Available online at www.sciencedirect.

com

ScienceDirect
Solar Energy 122 (2015) 1123–1137
www.elsevier.com/locate/solener

Impact of factors influencing the performance of hybrid energy system

Vinay Thapar ⇑
University Institute of Technology, Rajiv Gandhi Technical University (RGPV), Bhopal 462033, Madhya Pradesh, India

Received 6 March 2015; received in revised form 16 August 2015; accepted 28 September 2015
Available online 12 November 2015

Communicated by: Associate Editor Elias Stefanakos

Abstract

Objective of this study is to analyse the impact of main factors, viz., load being supplied, wind speed, global irradiance, ambient
temperature and battery bank capacity (BBC), on the performance of Hybrid Energy System (HES). Levelised Cost of Energy (LCE)
and Loss of Power Supply Probability (LPSP) have been considered as performance indicators. Energy balance approach is used for
calculating BBC and corresponding LCE at desired LPSP for the selected system configuration and inputs.
It is concluded that for a given load, with increase in BBC, the LCE increases linearly, whereas LPSP decreases exponentially. With
increasing load, the BBC required for supplying power at a given LPSP, and corresponding LCE initially increase gradually and then shar-
ply and linearly. With increase in wind speed, BBC and corresponding LCE initially decrease sharply and then gradually; this variation is
more prominent for higher loads. The BBC and corresponding LCE decrease with increase in global irradiance, the decrease is gradual for
low loads; whereas, for large loads the decrease is initially sharp and then gradual. The BBC required for supplying power to a given load
at a given LPSP, and corresponding LCE, increase almost linearly with increase in annual average of hourly ambient temperature.
Ó 2015 Elsevier Ltd. All rights reserved.

Keywords: Hybrid Energy System (HES); Loss of Power Supply Probability (LPSP); Levelised Cost of Energy (LCE); Wind Turbine Generator (WTG);
Photovoltaic (PV)

1. Introduction Fortunately, these problems can be partially overcome by

The need for generation of electricity through renewable
energy sources is growing rapidly due to their well per-
ceived advantages. Of these sources wind and solar have
received plenty of attention in recent years due to their
direct availability globally and numerous other advantages.
In general, the variations of solar and wind energy do not
match with the time distribution of demand. Therefore,
independent use of these resources requires a large battery
backup and results in considerable over-sizing for
system reliability, which in turn makes the design costly.
⇑ Address: Electrical Engineering Department, University Institute of
Technology, Rajiv Gandhi Technical University (RGPV), Airport By Pass
Road, Gandhi Nagar, Bhopal 462033, Madhya Pradesh, India.
Tel.: +91 9826182871; fax: +91 755 2678819.
E-mail address: vinaythapar@rgtu.net
integrating the two resources in a proper combination
along with a storage facility to form a Hybrid Energy Sys-
tem (HES), using the strengths of one source to overcome
the weaknesses of the other (Bayod-Rújula et al., 2013;
Belmili et al., 2014; Chen, 2013; Papaefthymiou and
Papathanassiou, 2014; Torreglosa et al., 2015).
The HES may or may not be grid connected. The
isolated HESs are being extensively used in feeding loads
like, a remotely located single family, a military post, a
microwave station, a telecommunication tower, a village
community or a small industry. These also find
applications in rural electrification and for supplying
electricity to far-flung areas having difficult access due to
geographical restrictions such as islands, tourist spots.
Optimum sizing of components of HES is required for
its techno-economic feasibility. A large number of research

http://dx.doi.org/10.1016/j.solener.2015.09.043
0038-092X/Ó 2015 Elsevier Ltd. All rights reserved.
1124 V. Thapar / Solar Energy 122 (2015) 1123–1137

Nomenclature

AM air mass OM 0k maintenance cost in the first year of k th compo-

BBC battery bank capacity (Wh) nent ($)
CRF capital recovery factor Pe electrical power produced by the WTG at any
C bat ðtÞ battery bank capacity at hour t (Wh) specific speed v (W)
C bat ðt1) battery bank capacity at hour t  1 (Wh) P L ðtÞ electrical load on the system at hour t (W)
C bat nominal capacity of the battery bank (Wh) P mp maximum output power of PV generator (W)
C bat max maximum capacity of the battery bank (Wh) P pv ðtÞ DC power generated by the PV generator at
C bat min minimum allowable capacity of the battery bank hour t (W)
(Wh) P w ðtÞ AC power generated by the wind turbine at
d discount rate (%) hour t (W)
DODmax maximum depth of discharge of the battery q magnitude of charge on an electron
dI sc
dT c current temperature coefficient of solar cell (1.69e19 C)
((A/A)/°C) RC pk present worth of replacement cost of k th compo-
dV oc
dT c voltage temperature coefficient of solar cell nent ($)
(V/°C) Rs series resistance of the solar cell (X)
Eg ðtÞ net AC energy generated by the system at hour t r is the number of replacements required by the
(Wh) k th component
EL ðtÞ load demand at hour t (Wh) rs normalised value of cell resistance
EXCE(t) excess or wasted energy (Wh) S 0k salvage value of k th component ($)
FF fill factor of PV cell S pk present worth of salvage value of k th component
FF 0 fill factor of an ideal PV cell ($)
G global irradiance under STC (1000 W/m2) STC Standard Test Conditions
Ge effective global irradiance (W/m2) T total operation time (h)
HES Hybrid Energy System Ta ambient/air temperature (°C)
i inflation rate Tc operating cell temperature (°C)
IC k initial investment cost of the k th component of v wind speed at any instant (m/s)
HES ($) vc cut-in speed of the wind turbine (m/s)
I sc short circuit current of the PV cell (A) vf furling (or cut-out) speed of the wind turbine
kB Boltzman’s constant (1.38e23 J/K) (m/s)
LCC life cycle cost ($) V oc open circuit voltage of the PV cell (V)
LCE Levelised Cost of Energy ($/kWh) voc normalised value of open circuit voltage of the
LPSP Loss of Power Supply Probability PV cell
LPS Loss of Power Supply Vt thermal voltage (V)
m ideality factor (varies between 1 and 2) W–PV–B Wind–PV–Battery
n life span of the system (years) Dt time interval used for computations (h)
nr life span of the component requiring replace- gbatch battery charging efficiency
ment (years) gbatdch battery discharging efficiency
N cs number of solar cells connected in series in each ginv efficiency of inverter or dual inverter
PV module goth efficiency of other control and converter circuits
N cp number of solar cells connected in parallel in of the PV generator
each PV module grec efficiency of rectifier/dual inverter
N ms number of PV modules connected in series rB daily discharging rate of the battery bank (% per
N mp number of PV modules connected in parallel day)
OM pk present worth of operation and maintenance
cost of k th component ($)

