See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/303488997

MATLAB/PSpice Hybrid Simulation Modeling of a Photovoltaic System and Its

Applications

Thesis · January 2012

CITATIONS READS

2 504

4 authors, including:

Soliman A. Mahmoud Maitham Alsari

University of Sharjah Ministry of Education, Bahrain
224 PUBLICATIONS   2,325 CITATIONS    2 PUBLICATIONS   28 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Field Programmable Analog Array View project

EEG Detection System View project

All content following this page was uploaded by Soliman A. Mahmoud on 25 May 2016.

The user has requested enhancement of the downloaded file.

1
1.1 Introduction

The increasing demand for electrical power has created many challenges for the energy
industry, which can play a vital role in the quality of the generated power in both short
and long terms. The limited supply of fossil fuels had grabbed the international
community attention towards the importance of renewable energy. Transforming the
sun’s potential into clean non-polluted, efficient energy has become an alternative of
many power sources creating a revolution in the energy industry. Today, solar-generated
electricity serves people living in the most isolated spots on earth as well as in the center
of many big cities. Photovoltaic systems are now generating electricity to pump water,
light up the night, power satellites, activate switches, charge batteries, and supply the
electric utility grid. The large number of various solar powered applications that have a
great impact on peoples’ lives led to a rise in the importance of photovoltaics globally.

1.2 History of the Solar Cell [1]

The term "photovoltaic" comes from the Greek words (phōs) meaning "light", and
"voltaic", from the name of the Italian physicist Volta. The photovoltaic effect was first
reported by the French physicist Edmund Bequerel in 1839 when he observed that the
action of light on a silver coated platinum electrode immersed in electrolyte produced an
electric current. Forty years later the first solid state photovoltaic devices were
constructed investigating the recently discovered photoconductivity of selenium. In 1876
William Adams and Richard Day found that a photocurrent could be created in a sample
of selenium when contacted by two heated platinum contacts. The photovoltaic action of
the selenium differed from its photoconductive action in that a current was produced
spontaneously by the exposure to light meaning that no external power supply was

1
needed. In this early photovoltaic device, a rectifying junction had been formed between
the semiconducting material and the metal contact. In 1894, Charles Fritts prepared what
was probably considered as the first large area solar cell by pressing a layer of selenium
between gold and another metal. This device was only around 1% efficient. In the
following years photovoltaic effects were observed in structures like copper-copper oxide
thin film, lead sulphide and thallium sulphide. These cells were thin film Schottky barrier
devices, where a semitransparent layer of metal deposited on top of the semiconducting
material provided both the asymmetric electronic junction, which is crucial for
photovoltaic action and access to the junction for the incident light. The photovoltaic
effect of these structures was associated with the existence of a barrier to current flow at
one of the semiconductor-metal interfaces (i.e., rectifying action) by Goldman and
Brodsky in 1914. Later, during the 1930s, Walter Schottky, Neville Mott and others
developed the theory of metal-semiconductor barrier layers. However, researchers were
interested in the photoconductivity of materials like selenium and not their photovoltaic
properties. The fact that the current produced was proportional to the intensity of light
and related to the wavelength in a definite way meant that photoconductive materials
were ideal for photographic light meters operating with no power supply. The
development of good quality silicon wafers for application in the latest solid state
electronics, that potentially useful quantities of power were created by photovoltaic
devices in crystalline silicon. In the 1950s, the growth of silicon electronics followed the
discovery of a technique to manufacture p-n junctions in silicon. These p-n junction
structures produced much improved rectifying action and better photovoltaic behavior
than Schottky barriers. The first silicon photovoltaic cell was reported by Chapin, Fuller
and Pearson in 1954 and converted sunlight with an efficiency of 6%, six times greater
than the best previous attempt. These cells were not considered for power generation for
several decades. Nevertheless, the first silicon solar cell did introduce the possibility of
power generation in rural areas where fuel could not easily be delivered. The apparent
application was to satellites where the condition of reliability and low weight made the
cost of the cells unimportant. Throughout the 1950s and 60s, silicon solar cells were
broadly developed for applications in space. In 1954, a cadmium sulphide p-n junction
was produced with efficiency of 6%, and in the subsequent years studies of p-n junction

2
photovoltaic devices in gallium arsenide, indium phosphide and cadmium telluride were
enthused by theoretical work indicating that these materials would offer a higher
efficiency. However, silicon remained and remains the leading photovoltaic material. In
the 1970s the crisis in energy supply led to an abrupt growth of interest in alternative
sources of energy, and support for research and development in those areas. Photovoltaics
was a subject of intense interest during this period, searching for strategies to improve the
performance of solar cells and lower their cost. Routes to lower cost included photo
electrochemical junctions, and several alternative materials including polycrystalline
silicon, amorphous silicon, other `thin film' materials and organic conductors. Routes to
higher efficiency included tandem and further multiple band gap designs. During the
1990s, interest in photovoltaics grew, along with the increasing awareness of the need to
secure sources of electricity alternative to fossil fuels. In the late 1990s the photovoltaic
production expanded at a rate of 15-25% per annum, leading to a reduction in cost.

1.3 Objective

The objective of this project is to learn the basic principles of a photovoltaic system and
carry out a hybrid simulation study of a PV cell/module and system using
MATLAB/Simulink and PSpice simulating both the I-V and P-V characteristic curves.
The project will also be concerned with the modeling of a PV module using MATLAB
based on three models, the single-diode model, the two-diode model, and the three-diode
model. In addition, the modeling of a PV grid-connected system is carried out.

1.4 Thesis Organization

This project will discuss the modeling of a photovoltaic module using both PSpice and
Simulink. The report is divided into six main chapters. Chapter 2 briefly discusses the
basic principles of photovoltaics. Chapter 3 presents an overall study of the basic
parameters that affect the performance of the solar cell model using PSpice and Simulink.
Chapter 4 studies the basic performance of the solar cell modules and arrays using PSpice
and Simulink. In chapter 5, a modeling method is investigated that finds the parameters

3
that are not available in the standard datasheets in order to obtain the desired PV model
using any circuit simulator uusing the single-diode model. Chapter 6, elaborate more on
the modeling method discussed in chapter 5, introducing the two-diode, and three-diode
models. The modeling of a PV grid-connected system is carried out in chapter 7. Finally,
the conclusion is stated in chapter 8.

4
2
Photovoltaic Principles

2.1 Introduction
This chapter talks about the basic principles of photovoltaics. The sections of this chapter
discuss the following topics: the physics of semiconductors highlighting few properties, a
brief description on the photovoltaic effect, the anatomy of the solar cell, how silicon can
be used to form a solar cell, the basic electrical parameters of the solar cell, the ideal
model of the solar cell, the efficiency of the solar cell, the non-ideal model of the solar
cell, and the different sun spectral distributions.

2.2 Semiconductors
2.2.1 Diamond Lattice and the Energy Band of Pure Semiconductors
The direct conversion of sunlight energy can be done using semiconducting materials.
Semiconductors are crystalline solids that have almost a perfect periodicity and regularity
of its atomic structure. Most atoms of intrinsic (pure) semiconductors have a unique
arrangement and structure that is called a diamond lattice. The diamond lattice of these
semiconductors occurs when four electrons in the outer shell of an atom from the forth
group of the periodic table forms a strong bond to the crystal structure [2]. In this
structure, electrons exist in electron states and as a result of having an interaction
between the atoms, these electron states are grouped in energy bands i.e. the valence band
and conduction band. Between these two bands there is an energy (band) gap [3]. The
band gap is categorized into two groups, direct and indirect band gap semiconductors.
Direct band gap semiconductors occur when the maximum point of the valence band has
the same k as the minimum point of the conduction band as shown in figure 2.1 (a). On

5
the other hand, indirect band gap semiconductors happen at different wavevectors as
shown in figure 2.1 (b) &figure 2.1 (c) [3].

2.2.2 Light Absorption

At zero temperature (no light emission), the valence band is completely full of electrons
while in the conduction band none of the energy levels are occupied by electrons. In this
case semiconductors behave as an insulator. As the temperature increases, some electrons
will jump from the valence band to occupy the energy levels in the conduction band,
which creates holes in the valence band. Both electrons and holes will contribute to
provide a current [2]. This can be explained in detail taking into consideration the types
of different band gaps. For direct band gap semiconductors the absorption of light will be
done if an empty space in the conduction band exists in which it has the same energy and
momentum of a photon and an electron existing in the valence band. In contrast, for the
indirect band gap semiconductors, the valence band is not aligned with the conduction
band so the interaction between the electrons in the valence band and the photons will not
grant the wanted energy and momentum of the corresponding empty space in the
conduction band. Therefore, another particle will be associated with photons, which is
called a phonon that assists in light absorption. As a consequence of having three
particles in the indirect band gap semiconductors (an electron, a photon and a phonon),
the possibility of the interaction of the three particles is lower that the electron-photon
interaction. As a result, the light absorption in the direct band gap semiconductors is
stronger as it only requires a simple interaction between an electron and a photon [3]. In
general, the conductivity, which depends on the number and mobility of charge carriers,
can be raised by increasing the temperature [2].

2.3 Photons In, Electrons Out

The photovoltaic effect is a physical process through which solar energy (light) is
converted into electrical energy directly. Light is made up of packets of energy that are
called photons. The energy of these photons, which depends upon the frequency, or color,
of light, is sufficient to excite electrons to higher levels of energy where they are free to

6
 ɛ ɛ ɛ

Conduction Band
phonon

Ec Ec Ec
phonon
photon photon
Eν Eν photon Eν

Valence Band
k k k
(a) (b) (c)

Fig. 2.1 Energy-wavevector diagram

7
move. This was explained through an experiment Einstein conducted in 1905 in which
blue or ultraviolet light provides enough energy to let electrons escape completely from
the surface of a metal. Although these electrons move to higher energy states, they relax
back to their ground state. However, in photovoltaic devices, there is a built-in
asymmetry that pulls the excited electrons away before they relax feeding them to an
external circuit, producing an electrical current [3].

2.4 Anatomy of a Solar Cell

The solar cell is the basic structure of photovoltaics. It can be considered as a two
terminal device that acts as a diode in the dark and generates electricity when exposed to
light. It is usually a thin layer of semiconductor material covered by a surface treated to
reflect as little visible light as possible. The surface is made of glass or other clear
material to seal the cell from the external environment. A pattern of metal contacts is
engraved on the surface to make electrical contact [3]. This is illustrated in figure 2.2 [4].

2.5 How Silicon Makes a Solar Cell

Silicon has special chemical properties, especially in its crystalline form. An atom of
silicon has 14 electrons, distributed in three different shells. The first two levels that hold
two and eight electrons respectively are completely full. However, the outer (valence)
level is only half full with just four electrons. Therefore, a silicon atom will share
electrons with four other nearby atoms to achieve a stable electron configuration. An
electron bond can be broken by the action of light or heat letting the electron moves
freely and leaving a hole in the crystal lattice. This is known as intrinsic conductivity.
Intrinsic conductivity cannot be used to generate electricity. This issue can be addressed
through introducing impurities, or what is called doping in which other atoms are
deliberately mixed with the silicon atoms. These impurities are classified into two
categories, donors and acceptors. Donors have one electron more (for example,
phosphorus), accepters have one electron less (for example, boron) than silicon in their
outermost electron level. In the case of phosphorus doping (n-doped), there is an extra
electron for every phosphorus atom in the lattice. This electron can move freely in the
crystal and hence transport an electric charge. However, with boron doping (p-doped),

8
Fig. 2.2 Solar cell anatomy

9
there is a hole (missing bonding electron) for each boron atom in the crystal lattice.
Electrons near silicon atoms can fill this hole, creating a new hole somewhere else. This
conduction method that is based on doping atoms is known as impurity conduction or
extrinsic conduction. If the n- and p- doped semiconductor layers are brought together
forming a p-n junction, extra electrons from the n-doped diffuse into the p-doped layer.
This creates a region with few free charge carriers, known as the space charge region.
Positively charged doping atoms stay in the n-region of the transition and negatively
charged doping atoms stay in the p-region of the transition. An electrical field is then
created that opposes the movement of the charge carriers. In the case of exposing the p-n
semiconductor (solar cell) to light, photons are absorbed by the electrons breaking their
bonds. The emitted electrons are pulled through the electrical field into the n-region,
While, the holes that are formed migrate in the opposite direction, into the p-region. This
whole process is called the photovoltaic effect which is shown in figure 2.3 and figure
2.4. The diffusion of charge carriers to the electrical contacts creates a voltage across the
solar cell terminals. In the case of an unloaded state, the open circuit voltage arises at the
solar cell. If the electrical circuit is loaded, a current flows. Some electrons do not reach
the contacts, they recombine instead. Recombination can be defined as the bonding of a
free electron to an atom lacking an outer electron (hole). Diffusion length is the average
distance, which an electron covers in the crystal lattice during its lifetime until it meets an
atom missing an electron and bonds with it. Free charge carriers can no longer contribute
to generating electricity [5]. The diffusion length depends upon the number of impurity
atoms in the crystal, which must be large enough so that a sufficient number of charge
carriers reach the contacts, as well as the material. In the space charge (depletion) region,
there is a high probability of charge carrier separation without recombination. Outside of
this region, the probability of recombination rises with the distance from the space charge
region. Figures 2.3 and 2.4 [5] show the sequence of this process.

10
 (a)

(b)

Fig. 2.3 (a) Crystalline structure of silicone and intrinsic conductivity.

(b) Extrinsic conduction in n- and p- doped silicon.

11
 (c)

Fig. 2.4 Formation of space charge region at the p – n junction

through diffusion of electron and holes.

