Renewable Energy 99 (2016) 1161e1180

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Employing demand side management for selection of suitable

scenario-wise isolated integrated renewal energy models in an Indian
remote rural 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 1 May 2016
Received in revised form
30 June 2016
Accepted 8 August 2016

The present study focuses on optimal sizing of an integrated renewable energy (IRE) system with battery
bank to meet the load demand of a cluster of four zones in Karnataka state of India. Hourly load demand
has been simulated using HOMER software based on three scenarios - high investment with low (HILR)
rating appliances, medium investment with moderate (MIMR) rating appliances, and low investment
with high (LIHR) rating appliances. Simulation results of optimal size, total net present cost (TNPC) and
cost of energy (COE) have been obtained using genetic algorithm (GA) considering different scenarios
with and without DSM strategy by keeping energy index at 1. Results of three scenarios with DSM
strategy were compared with those without DSM strategy. Based on the comparison of the results,
scenario of HILR with DSM strategy is found to have minimum TNPC and COE for zone 1 and zone 2,
while scenario of MIMR with DSM for zone 3 and LIHR with DSM for zone 4 are found to have least TNPC
and COE. Further, obtained results of TNPC of GA have been validated by comparing PSO technique. It has
been observed that, results of GA and PSO were found to be almost same.
2016 Elsevier Ltd. All rights reserved.

Keywords:
GA
IRES
Demand side management
Net present cost
Remote rural area

1. Introduction
Energy is generally accepted as a vital component to a nation's
growth and development of its citizens. Power generation across
the world is currently dependent to a large extent, on conventional
sources of energy such as coal, gas, oil, etc. However, the use of such
power sources has a damaging effect on environment and society,
and gives rise to various environmental and health problems.
Renewable energy resources such as biomass, micro hydro, wind,
solar, etc. offer a way of generating power with considerably less
damage to the environment. Various social and economic benets
also accompany by generating power from renewable energy resources [1e4]. It is noteworthy that economical and efcient utilization of renewable energy resources warrants development of
optimal models through which demand for energy can be met.
Further, IRE system modeling is a complex process that involves
formulation of mathematical models pertaining to each component
[5e7]. Efcient IRE system modeling depends on effective load
demand management. Demand-side management (DSM) offers

* Corresponding author.
E-mail address: srajannamce@gmail.com (S. Rajanna).
http://dx.doi.org/10.1016/j.renene.2016.08.024
0960-1481/ 2016 Elsevier Ltd. All rights reserved.

feasible solutions for enhancing IRE system efciency by adjusting

the time and the quantity of electricity use by shifting peak load to
off-peak periods. This method also gives a uniform daily peak and
evens out valleys in the load curve. Further, DSM minimizes the cost
involved in power generation by reducing system size [8e12].
Various authors have advocated the use of DSM for optimal
modeling of IRE systems. Matallanas et al. [13] proposed control
stratagem to enhance energy performance of PV energy system
planning by consumers and maximize usage of power produced
utilizing neural networks. Gudi N et al. [14] developed a binary
particle swarm optimization (BPSO) algorithm for optimizing DSM
strategy in a household function resulting in system cost savings.
Kyriakarakos G et al. [15] proposed an intelligent DSM strategy on
the basis of grey prediction algorithms which could successfully
meet system design principles and ensure efcacious functioning of
an independent poly-generation micro grid in remote areas. Stathopoulos M et al. [16] developed a sizing algorithm for a hybrid
renewable energy system in remote area using DSM strategy.
Upadhyay S and Sharma MP [17] proposed three energy management strategies employing particle swarm optimization, genetic
algorithm and biogeography based optimization techniques while
considering the index ratio at 1. They suggested cycle charging

1162

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Nomenclature
AC
ACnpv
ASPV
Aw
BGG
BMG
BATT/BS
BC
CFL
COE
CN

alternating current
net present value of appliances cost ($)
area of SPV (m2)
swept area of wind turbine rotor (m2)
biogas generator
biomass generator
battery system
bi-directional converter
compact uorescent
cost of energy ($/kWh)
runoff curve number for hydrological sail (CN 40e58
for dense forest)
Cb
initial capacity of battery system
CD
cow dung (ton/yr)
CP
power co-efcient (0.45)
COA
cost of appliances
CNPV
net present capital cost ($)
CRF
capital recovery factor
CVBG
caloric value of biogas (4700 kcal/kg)
CVBG
caloric value of biomass (4015 kcal/kg)
D
duration of unavailability of load (hour)
DC
direct current
DOD
depth of discharge
DSM
Demand side management
Egen
total annual generation (kWh)
EENS
expected energy not supplied (kWh)
EIR
energy index ratio
ESL
shiftable load (kWh)
ENSL
non-shiftable load (kWh)
FCNPV
net present fuel cost ($/ton)
g
acceleration due to gravity (m/s2)
GA
Genetic algorithm
hnet
net head (m)
HILR
high investment with low rating
HT
solar radiation (kW/m2)
I
monthly rainfall (mm)
IL
incandescent lamp
IRE
integrated renewable energy
L
average annual power load (kW)
LED
Light emitting diode
LIHR
low investment with high rating
Lw
length of the watershed (m)
MHG
micro hydro generator
MIMR
medium investment with moderate rating
NSPV
solar panels (Nos.)
NWTG
wind turbine (Nos.)
NBS
battery system (Nos.)
NREL
national renewable energy limited
OM
Operation and maintenance
OM npv Net present value of O&M ($)
PBMG(t) power output of BMG (kWh)

strategy to be more cost effective in comparison to load following

and peak shaving approaches. Chauhan A and Saini RP [18]
considered techno-economic aspects using energy management
for an integrated renewable energy system to fulll the energy
demand of a group of villages in Uttarakhand state. A load shifting
based DSM strategy was been considered for the proposed system.
They suggested that with DSM strategy was found to be most cost
feasible solution when compare to without DSM strategy. Patil et al.

PBGG(t)
PWTG(t)
PSPV(t)
Pr
POA
PSO
Q
Qd
RNPV
S
SONPT
SOFFPT
SOC
Tc
Tp
TNOA
TNPC
Vr
Vci
Vco
WTG

hMHP
hBGG
hBMG
hSPV
hWTG
hINV
hRECT
hBC
hBD
s
aBMG
aBGG
aMHG
aSPV
aWTG
aBS
aCONV
bBMG
bBGG
bMHG
bSPV
bWTG
bBS
bCS
rW
ra
5

m
lAFC
t
$

power output of BGG (kWh)

power output of WTG (kWh)
power output of SPV system (kWh)
rated output power of wind turbine (kW)
prices of appliances ($)
Particle swarm optimization
discharge (m3/s)
direct surface runoff depth (mm)
net present replacement cost
maximum potential retention (mm)
shift able load on peak time
shiftable load off peak time
State of charge
time of concentration (hour)
time of peak runoff (hour)
total number of appliances
total net present cost ($)
rated speed of the wind turbine (m/s)
cut in speed (m/s)
cutout speed (m/s)
wind turbine generator
Efciency of micro hydro generator (%)
efciency of biogas generator (%)
efciency of biomass generator (%)
efciency of SPV panel (%)
efciency of wind turbine generator (%)
efciency of inverter system (%)
efciency of rectier system (%)
battery charging efciency
battery discharge efciency
hourly self discharge rate
capital cost of BMG ($/kW)
capital cost of BGG ($/kW)
capital cost of MHG ($/kW)
capital cost of SPV ($/kW)
capital cost of WTG ($/kW)
capital cost of battery system (per battery)
capital cost of bi-directional converter
maintenance cost of BMG ($/yr)
maintenance cost of BGG ($/yr)
maintenance cost of MHG ($/yr)
maintenance cost of SPV ($/yr)
maintenance cost of WTG ($/yr)
maintenance cost of BS ($/yr)
O&M cost of BC ($/yr)
density of water (1000 kg/m3)
density of air (1.23 kg/m3)
interest rate (%)
escalation rate (%)
average biomass and biogas fuel price ($/ton)
project life time
U.S. Dollars

