Sie sind auf Seite 1von 9

Energy 36 (2011) 1214e1222

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Sizing optimization of grid-independent hybrid photovoltaic/wind power


generation system
type : hybride isol
But: optimisation du dimensionmt pour alimentation residentiel
Methd: technique itrative

A. Kaabeche a, *, M. Belhamel a, R. Ibtiouen b


a
b

Centre de Dveloppement des Energies Renouvelables, B.P. 62 16340 Bouzareah, Algiers, Algeria
Ecole Nationale Suprieure Polytechnique dEl Harrach, Algiers, Algeria

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 20 April 2010
Received in revised form
15 November 2010
Accepted 17 November 2010
Available online 23 December 2010

To allow a real penetration of the huge dispersed naturally renewable resources (wind, sun, etc.)
intermittent and more or less easily predictable, optimal sizing of hybrid renewable power generation
systems prove to be essential. This paper recommends an optimal sizing model based on iterative
technique, to optimize the capacity sizes of different components of hybrid photovoltaic/wind power
generation system using a battery bank. The recommended model takes into account the submodels of
the hybrid system, the Deciency of Power Supply Probability (DPSP) and the Levelised Unit Electricity
Cost (LUEC). The ow chart of the hybrid optimal sizing model is also illustrated. With this incorporated
model, the sizing optimization of grid-independent hybrid PV/wind power generation system can be
accomplished technically and economically according to the system reliability requirements. A case study
is conducted to analyze one hybrid project, which is designed to supply residential household located in

the area of the CDER (Center for Renewable Energy Development) situated in Bouzarah, Algeria (36

0
0
48 N, 3 1 E, 345 m).
2010 Elsevier Ltd. All rights reserved.

Keywords:
Hybrid PV/wind system
Unit sizing
Optimization
Economic evaluation

1. Introduction
Energy consumption in the last century has considerably
increased due to massive industrialization. The forecast of energy
needs for years to come only conrm this trend, especially given
demographic trends and development in some world regions,
particularly in Asia. On the one hand, the elds of traditional energy
resources of fossil origin can be exploited for several decades,
which suggests a situation of energy shortage globally which will
be imminent. On the other hand, nuclear waste poses further
problems in terms of pollution of radioactive waste, decommissioning of old plants and industrial hazard. To meet the energy
needs of todays society, it is necessary to nd solutions and to
diversify them.
Alternative energy resources such as solar and wind have
attracted energy sectors to generate power on a large scale. However,
common drawback with solar and wind energy is their unpredictable
nature and dependence on weather and climatic changes. Standalone photovoltaic (PV) or wind energy systems do not produce
usable energy for considerable portion of time during the year.
In order to efciently and economically utilize the renewable
energy resources, one optimum match design sizing method is
* Corresponding author. Tel.: 213 21 90 15 03; fax: 213 21 90 15.
E-mail address: h.kaabeche@hotmail.com (A. Kaabeche).
0360-5442/$ e see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.energy.2010.11.024

essential. The sizing optimization method can help to endorse the


lowest investment with adequate and full use of the solar system,
wind system and battery bank, so that the hybrid system can work
at optimum conditions in terms of investment and system power
reliability requirement.
Different optimization techniques for hybrid PV/wind systems
sizing have been reported in the literature such as dynamic programming, graphical construction technique, probabilistic approach, articial intelligence methods, multi-objective design, linear
programming and iterative technique [1].
Thus, Musgrove [2] presented a dynamic programming model,
RAPSODY, which is designed to determine optimal operating strategies for a hybrid wind power system incorporating battery storage
and an auxiliary diesel generator. The model takes capital, operating
and maintenance, and fuel costs into account to calculate the
average daily cost of satisfying an electrical load prole. The developed model is provided with an efcient optimizing routine which
allows the user to obtain optimal component sizes for a particular
load prole and wind or solar resource. A graphical construction
technique to follow the optimum combination of PV array and
battery for a hybrid solarewind system has been presented by
Borowy and Salameh [3]. For a given load and a desired LPSP, the
number of batteries and PV modules were calculated based on the
minimum cost of the system. The minimum cost will be at the point
of tangency of the curve that represents the relationship between