papers have been published dealing with the design and siz-
ing of HES. Borowy and Salameh (1994) demonstrated a
methodology for calculating the optimum size of the PV
array for a Wind/PV hybrid system for a specific wind
turbine operating at a given load. It has been concluded
that optimal mix of wind and photovoltaic generators
depends on the particular site and load profile. Kellog
et al. (1998) have developed a simple numerical algorithm
for finding the optimum component sizes for the HES, to
supply a known load pattern at a given site, taking into
account economic factors. Yang et al. (2007) have devel-
oped the Hybrid Solar-Wind System Optimization Sizing
model to optimise the capacity sizes of different compo-
nents of HES, and have carried out optimisation studies
for a telecommunication relay station at a remote island
in Guangdong province, China.
 V. Thapar / Solar Energy 122 (2015) 1123–1137 1125

Diaf et al. (2008) have proposed an iterative technique AC Bus

for the optimisation of HES, and have optimised a HES

Wind
for a given load distribution assumed to be installed at Generator
three sites located at Corsica island. Similar approach
has been used by Kaabeche and Ibtiouen (2014) to
optimise the capacity sizes of various stand-alone
PV/wind/diesel/battery hybrid system components for zero PV Generator Inverter

load energy deficit for a site at Ghardaı̈a (Algeria). Rohani

and Nour (2014), using HOMER software, modelled and AC Load

designed a HES for a remote area in Ras Musherib, located

Baery Bi-direconal
in the western region of Abu Dhabi. Bank Inverter
Alsayed et al. (2014) have proposed a comprehensive
method for optimal design of hybrid power generation sys-
Fig. 1. Basic configuration of HES selected for the analysis.
tems taking into account environmental, economic and
technical issues.
Bayod-Rújula et al. (2013), for a typical residential load HES configuration is selected for the present study due to
in Spain, simulated on hourly basis the yearly energy bal- its inherent superiority (Setiawan et al., 2009). Fig. 1 shows
ances for hybrid systems with different combinations and such a configuration for a Wind–PV–Battery HES.
sizes of photovoltaic, wind energy conversion and battery The basic components of HES used in this study are: a
storage systems. 10 kW Wind turbine generator (WTG) (manufactured by
Recent reviews on the optimisation of renewable energy Bergey Windpower, model-BWC Excel-S); a 10.32 kW
systems show that heuristic methods have proved them- PV generator (comprising of 48, 215 W photovoltaic mod-
selves as good candidates in finding the optimum solution ules manufactured by BP solar Ltd., model number BP
in a reasonable computational time, especially where 3215B); Battery bank comprising of required number of
multi-criterion evaluation is required (Askarzadeh, 2013; batteries, each having capacity of 1800 Wh (model No.
Maleki and Askarzadeh, 2014a,b; Perera et al., 2013). J-150 Trojan Batteries); an inverter (having a rating of
Literature review reveals that sizing of HES mainly 1.1 times the rated capacity of the PV generator); and a
depends upon load being supplied by the system, desired dual inverter (rating taken as 1.25 times the maximum load
reliability and the climatic conditions of the site. However, on the system).
most of the researchers have optimised the system for a
given climatic condition and for a fixed load requirement.
3. Modelling of components of hybrid energy system
The variability of these factors has not been considered.
Thorough understanding of the factors and their variability
3.1. Modelling of wind turbine
influencing the performance of HES is a prerequisite for
designing a techno-economic optimum HES. Therefore,
Various authors have used different models for predict-
this study has been undertaken with the objective of analys-
ing the performance of wind turbines. These models have
ing, in detail, the impact of factors influencing the perfor-
been critically analysed and validated by Thapar et al.
mance of Wind–PV–Battery (W–PV–B) HES. The impact
(2011). It has been concluded therein that the modelling
of variation of wind speed, global irradiance, ambient tem-
methods in which actual power curve of a wind turbine is
perature, battery bank capacity (BBC) and load on the per-
used for developing characteristic equations, by utilising
formance of HES has been investigated and analysed.
the curve fitting technique of the method of least squares,
The remaining part of this paper is organised as follows:
give accurate results. Following characteristic equations
in Section 2, HES configuration selected for the present
of a wind turbine, obtained by fitting its actual power curve
study is described. Models developed for wind turbine gen-
using the method of least squares, can be used to predict
erator, PV generator, battery performance, reliability anal-
the performance of the wind turbine:
ysis (based on the concept of Loss of Power Supply
9
Probability (LPSP)) and economic analysis (based on the Pe ¼ 0 ðfor v < vc Þ >
>
>
Levelised Cost of Energy (LCE)), are discussed in Section 3. P e ¼ a1 v2 þ b1 v þ c1 ðfor vc 6 v < v1 Þ >
>
>
>
>
P e ¼ a2 v þ b2 v þ c2 ðfor v1 6 v < v2 Þ =
The methodology adopted for the study is given in Sec- 2
tion 4. Results and discussions are given in Section 5 and ð1Þ
: : : : >
>
conclusions are presented in Section 6. >
>
P e ¼ an v2 þ bn v þ cn ðfor vn 6 v 6 vf Þ >
>
>
>
2. Hybrid energy system configuration ;
Pe ¼ 0 ðfor v > vf Þ

There are various ways to integrate different where a1 ; b1 and c1 , etc., are the coefficients of quadratic
renewable/alternative power generation sources to form a equations; v1 to vn are various wind speeds on the power
HES (Nehrir et al., 2011). Power frequency AC coupled curve of the selected wind turbine; vc and vf are cut in and
1126 V. Thapar / Solar Energy 122 (2015) 1123–1137

furling speeds of the wind turbine, respectively, and v is the here m; k B and q stand for ideality factor, Boltzman’s con-
wind speed at which output of the turbine is to be calculated. stant and magnitude of charge on an electron, respectively.
Based on the above method, the characteristic equations The value of Rs at STC can be determined by the
describing the selected wind turbine are shown in Appendix expression
A. These have been derived by using the output power vs  
FF  V oc
wind speed data of the selected wind turbine available Rs ¼ 1   ð8Þ
FF 0 I sc
in the resource file of National Renewable Energy
Laboratory’s HOMER software (2012). The values of V oc ; I sc and other parameters of the solar cell
operating on arbitrary conditions of irradiance and cell
3.2. Modelling of PV generator temperature can be found from their standard values given
in manufacturer’s data sheet, using following assumptions
Maximum power output of a PV generator P mp , operat- (Lorenzo, 2003):
ing at arbitrary conditions of effective solar irradiance Ge
and cell temperature T c , can be determined with the help i. The short circuit current of the solar cell depends lin-
of information available in manufacturer’s data sheet by early on solar irradiance and increases slightly with
using following set of equations (Thapar et al., 2012): increasing temperature, and is given by
P mp ðGe ; T c Þ ¼ FF ðGe ; T c Þ V oc ðGe ; T c Þ I sc ðGe ; T c Þ  
 Ge
 