[19] proposed IRE model for Uttarakhand state of India. Ramakumar et al. [20] proposed a linear program based IRE model
developed for developing country.
Randa Kallel et al. [21] highlighted advantages of the suggested
strategy for residential hybrid system functions under different
conditions and conducted a comparative study between energy
management strategies with and without DSM strategy. Rajanna S
and Saini RP [22] proposed a genetic algorithm based optimal IRE

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

1163

investment scenarios without and with DSM strategy is emphasized to develop optimal IRE models. Further, present study also
presents for a comparative analysis between the two strategies
investigated.

model for meeting energy demand in various sectors of an Indian

remote rural area. They worked out decit in energy and customer
interruption cost for different optimization power factor to quantied the system reliability. Ariel Miara et al. [23] proposed an
optimizing environmental and social benets through three DSM
scenarios based on capital investment costs for the Northeastern
US. Ramakumar et al. [24] proposed a design tool IRES-KB on the
basis of knowledge-based system and described scenario based
models remote rural Indian areas. Alireza and Leandro [10] developed three grid independent solar and wind based hybrid systems
for electrication to a small load area in Kerman Iron. They found
cost effective and reliable based PV/WT/battery hybrid system for
study area. In Refs. [25,26] develops an economic and technical
based hybrid model considering various metrological conditions.
Ardakani and Ardehali [27] developed electrical energy consumption (EEC) forecasting models based on forecasting period of
2010e2030 using data of 1967e2009 through ANN and other
promising optimization techniques. They analyzed the effects of
DSM on EEC forecasting and also determine how input data affect
EEC forecasting using two types of input data for 1967e2009 such
as EEC data and socio-economic indicators data, which comprises
of GDP, IMP, EXP, and POP.
Based on the literature review, it has been found that various
researchers carried out studies on DSM strategies. However, no
study has been attempted, considering investment based scenario
(electrical appliances with low price with high rating, medium
price with moderate rating and high price with low rating) with
peak load shifting based DSM approach which leaves a major gap in
existing literature. The present paper discusses size optimization of
IRE models comprising micro hydro, solar, wind, biomass, biogas
with battery storage system for a study area considering the economic and reliability worth. To maintain accuracy in the obtained
results, genetic algorithm is used for modeling of considered integrated renewable energy (IRE) system. Based on varying load demand and consumption of energy patterns in the study area, three

2. Load and resources assessment

Based on eld survey in the form of questionnaire, the load
demand estimation has been estimated considering the minimum
desirable load and renewable energy resources available in the
study area. The detail study of study area, load assessment and
resources assessment through mathematical modeling are discussed below.
2.1. Study area
Based on the comprehensive assessment on cluster of unelectried villages in Chamarajanagar district in Karnataka 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 study area.
Detailed description of zone wise demography data of the study
area has been discussed under earlier study carried out by the
authors [22]. The general information of the study area provided in
the earlier study are summarized and given in Table 1.
2.2. Load estimation
The present study considers three investment scenarios based
appliances - high investment with low rating (HILR), medium investment with moderate rating (MIMR) and low investment with
high rating (LIHR). Each scenario includes the following appliances
categorized into residential, community and commercial sectors:
high investment appliances - LED Light, LED TV, LED Leodan Street
light, high efciency refrigerator (Tier-3), BEE 5 star ceiling fan (R3);

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

Table 2
Technology key with three investments based appliances for residential, community and commercial sectors.
Scenarios

Residential

Community

Commercial

Lighting

Entrainment TV

Cooling fan

Lighting

Street lighting

Cooling fan

Refrigerator

Lighting

HILR
MIMR
LIHR

LED
CFL
IL

TV-3
TV-2
TV-1

R3
R2
R1

LED
CFL
IL

ST3
ST2
ST1

R3
R2
R1

Tier-3
Tier-2
Tier -1

LED
CFL
IL

Technology key and investment for residential, community and commercial sectors
Appliances

Prices ($)

Appliances

Prices ($)

IL-incandescent lamp (30 W)

CFL-compact uorescent lamp (18 W)
LED-light emitting diode (11 W)
R1-BEE 3 star ceiling fan ( 56 W)
R2-BEE 4 star ceiling fan (55 W)
R3-BEE 5 star ceiling fan (52)
TV-1-Regular (140 W)
TV-2-Lequid crystal diode (90 W)

0.38
5.71
7.04
22.53
28.38
29.73
84.10
112.63

TV-3-light emitting diode (40 W)

SL1-uoracent tube type (60 W)
SL2-leodan compact uorescent lamp (36 W)
SL3-leodan light emitting diode (20 W)
Tier-1- refrigerator (475 W)
Tier-2-high efciency refrigerator ( 434 W)
Tier-3- high efciency refrigerator (417 W)

127.65
3.00
11.26
22.53
300.35
435.50
585.67

1164

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

medium investment appliances - CFL, LCD TV, CFL street light, high
efciency refrigerator (Tier-2), BEE 4 star ceiling fan (R2); and low
investment appliances - incandescent bulb, regular TV, uorescent
lamp, regular refrigerator (Tier-1), BEE 3 star ceiling fan (R1).
Detailed description regarding technology key, ratings and electrical appliances prices is given in Table 2.
Hourly load demand has been calculated as hourly segments in a
day based on operating hours and considered investment scenarios
as given in Table 2. An hourly load has been estimated for different
investment scenarios of LIHR, MIMR and HILR and hourly load
variation of three investment scenarios is shown in Fig. 6(aec). 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 considers different electrical
appliances in three distinct scenarios. For each zone, total load
demand has been computed based on power ratings and operating
hours of appliances in a day. Detailed description of load sectors,
considered scenarios and total electrical appliances for all four
zones are given in Table 3.
The data have been synthesized through HOMER software
considering estimated values of three investment scenarios based
hourly load with and without DSM are given in Tables 4 and 5. The
total estimated annual loads of zone 1, zone 2, zone 3 and zone 4 are
shown in Fig. 1.
2.3. Resources assessment
Mathematical models of system components of the IRE system
are essential before optimizing size because they give information
about the performance of system components in different operating conditions. The mathematical modeling of solar system, micro hydro generator, wind turbine generator, biogas generator,
biomass generator and battery storage system are developed based
on the available energy potential in the study area are discussed as
follows:
2.3.1. Micro hydro generator (MHG)
Micro hydro generator is designed considering low volumetric
ow rate and low head conditions. 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.

Actual runoff is calculated by using following equations [28,29];

Qd

I  0:2S2
I 0:8S

when I > 0:2S

(1)

25400
 254
S
CN
The peak rate of runoff or discharge of a watershed is calculated
by using following Equations [28,29];

p
TP 0:6TC TC
r
0:77
.
TC 0:0195 Lw 3 Hnet

(2)

0:0208  A  Qd
TP

(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).
Based on the availability of hydro potential in study locations,
water ow rate of 115 l/s (60%) in zone 1, 410 l/s (50%) in zone 3 and
437 l/s (50%) in zone 4 have been estimated based on proposed
approach. The estimated values of stream ow for zone 1, zone 3,
and zone 4 are shown in Fig. 3 [28,29].
The hourly power output of MHG system (PMHG) in kW at an
hour t can be computed based on estimated value of water stream
potential in the study location by using following mathematical Eq.
(4) [22] as;

PMHP t 9:81  Q  hnet  hMHP  rw =1000

(4)

where Q is discharge (m3/s), hnet is net head (m), hMGH is overall

efciency of micro hydro generator (80%), and rW is density of
water (1000 kg/m3).
2.3.2. Biogas generator (BGG)
The size of biogas generator depends on availability of total gas
yield in m3/day at location of study area (m3/day) and operating
hours per day of the internal combustion engine. The biogas
generator is proposed to operate for 10 h per day. Hourly output

Table 3
Categorization of load sectors with power rating of considered investment scenario for study area.
Load sectors

Parameters

Domestic

Household

Community

Hospital

Primary School
Community
Community hall
Commercial
Industrial

Appliances

Lighting
Ceiling fan
TV
Radio
Refrigerator
Lighting
Ceiling fan
Lighting
Ceiling fan
pumping water
street Lighting
Lighting
Ceiling fan
Lighting
Flour mill
Saw mills

Switching point (nos.)

