Available online at [Link].

com

Solar Energy 84 (2010) 1124–1136

[Link]/locate/solener

Design of isolated renewable hybrid power systems

E.S. Sreeraj a, Kishore Chatterjee a, Santanu Bandyopadhyay b,*
a
Department of Electrical Engineering, Indian Institute of Technology Bombay, Mumbai 400 076, India
b
Department of Energy Science and Engineering, Indian Institute of Technology Bombay, Mumbai 400 076, India

Received 7 January 2010; received in revised form 15 March 2010; accepted 17 March 2010
Available online 10 April 2010

Communicated by: Associate Editor Mukund Patel

Abstract

Isolated electrical power generating units can be used as an economically viable alternative to electrify remote villages where grid
extension is not feasible. One of the options for building isolated power systems is by hybridizing renewable power sources like wind,
solar, micro-hydro, etc. along with appropriate energy storage. A method to optimally size and to evaluate the cost of energy produced
by a renewable hybrid system is proposed in this paper. The proposed method, which is based on the design space approach, can be used
to determine the conditions for which hybridization of the system is cost effective. The simple and novel methodology, proposed in this
paper, is based on the principles of process integration. It finds the minimum battery capacity when the availability and ratings of various
renewable resources as well as load demand are known. The battery sizing methodology is used to determine the sizing curve and thereby
the feasible design space for the entire system. Chance constrained programming approach is used to account for the stochastic nature of
the renewable energy resources and to arrive at the design space. The optimal system configuration in the entire design space is selected
based on the lowest cost of energy, subject to a specified reliability criterion. The effects of variation of the specified system reliability and
the coefficient of correlation between renewable sources on the design space, as well as the optimum configuration are also studied in this
paper. The proposed method is demonstrated by designing an isolated power system for an Indian village utilizing wind–solar photo-
voltaic-battery system.
Ó 2010 Elsevier Ltd. All rights reserved.

Keywords: Design space; Renewable hybrid system; Chance constrained method; Process integration; Battery sizing

1. Introduction India do not have access to electricity ( Govt. of India,

2001). As electricity is important for rapid economic
Isolated power systems using renewable energy sources growth and poverty alleviation, Indian government has
like wind, solar, biomass, micro-hydro, etc. can be utilized decided to provide electricity access to all households.
to provide electricity for remote locations where grid exten- Along with rapid expansion in conventional power genera-
sion is not feasible and/or economical. It was estimated tion, Indian government has also decided to go for power
that more than 1500 million people around the world had generation from new and renewable sources. For many
no access to electricity in 2005 (International Energy remote non-electrified rural areas, power generation from
Agency, 2006). A vast majority of them are from Sub-Sah- stand-alone systems is cheaper than grid extension.
aran Africa and South Asia, where electrification rates are The National Electricity policy of India states that wher-
only 25.8% and 51.8%, respectively (International Energy ever it is neither cost effective nor optimal to provide grid
Agency, 2006). In 2001, about 44% of the households in connectivity, decentralized distributed generation facilities
together with local distribution network would be provided
so that every household gets access to electricity (Govt. of
*
Corresponding author. Tel.: +91 22 25767894; fax: +91 22 25726875. India, 2005). Non-conventional sources of energy could be
E-mail address: santanu@[Link] (S. Bandyopadhyay).

0038-092X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/[Link].2010.03.017
 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136 1125

Nomenclature

AP total array area (m2) Pr rated electrical power of the wind power gener-
ACC annualised capital cost (Rs) ating unit (W)
AOM annualised operation and maintenance cost (Rs) Pw power generated by the wind turbine (W)
B battery capacity (kW h) Pj power generated by the jth power source (W)
a confidence level QB energy stored in the battery (kW h)
C0 capital cost of the component (Rs) Qmax1 maximum stored energy of the battery energy
COE cost of energy (Rs/kW h) before tref (W h)
CRF capital recovery factor Qmax2 maximum stored energy of the battery energy
D load demand (W) after tref (W h)
d discount rate Qmin minimum stored energy by the battery (W h)
Dactual the deterministic demand to be met taken in the qj(t) multiplication factor for finding out power pro-
chance constrain (W) duced
Dt time step for the simulation (h) qij coefficient of correlation between the power
f factor representing net charging/discharging available from ith and jth sources
efficiency rP j ðtÞ standard deviation of power available from ith
fi factor that represents inverter efficiency power source (W)
H wind turbine height (m) T time horizon for the simulation (h)
Hi specified meteorological mast height (m) t time (h)
IT total radiation incident on the array (W/m2) tmax1 time at which battery reaches maximum energy
lP j ðtÞ mean of power available from ith power source before tref (h)
(W) tmax2 time at which battery reaches maximum energy
n life of the component (years) after tref (h)
gc battery charging efficiency tref time at which battery reaches minimum energy
gd battery discharging efficiency (h)
gi inverter efficiency vc cut-in wind speed (m/s)
gP photovoltaic system efficiency vf cut-off wind speed (m/s)
Pnet net power available at the dc bus (W) vi wind speed at reference height Hi (m/s)
Pdu dumped excess power (W) vr rated wind speed (m/s)
Ppv power generated by the photovoltaic array (W) X depth of discharge of battery
z power law exponent

utilized even where grid connectivity exists, provided it is to complimentary nature of wind and solar power and
found to be cost effective (Govt. of India, 2003). the cost effectiveness of hybridizing these two systems, a
Isolated systems using renewables can be powered by a vast literature deals with wind–solar hybrid systems and
single or a combination of renewable power sources. The is recently reviewed by Deshmukh and Deshmukh (2008).
power available from the renewable sources is stochastic The performance of a hybrid system depends upon
in nature. However, some of the renewable resources like proper sizing of the system. Design and simulation fol-
solar and wind are complementary in nature. This means lowed by optimization are main steps involved in sizing
that during seasons of low insolation, the wind speed is an isolated hybrid system. The size of a system, that can
typically higher and the wind speed is generally low for sea- supply the required power demand, can be determined by
sons of high insolation. In Fig. 1, monthly average value of simulating the entire system using the resource and the
wind speed, monthly average total rainfall, and monthly demand data. Optimization of the entire system may be
average daily global insolation for an Indian town, Rat- performed to arrive at a sizing which satisfies certain cost
nagiri during each month is plotted (Mani and Rangarajan, and reliability criteria. This is typically achieved by mini-
1982). From Fig. 1, it can be observed that the amount of mizing the net present cost of the system or the levelized
rainfall and wind speed is negatively correlated to solar cost of generated energy. The reliability of the power pro-
insolation. During monsoon months, when the rainfall is duced by the hybrid system is also generally included in the
high, the wind speed is also high and solar insolation is optimization process either in the form of constrains or as
low. The wind speed and rainfall are low for summer another variable to be maximized. In the latter case, a
months when the insolation is high. Thus, it is apparent multi-objective optimization routine has to be revoked
from Fig. 1 that it may be advantageous to make a wind and the solution set generally consists of a set of Pareto-
solar or a micro-hydro-solar hybrid power system. Due optimal configurations, out of which a suitable one has
1126 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136