 dI sc
I sc ðGe ; T c Þ ¼ I sc  1 þ T C  T C ð9Þ
N ms N mp N cs N cp ginv goth ð2Þ G dT c
where ginv and goth are the efficiencies of the inverter circuit where dI sc
is the current temperature coefficient of the
dT c
and control & converter circuits associated with the PV solar cell.
generator, respectively. The PV generator consists of N ms ii. The open circuit voltage decreases linearly with the
modules in series and N mp modules in parallel, each module temperature and depends logarithmically on the illu-
having N cs and N cp cells in series and parallel, respectively. mination level, and is given by
V oc and I sc are, respectively, the open circuit voltage and
  dV oc
short circuit current of the PV cell and FF is its fill factor. V oc ðGe ; T c Þ ¼ V oc þ T c  T c
In Eq. (2) and equations to follow, a variable with Ge dT c
 
and/or T c in the bracket represents the value of that vari- Ge
þ V t ðT c Þ ln ð10Þ
able under arbitrary condition of solar irradiance Ge and/ G
or cell temperature T c , whereas * in the superscript (wher-
ever used) represents the value of that variable under Stan- where dVdTocc is voltage temperature coefficient, and is a
dard Test Conditions (STC) (Irradiance: 1000 W/m2; negative quantity.
Spectrum: AM 1.5; and Cell Temperature: 25 °C). iii. The operating temperature of the solar cell T c above
The fill factor of the PV cell at any arbitrary condition ambient temperature T a is roughly proportional to
can be satisfactorily found by suitably modifying the the incident irradiance and is given as
empirical formula developed by Green (1982): T c ¼ T a þ C t Ge ð11Þ
FF ðGe ; T c Þ ¼ FF 0 ðGe ; T c Þ ð1  rs ðGe ; T c ÞÞ ð3Þ
where the constant C t is related to Normal Operating
where FF0 is the fill factor of the ideal PV cell without resis- Cell Temperature (NOCT) by
tive effects and rs is the normalised value of cell resistance. NOCT  20
The values of FF0 and rs can be found by using Eqs. (4) Ct ¼ ð12Þ
800
and (5), respectively,
iv. The series resistance Rs is the property of solar cell,
voc ðGe ; T c Þ  lnðvoc ðGe ; T c Þ þ 0:72Þ
FF 0 ðGe ; T c Þ ¼ ð4Þ and is unaffected by operating conditions.
voc ðGe ; T c Þ þ 1
Rs I sc ðGe ; T c Þ Once the value of output power of the PV generator is
rs ðGe ; T c Þ ¼ ð5Þ known at any arbitrary condition of solar irradiance and
V oc ðGe ; T c Þ
ambient temperature, its energy yield for the required time
where voc and Rs are the normalised values of voltage and period can be calculated.
series resistance of the solar cell, respectively.
The value of voc is given by
3.3. Modelling of battery bank
V oc ðGe ; T c Þ
voc ðGe ; T c Þ ¼ ð6Þ
V t ðT c Þ Various researchers have adopted different techniques to
where V t is thermal voltage and is given by model the behaviour of battery bank (Berndt, 2006; Diaf
et al., 2008, 2007; Deshmukh and Deshmukh, 2008;
m k B ð273 þ T c Þ Dufo-López and Bernal-Agustı́n, 2005; Hongxing et al.,
V t ðT c Þ ¼ ð7Þ
q 2009). In most of these models the state of battery at a
 V. Thapar / Solar Energy 122 (2015) 1123–1137
given hour t, is related to its previous state of charge and to
the energy production and consumption of the system
during the time interval from t  1 to t. At a given
hour, the battery bank capacity is subject to following
constraints:
C bat min 6 C bat ðtÞ 6 C bat max
C bat min ¼ ð1  DODmax ÞC bat
where C bat max is the maximum capacity of the battery bank
and is equal to its nominal capacity, i.e., C bat ; C bat min is the
minimum allowable capacity of the battery bank, and is
determined by the maximum depth of discharge (DODmax),
which is 80% for the selected batteries.
The following conditions may exist depending upon the
relative magnitude of total power generated by the HES
and the load requirement:
i. When the total AC output power of the generating
system is greater than the load requirement, the bat-
tery bank is under charging state, and the battery
bank capacity at hour t is given by
 
rB Dt
C bat ðtÞ ¼ C bat ðt  1Þ 1 
24
 
þ Eg ðtÞ  EL ðtÞ grec gbatch
When the battery bank capacity reaches its maximum
value C bat max , the battery bank stops getting charged
and the remaining energy, called excess or wasted
energy, either goes as waste or is supplied to a dump
load, and is given by
 
C bat max  C bat ðt  1Þ
EXCEðtÞ ¼ Eg ðtÞ  EL ðtÞ þ
grec gbatch
ii. When the total AC output power of the generating
system is less than the load requirement, the battery
bank supplies the stored energy and is in the state
of discharging. The battery bank capacity at any hour
t is then expressed as
    RT
rB Dt EL ðtÞ  Eg ðtÞ
C bat ðtÞ ¼ C bat ðt  1Þ 1   LPSP ¼ R T0
24 ginv gbatdch
If the battery capacity reaches its minimum value
C bat min , the load is disconnected from the circuit as
the system is unable to meet the load demand.
The energy deficit between the load demand and the
power supplied by the HES (which also includes the
power supplied by the battery bank) is called Loss
of Power Supply (LPS), and can be expressed as
LPSðtÞ ¼ EL ðtÞ  ðEg ðtÞ þ ðC bat ðt  1Þ
 C bat min Þginv gbatdch Þ
1127
iii. When the total AC output power of the generating sys-
tem is equal to the load requirement, the storage capac-
ity of the battery bank remains almost unchanged
(other than the self discharge), and is given by
 