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

Power rating

Nos. used

HILR

MIMR

LIHR

Zone 1

Zone 2

Zone 3

Zone 4

9
52
40
10
417
9
52
9
52
3675
20
9
52
9
3750
5000

18
55
90
10
434
18
55
18
55
3675
36
18
55
18
3750
5000

30
56
140
10
475
30
56
30
56
3675
60
30
56
30
3750
5000

604
302
302
302
1
3
3
12
12
3
30
1
1
5
3
1

790
395
395
395
1
3
3
12
12
3
39
1
1
6
3
2

1526
763
763
763
1
3
3
12
12
4
76
1
1
8
4
3

1560
780
780
780
1
3
3
12
12
4
78
1
1
8
4
3

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

1165

Table 4
Zone wise hourly estimated load prole in different investment scenarios without DSM strategy.
HILR

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

MIMR

LIHR

Zone 1

Zone 2

Zone 3

Zone 4

Zone 1

Zone 2

Zone 3

Zone 4

Zone 1

Zone 2

Zone 3

Zone 4

16.75
16.75
16.75
16.75
19.45
19.45
25.03
25.03
11.29
11.29
12.24
40.57
40.57
40.57
12.24
12.24
12.24
20.35
45.34
45.34
45.344
34.31
16.75
16.75

21.76
21.76
21.76
21.76
25.32
25.32
29.22
29.22
14.64
14.64
15.59
52.64
52.64
52.64
15.59
15.59
15.59
26.49
55.75
55.75
55.75
44.73
21.77
21.76

42.56
42.56
42.56
42.56
49.58
49.58
50.22
50.22
28.5
28.5
29.45
90.65
90.65
90.65
29.45
29.45
29.45
51.9
102.58
102.58
102.58
87.88
42.56
42.56

41.64
41.64
41.64
41.64
48.51
48.51
49.45
49.45
27.89
27.89
28.83
89.35
89.35
89.35
28.83
28.83
28.83
50.78
100.67
100.67
100.67
85.97
41.64
41.64

18.18
18.18
18.18
18.18
23.62
23.62
28.22
28.22
11.76
11.76
12.87
56.3
56.3
56.3
12.87
12.87
12.87
35.92
67.35
67.35
67.35
56.32
18.19
18.18

23.62
23.62
23.62
23.62
30.73
30.73
33.38
33.38
15.25
15.25
16.36
73.16
73.16
73.16
16.36
16.36
16.36
46.85
84.48
84.49
84.49
73.46
23.62
23.62

45.19
45.19
45.19
45.19
58.92
58.92
57.48
57.48
29.04
29.04
30.16
123.83
123.83
123.83
30.16
30.16
30.16
90.08
156.14
156.14
156.14
141.44
45.19
45.19

46.2
46.2
46.2
46.2
60.24
60.24
50.62
58.42
29.68
29.68
30.8
126
126
126
30.8
30.8
30.8
92.08
159.28
159.28
159.29
144.59
46.2
46.2

19.28
19.28
19.28
19.28
28.34
28.34
32.04
32.04
11.95
11.95
13.24
71.77
71.77
71.77
13.24
13.24
13.24
51.21
90.85
90.85
90.85
79.83
19.28
19.28

25.03
25.03
25.03
25.03
36.88
36.88
38.36
38.36
15.49
15.49
16.77
93.32
93.32
93.32
16.77
16.77
16.77
66.84
115.2
115.2
115.2
104.16
25.03
25.03

47.85
47.85
47.85
47.85
70.74
70.74
67.06
67.06
29.47
29.47
30.76
158.83
158.83
158.83
30.76
30.75
30.75
128.66
215.33
215.33
215.33
200.63
47.85
47.85

48.93
48.93
48.93
48.93
72.33
72.33
68.23
68.22
30.12
30.12
31.4
161.85
161.85
161.85
31.4
31.4
31.4
131.52
219.81
219.81
219.815
205.11
48.93
48.93

Table 5
Zone wise hourly estimated load prole in different investment scenarios with DSM strategy.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

HILR

MIMR

LIHR

Zone 1

Zone 2

Zone 3

Zone 4

Zone 1

Zone 2

Zone 3

Zone 4

Zone 1

Zone 2

Zone 3

Zone 4

16.75
16.75
16.75
16.75
19.47
19.47
25.03
25.03
27.54
27.54
28.49
24.32
24.32
24.32
23.26
23.26
23.26
20.35
34.31
34.31
34.31
34.31
16.75
16.75

21.76
21.76
21.76
21.76
25.32
25.32
29.22
29.22
35.89
35.89
36.84
31.39
31.39
31.39
26.61
26.61
26.61
26.49
44.73
44.73
44.73
44.73
21.77
21.77

42.56
42.56
42.56
42.56
49.58
49.58
50.22
50.22
58.5
58.5
59.45
60.65
60.65
60.65
44.15
44.15
44.15
51.9
87.88
87.88
87.88
87.88
42.56
42.56

41.64
41.64
41.64
41.64
48.51
48.51
49.45
49.45
57.89
57.89
58.83
59.35
59.35
59.35
43.53
43.53
43.53
50.78
85.97
85.97
85.97
85.97
41.64
41.64

18.18
18.18
18.18
18.18
23.65
23.61
28.2
28.22
28.01
28.01
29.12
40.05
40.05
40.05
23.9
23.9
23.9
35.92
56.32
56.32
56.32
56.32
18.18
18.18

23.62
23.62
23.627
23.62
30.73
30.73
33.38
33.38
36.5
36.5
37.61
51.91
51.91
51.91
27.39
27.39
27.39
46.85
73.46
73.46
73.46
73.46
23.62
23.62

45.19
45.19
45.19
45.19
58.92
58.92
57.48
57.48
54.04
54.04
55.16
98.83
98.83
98.83
44.86
44.86
44.86
90.08
141.44
141.44
141.44
141.44
45.19
45.19

46.2
46.2
46.2
46.206
60.24
60.24
50.634
58.43
54.68
54.68
55.8
101
101
101
45.5
45.5
45.5
92.08
144.58
144.58
144.58
144.58
46.2
46.2

19.277
19.277
19.277
19.277
28.337
28.337
32.036
32.036
28.201
28.201
29.487
55.517
55.517
55.517
24.262
24.262
24.262
51.211
79.827
79.827
79.827
79.827
19.277
19.277

25.025
25.025
25.025
25.025
36.875
36.875
38.36
38.36
36.74
36.74
38.02
72.07
72.07
72.07
27.80
27.80
27.80
66.84
104.18
104.18
104.18
104.18
25.03
25.03

47.85
47.85
47.85
47.85
70.74
70.74
67.06
67.06
50.72
50.72
52.01
137.58
137.58
137.58
45.46
45.46
45.46
128.66
200.64
200.64
200.64
200.64
47.85
47.85

48.93
48.93
48.93
48.93
72.335
72.33
68.22
68.22
51.37
51.37
52.65
140.60
140.60
140.60
46.10
46.10
46.10
131.52
205.11
205.11
205.12
205.12
48.93
48.93

Annual Load
demand (kWh)

Hours

1050000
550000
50000
HILR MIMR LIHR
Zone 1

HILR MIMR LIHR

Zone 2

HILR MIMR LIHR

Zone 3

HILR MIMR LIHR

Zone 4

Fig. 1. Total annual load demand considering three investment scenarios for all zones.

