N (series block) so the relative advantage of the BL array over
the SP array increases with increasing the array size. In the BL array the surrounding
cells share the effect of reverse bias on the shadowed cell.
Figure 11.3d shows partially cross ties, which is obtained from the simple series-
parallel array by connecting ties across row of junctions; it may be characterized as the
scheme with higher cell interconnection redundancy. The experimental investigations of
electrical mismatches due to shadow effect have been carried out to find an array
configuration, which is comparatively more fault tolerant. Module of size 6  6
(Fig. 11.3) have been considered to carry out the investigations. Two cells of a string
were shadowed. Figures 11.4, 11.5 and 11.6 portray corresponding I-V curves for the

Fig. 11.4 I-V

characteristics when solar
cells are connected in series-
parallel with diode

Fig. 11.5 I-V

characteristics when solar
cells are connected in
bridge-linked configuration
11.6 System Reliability Improvements 137

Fig. 11.6 I-V

characteristics when solar
cells are connected in
partially cross-tied
configuration

interconnection schemes of Fig. 11.3. These are drawn under conditions of known
temperature and stable solar irradiance. The results correspond to 6  6 modules wherein
two cells of a string have been shadowed. It is obvious that the bridge linked solar cell
interconnection configuration turns out to be superior to both series parallel and partial
cross ties configurations in fault tolerance due to shadow effect.
Chapter 12
High Performance Solar Cells

12.1 Solar Cell Developments

In recent years research in solar cell physics and reliability aspects of solar photo-
voltaic systems has exhibited potential for significant cost reduction as well as
functionality amelioration. In this chapter high efficiency and cost-effective
approaches in solar energy conversion to make solar photovoltaics a commercial
perspective has been discussed.
Many solar cell technologies using different types of materials are available
today, and an even larger number is expected to be available in the near future.
These may be classified into three generations, depending on the basic material used
and the level of commercial maturity. The basis of classification is as follows:
First-generation PV systems (fully commercial) use the wafer-based crystalline
silicon technology, either single crystalline or multi-crystalline.
Second-generation PV systems are based on thin-film PV technologies and
generally include three main families: (1) amorphous (a-Si) and micromorph silicon
(a-Si/μc-Si); (2) Cadmium-Telluride (CdTe); and (3) Copper Indium-Selenide (CIS)
and Copper-IndiumGallium-Diselenide (CIGS).
Third-generation PV systems include technologies, such as advanced thin film
technologies (multi junction technologies), concentrating PV (CPV) technologies
and organic PV cells technologies that are still under demonstration. It also includes
novel concepts under development.

12.2 First Generation Solar Cells

It corresponds to fully commercial solar cells which include cells made from pure
mono-crystalline silicon with almost no defects or impurities. These cells in general
have a solar conversion efficiency of about 15–17%. The most efficient and

© Capital Publishing Company, New Delhi, India 2018 139

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_12
140 12 High Performance Solar Cells

expensive single-junction silicon cells have been reported to have about 22%
efficiency. The fabrication of these devices is energy, material and time intensive;
they are expensive to produce. As a follow up to these cells, the polycrystalline solar
cells were produced from slightly poorer grades of mono-crystalline silicon or
semiconductor grade silicon. The cells have white speckles on the surface due to
impurities. They are comparatively less expensive since simpler processes are
involved in their production. They have a solar conversion efficiency of about
10–12%.

12.3 Second Generation Solar Cells

Cost of electricity generation from solar cells depends on many factors but most
important are the cost of material used in the fabrication of solar cells and the cost of
manufacturing processes. The reduction of these costs is achieved in solar cell made
from thin film of material rather than from usual slice of crystalline silicon. The
thickness of the thin film can range between 10–100 micron (10 6 m); there are wide
variety of techniques which can be used to deposit thin films of amorphous silicon as
well as of several polycrystalline materials (the materials are intermediate between
amorphous and crystalline). They absorb sunlight more effectively than crystalline
cells and can therefore be thinner. Amorphous silicon solar cells have lower
manufacturing cost than crystalline cell but have a low conversion efficiency of
5–7%. However the reduced cost often overweighs the reduced efficiency leading to
the increase in ratio of performance to cost. Almost all polycrystalline film solar cells
are heterogeneous consisting of an n-type material and a different p-type material.
Two types of polycrystalline thin film cells have been shown to have characteristics
which make them attractive for commercial production. These are: the copper
indium diselenide/cadmium sulphide cell and cadmium telluride/cadmium sulphide
cell. Following three primary types of thin film solar cells have been commercially
developed:
1. Amorphous silicon (a-Si) and microcrystalline silicon (micro-Si);
2. Cadmium Telluride (Cd-Te); and
3. Copper-Indium-Selenide (CIS) and Copper Indium-Gallium-Diselenide (CIGS).
Thin film technology has been successfully used on rigid, flexible, curved and
foldable substrates. Amorphous silicon efficiencies are rather low 4–8% and amor-
phous silicon with layer of micro crystalline silicon may increase the efficiency up to
10%. The Cadmium-Telluride thin film cells have lower production cost and higher
efficiencies up to 16.7% (Green et al. 2011). However, the main disadvantages are
that cadmium is byproduct of zinc mining; tellurium is byproduct of copper
processing and cadmium has issues related to its toxicity. CIS and CIGS have the
highest efficiencies of all thin film cells and efficiency of 23.3% has been achieved in
the laboratory.
12.4 Third Generation Solar Cells 141

Multiple junction approach has also been proposed (Jackson 1955) for the
fabrication of high performance solar cells. The approach is depicted in
Chambouleyron (1991). The multiple gap devices consists of a series of single gap
cells fabricated one over the other; each one uses a different region of solar spectrum.
In principle, multijunction devices with any of single crystal, poly crystalline or
amorphous structure can be built. However, the single crystal structures involve the
problem of lattice mismatch between different materials. Current research addresses
such problems.

12.4 Third Generation Solar Cells

Over the years the cost of solar photovoltaic electricity has gradually decreased but it is
still far from proving to be affordable by masses. Third generation solar PV systems are
characterized mainly by the fact that they offer substantial improvement in the efficiency
of solar conversion or a large reduction in their cost of production as compared to
previous generation of system.
For example, in the last few years there has been an interest in thin film nano
crystalline silicon solar cells. Low temperature plasma enhanced chemical vapour
deposition (PECVD) has been considered as a suitable deposition method (Ray
2009). In university of Neuchatel Switzerland, Shah et al. (2003) have fabricated
nanocrytalline silicon solar cell using very high frequency glow discharge method
and attained 7% efficiency. Yet another class of thin film solar cells includes organic
solar cells (Pulfrey 1978; Hall et al. 2001) and dye-sensitized solar cell (Brian
O’Regan and Gratzel 1991). The potential of organic cells to be a viable candidate
for solar electricity generation hinges on various factors such as materials, produc-
tion techniques, working principles and the stability of solar cell device. Gunes et al.
(2007) have reviewed these aspects of organic solar cells. Materials having
delocalized π electron system can absorb sunlight, create excitons and transport
these carriers. The organic solar cells use conductive organic polymers or small
organic molecules materials for light aborption and production of excitons (Suresh
et al. 2009; Sharma 2009; Lokhande and Dhawla 2009).
The effective field for breaking up of excitons into free electron-hole pair is set up by
the heterojunction resulting from donor (D) and acceptor (A) material. Two types of
architectures namely, bilayer heterojunction (Hoppe and Sariciftci 2004) and bulk
heterojunction (BHJ) have been proposed and tested (Sharma 2009). BHJ solar cells
have essentially used soluble conjugate polymers and small molecules as donor mate-
rials. The solar conversion efficiency up to 7.4% has been reported for these cells
(Mikroyannidis et al. 2009; Sharma et al. 2010). Dye-sensitized solar cell (DSSC) is
based on the basic premise that organic dyes can generate electricity at oxide electrodes
and it is basically a photoelectron chemical system (Brian O’Regan and Gratzel 1991). It
is often referred to as Gratzel cells. In dye-sensitized solar cell the photoelectrons are
provided from the photosensitive dye and the charge separation occurs at surfaces
between the dye semiconductor and electrolyte (Sivakumar and Anandau 2009). Its
142 12 High Performance Solar Cells