12
2.6 Solar Cell Parameters
There are a couple of important parameters in the solar cell. The first parameter is the
open circuit voltage (Voc), which is defined as the voltage produced between the
terminals when there is no current flow or when the terminals of the cell are isolated (an
infinite load resistance). Also Voc is known as the crossing of the I-V curve with the
voltage axis. The second parameter is the short circuit current (Isc) which is known as the
current generated when the cell terminals are connected together (short circuit, zero load
resistance) .The short circuit current increases with more photons [4].

2.7 Ideal Solar Cell Model

The equivalent circuit for an ideal solar cell contains a single ideal diode as shown in
figure 2.5. Once the load is placed, a potential difference across the solar cell will exist.
As a result, a current which operates in the opposite direction of the light current
(photocurrent) will be generated. This current is called the dark current which basically
flows across the diode when a voltage is applied with no light emission (dark) [3]-[6]. It
can be expressed mathematically as follows [6]:

( ) ( ) (2.1)
Where [1]:

(2.2)

(2.3)

(2.4)

As the light is absorbed under no load conditions, the photocurrent will be produced
opposing the direction of the dark current as illustrated in figure 2.6. Therefore, to
calculate the overall current-voltage characteristic, the dark current density is added to

13
the illuminated characteristic as shown in equations 2.5 and 2.6 [6]. This is called the
superposition principle [2].

(2.5)

( ) (2.6)

For the open circuit voltage, Voc, equation (2.6) is set to zero.

( ) (2.7)

The V is then given by:

OC

( ) (2.8)

2.8 Efficiency

The power reaches its maximum at the cell's operating point or maximum power point.
The ratio of the maximum power from the solar cell to the product of Voc and Isc is
known as the fill factor (FF). This factor determines the maximum power delivered from
a solar cell. As shown in figure 2.7.The FF is the area of the largest rectangle that will fit
in the I-V curve. This happens at some voltage called Vm with a corresponding current Im.
The FF is expressed in equation 2.9 [6].

(2.9)

The efficiency η of the solar cell is the ratio of the solar cell output power to the solar cell
power affecting the solar cell surface (Pin) as illustrated in equation 2.10 [6].

(2.10)

14
 +
Isc Idark V

-
Fig. 2.5 Equivalent circuit of ideal solar cell

I(A) Isc
Light current

Dark current Voc

V (V)

Fig. 2.6 Current voltage characteristic of ideal diode in the light and the dark

15
 I
I, P MPP
Isc
Current
Im
Current ,Power

Power

Voltage Vm Voc V

Fig. 2.7 The current-voltage (black) and power-voltage (gray) characteristics of an ideal cell.

16
The efficiency is related to Isc, Voc and FF as shown in equation 2.11 [6].

= (2.11)

These four parameters: Isc, Voc, FF and η are the key performance characteristics of a
solar cell [3]-[6].

2.9 Solar Cell Equivalent Circuit Diagram

The equivalent circuit of a non-ideal solar cell comprises of two diodes. The first diode
describes the simple p-n junction behavior without electrons and holes generation and
recombination in the space charge region. The second diode that is connected in parallel
with the first diode describes the "non-ideality" of the real p-n junction due to the
recombination that occurs at the space charge region. The generalized model of a real
solar cell contains a shunt resistor that is parallel to both diodes. There is also an internal
series resistance that is in series to the parallel diodes and shunt resistor as it shown in
figure 2.8. The series resistance arises from the resistance of the cell material to current
flow, and from resistive contacts. At high current densities, series resistance becomes a
problem, for instance under concentrated light. On the other hand, the shunt resistance
arises from leakage of current through the cell, around the edges of the device and
between contacts of different polarity. It becomes a problem in poorly rectifying devices.
Series and shunt resistances reduce the fill factor. For an efficient solar cell Rs should be
small and Rsh should be as large as possible [6].

The mathematical expression of the resultant current in a non-ideal solar cell can be
written as follows [6]:

( ) ( )– (2.12)

17
 Rs

+
IL D1 D2 Rsh
V
_
Fig. 2.8 Two diode model of a solar cell

18
2.10 Sun Spectral Distribution
The radiation of the sun reaching the earth, distributed over a range of wavelengths from
300 nm to 4 micron approximately, is partly transmitted to the earth’s surface and partly
reflected by the atmosphere. At the surface of the earth the atmosphere selectively
absorbs the radiation at certain wavelengths. It is important to define two different sun
spectral distributions, the air masses (AM0) and (AM 1.5 G). The AM0 spectral
distribution is defined to be as the spectrum outside the atmosphere. However, the
AM1.5G is known as the spectrum at sea level under certain standard conditions. Several
magnitudes are essential concerning solar radiation, spectral irradiance, irradiance and
radiation. Spectral irradiance is known as the power received by a unit surface area in a
wavelength differential dʎ whose unit is W/m2 µm, where it is greater in the shorter
wavelengths than in the longer. Irradiance is regarded as the integral of the spectral
irradiance extended to all wavelengths of interest whose unit is W/m2. Radiation is the
time integral of the irradiance extended over a given period of time whose units are units
of energy. The irradiance received at a given surface depends on the time of the day, day
of the year, the site location at the earth’s surface (longitude and latitude) and on the
weather conditions [6].

2.10.1 Standard Sun Spectrum (AM1.5G)

The standard sun spectrum comes from Air Mass (AM) and from a number which is 0 for
outer space spectrum and 1.5 for the sea-level spectrum [6]. The sea-level spectrum can
be modeled using PSpice as given in appendix A.1. The PSpice subcircuit for this spectral
irradiance is shown in figure 2.9. The values of the AM1.5G spectral irradiance in W/m2 can
be seen in figure 2.10.

2.10.2 Standard Outer-space Spectrum (AM0)

The irradiance corresponding to the sun spectrum outside of the atmosphere is called
AM0 [6], this can be modeled using PSpice as listed in appendix A.2. A plot of this
spectrum is shown in figure 2.11.

19
2.10.3 Black Body Radiation
The sun radiation can also be approximated by the black body (a theatrical object that
appears black due to the absorption of all the radiation that hits it [7]) radiation at 5900K
[6]. PSpice can be used to simulate this spectrum as given in appendix A.3. Figure 2.12
shows the spectrum. Comparing the black body radiation plot with both AM1.5G and
AM0 plots, it is clearly shown that the black body radiation is smoother than both
AM1.5G and AM0.

20
 (12)

Output
(12)
+
Subcircuit
Vam1.5g
AM1.5G

(10) Reference

(10)

Fig. 2.9 PSpice subcircuit for the spectral irradiance AM1.5G.

21
 2.0
Spectral irradiance
(W/m2μm)

1.0

0
0 1.0 2.0 3.0
Wavelength(μm)
Fig. 2.10 Pspice plot of AM1.5G sun spectrum normalized to 1KW/m2 total irradiance.

3.0
Spectral irradiance

2.0
(W/m2μm)

1.0

0
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Wavelength(μm)
Fig. 2.11 PSpice plot of AM0 sun spectrum.

22
 2.0
Spectral irradiance
(W/m2μm)

1.0

0
0 1.0 2.0 3.0 4.0
Wavelength(μm)
Fig. 2.12 Black body spectral irradiance at 5900K.

23
3
PSpice/Simulink Based Study of Photovoltaic
Cell
3.1 Introduction
The most significant semiconductor properties that determine the effectiveness of
generating electricity when exposed to light are the absorption coefficient, the reflectance
of semiconductor surface, the drift-diffusion parameters and the surface recombination
velocities at the surface of the solar cell. The absorption coefficient depends upon the
value and the type of the bandgap of the semiconductor, indirect or direct bandgap, as
well as the semiconductor material. The second property, reflectance, is dependent upon
the structure of the solar cell in the shape and antireflection coating. If proper
antireflection design is used, the reflectance will be reduced allowing more photons to be
absorbed and hence improving power generation.

This chapter summarizes an overall study of the basic parameters that affect the
performance of the solar cell model using two software programs, namely, PSpice and
Simulink. The sections of this chapter discuss the following topics: The analytical solar
cell model and the parameters concerned with it, the ideal solar cell model, the open
circuit voltage, the maximum power point, the generalized model of the solar cell and the
effects of the different parameters, the behavioral solar cell model, and a general study of
the photovoltaic modules and arrays.

24
3.2 Analytical Solar Cell Model
The analytical solar cell model comprises of two regions that are known as the emitter
and base. As the solar cell is illuminated, the photocurrent will be generated in the
external electric short circuit due to the low reflection coefficient at the emitter surface
that is not covered by the metal grid. The current will flow out of the device by the base
layer as shown clearly in figure 2.2. The three significant magnitudes concerned with the
analytical solar cell are the spectral short circuit current density, the quantum efficiency
and the spectral response.

3.2.1 Spectral Short Circuit Current Density

The spectral short circuit current density can be defined as the photocurrent that is
collected by a surface area of the solar cell when it is exposed to light and circulating by
an external short circuit. Both the emitter and base layers contribute to generating this
current .The analytical expression of both emitter and base short circuit current densities
are expressed in equations 3.1 and 3.2 [6] respectively. The PSpice file describing the
model of the spectral short current density for both the emitter and base is listed in
appendix B.1.The structure of this file is shown in figure 3.1. The simulation results can be
seen in figure 3.2.

( ) ( )
( ) [ ] (3.1)
( )

( ) ( )
( ) [ ] (3.2)
( )

Where the unit of the spectral short circuit current density is A/cm2μm.

25
 (15)
Subcircuit Node(11)

Lemda.lib
(16)

ʎ
(105) Jsce (ʎ )
(101)
Node(12) α(ʎ) (106) Jscb (ʎ )
(102) Subcircuit
Subcircuit (12)
(107) Jsc
Iʎ
silicon_absorption_ (103) (108) QE(ʎ)
coeffeicient.lib
R(ʎ) Jshortcircuit.lib
(104) (109) SR(ʎ)
(10)
(100)

+
Dc
Node(13) voltage

Subcircuit (12)
AM15gSub.lib
(10)

Reference node

Fig. 3.1 Structure of the base and emitter short circuit currentdensity.

26
 80
Spectral short-current
density(mA/cm2μ)

40

0
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Wavelength(μm)
(a)

30
Spectral short-current
density(mA/cm2μ)

20

10

0
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Wavelength(μm)
(b)

Fig. 3.2 (a) Base spectral short circuit current density.

(b) Emitter spectral short circuit current density.

27
3.2.1.1 Spectral Photon Flux

It can be easily seen from equations 3.1and 3.2 that the short circuit current density
depends on several magnitudes that are functions of the wavelength such as the
absorption coefficient , the reflectance R ( ) and the spectral photon flux which is
received at the front surface of the emitter and base layers of a solar cell.
The spectral photon flux is dependent upon both spectral irradiance and wavelength and
it has a unit of photon/cm2μm s. Both equations of the spectral photon flux for the two
regions of the emitter and base are given in equations3.3 and 3.4 respectively [6].

(3.3)

(3.4)

3.2.1.2 Total Short Circuit Spectral Current Density

The total short circuit spectral current density at a certain wavelength can be calculated
by adding the base and emitter components of the spectral short circuit current density to
give the following equation [6]:

(3.5)

Where the total short circuit spectral current density has a unit of A/cm2μm.

3.2.1.3 Total Short Circuit Current Density

It is worth mentioning that the total spectral short circuit current density is a different
magnitude than the total short current density generate by a solar cell when illuminated
by a spectral light source, which is proportional to the value of the irradiance. The
relationship between these two magnitudes is shown in equation 3.6 [6].The total short
circuit current density is modeled in PSpice as given in appendix B.1. The simulation
results are shown in figure 3.3.

∫ ∫ ( ) (3.6)

Where the unit of the total short current density is A/cm2.

28
 40
Integral of the spectral short-
current density(mA/cm2)

20

0
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Wavelength(μm)
Fig. 3.3 The total short circuit current density.

29
3.2.2 Quantum efficiency (QE)
Quantum efficiency, which is known as the number of electrons generated in the external
circuit by the solar cell for every photon in the incident spectrum, is categorized into
internal and external efficiencies. The internal quantum efficiency is related to the non-
reflected part of the incident spectrum, while the external quantum efficiency is related to
the total spectral irradiance taking into account both the reflected and non-reflected parts.
The internal and external quantum efficiencies are generally given in percentage as
shown in equation 3.7 and 3.8 [6] respectively. The PSpice model for this quantity is
listed in appendix B.1. The simulation results are shown in figure 3.4.

(3.7)
( )

(3.8)

3.2.3 Spectral response (SR)

The ratio between the short circuit spectral current density and the spectral irradiance is
known as the spectral response which can be expressed as in equations3.9 and 3.10 [6]
for the internal and external components respectively. The PSpice file model for this
quantity is listed in appendix B.1. The simulation results are shown in figure 3.5.

(3.9)
( )

(3.10)

Where the unit of the spectral response is A/W.

30
 100
Quantum efficiency(%)

50

0
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Wavelength(μm)

Fig. 3.4 The total internal quantum efficiency of a solar cell.

800
Spectral response(mA/W)

400

0
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Wavelength(μm)
Fig. 3.5 The spectral response of a solar cell.