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Appliance cost
(millions)

1166

10
8
6
4
2
0
HILR MIMR LIHR

HILR MIMR LIHR

zone 1

HILR MIMR LIHR

zone 2

HILR MIMR LIHR

zone 3

zone 4

Fig. 2. Total appliances cost considering three investment scenarios.

power of a biogas generator (PBGG) in kW can be computed by using

mathematical Eq. (5) [32] as;



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

(5)

where PBGG is hourly output of biogas generator, hBGS is overall

energy conversion efciency of biogas generator (25%) and CVBG is
caloric value of biogas (4700 kcal/kg).
2.3.3. Biomass generator (BMG)
The size of biomass generator (PBMG in kW) depends on availability of total fuel wood in ton/yr of the study area and operating
hour of generator in a day. Hourly output power of a biomass
generator (PBMG) in kW has been estimated by using following
mathematical Eq. (6) [33] as;

Total fuel wood ton=yr  CVBM  hBMG  1000

PBMG t
365  860  operating hours per day
(6)
where PBMG is hourly power output of biomass generator, hBMG is
overall energy conversion efciency of biomass generator (20%)
and CVBM is caloric value of biomass (4015 kcal/kg).
2.3.4. Scheduling of bio-generators
The scheduling of biogas and biomass generators are xed based
on the maximum load in an hours. The output power of biogenerators of HILR, MIMR and LIHR based scenarios are shown in
Fig. 4(a.c).
2.3.5. Solar photovoltaic panel (SPV)
Solar energy potential has been estimated based on solar radiation obtained from National Renewable Energy Limited (NREL).
The size of SPV panel at an hour t can be estimated based on the
available solar radiation of the study area, active area of SPV panel
and efciency of the panel. Hourly power output of SPV panel can
be estimated using mathematical modeling Eq. (7) [30] as;

PSPV t hSPV  HT t  ASPV

(7)

where ASPV is the active area of SPV panel (1.63 m2), HT is the solar
radiation in kW/m2 and hspv denotes the efciency of the SPV
panels (14.4%).

2.3.6. Wind turbine generator (WTG)

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. The size of wind turbine depends on swept
area, rated speed, cut-in and cut-out speed of the wind turbine. At
an hour t, electrical power output of a wind turbine generator
(PWTG) in kW can be calculated based on following mathematical
modeling Eqs. [10];

8
<0
PW t Pr
:
Pr

v3 t  v3ci
v3r

v3ci

vt  vci or vt  vco
vci < vt < vr
vr < vt < vco

(8)

Rated power of the wind turbine can be calculated using

following equations;

1
Pr $Cp $ra $hg $Aw $Vr3
2

(9)

where Aw is swept area of wind turbine rotor, ra is air density

(1.23 kg/m3), Cp is power coefcient of the proposed wind turbine
(0.45), and hg is generator efciency (85%) and rated speed, cut-in
and cut-out speed respectively represent vr, vci and vco of wind
turbine generator.
2.3.7. Battery storage system
The battery bank should be permitted to discharge as per the
DOD (depth of discharge) as specied by the designer at the
beginning of optimal sizing process. It can be expressed by
following Eq. [33,38]

Cmin DOD  Cb

(10)

The SOC charging process can be expressed by Eq. (11) [10,38]

as:

SOCt 1 SOCt$1  s EPDC t hrect EPAC t

 PLD t$hBC

(11)

As maximum energy stored in the battery system cannot surpass maximum state of charge (SOC max), the following condition
must be met during optimization.

SOC t  SOC max

(12)

SOC max takes the value of nominal capacity of the battery bank
(Cb).
The discharge process of the SOC is expressed by Eq. (12) [10,38]
as:

SOCt 1 SOCt$1  s  hinv NPAC t=hDB

(13)

While discharging, following constraints must be met:

SOCt  SOCmin
Fig. 3. Flow duration curve [27,28].

(14)

Further, during discharge of the battery, if SOC (t) < SOC min, a
percentage of energy demand to the load is not met. In this case,

1167

Out put power of BGG in HILR scenario

Out put power of BMG in HILR scenario

30
20

(a)
10

11-12pm

10-11pm

8-9pm

9-10pm

7-8pm

6-7pm

5-6pm

4-5pm

3-4pm

2-3pm

12pm

121pm

1112am

910am

1011am

78am

89am

56am

67am

45am

34am

12am

23am

0
0-1am

Output power (kW)

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Hour (h)
Out put power of BGG in MIMR scenario

Output power (kW)

50

Out put power of BMG in MIMR scenario

40

(b)

30
20
10

10-11pm

7-8pm
7-8pm

11-12pm

6-7pm
6-7pm

9-10pm

5-6pm
5-6pm

8-9pm

3-4pm

4-5pm
4-5pm

2-3pm
2-3pm

3-4pm

12pm
12pm

121pm

1112am

1011am

89am

910am

78am

67am

56am

34am

45am

23am

0-1am

12am

Hour (h)

Out put power of BGG in LIHR scenario

Out put power of BMG in LIHR scenario

Output power (kW)

60
50
40

(c)

30
20
10
11-12pm

10-11pm

8-9pm

9-10pm

121pm

1112am

1011am

89am

910am

78am

67am

56am

45am

34am

23am

0-1am

12am

Hour(h)
Fig. 4. (aec). Output power of BMG and BGG in different investment scenarios for zone 1.

SOC (t) is set to SOC min and the loss of power supply is considered
as unmet load.
3. Problem formulation
Under the present study, problem formulation involves objective function, reliability and economic constraints, which are
described under this section.
3.1. Objective function
Net present cost (NPC) has been found as a useful indicator as it
includes all cost that occurs within the project lifetime. The total
net present cost (TNPC) comprises initial capital cost, replacement
costs, maintenance costs, fuel costs and investment cost of appliances. Therefore minimization of TNPC is considered as an economic indicator used for the tness function for the proposed
system. It can be expressed as [32e34];

TNPC CNPV OMNPV RNPV FCNPV ACNPV

(15)

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

cost; R is replacement cost, FC is the fuel cost and appliances cost.
The subscripts of npv represent the net present value of system
components.
3.1.1. Net present capital costs
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 the system components is
determined using following cost function Eq. (15) [10,38] as;

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

aMHP $PMHP aBGG $PBGG aBMG $PBMG g
(16)
3.1.2. Net present replacement costs
As the life of battery (Lb) and inverter is less than the life of SPV,
WTG, BGG, MHP and BMG. It is therefore, additional investment is
needed to replace these systems accordingly. Under present study,

1168

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

the lifetime of the project is considered for 20 years while the life of
the battery and inverter system is considered as 4 and 10 years
respectively. Therefore, 3 times replacement for battery and 1 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 batteries and inverters are
calculated based on the following Eq. (17) [10,38] as:

Xb

N
1
Lb

(17)

The total net replacement cost of the system components has

been determined using following cost function expressed Eq. (18)
[10,38] as:

RNPV NBATT $ aBATT



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

(18)

3.1.3. Net present operation and maintenance costs

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 O&M cost of all
the system components is determined using cost function expresses has Eq.(19) [10,38] as:

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

bMHG $PMHG bBGG $PBGG bBMG $PBMG

n 
X
1m j

1i
j1

3.1.4. Net present fuel costs

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 using
following cost function Eq.(20) [10,38] as:


n 
X
1m j
j1

1i

flAFR CDBGG ton=yr FFBMG ton=yrg

(22)

CRF

g1 gt
1 gt  1

(23)

where Egen is the total annual energy generated by the IRE system,
TNPC is total net present cost and CRF is capacity recovery factor.

3.2. Reliability criteria

Energy index ratio (EIR) is a probability index value considered
for reliability assessment of the IRE system. Expected energy not
supplied (EENS) is the amount of energy that is not supplied at the
condition when demand surpasses generation [35] and it is
expressed by Eq. (24) as;

EENS

8760
X

L:D

(24)

i1

where, EIR is dened as one minus the ratio of EENS to total energy
demand of the system (EO). It can be expressed as;

EENS
EO

(25)

In case of genetic algorithm, technical and cost data are

considered as an input data for optimization process. Cost parameter includes net present value of capital cost, maintenance cost,
replacement cost and fuel cost of the system components. Appliances cost also taken in to account in the optimization process to
estimate the TNPC of the IRE model. The technical parameters such
as 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. A summary of technical and
cost parameters of the system components considered under present the study is given in Table 6 [17,36-38].

4. Methodology

(20)

3.1.5. Net present value of appliances costs

Total appliances cost can be estimated based on price of appliances of three scenarios considered under present study. The
variation of the cost of appliances as shown in Fig. 2. The net present value of appliances cost has been calculated by using cost
function expressed by Eq. (21) as:

COA POA*TNOA

TNPC  CRFg; t
P8760
t1 Egen t

The capital recovery factor (CRF) converts the present value of

system components into annual payment over the project life time.
The CRF depends on annual interest rate ( ) and life of the plant (t)
and can be computed by using Eq. (23) as;

EIR 1 

(19)

FCNPV

COE Rs=kWh

(21)

Further, cost of energy of the IRE system has been estimated

based on annual net present cost (ANPC) over the project life time
to the total annual energy generated by the IRE system. The COE of
the proposed system can be estimated by using Eq. (22) as given
below [32-34];

Energy management system is essential in meeting the load

demand utilizing efcient system components making up IRES.
Various investments scenarios such as low, medium and high investment based appliances without DSM and peak load shifting
based DSM strategy have been considered for the considered IRE
System. The following section of the paper describes the IRES
conguration and steps involved in DSM strategy.