A. Kaabeche et al. / Energy 36 (2011) 1214e1222

the number of PV modules and the number of batteries. Then the


optimum sizing of the battery bank and the PV array can be achieved. Another graphical technique has been given by Bin et al. [4],
Kaabeche et al. [5] and Markvart [6], to optimally design a hybrid
solarewind power generation system. However, in both graphical
methods, only two parameters (either PV and battery, or PV and
wind turbine) were included in the optimization process.
Tina et al. [7] presented a probabilistic approach based on the
convolution technique [8] to incorporate the uctuating nature of
the resources and the load, thus eliminating the need for time-series
data, to assess the long-term performance of a hybrid solarewind
system for both stand-alone and grid-connected applications.
Disadvantage of this probabilistic approach is that it cannot represent the dynamic changing performance of the hybrid system.
A methodology for optimum design of a hybrid PV/wind system
has been proposed by Koutroulis et al. [9]. The purpose of the
proposed methodology is to suggest, among a list of commercially
available system devices, the optimum number and type of units
ensuring that the 20-year round total system cost is minimized by
Genetic Algorithms subject to the constraint that the load energy
requirements are completely covered, resulting in zero load rejection. Yang et al. [10] proposed one optimum sizing method based
on Genetic Algorithms by using the Typical Meteorological Year
data. This optimization model is proposed to calculate the system
optimum conguration which can achieve the desired LPSP with
minimum Annualized Cost of System. Another Heuristic technique
based on the evolutionary algorithms have been performed by
Ekren et al. [11] for optimizing size of a PV/wind integrated hybrid
energy system with battery storage. The proposed methodology
uses a stochastic gradient search for the global optimization. In the
study, the objective function is the minimization of the hybrid
energy system total cost.
Bernal-Agustn et al. [12] present a multi-objective optimization
(NPC versus CO2 emissions) for a hybrid solar/wind/diesel system
with battery storage based on Multi-Objective Evolutionary Algorithms (MOEAs). A triple multi-objective optimization to minimize
simultaneously the total cost throughout the useful life of the
installation, pollutant emissions (CO2) and unmet load has been
presented by Dufo-Lpez and Bernal-Agustn [13]. For this task,
a MOEAs and a Genetic Algorithm have been used in order to nd
the best combination of components and control strategies for the
hybrid system. According to the methods proposed by Chedid and
Rahman [14] and Yokoyama et al. [15] the optimal sizes of the PV
and wind power sources and the batteries are determined by
minimizing the system total cost function using linear programming techniques. The total system cost consists of both the initial
cost and yearly operation and maintenance costs.
Yang et al. [16,17] have proposed an iterative optimization
technique following the loss of power supply probability (LPSP)
model for a hybrid solarewind system. The number selection of the
PV module, wind turbine and battery ensures the load demand
according to the power reliability requirement, and the system cost
is minimized. Similarly, an iterative optimization method was
presented by Kellogg et al. [18] to select the wind turbine size and
PV module number needed to make the difference of generated and
demanded power (DP) as close to zero as possible over a period of
time. From this iterative procedure, several possible combinations
of PV/wind generation capacities were obtained. The total annual
cost for each conguration is then calculated and the combination
with the lowest cost is selected to represent the optimal mixture.
Another iterative optimization technique for a stand-alone
hybrid photovoltaic/wind system (HPWS) with battery storage is
presented by Diaf et al. [19]. The main objective of the presented
study is to nd the optimum size of system, able to fulll the energy
requirements of a given load distribution, for three sites located at

1215

Corsica island and to analyze the impact of different parameters on


the system size. In the proposed stand-alone system, the supply of
wind power via an uninterruptible power supply (UPS) is used and
therefore the energy produced by the wind generator can be sent
directly to the load.
In this paper, a grid-independent hybrid PV/wind system optimization model, which utilizes the iterative optimization technique
to follow the Deciency of Power Supply Probability (DPSP) model
and the Levelised Unit Electricity Cost (LUEC) model for power
reliability and system cost respectively, is presented. At the difference of the system conguration used by Diaf et al. [19] in which
the wind generator is taken as the primary source of energy and the
PV generator as the secondary source of energy. This conguration
is good especially for high wind potential regions. The conguration adopted in this study corresponds to the hybrid PV/wind
system in which both wind and PV generators present the primary
source of energy thereby represent the best complementarity
between the two renewable energy sources photovoltaic and wind
and leads to an energy management strategy different from that
presented in [19]. Also, the algorithm developed in this study does
not allow the unballasting of the production in the event of surplus
energy: power produced from natural sources intermittent (wind,
sun) greater than the amount of power consumed and the
maximum acceptable power storage device. Moreover, this algorithm permits the calculation of the excess energy. So, the surplus
energy produced could be used in the production of hydrogen from
an electrolyzer for Long-term energy storage, helping to improve
the total efciency of the hybrid system. Using the DPSP objective
function, the congurations of a hybrid system which can meet the
system reliability requirements can be obtained. There are three
sizing parameters in the simulation, i.e. the capacity of PV system,
the rated power of wind system and the capacity of the battery
bank. The optimum conguration can be identied from the set of
the above obtained congurations by reaching the lowest Levelised
Unit Electricity Cost (LUEC). couter Lire phontiquement.

2. Grid-independent system description


A schematic diagram of a stand-alone hybrid PV/wind system is
shown in Fig. 1. Battery chargers, connected to a common DC bus,
are used to charge the battery bank from the respective PV and
wind input power sources. Depending on the battery charger
technology, the maximum available power can be extracted from
the PV and wind power sources (Maximum Power Point Tracking,
MPPT). The battery bank is used to store the energy surplus and to
supply the load in case of low wind speed and/or irradiation

Fig. 1. Block diagram of a hybrid PV/wind system.

