Energy 111 (2016) 803e817

Contents lists available at ScienceDirect

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

Development of optimal integrated renewable energy model with

battery storage for a remote Indian area
S. Rajanna*, R.P. Saini
Alternate Hydro Energy Centre, Indian Institute of Technology, Roorkee, Uttarakhand, 247667, India

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 10 March 2016
Received in revised form
20 May 2016
Accepted 2 June 2016

Over the past few years, renewable energy has come to be seen as a possible solution to the energy
problems of people. The present work focuses on optimal sizing of an integrated renewable energy
system (IRES) considering locally available different renewable energy sources namely micro hydro, solar,
wind, biomass and biogas with battery system for electrication of a remote area in Karnataka state in
India. Genetic algorithm (GA) has been used to minimize the total net present cost (TNPC) and cost of
energy (COE) of the developed IRES model considering the three decision variables-total active sunshine
area occupied by the SPV modules, total swept area required to install wind mills and state of charge
(SOC) of battery system. Scenario based results of optimal sizes, TNPC and COE have been obtained based
on suitable device types and time schedule of biomass generator. Based on optimization results, three IRE
scenarios are proposed for the study area. Of the three, scenario-S1 for zone 2 and zone 3. While,
scenario-S2 for zone 1 and zone 4 are found to be most feasible for the study area. Further, optimal time
schedule, resource combination and device type for all zones have also been determined.
2016 Elsevier Ltd. All rights reserved.

Keywords:
Net present cost (NPC)
Generic algorithm
Integrated renewable energy optimization
model (IREOM)
Remote area electrication

1. Introduction
Rural electrication plays an important role in the growth of a
nation and living standard of its citizens. Grid based electricity
supply to remote regions has proved unfeasible due to challenges
involved in connecting the grid to these areas [1,2]. Diesel generator
based power supply proves costly due to high fuel prices and more
importantly, they increase green-house emissions. Renewable energy resources are clean and may be good alternatives to conventional fuel for meeting the electrical load demand of remote rural
areas. However, due to the random nature of renewable energy
resources, it is preferable to use integrated renewable energy (IRE)
systems to increase system reliability [3,4]. An IRE system utilizes
two or more locally available renewable energy resources and has
the potential to provide a cost effective solution to meet the variety
of energy needs in off grid applications [5-7]. Renewable energy
systems generally entail high capital costs, low operation and
maintenance (O&M) costs and fuel costs, due to which an economic
analysis required to determine the optimum cost and benet ratio,
to arrive at the least possible unit price of the system. In order to

* Corresponding author.
E-mail address: srajannamce@gmail.com (S. Rajanna).
http://dx.doi.org/10.1016/j.energy.2016.06.005
0360-5442/ 2016 Elsevier Ltd. All rights reserved.

utilize the available renewable energy resources efciently and

economically, optimal models need to be developed. However, the
modeling of an IRE system is a complicated task which requires the
development of mathematical models for each component [8].
Such modeling requires optimal designing of system components
to minimize the total annual cost of the IRE system. As an example
for a typical case the optimum designing process includes estimation of the appropriate number of wind turbine, SPV panels and
batteries so that the load demand is satised at minimum possible
total net present cost [9]. Yang hongxing et al. [10] recommended
an optimal model for designing hybrid system employing battery
storage system. They considered ve decision variables such as
wind turbine number, wind turbine installation height and battery
numbers in order to obtain the power supply for a telecommunications relay station. Koutroulis et al. [11] investigated optimal
number and type of system components for 20-years using genetic
algorithm. The objective was to minimize the total system cost
subjected to the constraint that the load power with zero load
rejection. Total system cost was compared with the system cost
obtained through conventional optimization methods. Askarzadesh and Leandro dos [5] developed three grid independent hybrid
renewable energy systems for electrication of a small load area in
Kerman Iron. Optimal model of hybrid system was achieved based
on the integer variables of system components such as total active

804

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

Nomenclature
ASPV
AWTG
BGG
BMG
CN
BS
COE
Com
Cb
CRF
CD
CP
CNPV
CS
DE
D
DOD
EBGG
EENS
EIR
EMHP
EBMG
EWTG
ESPV
Egen
EO
FCNPV
g
hnet
It
I
L
Lw
MHG
NSPV
NWTG
NBS
OMNPC
Pr
PSPV
PMHG
PBGG

area of SPV (m2)

swept area of wind turbine (m2)
biogas generator
biomass generator
runoff curve number for hydrological sail (CN 40e58
for dense forest)
battery system
cost of energy ($/kWh)
combination
initial capacity of battery system
capital recovery factor
cow dung (ton/yr)
power co-efcient
net present capital cost
converter system
dump energy (%)
duration of unavailability of load (hour)
depth of discharge
annual energy output of BGG (kWh)
expected energy not supplied (kWh)
energy index ratio
annual energy output of MHP (kWh)
annual energy output of BMG (kWh)
annual energy output of WTG (kWh)
annual energy output of SPV (kWh)
annual energy (kWh)
total energy demand (kWh)
net present fuel cost ($/ton)
acceleration due to gravity (m/s2)
net head (m)
solar radiation availability (W/m2)
monthly rainfall (mm)
average annual power load (kW)
length of the watershed (m)
micro hydro generator
solar panels (Nos.)
wind turbine (Nos.)
battery system (Nos.)
net present O &M cost ($/yr)
rated output power of wind turbine (kW)
power output of SPV system (kW)
power output of MHG (kW)
power output of BGG (kW)

area panels, total swept area of the wind turbine blades and
number of batteries using PSO based approach. They found that PV/
WT/battery hybrid system was most cost effective and reliable for
meeting the energy demand of the proposed area. Koutroulis et al.
[12] proposed genetic algorithms based optimization function to
minimize the sum of the capital and maintenance costs occurring
during the desalination system's life period. Ramakumar et al. [7]
developed a knowledge-based system design tool IRES-KB and reported scenarios based models for typical remote rural village in
India. The objective of the study was to minimize the total capital
cost at a pre-selected reliability level. Optimal sizes of energy
storage systems have been found to fulll the energy requirements
at the desired reliability level. Subho and Sharma [13] proposed a
particle swarm optimization based hybrid model with the cycle
charging strategy for a remote area demand of 7 un-electried
villages of Dhauladevi block of Almora district, with an account of

PWTG
power output of WTG (kW)
PBMG
power output of BMG (kW)
Q
discharge (m3/s)
Qd
direct surface runoff depth (mm)
RNPV
net present replacement cost
SPV
solar photovoltaic panel
S
maximum potential retention (mm)
S1, S2 and S3 three scenarios of IRE models
TS1 and TS2 time schedules of the biomass generator
TNPC
total net present cost ($)
UL
unmet load
Vr
rated speed of the wind turbine (m/s)
WTG
wind turbine generator
$
US Dollar
hMHP
efciency of micro hydro generator (%)
hBGG
efciency of biogas generator (%)
hBMG
efciency of biomass generator (%)
hSPV
efciency of SPV panel (%)
hWTG
efciency of wind turbine generator (%)
hinv
efciency of inverter system (%)
hRECT
efciency of rectier system (%)
hBC
battery charging efciency (%)
hBD
battery discharge efciency (%)
s
hourly self discharge rate
rw
density of water (kg/m3)
aBMG
capital cost of BMG ($/kW)
aBGG
capital cost of BGG ($/kW)
aMHG
capital cost of MHG ($/kW)
aSPV
capital cost of SPV ($/kW)
aWTG
capital cost of WTG ($/kW)
aBS
capital cost of battery system (per battery)
aCONV
capital cost of bi-directional converter (per converter)
bBMG
maintenance cost of BMG ($/yr)
bBGG
maintenance cost of BGG ($/yr)
bMHG
maintenance cost of MHG ($/yr)
bSPV
maintenance cost of SPV ($/yr)
bWTG
maintenance cost of WTG ($/yr)
bBS
maintenance cost of BS ($/yr)
bCS
maintenance cost of bi-directional converter ($/yr)
rW
density of water (kg/m3)
g
interest rate (%)
m
escalation rate (%)
lAFC
average bio-generators fuel price ($/ton)
t
project life time