function is based on photo induced charge separation by dye-sensitized interface

between a nano crystalline, mesoporous metal oxide electrode and redo electrolyte.
Overall peak power conversion efficiency for the dye-sensitized solar cell has been
reported 11% (Gao et al. 2008) and record efficiency for prototype lies at 12.3% (Yella
et al. 2011).
Furthermore, the major barrier is the low conversion efficiency of first and second
generation solar cells. The most efficient and expensive single-junction silicon cells
have been reported to have about 22% efficiency and the upper obtainable efficiency
limit (Shockley-Queasier) for a perfect silicon homojunction solar cell is 31%. It is
due to inefficient utilization of solar radiation spectrum and single electron band gap
of solar cell material. The definition of third generation solar cells has, therefore,
been set as the solar conversion efficiency greater than theoretical efficiency
(Shockley-Queasier limit) and low cost. The main approaches include multi junction
cells, quantum dot technologies and modification of the photonic energy distribution
prior to absorption in the solar cell.
In this category of high efficiency solar cells nanostructure solar cells have been
reported which include several advanced concepts of quantum dot solar cells (e.g. Nozik
2001). Quantum dot is a special type of nanocrystal. It enables several advanced concepts
of solar cell design to be implemented which includes intermediate band and multiple
exciton devices, hot carrier and up/down conversion devices.

12.5 Limited Demonstration of Concentrating Solar

Photovoltaic

We have come across limited demonstration of concentrating solar photovoltaic

system. This includes reflecting concentrating systems by Tata BP Solar India and
concentrating solar photovoltaic refracting type 300 kW and 20 kW Entech Solar PV
Systems, USA (Figs 12.1a, b).

Fig. 12.1 Concentrating solar PV system: reflecting type. (a) 300 kW solar PV system. (b) 20 kW
solar PV system
Chapter 13
Solar PV System Economics

13.1 General

In the present day world we come across various energy supplies or projects which
are based on many energy sources (conventional and new as well as renewable).
Consequently, there are various forms of utilizable energy, various utilization
technologies, and various applications at various locations. Presently solar photo-
voltaics is being recognized as a principal emerging technology growing well over
30% every year. Mass scale adoption of solar PV technology as an energy resource
will depend on how this is worth economically. Operationally the task of evaluation
consists of identifying the potential advantages and disadvantages of this option.
This constitutes the problem of system evaluation.

13.2 System Evaluation

The methodologies of system evaluation have been developed in three disciplines as

follows:
1. Engineering: Benefit-cost or cost-benefit analysis
2. Economics: Social and welfare economics or social cost-benefit analysis
3. Operations research: Decision analysis
Each of these disciplines focuses on a separate set of issues. Each is, therefore,
appropriate for a different kind of problem. So the first task of an analyst is to choose
the best method for a specific problem. The second task involves ‘precision’ in
decision making. It depends on the assumptions one makes about the situation. As a
rule precision is decreased by making more simplifying assumptions; conversely,
more simplifying assumptions make the evaluation less acceptable.

© Capital Publishing Company, New Delhi, India 2018 143

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1_13
144 13 Solar PV System Economics

To illustrate the relationship between simplification, precision and reality let us

consider an example for evaluation. Suppose we have savings and we want to invest,
following choices are available for different interest rates as follows:

Choices Different monetary return (in terms of interests)

Saving bank account 7%
Fixed deposits receipt (with bank) 10%
Fixed deposit receipt (Public sector) 14%
Fixed deposit receipt (Private sector) 18%
Investment in business 30%

In the above situation the decision making involves the following complicating
factors:
1. Risk
2. Our objective
(a) Make profit only
(b) Protects savings
(c) Flexibility in use
If we consider only the monetary gain the decision becomes pin pointed but it
involves elimination of risk and flexibility to use the capital. If we further eliminate
the consideration of risk the decision becomes more unrealistic for the results to be
acceptable. Further complicating factor is flexibility in use.

13.3 Time Value of the Money

The most basic approach to evaluation in engineering economics is to encompass the

situation wherein one considers comparing money over large span of time. For such
situations in which various projects have predominantly financial effects and invest-
ment at different times, it is, therefore, suitable either to reduce cost or produce
marketable goods at same time.
Furthermore, the engineering projects evolve over a long time. Costs incurred in
one period generate benefit for many years. The essential problem in the evaluation
of such projects comes from the fact that money has a time value. A dollar now and a
dollar later are not the same. We cannot estimate total benefit (costs) simply by
adding up the benefits (or costs) that occur at different times. This would be like
adding apples and oranges. We, therefore, need to translate all costs and benefit with
comparable amounts. This is by the use of parameter known as discount rate which
may be defined as follows:
• A dollar now is worth more than a dollar in future because it can be used
productively between now and then. For example, one can put the money in
saving bank account and get a large amount back.
13.4 Formulations for Evaluation of Money Value Over Time Span 145

• Furthermore, any given amount of money now is typically worth more than the
same amount in future because of inflation. As prices go up due to inflation, the
money you have now will buy less and less.
The discount rate represents the way money now is worth more than money later.
It determines by how much any future amount is discounted or reduced to make it
correspond to an equal amount today. Discount rate is similar to interest rate but with
a difference that discount rate represents real change in value to a person or group as
determined by their possibilities for production use of money and the effects of
inflation. As a general rule, discount rate is different than interest rate. For the time
being, it is sufficient to recognize that the discount rate is mathematically similar to
an interest rate and is related to yet another term escalation ratio, d through the
escalation rate i.
The discount rate appears in evaluation calculations as the discount multiplier
(1 + r) where r is a fraction. The influence of currency inflation and other commod-
ities market forces is generally exposed by a fraction referred to as escalation rate.
The escalation rate appears in evaluation calculation as the escalation multiplier
(1 + e) where e is a fraction.
The rate of escalation multiplier (1 + e) to discount multiplier (1 + r) is referred to
as escalation ratio, d and given by

1þe
d¼
1þr

13.4 Formulations for Evaluation of Money Value Over

Time Span

13.4.1 Notations

Present amount, P: It is either money spent or received now, or the value now of
future sums.
Future amount, F: It is some amount in future, N, periods from now.
The series of equal amount, R: It is a constant stream of equal amounts received at
the end of each of N periods.

13.4.2 Period

It is the fixed interval of time for which the discount rate is defined. It is typically but
not necessarily a year.
146 13 Solar PV System Economics

1. The Compound Amount Formula

Future value F a given amount P at present after N periods is given by

F ¼ P ð1 þ r ÞN ðr ¼ 0:1 for 10%Þ

Compound amount factor ¼ (1 + r)N

2. The Present Value Formula:

It is present value P of a future sum at N years. It is inverse of F.

P ¼ F ð1 þ r ÞN
Present value factor ¼ ð1 þ r ÞN

3. Capital Recovery Formula

It defines the stream of N constant payments that are equivalent to a present
sum, P.
The value of future sum that is equal to the series is given by
 
X
N
i R ð1 þ r ÞN  1 " --------- " -------- "
F¼ Rð1 þ r Þ ¼ "
i
ð1 þ r Þ  1 1 i N

The equivalent present value is obtained by dividing by (1 + r)N

F
As P¼
ð1 þ r ÞN

So after rearranging we have

Pr ð1 þ r ÞN
R¼ 
ð1 þ r ÞN  1

Capital recovery factor, c r f, is given by

r ð1 þ r ÞN
R ¼ P ðc r f Þ and c r f ¼  
ð1 þ r ÞN  1

If (1 + r)N > > 1 as N is large c r f ! r

4. Series Present Value Formula

13.5 Cost-Benefit Analysis in Engineering Systems/Projects 147

1
P ¼ R ðc r f Þ1 Series present value factor ¼
c rf

13.5 Cost-Benefit Analysis in Engineering Systems/Projects

The cost-benefit analysis in engineering systems is used for:

1. Identification of worthwhile projects
2. Ranking of projects.
The methodology involves the use of an index of merit based on accumulated
costs and benefits. In practice there are many indices and each one is used in
accordance with the nature of problem. These indices may be classified into two
categories.

13.5.1 Category I

It is based on the variety of ways of using discounted cash flow of benefits and costs
associated with any project. These are the net present value, the benefit-cost ratio and
internal rate of return.