31
3.3 Ideal Model of a Solar Cell
The circuit representation of an ideal solar cell is expressed in equation 2.6. The solar cell
receives a given irradiance value generating the light current that is proportional to this
irradiance and influenced by the temperature. This can be easily seen in following
mathematical expression of light current [8]:

( ( )) (3.11)

G is the irradiance that has a unit of W/m2. The value of JI,n is given at nominal
conditions (AM1.5G, Gn=1000 W/m2, Tn = 25oC). KI is the cell’s short circuit current
temperature coefficient and equal to 0.0017A/oC and A is the solar cell area. It is known
that the light current is equal to short circuit current in the ideal solar cell when setting V
to zero and for the sake of simplicity it is assumed that the cell temperature T is 25oC,
then the short circuit current expression can be simplified to [6]:

( ) (3.12)

This ideal cell can be modeled using PSpice as described in appendix B.2. A Simulink
model has also been obtained for this cell taking into account the assumption of the
following parameters:

A= 126.6 cm2

Jsc,n= 0.0343 A/cm2

G= Gn= 1000 W/m2

Substituting the assumed values in equation 3.12 gives Isc,n = 4.342A (Note that J01 =
1×10-11 A/cm2). Both PSpice and Simulink subcircuit models and block diagrams of the
cell are shown in figure 3.6 and 3.7 respectively. The I-V characteristics of an ideal solar
cell for both the PSpice and Simulink models are shown in Figure 3.8. Figure 3.9 shows
the I-V characteristics of an ideal solar cell for different irradiances which proves that Voc
depends logarithmically upon Isc which scales linearly with irradiance as seen in equation
3.12.

32
(103) (102)

G_irrad

(101) (101)

(a)

(12) (11)
(103) (102)

V_irrad V_bias
+ +
Subcircuit
ideal_cell.lib

(101)

(0)

(b)

Fig. 3.6 (a) PSpice subcircuit of a solar cell.

(b) PSpice block diagram of a subcircuit.

33
 Subsystem1
1 G G IL

q/nKT eu
I
1 V 1 IO1

(a)

A Jsc
IL

G 1/Gn

(b)

Fig. 3.7 (a) Simulink block diagram of an ideal solar cell.

(b) Simulink subsystem1 included in the main model.

34
 5.0

4.0
Current (A)

2.0

0
0 100 200 300 400 500 600

Voltage (mV)
(a)

5.0

4.0
Current (A)

2.0

0
0 100 200 300 400 500 600

Voltage (mV)
(b)

Fig. 3.8 (a) The I-V curve of the ideal PSpice solar cell model.
(b) The I-V curve of the ideal Simulink solar cell model.

35
 5.0
G=1000W/m2
G=800W/m2
G=600W/m2
4.0 G=400W/m2
Current (A)

G=200W/m2

2.0

0
0 100 200 300 400 500 600

Voltage(mV)
(a)

5.0 G=1000W/m2
G=800W/m2
G=600W/m2
4.0 G=400W/m2
Current (A)

G=200W/m2

2.0

0
0 100 200 300 400 500 600

Voltage (mV)
(b)

Fig. 3.9 (a) The I-V curve of the PSpice solar cell model under several irradiance values.

(b) The I-V curve of the Simulink solar cell model under several irradiance values.

36
3.4 Maximum Power Point
The output power of a solar cell can be defined as the product of the output current
delivered to the electric load and the voltage across the cell. In general positive power
indicates power being delivered to the load and negative power indicates power being
consumed by the solar cell. The power can be represented as follows [6]:

P=V×I= ( ) (3.13)

The maximum power generated by a solar cell happens at a point called the maximum
power point (MPP) having the coordinates of V = Vm and I = Im as follows [6]:

( ) (3.14)

( ) (3.15)

The coordinates of the maximum power point can be plotted in PSpice and Simulink as
the I-V product that is shown in figure 3.10 for different irradiances.

3.5 Generalized Model of a Solar Cell

The equivalent circuit of an ideal solar cell based on an ideal diode and an ideal current
source has been described earlier. Generally this level of model is inadequate to
accurately represent the maximum power delivered by the solar cell due to several factors
that should be taken into account some of which are the series resistance, shunt
resistance, recombination and non-ideality of the diffusion diode. Equation 2.10
expresses this model.

37
 2.5

G=1000W/m2
2.0
G=800W/m2
G=600W/m2
Power(W)

G=400W/m2
G=200W/m2

1.0

0
0 100 200 300 400 500 600
Voltage (mV)
(a)

2.5

2.0 G=1000W/m2
G=800W/m2
Power(W)

G=600W/m2
G=400W/m2
G=200W/m2

1.0

0 100
0 200 300 400 500 600
Voltage(mV)
(b)

Fig. 3.10 (a) PSpice solar cell model power for several irradiance values.
(b) Simulink solar cell model power for several irradiance values.

38
3.5.1 Effects of the Series Resistance
The PSpice file developed to study the effects of the series resistance on both the short
circuit current and the open circuit voltage is listed in appendix B.3. The analysis was
done for several series resistance values at an appropriate constant value of the shunt
resistance, namely, 1×105 Ω, and at equivalent values of the irradiance and the
temperature and J02 is chosen to be 1×10-9A/cm2. The PSpice and Simulink subcircuit
models and block diagrams are shown in figures 3.11 and 3.12 respectively. The
simulation results of the PSpice file and Simulink models are shown in figure 3.13. As
can be seen in this figure, the series resistance has a noticeable impact on the I-V
characteristics of the solar cell in which the increase of the series resistance value causes
a severe reduction in the short circuit current this can be mathematically proved by
setting V to zero in equation2.14 then the short circuit current will be equal to [6]:

( ) ( ) (3.16)

However, the open circuit voltage is independent of the series resistance value as all
curves cross the same point on the voltage axis as it is shown in following equation that is
derived from equation 2.14 by setting I to zero and neglecting the third and fourth terms
as follows [6]:

( ) (3.17)

The fill factor is one of the most affected electrical parameters of the solar cell by the
series resistance. As discussed earlier, the power delivered by the cell can be expressed as
[6]:

(3.18)

Where the maximum power delivered by the solar cell when non-zero Rs can be
calculated as follows [6]:

( ) (3.19)

39
 Rs
(301)
(302) (303)

IL=g_irrad D1 D2 Rsh

(300) (300)
(a)

(302) (303)
Subcircuit
non_ideal.lib

(300)

(b)

Fig. 3.11 (a) Subcircuit of a non-ideal solar cell including two diodes and
series and shunt resistors.
(b) PSpice block diagram of the subcircuit.

40
 q/nKT
eu
IL
1 V
1
I
Io2

1/Rsh

Rs

eu
q/nKT Io2

Fig. 3.12 Simulink developed model to study the effect of the series resistor.

41
 5.0

4.0
Current(A)

Rs= 0.0001Ω
2.0 Rs= 0.001Ω
Rs= 0.01Ω
Rs= 0.1Ω
Rs= 1Ω

0
0 100 200 300 400 500 600
Voltage(mV)

(a)

5.0

4.0
Current(A)

Rs= 0.0001Ω
2.0 Rs= 0.001Ω
Rs= 0.01Ω
Rs= 0.1Ω
Rs= 1Ω

0
0 100 200 300 400 500 600
Voltage(mV)

(b)

Fig. 3.13 (a) The series resistance effect on the PSpice cell model I-V characteristics for
several series resistance values.
(b) The series resistance effect on the Simulink cell model I-V characteristics for
for several series resistance values.

42
Where is the normalized value of the series resistance and is given by:

(3.20)

The fill factor with non-zero Rs can then be written as follows [6]:

( )
(3.21)

It can be concluded that increasing the value of the series resistance causes a great
reduction in the fill factor degrading the performance of the solar cell.

3.5.2 Effects of the Shunt Resistance

The effects of the shunt resistance on the solar cell parameters can be studied similarly as
done for the series resistance. The PSpice files developed for this purpose is listed in
appendix B.4. The Simulink model developed is shown in figure 3.14. Here, the series
resistance has an appropriate constant value of 1×10-6Ω. The second diode has been
eliminated for the analysis by setting its saturation current to 0. The simulation results for
several values of the shunt resistance are shown in figure 3.15, which shows a severe
reduction in the open circuit voltage due to the decrease in the shunt resistance value that
results in a reduction of the fill factor degrading the solar cell performance. On the other
hand, the short circuit current is independent upon the shunt resistance value as all curves
meet at the same point on the current axis. Substituting in equation 2.14, with Rs=0, J02=0
and V=0, results in:

(3.22)

That is independent of Rsh.

43
 q/nKT IL
u
e I
1 V
IO1
1

1/Rsh

Rs

Fig. 3.14 Simulink model developed to study the effect of the shunt resistor.

44
 5.0

Current (A) 4.0

Rsh= 10kΩ
2.0 Rsh= 1kΩ
Rsh= 0.1kΩ
Rsh= 10Ω
Rsh= 1Ω
Rsh= 0.1Ω

0
0 100 200 300 400 500 600

Voltage(mV)

(a)

5.0

4.0
Current (A)

Rsh= 10kΩ
2.0 Rsh= 1kΩ
Rsh= 0.1kΩ
Rsh= 10Ω
Rsh= 1Ω
Rsh= 0.1Ω

0
0 100 200 300 400 500 600
Voltage(mV)

(b)
Fig. 3.15 (a) The shunt resistance effect on the PSpice cell model I-V curve for several shunt
resistance values.
(b) The shunt resistance effect on the Simulink cell model I-V curve for several shunt.
resistance values.

45
3.5.3 Effects of the Recombination Diode
The PSpice file describing the effects of the recombination diode is shown in appendix
B.5. The Simulink model developed is shown in figure 3.16. Here, both series and shunt
resistances have appropriate values of 1×10-6Ω and 1000Ω respectively in order to cancel
their effects and isolate the recombination diode effect. The simulation results for
different values of diode recombination are shown in figure 3.17. This figure indicates
that when the recombination diode dominates, the open circuit voltage is heavily
degraded reducing the fill factor and performance of the solar cell. This is can be proved
mathematically by setting I to zero in equation 2.14 and neglecting the second and fourth
terms so that the open circuit voltage will be equal to [6]:

( ) (3.23)

By increasing the value of Jo2, Voc will logarithmically decrease and vice versa. However,
when setting V to zero in equation 2.14 the short circuit current can be expressed as
follows noting that the value of Jo1 is chosen to be very small so it can be neglected with
both Rs and Rsh [6]:

(3.24)

This clearly shows that the short circuit current is independent of J02 and remains
constant.

3.5.4 Effects of the Temperature

Temperature plays a critical role on the electrical response of a solar cell. This comes
from the strong relationship between the saturation current of the diode and temperature
as seen in equation 3.25 [6].

46
 q/nKT IL
u
e
1 V
1 I
IO

1/Rsh

Rs

eu
q/nKT 1
A Jo2

Fig. 3.16 Simulink model developed to study the effect of the recombination diode.

47
 5.0
Current (A)

2.5
Jo2=1x10-8 A/cm2
Jo2=1x10-7 A/cm2
Jo2=1x10-6 A/cm2
Jo2=1x10-5 A/cm2
Jo2=1x10-4 A/cm2

0
0 100 200 300 400 500 600
Voltage (mV)
(a)

5.0
Current(A)

2.5 Jo2=1x10-8 A/cm2

Jo2=1x10-7 A/cm2
Jo2=1x10-6 A/cm2
Jo2=1x10-5 A/cm2
Jo2=1x10-4 A/cm2

0
0 100 200 300 400 500 600

Voltage(mV)
(b)

Fig. 3.17 (a) Effect of the recombination diode on the PSpice cell model I-V curve for
several Jo2 values.
(b) Effect of the recombination diode on the Simulink cell model I-V curve for
several Jo2 values.

48
The temperature effects can be modeled in PSpice as shown in appendix B.6.The
Simulink model developed is shown in figure 3.18. The simulation was performed for
several temperature values at a small series resistance and a large shunt resistance. The
second diode has been removed from the model for simplicity. The simulation results are
shown in figure 3.19. As can be seen the open circuit voltage is heavily affected by the
temperature in which it decreases for increasing temperature values. The short circuit,
however, remains constant at different temperature values.

( ) ( ) (3.25)

Where Eg is a semiconductors band gap energy, K is the Boltzmann’s constant (1.381

10-23J/A) , q is the electron charge (1.602 10-19C) and In,0 is the short circuit current at
the nominal conditions (Tn=25oCand G=1000 W/m2) and n is the ideality factor of the
diodes.

The value of the energy band gap of the silicon is =1.17.

3.5.5 Effects of Space Radiation

The primary effect of space radiation concerns the minority carrier lifetime degradation
in the semiconductor bulk leading to an increase of the dark current density, and a
degradation of the photocurrent generated. All in all, these effects produce a significant
reduction of the maximum output power of the solar cell, which is described analytically
by empirical equations relating the values of the short circuit current, open circuit voltage
and the maximum power to the fluence of a given high energy particle Φ in equations
3.26, 3.27, and 3.28 [6].

49
 1 T T Io1
q/nKT
Subsystem1 IL

eu
I
1 V 1

1/Rsh

Rs
(a)

Io,n

T T^3 Io

1/Tn

q/nk eu
1/T

Ego
(b)

Fig.3.18 (a) Simulink developed model to study the effect of temperature.

(b) Simulink subsystem1 included in the main model.

50
 5

4
Current (A)

T=27oC
T=35oC
2 T=40oC
T=45oC
T=50oC
T=55oC
T=60oC
0
0 100 200 300 400 500 600

Voltage (mV)
(a)

4
Current (A)

T=27oC
T=35oC
2 T=40oC
T=45oC
T=50oC
T=55oC
T=60oC
0
0 100 200 300 400 500 600
Voltage (mV)
(b)

Fig.3.19 (a) Effect of temperature on the PSpice for several T values.

(b) Effect of temperature on the Simulink for several T values.

51
 ( ) ( ) ( ) (3.26)

( ) ( ) ( ) (3.27)

( ) ( ) ( ) (3.28)

Where the value of J0 of the BOL conditions can be calculated as follows [6]:
( )
( ) ( ) (3.29)

And the ideal value of FF in BOL is expressed as follows [6]:

( )
( ) (3.30)

Where

(3.31)

Moreover, the Rs in BOL is given as follows [6]:

(3.32)
( )

This effect can be modeled in PSpice as described in appendix B.7 and the Simulink
model developed is shown in figures 3.20 and 3.21. The simulation was done for several
space radiation fluencies at a temperature of 300 K and the results are shown in figure
3.22 that shows a severe degradation in both the short circuit current and open circuit
voltage indicating a great reduction in the maximum output power.