the available resources of solar, hydro, biomass and biogas energy,

along with the addition of diesel generator. In Refs. [14], developed
suitable hybrid energy model for the same area with different three
energy management strategies of cycle charging, load following
and peak shaving strategy. Minimum total net present cost and cost
of energy was achieved through three different optimization
techniques (genetic algorithm, particle swarm optimization and
biogeography based techniques) with considered same strategy.
They also found twelve different device type combination of hybrid
system and compared, out of that 10th combination was found to
be most feasible and suggested for the same study area. Ismail et al.
[15] proposed a hybrid system comprises solar PV panel and battery
system of for a small rural community. Based on the investigation,
the system has been found extremely benecial compared to utility
grid. Rajanna and Saini [16] developed scenario based integrated
renewable energy models with least net present cost and cost of

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

energy for a study area using HOMER. Chauhan and Saini [17]
developed an optimal size of integrated renewable energy system
from locally available renewable energy resources of Micro Hydro
Power, biogas, biomass, wind and solar energy system to supply
continuous power to meet energy needs of cluster of villages in
Uttarakhand State of India. Levelized cost of energy of different
resources scenarios have been computed based on net present
value of system components using discrete harmonic search (DHS)
based approach. In Ref. [18], techno-economic analysis was carried
out for proposed system through load shifting based demand
management strategy for the same study area. They suggested that
with DSM strategy was found to be most cost feasible solution
when compare to without DSM strategy.
Wang and Singh [19] developed an optimal hybrid model using
PSO based approach by considering the multi-criteria parameters.
Patil et al. [20] developed an integrated renewable energy model
based on the seasonal load prole and determined two system
sizes, i.e. the specied system sizes and the manufacturer specied
system sizes. They also worked out the reliability indices in terms of
the expected energy not supplied. The developed model was
optimized with the energy balance techniques by considering two
different probabilistic values of the EENS.
Maleki and Askarzadeh [21] proposed a hybrid energy system
with battery storage system based power supply to meet the load
demand of remote region in Iran. Optimal sizes of different resources combination of hybrid system was modeled using harmony
search while considering the variation load demand as well as solar
and wind resources. Heydari and Askarzadeh [22] developed an
efcient PV power plant to supply the electrical load of stand-alone
remote areas in Kerman, Iran. Biomass energy generation systems
used as a backup system to supply the demanded load in decit
conditions. In Ref. [23], discrete harmonic search method was used
to develop optimal size of hybrid energy systems consisting of solar
and wind energy system for meeting the energy need for the same
study area. Chang and Grace Lin [24] proposed a methodology to
develop optimal sizes of hybrid energy system comprises of PV,
wind and diesel power generators with battery bank using simulation optimization and Monte Carlo simulation based techniques.
They also estimate the cost of power shortage, energy storage,
power generation and carbon emission for the developed hybrid
system model. Matteo Ranaboldo et al. [25] proposed a deterministic heuristic for designing community off-grid electrication
projects through micro-scale renewable energy resource of independent generation points and micro grids. The Meta heuristic algorithm has been used to design tool that can efciently used for
stand-alone community electrication projects requiring of low
computational resources. Arabali et al. [26] proposed a hybrid energy system model with battery storage system to meet the various
energy needs of heating, ventilation, and air conditioning of HVAC
load considering stochastic model of wind generation, photovoltaic
generation, and load. Optimal cost and number of SPV, wind and
battery system has been obtained using genetic algorithm with two
point estimate method. The maximum capacity of energy storage
system and excess energy are computed as the important parameter for energy efciency assessment. Also developed a smart-grid
strategy for matching HVAC load through solar and wind renewable
energy generation. Finally, the economic indicators of levelized cost
of energy (LCE) [27], total life cycle cost (LCC) [28] and net present
cost (NPC) [13,29,30] along with reliability based model [31-33]
were considered during the optimal sizing of the hybrid energy
system.
Based on the literature review, it is found that only few studies
have been carried out to develop modeling of integrated renewable
energy system by combining resources such as micro hydro, solar,
wind, biomass, biogas without energy storage system [34]. Very

805

few studies are available for electrication of cluster/block/zone in

remote rural areas, considering cost and size optimization based on
objective function with economic and reliability indicators. The
present paper deals with the size optimization of IRE models
comprises of micro hydro, solar, wind, biomass, biogas with battery
storage system for a study area considering the economic, reliability worth. Finally, keeping in view the accuracy level in results,
the genetic algorithm is used for modeling of integrated renewable
energy (IRE) system for the study area considered.
2. Study area
Based on the comprehensive assessment on un-electried villages of a cluster of villages of Chamarajanagar district in Karnataka
state of India has been found to be most appropriate remote rural
areas for the present study. The objective of the study is to minimize the total net present cost and cost of energy of IRE system for
the proposed study areas. Detailed description of zone wise
demography data of the study area has been discussed under
earlier study carried out by the authors [34]. The general information of the study area provided in the earlier study are summarized and given in Table 1.
As given in Table 1, the villages are categorized into four zones
based on availability of energy resources and population density.
The total number of households in all four zones as 2240 having a
total population in all the four zones as 4440.
2.1. Load assessment
The load data of households and population of the villages have
been estimated based on eld survey in the form of questionnaire.
Hourly load demand has been estimated on the basis of operating
hours and power rating of the appliances used during 24 hourly
segments in a day. Hourly load variation for all zones during
different seasons (season I- Feb to April; season II- May to July;
season III - August to Oct; and season IVeNov. to Jan.) have been
estimated for all the four zone and are as shown in Fig. 1. The load
assessment for all four zones has been done on the basis of four
different load sectors (domestic, community, commercial and small
industrial). Each load sector consists of different electrical appliances. Estimation of the total energy demand of each load sector
was done on the basis of the total number of appliances used within
one zone, and power ratings and operating hours of those appliances in a day in every season. Details of load sectors for all the four
zones are shown in Fig. 2. The total number of appliances used and
their operating hours in different seasons in different sectors for all
zones are given in Table 2.
The estimation of total energy requirement of the villages was
made on the basis of minimum desirable load. The total estimated
annual loads of zone 1, zone 2, zone 3 and zone 4 are found to be as
291840 kWh/yr, 652072 kWh/yr, 737578 kWh/yr and 753010 kWh/
yr respectively. The data have been synthesized through HOMER
software considering estimated values of seasonal hourly load.

2.2. Resources assessment

In order to assess the potential of renewable energy resources,
an extensive survey was conducted and information collected
regarding the availability of biomass, solar irradiation, micro hydro
and wind speed. The potential assessment of SPV, MHG, BGG, BMG,
and WTG was made as per the available standard methodology.
Village wise energy potential in the study area and optimum system sizes are computed and discussed as follows:

806

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

Table 1
The general information of the study area.
Features

Zone 1

Zone 2

Zone 3

Zone 4

Electried villages (Nos.)