1000 10
25
insolation

monthly average of daily global insolation(kWh/m 2)

900 wind speed
rainfall
800 20 8

monthly average windspeed(km/h)

monthly average rainfall(cm)
700

600 15 6

500

400 10 4

300

200 05 2

100

0
0 2 4 6 8 10 12
month

Fig. 1. Monthly average rainfall, wind speed and solar insolation at Ratnagiri, India.

to be selected (Gavanidou and Bakirtzis, 1992). The indices the individual components due to hardware failures. For
that are generally used to evaluate system reliability are: (a) stand-alone photovoltaic–wind hybrid system, the stochas-
number of days of autonomy (Natarajan and Rajendra tic nature of the resources is modelled as a three event Mar-
Prasada, 2006), (b) loss of load probability (Gavanidou kov process by Bagul et al. (1996). This is an extension of
and Bakirtzis, 1992; Beyer and Langer, 1996), (c) loss of the two state Markov process, proposed by Bucciarelli
power supply probability (Ai et al., 2003; Yang et al., (1984). A transformation theorem based method for sizing
2003), and (d) unmet load (Chedid and Rahman, 1997). of a stand-alone photovoltaic–wind system has been pro-
There are primarily two different approaches to design posed by Abousdar and Ramkumar (1990, 1991) where
and simulate a hybrid system: deterministic and probabilis- solar insolation, wind speed and load are considered as
tic (Deshmukh and Deshmukh, 2008). In deterministic random variables.
approaches, the renewable energy resources and the The set of feasible configurations that can meet the given
demand are considered as deterministic quantities and their load demand forms the design space. The concept of design
variation with respect to time is assumed to be known. space is introduced for optimized sizing of solar hot water
Usually, time for which the system has to be analyzed (time systems by Kulkarni et al. (2007, 2008, 2009). For isolated
horizon), is divided into smaller time periods, during which power systems, the representation of the design space is
the resources availability and load are assumed to be con- used for diesel generator–battery systems (Arun et al.,
stant. In deterministic methods, the chronological sequence 2008), photovoltaic-battery systems (Arun et al., 2007),
of the data is extremely important. Sometimes the calcula- wind-battery systems (Roy et al., 2009) and wind–photo-
tion based on the worst case scenario, say worst month, voltaic hybrid systems (Roy et al., 2007). The design space
can also be used in designing of the system (Beyer and representation for a photovoltaic battery system incorpo-
Langer, 1996; Protogeropoulos et al., 1997; Celik, 2002a). rating uncertainty through chance constrained program-
Typical weather year for a particular location can also be ming has been presented by Arun et al. (2009). Chance
used to design a renewable power system (Yang et al., constrained programming approach (Charnes and Cooper,
2003). However, the typical year data for more than one 1959), dealing with stochastic programming, has been
renewable resource are difficult to obtain. Celik (2002b) applied in various fields of engineering to deal with uncer-
has employed synthetically simulated weather data to sim- tainty (Rao, 1980; Changchit and Terrell, 1993; Azaiez
ulate a hybrid system. These approaches are either compu- et al., 2005; Li et al., 2008).
tationally intensive (if it uses data for a long time period, As mentioned earlier, deterministic and probabilistic
say many years) or produce sub-optimal results (if it uses approaches for system design and simulation have their
the worst month scenario for calculation). own merits and demerits. Design methodology utilizing
In probability-based approaches, energy generated by deterministic methods are either computationally intensive
power sources, and load demand in some cases, are consid- or produce sub-optimal results depending upon the type
ered as random variables. Some methods may not consider and amount of data used. Whereas, design methodologies
the chronological sequence of the data, which makes them based on stochastic methods are simple. However, deter-
less accurate. Karaki et al. (1999) proposed a probabilistic ministic methods take into account the chronological
model for an autonomous photovoltaic–wind system with sequence of data, and hence, can produce accurate results
several wind machines, accounting uncertainties related compared to that of stochastic methods. In order to over-
to fluctuations in primary energy output, and outages of come the limitation of inaccuracy in stochastic methods
 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136 1127

reported in the literature (e.g., Bucciarelli, 1984), a new sto- system with power sources connected to either ac bus or
chastic method to optimally size a renewable hybrid system dc bus and load connected to the ac bus is shown in
with more than one power source and to determine the cost Fig. 2. During some instants the battery may be fully
of energy produced is proposed in this paper. The method charged and the net power produced by the sources is
combines the advantages of both the deterministic and greater than the demand. The excess power is dumped
probabilistic approaches while retaining inherent simplicity using the dump load and it is connected to the dc bus.
of stochastic method. The system chosen is constrained to The system considered has n power sources, with the
supply power to the load satisfying certain reliability crite- first m sources (1st to mth) connected to the ac bus and
rion. The probabilistic approach is used to incorporate remaining n  m sources (m + 1st to nth) connected to
uncertainty in available energy from each power source. the dc bus. Let Pj(t) represents the power generated by
The variability associated with power available from each the jth source. The net power available at the dc bus that
power source along with coefficient of correlation between can be used to charge the battery (Pnet(t)) is the sum of
them is used in the analysis as these quantities affect the the power generated by the sources connected to the dc
performance and reliability of the system. The emphasis bus and the power available from the ac bus.
is on the system design part and the optimum configuration !
that is found by doing a search in the design space. The X
n X
m
P net ðtÞ ¼ P j ðtÞ þ P j ðtÞ  DðtÞ fi ðtÞ ð1Þ
proposed sizing method incorporates a simple and novel j¼mþ1 j¼1
methodology to find the battery capacity when renewable
resource availability, ratings of renewable power sources
The term inside the bracket is the net power generated at
and load demand are known. This methodology also helps
the ac bus. D is the power required by the load and fi rep-
to plan further load growth, as it gives the durations for
resents the efficiency associated with the inverter (gi).
which load can be increased without additional capacity
Depending on the sign of the net power generated in the
growth. When the random nature of the resources is con-
ac bus, that is the difference of power generated by the var-
sidered, chance constrained programming is used to arrive
ious power sources and the load demand, the power flow
at the design space. The effectiveness of the proposed
will take place from ac bus to dc bus or vice versa. The
method is demonstrated by designing an isolated power
power loss in inverter is accounted by the factor, fi which
system for an Indian village with wind and solar power
represents the inverter efficiency and is given as follows:
as potential options.
P
m
fi ðtÞ ¼ gi when P j ðtÞ P DðtÞ
2. Design space generation with deterministic approach
j¼1 ð2Þ
¼ g1i otherwise
System sizing methodology for a renewable hybrid sys-
tem following a deterministic approach is discussed in this During some time periods, the net energy that can be
section. Renewable power sources can produce ac or dc produced by the system may be greater than the sum of
electric power depending on the type of generator and the energy consumed in the load and that can be stored
power electronic interfaces connected to the system. The in the battery. This excess energy has to be dumped or
hybrid system considered here has both ac and dc power lower amount of energy has to be produced by the power
sources. The average power available from each power sources. The hybrid system considered in this paper has a
source during every time step is assumed to be known from dump load connected to the dc bus and the excess energy
the resource data and the capacity of the power source. The is dumped using that load. The energy transfer across the
time series simulation, based on the energy balance of the battery bank ðdQB ðtÞ=dtÞ is proportional to the net power
overall system is performed to arrive at the required battery available at the dc bus (Pdc(t)) minus the dumped power
capacity. The schematic of an isolated renewable hybrid (Pdu(t)).