rB Dt
ð13Þ C bat ðtÞ ¼ C bat ðt  1Þ 1  ð19Þ
24
ð14Þ where
C bat ðtÞ and C bat ðt1) are the battery bank capacities
at hour t and t  1, respectively.
rB is the daily discharging rate of the battery bank
(the manufacturer’s documentation gives a self
discharging rate of 25% over 6 months at a storage
temperature of 20 °C, i.e., 0.14% per day
(Diaf et al., 2007)).
The load demand at hour t is given by
EL ðtÞ ¼ P L ðtÞDt, where P L ðtÞ and EL ðtÞ are the load
connected and energy consumed, respectively, and
Dt is the time interval used for the computations,
which is one hour for the present study.
Eg ðtÞ is the net AC energy generated by the
system at hour t and is given by Eg ðtÞ ¼ ðP w ðtÞ þ
ðP pv ðtÞ  ginv ÞÞDt, where P w (tÞ and P pv (tÞ are AC
and DC power generated by wind turbine and PV
generator, respectively.
ð15Þ ginv and grec are the inverter and rectifier/dual inverter
efficiencies, respectively, considered as 90–95%; gbatch
is the battery charging efficiency, which is between
0.65 and 0.85, depending on the charging current;
and gbatdch is the battery discharging efficiency, which
is usually assumed as 1 (Diaf et al., 2008, 2007;
 Hongxing et al., 2009).
ð16Þ 3.4. Reliability model based on concept of LPSP
It is customary to quantify reliability of a supply system
in terms of LPSP, defined as the ratio between the energy
deficit and the energy demand, both referring to the load,
over the total operation time (T Þ of the installation
(Lorenzo, 2003). Thus,
Energy deficit
ð20Þ
0
Energy demand
ð17Þ
Accordingly, LPSP, for a considered period T, can be
found by dividing all the LPS(tÞ values by the total load
requirement for that given period T, and is expressed as
PT
LPSðtÞ
LPSP ¼ PTt¼1 ð21Þ
t¼1 L ðtÞ Dt
P
3.5. Economy model based on the concept of LCE
LCE is one of the commonly used indicators for com-
ð18Þ paring financial performance of various power generating
1128 V. Thapar / Solar Energy 122 (2015) 1123–1137
 n
utilities, including HES (Diaf et al., 2008, 2007; Whisnant 1þi
S pk ¼ S 0k ð27Þ
et al., 2003). It is the unit cost of electricity in $ (or any 1þd
other currency) per kWh, and is given by
where S 0k is the salvage value of the k th component of the
LCC  CRF HES.
LCE ¼ ð22Þ
Annual electric production The capital recovery factor (CRF), which converts a pre-
where LCC is the life cycle cost, and is defined as the sum sent value into a stream of equal annual payments over a
of the present worth of all the cost components associated specified time, at a specified discount rate, is given by
n
with a project relating to procurement, erection, operation, d ð1 þ d Þ
maintenance and replacement activities, as also the salvage CRF ¼ n ð28Þ
ð1 þ d Þ  1
value.
The life cycle cost of the hybrid system can, in general, Using Eqs. (22)–(28), the LCE of a HES can be calculated.
be found by using the following equation:
Xj X j X
l X
s 3.5.1. Cost components associated with hybrid energy system
LCC ¼ ðICÞk þ ðOM p Þk þ ðRC p Þk  ðS P Þk The cost components in the equations for calculating
k¼1 k¼1 k¼1 k¼1 LCE depend upon the configuration of the HES, and are
ð23Þ shown in Table 1 for the HES considered in the present
study.
where ICk is the initial investment cost of the k th compo-
nent of HES; j; l and s are the limits of summation, being
4. Methodology
the number of components which, form the hybrid system,
require replacement and have salvage value, respectively.
4.1. The flow chart of the basic algorithm used for this
The present worth of operation and maintenance cost of
study is shown in Fig. 2, which for the selected ratings of
the k th component (OMpk) is given as under, if the mainte- wind and PV generators, by using the energy balance
nance amount is spent at the beginning of the year: approach, computes the BBC required for supplying power
   n 
1þd 1þi to a connected load at a desired LPSP. Inputs required for
OM pk ¼ OM 0k 1 ð24Þ
d i 1þd this algorithm are wind speed, solar radiation, ambient
temperature, load and LPSP. Once the BBC is known the
However, if the maintenance amount is spent at the end of
LCE of the selected system is computed.
the year then
   n 
1þi 1þi Programs have been developed for the relevant compu-
OM pk ¼ OM 0k 1 ð25Þ
d i 1þd tations and implemented using MATLAB. Models used for
where OM0k is the maintenance cost in the first year of the performance analysis of HES components, reliability and
k th component, i is the inflation rate, d is the discount rate economic analyses are given in Section 3. The characteristic
and n is the life span of the system. equations used for the selected wind turbine model are
The present worth of the replacement cost RCpk of the shown in Appendix A. The relevant technical specifications
for the selected PV module are given in Appendix B
k th component of the HES is given by
(Website of BP solar (2012)).
Xr  n y
1þi r
RC pk ¼ IC k ð26Þ
y¼1
1þd 4.2. Performance analyses of HES have been done with
following common conditions, unless otherwise specified:
where nr is the life span of the component requiring
replacement, and r is the number of replacements required i. Hourly weather data of Valkenburg, Netherlands
by the k th component. (latitude 52°90 N & longitude 4°250 E) for the year
The present worth of the salvage value of the k th 2004 is considered for the analysis (Royal
component S pk of the HES is given by Netherlands Meteorological Institute (2012)).
Table 1
Various cost components associated with HES.
Component Basic Installation Annual Salvage Life span
of HES cost cost O&M cost value (years)
Wind Energy Conversion Varies linearly from $ 3.6 to 3.2/W for rating 20% of the cost 1% of the cost of WECS 10% of the capital cost 25
System (WECS) varying from 1 to 10 kW of WECS of WECS
PV system $ 1.75/W 50% of the cost 0.2% of the total capital 10% of the capital cost 25
of PV array cost of PV system of PV system
Battery bank $ 0.1132/Wh – 1% of the cost of battery – 5
bank
Dual inverter with charger $ 0.76/W – 1% of the cost of inverter – 10
 V. Thapar / Solar Energy 122 (2015) 1123–1137
ii. The system is feeding to yearly hourly load of 800 W.
iii. The efficiency of inverter/dual inverter is assumed to
be 95%.
iv. Battery charging and discharging efficiencies are
assumed 85% and 100%, respectively.
v. Inflation and discount rates are assumed as 7% and
10%, respectively.
vi. PV module is mounted horizontally.
vii. Ideality factor for the solar cell is taken as 1.2.
4.3. Following procedures have been adopted to study
the impact of factors influencing the performance of
HES:
4.3.1. Impact of battery bank capacity (BBC)
Number of batteries connected in HES is increased from
0 to such a value where complete load requirement is met
by the renewable source/s. For each of these BBC, corre-
sponding values of performance indicators (LCE and
LPSP) are calculated.
4.3.2. Impact of load
The battery bank requirement and corresponding LCE
are calculated for supplying power to yearly hourly electri-
cal load varying from 0.1 to 10.0 kW (in steps of 100 W), at
a desired LPSP.
4.3.3. Impact of wind speed
The values of BBC required and corresponding LCE,
for supplying power to different loads at desired LPSP,
have been calculated for various values of annual average
of hourly wind speed (obtained by multiplying the hourly
wind speed at Valkenburg by a factor varying from 0.1
to 2, in steps of 0.1).
4.3.4. Impact of global irradiance
The values of BBC required and corresponding LCE,
for supplying power to different loads at desired LPSP,
have been computed for various values of annual average
of hourly global irradiance (obtained by multiplying the
hourly global irradiance at Valkenburg by a factor varying
from 0.1 to 3, in steps of 0.1, subject to a maximum hourly
irradiance of 1000 W/m2).
4.3.5. Impact of ambient temperature
The values of BBC required and corresponding LCE,
for supplying power to different loads at desired LPSP,
have been calculated for different annual average of hourly
ambient temperature (obtained by adding a factor, varying
in the range of 20 to 20 °C in steps of 1 °C, to the hourly
ambient temperature of Valkenburg).
1129
5. Results and discussions
5.1. Impact of battery bank capacity on the performance of
HES
The variation of LCE and LPSP with BBC is plotted in
Fig. 3.
It can be observed from Fig. 3 that:
i. With the increase in the BBC from 1.8 (single battery)
to 14.4 kWh (8 batteries), the LPSP decreases sharply
from 0.1950 to 0.0374; thereafter, it decreases gradu-
ally to 0 for BBC of 207 kWh (115 batteries).
ii. The LCE increases linearly from $ 0.2995/kWh for
BBC of 1.8 kWh to $ 0.3196/kWh for BBC of
14.4 kWh and thereafter to $ 0.6265/kWh for BBC
of 207 kWh (115 batteries); the relationship is
LCE ¼ 0:0015939ðBBC Þ þ 0:29661
It can be concluded from above that for a W–PV–B
system:
i. The LCE increases linearly with the increase in BBC.
ii. The LPSP decreases exponentially with the increase
in BBC.
5.2. Impact of load on the performance of HES
The BBC required for obtaining power at LPSP of 0 and
0.1, for various loads is plotted in Fig. 4; corresponding
variation in LCE is plotted in Fig. 5.
It can be seen from Figs. 4 and 5 that:
i. For increase in load from 0.1 to 3.4 kW, the BBC
required to maintain supply at a LPSP of 0 increases
gradually from 3.6 (2 batteries) to 3538.8 kWh (1966
batteries); beyond which, BBC required increases lin-
early and sharply to 75322.8 kWh (41,846 batteries)
for a load of 10.0 kW. Similar trend is observed for
variation of LCE with load. The LCE being $
0.2966, 5.9582 and 120.427/kWh for loads of 0.1,
3.4 and 10 kW, respectively.
ii. Similar pattern is observed for maintaining supply at
a LPSP of 0.1, with sharp increase in values of BBC
and LCE occurring beyond a load of 3.6 kW. The
LCE being $ 0.2938, 2.1556 and 101.5807/kWh for
loads of 0.1, 3.6 and 10 kW, respectively.
Following conclusions are drawn from the above
observations:
i. The BBC required for supplying power at a given
LPSP increases with load. The increase is initially
gradual but sharp and linear beyond a certain load.
1130 V. Thapar / Solar Energy 122 (2015) 1123–1137