Un-electried villages (Nos.)
Households (Nos.)
Household populations (Nos.)
Latitude of the area (Degree)
Longitude of the area (degree)

11
5
302
898
12.05
77 400 4800

23
6
395
1141
11 520 58.800
77 160 58.800

24
8
763
2718
11 580 58.800
77 70 58.800

10
7
780
1906
11 450 57.600
77 50 6000

Fig. 1. Seasonal variation of hourly load prole for all four zones.

2.2.1. Micro hydro generator

Micro hydro turbine generator is able to operate under low head
and low volumetric ow rate conditions, and is suitable for small
water streams. The potential of micro hydro power (PMHG) at a

particular site area depends on net head and dependability of the

ow rate (Q). The discharge was estimated for the considered zones
on the basis of topography and monthly average rain fall data using
curve number method [34e36].

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

807

Fig. 2. Sector wise load prole for all four zones.

Table 2
Details of load estimation for all the four zones.
Load sectors

Parameters

Domestic

Household

Appliances

Switching point (Nos.)

Nos. used

Power rating (W) Operating

hours in
different
seasons

Zone 1 Zone 2 zone 3 zone 4

Lighting
TV
Radio
Ceiling fan
Community
Primary health centre
Refrigerator
Lighting
Ceiling fan
Primary High School
Lighting
Ceiling fan
Pumping water
Community street lights Street lighting
Community hall
Lighting
Ceiling fan
Commercial
Lighting
Flour mill
Small Industrial
Saw mills

2 point
604
1 point
302
1 point
302
1 point
302
1 point
1
3 point
3
3 point
3
1 2 point
12
12 point
12
1 motor (5HP) set at cluster of 300 HH
1
1pole at cluster 6 HH
100
1 point
1
1 point
1
1 shop at cluster of 125 HH
5
1 our mill at cluster of 250HH
1
1 saw mill at cluster of 250 HH
1

Actual runoff is calculated by using Eq. (1) [34e36] as;

Qd
S

I  0:2S2
when I > 0:2S
I 0:8S

(1)

25400
 254
CN

TC 0:0195

L3w Hnet

0:0208  A  Qd
TP

1526
763
763
763
1
3
3
12
12
2
254
1
1
12
3
3

1560
780
780
780
1
3
3
12
12
2
260
1
1
12
3
3

I
30
100
15
55
1500
30
55
30
55
3675
60
30
55
30
5000
5000

II

III IV

8 9 8 8
8 9 8 9
4 5 4 5
24 12 18 0
24 12 18 10
4 5 4 5
5 4 4 0
4 5 4 4
6 4 5 0
4 4 4 4
12 12 12 12
4 4 4 3
5 4 5 0
3 4 4 4
4 5 4 5
4 5 4 5

on proposed approach. The estimated values of stream ow for

zone 1, zone 3, and zone 4 are shown in Fig. 3. The hourly power
output of micro hydro power (MHG) plant has been calculated
based on the dependability, ow rate and head by using Eq.(4)
[34,37] as;

PMHP t 9:81  Q  hnet  hMHP  rw =1000

The peak rate of runoff or discharge of a watershed is calculated

by using Eq.(2) and (3) [34e36] as;

p
TP 0:6TC TC
0:77
r
.

790
395
395
395
1
3
3
12
12
1
132
1
1
8
2
2

(4)

Where Q is discharge (m3/s); hnet is net head (m), hmhg is overall

efciency of micro hydro generator, and ra is density of water
(1000 kg/m3).

(2)

(3)

Where; Q is the rate of runoff or discharge (m3/s); TP is the time to

peak runoff (hour); TC is time of concentration (hour); LW is the
length of the watershed (m); Hnet is the difference in elevation
between the highest point and the outlet point dened as head of
the watershed (m); A is the area of the watershed (hectare).
Water stream potential of 115 l/s (60%) for zone 1, 410 l/s (50%)
for zone 3 and 437 l/s (50%) for zone 4 have been estimated based

2.2.2. Solar photovoltaic panel

Solar radiation data has been taken from National Renewable
Energy Limited (NREL) for the latitude and longitude of the study
areas. Solar energy potential of 1944 kWh/m2/yr for zone 1,
1919 kWh/m2/yr for zone 2, 1975 kWh/m2/yr for zone 3 and
1925 kWh/m2/yr for zone 4 is estimated in the study area. The
output power of SPV array (PSPV) at hour t has been calculated
based on modeling equation developed under earlier studies [34]
and is expressed by Eq.(5) as.

PSPV t hSPV  HT t  ASPV

(5)

Where ASPV is active area of SPV panel, HT is solar radiation in kW/

808

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

Fig. 3. Flow duration curve.

m2, h

spv

is efciency of the SPV panels.

2.2.3. Wind turbine generator

The power density of the wind speed has been computed based
on the wind speed data obtained from nearby wind monitoring
station of all the zones. A wind potential of 1097 kWh/m2/yr for
zone 1, 2730 kWh/m2/yr for zone 2, 2410 kWh/m2/yr for zone 3 and
2816 kWh/m2/yr for zone 4 has been estimated for power generation. At an hour t, electrical power output of a wind turbine
generator (PWTG) in kW is calculated by using mathematical
modeling Eq. (6) [5] as.

8
0
>
>
>
>
!
>
<
v3 t  v3ci
PW t Pr

>
v3r  v3ci
>
>
>
>
:
Pr

vt  vci or vt  vco
vci < vt < vr

(6)

vr < vt < vco

Rated power of the wind turbine can be calculated by using Eq.

(7) as;

1
Pr $Cp $ra $hg: Aw $v3r
2

(7)

Where Aw is swept area of wind turbine rotor; ra is air density; Cp is

power coefcient of the proposed wind turbine; hg is generator
efciency; vr is rated speed of wind turbine;vci is cut-in speed of
wind turbine generator; vco is cut-out speed of wind turbine
generator.

biomass potential of 135 tons/yr for zone 1, 340 tons/yr for zone 2,
343 tons/yr for zone 3 and 348 tons/yr for zone 4 has been
considered for electricity generation. The hourly potential of biogas
generator (PBMG) in kW, considering system operating hours as 12
per day has been estimated using mathematical modeling Eq. (9)
[38] as;

PBMG t

Total fuel wood ton=yr  CVBM  hBMG  1000

365  860  operating hours per day

Where PBMG is hourly power output of biomass generator; hBMG is

overall energy conversion efciency of biomass generator; CVBM is
caloric value of biomass (4015 kcal/kg).
2.2.6. Battery storage system
The nominal capacity of battery storage system can be dened
as the product of initial capacity of battery (Cn) system and ampere
hour (Ah) of the battery. The maximum permissible depth of
discharge DOD (%) is specied by the system designer at the
beginning of optimal sizing process and determined by using Eq.
(10) as;

SOCmin 1  DOD$Cb



Total gas yield m3 =day  CVBG  hBGS
PBGG t
860  operating hours per day

(8)

Where PBGG is hourly output of biogas generator; hBGS is overall

energy conversion efciency of biogas generator, CVBG is caloric
value of biogas (4700 kcal/kg).
2.2.5. Biomass generator
Based on the availability of forest foliage in the study area,

(10)

The state of charging and discharging of the battery is based on

difference in the load demand and generating power of that
particular hour and is calculated by using Eqs.(11) and (12) [5] as.