13.5.1.1 Net Present Value (NPV)

The NPV of a project is simply the difference between the discounted benefits B and
cost C associated with the project:

NPV ¼ B  C

Advantages
The most important advantage of NPV criteria is that it focuses attention on
quantity of money. Such an advantage is not available with methods which rank
projects by ratios.
Disadvantages
1. At the first glance NPV suggests profit. Let us see whether it really implies profit.
In ordinary sense profit is difference between what we receive and what we pay
out. Consider an example: an investment now for a lump sum of revenue later. So
Profit ¼ Money received  Money invested.
Profit ¼ Money received  Money invested (1 + i)N
148 13 Solar PV System Economics

where i is interest rate.

Present value of profit ¼ (Money received) (1 + r)N – (Money invested) (1 + r)N
(1 + r)–N
So Net Present Value is not equal to present value of profit, because discount
2. Secondly, it gives no indication of the scale of effort required to achieve the
result. For example see the table below where NPV criterion marks the scale of
effort required:

Project Benefits (Rs.) Cost (Rs.) Net value (Rs.) NPV % of cost
S 2002,000 2000,000 2000 0.1
T 2000 1000 1000 100

So, net present value method by itself is not a good criterion for ranking projects.
The essential conditions for NPV methods are:
We have a fixed budget to invest.
Projects require the same investment.

13.5.1.2 Benefit-Cost (Cost-Benefit Ratio)

Benefit–cost or cost-benefit ratio is the quotient of monetary value of benefits of a

project divided by its costs.

Benefit B
¼
Cost C
Advantages
• It compares projects on a common scale.
• It directly provides an indication of whether a project is worthwhile.
• It provides an easy means to rank projects in order of relative merit.
The major analytic weakness of the benefit-cost ratio lies with the ambiguity of
the treatment of recurring costs and consequent bias in favour of capital intensive
projects. This is illustrated in Table 13.1.
Benefit/Cost 1.34 for K-project is better than 1.17 for R-project
ratio:
Annual return: 15% for K-project is worse than 20% for R-project
NPV: Rs. 500,000 for K-project is worse than Rs. 1000,000 for
R-project
13.5 Cost-Benefit Analysis in Engineering Systems/Projects 149

Table 13.1 A comparison of Project K-project R-project

capital intensive and
Investment (Ck) Rs. 1000,000 Rs. 1000,000
operations projects (All costs
and benefits at present value). Annual cost (Cr) Rs. 50,000 Rs. 500,000
Annual benefits Rs. 200,000 Rs. 700,000
Annual net return Rs. 150,000 Rs. 200,000
Usual life 10 years 10 years
Total benefits Rs. 2000,000 Rs. 7000,000
Total cost (Ck + Cr) Rs. 1500,000 Rs. 6000,000

Fig. 13.1 NPV as a function of discount rate.

13.5.1.3 Internal Rate of Return (IRR)

It is the discount rate for which the NPV is equal to zero and the project should be
thus ranked from the highest to lowest internal rate of return.
Advantage: No necessity to find the discount rate. Instead one can plot NPV as a
function of discount rate (Fig. 13.1) and determine IRR.

13.5.2 Category II

It has been devised to circumvent the problems in category I. It includes cost

effectiveness ratio payback period.
150 13 Solar PV System Economics

Payback period may be defined as the number of years required for the capital to be
recovered from the net cash flow of benefits. For example, in solar PV systems saving of
conventional energy accrue from the use of solar energy in solar PV system.

13.5.2.1 Mathematical Formulation of Cost-Benefit Analysis

Let us consider a system involving an initial capital I which saves energy S. The
present cost of this energy be C0 per unit of time and let this cost of energy escalate
e per unit of time; then let the initial capital investment be I.Then at the end of N units
of time
The future value of the investment ¼ I (1 + r)N
As energy saved ¼ S
Present cost of energy ¼ C0
At any period Ci ¼ C0 (1 + e)i
where e is escalation rate of energy cost.
Accumulated benefits (B) at the end of the payback period, N is given by

X
N
B¼ S Ci ð1 þ r ÞNi
i¼1

For N to be payback period of the system we have

N 
X 
1þe i
I ð1 þ r ÞN ¼ C 0 S ð1 þ r ÞN
i¼1
1þr

So,

XN  
1 1þe i
¼
C0 S i¼1
1þr

or,
   N
1þe
1þr
1þe
1þr 1
I
¼ ð13:1Þ
C0 S 1þe
1þr 1

If we put
re
¼W
1þe
or
13.5 Cost-Benefit Analysis in Engineering Systems/Projects 151

rþ1
1þW ¼
1þe
∴
 N
1
1þW
1
1þW 1
1
¼
C0 S 1
1þW 1

∴
"  N #
I 1 1
¼ 1
C0 S W 1þW
   N
WI
1 ¼ 1þW
1
C0 S

or,
 
WI
Nlogð1 þ W Þ ¼ log 1 
C0 S

or,


log 1  ROI
W
N¼
logð1 þ W Þ

where

C0 S
ROI ¼
I
As C0S is return on investment in first unit of time and I is initial investment.
Alternatively, the Eq. (13.1) can be expressed as
  N
1þe
1þr
1þe
1þr 1
1
¼
C0 S 1þe
1þr 1

Let

1þe
d¼
1þr
152 13 Solar PV System Economics

 
I d ðd ÞN  1
¼
C0 S d1
or,

I d1
¼ ðd ÞN  1
C0 S d
or,

I d1
þ 1 ¼ ðd ÞN
C0 S d

or,
  
I d1
Nlogd ¼ log þ1
C0 S d

or,
hn 
o i
log I
C0 S
d1
d þ1
N¼
logd

In practice the life cycle costing is often used for determining the most cost-
effective energy solution. In the life cycle cost, there are various levels of sophisti-
cation and accounting ranging from payback period to net present cost for the
project. There are several assumptions such as future values of discount rate and
the energy price escalation as well as several formats for expressing the figure of
merit for decision making. In what follows we present some illustrative numerical
examples.

13.5.2.2 Problem

A roof top integrated solar PV system 100 kWp capacity is installed at Delhi
characterized by solar insolation of 4.2 kWh per day. The power generated is
injected into Low Tension (LT) electricity network. The power cost obtainable
from the grid owner is Rs. 8.00 with 10% escalation every year. The initial cost is
Rs. 70 lacs. Calculate the payback period of the system.
Using the expression for payback period, N


log 1  ROI
W
N¼
logð1 þ W Þ

Here,
13.5 Cost-Benefit Analysis in Engineering Systems/Projects 153

C0 S
ROI ¼
I
From the given problem

I ¼ 70 lacs
C0 S ¼ 4:2  100  360  8 ¼ 1209, 600
1209, 600
ROI ¼ ¼ 0:1728
7000, 000
W ¼ 0:1
W 0:1
1 ¼1 ¼ 0:421
ROI 0:1728
logð0:421Þ ¼ 0:375
logð1 þ W Þ ¼ logð1:1Þ ¼ 0:0413

Hence,

0:375
N¼ ¼ 9:059
0:0413
The payback period of the system is 9.06 years.