52
 1 Φ Φ IL

Subsystem1

q/nKT eu
1 I
1 V
Subsystem2
1 Φ Φ Io1

Voc

Jsc X A

Jsc2 X A
νoc
Rs
Pmax
νoc
ln
0.72
1

νoc

Fig.3.20 Simulink model developed to study the effect of space radiations fluencies.

53
 Jsc(BOL)
Gn A
1 Ki IL
log10
Gn
Φ 1/Φi

(a)

Jsc(BOL) A Io1

Φ 1
log10 Ki
1/Φi

1 Voc(BOL) -1
eu
1/Φν log10 Kν
VT

(b)

Fig. 3.21 (a) Simulink subsystem1 included in the main model.

(b) Simulink subsystem2 included in the main model.

54
 0.5
Current (A)

0.2 f=1x1010 elecrons/cm2

f=1x1011 elecrons/cm2
f=1x1012 elecrons/cm2
f=1x1013 elecrons/cm2
f=1x1014 elecrons/cm2
f=1x1015 elecrons/cm2
f=1x1016 elecrons/cm2

0
0 100 200 300 400 500 600
Voltage(mV)
(a)
0.5
Current(A)

0.25

f=1x1010 elecrons/cm2
f=1x1011 elecrons/cm2
f=1x1012 elecrons/cm2
f=1x1013 elecrons/cm2
f=1x1014 elecrons/cm2
f=1x1015 elecrons/cm2
f=1x1016 elecrons/cm2
0
0 300 400 500 600
100 200
Voltage(mV)
(b)
Fig. 3.22 (a) Effect of space radiation on the I-V characteristic of a Si PSpice cell model for space
Radiation fluencies.
(b) Effect of space radiation on the I-V characteristic of a Si Simulink cell model for space
Radiation fluencies.

55
3.6 Behavioral Solar Cell Model
Solar cells that are not made of famous semiconductor materials such as Si, GaAs and
Ge, are difficult to study their behaviors for changing temperature and irradiances as the
required data to model the effects of temperature is not easily accessible. Therefore, the
typical solar cell model that includes a diode is not very practical and is restricted in
terms of use. As a result, a behavioral model of a solar cell is introduced. The behavioral
model presumes that the solar cell can be modeled by a series resistance and two current
sources. The first current source is expressed as follows [6]:

( ) ( )( ) (3.33)

( )
( ) ( ) (3.34)
( )

The cell operating temperature is derived from the NOCT (nominal operating
conditions temperature) concept at 800 W/m2 irradiance and 20°C of ambient temperature
[6].

(3.35)

The open circuit voltage is approximately can be given as [6]:

( )( ) (3.36)

The series resistance can be calculated as follows [6]:

(3.37)

The current of the maximum power point can be expressed as follows [6]:

( ) ( ) (3.38)

Then, the voltage of maximum power point is written as follows [6]:

56
 ( ( )) (3.39)

The model can be implemented in PSpice using equations 3.33, 3.34, 3.35, 3.36, 3.37,
3.38 and 3.39 as listed in appendix B.8. The schematic of the solar cell behavioral model
and block diagram are shown in figure 3.23. Figure 3.24 shows that the model is able to
simulate the I-V characteristics of a solar cell just like the ordinary solar cell model.

57
 R
(201)
(202) (204)
G_irrad eisc (205)
(203) (206)
evoc
(207)
evtemp
G_irrad (208)
gim
evm
(209)

(200)

(a)

Irradiance Cell output

(202) (204)
Ambient Short Circuit
Temperature Current
(203) (205)
Open Circuit
Voltage
Subcircuit (206)
cell_behavioural.lib Cell
Temperature
(207)
MPP current
(208)

MPP voltage
(209)
(200)

(b)
Fig. 3.23 (a) Schematic of the solar cell behavioral model.
(b) Block diagram of the solar cell behavioral model.

58
 20

15
Current (mA)

10

0
0 100 200 300 400 500 600 700 800

Voltage (mV)

Fig. 3.24 The I-V characteristics of a CIGS solar cell simulated using the behavioral model.

59
4
PSpice/Simulink Based Study of Photovoltaic
Module/Array

4.1 Introduction
In most photovoltaic applications, high voltage levels are mandatory. This cannot be
achieved by a single solar cell that has a limited potential in which the open circuit
voltage is independent upon the area of the cell and is limited to the semiconducting
material properties. This suggests the need to connect more than one solar cell in series in
order to scale-up the voltage produced by a PV generator. Not only high voltage levels
are required in most applications but also high current levels that can be achieved through
the parallel connection of the solar cells connected in series. These connections will be
described in detail in the following sections.

4.2 Series Connection of Solar Cells

Several cases will be discussed here, depending on the irradiance levels and internal
parameter values of the different cells.

4.2.1 Connection of Identical Solar Cells

The corresponding PSpice file for this connection is given in appendix C.1. The PSpice
block diagram of the association of two solar cells in series are shown in figure
4.1.Figure 4.2 shows the Simulink block diagram. The simulation results are shown in
figure 4.3 as can be seen the short circuit current is the same current obtained from a
single solar cell; however, the open circuit voltage is twice the value indicating the series
connection of two solar cells exposed to the same irradiance level with equal series and
shunt resistances.

60
 (32) (33)
(302) (303)

Virrad1
+ Subcircuit
non_ideal.lib

(300) (35)

(34)
(302) (303)
Vbias

Virrad2
+
Subcircuit
non_ideal.lib

(300) (0)

Fig. 4.1 PSpice block diagram of the association of two solar cells in series.

61
 IL

q/nKT
eu
1 V
1 Io1

1/Ns
eu

1 Io2
q/nKT

Fig. 4.2 Simulink block diagram of the association of two solar cells in series.

62
 5

4
Current (A)

0
0 200 400 600 800 1000 1200
Voltage (mV)
(a)

4
Current (A)

00
200 400 600 800 1000 1200
Voltage (mV)
(b)
Fig. 4.3 (a) The I-V curve of the series connection of two identical PSpice solar cells.
(b) The I-V curve of the series connection of two identical Simulink solar cells.

63
4.1.2Connection of identical solar cells with different irradiance levels: ‘hot spot’
problem

Here the two solar cells described earlier are exposed to different irradiance values. This
common situation may occur due to the presence of dirt in one of the solar cells. The first
cell is exposed to an irradiance of 1000 W/m2 while the second cell is exposed to 700
W/m2 irradiance value. The corresponding PSpice file for this condition is given in
appendix C.2.The I-V characteristics are shown in figure 4.4. As can be seen the total
short circuit current of the series connection has taken a similar value to that of the less
illuminated solar cell (700 W/m2) conditions. A comparison between the power delivered
by each cell and the power delivered by the series connection is shown in figure 4.5. It
can be concluded that some of the power produced by the first solar cell is dissipated by
the second solar cell thereby reducing the available output power. This phenomenon is
called the ‘hot spot’.

4.2.3Bypass diode in series strings of solar cells

The ‘hot spot’ problem can be avoided using a bypass diode across a part of a series
string of solar cells or across each solar cell. The corresponding PSpice file is described
in appendix C.3. The simulation results are illustrated in figures 4.6. It can be concluded
that in the case of a shadowed solar cell in an array the output current is similar to that of
an illuminated array due to the bypass diode placed across the cell at the price of lower
open circuit voltage value.

4.3Shunt Connection of Solar Cells

As have been discussed earlier higher voltage levels can be obtained through the series
connection of solar cells. On the other hand, scaling of current can be achieved by
scaling-up the solar cell area that is by the parallel connection of series strings of solar
cells. The PSpice file for this type of connection of two identical solar cells is listed in
appendix C.4 and the Simulink developed model is illustrated in figure 4.7. Figure 4.8
shows the I-V characteristics of the two solar cells individually in comparison with the
parallel connected cells.

64
 5

4
Current (A)

0
0 200 400 600 800 100 1200
Voltage (mV)
(a)

4
Current (A)

00
200 400 600 800 1000 1200

Voltage (mV)
(b)

Fig. 4.4 (a) The I-V curve for the shunt connection of two identical PSpice cells with
different irradiances.
(b) The I-V curve for the shunt connection of two identical Simulink cells with
different irradiances.

65
 5
Total power
Cell#1 power
Cell#2 power
Power (W)

-5
0 200 400 600 800 1000 1200
Voltage(mV)
Fig. 4.5 Power of the unevenly illuminated solar cells

66
 5.0
Current (A)

2.5

0
0 1000 2000 3000 4000 5000 6000 7000 8000

Voltage (mV)
(a)

5.0
Current (A)

2.5

0
0 1000 2000 3000 4000 5000 6000 7000 8000
Voltage (mV)
(b)

Fig. 4.6 The I-V curve of a string of PSpice solar cell Module without the bypass diode (a) and
with it.

67
 IL

q/nKT eu

1 V
1 NpIo1
1

eu

NpIo2
1
q/nKT

Fig. 4.7 Simulink block diagram of the association of solar cells in parallel.

68
 10
Total
Cell #1 at 1000W/m2
Cell #2 at 700W/m2
Current (A)

0
0 100 200 300 400 500 600
Voltage (mV)
(a)

10
Total
Cell #1 at 1000W/m2
Cell #2 at 700W/m2
Current (A)

0
0 100 200 300 400 500 600
Voltage (mV)
(b)

Fig. 4.8 (a) The I-V curve of the parallel connection of two PSpice solar cell models.
(b) The I-V curve of the parallel connection of two Simulink solar cell models.

69
It can be seen that the short circuit current for the parallel connection is the addition of
the two short circuit currents of the individual cells.

4.4 The Terrestrial PV Module

In terrestrial applications the PV standard modules are composed of a number of solar
cells connected in series, usually 33 to 36 cells. As have been discussed in section 4.2 the
voltage scaling of the I-V characteristic of a single solar cell represents the characteristic
of the PV module. Usually, PSpice allows a certain number of circuit nodes to perform
simulation. As a result, a more compact version of a PV module is needed. Using the
general scaling rules of current and voltage this model could be used for a single PV
module, and then scaled-up to build a PV plant. Considering the I-V characteristic of a
single solar cell mentioned in chapter 1, the scaling rules of voltages, currents and
resistances when Ns solar cells are connected in series and Np cells are connected in
parallel are shown in equations 4.1 and 4.2 with taking in consideration that the values of
Rsh =10000 Ω and Rs=0.0001 Ω. Moreover, the effects of the second diode is neglected
(J02=0) for simplification purposes. The assumption of = has been made.

(4.1)
(4.2)
Where the subscripts M stands for module.

The PSpice file developed to simulate a 36 1 MSX-60 PV module and the provided
parameters in datasheet that is included in appendix C, which comprises of 36 solar cells
connected in series, is listed in appendix C.5. The value of Io is assumed to be 1×10-9.
The PSpice block diagram and Simulink representation are shown in figures 4.9 and 4.10
respectively. This model reproduces the whole standard AM1.5G I-V and P-V
characteristics of a commercial module having values of 3.8 A short circuit current,
20.39V open circuit voltage, 63.6 W maximum power. Those values are close to the
given values in the datasheet of the module. The PSpice and Simulink I-V and P-V
simulation results for this model are shown in figures 4.11 and 4.12 respectively.

70
4.5 PV Arrays
In general PV generators, PV arrays, are formed by combinations of series and parallel
connections of solar cells or PV modules. In the case of outer-space applications, arrays
are designed especially for a given space satellite or station. However, in terrestrial
applications arrays are formed by connecting PV modules containing a certain number of
series-connected solar cells and bypass diodes if needed. In this section the arrays used
for outer-space applications will be discussed. The PSpice model developed is listed in
appendix C.6. Note that Rsh =10000 Ω and Rs=0.0001 Ω with neglecting the effects of
the second diode (J02=0) for simplification purposes and assuming that = . This
model simulates the results of a space photovoltaic array composed of 504 solar cells
series which is an association of 72 sets of 7 parallel each. Figure 4.13 shows the
Simulink developed models of the MSX-64 found in appendix C note that the Rsh and Rs
are neglected for simplification purposes. The simulation results showing the I-V
characteristics of the array using PSpice and Simulink are shown in figure 4.14. Figure
4.15 shows the P-V characteristics of the array using PSpice and Simulink .This model
reproduces I-V and P-V characteristics of a commercial module having values of 28A
short circuit current, 40.89V open circuit voltage, 940.4 W maximum power.

71
 (601) RsM
(602) (603)

g_irradmodule D1 RshM

(600) (600)

(a)

(602) (603)
Subcircuit
Module.lib
(600)

(b)

Fig. 4.9 (a) Subcircuit of a non-ideal solar module including one diode and
series resistance.
(b) PSpice block diagram of the subcircuit.

72
 IL

1 V u
e
I

I0
q/nKT
P
1/Ns

Fig. 4.10 Simulink developed model to simulate the PV module.

73
 5.0

4.0
Current (A)

2.0

0
0 5 10 15 20 25
Voltage (V)

(a)

4
Current (A)

0
0 5 10 15 20 25
Voltage (V)
(b)

Fig. 4.11 (a) The I-V curve of MSX-60 photovoltaic module obtained by PSpice.
(b) The I-V curve of MSX-60 photovoltaic module obtained by Simulink.

74
 80

60
Power (W)

40

20

0
0 5 10 15 20 25
Voltage (V)

(a)

80

60
Power (W)

40

20

0
0 5 10 15 20 25
Voltage (V)
(b)

Fig. 4.12 (a) The P-V curve ofMSX-60 photovoltaic module obtained by PSpice
(b) The P-V curve of MSX-60 photovoltaic module obtained by Simulink.