4.1. IRE conguration

The IRE conguration includes AC-generators such as micro
hydro power, biogas generator and biomass generator while DC
generators are SPV panels and wind turbine generator. Battery
bank, charge controller and bidirectional converter are other
important components to constituting IRE system. The schematic
diagram of a developed IRE model is shown in Fig. 5

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

1169

Table 6
Technical and cost parameters of system components.
Parameters

Units

MHP

BGG

BMG

SPV

WTG

BATT

BC

Capital cost
O&M cost
Power rating
Area of SPV panel
Panel efciency
Rated speed
Swept area
Cut-in speed
Cut-out speed
DOD
Voltage
Efciency
Nominal capacity

$/kW
$/yr
kW
m2
%
m/s
m2
m/s
m/s
Volt
%
ah

1201.38
30.03
1
-

750.86
18.77
1
-

901.04
27.03
1
-

590.03
11.80
0.230 Wp
1.63
0.144
-

3075.54
92.27
1
10
4.26
1.5
25
-

187.72
7.51
50
24
-

170.60
3.41
1.5
90
150

AC Bus

DC Bus

DC Renewable
generators

Load

AC Renewable
generators

Charge controoler
Bidirectional
converter
Battery bank

Fig. 5. Schematic diagram of a developed IRE model [38].

(a) Low investment scenario

4.2. DSM strategy

Demand side management encourages consumers to utilize
lesser power during peak hours, or shift the time of energy usage to
off-peak hours. While level of load may not necessarily reduce total
energy consumed, it may reduce investment needs in networks
and/or power plants for fullling peak demands. Peak load shifting
based demand side management (DSM) strategy has been recognized as an eminent technique to reduce the peak load level by
smooth out the daily peaks and ll valleys in the load curve. In this
view, three investment scenario based appliances are proposed for
the present study. The details of technology key with three investments based appliances for residential, community and commercial sectors is given in Table 2. The various investments
scenarios considered and steps involved in DSM strategy are discussed below:

4.2.1. Capital investment scenario

Investments cost of the appliances are categorized based on
high, medium and low power rating of the appliances.

The low cost with high power rating based appliances are
considered in this scenario.
(b) Medium investment scenario
The moderate costs with power rating based appliances are
considered in this case.
(c) High investment scenario
In this scenario, high expensive with low power rating based
appliances are considered.

4.2.2. Load type, denition and identication

Two types of loads are considered during the demand assessment of the study area viz. shiftable and non-shiftable loads. Nonshoftable load includes priority and permanent loads. It involves
domestic lights, TV, Fan, radio, street lights and refrigerators.

Table 7
Load type and load identication.
Sl. no.

Load type

Load identication under present study

1
2

Non-shiftable
Shiftable

Residential lights, TV, Fan, Radio, Street lights, Refrigerators.

Water pumps, our mills and sawmills

1170

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Table 8
Peak shifting strategy based DSM strategy.
Time segment

Water pump

0:0e1:0
1:0e2:0
2:0e3:0
3:0e4:0
4:0e5:0
5:0e6:0
6:0e7:0
7:0e8:0
8:0e9:0
9:0e10:0
10:0e11:0
11:0e12:0
12:0e13:0
13:0e14:0
14:0e15:0
15:0e16:0
16:0e17:0
17:0e18:0
18:0e19:0
19:0e20:0
20:0e21:0
21:0e22:0
22:0e23:0
23:0e0:0
Duration of time length

Flourmills

Saw mills

SONPT

ROFFPT

SONPT

ROFFPT

SONPT

ROFFPT

OFF
OFF
OFF
OFF
OFF
OFF
ON
ON
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
ON
ON
ON
OFF
OFF
OFF
5 h/day

OFF
OFF
OFF
OFF
OFF
OFF
ON
ON
OFF
OFF
OFF
OFF
OFF
OFF
OFF
ON
ON
ON
OFF
OFF
OFF
OFF
OFF
OFF
3 h/day

OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
ON
ON
ON
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
3 h/day

OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
ON
ON
ON
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
3 h/day

OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
ON
ON
ON
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
3 h/day

OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
ON
ON
ON
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
OFF
3 h/day

(a)

HILR without dsm

HILR with dsm

Load demand (kW)

50
40
30
20
10
0
1

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour (h)

Load demand (kW)

(b)

MIMR without DSM

MIMR with dsm

80
70
60
50
40
30
20
10
0
1

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour (h)

Load demand (kW)

(c)

LIHR without DSM

LIHR with DSM

100
80
60
40
20
0
1

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour (h)
Fig. 6. (aec). Hourly load prole with and without DSM strategy for zone 1; (a) HILR (b) MIMR (c) LIHR.

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Operation of these appliances is strictly dominated by user comfort

and convenience. Shiftable loads are water pump, ourmill and saw
mills, whose operation can be scheduled during certain predened
periods. The end-users can tolerate the shifting operations within
predened time period. Load type and load identications are given
in Table 7.
(a) Range of time interval of shiftable loads
The range of time intervals of shiftable loads has been specied
by the end-user. Shifting of load with regard to usage of pumping
water, our mills and saw mills can be done by switching the usage
from peak load hours to non peak load hours. It is clear that peak
hours of power use are 18:00e21:00 h can be shifted to non peak
hours e 15:00e18:00 h through DSM strategy. The same procedure
can be adopted for Flour mills and Saw mills as given in Table 8.

1171

(b) Investment scenario with allocation of shiftable loads

The allocation of time for operation of water pump, four mill or
saw mill can be shifted from peak hours to non peak hours in a day.
Peak hours constitute those times in the day when electrical appliances are used most whereas non peak hours include time slots
when electrical appliances are not in use by most users. Based on
the hourly load distribution and considering different investment
scenarios - LIHR, MIMR and HILR with and without DSM strategy as
shown in Fig. 6(aec) - peak load can be shifted from peak hours to
non peak hours on the basis of preference criteria.
The total load demand Eload under without DSM strategy is the
sum of non-shiftable loads (ENSL) and shiftable load (ESL) multiplied
by shiftable load operating at on peak time while total load demand
Eload under with DSM strategy is the sum of non-shiftable loads
(ENSL) and shiftable load (ESL) multiplied by shiftable load operating

Resources assessment
DC Renewable energy
potentials
(SPV panels and
WTG)

AC Renewable energy
potentials
(MHG,BMG and BGG)

Calculate power output from AC and DC

renewable energy sources
Initialization
t=0 & d=0

d=d+1
t= t+1

Flow chart of Without (Fig. 8)

and
With DSM (Fig.9) strategy

A
Yes
t=24
Yes

Simulation
Failed

d=d+1

d=365
No
Yes
Simulation successful

Finish

Fig. 7. Methodology for optimizing IRE system using with and without DSM strategy.

1172

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

at off peak time. Load demand without DSM strategy can be

expressed by Eq. (26) [18] as;

ELoadWITHOUT:DSM t ENSL t

n
X

ESL t  SONPT t

(26)

i1

Load demand with DSM strategy can be expressed by Eq. [18] as;

ELoadWITH:DSM t ENSL t

n
X

ESL t  SOFFPT t

(27)

i1

Finally, hourly annual load demand of the study area with and
without DSM strategy has been simulated using HOMER software
based on low, medium and high investment scenarios, the ow

charts of shifting methodology with and with DSM strategy are

shown in Figs. 7e9.
5. Solution technique for the developed formulation
In this paper, operational strategy of optimization of the proposed IRE system has been performed using GA based approach
[31]. Methodologies adopted for optimization for with and without
DSM strategy are described through ow charts as shown in
Figs. 7e9. The optimization solution methodology involves
numerous data inputs such as annual electrical energy demand,
locally available resources, economical and technical data of system
components, battery specication and other decision variables
imposed by the proposed model.