1216

A. Kaabeche et al. / Energy 36 (2011) 1214e1222

conditions. A DC/AC inverter is used to interface the DC battery


voltage to the consumer load AC requirements. The energy
produced from each PV or wind source is transferred to the
consumer load through the battery charger and the DC/AC inverter,
while the energy surplus is used to charge the battery bank.
3. Hybrid PV/wind system model
3.1. PV generator model
The hourly output power of the PV generator with an area Apv
(m2) at a solar radiation on tilted plane module Gt (W/m2), is given
by [20]:

Ppv hpv Apv Gt

(1)

Where hpv represents the PV generator efciency and is given by


[21,22]:

hpv hr hpc

h

i
1  b Tc  Tcref

(2)

Where hr is the reference module efciency, hpc is the power


conditioning efciency which is equal to 1 if a perfect maximum
power tracker (MPPT) is used. b is the generator efciency
temperature coefcient, it is assumed to be a constant and for
silicon cells the range of b is 0.004e0.006 per ( C), Tcref is the
reference cell temperature ( C) and Tc is the cell temperature ( C)
and can be calculated as follows [23]:

Tc Ta NOCT  20=800Gt

(3)
( C)

Where Ta is the ambient temperature


and NOCT is the nominal
cell operating temperature ( C). hpc b, NOCT and Apv, are parameters
that depend upon the type of module used. The data are obtained
from the PV module manufacturers.

3.3. Battery bank model


Battery bank storage is sized to meet the load demand during
non-availability period of renewable energy source, commonly
referred to as days of autonomy. Normally the number of days of
autonomy is taken to be 2 or 3 days. Battery sizing depends on
factors such as maximum depth of discharge, temperature correction, rated battery capacity and battery life. The total capacity of the
battery bank that is to be employed to meet the load is determined
using the following expression [28]:

CB

EL SD
VB DODmax Tcf hB

(6)

Where EL is the load in Wh; SD is the battery autonomy or storage


days; VB is the battery bank voltage; DODmax is the maximum
battery depth of discharge; Tcf is the temperature correction factor
and hB is the battery efciency.
Depending on the PV and wind energy production and the load
power requirements, the state of charge of battery can be calculated
from the following equations:
Battery charging,

SOCt SOCt  1$1  s EGen t  EL t=hinv $hB

(7)

Battery discharging,

SOCt SOCt  1$1  s EL t=hinv  EGen t

(8)

Where SOC(t) and SOC(t-1) are the states of charge of battery bank
(Wh) at the time t and t e 1, respectively; s is hourly self-discharge
rate; EGen(t) is the total energy generated by PV array and wind
generators after energy loss of controller; EL(t) is load demand at
the time t; hinv and hB are the efciency of inverter and charge
efciency of battery bank, respectively. At any time t, the charged
quantity of the battery bank is subject to the following two
constraints:

3.2. Wind turbine system model

SOCmin  SOCt  SOCmax

The wind speed distribution for selected sites as well as the


power output characteristic of the chosen wind turbine are the
factors that have to be considered to determine the wind energy
conversion system power output. Choosing a suitable model is very
important for wind turbine power output simulations. The most
simplied model to simulate the power output of a wind turbine
can be described by [24]:

The maximum charge quantity of battery bank SOCmax takes the


value of nominal capacity of battery bank CB, and the minimum
charge quantity of battery bank SOCmin, is determined by the
maximum depth of discharge (DOD): SOCmin (1 e DOD)$CB.
According to the specications from the manufacturers, the batterys lifetime can be prolonged to the maximum if DOD takes the
value of 30e50%. In this paper, the DOD takes the value of 50%.

.
i
8 h
< PR V2  VC2
VR2  VC2 ;
Pw V
: PR ;
0;

VC  V  VR
VR  V  VF
Otherwise

4. Optimal sizing criteria for hybrid renewable energy system

(4)

Where PR is the rated electrical power; VC is the cut-in wind speed;


VR is the rated wind speed; and VF is the cut-off wind speed. In this
study, the adjustment of the wind prole for height is taken into
account by using the power law that has been recognized as
a useful tool to model the vertical prole of wind speed. The
equation can be described by [25,26]:

VH


V Href

H
Href

(9)

!f

In the existing literature there are various methods to evaluate


the hybrid PV/wind energy system (HPWES) such as energy to load
ratio, battery to load ratio, and non-availability of energy [28]. In
order to select an optimal combination of a HPWES to satisfy the
load demand, evaluation may be carried on the basis of reliability
and economy of power supply. The proposed methodology for
evaluation of HPWES is described in the next section.
4.1. Reliability criteria based on DPSP technique

(5)

Where V(H) is the wind speed at hub height H, m/s; V(Href) is the
wind speed measured at the reference height Href , m/s; a is the
power law exponent. The determination of a becomes very
important. The value of 1/7 is usually taken when there is no
specic site data [26,27].