SOCt 1 SOCt$1  s EPDC t hrect EPAC t

 PLD t$hBC

2.2.4. Biogas generator

Based on the availability of the cow dung in the study area,
biogas potential of 1928 tons/yr for zone 1, 7615 tons/yr for zone 2,
7199 tons/yr for zone 3 and 7946 tons/yr for zone 4 is estimated for
electricity generation. The hourly potential of biogas generator
(PBGG) in kW based on available biogas potential and considering
operating hour as 10 per day has been determined using mathematical modeling Eq. (8) [ 13] as;

(9)

SOCt 1 SOCt$1  s  hinv NPAC t=hDB

(11)
(12)

Where, s is hourly self discharge rate; hBC and hDB are respectively
charging and discharging efciency of the battery system.
3. Problem formulation
An IRE system has been proposed to meet energy demand in
different load sectors. Optimization involves the optimal sizes and
cost of an IRE system with reliability parameters. The objective of
the proposed study to minimizes total net present cost and cost of
energy of the integrated renewable energy system with reliability
constraints as EIR. It can be expressed by Eq. (13) [13] as:

COERs=kWh

TNPC  CRFg; t
P8760
t1 Egen t

(13)

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

809

Where Egen is the total annual energy generated by the IRE system;
g is the annual interest rate 11% and t is the life of the plant
considered as 20 years in the present study.

The life of the battery (Lb) and inverter is less than the life of
SPV, WTG, BGG, MHP and BMG. Therefore additional investment is
needed to replace these systems accordingly,. Here, the lifetime of
the project is selected for 20 years but the life of the battery and
inverter system is considered as 4 and 10 years respectively.
Therefore, 3 times replacement for battery and 1 time for inverter
have been considered during the projects lifetime. As results, the
net present value and number of times within the N year's horizon
that need to replacement of the battery and inverters are calculated
by using Eqs.(17) and (18) [5] as;

3.1. Constraints

Xb

In order to convert the total net present cost (TNPC) to the

annual capital cost, capital recovery factor (CRF) dened by the Eq.
(14) as

CRF

g1 gt
1 gt  1

(14)

For optimal sizes of an IRE system to meet the load demand,

assessment can be made based on reliability and economic
parameter as illustrated below;

N
1
Lb

(17)

RNPV NBATT $ aBATT



X 1 mj
1 m 10
aINV $PINV
1i
1i
j5;10;15
(18)

3.1.1. Economic criteria

In case of GA based modeling, cost parameter varies for different
specications with system components. Cost parameter comprises
capital cost, maintenance cost, replacement cost and fuel cost of the
system components. The technical parameters considered as input
to GA such as sizes based on mathematical modeling, power ratings
of the system components, rated speed, cut-out and cut-in speed
and swept area of wind turbine, area of SPV panels, nominal capacity, voltage and DOD of battery systems, efciency of SPV panel
and converters. The summary of technical and cost parameters of
different device types of the system components considered under
present study are given in Tables 3e7 [13,34,38].
The total net present cost (TNPC) is used as an economic indicator for sizing of IRE system. It involves net present capital, O&M,
replacement and fuel cost of all the expenses occur over the life
span of the system and can be expressed by Eq.(15) [5] as;

TNPC CNPV OMNPV RNPV FCNPV

c. Net present operation and maintenance cost

Net present operation and maintenance (O&M) cost of all the
system components such as SPV, WTG, MHP, BGG, BMG, BS and
inverter system has been computed based on total cost incurred on
operation and maintenance during the year. If the cost grows at an
annual rate of m, sum of net present value (NPV) of all the system
components can be expressed by Eq. (19) [5] as;

O&MNPV bSPV $NSPV bWTG :NWTG bBATT :NBAT bINV $PINV


n 
X
1m j
bMHG :PMHG bBGG :PBGG bBMG $PBMG 
d
1i
j1
(19)

(15)

Where, C is the capital cost; OM is the operation and maintenance

cost; R is replacement cost and FC is the fuel cost. The subscripts of
npv represent the net present value of system components.
a. Net present capital cost
In order to calculate total net present value of capital cost, initial
cost and optimal sizes of all the system components such as SPV,
WTG, MHP, BGG, BMG, BS and inverter system have been considered. The total net capital cost of all the system components is
expressed by Eq. (16) [5] as;

d. Net present fuel cost

Net present fuel cost of biomass and biogas has been calculated
based on availability of forest foliage and cattle dung in ton/yr. If the
cost grows at an annual rate of m, then the sum of net present value
(npv) of fuel cost of biomass and biogas is computed by using
Eq.(20) as;

FCNPV


n 
X
1m j
j1

1i

flAFR CDBGG ton=yr FFBMG ton=yrg

(20)

CNPV faSPV $NSPV aWTG $NWTG aBATT $NBATT aINV $PINV

aMHP $PMHP aBGG $PBGG aBMG $PBMG g
(16)

b. Net present replacement cost

Table 3
Technical and cost parameters of micro hydro, biomass and biogas system.
Resources

Type

Power rating (kW)

Capital cost ($)

O &M cost ($/yr)

MHP
BGG
BMG

1
1
1

1
1
1

1201
751
901

30
19
27

3.1.2. Reliability criteria

The developed model has been optimized based on the concept
of energy index ratio (EIR) and Expected Energy Not Supplied
(EENS). EENS represents expected energy not supplied when load
surpasses available generation. The EENS has been calculated [20]
as;

EENS

8760
X

L$D

(21)

i1

Where, L is the average annual power load (kW); D is duration of

unavailability of load (hour). OPF is considered as an energy index
ratio used to assess reliability of the IRE system which can be

810

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

Table 4
Technical and cost parameters of wind turbine system.
Type

Power rating (1 kW)

Rated speed (m/s)

Cut-out speed (m/s)

Cut-in speed (m/s)

Swept area (m2)

Capital cost ($)

O &M cost ($/yr)

1
2

1
1

11
10

13
25

2.5
1.5

3.14
4.26

3153
3076

$95
$92

Table 5
Technical and cost parameters of SPV panels.
Type

Power rating (Wp)

Area of SPV panel (m2)

Panel Efciency

Capital cost ($)

O &M cost ($/yr)

1
2

230
120

1.63
1.07

0.144
0.12

590
541

12
11

Table 6
Technical and cost parameters of battery system.
Type

Nominal capacity (ah)

Voltage (v)

DOD (%)

Capital cost ($)

O &M cost ($/yr)

1
2

150
150

24
12

50
50

188
270

$8
$11

Table 7
Technical and cost parameters of bidirectional converter.
Type

Efciency (%)

Power rating (kW)

Capital cost ($)

O &M cost ($/yr)

90

1.5

171

dened as one minus the ratio of EENS to total energy demand of

the system (EO) and can be expressed by Eq.(22) [20] as;

EIR 1 

EENS
EO

(22)

4. Energy management strategy

Energy management system plays an important role in satisfying energy demand by efciently utilizing the components present in IRE system. The main objective of the study is to develop an
integrated renewable energy system optimization model (IRESOM)
while considering locally available renewable energy resources
such as solar radiation, wind speed, biomass (forest foliage), biogas
(animal dung), micro hydro power and battery storage system to

meet energy needs in remote rural areas at least possible life cycle
cost and COE.
Fig. 4 shows a proposed off grid IRESOM model consisting of
solar photovoltaic system (SPV), wind turbine system (WTS),
biomass generator (BMG), biogas generator (BGG), micro hydro
generator (MHP), battery bank system and dump load. Out of ve
generators employed in the model, MHG, BMG and BGG generate
AC power whereas SPV and WTG generate DC power. Thus, the
power generation by renewable generators is divided into two
branches - one providing renewable generator energy directly to
load and the other powering the battery. The load demand can be
met either through the power generated by AC generators only, or a
combination of AC and DC generators. If the combination of AC and
DC generators also fails to meet the load demand, the battery system can be used to supply the required energy. These combinations
can mathematically expressed by Eqs. (23-25) [39] as;

PAC t PMHG t PBGG t PBMG t

(23)

PDC t PSPV t PWTG t

(24)

PACDC t PAC t hINV PDC t

(25)

Fig. 4. Schematic of the developed IRESOM Model.