13.5.2.3 Alternate Expression for Payback Period, N Calculation

hn 
o i
log I
C0 S
d1
d þ1
N¼
logd

Here,

1þe
d¼
1þr
1þ0 1
d¼ ¼
1 þ 0:1 1:1
or,

d1
¼ 0:1
d
I ¼ 70 lacs
C0 S ¼ 4:2  100  360  8 ¼ 1209, 600
 
  
log 7000000 ð0:1Þ þ 1 log½0:578 þ 1 0:3754
N¼ 1209600
¼ ¼
1
log1:1 logð0:90909Þ 0:04139

or,

N ¼ 9:069 years
154 13 Solar PV System Economics

Table 13.2 The variation of the payback period, N in years, for different daily insolation at
discount rate of 8%, 10% and 12%
Insolation/day r ¼ 8% r ¼ 10% r ¼ 12%
4.2 kWh 7.28 7.99 8.96
4.8 kWh 6.19 6.70 7.36
5.4 kWh 5.39 5.77 6.23

The variation of the payback period, N in years for the daily insolation as
4.2 kWh, 4.8 kWh and 5.4 kWh per day at discount rate of 8%, 10% and 12%
with 2% escalation rate of energy cost is illustrated in Table 13.2. These values are
the characteristics of Indian region.
References

Adams, W.G. and Day, R.E. (1877). The action of light on selenium. Proc. R. Soc. London, ser. A,
25: 113.
Arab, A.H., Chenlo, F. and Benghanem, M. (2004). Loss of load probability of photovoltaic water
pumping system. Solar Energy, 76: 713.
Arthur, D. Little Inc. (1995). Building-Integrated Photovoltaics (BIPV)—Analysis and US Market
Potential. Acorn Park, Cambridge, Massachusetts.
Backus, C.E. (1977). Photovoltaic conversion. In: Solar Energy Engineering. Academic Press,
New York.
Barbaro, S., Coolino, S., Leone, C. and Sinagra, E. (1978). Global solar radiation in Italy. Solar
Energy, 20: 431.
Becquerel, E. (1839). On electric effects under the influence of solar radiation. Comt. Rend., 9: 561.
Bhattacharya, A.B. and Gangadhar, R.B. (1971). Admittance characteristics of a p-n junction diode
with narrow diffused region, Solid-State Electron, 14: 1113–1118.
Bhattacharya, T. (1998). Terrestrial Solar Photovoltaic. Narosa Publishing House, New Delhi.
Billinton, R. and Allan, R.N. (1992). Reliability evaluation of engineering systems: Concepts and
techniques. Plenum Press, New York, USA.
Bird, R.E., Hulstrom, R.L. and Lewis, L.J. (1983). Terrestrial solar spectral data sets. Solar Energy,
30(6): 563–573.
Bishop, J.W. (1988). Computer simulation of the effects of electrical mismatches in photovoltaic
cell interconnection circuits. Solar Cells, 25: 75.
Blakers, A.W. and Green M.A. (1986). 20-percent efficient solar cell. Appl Physics Lett, 48:
215–217.
Blakers, A.W., Wang, A., Milne, A.M., Jhao, J. and Green, M.A. (1989). 22.8-percent efficient
silicon solar cell. Appl Physics Letter, 55(13): 1363–1365.
Boes, E. (1981). Fundamental of solar radiation. In: Solar Energy Handbook. J.F. Kreider and
F. Kreith (eds). McGraw Hill Book Company, New York.
Borowy, B.S. and Salameh, Z.M. (1994). Optimum photovoltaic array size for a hybrid wind/PV
system. IEEE Transaction on Energy Conversion, 9(3): 482.
Borowy, B.S. and Salameh, Z.M. (1996). Methodology for optimally sizing the combination of a
battery bank and a PV array in a wind/PV hybrid system. IEEE Transaction on Energy
Conversion, 11(2): 367.
Brandstetter, A. (1984). Solar cell arrays studies. Proceedings of the Solar World Congress,
Australia, Permagon Press, 4: 1549.
Brian O’ Regan and Gratzel, M. (1991). A low cost high efficiency solar cell based on
dye-sensitized colloidal TiO2 films. Nature, 353: 737–740.

© Capital Publishing Company, New Delhi, India 2018 155

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1
156 References

Bucciarelli, Jr. L.L. (1979). Power loss in photovoltaic array due to mismatch in cell characteristics.
Solar Energy, 23: 277.
Bucciarelli, Jr. L.L. (1984). Estimating loss of power probabilities of stand-alone photovoltaic solar
energy systems. Solar Energy, 32: 205.
Bucciarelli, Jr. L.L. (1986). The effect of day-to-day correction in solar radiation on the probability
of loss of power in standalone photovoltaic solar energy systems. Solar Energy. 36(1): 11.
Casacca, M.A. and Salameh, Z.M. (1992). Determination of Lead-Acid Battery via Mathematical
Modeling Techniques. IEEE Transaction on Energy Conversion, 7(3): 442.
Chamberlin, C.E., Lehman, P., Zoellick, J. and Pauletto, G. (1995). Effect of mismatch losses in
photovoltaic arrays. Solar Energy, 54(5): 165.
Chambouleyron, I. (1991). Light induced meta-stability in a-Ge:H. Journal of Non-Crystal Solids,
128(3): 243–254.
Chaplin, D.M., Fuller, C.S. and Pearson, G.L. (1954). A new silicon p-n junction photocell for
converting solar radiation into electrical power. Journal of Applied Physics, 25: 676–677.
Chedid, R., Akiki, H. and Rahman, S. (1998). A decision support technique for the design of hybrid
solar-wind power systems. IEEE Transaction on Energy Conversion, 13(1): 76.
Chemmangot, V.N., Ashari, M. and Keerthipala, W.W.L. (2000). A Grid-Interactive
Uninterruptible Power Supply System Using Battery Storage and a Back up Diesel Generator.
IEEE Transaction on Energy Conversion, 15(3): 348.
Chenni, R., Makhlour, M., Kerbache, T. and Bouzied, A. (2007). A detailed modeling method for
photovoltaic cell. Energy, 32(9): 1724–1730.
Chowdhary, B.H. (1992). Optimizing the integration of photovoltaic systems with electric utilities.
IEEE Transaction on Energy Conversion, 7(1): 72.
Chowdhary, B.H. and Rahman, S. (1988). Analysis of Interrelationship between photovoltaic
power and battery storage for Electric Utility Load Management. IEEE Transaction on Energy
Conversion, 3(3): 900.
Dorvlo, A.S.S., Jervase, J.A. and Al-Lawati, A. (2002). Solar radiation assessment using artificial
neural network. Applied Energy, 71: 307–319.
Duffie, J.A. and Beckman, W.A. (1980). Solar Engineering Processes. John Wiley and Sons,
New York.
El-Nashar, A.M. (1981). Solar radiation characteristics in Abu Dhabi. Solar Energy, 47(1): 49.
Endo, E., Kurokawa, K. and Hyaashi, M. (1992). Design aid expert system for photovoltaic power
generation system. 2nd Renewable Energy Congress, Reading England, Sept. 13–18.
Enslin, J., Wolf, M., Snyman, D. and Swiegrs, W. (1997). Integrated photovoltaic maximum power
tracking converter. IEEE Trans. Ind. Electron, 44(6): 769.
Esram, T. and Chapman, P.L. (2007). Comparison of Photovoltaic Array Maximum Power Point
Tracking Techniques. IEEE Transaction on Energy Conversion, 22(2): 439–449.
Fahrenbruch, A.L. and Bube, R.H. (1983). Application of the transport equation in fundamentals of
solar cells. Academic Press, New York.
Fam, W.Z. and Balachander, M.K. (1988). Dynamic performance of a dc shunt motor connected to
photovoltaic array. IEEE Transactions on Energy Conversion, 3: 613.
Feldman, J., Singer, S. and Braunstein, A. (1981). Solar cell interconnections and the shadow
problem. Solar Energy, 26: 419–428.
Fraunhofer, I.S.E. (2016). Photovoltaics Report, 17 November 2016 (www.ise.fraunhofer.de).
Francois, G. and Salameh, Z.M. (2001). Steady-State Performance of a Grid-Connected Rooftop
Hybrid Wind-Photovoltaic Power System with Battery Storage. IEEE Transaction on Energy
Conversion, 16(1): 1.
Gangadhar, R.B., Mathur, R.S., Sehgal, H.K. and Kulshreshta, A.P. (1967). Investigations of
silicon p-n junction for solar energy conversion. Indian Journal of Pure and Applied Physics,
5: 593–595.
Gangadhar, R.B. (1997). Investigation of silicon solar cells and their power supply systems. Ph.D.
Thesis (Supervisor N.D. Kaushika).
References 157