1.Input hourly weather data of wind speed; solar irradiance and ambient temperature.
2. Input hourly load (EL).
3.Input all relevant technical data pertaining to different components used in the HES.
4.Input all relevant information pertaining to economical analysis of the system.
5.Input required LPSP (LPSPreq).

NB = 1

t= 1

Calculate AC output
generated by the system
for that hour Eg(t)

Battery charges, Battery

calculate its discharges,
Yes Eg(t)>EL(t) No Eg(t)=EL(t) No
new capacity calculate its
Cbat(t) new capacity
Cbat(t)
Yes
Calculated new battery capacity Cbat(t)
after considering self discharging
Yes If No
Cbat(t) >Cbatmax Cbat(t) = Cbat(t)

Cbat(t) = Cbatmax LPS (t) = 0 Yes If

Calculate Excess Cbat(t) ≥Cbatmin
Energy
No
Cbat(t) = Cbatmin

Calculate
LPS (t)

t = t +1 No t = 8760

Yes
Calculate LPSP & Excess Energy

No LPSP ≤
NB = NB + 1 LPSPreq

Yes
Do economic analysis

Output LCE; Excess Energy, No.

of Batteries required (NB), LPSP
and store hourly outputs

Fig. 2. Basic flow chart for the analysis of HES.

ii. The Increase in LCE with the load follows the
same general pattern as that of variation of BBC with
load.
iii. The difference between the BBC required (and the
corresponding LCE) for supplying power at different
LPSPs, increases with load.
5.3. Impact of wind speed on the performance of HES
The variation of BBC with annual average of hourly
wind speed required for maintaining supply at a LPSP of
0, for different yearly hourly loads are plotted in Fig. 6,
corresponding variation of LCE is plotted in Fig. 7.
It can be observed from Figs. 6 and 7 that at a LPSP of 0:
i. When feeding power to a yearly hourly load of
2.0 kW, the requirement of BBC initially decreases
sharply from 11244.6 (6247 batteries) to 1134 kWh
(630 batteries) for increase in wind speed from 1.31
to 5.91 m/s. Beyond this wind speed the reduction is
gradual, decreasing to 126 kWh (70 batteries) at a
wind speed of 13.15 m/s. The requirement of BBC
decreases with decrease in load, and the curve of
 V. Thapar / Solar Energy 122 (2015) 1123–1137 1131