Source-m+1
Source-1 .
.
.
. .
. .
. .
. .
. .
. .
. .
. .

Converter
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .

Source-m Source-n
. .
.
.
.

AC Load Battery bank

Dump load
AC Bus DC Bus

Fig. 2. Schematic of renewable energy based isolated power system.

1128 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136

dQB ðtÞ
¼ ðP net ðtÞ  P du ðtÞÞf ðtÞ ð3Þ
dt Qmax2

where f(t) represents the efficiencies associated with the

charging (gc) and discharging processes (gd) of the battery. Qmax1

f ðtÞ ¼ gc when P net P 0

Battery energy
ð4Þ
¼ g1d when P net < 0 Dump
energy
A B tmax1 C tmin D tmax2 E
The change in stored energy for the time period (Dt) may time
be expressed as follows:
Z tþDt
QB ðt þ DtÞ ¼ QB ðtÞ þ ðP net ðtÞ  P du ðtÞÞf ðtÞdt ð5Þ
t No dumping
Qmin
during this
For a relatively small time period, the above equation period
can be approximated as
Fig. 3. Typical variation of stored energy level.
QB ðt þ DtÞ ¼ QB ðtÞ þ ðP net ðtÞ  P du ðtÞÞf ðtÞDt ð6Þ
Combining Eq. (1) and Eq. (6), we get:
QB ðt þ DtÞ ¼ QB ðtÞ cated by the stored energy level at the end of the time hori-
! ! zon, QB(t = T).
X
n X
m
þ P j ðtÞ þ P j ðtÞ  DðtÞ fi ðtÞ  P du ðtÞ f ðtÞDt ð7Þ Fig. 3 shows the variation of stored energy during the
j¼1þm j¼1 entire time horizon, T. Let, Qmin is the minimum stored
Eq. (7) represents the energy balance across the battery energy and it occurs at tref (point ‘C’ in Fig. 3). On the
and is solved to find the battery capacity required to meet other hand, Qmax1 is the maximum stored energy level
the load. The following additional constrains are to be before tref and Qmax2 is the maximum stored energy level
satisfied: after tref. They occur at tmax1 (point ‘B’ in Fig. 3) and tmax2
(point ‘D’ in Fig. 3), respectively. Since no energy is
QB ðtÞ P 0 8 t ð8Þ dumped in this initial process, the extra energy generated
QB ðt ¼ 0Þ ¼ QB ðt ¼ T Þ ð9Þ has to be dumped at appropriate time so that Eq. (9) is sat-
isfied. This is done to reduce the required battery capacity
Eq. (8) ensures that the battery charge level is non-neg- to its minimum value. The excess energy, QB(t = T) can be
ative. Eq. (9) represents the repeatability of the battery dumped at various instants to reduce the peak of the stored
state of energy over the time horizon, T. The power avail- energy of the hypothetical battery keeping the minimum of
ability and the load at each time step are repeated in the the stored energy of the battery to occur at reference time
next cycles, and if the initial state of charge of the battery itself. The peak of the stored energy level that occurs after
is not equal to the final, the battery charge will either accu- tref, i.e. Qmax2, can be reduced by dumping the excess
mulate or deplete over the time. A methodology is pro- energy between tref and tmax2, and hence the peak of the
posed in Section 2.1 to find the capacity of the battery energy stored by the actual battery after tref can be lowered
when the ratings of each power source are known. by the amount equal to the dumped energy. This is because
of the excess energy QB(t = T) that is available between
2.1. Methodology for finding the battery capacity tmax2 and tref. On the other hand, the peak of the stored
energy level that occurs before tref, i.e. Qmax1, cannot be
Since the power available from the sources are known, reduced as there is no excess energy between tmax1 and tref.
Eq. (7) is solved to find the stored energy at the end of each The energy stored in an actual battery cannot go negative
time step over the time horizon by assuming zero dump (8), and a real battery should not be discharged beyond a
energy (Pdu = 0). A typical variation of stored energy in certain limit. Discharging the battery beyond the specified
the battery is shown in Fig. 3. For finding the minimum minimum limit will affect the life of the battery. Typically,
battery capacity, the stored energy level is determined for a battery is allowed to be discharged till its depth of dis-
a hypothetical battery with an initial zero state, charge (X). Combining these, the minimum battery capac-
QB(t = 0) = 0 (point ‘A’ in Fig. 3). If the final stored energy ity (B) required can be determined as:
level is negative at the end of the time horizon (at point ‘E’
in Fig. 3), i.e. if QB(t = T) < 0, then the power sources are
insufficient to meet the load. Starting from zero energy  
Qmax  Qmin Qmax1  Qmin  QB ðt ¼ T Þ
level to positive stored energy at the end of the time hori- B ¼ max ;
X X
zon is necessary for meeting the demand, incorporating
various system losses. The energy to be dumped is indi- where QB ðt ¼ T Þ P 0 ð10Þ
 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136 1129