In this study, reliability of the system is expressed in terms of


deciency of power supply probability (DPSP) which is the probability that an insufcient power supply results when the hybrid
system (PV array, wind power and energy storage) is unable to
satisfy the load demand. The DPSP technique is considered to be the
technical implemented criteria for sizing and evaluating a hybrid
PV/wind system employing a battery bank. The technical model for

A. Kaabeche et al. / Energy 36 (2011) 1214e1222

Fig. 2. Flow chart of the optimal sizing model.

1217

1218

A. Kaabeche et al. / Energy 36 (2011) 1214e1222

hybrid system sizing is developed using the DPSP technique. The


methodology used can be summarized in the following steps:
(a) If the power generated from the PV/wind system is greater
than the load for a particular hour. In this case, the energy surplus is
stored in the battery bank and the new state of charge is calculated
using Eq. (7) until the full capacity is obtained; the remainder of the
available energy is not used.
(b) When the energy demand of the load is greater than the
available energy generated by the PV/wind system, the battery
bank will be used to assure the load demand. In this case, the new
state of charge at hour t is calculated using Eq. (8).
In case (a) when the state of charge of the battery bank reaches
a maximum value, SOCmax, the control system stops the charging
process. The excess power generated (EPG) is an important
parameter, which gives the excess in power generated and unutilized by the system. This value can vary due to the variation of
hourly average demand, insolation, wind velocity and state of
charge of the battery bank. At hour t, the excess power generated
(EPG) can be expressed as follows:

EPGt EGen t  EL t=hinv SOCmax  SOCt  1=hB 


(10)
The relative excess power generated (REPG), expressed as the
ratio of power excess to the sum of load demand during the
considered period is calculated by the following equation:

REPG

T
X

EPGt=

t1

T
X

EL t

(11)

t1

In case (b), if the state of charge of the battery bank decreases to


its minimum level, SOCmin, the control system disconnects the load
and that decit called deciency power supply (DPS) at hour t, can
be expressed as:

DPSt EL t  EGen t SOCt  1  SOCmin hinv

(12)

The deciency of power supply probability (DPSP), for


a considered period T (1 year in this study), can be dened as the
ratio of all the (DPS(t)) values for that period to the sum of the load
demand. This can be dened as [29]:

DPSP

T
X

DPSt=

t1

T
X

EL t

(13)

t1

A DPSP of 1 means that the load will never be satised and the
DPSP of 0 means that the load will be always satised. From the
above-described situations, a program is developed in MATLAB to size
the components for each conguration, for a particular DPSP specied
by the user. The ow chart of HPWES model is illustrated in Fig. 2.
In this program, Ppv,min, Ppv,max and Pw,min, Pw,max represent the
lower and higher limits of the variation interval of the PV and wind
generator rated power, respectively. DPPV and DPW represent the
variation step of the PV and wind power, Dt the simulation step and
DSD is the step of storage days. In this study the maximum number
of storage days, NSD 5.
Table 1
The costs and lifetime aspect for the system components.
Component

Unit Price Maintenance cost Lifetime Real interest Ination


rate f (%)
(US$/W) in the rst year % (year)
rate kd (%)

PV arraya
Wind turbinea
Battery banka
Invertera

4.84
3.000
0.207
0.715

1%
3%
1%
0%

of
of
of
of

price
price
price
price

Mean value of the literature data.

25
20
4
10

Table 2
Specications of the PV module.
Type

Voc (V)

Isc (A)

Vmax (V)

Imax (A)

Pmax (W)

Arco-Solar

21.7

2.7

17.3

2.49

43

The program input data set consists of hourly solar irradiation


on a tilted plane, hourly mean values of ambient temperature and
wind speed, desired DPSP, load power requirements during the
year and specications of the system devices. Using the developed
program, a set of system congurations, which satisfy the system
power reliability requirements, can be obtained. The optimal one is
subsequently predicted on the basis of the minimum LUEC.
4.2. Economic criteria based on LUEC concept
It is pertinent that economic analysis should be made while
attempting to optimize the size of integrated hybrid PV/wind
generation systems favouring an affordable unit price of power
produced. The economical approach, according to the concept of
Levelised Unit Electricity Cost (LUEC), is developed to be the best
indicator of economic protability of system cost analysis in this
study. The LUEC is dened as the total cost of the whole hybrid
system divided by the energy supplied from the hybrid system.
Four main parts are considered: PV array, wind turbine, battery
bank, and the inverter. Then, the LUEC can be expressed by:

LCC  CRF
LUEC$=kWh P8760
t 1 EGen t

(14)

Where LCC is the life cycle cost of the hybrid system; EGen is the
hourly total energy generated by both the wind and PV system; CRF
is the capital recovery factor, a ratio to calculate the present value of
an annuity (a series of equal annual cash ows). The equation for
the capital recovery factor is:



k 1  kd Lp
CRF kd ; Lp d
1 kd Lp 1

(15)

In which kd is the annual real interest rate (8e10%), Lp is the


system life period in years (25 years) [30]. The annual real interest
rate kd is related to the nominal interest rate k0d (the rate at which
you could get a loan) and the annual ination rate f by the equation
given below.

kd

k0d  f
1f

(16)

According to the studied system, the life cycle cost (LCC) takes
into account the capital cost (Ccap), the present value of replacement cost (Crep) and the present value of maintenance cost (Cmain).
4.2.1. The initial capital cost
The initial capital cost of each system component consists of the
component price, the cost of civil work, installation and the
connections. In this study, the civil work and installation costs are
taken as 40% of PV generator price for PV part and 20% of wind

Table 3
Specications of the wind turbine.