Gangadhar, R.B. and Bhattacharya, A.B. (1968). Drift field photovoltaic cell performance with bulk
and surface recombination. International Journal of Electron, 25: 17–26.
Gangadhar, R.B. and Bhattacharya, A.B. (1969). On the transport characteristics of a p-n junction
with narrow diffused region. Solid State Electronics, 12: 887–891.
Gangadhar, R.B. and Kaushika, N.D. (1990). Solar spectral response of graded base silicon cell.
Proc. National Energy Solar Convention, Udaipur, pp. 397–401.
Gangadhar, R.B. and Kaushika, N.D. (1992a). Physics and Technology of Solar Cells. In: Energy
Resources and Technology. N.D. Kaushika, S. Kaneff and C.S. Jha (eds). Scientific Publishers,
Jodhpur.
Gangadhar, R.B. and Kaushika, N.D. (1992b). Spectral response of graded base solar cells. Energy
Conversion Management, 33(4): 275–278.
Gao, F., Wang, Y., Zhang, J. Shi, D., Wang, M., Humphry-baker, R., Wang, P. and Zakeeruddin,
M. Gratezel (2008). A new heteroleptic ruthenium sensitizer enhances the absorptivity of
mesoporous titania film for a high efficiency dye-sensitized solar cell. Chemical Communica-
tions, 23: 2635–2637.
Gautam, N.K. and Kaushika, N.D. (2001). Network analysis of fault-tolerant solar photovoltaic
arrays. Solar Energy Material and Solar Cell, 69: 25.
Gautam, N.K. and Kaushika, N.D. (2002a). An efficient algorithm to simulate the electrical
performance of solar photovoltaic arrays. Energy, 27: 347.
Gautam, N.K. and Kaushika, N.D. (2002b). Reliability evaluation of solar photovoltaic arrays.
Solar Energy, 72(2): 129.
Gavanidou, E.S. and Bakirtzis, A.G. (1992). Design of stand-alone system with renewable energy
sources using trade-off methods. IEEE Transaction on Energy Conversion, 7(1): 42.
Glasnovic, Z. and Margeta, J. (2006). A model for optimum sizing of photovoltaic irrigation water
pumping system. Solar Energy, 81(7): 904–916.
Goh, T.N. (1979). Statistical study of solar radiation in formation in an equatorial region (Singa-
pore). Solar Energy, 22, 105.
Gonzalez, C. and Weaver, R. (1980). Circuit design considerations for photovoltaic modules and
systems. Proceedings of the 14th IEEE Photovoltaic Specialists Conferences, San Diego,
New York, USA.
Gover, A. and Stella, P. (1974). Vertical multi-junction solar-cell one-dimensional analysis. IEEE
Transaction on Electron Devices, ED-21: 351.
Gow, J. and Manning, C. (2000). Controller management for boost converter systems source from
solar photovoltaic arrays or other maximum power sources. IEE Proc. Electronics Power Appl.,
147(1): 15.
Green, M.A. (1979). Solar cells—Operating principle, technology and system applications. A short
course on solar photovoltaic systems at University of Queensland, Aug. 21–23.
Green, M.A., Gauja, E. and Withayachamnankul, W. (1981). Silicon solar cells with integral bypass
diode. Solar Cells, 3: 233–244.
Green, M.A. (1982). Solar cells. Prentice-Hall Inc., Englewood Cliffs, NJ.
Green, M.A. (1984), Limits on the open circuit voltage and efficiency of silicon solar cells imposed
by intrinsic auger processes. IEEE Trans Electron Devices, ED-31: 671–678.
Green, M.A. (1987). High efficiency silicon solar cells. Aedermaunsdorf: Trans. Tech.
Publications.
Green, M.A., Blaker, A.W., Zhao, J., Milne, A.M., Wang, A. and Dai, X. (1990). Characterization
of 23 percent efficient silicon solar cells. IEEE Trans. on Electron Devices, 37(2): 331–335.
Green, M.A. et al. (2011). Solar Cell Efficiency Tables (Version 37). Progress in Photovoltaics:
Research and Applications, Vol. 19. John Wiley & Sons Ltd., N.J.
Grondahl, L.O. (1933). The copper-cuprous oxide rectifier and electric cell. Rev. Mod Physics, 5:
141.
Gunes, Serap., Neugebaucer, H. and Sariciftci, N.S. (2007). Conjugated polymer based organic
solar cells. Chem. Rev., 107: 1324–1338.
158 References

Halls, J.J.M., Friend, R.H., Archer, M.D. and Hill, R.D. (2001). Clean electricity from photovol-
taics. Imperial College Press, London.
Hilloowala, R.M. and Sharaf, A.M. (1992). A Rule-Based Fuzzy Logic Controller for a PWM
Inverter in Photovoltaic Energy Conversion Scheme. Proc. IEEE IAS Conf., Houston.
Hontoria, L., Anguilera, J. and Zufiria, P. (2005). A new approach for sizing stand alone photo-
voltaic systems based in neural network. Solar Energy, 78(2): 313.
Hoppe, H. and Sariciftci, N.S. (2004). Organic solar cells: An overview. J. Mater. Res., 19:
1924–1945.
Hottel, H.C. (1976). A simple model for estimating the transmittance of direct solar radiation
through clear atmosphere. Solar Energy, 18: 129.
Hussein, H.M.S., Ahmad, G.E. and El-Ghetany, H.H. (2004). Performance evaluation of photo-
voltaic modules at different tilt angles and orientation. Energy Conversion and Management,
45: 2441–2452.
Jackson, E. (1955). Areas for improvement of the semiconductor solar energy converter. In:
Transactions of the conference on the use of solar energy, Vol. 5.
Jain, G.C. and Al-Rifai, R.M.S. (1967). Effects of Drift Fields and Field Gradients on the Quantum
Efficiency of Photocells. Journal of Applied Physics, 38: 768.
Jain, S.C., Sehgal, H.K. and Gangadhar, R.B. (1968). Silicon Solar Cells and Their Terrestrial
Applicatins. Journal of Scientific & Industrial Research, 27(8): 295–305.
Jordan, A.G. and Milnes, A.G. (1960). Photoeffect on diffused P-N junctions with integral field
gradients. I R E Transaction on Electron Devices, ED-7 (4): 242–251.
Kalogivou, S.A. and Tripanagnostopoulos, Y. (2006). Hybrid PV/T systems for domestic hot water
and electricity production. Energy Conversion and Management, 47(18–19): 3368.
Kaneff, S. (1983). On the practical and potential viability of paraboloidal dish solar thermal electric
power systems. Proc. Solar World Congress (ISES) Perth, 4, p. 2767.
Karatepe, E., Mutlu, B. and Colak, M. (2006). Neural network based solar cell model. Energy
Conversion and Management, 47(9–10): 1159.
Karatepe, E., Mutlu, B. and Colak, M. (2007). Development of suitable model for characterizing
photovoltaic arrays with shaded solar cell. Solar Energy, 81: 977–992.
Kaushika, N.D. (1986). Solar parabolic dish power production for rural electrification. Invention
Intelligence, May-June, 182.
Kaushika, N.D. (1996). 50 kW solar PV power plant in Rajasthan TEFR. Sun Source of India Ltd.
Kaushika, N.D. (1999). Design and development of fault tolerant circuitry to improve the reliability
of solar PV modules and arrays. Final Technical report of the Department of Science and
Technology, the Government of India Research project no. III 5 (98) / 95-ET.
Kaushika, N.D. and Gautam, N.K. (2001). Mismatch losses and time to failure of solar PV arrays.
Proceedings International Solar Energy Society Meeting, Adelaide, Australia.
Kaushika, N.D. (2002). Solar PV design aid expert system. In: Final Technical report of the
Department of Science and Technology, the Government of India Research project no. III
5 (81) / 99-ET.
Kaushika, N.D. and Gautam, N.K. (2003). Energy Yield Simulations of Interconnected solar PV
Arrays. IEEE Transaction on Energy Conversion, 18(1): 127.
Kaushika, N.D., Gautam, N.K. and Kaushik, K. (2005). Simulation model for sizing of standalone
solar PV system with interconnected array. Solar Energy Materials and Solar Cell, 85(4): 499.
Kaushika, N.D. and Rai, Anil K. (2006). Solar PV design aid expert system. Solar Energy Materials
and Solar Cells, 90: 2829–2845.
Kaushika, N.D. and Rai, Anil K. (2007). An investigation of mismatch losses in solar Photovoltaic
cell networks. Energy, 32: 755–759.
Kaushika, N.D., Tomar, R.K. and Kaushik, S.C. (2014). Artificial neural network model based on
interrelationship of direct, diffuse and global solar radiations. Solar Energy, 103: 327–342.
Kazmerski, L.L. and Emery, K.A. (1992). Proc. 2nd World Renewable Energy Congress, Reading, U.K.
Kellogg, W.D., Nehrir, M.H., Venkataramanan, G. and Gerez, V. (1998). Generation unit sizing
and cost analysis for stand-alone wind, photovoltaic and hybrid wind/PV systems. IEEE
Transaction on Energy Conversion, 13(1): 70.
References 159