In terms of pinch analysis, the variation of stored energy Table 1

level is equivalent to the grand composite curve in heat Time interval for dumping excess energy.
exchange networks and the point of minimum stored energy Condition Dumped power Time for
is equivalent to the pinch point. Pinch analysis began as a dumping
thermodynamic-based approach to energy conservation Qmax2  Qmax1 > QB ðt ¼ T Þ Qmax1  Qmax2 þ QB ðt ¼ T Þ AB, CD, DE
(Linnhoff et al., 1982), and later evolved over the years to Qmax2  Qmax1 CD
become a powerful tool for resource optimization (Linn- Qmax2  Qmax1 6 QB ðt ¼ T Þ QB ðt ¼ T Þ CD
hoff, 1993; Shenoy, 1995; Smith, 1995). Pinch analysis has
been fruitfully used in analyzing heat exchanger networks
(Shenoy, 1995), utility systems (Shenoy et al., 1998), mass 2.2. System optimization
exchanger networks (El-Halwagi and Manousiouthakis,
1989), water networks (Wang and Smith, 1994; Bandyopad- The set of all configurations, which can meet the load,
hyay et al., 2006; Pillai and Bandyopadhyay, 2007), distilla- forms the design space and it can be used to select the opti-
tion column (Bandyopadhyay, 2002; Bandyopadhyay et al., mum configuration based on the specified objective. The
2003, 2004), production planning (Singhvi and Shenoy, objective function, used in this paper, is the minimum cost
2002; Singhvi et al., 2004), etc. Pinch analysis recognizes of energy (COE) produced by the hybrid system. The cost
the importance of setting targets before design. This allows of energy depends upon the capital cost, operation and
different process design objectives to be screened prior to maintenance cost and the amount of energy delivered.
the detailed design of the process. Pinch analysis provides COE is represented as:
graphical representation tools and full control to the pro- ACC þ AOM
cess designer over decision making processes. COE ¼ ð11Þ
E
Along with finding the minimum battery capacity
requirement, the proposed methodology can be extended where ACC is the annualized capital cost, AOM is the an-
to find the maximum possible amount of energy that can nual operating and maintenance cost, and E is the amount
be dumped at various instants. This helps to plan further of energy delivered to meet the demand. The annualized
load expansion. A load which consumes less power than capital cost is calculated as:
the dumped power at that time can be added to the system X
without any addition in system capacity. The time horizon ACC ¼ C 0i  CRFi ð12Þ
i
(T) is divided into four parts, (AB, BC, CD and DE in
Fig. 3), using the points A (t = 0), B (t = tmax1), C (t = tref), where the capital recovery factor (CRF) is a function of life
D (t = tmax2) and E (t = T) as shown in Fig. 3. The excess (n) and the discount rate (d).
energy QB(t = T) has to be dumped such that the difference n
between the maximum and the minimum of the stored dð1 þ dÞ i
CRFi ¼ n ð13Þ
energy level should be minimum. The difference between ð1 þ dÞ i  1
Qmax1 and Qmin cannot be reduced further because there
is no excess energy between tmax1 and tref. The peak of C0i is the capital cost of the ith system component corre-
the stored energy curve that occurs after tref (Qmax2) can sponding to different power sources (photovoltaic module,
be reduced by dumping the excess energy between tref wind turbine, etc.) battery bank, inverter, and balance of
and tmax2. A part of the excess energy, (min(- system. Generation of the design space is illustrated for a
QB(t = T),Qmax2  Qmax1)), has to be dumped during time wind–solar hybrid system in the following section.
interval CD. Since the reduction of Qmax2 beneath Qmax1
will not further reduce the battery capacity the remaining 2.3. Illustrative example
excess energy can be dumped during other time intervals.
The minimum amount of excess energy that has to be Sukhalai in Hoshangabad district of Madhya Pradesh,
dumped during interval CD is QB(t = T), if Qmax2  India is the location considered for this example. The load
Qmax1 P QB ðt ¼ T Þ and Qmax2  Qmax1 , if Qmax2  Qmax1 profile is given in Fig. 4. The hourly variation of wind
< QB ðt ¼ T Þ. The remaining excess energy, if any, can be speed and solar insolation are taken from Roy et al.
dumped at any time interval except BC. If we dump the (2007). Various parameters used for sizing calculations
excess energy during BC, battery capacity increases as the are given in Table 2 (Roy et al., 2007). In this example
difference between Qmax1 and Qmin increases. Table 1 sum- the wind system is connected to the ac bus and the solar
marizes the results discussed in this paragraph and can be system is connected to the dc bus. The economic
used to allocate timing for dumping excess energy or plan parameters considered for system optimization are given
further load growth without increasing the size of the sys- in Table 3 (Kolhe et al., 2002). Even though the power gen-
tem. The amount of excess energy to be dumped at any erated by the photovoltaic array is a function of the ambi-
instant should always be non-negative, and a negative term ent temperature, operating voltage and current, wind
in the table indicates no dumping during that period. speed, etc., a simple constant efficiency model is used to
1130 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136

6 specified meteorological mast height, Hi. The wind speed at

turbine height H is calculated by the following correlation
5
Maximum load = 3.8 kW (Justus, 1978):
4  z
H
Load (kW)

v ¼ vi ð16Þ
3 Hi
Average load = 1.8 kW
2 where v is wind speed at turbine hub height H, vi is wind
speed at reference height Hi, and z is the power law expo-
1 Minimum load = 0.6 kW
nent. Figs. 5 and 6 show the power produced by a 1 kWp
0 solar PV array and the power produced by a 1 kW rated
0 2 4 6 8 10 12 14 16 18 20 22 24 wind turbine respectively kept in the location of case study.
Time (hour of the day)

Fig. 4. Load variations on a typical day at Sukhalai in Hoshangabad. 2.3.1. PV–battery system
The solar array rating is varied from zero to 1000 kWp
Table 2 and the minimum battery capacity required to meet the
Input parameters used in the system sizing and optimization. load is determined to generate the sizing curve of the
Photovoltaic system efficiency, gP (%) 10
Mast height, Hi (m) 10 0.8
Turbine hub height, H (m) 20

Output power from 1kWp PV array (kW)

Power law index, z 0.14
0.7
Cut-in wind speed, vc (m/s) 3
Rated wind speed, vr (m/s) 8
0.6
Cut-off wind speed, vf (m/s) 15
Net charging efficiency, gc (%) 90
Net discharging efficiency, gd (%) 90 0.5
Depth of discharge, gD (%) 60
Inverter efficiency, gi (%) 90 0.4

0.3

Table 3
Economic parameters considered for system optimization. 0.2

Discount rate (d%) 10

0.1
Wind generator life (years) 20
Photovoltaic system life (years) 20
0
Battery bank life (years) 5 0 5 10 15 20 25
Converter life (years) 10 time (hour of the day)
Cost of photovoltaic system ($/kWp) 3213
Cost of wind generator ($/kW) 3427 Fig. 5. Hourly average power available from a 1 kWp PV array at
Cost of battery bank ($/kW h) 86 Sukhalai in Hoshangabad.
Cost of converter ($/kW) 386
Operation and maintenance cost as a % of total capital cost 1
0.45
output power from a 1kW wind turbine(kW)

model the PV array. The power generated by the photovol- 0.4

taic array (Ppv) is given as: 0.35
P pv ¼ gp AP I T ð14Þ
0.3
where gP is the photovoltaic system efficiency, IT is the total
0.25
radiation incident on the array (W/m2) at that time step,
and AP is the total array area (m2). The power generated 0.2
by the wind turbine (Pw) is obtained from the power curve
0.15
of the wind turbine (Powell, 1981).
8 v2 v2 0.1
>
< P r v2r v2c for vc < v < vr
c

Pw ¼ Pr for vr < v < vf ð15Þ 0.05

>
:
0 otherwise 0
0 5 10 15 20
Pr is the rated electrical power (W), vc is cut-in wind speed time(hour of the day)

(m/s), vf is cut-off wind speed, (m/s) and vr is the rated wind Fig. 6. Hourly average power produced by 1 kW rated wind turbine at
speed (m/s). The wind speed data is generally available at a Sukhalai in Hoshangabad.
 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136 1131

system. The design space for the solar alone system is the 25
portion above the sizing curve and is shown in Fig. 7. As
the rating of photovoltaic array is increased, the minimum
battery capacity required to meet the load decreases till a 20 Optimum configuration
certain point, beyond which the battery capacity remains