BBC versus wind speed becomes flatter. For a load of iv. The variation of LCE with wind speed follows
0.2 kW, the requirement of BBC decreases from 59.4 the same general pattern as that of BBC with wind
(33 batteries) to 5.4 kWh (3 batteries) for increase in speed.
annual average of hourly wind speed from 1.31 to
13.15 m/s.
ii. The decrease in LCE with increase in annual average 5.4. Impact of global irradiance on the performance of HES
of hourly wind speed follows the same general pattern
of variation as that of BBC with wind speed. For a The variation of BBC with annual average of hourly
load of 2.0 kW, the LCE decreases sharply from global irradiance required to maintain supply at a
$ 57.4287 to 2.4610/kWh for increase in wind speed LPSP of 0, for different yearly hourly loads, is plotted in
from 1.31 to 5.91 m/s, and then gradually to Fig. 8; corresponding variation of LCE is plotted in
$ 0.2829/kWh for a wind speed of 13.15 m/s. Fig. 9.
Whereas for a load of 0.2 kW, the LCE decreases It can be observed from Figs. 8 and 9 that when W–PV–
from $ 1.2066 to 0.1675/kWh for increase in annual B system supplies power at a LPSP of 0:
average of hourly wind speed from 1.31 to 13.15 m/s.
i. For yearly hourly load of 0.5 kW, the requirement of
It can be concluded from above observations that for a BBC decreases from 120.6 (67 batteries) to 12.6 kWh
W–PV–B system supplying power at a given LPSP: (7 batteries) with increase in annual average of hourly
global irradiance from 12.24 to 270.05 W/m2. The
i. For large loads the requirement of BBC at lower corresponding decrease in BBC for loads of 1.0, 1.5
speed range (of about 3–5 m/s) decreases sharply with and 2.0 kW is from 428.4 (238 batteries) to
increase in annual average of hourly wind speed, 172.8 kWh (96 batteries), 763.2 (424 batteries) to
beyond which requirement of BBC decreases gradu- 471.6 kWh (262 batteries) and 1738.8 (966 batteries)
ally, whereas for small loads the decrease is gradual to 804.6 kWh (447 batteries), respectively.
for all wind speeds. ii. For loads of 0.5, 1.0 and 1.5 kW the BBC
ii. The rate of change of BBC with wind speed increases decreases gradually with the increase in annual
with load. average of hourly global irradiance from 12.24 to
iii. At a given wind speed requirement of BBC increases 270.05 W/m2. Whereas for a load of 2.0 kW the
with load. However, as the wind speed increases, the BBC decreases sharply for increase in global
difference between BBC required for different loads irradiance from 12.24 to 36.73 W/m2, thereafter the
decreases. decrease is gradual.

1 0.2
LCE
LPSP
0.9 0.18
Levelised Cost of Energy in $/Wh -------------------->

0.8 0.16
Loss of Power Supply Probability -------------------->

0.7 0.14

0.6 0.12

0.5 0.1

0.4 0.08

0.3 0.06

0.2 0.04

0.1 0.02

0 0
0 20 40 60 80 100 120 140 160 180 200
Battery Bank Capacity in kWh -------------------->

Fig. 3. Variation of LCE and LPSP with BBC for a W–PV–B system.
1132 V. Thapar / Solar Energy 122 (2015) 1123–1137

4
x 10
8
LPSP = 0
LPSP = 0.1

7
Battery Bank Capacity in kWh -------------------->

0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Hourly Electrical Load in W --------------------->

Fig. 4. Variation of BBC with load for a W–PV–B system.

140

LPSP = 0
LPSP = 0.1
Levelised Cost of Energy in $/kWh -------------------->

120

100

80

60

40

20

0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Hourly Electrical Load in W -------------------->

Fig. 5. Variation of LCE with load for a W–PV–B system.

iii. For a load of 0.5 kW, the value of LCE decreases decrease in LCE, for yearly hourly load of 1.0,
from $ 0.6816 to 0.2378/kWh, with the increase 1.5 and 2.0 kW, is from $1.3749 to 0.4342,
in annual average of hourly global irradiance 2.1284 to 0.7978 and 4.3134 to 1.2026/kWh,
from 12.24 to 270.05 W/m2. The corresponding respectively.
 V. Thapar / Solar Energy 122 (2015) 1123–1137 1133

12000
Load = 0.2 kW
Load = 0.5 kW
LPSP = 0 Load = 1.0 kW
Load = 1.5 kW
Battery Bank Capacity in kWh -------------------->

10000 Load = 2.0 kW

8000

6000

4000

2000

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Average Wind Speed in m/s -------------------->

Fig. 6. Variation of BBC with wind speed for a W–PV–B system.

60
Load = 0.2 kW
LPSP = 0 Load = 0.5 kW
Levelised Cost of Energy in $/kWh -------------------->

Load = 1.0 kW
Load = 1.5 kW
50 Load = 2.0 kW

40

30

20

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Average Wind Speed in m/s --------------------->

Fig. 7. Variation of LCE with wind speed for a W–PV–B system.

It can be concluded from above that for a W–PV–B sys- the requirement of BBC initially decreases sharply
tem supplying power at a LPSP of 0: and thereafter gradually with the increase in global
irradiance.
i. For low values of load the BBC decreases gradually ii. The variation of LCE with global irradiance follows,
with the increase in annual average of hourly global in general, the same pattern as that for the variation
irradiance. However, beyond a certain value of load, of BBC with global irradiance.
1134 V. Thapar / Solar Energy 122 (2015) 1123–1137

1800
Load = 0.5 kW
LPSP = 0
Load = 1.0 kW
1600 Load = 1.5 kW
Battery Bank Capacity in kWh -------------------->

Load = 2.0 kW

1400

1200

1000

800

600

400

200

0
0 25 50 75 100 125 150 175 200 225 250 275
Annual Average of Hourly Global Irradiance in W/sq. meter -------------------->

Fig. 8. Variation of BBC with global irradiance for a W–PV–B system.

4.5
Load = 0.5 kW
LPSP = 0 Load = 1.0 kW
Load = 1.5 kW
Levelised Cost of Energy in $/kWh -------------------->

4
Load = 2.0 kW

3.5

2.5

1.5

0.5

0
0 25 50 75 100 125 150 175 200 225 250 275 300
Annual Average of Hourly Global Irradiance in W/sq. meter -------------------->

Fig. 9. Variation of LCE with global irradiance for a W–PV–B system.