Without DSM
strategy

Consider the without DSM based

load modeling equation
As given in equation (19)
Simulate Hourly (24h) load
demand to year load (8760h )
using HOMER
( without shifting peak load to off
peak time)

Calculate total load demand PL(t)

No
Dump Load

Yes
SOC < SOC max

P (t) > P (t)

No
P

(t) > P (t)

Battery charging
Calculaion of new SOC

No
Yes

Yes
P

(t) < P (t)

(t) +P

< P (t)

No

Yes

SOC< SOC min

No
Battery discharge
Calculation of new SOC

Calculate total net present cost(TNPC) of the system (fitness function) f

For all the LIS,MIS and HIS Using GA based approach

To

Fig. 8. Flowchart for without DSM strategy.

Set SOC =SOCmin

Set needed SOC as unmet load
Set fitness function value f=1e9

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

With DSM
strategy

Consider the with DSM based

load modeling equation
As given in equation (20)
Shifting peak load demand to
off peak time by ON and OFF of
shiftable load appliances
(Table 5)

Calculate total load demand PL(t)

No
Dump Load

Yes
SOC < SOC max

1173

Simulate Hourly (24h) load

demand to year load (8760h )
using HOMER

P (t) > P (t)

No
P

(t) > P (t)

Battery charging
Calculaion of new SOC

No
Yes

Yes
P

(t) < P (t)

(t) +P

< P (t)

No

Yes

SOC< SOC min

No
Battery discharge
Calculation of new SOC

Set SOC =SOCmin

Set needed SOC as unmet load
Set fitness function value f=1e9

Calculate total net present cost(TNPC) of the system (fitness function) f

For all the LIS,MIS and HIS Using GA based approach

To

Fig. 9. Flowchart for with DSM strategy.

a) In without DSM strategy (Fig. 8), hourly electrical load has been
computed based on load modeling equation as given in Eq. (26).
b) In with DSM strategy (Fig. 9), hourly electrical load have been
estimated based on modeling equation as given in Eq. (27).

parameters of the system components as listed in Table 2 are

considered while designing the IRE model.

6. Results and discussion

The GA based optimization process includes three decision
variables such as active area of solar panel, swept area of wind
turbine and SOC of the battery system (ASPV, AWTG and NBS). In order
to handle the three decision variables, GA probes the search space
and best tness function considered based on the objective function. The parameters of the GA algorithm have been adjusted as;
population size (n 25), crossover rate (crate 0.5) and mutation
rate (crate 0.5). MATLAB (R20013 b) and are used for coding and
simulate the proposed methodology. The technical and cost

Optimization results in terms of net present cost and cost of

energy with and without DSM strategies have been obtained using
GA based coding through MATLAB R2013b environment. Table 6
represents the input data considered for GA simulation, which includes technical and economical data of system components of
IRES. Hourly load demands for all zones have been estimated based
on of low, medium and high investment based appliances. Estimated values of hourly load demand of LIHR, MIMR and HILR for

1174

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Table 9
Zone wise optimization results considering investment scenarios without DSM strategy.
Zones

Zone 1

Zone 2

Zone 3

Zone 4

Scenarios

HILR
MIMR
LIHR
HILR
MIMR
LIHR
HILR
MIMR
LIHR
HILR
MIMR
LIHR

System components (kW)

System components (nos.)

MHG

BGG

BMG

SPV

WTG

BATT

6
6
6
0
0
0
15
15
15
16
16
16

11
15
27
27
38
46
26
44
60
34
45
62

18
28
30
36
50
58
45
63
75
41
68
77

43
64
75
32
49
40
65
53
71
61
76
73

20
24
31
13
6
20
15
24
33
25
28
46

109
146
190
190
238
283
244
308
380
273
340
397

TNPC ($)

COE ($/kWh)

265355
332633
408620
370176
414777
537318
531762
649497
758672
587370
691095
790509

0.105
0.108
0.108
0.124
0.125
0.130
0.118
0.114
0.114
0.129
0.113
0.119

TNPC ($)

COE ($/kWh)

238624
298243
368974
343145
388347
528308
504730
622316
725634
565550
632678
768584

0.103
0.104
0.105
0.122
0.123
0.127
0.116
0.112
0.112
0.124
0.107
0.105

TNPC ($)

COE ($/kWh)

265355
370176
649497
790509

0.105
0.124
0.114
0.119

Table 10
Zone wise optimization results considering investment scenarios with DSM strategy.
Zones

Zone 1

Zone 2

Zone 3

Zone 4

Scenarios

HILR
MIMR
LIHR
HILR
MIMR
LIHR
HILR
MIMR
LIHR
HILR
MIMR
LIHR

System components (kW)

System components (nos.)

MHG

BGG

BMG

SPV

WTG

BATT

6
6
6
0
0
0
15
15
15
16
16
16

11
15
27
27
38
46
26
44
60
34
45
62

18
28
30
36
50
58
45
63
75
41
68
77

27
46
69
30
40
36
42
42
68
47
52
45

18
22
27
9
4
18
12
21
25
19
27
31

101
139
183
185
215
256
228
306
378
243
315
348

Table 11
Possible conguration of IRE model without DSM strategy.
Zones

Zone
Zone
Zone
Zone

1
2
3
4

Scenarios

HILR
HILR
MIMR
LIHR

System components (kW)

System components (nos.)

MHG

BGG

BMG

SPV

WTG

BATT

6
0
15
16

11
27
44
62

18
36
63
77

43
32
53
73

20
13
24
46

109
190
308
397

zone 1 without and with DSM are shown in Fig. 6(aec). The hourly
load demand is further synthesized to hourly annual data (8760 h)
using HOMER software. Total annual load demand for all the three
scenarios of all the zones are shown in Fig. 1. Cost parameter
comprises capital cost, maintenance cost, replacement cost, fuel
cost of the system components and appliances cost of various investment scenarios used in GA based approach. Also, project lifetime and annual real interest rate are 20 years and 0.11 respectively
considered in the present study.
Optimization results in terms of net present cost and cost of

energy under with and without DSM strategies and different values
of parameters have been considered under the present study. Three
scenarios based results of IRE system with and without DSM
strategy has been obtained for all the four zones considering three
investment scenarios of low, medium and high cost of appliances
are presented in Tables 9 and 10. Out of these three, the best
possible congurations of IRE system with and without DSM
strategy are presented in Tables 11 and 12. Further, results of
possible conguration without DSM were compared with results
with DSM strategy. It has been observed that results with DSM

Table 12
Possible conguration of IRE model with DSM strategy.
Zones

Zone
Zone
Zone
Zone

1
2
3
4

Scenarios

HILR
HILR
MIMR
LIHR

System components (kW)

System components (nos.)

MHG

BGG

BMG

SPV

WTG

BATT

6
0
15
16

11
27
44
62

18
36
63
77

27
30
42
45

18
9
21
31

101
185
306
348

TNPC ($)

COE ($/kWh)

238624
343145
622316
768584

0.103
0.122
0.112

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Table 13
Comparison of tech-economic results based on selected scenarios with and without
DSM strategy.
Zones

Parameters

Zone 1

ScenarioeI: HILR
SPV
Nos.
WTG
Nos.
BATT
Nos.
TNPC
$
COE
$/kWh
ScenarioeI: HILR
SPV
Nos.
WTG
Nos.
BATT
Nos.
TNPC
$
COE
$/kWh
ScenarioeII: MIMR
SPV
Nos.
WTG
Nos.
BATT
Nos.
TNPC
$
COE
$/kWh
ScenarioeIII: LIHR
SPV
Nos.
WTG
Nos.
BATT
Nos.
TNPC
$
COE
$/kWh

Zone 2

Zone 3

Zone 4

Units

1175

achieved. Similar line, saving for zone 2, zone 3 and zone 4 are given
in Table 13.