Battery rating(kWp)
the same. This is because of the non-availability of solar Design
15 space
insolation during night time and a minimum amount of
storage is required for any amount of PV array to supply Feasible
the load at night. The optimum configuration corresponds region
10
to a system with PV array of 10.2 kWp, battery capacity of Sizing curve
57.5 kW h, and the cost of energy is $0.38/kW h. It is seen Infeasible region
from Fig. 7 that the optimum configuration corresponds to 5
the one with the maximum amount of storage, with the
minimum PV array rating.
0
0 10 20 30 40 50 60 70 80
2.3.2. Wind–battery system Wind trbine rating(kW)
The wind turbine rating is varied from 0 to 80 kW to
Fig. 8. Design space for wind-battery system.
determine the sizing curve and the design space for the sys-
tem (Fig. 8). Optimum system configuration calls for a
wind turbine capacity of 8 kW and battery capacity of imum battery size required to meet the load. The sizing-
19.7 kW h. The minimum cost of energy is determined to curve, shown in Fig. 9, is a surface in three-dimensional
be $0.24/kW h. space formed by the PV array rating, wind turbine rating
Comparing the sizing curves (Figs. 7 and 8), it is seen that and the minimum battery storage capacity required to meet
there is always a requirement of storage for any large solar the load. Every point above the sizing curve in the figure is
array capacity. This is primarily due to non-availability of a feasible design point and collection of all of them consti-
power during night. As wind is always available, for large tutes the design space. The sizing curve is also shown on the
values of wind turbine capacity, the storage capacity falls battery capacity vs. total generator rating diagram for var-
to zero and the system can be designed with no storage ious values of solar PV and wind turbine ratings in Fig. 10.
(Fig. 8). However, the optimum configuration based on low- For a typical configuration with solar and wind turbine rat-
est cost of energy contains a large amount of storage for both ing of 5 kWp and 4 kW, respectively, the minimum battery
the cases. The amount of storage required for the minimum capacity required is determined to be 36.6 kW h and the
PV array is larger than the amount of storage required for corresponding cost of energy is $0.31/kW h. From storage
the minimum wind turbine capacity and this is also due to capacity vs. total generation graph for various fixed values
the non-availability of insolation for the entire night. of wind turbine/PV array rating, it is seen that all optimum
points lie towards the maximum storage side. As the cost of
2.3.3. Wind–solar hybrid system generators is comparatively more than storage cost and the
The wind turbine rating is varied from 0 to 20 kW and configuration with largest battery storage is the optimum
PV array peak rating from 0 to 100 kWp to obtain the min- one. The overall optimum configuration corresponds to

58
Optimum configuration
56

54
Battery capacity(kWh)

60
52 Design
Battery rating(kWp)

Design space
40
50 Feasible space Feasible region
region
48 20

46
0
44 20
Sizing curve
Sizing curve
42 15 Infeasible region
100
10 80
40 Infeasible region 60
5 40
38 Wind rating(kW) 20
0 100 200 300 400 500 600 700 800 900 1000 0 0 PV array rating(kWp)
Solar array rating (kWp)
Fig. 9. Three dimensional plot showing the sizing surface and the design
Fig. 7. Design space for solar-battery system. space.
1132 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136

60 3. Generation of design space and optimization with

probabilistic approach
optimum
50 No wind turbine
configurations The methodology used to arrive at the design space in
for each sizing curve
Battery capacity(kWh)

deterministic approach is extended to incorporate the

40
Wind turbine=2 kW effects of uncertainty in the power available from the
resources. The hourly values of power available from the
30 power sources are considered as random variables. The sys-
Wind turbine=4 kW
tem is sized for a specified reliability level using chance con-
20 Wind turbine=6 kW strained programming approach. This approach is used by
Arun et al. (2008) for sizing a photovoltaic battery system.
It is assumed that the power available from the ith power
10
source (Pi(t)) during each hour follows a normal distribu-
No PV array
tion with mean lP i ðtÞ and standard deviation rP i ðtÞ . qij is
0 the coefficient of correlation between the power available
0 20 40 60 80 100 120
from ith and jth power sources. The load is assumed to
total generation capacity(kW)
be deterministic over the time step. The source uncertainty
Fig. 10. Sizing curves for different options for isolated power generation. is expressed as a probabilistic constraint. The system has to
cater to the specified deterministic demand with probability
greater than a specified value. The chance constrain relat-
the wind-battery system with no photovoltaic array. It is ing the probability of the demand, D(t) being met by the
economical to go for a stand-alone wind-battery system system is:
than the hybrid system for this location.
Variation of cost of generators and battery bank affect P ½DðtÞ P Dactual ðtÞ P a ð17Þ
the optimum system configuration and the COE. Fig. 11
shows the type of hybrid system which is cost effective as where Dactual(t) is the deterministic demand to be met over
the cost of solar PV array and wind turbine varies. It shows the time step and a is the specified reliability of compliance
that for most small variations around the price considered, of the constraint or the confidence level. Combining (17)
the optimum configuration is a wind alone system. Hence with the energy balance (7), the overall chance constraint
for these cases the optimum cost is not dependent on the can be written as:
cost of PV array. But as both cost of PV array decreases !
Xn
QB ðt þ DtÞ QB ðtÞ Xm
by a large amount and cost of wind turbine increases by P P j ðtÞ  þ þ P j ðtÞ  Dactual ðtÞ
a large amount, the system moves from wind alone to j¼1þm
f ðtÞDt f ðtÞDt j¼1

wind–solar. Solar alone system is not cost effective, even fi ðtÞ  P du ðtÞÞ P 0Þ P a ð18Þ
if the solar cost decreases by 50% and wind turbine cost
increases by 50% from their respective base cost. Separating stochastic and deterministic variables and
modifying the above equation, we get:
X
n X
m
QB ðt þ DtÞ QB ðtÞ
1.5
P P j ðtÞ þ P j ðtÞfi ðtÞ 6 
j¼1þm j¼1
f ðtÞDt f ðtÞDt
1.4
actual cost of PV array/cost considered

þDactual ðtÞfi ðtÞ þ P du ðtÞÞ 6 1  a ð19Þ

1.3
Since the random variables (Pj(t)) are normally distrib-
1.2
uted, the sum of the stochastic variables in the above
1.1 expression is a new random variable Pt(t) with mean lP j ðtÞ
Wind only and standard deviation rP j ðtÞ such that
1