75
 Np*IL

1 V eu
I

Np*I0
q/nKT
P
1/Ns

Fig. 4.13 Simulink developed model to simulate the PV array.

76
 30

25
Current (A)

20

15

10

0
0 5 10 15 20 25 30 35 40 45 50
Voltage (V)
(a)

30

25
Current (A)

20

15

10

0
0 5 10 15 20 25 30 35 40 45 50
Voltage (V)
(b)

Fig. 4.14 (a) The I-V curve of the 72 7 array using MSX-64 photovoltaic module
obtained by PSpice.
(b) The I-V curve of the 72 7 array using MSX-64 photovoltaic module
obtained by Simulink.

77
 1000

900

800

700
Power(W)

600

500

400

300

200

100

0
0 5 10 15 20 25 30 35 40 45 50
Voltage (V)

(a)

1000
900
800
Current (A)

700
600
500
400
300
200
100
0
0 5 10 15 20 25 30 35 40 45 50
Voltage (V)
(b)
Fig. 4.15 (a) The P-V curve of the 72 7 array using MSX-64 photovoltaic module obtained
by PSpice.
(b) The P-V curve of the 72 7 array using MSX-64 photovoltaic module obtained
by Simulink.

78
5
MATLAB Modeling and Simulation of
Photovoltaic Modules: the Single-Diode
Model

5.1 Introduction
Manufacturers of PV modules provide only a few experimental data regarding the
electrical and thermal characteristics of these modules. The parameters provided in the
PV datasheets, which are obtained under nominal conditions of temperature and solar
irradiation, are not enough to build an accurate PV circuit model with any circuit
simulator using basic math blocks. Therefore, a modeling method should be investigated
that would find the rest of these parameters in order to obtain the desired PV model using
any circuit simulator. Different techniques have been developed in the literature
regarding this issue. One particular method finds the non-linear equation parameters by
adjusting the I-V curve at three remarkable points: the open circuit voltage, the short
circuit current, and the maximum power point [9]. Three models will be investigated in
chapters 5 and 6 using this technique namely, the single-diode model, the two-diode
model, and the three-diode model. The developed models will be suitable to simulate
several homogenous or/and heterogeneous PV cells or PV panels connected in series
or/and parallel. An evaluation study of the accuracy of these three models will be carried
out in order to obtain the best possible results.

79
5.2 Modeling the PV module

Equation 5.1 [9] represents the I-V characteristics of the single-diode model of a PV cell.

( )– (5.1)

Where (5.2)

The I-V curve that represents equation 5.1 is shown in figure 5.1 in which three
significant points are highlighted: the open circuit voltage (Voc), the short circuit current
(Isc), and the maximum power point (Pmax). In practical PV cells the generated light
current depends upon the solar radiation and the temperature according to equation 5.3
[9].

( ) (5.3)
where (5.4)

The second parameter in the I-V equation is the diode saturation current that is
represented by equation 5.5. For simplification purposes equation 5.5 can be improved to
equation 5.6 [9] showing the dependence of the saturation current on the temperature in a
different manner. This alteration aims to match the open circuit voltage of the model with
the available parameters in the datasheet, namely, KI and KV, for a wide range of
temperatures.

( ) ( ) (5.5)
(5.6)
( )

The third and fourth parameters to be considered are the series and parallel resistances.
The modeling method used here chooses only one pair (Rs,Rsh) that makes the peak
output power available in the datasheet equal to the peak output power of the proposed
model. The relationship between Rs and Rsh can be obtained by setting Pmax,e = Pmax,m as
shown in equation 5.7 [9]. The resultant relationship is expressed in equation 5.8 [9] that
shows a great dependence between the two resistances.

80
 MPP
Isc
Current (A)

Voltage (V) Voc

Fig. 5.1 The I-V curve with the three remarkable points

81
 { [ ( ) ] } (5.7)

( ) { [ ] } (5.8)

Rs and Rsh can be obtained using an iterative method that assumes an initial value of Rs,
namely, zero, and incrementing it until the maximum power of the model matches the
experimental maximum power. The initial value of Rsh is given in equation 5.9 [9].

(5.9)

As for the fifth and last parameter, the diode ideality factor (n), a set of n values have
been assumed ranging from 1 to 2 with an increment of 0.1. This represents the usual
assumption for n [10].

The developed model can be further improved taking into consideration the effect of Rs
and Rsh on the light current as shown in equation 5.10 [9]. This shows a significant
reliance where the light current has a different value from the short circuit current.

(5.10)

The I-V curve can be obtained using Newton-Raphson method for solving non-linear
systems. The I-V curve is obtained by numerically solving g(V,I) = I-f(V,I)=0 for a set of
V values obtaining the corresponding set of I values. The algorithm is shown in figure
5.2.This modeling method was applied to three modules namely, the KC200GT, the
MSX60, and the H85 (the experimental data of this module was obtained in the
university) PV modules whose datasheets are listed in appendix D. The three-point
adjusted I-V curves for the three modules are shown in figures 5.3, 5.4, and 5.5
respectively. The MATLAB codes used to plot these curves are listed in appendixes D.1,
D.2, and D.3. The simulation was based on comparing the absolute error |I,m-I,e| for the
purpose of matching the calculated current with the experimental value. The obtained
parameters for each module of the proposed model are listed in tables 5.1, 5.2, and 5.3
achieving the least error. The objective of this method, to fit the I-V equation to the three
distinct points, was successfully achieved.

82
 Input :T , G
I0,eq(5.6)
Rs=0
Rsh = Rsh,min ,eq(5.9)

NO Epmax >
tolerance

YES

IL,n , eq(5.10)
IL and Isc , eq (5.3)
Rsh , eq(5.8)
solve eq (5.1) for 0 ≤ V ≤ Vocn
Calculate P for 0 ≤ V ≤ Vocn

find Pmax

Epmax = || pmax – pmax,e ||

Increment Rs

Fig. 5.2 The modeling algorithm for the single-diode model

83
 9

7 (26.3,7.61)
Current (A)

6 (0,8.197)

1 (32.82,0)

0
0 5 10 15 20 25 30 35
Voltage (V)

Fig. 5.3 The I-V curve adjusted to three points for the KC200GT PV module

3.5
(17.14,3.492)
3
(0,3.788)
Current (A)

2.5

1.5

1
(21.02,0)
0.5

0
0 5 10 15 20
Voltage (V)

Fig. 5.4 The I-V curve adjusted to three points for the MSX60 PV module

5
(0,5.188) (18,4.85)
4
Current (A)

1 (20.6,0)

0
0 5 10 15 20

Voltage (V)

Fig. 5.5 The I-V curve adjusted to three points for the H85 PV module

84
 n=1
Ion 4.12801 10-10

Rsh 93.856511

Rs 0.276600

IL 8.234202

Table 5.1 Obtained parameters from the proposed model for the KC200GT PV module

n=1
Ion 4.70398 10-10

Rsh 160.602702

Rs 0.357000

IL 3.808481

Table 5.2 Obtained parameters from the proposed model for the MSX60 PV module

n=1
Ion 8.88611 10-10

Rsh 206.412004

Rs 0.001000

IL 5.190000

Table 5.3 Obtained parameters from the proposed model for the H85 PV module

85
5.3 Accuracy Evaluation of the Proposed Model
The accuracy of the proposed single-diode model is evaluated by comparing the
experimental data with the measured data obtained from the modeling algorithm. The
error evaluation was done for the PV output current and was applied to the KC200GT,
the MSX60, and the H85 PV modules. The error curves of the proposed model for the
three modules are shown in figures 5.6, 5.7, and 5.8 respectively and were obtained from
the MATLAB codes listed in appendixes D.1, D.2 and D.3 as mentioned previously.
These three curves display a wide range of variations in the error for different voltage
values. Consequently, a method that would reduce such error variations should be
investigated in order to obtain and achieve accurate results.

5.4 Accuracy Improvement of the Proposed Model

The accuracy of the proposed single-diode model can be further improved by finding the
five unknown parameters (IL, Io, n, Rs, Rsh) that are not available in the typical PV
datasheets using five points from the I-V curve instead of only three points (remarkable
points). The specific five points on the I-V curve are the three remarkable points (Voc, Isc,
Pmax) and an additional two points (Vx1, Ix1) where Vx1=0.5*Voc and (Vx2, Ix2) where
Vx2=0.5*(Voc+Vm) [11]. These points are listed in appendix D.4. This evaluation aims at
improving the accuracy of the obtained results based on five nonlinear equations 5.11,
5.12, 5.13, 5.14, 5.15 that are solved in the nominal condition in order to find the targeted
parameters.

(5.11)

(5.12)

(5.13)

(5.14)

(5.15)

86
 1
Absolute error (A)

0.8

0.6

0.4

0.2

0
0 5 10 15 20 25 30 35
Voltage (V)

Fig. 5.6 The error curve of the proposed model for the KC200GT PV module

0.5
Absolute error (A)

0.4

0.3

0.2

0.1

00
2 4 6 8 10 12 14 16 18 20 22
Voltage (V)

Fig. 5.7 The error curve of the proposed model for the MSX60 PV module

1.8
Absolute error (A)

1.6

1.4

1.2

0.8

0.6

0.4

0.2

00
5 10 15 20
Voltage (V)

Fig. 5.8 The error curve of the proposed model for the H85 PV module

87
These equations can be solved using the quasi-Newton method. At first, the main five
points in the I-V curve are extracted as listed in appendix D.4 [11] . Then, these points
are used as input parameters for the five nonlinear equations that are solved using
MATLAB. Finally, the five unknown parameters are determined. This process is shown
in figure 5.9. The MATLAB code that solves these nonlinear equations is listed in
appendix D.5. The new obtained parameters are used in the previous MATLAB code
listed in appendix D.1,D.2 and D.3 that adjusts the I-V curve to five points instead of
only three. We also applied this method to five points listed in appendix D.6 that are
different than the ones mentioned earlier depending on our observation of the error
curves in order to obtain more accurate results than the ones obtained from the earlier
five points. These five points are the three remarkable points and another two points in
which a large error occurs. In order to compare the previous results obtained in section
5.2 with the new results, three error curves are plotted for the new five-point adjusted
curves and compared with the error curve for the three-point adjusted curve. Figures 5.10,
5.11, and 5.12 show this comparison for the three PV modules. It can be seen that a great
reduction in the error variation is obtained from applying this method. The last approach
achieved more accurate results than the ones obtained earlier [11] since the additional
two points were chosen according to where a large error occurs. As a verification of our
results, the parameters obtained from adjusting the I-V curve to five points in which a
large error occurs, which are shown in tables 5.4, 5.5, and 5.6 for each module, are used
in a model developed in PSpice to obtain the I-V curves that are shown in figures 5.13,
5.14, and 5.15. The PSpice files simulations are listed in appendix D.7.

88
 Extracting the Specified Five Points on the I-V Curve

Voc
Isc
Pmax
(Vx1,Ix1)
(Vx2,Ix2)

Solving Five Nonlinear Equations

(5.11)
(5.12)
(5.13)
(5.14)
(5.15)

Finding the Five Unknown Parameters of the Proposed Model

IL
Io
n
Rs
RSH

Fig. 5.9 The model improvement process

89
 1 3-point adjusted error curve
Absolute error (A)

5-point adjusted error curve (previous work) [11]

0.8
5-point adjusted error curve
0.6

0.4

0.2

0
0 5 10 15 20 25 30 35
Voltage (V)
Fig. 5.10 Three error curves of the proposed model for the KC200GT PV module

3-point adjusted error curve

0.5
Absolute error (A)

5-point adjusted error curve

0.4 (previous work) [11]
5-point adjusted error curve
0.3

0.2

0.1

00
2 4 6 8 10 12 14 16 18 20 22
Voltage (V)

Fig. 5.11 Three error curves of the proposed model for the MSX60 PV module

1.8 3-point adjusted error curve

Absolute error (A)

1.6
5-point adjusted error curve
1.4 (previous work) [11]

1.2 5-point adjusted error curve

0.8

0.6

0.4

0.2

00
5 10 15 20
Voltage (V)

Fig. 5.12 Three error curves of the proposed model for the H85 PV module

90
 n=1.2450
Ion 4.20198 10-8

Rsh 90.118899

Rs 0.132835

IL 8.222102

Table 5.4 Obtained parameters from adjusting the proposed model to five points for the KC200GT PV module

n=1.4138
Ion 3.7278 10-7

Rsh 27067.846214

Rs 0.194933

IL 3.799962

Table 5.5 Obtained parameters from adjusting the proposed model to five points for the MSX60 PV module

n= 1
Ion 9.085 10-10

Rsh 209.912002

Rs 0.000154

IL 5.189979

Table 5.6 Obtained parameters from adjusting the proposed model to five points for the H85 PV module

91
 9

7.5
Current (A)

2.5

0
0 5 10 15 20 25 30 35
Voltage (V)

Fig. 5.13 I-V curve of the five-point adjusted curve of the proposed model for the KC200GT module using PSpice
4

3
Current (A)

0
0 2 4 6 8 10 12 14 16 18 20 22
Voltage (V)

Fig. 5.14 I-V curve of the five-point adjusted curve of the proposed model for the MSX60 module using PSpice

5
Current (A)

0
0 5 10 15 20 25
Voltage (V)

Fig. 5.15 I-V curve of the five-point adjusted curve of the proposed model for the H85 module using PSpice

92
6
MATLAB Modeling and Simulation of
Photovoltaic Modules: the Two-Diode and
Three-Diode Models

6.1 Introduction
The generalized model of a non-ideal solar cell comprises of two diodes. The first diode
describes the simple p-n junction behavior without electrons and holes generation and
recombination in the space charge region. The second diode that is connected in parallel
with the first diode describes the "non-ideality" of the real p-n junction due to the
recombination that occurs at the space charge region. Therefore, it is critical to take into
consideration the non-idealities of a solar cell by means of modeling and simulation. This
analysis can be further extended by taking into account the non-ideality that comes from
the leakage current. The same process implemented in chapter 5 will be implemented
here for comparison purposes.