Without DSM

With DSM

Saving

6.2. Simulation results with DSM strategy

43
20
109
265355
0.105

27
18
101
238624
0.103

16
2
8
26731
0.002

32
13
190
370176
0.124

30
9
185
343145
0.122

2
4
5
27031
0.002

53
24
308
649497
0.114

42
21
306
622316
0.112

11
3
2
27181
0.002

73
46
397
790509
0.119

45
31
348
768584
0.105

28
15
49
21925
0.014

Based on the comparison of the results of possible congurations with and without DSM strategy, results of possible conguration with peak load shifting based DSM strategy are found to be
better results with ensuring EIR of 1. It involves optimal size of 6 kW
of MHG, 18 kW of BMG, 11 kW of BGG, 27 solar modules with a area
of 1.63m2, 18 wind turbine with a swept area of 4.26 m2 and 101
batteries with a rating of 3.6 kVA. The contribution of electricity
generation from AC and DC renewable generators are found to be as
171550 kWh/yr and 111735 kWh/yr respectively. Battery contributes total discharge energy of 25205 kWh/yr during a decit in the
energy generation from AC and DC generators. Total surplus power
generates from the generators was found to be 10694 kWh/yr.
Further, the minimum NPC and COE for HILR with DSM strategy are
found to be as 238624 $ and $ 0.103 per kWh respectively. Similarly,
HILR based scenario is found to be cost effective for zone 2, while
MIMR based scenario and LIHR based scenario and LIHR based
scenario are found cost effective for zone 3 and zone 4 respectively.
Summary of results for zone 2, zone 3 and zone 4 obtained and are
presented in Tables 12e14.
6.3. Hourly simulated results with and without DSM strategy

strategy are found to be most cost effective compared to results

without DSM strategy. Detail discussions of comparison of simulated results of possible conguration of IRE model without and
with DSM strategy, scenario with peak load shifting based DSM and
nally hourly simulated results are described below:
6.1. Comparative results on the basis of economic and technical
parameters
Optimization results with and without DSM strategies have
been comparatively evaluated on the basis of economic parameters.
Results with DSM strategy were found to be better as compared to
results of without DSM strategy. Savings of total NPC stood at 26731
$ and corresponding COE as $0.02 per kWh was achieved through
DSM strategy for zone 1. Similarly, cost savings for zone 2, zone 3
and zone 4 were obtained. Results are presented in Table 13.
Further, the simulation results obtained on the basis of technical
parameters are also compared. It was found that savings of 8 units
of 3.6 kVA rated battery bank storage, 2 wind turbines with a swept
area of 4.26 m2, 16 solar modules with a active area of 1.63 m2 is

Fig. 10(a,b) shows load demand variation without shifting peak

load and power generation for a typical day on 13 Feb of zone 1. It
includes AC-power, INV-DC power, discharge power and excess
power and HILR based load. It has been observed that, 1st to 8th
hours, load demand becomes more than the power generated and
load demand during this period is satised by AC generators, DC
generators and discharge power from battery system. From 9th to
13th hours, AC and DC renewable generators are capable to supply
the load demand and excess power generated is used to charge the
batteries. From 15th to 18th hours (off peak time), the generation is
more than the load. This excess generation used by the deferrable
load. The percentage of total excess power was found to be 12.5%
during this hour. From 19th to 21st hours, load demand is unable to
satisfy by the output from AC and DC generators to meet the load
demand; hence, excess demand is fullled by the batteries. The
operation of the battery show discharging at 1st to 8th hours and
SOC at 9th hour becomes minimum and then starts charging up to
13th hours. Surplus energy in the 16th and 18th hours is utilized by
deferrable loads. Further, during 19th to 21st hours, the batteries

Table 14
Possible conguration of IRE model with and without DSM strategy.
Zones

Scenario

Parameters

Units

Without DSM

With DSM

Zone 1

ScenarioeI: HILR

Zone 2

ScenarioeI: HILR

Zone 3

ScenarioeII: MIMR

Zone 4

ScenarioeIII: LIHR

AC power
INV-DC power
Discharge
Excess power
AC power
INV-DC power
Discharge
Excess power
AC power
INV-DC power
Discharge
Excess power
AC power
INV-DC power
Discharge
Excess power

kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr
kWh/yr

171550
130153
28805
32529
256230
52244
69341
32989.04
567940
95334
104563
27829.71
703720
155920
135147
37295

171550
111735
25205
10694
256230
41431
67394
15304
567940
81368
93278
13107
703720
124785
111502
26505

1176

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Power (kW)

(a): Hourly electricity producon of HILR without DSM for zone 1

60
50
40
30
20
10
0
1

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour (h)

Load

AC-Power

INV-DC Power

Discharge

Excess power

Power (kW)

(b): SOC of HILR without DSM for zone 1

120
100
80
60
40
20
0
1

Discharge

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour(h)
SOC
SOC-MIN

Excess power

SOC-MAX

Fig. 10. (a) Hourly electricity production versus load and (b) SOC of a battery system for a day of 13 of Feb for a zone 1 (without DSM).

begin to discharge and then charge at 22nd hour and again

discharge during 23 and 24th hours.
Fig. 11(a,b) depicts a peak shifting load and power generation
diagram for same day on 13 Feb of zone 1 which includes, AC power
production, INV- DC power, battery discharge and excess power
generation. In this case, 1st to 8th hours, load demand is same as in
the case of without DSM and load demand is met by AC generators,
DC generators and discharge power from the battery system. From
15th to 18th hours, generation is capable to supply the power to
load and excess power is used as a deferrable load. But there is a

reduction in power generation and excess power during this hour.

The percentage of total excess power is found to be 5% after the
peak load shifting. The operating characteristics of the battery is
almost same as in the case of without load shifting strategy, however there is a more reduction in state of charge of the battery
system. Based on the comparison of with and without DSM strategy, it has been found that, total excess power in case of without
DSM strategy is more than the total excess power generation with
DSM strategy. Also SOC of the battery without DSM is found to be
more than that of SOC of battery with DSM strategy. Similarly,

Power (kW)

(a): Hourly Electricity Producon of HILR with DSM for zone 1

50
40
30
20
10
0
1

Load

AC-Power

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

INV-DC

Hour (h)
Discharge
Excess power

(b):SOC of HILR with DSM for zone 1

SOC

120
100
80
60
40
20
0
1

Discharge

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour(h)
Excess power
SOC
SOC-MIN
SOC-MAX

Fig. 11. (a) Hourly electricity production versus load and (b) SOC of a battery system for a day of 13 Feb (with DSM) of zone 1.

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

1177

Table 15
Hourly simulation results of HILR scenario with and without DSM strategy for a day on 13 Feb of zone 1.
Hours

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

Without DSM strategy

With DSM strategy

Load

AC power

INV-DC power

Discharge

SOC

Excess power

Load

AC power

INV-DC power

Discharge

SOC

Excess power

14.43
10.18
14.1
13.75
18.61
16.35
18.23
28.28
9.14
9.66
11.66
24.09
38.16
33.79
6.33
9.44
12
13.64
56.82
47.88
36.66
19.82
10.29
11.27

6
6
6
6
6
6
6
24
24
24
35
35
35
17
17
17
24
24
35
35
35
35
6
6

4.88
0.23
0.63
0.98
0.24
0.03
0
0.4
1.77
2.56
5.82
8.74
21.19
17.48
15.97
25.39
20.83
10.92
5.94
1.87
0.12
0
0
0

3.81
3.96
7.5
6.82
12.38
10.32
12.23
3.9
0
0
0
0
0
0.18
0
0
0
0
16.19
11.1
1.54
0
4.29
5.27

95
95
90
80
70
50
40
25
20
25
30
35
65
85
80
95
99
99
99
95
95
99
99
95

0
0
0
0
0
0
0
0
0
0
0
0
0
0
17.46
29.31
29
18.7
0
0
0
0
0
0

14.43
10.17
14.08
13.73
18.58
16.33
18.2
28.25
22.26
23.54
27.12
14.42
22.85
20.23
12.02
17.93
22.79
13.62
42.95
36.2
27.71
19.8
10.28
11.26

6
6
6
6
6
6
6
24
24
24
35
35
35
17
17
17
24
24
35
35
35
35
6
6

4.39
0.21
0.56
0.88
0.22
0.03
0
0.36
1.11
1.73
3.82
6.12
16.88
13.35
12.02
20.82
17.05
9.12
5.09
1.69
0.11
0
0
0

4.26
3.97
7.55
6.89
12.38
10.3
12.2
3.9
0
0
0
0
0
0
0
0
0
0
3.11
0
0
0
4.28
5.26

95
95
90
85
65
55
40
25
20
35
55
85
95
99
99
99
99
99
99
90
85
85
95
90

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
10.38
16.38
17.08
0
0
2.25
13
0
0

hourly simulation results of different day for zone 2 (HILR), zone 3

(MIMR) and zone 4 (LIHR) with and without DSM strategies have
also been simulated and presented in Tables 15e18.