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

In any hour t, the energy available to charge the battery bank

represented by Eq. (26) [39] as;

PBatt;IN t hBC  hCC  PSPV ;Batt t PWTG;Batt t hREC



 PMHG;Batt t PBMG:Batt t PBGG:Batt t

(26)

Further, the energy available from the battery to serve the load
expressed by Eq. (27) [39] as;



PBatt ;Load t hBD  PBatt;IN t

(27)


PLoad t PMHG;Load t PBMG$Load t PBGG$Load t hINV


 PSPV ;Load t PWTG;Load t PBatt;Load t

(32)

The discharge process of the SOC is expressed by the Eq. (33) [5]

Where Pj (t) is the energy output from system components and j

represents system components such as SPV, WTS, MHG, BMG, BGG,
and BATT.
Pj, Load (t) is energy supply from system components to load, Pj,
Batt (t) is energy supply from sources to battery, P j, Dump (t) is excess
energy output from system components to dump load and EBatt, IN is
energy input to battery.

4.1. Battery charging strategy

The deep-cycle lead-acid type battery is commonly used in integrated system. The optimal size of the battery bank can be
determined based on charging and discharging of the battery.
Fig. 5 shows ow chart of SOC of battery. It is clear from the
Figure that the power produced by renewable generators bifurcates
into two channels-one provides energy from renewable generator
directly to load while the other powers from the battery. The load
demand can be fullled either by solely using AC generators or a
combination of AC and DC generators if only AC generators are
unable to meet the load. If AC and DC generators are also unable to
produce the required amount of energy, the battery system can be
used to supply power to the required load.
During load demand from the system, the system has three
conditions of operation to supply energy to the required load. These
conditions are discussed below.
Condition (1): PAC t > PLD t
When the energy generated from AC generators (MHP, BGG and
BMG) is more than the load demand, the surplus energy along with
DC power go on to charge the battery bank provided they are not
already fully charged. Thus, AC renewable generators can be used
under the conditions when the load demand is low. The SOC
charging process can be expressed by Eq. (29) [5] as.

SOCt 1 SOCt$1  s EPDC t hrect EPAC t

(29)

Condition (1I): PACDC t > PLD t

In cases when the load demand is more, total energy generated
from AC and DC renewable generators can be used to power to
meet the total demand. If the total energy is in excess of total demand then it is used to charge the battery. If the battery is already
charged to maximum level, then this energy is send to the dump
load. The SOC charging process can be expressed by Eq. (30) [5] as.

SOCt 1 SOCt$1  s PACDC t  PLD t$hBC

(31)

SOC max takes the value of nominal capacity of the battery bank
(Cb).
Condition (1II): PACDC t < PLD t
In cases where the total power generated by AC and DC generators is less than the load demand, the needed energy can be obtained by discharging the stored energy of the battery system,
provided SOC is maximum. The required energy under such condition can be expressed by the Eq. (32) as;

NPAC t PLD t  PACDC t

(28)

 PLD t$hBC

SOC t  SOC max

811

(30)

Since the maximum energy stored in the battery bank cannot

exceed the maximum state of charge (SOC max), the following
condition should be satised during optimization.

as.

SOCt 1 SOCt$1  s  hinv NPAC t=hDB

(33)

During discharging, the following constraints should be

satised.

SOCt  SOCmin

(34)

Further, during discharge of the battery, if SOC (t) < SOC min, a
percentage of energy demand to the load is not satised. In this
case, SOC (t) is set to SOC min and the loss of power supply is
considered as unmet load.
5. Optimization methodology
Genetic algorithm is a heuristic based technique for solving for
both non-linearity and non-convexity of the sizing problems
[25,40]. A random generation of population members (possible
solutions) by GA led to the generation of the initial population.
Every possible solution is a decision vector code having upper and
lower constraints. The initial generation develops in the course of
consecutive iterations, and the tness function is estimated by
evaluating member components of each generation. Fig. 6 shows a
owchart of the genetic algorithm for the optimization process.
This problem includes three decision variables (ASPV, AWTG and NBS)
in the optimization process. In order to handle the decision variables, GA continuously probes the search space and best tness
function based on objective function. The parameters of the GA
algorithm is adjusted as; population size (n 20), crossover rate
(crate 0.5) and mutation rate (crate 0.5). MATLAB (R20013 b)
has been used for coding and simulate the proposed methodology.
The technical and cost parameters of the system components as
listed in Tables 3e7 are considered while designing the IREOM
model. Although a population size of 90 was found adequate for a
lower degree of optimization of the system after several iterations,
100 generations gave the optimal solution. In a majority of cases
however, the requisite number of generations came out to be
smaller than 60. The problem and the parameters were optimized
within the factors that specify used functions.
GA optimization procedure consists of following steps;
1) Initialization
a) Enter all metrological data such as hourly solar irradiance
and wind speed over a year.
b) Specication of the system components are initialized
(Tables 3e7).
c) Set the chromosome rate (C rate 0.5) and migration rate (M
rate 0.5).
d) Set EIR as 1.
e) Set population size of the variables as 20, with 100 as the
maximum iteration count.

812

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

Fig. 5. Flow chart for charging SOC of Battery system.

f) Set lower and upper bounds variable of SPV, WTG and BS

between 0 and 500.
g) Set iteration k as 1.
h) Calculate tness of each chromosomes i as Fi f Xi;j and
nd best t chromosome.
i) Print objective function value.

5.1. Results and discussion

Following the optimization procedure described above, the

results are obtained for different values of parameters considered

under the present study; project life time considered as 20 years,
discount rate and ination rate are considered as 0.11 and 0.07
respectively in the proposed IREOM model as shown in earlier
Fig. 4. The simulation time step Dt is taken as 15 min. The input
values of load prole for 24 h and monthly averages of hourly wind
speed and solar irradiation are used in HOMER which generated
simulated data for 8760 h (one year). Using genetic algorithm (GA)
based approach, optimization results in terms of optimal size, net
present cost (NPC) and cost of energy (COE) have been obtained on
the basis of the specications given in Tables 3e7.

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

813

Fig. 6. Flow chart for a genetic algorithm based approach.

The proposed methodology has been used to design an optimal

IREOM model to supply energy to four zones in the Chamarajanagar
district of Karnataka state in India. Description of results of proposed approach for all considered scenarios is given below.