0.9
X
n
lP t ðtÞ ¼ lP j ðtÞ qj ð20Þ
0.8 j¼1

0.7
Wind-solar
n X
X n
r2P t ðtÞ ¼ qjk rP j ðtÞ rP k ðtÞ qj ðtÞqk ðtÞ ð21Þ
0.6
j¼1 k¼1

0.5
0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 where
actual cost of wind turbine/cost considered

fi ðtÞ if j 6 m
Fig. 11. Type of hybrid system which is cost effective as the cost of PV qj ðtÞ ¼ ð22Þ
array and wind turbine varies. 1 otherwise
 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136 1133

where qjk is the coefficient of correlation between the power deviation. The standard deviation of power produced by
available from jth and kth sources. The deterministic equiv- each generator is assumed to be a fixed percentage of the
alent of (19) can now be expressed as: mean value of it. The PV array is considered to be a rela-
tively less variant power source with standard deviation
QB ðt þ DtÞ QB ðtÞ
 þ Dactual ðtÞfiðtÞ þ P du ðtÞ equal to 45% of the mean of the power available. The wind
f ðtÞDt f ðtÞDt
is taken as a high variable source with standard deviation
¼ lP t ðtÞ  rP t ðtÞ Z a ð23Þ equal to 80% of the mean power available from wind. The
reliability requirements for rural applications are consid-
where za is the inverse of the cumulative normal probability ered to be not very high and the specified reliability of com-
distribution corresponding to the required confidence level pliance is varied from 50% to 90%. The coefficient of
a with zero mean and unity standard deviation. Expressing correlation between the power sources is varied in full pos-
the deterministic equivalent in terms of the battery energy sible range to study its effect.
values for any time step, we get: The wind turbine rating is varied from 0 to 100 kW and
QB ðt þ DtÞ ¼ QB ðtÞ þ ðlP t ðtÞ  rP t ðtÞ za  Dactual ðtÞfi ðtÞ PV array peak rating from 0 to 100 kWp, respectively and
the minimum battery size required to meet the load for
 P du ðtÞÞf ðtÞDt ð24Þ
each case is determined. Similar to Fig. 10, sizing curves
It may be noted that the deterministic energy balance (7) for a reliability of 80% are shown in Fig. 12, where the
and the probabilistic energy balance (24) are equivalent. As power produced from the two sources are assumed to be
the random variable Pt(t) in the stochastic case is repre- independent. Optimum configuration (shown in Fig. 12)
sented by its deterministic equivalent, the methodology with 8 kWp of solar PV, 7 kW of wind turbine, and battery
developed to find the design space for the deterministic case capacity of 44.29 kW h, compliance with a reliability of
can be applied directly. 80% and the corresponding COE is $0.47/kW h. It may
Even though we have assumed that the power available be noted that the COE almost doubled to enhance the reli-
from the sources follows normal distribution, the power ability of the system from 50% to 80%. It may also be noted
sources cannot produce negative power. Eq. (24) is cor- that the optimum system is hybrid system, comprising of
rected when the deterministic equivalent of the power pro- both the PV and wind generators unlike wind-battery sys-
duced by any power source becomes negative. If the tem for deterministic case. Effect of reliability and coeffi-
deterministic equivalent of power produced by any source cient of correlation of sizing curve of the overall system
is negative, it is considered that particular source is not are shown in Figs. 13 and 14, respectively. From the sizing
producing any power. The mean and standard deviation curves shown in Fig. 13, the increase in storage capacity
of the power produced by such sources are zero. This can required with the increase in the system reliability require-
be incorporated by modifying only the expression of qj(t) ment can be observed for a hybrid wind–solar system with
in (22). 6 kW wind turbine. The power available from the two
8 sources is considered as independent. It can be noted from
< 0;
> if lP j ðtÞ  rP j ðtÞ za 6 0 Fig. 14 that as the coefficient of correlation between the
qj ðtÞ ¼ fi ðtÞ; if j 6 m and lP j ðtÞ  rP j ðtÞ za > 0 ð25Þ power sources varies from 1 to +1, the system size
>
: 1; if j > m and lP j ðtÞ  rP j ðtÞ za > 0 increases. The variation of sizing curve for 80% reliable

3.1. Illustrative example 60

No wind turbine
Illustrative example, discussed in Section 2.3, is extended
50
to incorporate the uncertainty associated with the power
wind turbine=2.5kW
produced by different sources. For energy related applica-
40 wind turbine=5kW
tions, wind speed is generally described using gamma distri-
Battery capacity

wind turbine=7.5kW
bution function, or its special cases like Weibull or Rayleigh
Optimum configuration
distribution functions, or using bivariate normal distribu- 30 (8kWp,7kW,44.29kWh)
wind turbine=10kW
tion function (Hennessey, 1977). The solar insolation gener-
ally forms a bimodal distribution function and can be
20
approximated by the superposition of two distribution
No solar array
functions. Those two distribution functions can be from
normal, beta or Weibull distributions (Khallat and Rah- 10
man, 1986). However, if the number of days for which the
data taken is sufficiently large, using central limit theorem 0
(Papoulis and Pillai, 2002), the power produced by the pho- 0 10 20 30 40 50 60 70 80 90
Total generation capacity(kW)
tovoltaic array and the wind turbine tends to follow a bivar-
iate normal distribution with known mean and standard Fig. 12. Sizing curves for a reliability of 80%.
1134 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136

60 1

55 0.8

alpha=0.90 0.6
50
Battery capacity (kWh)

coefficient of correlation
0.4
45
0.2 wind
40 alone solar
alpha=0.80
alone
0
35 Wind solar
alpha=0.70 -0.2
30
-0.4
alpha=0.60
25 -0.6
alpha=.50,deterministic
20 -0.8

15 -1
0 20 40 60 80 100 120 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95
Total generation (kW) alpha

Fig. 13. Variation of design space with the confidence level (wind turbine Fig. 15. Type of hybrid system which is costeffective for different
rating of 6 kW and zero coefficient of correlation). reliability of compliance and coefficient of correlation.

correlated power sources. So hybridisation is more effective

48 in such cases.
Optimum configuration
Hybridisation is most effective for
46 for each sizing curve

(a) Moderate values of reliability of compliance.

Battery capacity (kWh)

44 (b) High negative values of coefficient of correlation.