6.2 Modeling the PV module: the Two-Diode model

Equation 6.1 [12] represents the I-V characteristics of the two-diode model of the PV
cell.

( ) ( )– (6.1)

Equation 6.1 is represented by figure 5.1 which highlights the three remarkable points
mentioned in chapter 5. In practical PV cells the light generated current depends upon the
solar radiation and the temperature according to equation 5.3.

93
The second parameter in the I-V equation is the diode saturation current that is
represented by equation 5.5 and simplified to equation 5.6 as discussed in chapter 5. For
further simplification purposes, the saturation currents of both the first and second diodes,
and , are set to be equal. This is clarified mathematically in equation 6.2 [12].

( )
(6.2)
[ ]
( )

As for the diode ideality factor (n), according to the Shockley’s diffusion theory, the
value of n1 must be unity [10]. In addition, the value of n2 should be greater than or equal
to 1.2 (>=1.2) to obtain the best possible match between the proposed model and the
experimental model as extensive simulations showed. Since the term ( ) in
equation 6.2 should be equal to 1 and must be equal to 1 as mentioned earlier, the
variable can be selected to be greater than or equal to 2.2 [12]. As a consequence, the
mathematical expression for and is given as in equation 6.3 [12].

( )
(6.3)
[ ]

As for the series and shunt resistances, the modeling method used here chooses only one
pair (Rs,Rsh) that makes the peak output power available in the datasheet equal to the
peak output power of the proposed model. The relationship between Rs and Rsh can be
obtained by setting Pmax,e = Pmax,m as mentioned in chapter 5. The resultant expression for
the shunt resistance is shown in equation 6.4 [12].

( )
(6.4)

Where [ ( ) ] (6.5)

And [ ( ) ] (6.6)

94
Rsand Rsh can be obtained using an iterative method that assumes an initial value of Rs,
namely, zero, and incrementing it until the maximum power of the model matches the
experimental value. The initial value of Rsh is given in equation 5.9 as discussed in
chapter 5.

As mentioned in chapter 5, the I-V curve can be obtained using Newton-Raphson

method. The algorithm used is similar to that shown in figure 5.2. This modeling method
was applied to the KC200GT, the MSX60, and the H85 PV modules used in chapter 5.
The I-V curves for each module are shown in figures 6.1, 6.2, and 6.3 respectively. The
MATLAB code used to plot these curves is listed in appendixes E.1, E.2 and E.3. The
simulation was done based on the modeling done in chapter 5 for each module for the
purpose of obtaining the least error. Tables 6.1, 6.2, and 6.3 show the obtained
parameters from the proposed two-diode model for each module. The objective of this
method was successfully achieved by fitting the I-V equation to the three remarkable
points.

6.3 Accuracy Evaluation of the Proposed Model

As discussed in chapter 5, the accuracy of the proposed two-diode model can be
evaluated by comparing the experimental data with the measured data obtained from the
modeling algorithm. The error evaluation was done for the PV generated output current
and was applied to the three modules. The error curves of the proposed two-diode model
for each module are shown in figures 6.4, 6.5, and 6.6 and were obtained from the
MATLAB code is listed in appendixes E.1, E.2 and E.3.

95
 9

7
Current (A)

6 (0,8.177) (26.3,7.61)
5

2
(32.77,0)
1

0
0 5 10 15 20 25 30

Voltage (V)
Fig. 6.1 The I-V curve adjusted to three points for the KC200GT PV module

3.5
(17.14,3.492)
3 (0,3.788)
Current (A)

2.5

1.5

1
(21.02,0)
0.5

0
0 5 10 15 20
Voltage (V)

Fig. 6.2 The I-V curve adjusted to three points for the MSX60 PV module

(18,4.85)
(0,5.188)
Current (A)

1
(20.6,0)

0
0 2 4 6 8 10 12 14 16 18 20

Voltage (V)

Fig. 6.3 The I-V curve adjusted to three points for the H85 PV module

96
 n1=1
n2=1.2
Ion 4.128010e-010

Rsh 108.672034

Rs 0.280000

IL 8.210000

Table 6.1 Obtained parameters from the proposed model for the KC200GT PV module

n1=1
n2=1.2
Ion 4.703980 10-10

Rsh 176.392748

Rs 0.350000

IL 3.800000

Table 6.2 Obtained parameters from the proposed model for the MSX60 PV module

n1=1
n2=1.2
Ion 8.886108e-010

Rsh 232.707561

Rs 0.001000

IL 5.190000

Table 6.3 Obtained parameters from the proposed model for the H85 PV module

97
 1
0.9
Absolute error (A) 0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 5 10 15 20 25 30 35
Voltage (V)
Fig. 6.4 The error curve of the proposed model for the KC200GT PV module

0.5
Absolute error (A)

0.4

0.3

0.2

0.1

0
0 2 4 6 8 10 12 14 16 18 20 22

Voltage (V)

Fig. 6.5 The error curve of the proposed model for the MSX60 PV module

1.5
Absolute error (A)

0.5

0
0 2 4 6 8 10 12 14 16 18 20 22
Voltage (V)
Fig. 6.6 The error curve of the proposed model for the H85 PV module

98
6.4 Accuracy Improvement of the Proposed Model
The accuracy of the proposed two-diode model can be further improved by following the
same procedure discussed in chapter 5 through finding the seven unknown parameters
(IL, Io1, Io2, n1, n2, Rs, Rsh) using seven points from the I-V curve instead of only three
points (remarkable points) as mentioned in section 5.2. The specific seven points on the
I-V curve are the three remarkable points (Voc, Isc, Pmax) and an additional four points
(Vx1, Ix1) where Vx1=Vm/3, (Vx2, Ix2) where Vx2=(2*Vm)/3, (Vx3, Ix3) where
Vx3=(Voc+2*Vm)/3, and (Vx4, Ix4) where Vx4=(2*Voc+Vm)/3 [11]. These points are listed
in appendix E.5. The seven nonlinear equations used for accuracy improvement are 6.7,
6.8, 6.9, 6.10, 6.11, 6.12, and 6.13.

(6.7)

(6.8)

(6.9)

(6.10)

(6.11)

(6.12)

(6.13)

99
The process used to improve the accuracy of the proposed model is similar to that shown
in figure 5.9. The MATLAB code that solves these nonlinear equations is listed in
appendix E.5. The new obtained parameters are used in the previous MATLAB code that
adjusts the I-V curve to seven points instead of only three which are listed in appendix
E.4 [11] . We also applied this method to seven points listed in appendix E.6 depending
on our observation of the error curves to achieve more accurate results. The two error
curves for the new seven-point adjusted curve compared with the error curve for the
three-point adjusted curve are shown in figures 6.7, 6.8, and 6.9 for the three modules. A
great reduction in the error variation is obtained from applying this method to the
proposed two-diode model. To verify our results for each module, the parameters
obtained from adjusting the I-V curve to seven points in which a large error occurs,
which are shown in tables 6.4, 6.5, and 6.6, are used in a model developed in PSpice
listed in appendix E.7 to obtain the I-V curves that are shown in figures 6.10, 6.11, 6.12.

100
 1
0.9
Absolute error (A)

3-point adjusted error curve

0.8
0.7 7-point adjusted error curve (previous work) [11]

0.6
7-point adjusted error curve
0.5
0.4
0.3
0.2
0.1
0
0 5 10 15 20 25 30 35
Voltage (V)
Fig. 6.7 Three error curves of the proposed model for the KC200GT PV module

3-point adjusted error curve

0.5
7-point adjusted error curve (previous work) [11]
Absolute error (A)

0.4 7-point adjusted error curve

0.3

0.2

0.1

0
0 2 4 6 8 10 12 14 16 18 20 22

Voltage (V)

Fig. 6.8 Three error curves of the proposed model for the MSX60 PV module
1.5

3-point adjusted error curve

Absolute error (A)

7-point adjusted error curve (previous work) [11]

1
7-point adjusted error curve

0.5

0
0 2 4 6 8 10 12 14 16 18 20 22
Voltage (V)

Fig. 6.9 Three error curves of the proposed model for the H85 PV module

101
 n1=1.4363
n2=1.4363
Ion1 2.750103 10-7

Ion2 2.344898 10-7

Rsh 108.600019

Rs 0.141532

IL 8.218009

Table 6.4 Obtained parameters from adjusting the proposed model to seven points for the KC200GT PV module

n1=0.9122
n2=1.2971
Ion1 3.570000 10-11

Ion2 2.499580 10-8

Rsh 232.850539

Rs 0.336428

IL 3.799735

Table 6.5 Obtained parameters from adjusting the proposed model to seven points for the MSX60 PV module

n1=1
n2= 1.1990
Ion1 6.385 10-10

Ion2 8.7442 10-9

Rsh 232.707561

Rs 0.001000

IL 5.190000

Table 6.6 Obtained parameters from adjusting the proposed model to seven points for the H85 PV module

102
 9

Current (A) 7.5

5.0

2.5

0
0 5 10 15 20 25 30 35

Voltage (V)

Fig. 6.10 The I-V curve of the seven-point adjusted curve of the proposed model for the
KC200GT module using PSpice

4.0

3.0
current (A)

2.0

1.0

0
0 2 4 6 8 10 12 14 16 18 20 22 24

Voltage (V)

Fig. 6.11 The I-V curve of the seven-point adjusted curve of the proposed model for the MSX60
module using PSpice

5
Current (A)

0
0 2 4 6 8 10 12 14 16 18 20 22
Voltage (V)
Fig. 6.12 The I-V curve of the seven-point adjusted curve of the proposed model for the H85
module using PSpice

103
6.5 Modeling and Simulation of the PV module: the Three-
Diode model
The modeling of the three-diode model of a PV module can be done following the same
procedure mentioned in section 6.1. In addition, the saturation current of the third diode
can be set to be equal to the saturation currents of the two diodes. As for the ideality
factor of the third diode, the simulation was done for several values of n3 for the sake of
achieving the least error. The I-V curves adjusted to three points are shown in figure 6.13,
6.14, and 6.15 for the three modules and the corresponding parameters obtained are
shown in tables 6.7, 6.8, and 6.9. As done in section 6.3, the error curves of the proposed
three-diode model for each module are shown in figures 6.16, 6.17, and 6.18. The points
chosen to improve the accuracy of the model are listed in appendix E.8. Similarly, the
two error curves that were discussed earlier are shown in figures 6.19, 6.20, and 6.21 for
each module. The parameters obtained from the improved three-diode model are shown
in tables 6.10, 6.11, and 6.12 and used as input parameters in the PSpice model listed in
appendix E.9 for verification purposes as illustrated in figures 6.22, 6.23, and 6.24.

104
 9

7 (0,8.197) (26.3,7.61)
Current (A)

2
(32.77,0)
1

0
0 5 10 15 20 25 30 35
Voltage (V)
Fig. 6.13 The I-V curve adjusted to three points for the KC200GT PV module

3.5
(17.14,3.492)
3
(0,3.788)
Current (A)

2.5

1.5

1
(21.02,0)
0.5

0
0 5 10 15 20
Voltage (V)

Fig. 6.14 The I-V curve adjusted to three points for the MSX60 PV module

5
(0,5.188) (18,4.85)
4
Current (A)

1 (20.6,0)

0
0 5 10 15 20

Voltage (V)

Fig. 6.15 The I-V curve adjusted to three points for the H85 PV module

105
 n1=1
n2=1.2
n3=1.2
Ion 4.128010 10-10

Rsh 94.811428

Rs 0.266400

IL 8.233074

Table 6.7 Obtained parameters from the proposed model for the KC200GT PV module

n1=1
n2=1.2
n3=1.2
Ion 4.703980 10-10

Rsh 162.612957

Rs 0.339800

IL 3.807944

Table 6.8 Obtained parameters from the proposed model for the MSX60 PV module

n1=1
n2=1.2
n3=1.2
Ion 8.886108 10-10

Rsh 261.703349

Rs 0.000100

IL 5.190000

Table 6.9 Obtained parameters from the proposed model for the H85 PV module

106
 1
0.9
Absolute error (A) 0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 5 10 15 20 25 30 35
Voltage (V)
Fig. 6.16 The error curve of the proposed model for the KC200GT PV module

0.6

0.5
Absolute error (A)

0.4

0.3

0.2

0.1

0
0 2 4 6 8 10 12 14 16 18 20 22

Voltage (V)

Fig. 6.17 The error curve of the proposed model for the MSX60 PV module

1.5
Absolute error (A)

0.5

0
0 2 4 6 8 10 12 14 16 18 20 22
Voltage (V)

Fig. 6.18 The error curve of the proposed model for the H85 PV module

107
 0.9

Absolute error (A)

0.8 3-point adjusted error curve

0.7 9-point adjusted error curve

0.6
0.5
0.4
0.3
0.2
0.1
0
0 5 10 15 20 25 30 35

Voltage (V)
Fig. 6.19 Two error curves of the proposed model for the KC200GT PV module

0.6
3-point adjusted error curve

0.5 9-point adjusted error curve

Absolute error (A)

0.4

0.3

0.2

0.1

0
0 2 4 6 8 10 12 14 16 18 20 22

Voltage (V)
Fig. 6.20 Two error curves of the proposed model for the MSX60 PV module

1.5

3-point adjusted error curve

9-point adjusted error curve

Absolute error (A)

0.5

0
0 2 4 6 8 10 12 14 16 18 20 22
Voltage (V)