4
Type

Rated
power
(W)

Cut-in
speed
VC (m/s)

Rated
speed
VR (m/s)

Cut-off
speed
VF (m/s)

Tower
high (m)

AIR 403

400

12

25

10

A. Kaabeche et al. / Energy 36 (2011) 1214e1222

Table 4
Specications of the single battery.
Type

Nominal
capacity (Ah)

Voltage (V)

Round-trip
efciency

DOD (%)

Varta Solar

100

12

0.85

50

(17)

Where (CPV, Cunit,PV) are the total capacity (W) and unit cost
($/W) of PV array respectively; (CW, Cunit,W)are the total capacity
(W) and unit cost ($/W) of the wind machine respectively; (CB,
Cunit,B) are the total capacity (Wh) and unit cost ($/Wh) of the
battery bank respectively; and (CINV, Cunit,INV)are the nominal
capacity (W) and unit cost ($/W) of the inverter respectively; and C0
is the total constant cost including the cost of civil work and
installation.
4.2.2. The present value of replacement cost
The present value of replacement cost of a system component is
the present value of all the replacement costs occurring throughout
the system lifetime. As the life period of wind generator, battery
banc and inverter are shorter than PV system; the replacement cost
of the wind generator, the batteries and the inverter have to be
included in the cost analysis of the hybrid system. Considering the
ination rate of component replacements f0 and real interest rate
(kd), the present value of replacement cost (Crep) can be determined
as follows [31]:


Nrep 
X
1 f0 Ni =Nrep 1
i1



1

1f1
1kd

L p
;

CO&M0  Lp ;

for kd sf1
for kd f1

900
800
700
600
500
400
300
200
100
0

1000

2000

3000

4000

5000

6000

7000

8000 8760

6000

7000

8000 8760

6000

7000

8000 8760

Time (hours)

14

12

10

(18)

1 kd

Where Cnom is the nominal capacity of the replacement system


component (wind generator in (W); battery bank in (Wh) and
inverter in (W)); Cunit is the unit component cost (wind generator
($/W); battery bank ($/Wh) and inverter ($/W)) and Nrep is the
number of component replacements over the system life period.

4.2.3. The present value of operation and maintenance cost


In its general form, the present value of operation and maintenance cost of the hybrid system CO&M,Hyb is expressed as [32]:

0
0

1000

2000

3000

4000

5000

Time (hours)

c
350

45
40

Ambient temprature (C)

300
250

Load power (W)

1f1
kd f1

1000

Wind speed (m/s)

Crep Cunit Cnom

Where f1 is the ination rate for operations; kd is the annual real


interest rate and Lp is the system life period in years. CO&M is the
operation and maintenance cost in the rst year. It can be given as
a fraction "k" of the initial capital cost (CIC ).CO&M is expressed as:

Hourly solar irradiation (Wh/m2)


 
 

CPV  CUnit;PV CW  CUnit;W CB  CUnit;B



CINV  CUnit;INV C0

CO&M0

(19)

generator price for wind part. Then the initial capital cost for the
hybrid system, (CIC) is given by:

CIC

CO&M;Hyb

1219

200
150
100

35
30
25
20
15
10

50
5
0

0
2

10

12

14

16

Time (hours)
Fig. 3. Hourly load prole.

18

20

22

24

1000

2000

3000

4000

5000

Time (hours)
Fig. 4. Meteorological conditions for optimal design. (a) solar irradiation on horizontal
plane, (b) wind speed and (c) ambient temperature.

1220

A. Kaabeche et al. / Energy 36 (2011) 1214e1222


45

70
DPSP=1 %
DPSP=3 %
DPSP=5 %

35

1 day storage

50

Num ber of PV M odules

Num ber of PV M odules

60

DPSP=1 %
DPSP=3 %
DPSP=5 %

40

40

30

20

5 days storage

30
25
20
15
10

10

0
400

800

1200

1600

2000

2400

2800

3200

3600

0
400

4000

Rated Power of Wind Turbine (W)

800

1200

1600

2000

2400

2800

3200

3600

4000

Rated Power of Wind Turbine (W)

Fig. 5. System congurations for different DPSP for 1 day of autonomy of the battery
bank.