Khouzem, K. and Khouzam, L. (1993). Optimum matching of Direct coupled Electromechanical

Loads to a photovoltaic generator. IEEE Transaction on Energy Conversion, 8(3): 343.
Kislovski, A. and Redl, R. (1994). Maximum power tracking using positive feedback. Proc. IEEE
Power Electronics Spec. Conf. 1065.
Klein, S.A. and Beckman, W.A. (1987). Loss-of-load probabilities for standalone photovoltaic
systems. Solar Energy, 39(6): 499.
Konadu, R.A. and Mingyua, C. (1999). An Expert System tool for designing hybrid energy
systems. Power and Energy Systems, 19(3): 239.
Koutroulis, E., Kalaitzakis, K. and Voulgaris, N.C. (2001). Development of a micro controller
based photovoltaic maximum power point tracking control system. IEEE Transaction Power
Electron, 16(21): 46–54.
Koutroulis, E., Kolokotsa, D., Potivakis, A. and Kalaitzakis, K. (2006). Methodology for optimal
sizing of stand-alone photovoltaic/wind generation system using genetic algorithm. Solar
Energy, 80(9): 1072.
Landsman, E.E. (1978). Maximum power trackers for photovoltaic array. 13th IEEE PVSC,
Washington DC, June 5–8.
Lange, B. (1930). New photoelectric cell. Zeit Physics, 31: 139.
Lee, Tsung-Ying and Chen, N. (1993). Optimal capacity of the battery energy storage system in a
power system. IEEE Transaction on Energy Conversion, 8(4): 667.
Liu, B.Y.H. and Jordan, R.C. (1960). The Interrelationship and Characteristic Distribution of
Direct, Diffuse and Total Solar Radiation. Solar Energy, 4: 1–19.
Liu, B.Y.H. and Jordan, R.C. (1962). Daily Insolation on Surfaces Tilted towards the Equator.
ASHRAE Transactions, 67: 526–541.
Lokhande, C.D. and Dhawla, D.S. (2009). Chemically deposited thin film for solar cells. Proc.
National Conference on Renewable Energy (NACORE-2009), Jodhpur, India.
Mack, M.R. (1979). Solar Power for Telecommunication. Telecommunication Journal of Australia,
29(1): 120.
Mani, A. (1980). Handbook of Solar Radiation Data for India. Allied Publishers, India.
Marwali, M.K.C., Shahidetipour, S.M. and Daneshdoot, M. (1997). Probabilistics-utility systems
with battery storage. IEEE Transaction on Energy Conversion, 12(2): 175.
McDaniels, D.K. (1979). The Sun: Our future energy source. John Wiley and Son Inc., New York.
Mikroyannidis, John A., Stylianakis, Minas M., Suresh, P. and Sharma, G.D. (2009). Efficient
hybrid bulk heterojunction solar cells based on phenylenevinylene copolymer, perylene
bisimide and TiO2. Solar Energy Material & Solar Cells, 93: 1792–1800.
Milstein, J.B. and Tsno, Y.S. (1984). Research on crystalline silicon solar cells. Proc. 17th IEEE
Photovolt. Special Conf. Orlando, FL.
Mohan, N. (2004). Power Electronics Converters, Application and Design. Tore M. Undeland and
William P. Robbins (eds). John Wiley & Sons (Asia).
Mondol, J.D., Yohanis, Y.G. and Norton, B. (2006). Optimal sizing of array and inverter for grid
connected photovoltaic systems. Solar Energy, 80(12): 1517.
Nozik, A.J. (2001). Spectroscopy and hot election relaxation dynamics in semiconductor quantum
wells and quantum dots. Annual Reviews of Physics and Chemistry, 52, 193.
Ofry, E. and Braunstein, A. (1983). The Loss of Power Supply Probability as a Technique for
Designing Stand-alone Solar Photovoltaic systems. IEEE Transaction on Power Apparatus &
Systems, 102(5): 1171.
Ogeman, H., Ecevit, A. and Tasdemiroglu, E. (1984). A new method for estimating solar radiation
from bright sunshine data. Solar Energy, 19: 307.
Pulfrey, L.D. (1978). Photovoltaics Power Generation. Van Nostrand Reinhold Co., New York.
Quaschning, V. and Hanitsch, R. (1996). Numerical simulation of current-voltage characteristics of
photovoltaic systems with shaded solar cells. Solar Energy, 56: 513.
160 References

Rai, Anil K., Kaushika, N.D., Singh, B. and Agarwal, N. (2011). Simulation model of ANN based
maximum power point tracking controller for solar PV system. Solar Energy Materials & Solar
Cells, 95: 773–778.
Ralph, E.L. (1965). Use of concentrated sunlight with solar cells for terrestrial uses. Presented at the
Solar Energy Society Annual Meeting, Arizona.
Ramkumar, R., Shetty, P.S. and Ashenayi, K. (1986). A Linear Programming Approach to the
Design of Integrated Renewable Energy Systems for Developing Countries. IEEE Transaction
on Energy Conversion, 7 EC-1(4): 18.
Ramkumar, R., Abouzahr, I. and Ashenayi, K. (1992). A knowledge-based approach to the design
of Integrated Renewable Energy Systems. IEEE Transaction on Energy Conversion, 7(4): 648.
Raushenbach, S.H. (1971). Electrical output of shadowed solar cell arrays. IEEE Transaction on
Electron Devices, 18: 483.
Ravich, Yu. L. (1962). Radiotekh. Elektron, 7: 1002.
Ray, S. (2009). Nanocrystalline silicon based thin film solar cell. Proc National Conference on
Renewable Energy (NACORE-2009), Jodhpur, India.
Reddy, S.J. (1971). An empirical method for estimation of the total solar radiation. Solar Energy,
14: 289.
Reynolds, D.C., Leies, G., Antes, L.L. and Marburger, R.E. (1954). Photoelectric effect in cadmium
sulphide. Physics Rev., 96: 533–534.
Riad, C. and Rahman, S. (1997). Unit sizing and control of hybrid wind-solar power systems. IEEE
Transaction on Energy Conversion, 12(1): 79.
Robbins, T.P. (1987). Simulation of a Stand-alone Power System with Battery Storage. Annual
Conference and Annual General Meeting of Anzses, Australian National University, 27.
Ross, R.G. (1981). Photovoltaic module and array reliability. Proc. of 15th IEEE Photovoltaic
Specialists Conf., Orlando, FL, New York, USA.
Runyan, W.R. (1965). Silicon Semiconductor technology. McGraw Hill, NewYork.
Runyan, W.R. and Alexander, E.G. (1967). IEEE Trans. Electron Devices, 14: 3.
Salameh, Z. and Taylor, D. (1990). Step-up maximum power point tracker for photovoltaic arrays.
Solar Energy, 44: 57–61.
Sabbagh, J.A., Sayigh, A.A.M. and El-Salam, E.M.A. (1977). Estimation of the total solar radiation
from meteorological data. Solar Energy, 68(2): 169–178.
Salameh, Z.M., Borowy, B.S. and Amin, A.R.A. (1995). Photovoltaic module-site matching based
on the capacity factors. IEEE Transaction on Energy Conversion, 10(2): 326.
Sayed, M. and Partin, L. (1975). Effect of shading on CdS/CuxS solar cell array design. IEEE Trans
Energy Conversion, 14: 61.
Schottky, W. (1930). Cuprous oxide photoelectric cells. Zeit Physics, 31: 913.
Sfetsos, A. and Coonick, A.H. 2000. Univariate and multivariate forecasting of hourly solar
radiation with artificial intelligence techniques. Solar Energy, 68(2): 169–178.
Shah, A.V., Meir, J., Vallat-Sauvair, E., Wyrsch, N., Kroll, U., Droz, C. and Graf, U. (2003).
Material and solar cell research in microcrystalline silicon. Solar Energy Material & Solar Cells,
78: 469–491.
Sharma, G.D., Suresh, P., Mikroyannidis, John A. and Stylianakis, Minas M. (2010). Organic PV
devices based on conjugate urinyle copolymer: PCBM bled with incorporation of low band gas
small molicunk. J. Mater. Chem, 20: 561–567.
Sharma, G.D. (2009). Advances in Nano Structured Organic Solar Cells. Proc National Conference
on Renewable Energy (NACORE-2009), Jodhpur, India.
Shephard, N.F. and Sugimura, R.S. (1984). The Integration of bypass diode with Terrestrial
Photovoltaic Modules and Arrays. Proc. of the 17th IEEE Photovoltaic Specialists Conf.
Shrestha, G.B. and Goel, L. (1998). A study on optimal sizing of stand-alone photovoltaic stations.
IEEE Transaction on Energy Conversion, 13(4): 373.
Singer, S., Rosenshtein, B. and Surazi, S. (1984). Characterization of PV array output using a small
number of measured parameter. Solar Energy, 32: 603.
References 161