5.1.1. Simulation results based on device types

Renewable energy potential differs across geographical regions;
a power source may be more dominant in one region than another.
Manufacturers also consider the random potential in a geographical region while manufacturing systems for the area. For example,
the area of the blade in a wind turbine would be dependent on the
wind density of the region. The present study considers three device types as solar panel, wind turbine and battery system, each
device type are taken for two different specications. The eight
possible combinations is obtained from 2n base as 2 and integer n as
the exponent of value 3. For example, combination 3 (1,2,1) means
solar panel type 1, wind turbine type 2 and battery system type 1.
Numerical value 1 represents device type 1 while 2 signies device
type 2. Results of eight such combinations are obtained in terms of
optimal size, total net present cost and cost of energy for the proposed study area.
Table 8 gives the results of eight combinations for zone 4 in
terms of optimal size, net present cost and cost of energy ensuring

EIR to be 1. It has been observed that combination 3 (1,2,1) is most

feasible for zone 4 because of minimum values of TNPC of $ 606397
and COE of $ 0.092 per kWh have been found. This combination
involves optimal size of 16 kW of MHG, 77 kW of BMG, 50 kW of
BGG, 165 SPV modules each having area of 1.63 m2, 13 wind turbines each having swept area of 4.26 m2 and 236 batteries each
having rating of 3.6 kVA. Similar combinations are also made for
zones 1, zone 2 and zone 3 and obtained results in terms of size is
shown in Fig. 7. The TNPC and COE considering different device
types combination are obtained and as given in Table 9.

5.1.2. Simulation results based on time scheduling of biomass

generator
As load demand varies during the 24 h in a day, the total hours in
a day categorized as peak and off peak hours to represent load
demand. The period between 6:00 p.m. 11:00 p.m. is considered
maximum peak hours and the hours between 1:00 a.m. - 6:00 a.m.
are considered minimum peak hours. For optimal power scheduling purpose, the 24 h of the day categorized into two shifts of 12 h
each are considered in the present study. Therefore, majority of the
load demand during a 12 h shift can be met by the bio-mass
generator and, the remaining demand can be fullled through
other generators such as SPV, WTG, BGG and MHP. To arrive at the

Table 8
Results of optimal size of system components for zone 4.
Combinations

1
2
3
4
5
6
7
8

Device types

Optimal Nos.

SPV

WTG

BS

SPV

WTG

BS

UL

EIR

DE

1
1
1
1
2
2
2
2

1
1
2
2
1
1
2
2

1
2
1
2
1
2
1
2

174
268
165
204
5
12
7
157

16
12
13
21
46
67
43
46

233
350
236
358
325
514
333
379

0
0
0
0
0
0
0
0

1
1
1
1
1
1
1
1

2.24
5.84
2.33
6.54
2.39
5.17
2.42
8.32

TNPC ($)

COE ($/kWh)

616759
762126
606397
748160
641237
841718
649046
792762

0.094
0.111
0.092
0.108
0.099
0.123
0.102
0.112

814

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

Fig. 7. Optimal sizes of SPV, WTG and battery system for all four zones.

Table 9
Results of total net present cost and cost of energy for all four zones.
Zone 1
Combinations
1
2
3
4
5
6
7
8

Zone 2

Zone 3

Zone 4

TNPC ($)

COE ($/kWh)

TNPC ($)

COE ($/kWh)

TNPC ($)

COE ($/kWh)

TNPC ($)

COE ($/kWh)

276768
345247
273164
327376
289082
381739
290134
337738

0.103
0.124
0.099
0.114
0.108
0.141
0.11
0.108

665716
840517
661060
824,148
778345
966061
761225
960955

0.112
0.133
0.108
0.128
0.133
0.16
0.13
0.161

632077
784052
642289
775942
726536
919507
715423
902238

0.094
0.115
0.097
0.114
0.113
0.135
0.112
0.121

616759
762126
606397
748160
641237
841718
649046
792762

0.094
0.111
0.092
0.108
0.099
0.123
0.102
0.112

optimal power generating schedule of the biomass generator, two

different time schedules as 1:00 a.m.e1:00 p.m. (TS1) and 1:00
p.m.e1:00 a.m. (TS2) are considered. The schematic of biomass and
biogas time schedules (TS1 and TS2) are shown in Figs. 8 and 9.
Based on the time schedules of TS1 and TS2 of biomass generator, the time schedule during 1:00 a.m. e 1:00 p.m. (TS2) was
found to be most feasible for all four zones. This is because during
this time, power generation matches peak load demand and requires less power for energy storage. This results in decrease in the
system cost as number of batteries required is less.
Table 10 gives the optimized results of optimal number of system components with battery for zone 4 considering different time
schedule of biomass generator. It reveals that time schedule of TS2
is found to be most feasible because the generator in operations
during peak hours to met the peak load demand during this time. It
involves optimal storage size of 236 batteries each having capacity
of 3.6 kVA to fulll the load requirement. The minimum COE and
NPC have been found as $ 0.092 per kWh and $ 606247 respectively.
Similarly, optimum numbers of batteries for zone 1, zone 2 and
zone 3 are obtained as presented in Fig. 10. The TNPC and COE
considering different time scheduling of biomass generator are
obtained and as given in Table 11.

5.1.3. Simulation results based on resource combination

Characteristics of renewable energy resources are random in
nature. It means that one resource may be more readily available in
one area than in another. Thus, to arrive at a feasible power
generating solution in a region, it is necessary to consider the
availability of resources available, and combine them (resource
combination) optimally to achieve the desired power generation at
the least possible cost. Out of the ve resources considered in this
study, MHP, BMG and BGG generators are considered non decision
variables and the total annual energy generated by these sources
are xed. On the other hand, solar and wind energy are considered
decision variables as the degree of their inclusion in the system can
be controlled. Their total annual energy generation depends on
active area of PV panels and swept area of wind turbine. The
combination of MHP, BMG, BGG and battery system with solar and/
or wind system are considered under the present study.
Based on the resource combination, Scenario based results in
terms of optimum number of system components; TNPC and COE
have been computed for all the four zones using GA based
approach. Three different scenarios of IRE models for all four zones
have been considered for the proposed study area. Table 12 shows
three different scenario results of S1, S2 and S3. It is seen from
Table 12 that, the scenario S2 (MHP-BMG-BGG-WTG-BS) have been

Fig. 8. Time schedule TS1 of BMG and BGG.

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

815

Fig. 9. Time schedule TS2 of BMG and BGG.

Table 10
Results of optimal numbers, total net present cost and cost of energy of system components for zone 4.
Time schedules of BMG

SPV (Nos.)

WTG (Nos.)

BS (Nos.)

UL

EIR

DE

TNPC ($)

COE ($/kWh)

TS1
TS2

173
165

31
13

455
236

0
0

1
1

7.96
2.33

854933
606247

0.121
0.092

Fig. 10. Optimal number of SPV, WTG and battery system for all four zones.

Table 11
Results of total net present cost and cost of energy for zone 4.
Time schedules

TS1
TS2

Zone 1

Zone 2

Zone 3

Zone 4

TNPC ($)

COE ($/kWh)

TNPC ($)

COE ($/kWh)

TNPC ($)

COE ($/kWh)

TNPC ($)

COE ($/kWh)

346599
273164

0.099
0.099

910347
661060

0.14
0.108

910797
661060

0.127
0.108

854933
606247

0.121
0.092

Table 12
Results of optimum sizes, total net present cost and cost of energy for zone 4.
Scenarios

Resource combination

MHG (kW)

BMG (kW)

BGG (kW)

SPV (Nos.)

WTG (Nos.)

BS (Nos.)