Fig. 7. System congurations for different DPSP for 5 days of autonomy of the battery
bank.

CO&M0 k  CIC

research is that represented on Fig. 3. This hourly energy distribution is considered identical for every day of the year and corresponds
to the load prole generally encountered in remote areas in Algeria.
Hourly data of solar irradiation on the horizontal plane, wind
speed as well as ambient temperature, plotted in Fig. 4 during the
year 2003, were recorded using a properly data-acquisition system
installed at the CDER [33]. The annual wind energy potential for
Bouzarah at 10 m height is 187 kWh/m2 and the annual total solar
radiation on the horizontal surface is 1626 kWh/m2. On a comparative basis between the solar insolation and wind distribution of
the site, there is a great scope for generating power from solar for
longer periods in a year. Thus, the data recorded are used in system
unit sizing and the generation is assumed to keep constant in each
hour interval.

(20)

In this study it is assumed that all prices escalate at the same


rate, and use annual real interest rate rather than the nominal
interest rate.
The following unit price, maintenance cost and lifetime of each
component (PV array, wind generator, battery bank and inverter) in
this study are assumed as mentioned in Table 1. The conguration
with the lowest (LUEC) is taken as the optimal one from the set of
congurations which guarantee the required reliability of power
supply.

5. Results and discussion


5.1. Case study

5.2. Impact of power reliability on system congurations

The recommended methodology has been applied to analyze


a stand-alone hybrid PV/wind energy system, which is designed to
supply residential household located in the area of the CDER (Center
for Renewable Energy Development) situated in Bouzarah, Algeria


(36 480 N, 3 10 E, 345 m). The technical characteristics of the PV
module and wind turbine as well as the battery used in the studied
project are listed in Tables 2, 3 and 4. The load prole adopted in this

The relationships between system reliabilities and system


congurations are studied. Figs. 5, 6 and 7 show the results of the
relationship between system reliabilities or DPSP values and
system congurations for different days of autonomy of the battery
bank. Fig. 5 shows the relationships for a one day-storage battery
bank. In this gure, the curves are hyperbolic nature. Each point of
them represents a couple (Number of PV modules, wind turbine

45

Levelised unit electricity cost ($ / kWh)

DPSP=1 %
DPSP=3 %
DPSP=5 %

40

Num ber of PV M odules

35

3 days storage
30
25
20
15
10
5
0
400

800

1200

1600

2000

2400

2800

3200

3600

4000

Rated Power of Wind Turbine (W)


Fig. 6. System congurations for different DPSP for 3 days of autonomy of the battery
bank.

LUEC for 1 day storage


LUEC for 3 days storage
LUEC for 5 days storage

6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
1

400 800 1200


1600 2000

Rat ed Power
o

2400 2800 3200

f W ind Turbin

3600 4000

e (W)

20

Num

40

er o

80

u
Mod
f PV
60

les

Fig. 8. System congurations and Levelised Unit Electricity Cost for DPSP 1%.

A. Kaabeche et al. / Energy 36 (2011) 1214e1222

1221

LUEC for 1 day storage


LUEC for 3 days storage
LUEC for 5 days storage

5
4.5
4
3.5
3
2.5
2
1.5
1
0.5

4
3.5
3
2.5
2
1.5
1

0
400 800
1200 1600
2000

Rat ed Pow
er

2400 2800

of W ind Tu

3200 3600 4000 10

b
Num

rbine (W)

30

20

50

40

PV
r of

60

du
Mo

les

5.3. Impact of system congurations on the LUEC


The congurations meeting different desired DPSP requirements
under different battery capacities are obtained from the simulation
results. After the technical criteria, the Levelised Unit Electricity Cost
(LUEC) is utilized as the economic benchmark. The simulation results
are demonstrated, and the relationships between the LUEC and
system congurations are analyzed. In Figs. 8, 9 and 10, the curves
given by the solid symbols represent the Levelised Unit Electricity
Cost (LUEC) under different congurations. Obviously, one point

4.5

10

Deficiency of power supply probability (%)

power) that guarantees the desired energy autonomy. In the case of


a zero value of the DPSP, the corresponding curve is called curve of
autonomy of the system: each point of this curve represents
a combination which ensures the total autonomy of the system. The
areas above the curves are also congurations that can ensure the
required power reliability. It also shows that when the system
reliability is higher; the system conguration (PV module and wind
turbine power) is higher too for the same capacity of battery bank.
A similar situation happens to the system for three and ve daysstorage battery bank (Figs. 6 and 7), but compared to the system
with one day-storage battery bank, the PV module and wind
turbine power are more moderate. It means the hybrid system with
more batteries (5 days of storage capacity) can meet the load
demand with less supply failure.

0.5
0

Fig. 9. System congurations and Levelised Unit Electricity Cost for DPSP 3%.