Coefficient of correlation
increasing from-1 to 1
42 steps of 0.333, Constant alpha curves are plotted between PV array rat-
6kW wind turbine,
alpha=80% ing and wind turbine rating in Fig. 16 with coefficient of
40 correlation equal to zero. It can be observed that for a
wind-battery (y-axis) system, the increase in the wind tur-
38 bine rating is more than three times when the required reli-
ability changes from 50% to 80%, while for a solar-battery
36 system (x-axis), the increase in PV rating is only about 1.5
times.
34 The effect on the type of hybrid system with change in
0 20 40 60 80 100 120
the cost of individual components is studied. The change
Total generation capacity (kW)

Fig. 14. Variation of sizing curves with coefficient of correlation. 30

alpha=0.8
alpha=0.7
25
wind–solar system with 6 kW wind turbine is shown in alpha=0.6
Fig. 14. The optimum configuration corresponding to each
Wind turbine rating (kW)

alpha=0.5
sizing curve is also shown. 20
The type of hybrid system which is cost effective for dif-
ferent reliability of compliance and coefficient of correla- 15
tion is given in Fig. 15. As the reliability of compliance
increases, a predominantly wind system shifts to a predom-
10 alpha=0.8
inantly solar system. This can be attributed to the differ- alpha=0.6
ence in variability of the power available from the two
sources. For low values of alpha (0.5–0.6), the optimal sys- 5 alpha=0.5 alpha=0.7
tem is a wind-battery system (variability of power from
wind turbine is more than that of solar PV) while for a high
0
reliable system with confidence level of 90%, it’s a solar 0 2 4 6 8 10 12 14 16 18
alone system. Wind–solar hybridisation is a cost effective solar array rating (kWp)
solution for moderate values of alpha. The cost and size Fig. 16. PV array capacity vs. wind capacity for various values of
of power sources and battery reduces for highly negatively reliability of compliance.
 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136 1135

1.5 find out the durations for which more load can be added
1.4
without any change in system sizing.
A case study from Sukhalai in Hoshangabad district of
actual cost of PV array/cost considered

Wind only
1.3 Madhya Pradesh, India to illustrate the proposed method-
1.2 ology for a wind–solar hybrid system, is presented. Using
deterministic procedure (as described in Section 2), the cost
1.1
of energy from a PV–battery system is determined to be
1 $0.38/kW h and the same for wind-battery system is
0.9
$0.24/kW h. The wind power generating system is a cheap
Wind-solar source of electric power compared to solar PV, but the var-
0.8 iability of the power available from the wind is greater than
0.7 that of the solar PV. Hence, if the reliability requirement is
low, wind-battery system is cost effective. On the other
0.6
hand, if the reliability requirement of the power produced
0.5 PV only is high, solar-battery system is cost effective. For moderate
0.6 0.8 1 1.2 1.4 1.6
actual cost of wind turbine/cost considered values of reliability requirement, the optimum choice is a
wind–solar hybrid system. Applying the chance constraint
Fig. 17. Type of hybrid system which is cost effective as the cost of PV programming, the optimum configuration for 80% reliabil-
array and wind turbine varies (alpha is 0.7 and zero coefficient of
correlation).
ity consists of 8 kWp of solar PV, 7 kW of wind turbine, and
battery capacity of 44.29 kW h, and the corresponding
COE is $0.47/kW h.
in cost of power sources affects the type of system which Among the various available renewable options for
can produce power at the lowest possible cost and hence, hybridization at a location, a proper choice of combination
the effect of variation of cost of wind turbine and PV array of power sources and their sizing depends upon various fac-
is studied. The type of hybrid system which is cost effective tors. From the case study performed in this paper, it can be
as the cost of PV array and wind turbine varies is shown in concluded that the characteristics of the available resources
Fig. 17 with alpha of 0.7 and coefficient of correlation of 0. at that location, the quality and reliability of the power sup-
For the base cost (considered in Table 3), the optimum ply required and the cost of various components decide the
configuration is a wind-battery system. The ratio of the optimum configuration. The characteristics of the available
varied cost of the wind turbine and PV array to the base resources can be captured by two parameters, the variability
cost is plotted in x- and y-axis of the figure respectively. associated with the power available from each source and
It shows that the system varies from a wind only system the coefficient of correlation between the powers available
(when solar cost is high and wind cost is low) to a solar from various sources. Negatively correlated power sources,
alone system (when solar cost is low and wind cost is high). as shown in Fig. 1, are good choices for hybridization. A
For small cost variations around the actual values of cost highly variable power source with low cost can be a better
of power sources considered, it’s a hybrid system. choice compared to a low variable higher cost power source
if the power supply reliability requirement is small and the
reverse is true for high reliability requirement. For moder-
4. Conclusions ate values of reliability, hybridization of these sources will
lead to cost effective solution.
Renewable energy based isolated hybrid systems are
potential alternatives to grid extension for remote electrifi-
References
cation. The random nature of the renewable resources and
the non-linearities involved in the modeling of power Abousdar, I., Ramkumar, R., 1990. Loss of power supply probability of
sources makes the design and sizing of hybrid systems chal- stand alone wind electric conversion systems: a closed form approach.
lenging. Deterministic or probabilistic approach is gener- IEEE Trans. Energy Convers. 7 (3), 445–452.
ally used to size the system and both have its own merits Abousdar, I., Ramkumar, R., 1991. Loss of power supply probability of
stand alone photovoltaic systems: a closed form approach. IEEE
and demerits. The method proposed in this paper combines
Trans. Energy Convers. 6 (1), 445–452.
the advantages of both these approaches and takes uncer- Ai, B., Yang, H., Shen, H., Liao, X., 2003. Computer-aided design of PV/
tainty of resources into account. Design space approach wind hybrid system. Renew. Energy 28 (10), 1491–1512.
for designing and optimizing stand alone wind–solar Arun, P., Banerjee, R., Bandyopadhyay, S., 2007. Sizing curve for design
hybrid power generating system has been demonstrated of isolated power systems. Energy Sustain. Dev. 11 (4), 21–28.
Arun, P., Banerjee, R., Bandyopadhyay, S., 2008. Optimum sizing of
in this paper. A methodology is proposed to find the min-
battery integrated diesel generator for remote electrification through
imum battery capacity required when the ratings of renew- design-space approach. Energy 33, 1155–1168.
able energy sources are known. This methodology is used Arun, P., Banerjee, R., Bandyopadhyay, S., 2009. Optimum sizing of
to derive the sizing curve and the design space. The meth- photovoltaic battery systems incorporating uncertainty through design
odology also helps to plan further load growth as it can space approach. Sol. Energy 83 (7), 1013–1025.
1136 E.S. Sreeraj et al. / Solar Energy 84 (2010) 1124–1136