Fig. 6.21 Two error curves of the proposed model for the H85 PV module

108
 n1=0.9658
n2=1.0451
n3=0.9370
Io1 5.274300 10-11

Io2 4.392100 10-10

Io3 4.654500 10-11

Rsh 94.291213

Rs 0.274100

IL 8.233872

Table 6.10 Obtained parameters from adjusting the proposed model to seven points for the KC200GT PV module

n1=1.0812
n2=1.1893
n3=1.0812
Io1 1.309100e-009
Io2 7.931000e-010
Io3 1.035200e-009
Rsh 162.967070
Rs 0.327600
IL 3.808464

Table 6.11 Obtained parameters from adjusting the proposed model to seven points for the MSX60 PV module

n1= 1.1028
n2= 1.1514
n3= 1.1514
Io1 5.348000 10-9
Io2 3.241000 10-10
Io3 3.170300 10-9
Rsh 261.7
Rs 0.000100
IL 5.190000

Table 6.12 Obtained parameters from adjusting the proposed model to seven points for the H85 PV module

109
 9

Current (A) 7.5

2.5

0
0 5 10 15 20 25 30 35

Voltage (V)

Fig. 6.22 I-V curve of the nine-point adjusted curve of the proposed model for the KC200GT module using PSpice

4.0

3.0
current (A)

2.0

1.0

0
0 2 4 6 8 10 12 14 16 18 20 22 24

Voltage (V)

Fig. 6.23 I-V curve of the nine-point adjusted curve of the proposed model for the MSX60 module using PSpice

5
Current (A)

0
0 5 10 15 20 25
Voltage (V)

Fig. 6.24 I-V curve of the nine-point adjusted curve of the proposed model for the H85 module using PSpice

110
7
Modeling and Simulation of Grid-Connected
Photovoltaic Systems

7.1 Introduction
A photovoltaic system is a system that uses one or more PV modules to convert sunlight
into electricity. It comprises of multiple components, including the photovoltaic modules,
the mechanical and electrical connections, and means of regulating and/or modifying the
electrical output. Several applications are concerned with these types of systems some of
which are standalone systems, grid connected systems, and building systems. In the last
few years an important growth of grid-connected PV systems has been observed,
especially in industrialized countries. A grid connected system is a system that is
connected to an electrical grid and feeds power into the grid. They vary in size from
residential to solar power stations. The advantages of these systems lie in its ability to
reduce the power bill, easy to install due to no battery system is required and effective
utilization of generated power as there are no storage losses elaborated. This chapter will
discuss the overall PV grid-connected system, some technical considerations, the
modeling of each subsystem in the entire system, and the energy balance concept.

7.2 System Description [6]

Solar electricity is fed directly to a utility power grid through systems called grid-
connected PV systems. The main subsystems that make up a grid-connected PV system
are: a PV generator consisting of several PV modules connected together forming an
array that converts solar energy into DC electricity, an inverter, also known as a power
conditioning unit (PCU) that converts DC into AC electricity, power balance including

111
wiring and mounting structure, surge and ground fault protection and metering
components interconnected to the AC grid, and AC loads. The DC electricity obtained
from the PV generator is converted to AC electricity by the inverter, which is supplied to
the AC loads directly or fed to the power utility grid in the case of having excess power.
During dark hours, the power supplied to the AC loads comes from the utility grid that
works as an energy storage element saving excess power supplied by the PV generator.
The role played by batteries in standalone PV applications is replaced, in grid-connected
systems, by the utility grid itself acting as an energy storage element for the system,
resulting in cost and maintenance reduction. This entire system is shown in figure 7.1.
This system may vary from country to country depending upon different national
regulations including safety and protection circuitry. The schematic shown in figure 7.1
shows the use of fuses and blocking diodes as protection devices. It also shows surge
protection devices that are connected to the input and output of the inverter. In addition,
utility safety devices and guard relays are placed between the inverter output and the
switching point with the utility grid. Grid-connected PV systems can be classified
according to size into three main categories: small size, having a power ranging from 1 to
10 kW, medium size, having a power ranging from 10kW to several hundreds of kW, and
large size, having a power ranging from 500 kW to MW range. Each category is used for
certain applications that correspond to the amount power required.

7.3 Technical Considerations [6]

Protection systems are required to prevent damage to the PV system and also to avoid the
reduction of the quality of the grid electricity. Therefore, some important issues
associated with the connection of the PV system to the utility grid are discussed in this
section.

7.3.1 Islanding protection

The islanding operation mode can be defined as the continued operation of a grid-
connected inverter when the utility grid, or a component of it, has been switched off or
there is no electric energy that can be delivered from the utility system. Islanding can

112
 impp

vmpp
Grid

Input irradiance
profile Vgrid

PV Generator
Temperature
profile

Fig. 7.1 Schematic of the grid-connected system.

113
strongly influence the equipment and loads connected to the network, and can cause
electric shocks to consumers or utility grid workers. Consequently, inverters must
recognize a grid fault or disconnection and must immediately disconnect its output itself.
Basically, inverters detect the islanding condition by looking for some combination of the
following: an unexpected change in system frequency, an abrupt change in voltage
magnitude, a sudden change in the rate of change of frequency, and a sudden change in
the active output power (kW) and reactive output power (kVAR) beyond the expected
normal level. Nowadays most commercial inverters include acceptable islanding
prevention capabilities obtained by a combination of different control algorithms.

7.3.2 Voltage disturbances

A suitable voltage level must be maintained at the client's input connection to the grid.
Different limits for the voltage levels have been recognized in different countries. The
fact that grid-connected PV systems can contribute to disturbances, line voltage
distinction has to be taken into account. The voltage variations at the connecting point to
the grid are sensed through inverters. These inverters also include the capability of
disconnection and automatic reconnection after confirmation of utility immovability
recovery.

7.3.3 Frequency disturbances

Inverters usually provide internal over/under line voltage frequency shutdown. Internal
shutdown should be produced if, within a few cycles, the frequency falls outside the
prearranged limitations.

7.3.4 Disconnection
A switch (separator) is normally included in grid-connected PV systems to provide
safety, especially for the utility workers. Taking into account the protections that
inverters must implement, this main disconnection switch can be considered redundant,
however, personal safety is the most essential issue concerning grid-connected PV
systems.

114
7.3.5 Reconnection after grid failure
The correct operation of the grid for a sensible interval of time before reconnecting the
inverter must be taken into account. This is implemented in most commercial inverters by
means of grid sensing and auto-reconnection. Nearly all inverters are able to reconnect
themselves to the grid after observation of a number of grid cycles with proper values of
voltage, amplitude, and frequency.

7.3.6 DC injection into the grid

Some inverter designs include transformers inside the inverter, repressing any DC
injection into the grid. The Advantages of keeping DC electricity out of the grid are to
improve the personal safety, to protect the disturbances on the utility grid and saturation
effects in local distribution transformers, and to avoid the saturation on inductive loads.
However, the number of transformer-less inverters in the market today, has been growing
because of technical and cost advantages. In most countries isolation transformers are not
required for small PV grid-connected systems, but they are necessary for medium or large
size PV systems.

7.3.7 Grounding
Grounding is necessary to conduct current when a ground fault happens. This will reduce
electrical shock hazards, fire hazards, and damage to the loads and system equipment.
The components of a PV system must be grounded in accordance with the appropriate
national regulations of each country.

7.3.8 EMI
The components of a PV system, particularly inverters, are subjected to varying in high
frequency noise emission/immunity requirements that bound the tolerable radiation
spectrum for a range of frequencies, usually between 150 kHz and 30 MHz.

115
7.4 Subsystem Description [6]

In this section, each subsystem in the entire grid-connected system will be discussed.

7.4.1 PV Generator
In general PV generators, PV arrays, are formed by combinations of series and parallel
connections of solar cells or PV modules. In terrestrial applications arrays are formed by
connecting PV modules containing a certain number of series-connected solar cells. The
PV generator used in the overall grid-connected system will be the behavioral model and
hence the behavioral solar cell model must be discussed.

7.4.1.1 Behavioral Solar Cell Model

Solar cells that are not made of famous semiconductor materials such as Si, GaAs and
Ge, are difficult to study their behaviors for changing temperature and irradiances as the
required data to model the effects of temperature is not easily accessible. Therefore, the
typical solar cell model that includes a diode is not very practical and is restricted in
terms of use. As a result, a behavioral model of a solar cell is introduced. The behavioral
model presumes that the solar cell can be modeled by a series resistance and two current
sources. The first current source is expressed as follows [6]:

( ) ( )( ) (7.1)

( )
( ) ( ) (7.2)
( )

The cell operating temperature is derived from the NOCT (nominal operating
conditions temperature) concept at 800 W/m2 irradiance and 20°C of ambient
temperature. Furthermore, is the reference temperature which is usually considered
25°C [6].

(7.3)

116
The open circuit voltage is approximately given as [6]:

( )( ) (7.4)

Where is the voltage under arbitrary radiance and temperature conditions.

is the current under arbitrary radiance and temperature conditions.

The series resistance can be calculated as follows [6]:

(7.5)

The current of the maximum power point can be expressed as follows [6]:

( ) ( ) (7.6)
where is the Arbitrary radiance.

Then, the voltage of maximum power point is written as follows [6]:

( ( )) (7.7)

This model can be implemented in PSpice using equations 7.1, 7.2, 7.3, 7.4, 7.5, 7.6 and
7.7 as listed in appendix F.1. The schematic of the solar cell behavioral model and block
diagram are shown in figure 7.2.

7.4.2 Inverter
An inverter can be defined as a power electronic device that transforms a DC signal into
an AC signal. AC power is necessary in PV systems because most loads are AC loads. In
addition, many PV systems are connected to the grid and hence the output energy of the
complete system has to be AC. Thus, the inverter is the key component of the overall
system. Some of the most important characteristics of inverters are as follows: the voltage
range, which is defined to be the range of DC input voltage values satisfactory for the
nominal operation of the inverter, and the output power that is known as the nominal and
maximum values of AC output power that can be supplied by the inverter to the loads.
Some important definitions must be addressed regarding inverters as discussed in the
following sections.

117
 R
(201)
(202) (204)
G_irrad eisc (205)
(203) (206)
evoc
(207)
evtemp
G_irrad (208)
gim
evm
(209)

(200)

Irradiance Cell output

(202) (204)
Ambient Short Circuit
Temperature Current
(203) (205)
Open Circuit
Voltage
Subcircuit (206)
cell_behavioural.lib Cell
Temperature
(207)
MPP current
(208)

MPP voltage
(209)
(200)

Fig. 7.2 (a) Schematic of the solar cell behavioral model.

(b) Block diagram of the solar cell behavioral model.

118
7.4.2.1 Total Harmonic Distortion
Total harmonic distortion (THD) can be defined as the ratio of the root mean square
(RMS) value of the signal composed of all harmonics excluding the first, to the RMS
value of the first harmonic as represented in equation 7.8.

√∑
(7.8)

Where V1 is the RMS value of the first harmonic,

Vn is the RMS value of the nth harmonic.

7.4.2.2 Inverter Efficiency

Inverter efficiency can be defined as the ratio of the RMS value of the output power to
the value of the DC input power as shown in equation 7.9.

(7.9)

Inverters have to meet very high standards to set narrow limits for tolerable voltage and
frequency deviations from the nominal. The majority of the existing inverters in the
market exhibit levels of total harmonic distortion below 3% and efficiencies ranging from
90% to 96%. Inverters are available with different characteristics, properties,
performances, and can be classified according to the inverter output, either single-phase
or three-phase inverter, and its power rating. A more common inverter classification can
be made according to the internal topology. Thus, inverters can be classified into three
main categories as follows: Pulse-width modulated (PWM) inverters, which have the best
performances in terms of efficiency (90-97%) and harmonic distortion (below 3%),
square wave inverters that include solid-state devices such as power bipolar transistors or
MOSFETs working as switches to produce a square wave at the output of the inverter
from a single DC source at the input, and modified sine wave inverters , which have an
efficiencies ranging from (70–80%) and THD levels are (around 5%).

119
7.5 PSpice Modeling of the Overall Grid-Connected PV
Systems [6]

7.5.1 Modeling of Inverters

In grid connected PV systems, commercial inverters are used which are basically
modeled as dependent current sources. Some of the design considerations in standalone
inverter models can be neglected here, particularly the power control implemented in
these models which has to be changed for inverters connected to the grid. The inverter
model includes the capability of the maximum power point tracking, MPPT, to attain the
maximum power from the PV array. For simplification purposes, no restrictions have
been considered on the maximum power tracking range and minimum inverter input DC
voltage in this model. The simplified proposed model of the inverter for grid-connected
PV systems is show in figure 7.3 and the PSpice netlist that implements this connectivity
is listed in Appendix F.2. Figure 7.4 illustrates the schematic of the equivalent circuit of
the inverter model. As can be seen from the two figures, the inverter has two inputs,
namely nodes 4 and 5 which represent the values of the maximum power point
coordinates of the PV generator. At these inverter inputs, two resistances are connected.
The inverter input, node 1, forces the working point of the PV generator to be at the
maximum power point, MPP, which is accomplished by the connection of the controlled
voltage source (empp) that replicates the value of the MPP voltage at the irradiance and
temperature values at a given time. A correction of 0.8V is considered to take into
account the voltage drop of a blocking diode that is connected between the PV generator
output and the inverter input for safety purposes. In order to calculate the amplitude of
the sinusoidal output current, the controlled voltage source (eVc2) is used taking into
consideration the inverter efficiency (nf), the DC power at the inverter input, and the
amplitude of the output. The value of the amplitude is used at the controlled current
source (giout) that places the corresponding AC source at the inverter output, node 3, in
order to multiply a sinusoidal signal with 1V of amplitude and 50 Hz frequency which is
supplied by the voltage source (v2).