Levelised unit electricity cost ($ / kWh)

1 day storage
3 days storage
5 days storage

4.5

Minimal valiues of the LUEC

Levelised unit electricity cost ($ / kWh)

LUEC for 1 day storage


LUEC for 3 days storage
LUEC for 5 days storage

Fig. 11. Minimal values of LUEC vs. DPSP [fx3]for different days of autonomy of the
battery bank.

with the minimum LUEC value occurs in each curve which means
the best conguration for one certain DPSP value and one certain
battery bank. This conguration is considered as the optimal one
which meets the system reliability requirement with the lowest
LUEC value. On the other hand, a meticulous examination into Figs.
8, 9 and 10 shows that the lowest LUEC is found when the capacity of
wind turbine and the number of PV modules are both moderate. It is
also shown that the LUEC for one day battery storage is lower than
three and ve days for the desired DPSP of 1%, 3% and 5% for the
studied case because batteries are much more expensive with
a short lifespan.
5.4. Impact of power reliability on the LUEC
The minimal values of the Levelised Unit Electricity Costs for
different DPSP (power reliability requirements) are calculated by
the proposed optimal sizing method. The results for hybrid system
are demonstrated in Fig. 11. The best congurations for one, three
and ve days of autonomy of the battery bank happen for a DPSP of
10% (lowest value). It conrms that higher power reliable systems
are more expensive than lower requirement systems. On the other
hand, a deeper examination into Fig. 11 shows that the optimal
congurations for one day-storage capacity have the lowest LUEC
than the three and ve days-storage capacity because batteries are
much more expensive with a short lifespan. Thus, choosing an
optimal system conguration according to system power reliability
requirements can help save investment and avoid blind capital
spending sizing.
6. Conclusions

4
3.5
3
2.5
2
1.5
1
0.5
400 800 1200
1600 2000

Rat ed Power of W

2400 2800 3200

3600 4000

ind Turbine (W)

10

Num

20

ber o

30

40

es
odul
PV M

Fig. 10. System congurations and Levelised Unit Electricity Cost for DPSP 5%.

In this paper, a grid-independent hybrid PV/wind system optimization model, which utilizes the iterative optimization technique
to follow the Deciency of Power Supply Probability (DPSP) model
and the Levelised Unit Electricity Cost (LUEC) model for power
reliability and system cost respectively, is presented. The conguration adopted in this study corresponds to the hybrid PV/wind
system in which both wind and PV generators present the primary
source of energy thereby represent the best complementarity
between the two renewable energy sources photovoltaic and wind
and leads to an energy management strategy different from that
presented in the previous studies. The recommended model
consists of three main parts: the submodel of the hybrid system, the
technical submodel developed according to the Deciency of Power

1222

A. Kaabeche et al. / Energy 36 (2011) 1214e1222

Supply Probability (DPSP) technique for system reliability evaluation and the economic submodel developed based on the concept
of the Levelised Unit Electricity Cost (LUEC) which is considered as
a good indicator of economic protability in the eld of renewable
energy. A set of congurations meeting the desired DPSP can be
obtained by using the DPSP submodel. The conguration with the
lowest LUEC gives the optimal one.
A case study is conducted to analyze one hybrid project, which is
designed to supply residential household located in the area of the
CDER (Center for Renewable Energy Development) situated in


Bouzarah, Algeria (36 480 N, 3 10 E, 345 m). The algorithm input
data set consists of hourly solar irradiation on the horizontal plane,
wind speed as well as ambient temperature recorded at Bouzarah
(Algeria) for the year 2003, the desired DPSP, load power requirements during the year and specications of the system devices.
The grid-independent hybrid PV/wind system is simulated by
running the developed program and the relationships between
system power reliability and system congurations have been
studied. The optimal congurations of the hybrid system are determined in terms of different desired system reliability requirements (DPSP) and the Levelised Unit Electricity Cost (LUEC).
References
[1] Zhou W, Lou C, Li Z, Lu L, Yang H. Current status of research on optimum sizing
of stand-alone hybrid solarewind power generation systems. Applied Energy
2010;87(2):380e9.
[2] Musgrove ARD. The optimization of hybrid energy conversion system using
the dynamic programming model-RAPSODY. International Journal of Energy
Research 1988;12:447e57.
[3] Borowy BS, Salameh ZM. Methodology for optimally sizing the combination of
a battery bank and PV array in a wind/PV hybrid system. IEEE Transactions on
Energy Conversion 1996;11(2):367e73.
[4] Bin A, Hongxing Y, Hui S, Xianbo L. Computer aided design for PV/wind hybrid
system. Renew Energy 2003;28:1491e512.
[5] Kaabeche A, Belhamel M, Ibtiouen R, Moussa S, Benhadadi MR. Optimisation
dun systme hybride (Eolien e Photovoltaque) totalement autonome. Revue
des Energies Renouvelables 2006;(3):199e209.
[6] Markvart T. Sizing of hybrid PV/wind energy systems. Solar Energy 1996;59
(4):277e81.
[7] Tina G, Gagliano S, Raiti S. Hybrid solar/wind power system probabilistic
modeling for long-term performance assessment. Solar Energy 2006;80:578e88.
[8] Karaki SH, Chedid RB, Ramadan R. Probabilistic performance assessment of
autonomous solar-wind energy conversion systems. IEEE Transactions on
Energy Conversion 1999;14(3):766e72.
[9] Koutroulis E, Kolokotsa D, Potirakis A, Kalaitzakis K. Methodology for optimal
sizing of stand-alone photovoltaic/wind-generator systems using genetic
algorithms. Solar Energy 2006;80(9):1072e88.
[10] Yang HX, Zhou W, Lu L, Fang ZH. Optimal sizing method for stand-alone
hybrid solarewind system with LPSP technology by using genetic algorithm.
Solar Energy 2008;82(4):354e67.