Siva Kumar, R. and Anandan, S. (2009). New type of inorganic-organic Hybrid (Heteropoly
tungstic acid-Polyepich lorohydrin) Polymer electrolyte with TiO2 nano filler for solid state
dye sensitized solar cells. Proc National Conference on Renewable Energy (NACORE-2009),
Jodhpur, India.
Smith, K.D., Gummel, H.K., Bode, J.D., Cuttriss, D.B., Nielsen, R.J. and Rosenzweig, W. (1963).
The Solar Cells and Their Mounting. Bell System Technical Journal, 42: 1765–1816.
Solar Directory (1989). Prospectus for Renewable Energy in Australia. ANZSES Report.
Stember, L.H. (1981). Reliability considerations in the design of solar photovoltaic power system.
Solar Cells, 3: 269–285.
Sukhatme, S.P. (1984). Solar Energy Principles of Thermal Collection and Storage. Tata McGraw-
Hill, New Delhi.
Supit, I. and Van Kappel, R.R. (1998). A simple method to estimate global radiation. Solar Energy,
63: 147.
Suresh, P., Sharma, G.D., Mikroyannidis, John A. and Stylianakis, Minas M. (2009). Organic
Photovoltaic devices based on conjugated rinyle copolymer: PCBM blend with incorporation of
low band gap small molecule. Proc National Conference on Renewable Energy (NACORE-
2009), Jodhpur, India.
Swalesh, M.S. and Green, M.A. (1982). Effect of shunt resistance and bypass diode on the shadow
tolerance of solar cell modules. Solar Cells, 5: 183.
Theraja, B.L. and Theraja, A.K. (1988). A text-book of electrical technology. Nirja Construction &
Development Co. (P) Ltd, New Delhi.
Tiedje, T., Vablonowitch, E., Cody, G.D. and Brooks, B.G. (1984). Limiting efficiency of silicon
solar cells. IEEE Trans. on Electron Devices, ED-31: 711–761.
Treble, F.C. (1991). Generating electricity from Sun. Pergamon Press, New York.
Veerachary, M., Senjyu, T. and Uezeto, K. (2002a). Feed forward maximum power point tracking
of PV systems using fuzzy controller. IEEE Transactions on Aerospace and Electronic System,
38(3): 969.
Veerachary, M., Senjyu, T. and Uezeto, K. (2002b). Voltage based maximum power point tracking
of PV systems. IEEE Transactions on Aerospace and Electronic System, 38(1): 262.
Wang, A., Zhao, J. and Green, M.A. (1990). 24-percent efficient silicon solar cells. Appl. Phys.
Lett., 57: 602–604.
Wasynczuk, O. (1989). Modeling and Dynamic Performance of a Line-Commutated Photovoltaic
Inverter System. IEEE Transaction on E.C., 4(3): 37.
Winter, C.J., Sizmann, L.L. and Vant-Hull (1991). Solar Power Plants. Springer-Verlag Publishers.
Wolf, H.F. (1969). Silicon semiconductor data. Pergamon Press, New York.
Wolf, M. (1960). Limitations and possibilities for improvement of photovoltaic energy converters,
Part I: Considerations for Earth’s surface operation. Proc. Inst Radio Engg, 48: 1246–1263.
Wolf, M. (1963). Drift fields in photovoltaic solar energy converter cell. Proceedings IEEE, 5: 674.
Wolf, M. (1971). A new look at silicon solar cell performance. Energy Conversion, 11: 63–73.
Wolf, M. and Ralph, E.L. (1965). Effects of thickness on short-circuit current of silicon solar cells.
IEEE Transactions on Electron Devices, 12(8): 470.
Wolf, S. and Enslin, J. (1993). Economical PV maximum power point tracking regulator with
simplistic controller. Proc. IEEE Power Electronics Spec. Conf. 581.
Yella, A., Lee, H.W., Tsao, H.N., Yi, C., Chandiran, A.K., Nazeeruddin, M.K., Diau, EW-D., Yeh,
C.Y., Zakeeruddin, S.M. and Gratzel, M. (2011). Porphyrin-sensitized solar cells with cobalt
(II/III)-based redox electrolyte exceed 12 percent efficiency. Science, 6056: 629–634.
Yokoyama, R., Ito, K. and Yuasa, Y. (1994). Multiobjective optimal unit sizing of hybrid power
generation systems utilizing photovoltaic and wind energy. International Journal of Solar
Energy Engineering, 116: 167.
Index

A Continuity equation, 61–62

Agro waste and residues constant field method/single exponential
conversion into useful form of fuels/energy, approximation method, 63
5 solution of, 63–66
Amorphous silicon (a-Si) thin-film modules, 81 Conventional power plants engineering design,
Array network, mismatch losses in, 83–84 118
Average solar radiation, estimation of, 23–26 Crystalline silicon
atmospheric clearness index, 25 cells, categories of, 10
beam radiation falling on a horizontal modules, 81
surface, 25 solar cells, 10
solar beam radiation on earth under clear Crystalline solar cell, performance of, 77–79
sky condition, 24 I-V characteristics of, 79
solar radiation data, 24 I-V characteristics, measurement of, 77

B D
Balance of System (BOS), 93–103 Deep dish, 9
charge controller, 93
classification of, 95–102
maximum power point tracker (MPPT), 93, E
94 Electrical BOS, 95–96
SPV power systems, 94 captive-electrolyte, or gel-cell, batteries, 95
standalone PV power supply, 94 deep-cycle batteries, 95
utility-interactive PV systems, 94 disconnect switches, 96
Biogas, 5 lightning protection devices, 95
Biomass, 4 liquid-electrolyte, lead-acid storage
Building-integrated PV systems (BIPV), 108 batteries, 95
power systems, 108 National Electrical Code (NEC), Article
projects in India, 111 690 of, 95
over current protection, 96
string combiners, 95
C Electrical energy
Collector, 8 per capita consumption of, 1
Collector technologies, variations in, 8 Electrical generation, 2, 5

© Capital Publishing Company, New Delhi, India 2018 163

N. D. Kaushika et al., Solar Photovoltaics,
https://doi.org/10.1007/978-3-319-72404-1
164 Index

Electricity, 1 L
photovoltaic generation of, 27 Laboratory cell efficiency, 11
Electronic BOS, 96–102 Laboratory solar cells, conversion efficiency of,
dc/ac inverters, 99–101 52
dc/dc converters and MPPTs, 96–99 Large scale solar PV application, 106
energy control and management subsystem,
101, 102
energy conversion subsystem, 96–101 M
pulse-width modulation (PWM), 100 Mechanical BOS, 95
waveform synthesis, 101 Mismatch loss, 83
Engineering systems/projects, cost-benefit computational results, 89–91
analysis in, 147–154 Multicrystalline and EFG ribbon silicon cells,
benefit-cost (cost-benefit ratio), 148, 149 11
category I, 147–149 Municipal and industrial wastes, 4
category II, 149–154
internal rate of return (IRR), 149
net present value (NPV), 147, 148 N
Extrinsic or doped semiconductors, 34 Nano-architectured solar cells, 11
Next-generation photovoltaics, 11
N-type extrinsic semiconductor, 35
F N type semiconductor
Fan-mill, 7 junction region free electrons, 44
First-generation PV systems (fully
commercial), 139
First generation solar cells, 139–140 O
First silicon solar cell, solar conversion Ocean gradient, 5, 6
efficiency of, 105 mechanical energy from the tides and
Fuel gas, 5 waves, 6
ocean thermal energy conversion
technology, 6
G thermal energy from ocean surface warm
Gasification, 5 water, 6
Gasified-based power units, 5 tide induced currents, 6
Global cumulative solar PV installations, 12 Oceans, 5
Global SPV module, 1989, 107 OTEC turbo generators, 6
Grid-connected or utility-interactive
(UI) system, 108
P
Parabolic dish systems, 9
H Parabolic trough, 8
High performance solar cells, 139–142 Parabolic trough solar technology, 9
Hydro power, 4 Passivated emitter and rear cell (PERC), 52
Hydro power generation, 2 Passivated emitter rear locally (PERL) diffused
Hydro power project, 4 cell, 53
Passivated emitter solar cell (PESC), 52
Photovoltaic generation, fundamentals of, 27
I atomic bonding, 27
India conduction band, 30
mono-crystalline silicon solar cell, 53 covalent bonding, 28
solar cell fabrication, 53 crystalline solid, 29
Intrinsic semiconductor, 33, 34 crystal structure, 29
Index 165