5.5. Impact of ambient temperature on the performance of It can be observed from Figs. 10 and 11 that for a
HES W–PV–B system supplying power at a LPSP of 0:

The variation of BBC with annual average of hourly i. For a yearly hourly load of 0.5 kW, the BBC required
ambient temperature required to maintain supply at a increases from 55.8 (31 batteries) to 73.8 kWh (41
LPSP of 0, for different yearly hourly loads, is plotted in batteries) with the increase in the annual average of
Fig. 10; corresponding variation in LCE is plotted in hourly ambient temperature from 9.48 to 30.52 °C.
Fig. 11. The corresponding increase in BBC for loads of
 V. Thapar / Solar Energy 122 (2015) 1123–1137 1135

1200
Load = 0.5 kW
Load = 1.0 kW LPSP = 0
1100
Load = 1.5 kW
Load = 2.0 kW
Battery Bank Capacity in kWh -------------------->

1000

900

800

700

600

500

400

300

200

100

0
-10 -7.5 -5 -2.5 0 2.5 5 7.5 10 12.5 15 17.5 20 22.5 25 27.5 30 32.5
Annual Average of Hourly Ambient Temperature in deg. C -------------------->

Fig. 10. Variation of BBC with ambient temperature for a W–PV–B system.

2.2
Load = 0.5 kW LPSP = 0
Load = 1.0 kW
2
Levelised Cost of Energy in $/kWh -------------------->

Load = 1.5 kW
Load = 2.0 kW

1.8

1.6

1.4

1.2

0.8

0.6

0.4

0.2
-10 -7.5 -5 -2.5 0 2.5 5 7.5 10 12.5 15 17.5 20 22.5 25 27.5 30 32.5
Annual Average of Hourly Ambient Temperature in deg. C -------------------->

Fig. 11. Variation of LCE with ambient temperature for a W–PV–B system.

1.0, 1.5 and 2.0 kW is from 318.6 to 347.4 kWh (177 It can be concluded from the above observations that
to 193 batteries), 646.2 to 676.8 kWh (359 to 376 bat- for W–PV–B system supplying power at a LPSP of 0:
teries) and 993.6 to 1027.8 kWh (552 to 571 batter-
ies), respectively. i. The BBC required, for supplying power to a given
ii. With the increase in BBC the value of LCE also load at a given LPSP, and the corresponding LCE
increases. For increase in annual average of hourly increase almost linearly with the increase in annual
ambient temperature from 9.48 to 30.52 °C, the value average of hourly ambient temperature.
of LCE increases from $ 0.3643 to 0.4343, 0.7665 to ii. The rate of increase in LCE and BBC with
0.8986, 1.2669 to 1.4567 and 1.7973 to 2.0510/kWh ambient temperature increases slightly with increase
for loads of 0.5, 1.0, 1.5 and 2.0 kW, respectively. in load.
1136 V. Thapar / Solar Energy 122 (2015) 1123–1137