[11] Ekren O, Ekren BY. Size optimization of a PV/wind hybrid energy conversion
system with battery storage using simulated annealing. Applied Energy
2010;87(2):592e8.
[12] Bernal-Agustn JL, Dufo-Lpez R, Rivas-Ascaso DM. Design of isolated hybrid
systems minimizing costs and pollutant emissions. Renew Energy 2006;31
(14):2227e44.
[13] Dufo-Lpez R, Bernal-Agustn JL. Multi-objective design of PVewinde
dieselehydrogenebattery systems. Renew Energy 2008;33(12):2559e72.
[14] Chedid R, Rahman S. Unit sizing and control of hybrid wind-solar power
systems. IEEE Transactions on Energy Conversion 1997;12(1):79e85.
[15] Yokoyama R, Ito K, Yuasa Y. Multi-objective optimal unit sizing of hybrid
power generation systems utilizing PV and wind energy. Journal of Solar
Energy Engineering 1994;116:167e73.
[16] Yang HX, Burnett J, Lu L. Weather data and probability analysis of hybrid
photovoltaic/wind power generation systems in Hong Kong. Renewable
Energy 2003;28:1813e24.
[17] Yang HX, Lu L, Zhou W. A novel optimization sizing model for hybrid
solarewind power generation system. Solar Energy 2007;81(1):76e84.
[18] Kellogg WD, Nehrir MH, Venkataramanan G, Gerez V. Generation unit sizing
and cost analysis for stand-alone wind, photovoltaic and hybrid wind/PV
systems. IEEE Transactions on Energy Conversion 1998;13(1):70e5.
[19] Diaf S, Belhamelb M, Haddadic M, Louchea A. Technical and economic
assessment of hybrid photovoltaic/wind system with battery storage in Corsica Island. Energy Policy 2008;36(2):743e54.
[20] Marvart T. Solar Electricity. USA: John Wiley; 2000.
[21] Habib MA, Said S, El-Hadidy MA, Al-Zaharna I. Optimization procedure of
a hybrid photovoltaic/wind energy system. Energy 1999;24:919e29.
[22] Kolhe M, Agbossou K, Hamelin J, Bose TK. Analytical model for predicting the
performance of photovoltaic array coupled with a wind turbine in a standalone renewable energy system based on hydrogen. Renew Energy 2003;28
(5):727e42.
[23] Diaf S, Notton G, Belhamel M, Haddadi M, Louche A. Design and techno economical optimization for hybrid PV/wind system under various meteorological conditions. Applied Energy 2008;85(10):968e87.
[24] Pallabazzer R. Evaluation of wind generator potentiality. Solar Energy
1995;55:49e59.
[25] Lu L, Yang HX, Burnett J. Investigation on wind power potential on Hong Kong
islands e an analysis of wind power and wind turbine characteristics.
Renewable Energy 2002;27:1e2.
[26] Ilinka A, McCarthy E, Chaumel JL, Rtiveau JL. Wind potential assessment of
Quebec Province. Renewable Energy 2003;28(12):1881e97.
[27] Mostafaeipour A. Feasibility study of harnessing wind energy for turbine
installation in province of Yazd in Iran. Renewable Sustainable Energy Rev.
2010;14(4):93e111.
[28] Deshmukha MK, Deshmukh SS. Modeling of hybrid renewable energy
systems. Renewable and Sustainable Energy Reviews 2008;12:235e49.
[29] Prasad AR, Natarajan E. Optimization of integrated photovoltaic/wind power
generation systems with battery storage. Energy 2006;31(12):1943e54.
[30] Nandi KS, Ghosh HR. Prospect of windePVebattery hybrid power system
as an alternative to grid extension in Bangladesh. Energy 2010;35
(7):3040e7.
[31] Soras C, Makios V. A novel method for determining the optimum size of standalone photovoltaic systems. Solar Cells 1988;25:127e42.
[32] Groumpos PP, Papageorgiou G. An optimal sizing method for stand-alone
photovoltaic power systems. Solar Energy 1987;38(5):341e51.
[33] Division Energie Eolienne, Centre de Dveloppement des Energies Renouvelables, B.P. 62 16340 Bouzareah, Algiers, Algeria.