Fermi level, 31 Short circuit current, 80

forbidden energy gap, 30 Silicon
ionic bonding, 28 properties of, 32, 33
metallic bonding, 28 as a semiconductor material, 32
metals (conductors), valence-band energy, Silicon solar cell, I-V characteristics, 79
31 Silicon solar cells, fabrication of, 49–50
non-crystalline solid, 29 boron, diffusion of, 50, 52, 53
Pauli’s exclusion principle, 27 minority carrier lifetime, 49
semiconductor material, 32 phosphorus, diffusion of, 50
valence band, 30 silicon slices, 49
van der Waals bonding, 28 Silicon solar cell wafer technology, 47–52
Photovoltaic generators, 12 electrical contacts and encapsulation, 50–52
Photovoltaics as glazing material, 109–111 Haunsa Resist/SS, 51
BIPV, 110 Kodak Photo Resist (KPR), 51
mounting PV panels on buildings, 109 metallurgical to semiconductor grade
Photovoltaic systems for buildings, 108 silicon, 48, 49
Piecewise exponential computational scheme, polycrystalline silicon to single crystal
67–71 silicon wafers, 49
average field values in the diffused region, sand to metallurgical grade silicon, 48
68 Silicon wafer-based PV technology, 11
P-N junction, 37–38 Single crystal silicon homojunction solar cell,
junction or barrier potential, 38 10
P-N junction solar cell, 39–40 Single crystal silicon solar cells, conversion
crystalline silicon, electron-hole generation, efficiency, 53
40 Small hydro power plant, development of, 4
one dimensional continuity equation for Small hydro power projects, 4
electrons in the p-region, 61 Small solar PV power unit of 6 mW, United
Power generations, non-conventional methods States Army Signal Corps (USASC),
of, 3 105
Power supplies Solar cell, 8
environmental concerns, 2 and modules, 10–11
fuel security, 1, 2 photoionic processes in, 43, 44
rising prices, 3 Solar cell current-voltage (I-V) relationship, 76
supply and demand, gap in, 1 Solar cell equivalent circuit I-V characteristics,
Producer gas, 5 75
P-type extrinsic semiconductor, 35 Solar cell materials, 44–47
Pumping wind mills, 7 cell fabrication technologies, 45
Pyranometer, 20 photovoltaic conversion, efficiency of, 46
Pyrheliometer, 21 PV materials with their efficiencies, 44
Solar cell materials and fabrication steps,
research for, 52, 53
R antireflection coatings, 53
Rankine cycle steam turbine, 9 Solar cells and modules, 10
Renewable resources, power generation from, silicon as material used, 10
3–13 silicon PV modules, manufacturing process
of, 10
state and quality of the silicon material, 11
S Solar cells arrays, 83
Schockley-Quetsser efficiency limit, 11 Solar cells, electrical characteristics of, 73–80
Second-generation PV systems, 139 concentration of minority carriers, 73
Second generation solar cells, 140–141 current-voltage characteristics, 75–77
Semiconductor materials equivalent circuit, 75
movement of charge carriers, 35–37 forward bias voltage, 73
Semiconductors: doping, 33–35 temperature and solar irradiance effects, 80
166 Index

Solar cells, physical model of, 55–61 Solar PV power, cost effective applications,
built-in field, 57–61 107–112
excess carrier generation: photo generation, distributed PV systems on buildings, 107
55–57 solar photovoltaics for buildings, 107–111
Solar constant, 15–16 Solar PV power system, reliability, 127
radiant energy flowing out of the sun, 15 active redundancy, 131
solar energy intercepted by the earth (SE), 15 array operation at low voltage, 131
Solar parabolic trough collectors, 9 bridge-linked configuration, 135
Solar photovoltaic effect, 10 bypass diode, 133
Solar photovoltaic power, 10–12 cell characteristics, manufacturer’s
Solar photovoltaic (PV) technology, 10 tolerances in, 130
systems and applications, 12 environmental stresses, 130
Solar photovoltaics, 8 failure modes and effects analysis (FMEA),
direct conversion technology based on, 13 132
indirect method, 8 fault tolerance at component level, 132
Solar photovoltaic (SPV) system design, fault tolerant circuitry, 132, 133
115–126 fault tree model, 132
conventional power plants engineering improvements, 131–137
design, 118 parallel redundancy, 131
grid-interactive PV system, 118 power loss and hot spot effects, 129–131
power conditioning unit (PCU), 115 quality and redundancy, 131
PV-hybrid system, 116 series-paralleling solar cell interconnection
PV-wind hybrid system, 118 scheme, 133
stand-alone solar PV system, 116 shadow problem, 130
stand-alone solar PV system with storage site dependent performance, 128
battery, 116 standby redundancy, 131
Solar power conversion technologies, 8–12 ‘sun shine days’, measure of, 129
Solar power tower concept, field experiments system availability, 128–129
on, 9 system classification, 112–113
Solar PV Solar PV system economics, 143–154
concentrating applications, 112 capital recovery formula, 146
generation, 12 compound amount formula, 146
metal roofing, 111 costs and benefit with comparable amounts
Solar PV array in engineering projects, 144
electronic regulators, 102–103 discount rate, 145
environmental stresses in, 83 future amount, 145
Solar PV design expert system, 121–125 money value over time span, formulations
EUAO, 121–124 for evaluation, 145–147
flow chart, 126 present amount, 145
knowledge base for Indian region, 121–125 present value formula, 146
web access to, 125, 126 series of equal amount, 145
Solar PV module, 81–82 system evaluation, 143–144
SPV systems, 81 Solar PV system performance, 118–121
Solar PV module and array, fractional power Solar PV systems, USA, 142
loss in, 84–ENF Solar radiant energy, 8
fractional power loss in a module, 87–89 Solar radiation
Kirchhoff current law, 84 atmospheric effects on, 16, 17
solar cell model, 84–87 photoelectric effect, 43
Solar PV power Solar radiation geometry, 17–20
for illuminating hamlets in scorching solar declination angle, δ, 18
deserts, 106 solar incidence angle, 17
for remote refrigeration, 106 sun-earth angles, 19
Index 167

Solar radiation on a tilted surface, 21–23 Tetravalent (group IV of Periodic Table)

beam radiation and diffuse radiations on materials, 32
horizontal surface, 22 Thin film cadmium telluride (CdTe) cells, 14
beam radiation of horizontal surface, 22 Thin film solar cells, 11
beam radiation of tilted surface, 23 Third generation PV systems, 139
diffuse radiation, 22 Third generation solar cells, 141–142
total radiation incident on a surface, 21
Solar spectral distribution, 16, 17
Solar thermal electric system, 8 U
thermodynamic process, 8 USSR solar PV units, 105
Solar thermal power systems, 8
Space solar PV applications, 106
SPV power plant system, 12 V
Sun, 15 Vanguard-I satellite, 105
Sunshine recorder, 21
Surface receiver, 9
W
Wind energy utilization, 7
T Wind farms, 7
Tata BP Solar India, 106, 142 Wind generators, 7
Terrestrial solar PV Wind machines, 7
applications, 106–107 Wind power, 7
system grouping, 113 Wind power system, capital costs of, 7