TNPC ($)

COE ($/kWh)

S1
S2
S3

MHG-BMG-BGG-SPV-WTG-BS
MHG-BMG-BGGWTG-BS
MHG-BMG-BGG-SPV-BS

16
16
16

77
77
77

50
50
50

165
e
278

13
49
e

236
260
196

606367
605376
623547

0.092
0.087
0.093

Table 13
Results of total net present cost and cost of energy for zone 4.
Scenarios

S1
S2
S3

ZONE 1

ZONE 2

ZONE 3

ZONE 4

TNPC ($)

COE ($/kWh)

TNPC ($)

COE ($/kWh)

TNPC ($)

COE ($/kWh)

TNPC ($)

COE ($/kWh)

273149
252440
312104

0.099
0.089
0.113

661030
860880
728398

0.108
0.122
0.12

632107
826325
661105

0.094
0.119
0.098

606367
605376
623547

0.092
0.087
0.093

816

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

Fig. 11. Optimal number of SPV panels, WTG and BS for the optimized IRE models considering different resources combinations.

micro hydro, biomass, biogas and wind turbine generators are used
to supply the required load demand. SPV has not been considered in
the model and battery bank is used to store excess energy and deliver
it in decit conditions. Optimal values of 49 wind turbines, each
having swept area of 4.26 m2 and 260 batteries, each having capacity
of 3.6 kVA have been obtained using GA based approach. The model
was found to have minimum TNPC of $ 605376 and minimum COE of
$ 0.08 per kWh for the proposed study area. The total annual energy
produced by system components was found to be as 140160 kWh/yr
of MHG, 337260 kWh/yr of BMG, 182500 kWh/yr of BGG and
143950 kWh/yr of WTG with battery storage energy of 80104 kWh/
yr against the load demand of 753010 kWh/yr. Battery storage energy was found to be 80104 kWh which could be utilized in power
decit conditions. The total surplus energy generated from all system components was found to be as 5.37%. This excess energy can be
used as a dump load for cooking purpose. Thus, the model is capable

found minimum NPC and COE. This model involves optimal sizes of
16 kW of MHG, 77 kW of BMG, 50 kW of BGG, 49 wind turbines
each having swept area of 4.26 m2 and 260 batteries each having
rating of 3.6 kVA. The TNPC and COE of system was found to be $
605376 and $ 0.087 per kWh respectively. Similarly, result of
optimal size for zone 1, zone 2 and zone 3 are obtained and presented in Fig. 11. The results of TNPC and COE considering different
resources combination are obtained and are as given in Table 13.
Considering the different device type combinations, time
schedule of biomass generator and resources combinations for all
the zones, optimal solution are obtained as given in Table 14. It can
be seen that for zone 4, device type combination 3(1,2,1) has been
found to be optimal for time schedule of biomass generator TS2 and
resource combination S2.
For zone 4, MHP/BMG/BGG/WTG and battery system is considered for electrication of the proposed study area. In this model,

Table 14
Results showing all the three scenarios and their optimum solution by considering TNPC, COE for all four zones.
Zones

Device type combinations

Time schedules of biomass generator

Resource combinations

Zone
Zone
Zone
Zone

3(1,2,1)
3(1,2,1)
1(1,1,1)
3(1,2,1)

TS2
TS2
TS2
TS2

S2
S1
S1
S2

1
2
3
4

Table 15
Summarized results in terms optimal size, TNPC, COE and other system parameters.
Parameters

Zone 1

Zone 2

Zone 3

Zone 4

Total annual load (kWh)

IRE models

291840
MHP-BMG-BGG-WTG-BS

652072
BMG-BGG-SPV-WTG-BS

737578
MHP-BMG-BGG-SPVWTG-BS

753010
MHG-BMG-BGG-WTG BS

6
30
15
e
22
121

e
76
56
287
17
202

15
78
50
244
8
213

16
77
50
e
49
260

52560 (14.33% )
131400 (35.83%)
54750 (14.93% )
e
95323 (25.99% )
32654 (8.92%)
334033
32654
366688
0
10.73
252440
0.089

e
332880 (46.13%)
204400 (28.32% )
134447 (18.63%)
49942 (6.92%)
59629 (7.63%)
721669
59629
781298
0
8.19
661030
0.108

131400 (16.71%)
341640 (43.44%)
182500 (23.20%)
114303 (14.53%)
16588 (2.11%)
65882 (7.73%)
786431
65882
852313
0
5.28
632107
0.094

Sizes
MHP (kW)
BMG (kW)
BGG (kW)
SPV (Nos.)
WTG (Nos.)
BS (Nos.)
Energy generation
MHG (kWh)
BMG (kWh)
BGG (kWh)
SPV (kWh)
WTG (kWh)
Battery system (kWh)
Total energy (kWh)
Battery-INV (kWh)
Total energy (kWh) (RES Battery-INV)
Unmet load
Excess energy (%)
TNPC ($)
COE ($/kWh)

140160 17.14%)
337260 (41.26%)
182500 (22.32%)
e
157581 (19.28%)
80104 (8.92%)
817501
80104
897605
0
5.37
605376
0.087

S. Rajanna, R.P. Saini / Energy 111 (2016) 803e817

of meeting the load demand of the area without inclusion of SPV.

Similar procedure has been followed for zone1, zone 2 and zone 3
and obtained results in terms of optimal size, TNPC and COE and
other parameters are given in Table 15.
6. Conclusions
The present study has been carried out to develop an optimized
IRE model with battery storage system to supply power to different
load sectors in four different zones of a remote area in Karnataka,
India. GA has been used to optimally size the SPV panels, wind
turbines and battery storage systems on the basis of total sunshine
area required by the solar panel, total swept area occupied by the
wind turbine and SOC of the battery system. EIR is taken to be 1 in
all cases, and minimizing COE is considered the main objective
function. Device combination 3(1, 2, 1), scheduling time of 1:00
p.m. 1:00 a.m. (ST2) and scenario S2 (MHG-BMG-BGG-WTG BS)
have been found to be optimum for zone 4. Based on these optimum solutions, optimal sizes of system components came out to be
16 kW of MHG, 77 kW of BMG, 50 kW of BGG, 49 WTG each having
swept area of 4.26 m2 and 260 batteries each having capacity of
3.6 kVA with minimum COE of $ 0.087 per kWh and TNPC of $
605376. Similarly, optimum sizes, least NPC and COE for zones 1,
zone 2 and zone 3 are worked out and presented in Table 15. Based
on the results obtained for the three scenarios, following conclusion are drawn:
 230 WP solar modules each having area of 1.63 m2 (Type 1),
1.4 kW wind turbines each having swept area of 4.26 m2 (Type
2) and 24 V battery rating of 150 Ah, 24 V are suggested for zone
1, zone 2 and zone 4. Wind turbine generator (WTG) of 1 kW
having swept area of 3.14 m2 along with Type 1 solar and battery
is suggested for zone 3. SPV and BS are common for all zones
while wind turbine varies across zones.
 Number of batteries can be reduced if biomass generators are
operated during time scheduling TS2, thereby reducing the
number of batteries and cost of the system.
 Scenario S2 (MHP-BMG-BGG-WTG-BS) for zones 1 and 4, scenario S1 (BMG-BGG-SPV-WTG-BS) for zone 2, and scenario S1
(MHP-BMG-BGG-SPV-WTG-BS) for zone 3 was found most
feasible to supply power.
References
[1] Ramakumar R, Sudhakara Shetty P, Ashenayi K. A linear programming
approach to the design of integrated renewable energy systems for developing countries. IEEE Trans Energy Convers 1986;1(4):18e24.
[2] Joshi BS, Bansal NK. Decentralized energy planning model for a typical village
in India. Energy 1992;17(9):869e76.
[3] Yap Wai Kean, Karri Vishy. An off-grid hybrid PV/diesel model as a planning
and design tool, incorporating dynamic and ANN modeling techniques. Renew
energy 2015;78(1):42e50.
[4] Ramakumar R, Butler NG, Rodriguez AP, Mani S. Economic aspects of
advanced energy technologies. Proc IEEE 1993;81(3):317e33.
[5] Askarzadesh A, Leandro dos SC. A novel framework for optimization of a grid
independent hybrid renewable energy system: a case study of Iran. Sol Energy
2015;112(1):383e96.
[6] Ramakumar R. Renewable energy sources and developing countries. IEEE
Trans power App Syst 1983;102(2):502e10.
[7] Ramakumar R, Abouzahr M, Ashenay K. A knowledge-based approach to the
design of integrated renewable energy systems. IEEE Trans Energy Convers
1992;7(4):648e59.
[8] Cano Antonio, Francisco J, Higinio S, Fernandez M. Optimal sizing of standalone hybrid systems based on PV/WT/FC by using several methodologies.
J Energy Inst 2014;87(1):330e40.
[9] Diaf S, Notton G, Belhamel M, Haddadi M, Louche A. Design and technoeconomical optimization for hybrid PV/wind system under various meteorological conditions. Appl Energy 2008;85:968e87.
[10] Hongxing Yang, Wei Zhou, Chengzhi Lou. Optimal design and technoeconomic analysis of a hybrid solarewind power generation system. Appl
Energy 2009;86(1):163e9.