Azaiez, M.N., Hariga, M., Al-Harkan, I., 2005. A chance-constrained Kolhe, M., Kolhe, S., Joshi, J.C., 2002. Economic viability of stand-alone
multi-period model for a special multi-reservoir system. Comput. solar photovoltaic system in comparison with diesel-powered system
Oper. Res. 32, 1337–1351. for India. Energy Econ. 24 (2), 155–165.
Bagul, A.D., Salameh, Z.M., Borowy, B., 1996. Sizing of stand-alone Kulkarni, G.N., Kedare, S.B., Bandyopadhyay, S., 2007. Determination
hybrid wind–photovoltaic system using a three-event probability of design space and optimization of solar water heating systems. Sol.
density approximation. Sol. Energy 56 (4), 323–335. Energy 81 (8), 958–968.
Bandyopadhyay, S., 2002. Effect of feed on optimal thermodynamic Kulkarni, G.N., Kedare, S.B., Bandyopadhyay, S., 2008. Design of solar
performance of a distillation column. Chem. Eng. J. 88 (1–3), 175–186. thermal systems utilizing pressurized hot water storage for industrial
Bandyopadhyay, S., Malik, R.K., Shenoy, U.V., 2003. Feed precondi- applications. Sol. Energy 82, 686–699.
tioning targets for distillation through invariant rectifying–stripping Kulkarni, G.N., Kedare, S.B., Bandyopadhyay, S., 2009. Optimization of
curves. Ind. Eng. Chem. Res. 42 (26), 6851–6861. solar water heating systems through water replenishment. Energy
Bandyopadhyay, S., Mishra, M., Shenoy, U.V., 2004. Energy-based targets Convers. Manage. 50 (3), 837–846.
for multiple-feed distillation columns. AIChE J. 50 (8), 1837–1853. Li, P., Arellano-Garcia, H., Wozny, G., 2008. Chance constrained
Bandyopadhyay, S., Ghanekar, M.D., Pillai, H.K., 2006. Process water programming approach to process optimization under uncertainty.
management. Ind. Eng. Chem. Res. 45, 5287–5297. Comput. Chem. Eng. 32 (1–2), 25–45.
Beyer, H.G., Langer, C., 1996. A method for the identification of Linnhoff, B., 1993. Pinch analysis: a state-of-the-art overview. Chem. Eng.
configurations of PV/wind hybrid systems for the reliable supply of Res. Des. 71, 503–522.
small loads. Sol. Energy 57 (5), 381–391. Linnhoff, B., Townsend, D.W., Boland, D., Hewitt, G.F., Thomas,
Bucciarelli Jr., L.L., 1984. Estimating loss-of-power probabilities of B.E.A., Guy, A.R., Marsland, R.H., 1982. User Guide on Process
standalone photovoltaic solar energy systems. Sol. Energy 32 (2), 205– Integration for the Efficient Use of Energy. The Institution of
209. Chemical Engineers, Rugby, UK.
Celik, A.N., 2002a. Optimization and techno-economic analysis of Mani, A., Rangarajan, S., 1982. Solar Radiation Over India. Allied Pub.,
autonomous photovoltaic wind hybrid energy systems in comparison New Delhi.
to single photovoltaic and wind systems. Energy Convers. Manage. 43 Natarajan, E., Rajendra Prasada, A., 2006. Optimization of integrated
(18), 2453–2468. photovoltaic–wind power generation systems with battery storage.
Celik, A.N., 2002b. The system performance of autonomous photovol- Energy 31 (12), 1943–1954.
taic–wind hybrid energy systems using synthetically generated weather Papoulis, A., Pillai, S.U., 2002. Probability, Random Variables and
data. Renew. Energy 27 (1), 107–121. Stochastic Processes, fourth ed. Tata McGraw-Hill.
Changchit, C., Terrell, M.P., 1993. A multiobjective reservoir operation Pillai, H.K., Bandyopadhyay, S., 2007. A rigorous targeting algorithm for
model with stochastic inflows. Comput. Ind. Eng. 24 (2), 303–313. resource allocation networks. Chem. Eng. Sci. 62, 6212–6221.
Charnes, A., Cooper, W.W., 1959. Chance-constrained programming. Powell, W.R., 1981. An analytical expression for the average power
Manage. Sci. 6, 73–79. output of a wind machine. Sol. Energy 26 (1), 77–80.
Chedid, R., Rahman, S., 1997. Unit sizing and control of hybrid wind– Protogeropoulos, C., Brinkworth, B.J., Marshall, R.H., 1997. Sizing and
solar power systems. IEEE Trans. Energy Convers. 12 (1), 79–85. techno-economical optimization for hybrid solar photovoltaic/
Deshmukh, M., Deshmukh, S., 2008. Modeling of hybrid renewable wind power systems with battery storage. Int. J. Energy Res. 21,
energy systems. Renew. Sustain. Energy Rev. 12 (1), 235–249. 465–479.
El-Halwagi, M.M., Manousiouthakis, V., 1989. Synthesis of mass Rao, S.S., 1980. Structural optimisation by chance constrained program-
exchange networks. AIChE J. 8, 1233–1244. ming techniques. Comput. Struct. 12 (6), 777–781.
Gavanidou, E.S., Bakirtzis, A.G., 1992. Design of stand-alone system with Roy, A., Arun, P., Bandyopadhyay, S., 2007. Design and optimization of
renewable energy sources using trade off methods. IEEE Trans. Energy renewable energy based isolated power systems. SESI J. 17, 54–69.
Convers. 7 (1), 42–48. Roy, A., Kedare, S.B., Bandyopadhyay, S., 2009. Application of design
Government of India, 2001. Tables on Houses, Household Amenities and space methodology for optimum sizing of wind-battery systems. Appl.
Assets (H Series Tables). Census of India, New Delhi. Energy 86 (12), 2690–2703.
Government of India, 2003. The Electricity Act, 2003, The Gazette of Shenoy, U.V., 1995. Heat Exchanger Network Synthesis: Processes
India, Extraordinary, 2003, Part II Section 3 Sub-section (ii). Ministry Optimization by Energy and Resource Analysis. Gulf Publishing
of Power, New Delhi. Company, Houston.
Government of India, 2005. National Electricity Policy, The Gazette of Shenoy, U.V., Sinha, A., Bandyopadhyay, S., 1998. Multiple utilities
India, Extraordinary, Part I Section 1. Ministry of Power, New Delhi. targeting for heat exchanger networks. Chem. Eng. Res. Des. A 76,
<[Link] 259–272.
national_electricity_policy.htm>. Singhvi, A., Shenoy, U.V., 2002. Aggregate planning in supply chains by
Hennessey Jr., J., 1977. Some aspects of wind power statistics. J. Appl. pinch analysis. Chem. Eng. Res. Des. 80 (6), 597–605.
Meteorol. 16 (2), 119–128. Singhvi, A., Madhavan, K.P., Shenoy, U.V., 2004. Pinch analysis for
International Energy Agency, 2006. World Energy Outlook, second ed. aggregate production planning in supply chains. Comput. Chem. Eng.
IEA Publications. 28 (6–7), 993–999.
Justus, C.G., 1978. Wind energy statistics for large arrays of wind turbines Smith, R., 1995. Chemical Process Design. McGraw-Hill, New York.
(New England and central US regions). Sol. Energy 20, 379–386. Wang, Y.P., Smith, R., 1994. Wastewater minimization. Chem. Eng. Sci.
Karaki, S.H., Chedid, R.B., Ramadan, R., 1999. Probabilistic perfor- 49 (7), 981–1006.
mance assessment of autonomous solar–wind energy conversion Yang, H.X., Lu, L., Burnett, J., 2003. Weather data and probability
systems. IEEE Trans. Energy Convers. 14 (3), 766–772. analysis of hybrid photovoltaic–wind power generation systems in
Khallat, M.A., Rahman, S., 1986. A probabilistic approach to photovol- Hong Kong. Renew. Energy 28 (11), 1813–1824.
taic generator performance prediction. IEEE Trans. Energy Convers. 1
(3), 34–40.