5.6. The results of this study have been obtained by using ¼ 0:064 v2  0:018 v  1:332 ðfor 7 < v 6 9Þ ðI:4Þ
well established mathematical models, which have been val- ¼ 0:17 v  2:17 v þ 9:45
2
ðfor 9 < v 6 11Þ ðI:5Þ
idated by various authors, some of them have been cited in
the appropriate sections. As such, these results can be consid- ¼ 0:085 v  0:775v þ 4:39
2
ðfor 11 < v 6 13Þ ðI:6Þ
ered as validated for the purpose of achieving the objective of ¼ 0:14 v þ 5:23 v  35:65 ðfor 13 < v 6 15Þ
2
ðI:7Þ
the present study.
¼ 0:38 v þ 12:48 v  90:4 ðfor 15 < v 6 17Þ
2
ðI:8Þ
¼ 0:11 v þ 3:51 v  15:94 ðfor 17 < v 6 19Þ
2
ðI:9Þ
6. Conclusions
¼ 0:06 v þ 1:78 v  1:12 ðfor 19 < v 6 21Þ
2
ðI:10Þ
The first part of this paper discusses simple mathemati- ¼ 4:921e  15 v  0:68 v þ 24:08
2
ðfor 21 < v 6 23Þ ðI:11Þ
cal models for the modelling of various components of ¼ 5:181e  16 v  0:68v þ 24:08 ðfor 23 < v 6 25Þ
2
ðI:12Þ
HES. These models use only limited data which are nor-
¼0 ðfor v > 25Þ ðI:13Þ
mally available in manufacturer’s data sheet. Reliability
and economic models based on the concept of Loss of
Power Supply Probability and Levelised Cost of Energy, Appendix B. 215 W Photovoltaic module manufactured
respectively, have also been discussed. by BP solar Ltd
In the second part of the paper the impact of variation
in battery bank capacity, load, wind speed, global irradi- Sr. Item Specification
ance and ambient temperature on the performance of the number
selected hybrid energy system have been investigated and 1 Model number BP 3215B
analysed. The LCE and LPSP have been considered as per- 2 Maximum power under 215 W
formance indicators. STC
In view of varied nature of investigations carried out, 3 Voltage at maximum 29.1 V
the conclusions for each of the investigations have been power under STC
given in ‘‘Results and Discussions” section of the respective 4 Current at maximum 7.4 A
study for better comprehension. The main findings are power under STC
summarised below: 5 Short circuit current 8.10 A
It has been observed that for a given load, with increase under STC
in BBC, the LCE increases linearly; whereas, LPSP 6 Open circuit voltage 36.5 V
decreases exponentially. With increase in load, the BBC under STC
required for supplying power at a given LPSP and corre- 7 Temperature coefficient (0.065 ± 0.015)%/°C
sponding LCE initially increase gradually and then sharply of Isc
and linearly. With increase in wind speed, BBC and corre- 8 Temperature coefficient (0.36 ± 0.05)%/°C
sponding LCE initially decrease sharply and then gradu- of Voc
ally; this variation is more prominent for higher loads. 9 Normal Operating Cell 47 ± 2 °C
The BBC and the corresponding LCE decrease with the Temperature (NOCT)
increase in global irradiance, the decrease is gradual for 10 Solar cells connected in 60
low loads, whereas for large loads the decrease is initially series
sharp and then gradual. The BBC required for supplying
power to a given load at a given LPSP and the correspond-
ing LCE increase almost linearly with the increase in
annual average of hourly ambient temperature. References
It is expected that the findings of this study would help
Alsayed, M., Cacciato, M., Scarcella, G., Scelba, G., 2014. Design of
in designing an optimum HES, with limited data which are
hybrid power generation systems based on multi criteria decision
normally available. analysis. Solar Energy 105, 548–560.
Askarzadeh, A., 2013. A discrete chaotic harmony search-based simulated
annealing algorithm for optimum design of PV/wind hybrid system.
Appendix A Solar Energy 98, 190–195.
Bayod-Rújula, Á.A., Haro-Larrode’, M.E., Martı́nez-Gracia, A., 2013.
Sizing criteria of hybrid photovoltaic–wind systems with battery
Equations based on the method of least squares describing
storage and self-consumption considering interaction with the grid.
wind turbine model BWC Excel-S, manufactured by Ber- Solar Energy 98, 582–591.
gey Windpower, are as under: Belmili, H., Haddadi, M., Bacha, S., Almi, M.F., Bendib, B., 2014.
Sizing stand-alone photovoltaic–wind hybrid system: techno-economic
Pe ¼ 0 ðfor v < 3Þ ðI:1Þ analysis and optimization. Renew. Sust. Energy Rev. 30, 821–
¼ 0:0695 v  0:2855v þ 0:277
2
ðfor 3 6 v 6 5Þ ðI:2Þ 832.
Berndt, D., 2006. Maintenance-Free Batteries: Based on Aqueous
¼ 0:1045 v  0:7085 v þ 1:517
2
ðfor 5 < v 6 7Þ ðI:3Þ Electrolyte, third ed. Overseas Press, New Delhi.
 V. Thapar / Solar Energy 122 (2015) 1123–1137
Borowy, B.S., Salameh, Z.M., 1994. Optimum photovoltaic array size for
a hybrid wind/PV system. IEEE Trans. Energy Convers. 09 (3), 483–
488.
Chen, H.C., 2013. Optimum capacity determination of stand-alone hybrid
generation system considering cost and reliability. Appl. Energy 103,
155–164.
Deshmukh, M.K., Deshmukh, S.S., 2008. Modeling of hybrid renewable
energy systems. Renew. Sust. Energy Rev. 12, 235–249.
Diaf, S., Belhamel, M., Haddadi, M., Louche, A., 2008. Technical and
economic assessment of hybrid photovoltaic/wind system with battery
storage in Corsica island. Energy Policy 36 (2), 743–754.
Diaf, S., Diaf, D., Belhamel, M., Haddadi, M., Louche, A., 2007. A
methodology for optimal sizing of autonomous hybrid PV/wind
system. Energy Policy 35 (11), 5708–5718.
Dufo-López, R., Bernal-Agustı́n, J.L., 2005. Design and control strategies
of PV–diesel systems using genetic algorithms. Solar Energy 79, 33–46.
Green, M.A., 1982. Solar Cells: Operating Principles, Technology and
System Applications. Prentice-Hall, New Jersey.
Hongxing, Y., Wei, Z., Chengzhi, L., 2009. Optimal design and techno-
economic analysis of a hybrid solar-wind power generation system.
Appl. Energy 86 (2), 163–169.
Kaabeche, A., Ibtiouen, R., 2014. Techno-economic optimization of
hybrid photovoltaic/wind/diesel/battery generation in a stand-alone
power system. Solar Energy 103, 171–182.
Kellog, W.D., Nehir, M.H., Venkataramanan, G., Gerez, V., 1998.
Generation unit sizing and cost analysis for stand-alone wind,
photovoltaic, and hybrid wind/PV systems. IEEE Trans. Energy
Convers. 13, 70–75.
Lorenzo, E., 2003. Energy collected and delivered by PV modules. In:
Luque, A., Hegedus, S. (Eds.), Handbook of Photovoltaic Science and
Engineering. John Wiley, Sussex, pp. 905–970.
Maleki, A., Askarzadeh, A., 2014a. Artificial bee swarm optimization for
optimum sizing of a stand-alone PV/WT/FC hybrid system consider-
ing LPSP concept. Solar Energy 107, 227–235.
Maleki, A., Askarzadeh, A., 2014b. Comparative study of artificial
intelligence techniques for sizing of a hydrogen-based stand-alone
photovoltaic/wind hybrid system. Int. J. Hydrogen Energy 39, 9973–
9984.
Nehrir, M.H., Wang, C., Strunz, K., Aki, H., Ramakumar, R., Bing, J.,
Miao, Z., Salameh, Z., 2011. A review of hybrid renewable/alternative
1137
energy systems for electric power generation: configurations, control,
and applications. IEEE Trans. Sust. Energy 2 (4), 392–403.
Papaefthymiou, S.V., Papathanassiou, S.A., 2014. Optimum sizing of
wind-pumped-storage hybrid power stations in island systems. Renew.
Energy 64, 187–196.
Perera, A.T.D., Attalage, R.A., Perera, K.K.C.K., Dassanayake, V.P.C.,
2013. A hybrid tool to combine multi-objective optimization and
multi-criterion decision making in designing standalone hybrid energy
systems. Appl. Energy 107, 412–425.
Rohani, G., Nour, M., 2014. Techno-economical analysis of stand-alone
hybrid renewable power system for Ras Musherib in United Arab
Emirates. Energy 64, 828–841.
Setiawan, A.A., Zhao, Y., Nayar, C.V., 2009. Design, economic analysis
and environmental considerations of mini-grid hybrid power system
with reverse osmosis desalination plant for remote areas. Renew.
Energy 34, 374–383.
Thapar, V., Agnihotri, G., Sethi, V.K., 2011. Critical analysis of methods
for mathematical modelling of wind turbines. Renew. Energy 36 (11),
3166–3177.
Thapar, V., Agnihotri, G., Sethi, V.K., 2012. Estimation of hourly
temperature at a site and its impact on energy yield of a PV module.
Int. J. Green Energy 9 (6), 553–572.
Torreglosa, J.P., Garcı́a, P., Fernández, L.M., Jurado, F., 2015. Energy
dispatching based on predictive controller of an off-grid wind turbine/
photovoltaic/hydrogen/battery hybrid system. Renew. Energy 74, 326–
336.
Web Site of HOMER Software. <http://www.homerenergy.com> (last
accessed on: 30 January 2012).
Web Site of Royal Netherlands Meteorological Institute. <http://www.
knmi.nl> (last accessed on: 30 January 2012).
Web Sites of bp Solar. <http://www.bp.com/sectiongenericarticle.do?cat-
egoryId=9019633&contentId=7036926 https://na3.salesforce.com/sfc/
p/3000000001pzusEuSrbJQE.wKFH1JIDeow3fqPY> (last accessed
on: 30 January 2012).
Whisnant, R.A., Johnston, S.A., Hutchby, J.H., 2003. Economic analysis
and environmental aspects of photovoltaic systems. In: Luque, A.,
Hegedus, S. (Eds.), Handbook of Photovoltaic Science and Engineer-
ing. John Wiley, Sussex, pp. 971–1003.
Yang, H., Lu, L., Zhou, W., 2007. A novel optimization sizing model for
hybrid solar-wind power generation system. Solar Energy 81, 76–84.