817

[11] Koutroulis E, Dionysian K, Antonis P, Kostas K. Methodology for optimal sizing

of stand-alone photovoltaic/wind-generator systems using genetic algorithms. Sol Energy 2006;80:1072e88.
[12] Koutroulis E, Kolokotsa D. Design optimization of desalination systems
power-supplied by PV and wind energy sources. Desalination 2010;258(1-3):
171e81.
[13] Upadhyay Subho, Sharma MP. Development of hybrid energy system with
cycle charging strategy using particle swarm optimization for a remote area in
India. Renew Energy 2015;77:586e98.
[14] Upadhyay Subho, Sharma MP. Selection of a suitable energy management
strategy for a hybrid energy system in a remote rural area of India. Energy
2016;94:352e66.
[15] Ismail MS, Moghavvemi M, Mahlia TMI. Genetic algorithm based optimization
on modeling and design of hybrid renewable energy systems. Energy Convers
Manag 2014;85(1):120e30.
[16] S. Rajanna, R.P. Saini. Optimal modeling of solar/biogas/biomass based IRE
system for a remote area electrication. Proceeding of IEEE conference PIICON-2014.
[17] Chauhan A, Saini RP. Discrete harmony search based size optimization of
Integrated Renewable Energy System for remote rural areas of Uttarakhand
state in India. Renew Energy 2016;94:587e604.
[18] Chauhan A, Saini RP. Techno-economic optimization based approach for energy management of a stand-alone integrated renewable energy system for
remote areas of India. Energy 2016;94:138e56.
[19] Wang Lingfeng, Sigh Chanan. Multi-criteria design of hybrid power generation systems based on a modied particle swarm optimization algorithm. IEEE
Trans Energy Convers 2009;24(1):163e72.
[20] Patil AB, Saini RP, Sharma MP. Development of IREOM model based on
seasonally varying load prole for hilly remote areas of Uttarakhand State in
India. Energy 2011;36. 5690e5702.
[21] Maleki Akbar, Askarzadeh Alireza. Optimal sizing of a PV/wind/diesel system
with battery storage for electrication to an off-grid remote region: a case
study of Rafsanjan, Iran. Sustain Energy Tech Assess 2014;7:147e53.
[22] Heydari Ali, Askarzadeh Alireza. Optimization of a biomass-based photovoltaic power plant for an off-grid application subject to loss of power supply
probability concept. Appl Energy 2016;165:601e11.
[23] Askarzadeh Alireza. Developing a discrete harmony search algorithm for size
optimization of windephotovoltaic hybrid energy system. Sol Energy
2013;98:190e5.
[24] Chang Kuo-Hao, Lin Grace. Optimal design of hybrid renewable energy systems using simulation optimization. Simul Model Pract Theory 2015;52:
40e51.
[25] Ranaboldo Matteo, Garca-Villoria Alberto, Ferrer-Mart Laia, Moreno Rafael
Pastor. A meta-heuristic method to design off-grid community electrication
projects with renewable energies. Energy 2015;93:2467e82.
[26] Arabali A, Ghofrani M, Amoli ME, Fadali MS, Baghzouz Y. Genetic algorithm
based optimization approach for energy management. IEEE Trans Power Deliv
2009;28(1):162e70.
[27] Yang Hongxing, Lu Lin, Zhou Wei. A Novel optimization sizing model for
hybrid solar-wind power generation system. Sol Energy 2007;81:76e84.
[28] Kaabeche A, Belhamel M, Ibtiouen R. Sizing optimization of grid-independent
hybrid photovoltaic/wind power generation system. Energy 2011;36:
1214e22.
[29] Olatomiwa Lanre, Mekhilef Saad, Huda ASN, Ohunakin OS. Economic evaluation of hybrid energy systems for rural electrication in six geopolitical
zones of Nigeria. Renew Energy 2015;83:435e46.
[30] Lujano-Rojas JM, Dufo-Lopez Rodolfo, Bernal-Agustin, Bernal-Agustn. Technical and economic effects of charge controller operation and columbic efciency on stand-alone hybrid power system. Energy Convers Manag 2014;86:
709e16.
[31] Borowy BS, Salameh ZM. Methodology for optimally sizing the combination of
a battery bank and PV array in a wind/PV hybrid system. IEEE Trans Energy
Convers 1996;11(2):367e75.
[32] Karaki SH, Chedid RB, Ramadan R. Probabilistic performance assessment of
autonomous solar-wind energy conversion systems. IEEE Trans Energy
Convers 1999;14(3):766e72.
[33] Tina G, Gagliano S, Raiti S. Hybrid Solar/Wind power system probabilistic
modeling for long-term performance assessment. Sol Energy 2006;80:578e88.
[34] Rajanna S, Saini RP. Modeling of integrated renewable energy system for
electrication of a remote area in India. Renew energy 2016;90:175e87.
[35] Michael AM, ojha TP. Principles of agricultural engineering. Jain Brothers;
2013.
[36] Subramanian K. Engineering hydrology. McGraw-Hill Companies; 2012.
[37] Gupta A, Saini RP, Sharma MP. Steady-state modeling of hybrid energy system
for off grid electrication of cluster of villages. Renew Energy 2010;35(1):
520e35.
[38] Chauhan A, Saini RP. Renewable energy based off-grid rural electrication in
Uttarakhand state of India: technology options, modeling method, barriers
and recommendations. Renew Sustain Energy Rev 2015;51:662e81.
[39] Gupta Ajai, Saini RP, Sharma MP. Modeling of hybrid energy system Part I:
problem formulation and model development. Renew Energy 2011;36:
459e65.
[40] Rajanna S, Saini RP. A review on planning, congurations, modeling and
optimization techniques of hybrid renewable energy systems for off grid
applications. Renew Sustain Energy Rev 2016;58:376e96.