120
 impp
vmpp Node 4
Node 5

Node 1 Node 3
L1

Vmod

Node 2 Node 2
Neutral

Fig. 7.3 Inverter model.

5 R5

4 R4

7 3

evc2 +
1
giout
9

v2
empp

2
Fig. 7.4 Schematic of the inverter model.

121
7.5.2 Modeling of Grid-Connected PV systems

The behavioral PV generator is composed of 2 3 PV modules each having 36 solar cells

connected in series. The modules have a peak power of 85W, the MPP coordinates are:
immr=4.9 A and Vmmr =17.3 V, and the open-circuit voltage is vocmr = 21.2 V. A constant
ambient temperature of 12 C is selected and an inverter efficiency of 90%.The grid can
be modeled as an ideal sine wave voltage source, having an RMS voltage value of 220 V
and frequency of 50 Hz, connected to a parallel load formed by the series connection of a
resistor and an inductor each having values of 14.4 and 0.5 mH respectively. The
PSpice netlist for the simulation of the described PV system is included in Appendix F.3.
The PV generator has been implemented as described in section 7.4.1. The inverter model
presented in section 7.5.1 has been used for the inverter implementation and the grid has
been included as a sub circuit. Figures 7.5, 7.6, 7.7, and 7.8 show the simulation results
for the inverter output current connected to the grid point, and the AC inverter output
power respectively for two PV modules, the KC200GT and MSX60. As the PV generator
(for the MSX60) has a 272.723W peak and a 1.0606 KW (for the KC200GT) for an
irradiance value of 1000 W/m2, the AC inverter output power is around 245.589W (for
the MSX60) and 955.058 W (for the KC200GT) taking into account an inverter
efficiency of 90%. Temperature effects in the generator model also explain that the
output power is below the maximum available.

7.5.3 Modeling of AC-PV Generators

AC-PV generators used in grid-connected applications consist of a standard PV generator

and a small DC–AC inverter built into the array packaging. The major advantage of the
use of AC generators lies in their ability to eliminate the DC wiring and some of the other
components of a grid connected PV system, which results in a reduction of power losses
and system complexity. On the other hand, one of the most important disadvantages is
that each inverter used must incorporate the necessary electronic control such as MPPT,
and must also satisfy safety standards which result in a possible increased cost. Taking

122
 8.0

4.0
Current (A)

-4.0

-8.0
0s 20 40 60 80 100 120 140 160 180 200 220 240

Time (ms)

Fig. 7.5 Output current of the inverter for the KC200GT PV module.

1.5
Power (KW)

1.0

0.5

0
0 20 40 60 80 100 120 140 160 180 200 220 240

Time (ms)

Fig. 7.6 Output power of the inverter for the KC200GT PV module.

123
 2.0

1.0
Current (A)

-1.0

-2.0
0 20 40 60 80 100 120 140 160 180 200 220 240

Time (ms)

Fig. 7.7 Output current of the inverter for the MSX60 PV module.

300

200
Power (W)

100

0
0 20 40 60 80 100 120 140 160 180 200 220 240

Time (ms)

Fig. 7.8 Output power of the inverter for the MSX60 PV module.

124
into consideration the inverter model described in section 7.5.1 and the model presented
for a behavioral PV generator, these two models can be easily grouped to get the
corresponding model for an AC generator. The AC generator model is basically formed
by the PV generator directly connected to the inverter. The AC generator contains the
MPPT function, in the same way discussed for the inverter model. Figure 7.9 illustrates
the schematic of the AC generator connected to the grid. The PSpice netlist for an AC
generator model is listed in Appendix F.3. Figures 7.10, 7.11, 7.12, and 7.13 show the
AC module output current and output power for the two PV modules, the KC200GT and
MSX60 respectively under the nominal condition. The values of the maximum AC power
can be calculated from the amplitudes at the inverter output for an irradiance level of
1000 W/m2 as described in equation 7.10 which is found to be 970.086W (for the
KC200GT) and 273.054 W (for the MSX60) for a 1000 W/m2 irradiance level.

(7.10)

In order to show the relationship between the ambient temperature and irradiance level on
the output power of the AC PV generator, a study was carried out to show these
relationships as shown in figures 7.14 and 7.15.

125
 7 419

50
evc2 +
404
giout
9
empp + grid
vcl
Generator_be.lib

0
400

Acgenerator_be.lib
402

1
virrad

Fig. 7.9 Schematic of the AC generator model.

126
 8.0
Current (A)

4.0

-4.0

-8.0
0s 20 40 60 80 100 120 140 160 180 200 220 240
Time (ms)

Fig. 7.10 Output current of the AC generator for the KC200GT PV module.

1.5
Power(KW)

1.0

0.5

0
0s 20 40 60 80 100 120 140 160 180 200 220 240

Time(ms)

Fig. 7.11 Output power of the AC generator for the KC200GT PV module.

127
 2.0

1.0
Current (A)

-1.0

-2.0
0 20 40 60 80 100 120 140 160 180 200 220 240

Time (ms)

Fig. 7.12 Output current of the AC generator for the MSX60 PV module.

300

200
Power (W)

100

0
0 20 40 60 80 100 120 140 160 180 200 220 240

Time (ms)

Fig. 7.13 Output power of the AC generator for the MSX60 PV module.

128
 310

300
Power (W)

290

280

270

260

250
0 10 15 20
Temperature (°C)

Fig. 7.14 the relationship of the ambient temperature on the output power
of the AC PV generator

350

300

250
Power (W)

200

150

100

50

0
0 300 400
100 500 600 200 700 800 900 1000 1100

Irradiance (W/m2)
Fig. 7.15 the relationship of the Irradiance level on the output power of the
AC PV generator

129
7.6 Energy Balance in a PV System: PV Array Sizing Procedure [6]

The random nature of the sun’s radiation at a certain place, the orientation of the
collecting surface, and the unreliable data on the energy demand by the user urge the
need to apply the energy balance concept in a PV system in order to find the correct size
of the PV array or generator that would produce the desired amount of energy. There are
a number of reasons why the real energy generation of the PV system might not be equal
to the energy demanded. Some of these reasons are related to the efficiency of the MPP
tracker, the efficiency of the charge–discharge cycle of the battery itself, wiring losses,
and DC/AC converter efficiency to supply AC loads. The sizing method described here
considers the energy balance of the PV system taking into consideration the equivalent
peak solar hours (PSH) concept. The energy balance in a PV system is established by a
general equation stating that the energy consumed by a certain load in a given period of
time equals the energy generated by the PV system in the same period of time. This
period of time varies depending on the application that the PV system is used for. For
instance, if the application is a summer water pumping system, we are interested in the
energy balance in the summer season, whereas if the application is a year-round plant, we
are interested in the energy balance for a period of one year. Generally, two cases will be
considered here, the month having less solar radiation (worst-case design), and the
average monthly radiation value (average-case design).
The energy balance equation in a given day is as follows:

(7.11)

Where:
PmaxGr: the nominal output power of the PV generator at standard conditions.
PSH: the value of the peak solar hours (kWh/m2day).
L: the energy consumed by the load over a certain period of time.

The energy balance equation can be rewritten for the two cases described earlier in this
section. For the worst-case design, equation 7.11 becomes as shown in equation 7.12.

130
However, for the average-case design, the energy balance equation is represented by
equation 7.13.

( ) (7.12)

Where:
(PSH) min: the value of PSH in the worst month.

(̅̅̅̅̅̅) (7.13)

Where:
(̅̅̅̅̅): the average value of the 12 monthly PSH values.

The energy balance equation of the average-case design can be rewritten as shown in
equation 7.14 substituting the nominal maximum power of the PV generator by its
corresponding value.

(̅̅̅̅̅̅) (7.14)

Where:
VmMr: the voltage at the maximum power point of one PV module under standard
conditions.
ImMr: the current at the maximum power point of one PV module under standard
conditions.

Considering that a PV generator is composed of NsG series and NpG parallel PV modules,
the energy balance equation becomes:

(̅̅̅̅̅̅) (7.15)

Therefore, the basic design equation can be represented by equation 7.16.

(̅̅̅̅̅̅ )
(7.16)

131
Usually the loads in a PV system are connected to a DC voltage, namely Vcc. Therefore,
the load L can now be written as follows:

(7.17)

Where:
Ieq: the equivalent DC current drawn by the load over the whole day.

Now the design equation can be written as follows:

(̅̅̅̅̅̅) (7.18)

Here an engineering over sizing factor or ‘safety factor’ (SF) is introduced in such a way
that the general design equation can be written as follows:

(̅̅̅̅̅̅) ( ) (7.19)

Finally equation 7.19 can be split into two equations. Equation 7.20 represents the
number of PV modules connected in series in a PV generator and equation 7.21
represents the number of PV modules connected in parallel. As can be seen from the two
equations, the array safety factor (SF) has also been divided into a voltage safety factor
(VSF) and a current safety factor (CSF).

( ) (7.20)

( ) (̅̅̅̅̅̅ )
(7.21)

To comprehend this energy balancing concept, a practical example is introduced.

Considering a PV system connected to a DC load of 2000 Wh/day at 24 V in Abu Dhabi,
UAE. The solar radiation data available for Abu Dhabi [13] is shown in Table 7.1 in
Wh/m2 day. The PV modules used in this system are the MSX60 modules having the
characteristics that are provided in the datasheets at standard conditions.

132
We first compute the value of Ieq as follows:

The value of PSH is computed by calculating the average daily radiation value from
Table 7.1, taking into account that January, March, May, July, August, October and
December have 31 days each; that April, June, September and November have 30 days
each, and that February has 29 days each.

Daily radiation
Month
(Wh/m2 day)
Jan 4.3

Feb 5

Mar 5.7

Apr 6.7

May 7.8

Jun 7.6

Jul 7

Aug 6.7

Sep 6.5

Oct 5.7

Nov 4.8

Dec 4

Table 7.1 Radiation data for Abu Dhabi, UAE

133
The average value of the average monthly daily radiation data in Table 7.1 is 5.98 Wh/m2
day. Finally, the number of series and parallel modules connected in the PV generator are
computed as shown in equations. This implies that the recommended size of the PV array
in this system will be 2 modules connected in series with 4 modules connected in parallel
(2×4).

( )

( )
( )

134
8
Conclusion

8.1 Summary
In this project, Matlab/PSpice modeling of photovoltaic systems was presented. The
basics of photovoltaic principles and the history of solar cells were briefly discussed.
Moreover, PSpice simulation of different sun spectral distributions was illustrated. In
addition, a Study of Photovoltaic cell/module/array using PSpice and Simulink was
carried out. Moreover, MATLAB modeling and simulation of photovoltaic modules were
developed for the single-diode, two-diode and three-diode models of a solar system.
Finally, the modeling of a grid connected system was carried out.

8.2 Related issues

This project has distinct environmental advantages over conventional technologies for
electricity generation. In other words, the operation of photovoltaic systems does not
produce any noise or toxic-gas radiations. Photovoltaic energy can help meet the growing
global demand for electricity. In addition, this is accomplished without regarding the high
economic and environmental costs of installing power lines or burning fossil fuels [14].
Regarding the commercial issues, the simulation and modeling of photovoltaic
cells/module/arrays using PSpice and MATLAB did not cost anything since the
university offers the software programs needed. According to the Student Code of Ethics,
all preceding works used in this thesis were referred to their authors.

135
 References

[1] J. Nelson, the Physics of Solar Cells. London, U.K.: Imperial College Press, 2003.

[2] A. Geotzberger et.al, Crystalline Silicon Solar Cells. Chichester, U.K.: John Wiley &
Sons, 1998.

[3] B. V. Zeghbroek. (2011). Principles of Electronic Devices [Online]. Available:

www.ecee.colorado.edu/~bart/book/book/chapter4/ch4_6.htm.

[4] Shivaz (2010). Solar Cell Structures [Online]. Available:

www.imprintedimensions.blogspot.com/2010/08/solar-cell-structures.html.

[5] D. G. Sonnenenergie, Planning and installing photovoltaic systems: a guide for

installers, architects and engineers. London, U.K.: Earthscan, 2008.

[6] L. Castan˜er and S. Silvestre, Modelling Photovoltaic Systems using PSpice.

Chichester, U.K.: John Wiley & Sons, 2002.

[7] G.Pattison. (2001). Black Body Radiation [online]. Available:

http://www.egglescliffe.org.uk/physics/astronomy/blackbody/bbody.html

[8] S. Nema et.al, "Matlab/simulink based study of photovoltaic cells/modules/array and

their experimental verification" International Journal of Energy and Environment vol. 1,
NO. 3, pp. 487-500, 2010.

[9] M. G. Villalva et.al, “Comprehensive Approach to Modeling and Simulations of

Photovoltaic Arrays” IEEE Trans. Power Electron vol. 24, NO. 5, May 2009.

[10] P Mialhe et. al, "The diode quality factor of solar cells under illumination" Journal
of Physics D: Applied Physics vol. 19, NO. 3, 1986.

[11] M. Taherbaneh et.al, "Evaluation the Accuracy of One-Diode and Two-Diode

Models for a Solar Panel Based Open-Air Climate Measurements" Solar Cells-Silicon
Wafer-Based Technologies. Craotia: InTech, November 2011.

[12] K. Ishaque et.al, "Accurate MATLAB Simulink PV System Simulator Based on a

Two-Diode Model" Journal of Power Electronics, vol. 11, NO. 2, March 2011.

[13] M.D. Islam et.al, "Measurement of solar energy radiation in Abu Dhabi, UAE"
Applied Energy, vol. 86, 2009.

[14] V. M. Fthenakis and P. D. Moskowitz, Progress In Photovoltaics: Research and

Applications, Chichester, U.K.: John Wiley & Sons, 2000.

136

View publication stats