TNPC of GA & PSO found to be almost same and difference is

negligible. A details result of TNPC of all three scenarios for both GA
& PSO is given in Table 19 and convergence curves of GA and PSO for
all three are shown in Figs. 12 and 13.

6.4. Comparative analysis of GA and PSO

7. Conclusions
GA results of TNPC of three scenarios (HILR, MIMR & LIHR) of
possible IRE conguration with DSM have been compared with PSO
results of TNPC of three scenarios (HILR, MIMR & LIHR) of IRE
conguration with DSM strategy. It has been observed that, the

The objective of the present study was to develop optimal sizing

of IRE models to minimize TNPC and COE for the study area
comprising MHG, SPV, WTG, BMG, and BGG with battery for

Table 16
Hourly simulation results of HILR scenario with and without DSM strategy for a day on 16 Jan of zone 2.
Without DSM strategy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

With DSM strategy

Load

AC-power

INV-DC

Discharge

SOC

Excess power

Load

AC-power

INV-DC

Discharge

SOC

Excess power

16.67
17.45
11.62
9.84
24.86
17.57
8.88
22.27
7.51
10.15
10.48
25.37
36.52
12.08
9.92
6.86
6.26
20.03
40.75
46.24
40.23
20.5
13.79
11.53

0
0
0
0
0
0
0
36
36
36
63
63
63
27
27
27
36
36
63
63
63
63
0
0

16.51
17.02
11.37
9.84
24.86
17.55
8.44
1.2
3.31
7.44
9.94
10.31
16.99
15.14
5.73
2.79
3.22
3.24
1.13
1.51
1.15
0.19
13.75
11.53

16.51
17.02
11.37
9.84
24.86
17.55
8.44
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
13.75
11.53

65
60
55
50
45
35
25
20
25
35
45
55
65
70
75
85
99
99
99
99
99
99
99
98

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.4
27.38
15.53
19.69
15.69
20.5
36.51
0
0

16.67
17.45
11.63
9.84
24.86
17.57
8.88
22.28
18.41
24.89
24.77
15.13
21.78
7.21
16.93
11.72
10.69
20.03
32.7
37.1
32.28
20.5
13.79
11.53

0
0
0
0
0
0
0
36
36
36
63
63
63
27
27
27
36
36
63
63
63
63
0
0

0.11
0.32
0.18
0
0
0.01
0.32
0.83
1.9
4.28
6.09
6.42
10.34
8.96
3.52
1.69
1.89
1.87
0.66
1.05
0.79
0.13
0.03
0

16.56
17.15
11.45
9.84
24.86
17.56
8.58
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
13.76
11.53

60
55
50
45
40
30
25
20
25
30
40
50
60
65
70
80
90
99
99
99
99
99
99
95

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2.67
15.34
26.51
23.09
26.98
36.46
0
0

1178

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Table 17
Hourly simulation results of MIMR scenario with and without DSM strategy for a day on 10 March of zone 3.
Without DSM strategy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

With DSM strategy

Load

AC-power

INV-DC power

Discharge

SOC

Excess power

Load

AC-power

INV-DC power

Discharge

SOC

Excess power

36.18
47.9
36.36
33.55
49.89
58.93
31.13
40.83
53.49
44.87
58.57
104.13
79.77
68.42
42.05
28.63
37.36
58.74
97.01
103.85
96.81
82.06
25.05
40.11

15
15
15
15
15
15
15
78
78
78
122
122
122
59
59
59
78
78
122
122
122
122
15
15

0
1.19
0.39
5.2
3.45
2.32
1.28
3.82
1.94
5.96
5.14
7.93
8.03
10.69
10.22
9.89
8.2
3.91
1.86
0.31
1.35
7.45
6.99
7.05

21.18
31.77
20.99
13.61
31.61
41.73
14.91
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3.41
18.41

85
80
70
60
55
40
25
20
30
40
50
70
75
80
85
90
95
99
99
99
99
99
99
99

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.97
19.99
23.03
15.8
22.75
40.85
0
0

36.18
47.9
36.36
33.56
49.89
58.94
31.13
40.83
28.75
24.11
32.03
130.47
99.95
85.73
28.27
19.25
25.12
58.74
107.1
114.65
106.87
82.06
25.05
40.11

15
15
15
15
15
15
15
78
78
78
122
122
122
59
59
59
78
78
122
122
122
122
15
15

0
1.36
0.45
5.95
3.94
2.65
1.47
4.36
2.41
7.31
6.45
9.87
9.87
13.38
12.81
12.46
10.35
4.94
2.31
0.36
1.54
8.51
7.99
7.05

21.18
31.61
20.94
12.91
31.15
41.42
14.73
0
0
0
0
0
0
14.02
0
0
0
0
0
0
0
0
2.46
18.41

70
65
60
55
50
40
25
20
30
45
60
65
65
75
70
95
99
99
99
99
99
99
99
99

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
41.38
20.91
14.81
6.6
14.32
41.81
0
0

Table 18
Hourly simulation results of LIHR scenario with and without DSM strategy for a day on 28 Jan of zone 4.
Without DSM strategy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

With DSM strategy

Load

AC-power

INV-DC

Discharge

SOC

Excess energy

Load

AC-power

INV-DC

Discharge

SOC

Excess power

39.73
43.78
44.64
38.09
56.58
54.68
52.49
54.42
24.48
28.55
17.92
152.56
134.6
103.85
21.35
22.22
26.22
116.34
142.6
144.94
242.51
130.98
41.12
33.81

16
16
16
16
16
16
16
93
93
93
155
155
155
78
78
78
93
93
155
155
155
155
16
16

0
0.22
0
0
0.13
0
0.26
0.71
3.96
5.78
13.1
15.54
10.72
18.89
18.96
15.97
9.52
6.27
11.58
15.26
8.94
0.88
0.48
0.59

23.73
27.58
28.64
22.09
40.46
38.68
36.24
0
0
0
0
0
0
7.9
0
0
0
17.39
0
0
79.01
0
24.67
17.25

80
75
65
55
50
40
30
20
30
40
50
55
60
85
80
90
95
95
95
100
100
80
90
85

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
17.31
65.67
0
0.04
22.33
0
0
0
0

39.71
43.77
44.62
38.07
56.56
54.66
52.46
54.4
41.73
48.67
30.04
132.48
116.88
90.18
31.33
32.61
38.48
116.29
133.01
135.19
226.2
130.93
41.11
33.8

16
16
16
16
16
16
16
93
93
93
155
155
155
78
78
78
93
93
155
155
155
155
16
16

0
0.32
0
0
0.2
0
0.39
1.05
4.18
4.15
10.69
10.76
7.88
15.92
16.69
12.43
7.81
6.25
14.97
22.64
13.27
1.31
0.71
0.88

23.71
27.46
28.62
22.07
40.37
38.66
36.09
0
0
0
0
0
0
0
0
0
0
17.35
0
0
58.6
0
24.44
16.96

85
80
70
60
50
40
30
20
30
40
50
70
85
90
90
95
100
100
95
100
100
90
95
90

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
20.93
53.65
0
9.26
37.31
0
0
0
0

Table 19
Comparative results of GA & PSO considering three investment scenarios with DSM
strategy for zone 1.
Scenarios

Parameters

GA

PSO

HILR

TNPC ($)
Elapsed time (seconds)
No. of Iteration
TNPC($)
Elapsed time (seconds)
No. of Iteration
TNPC($)
Elapsed time (seconds)
No. of Iteration

238624
27.38
96
298243
49.94
176
368974
123
35.874

238624
28.258
114
298093
28.258
113
368974
119
29.796

MIMR

LIHR

supplying power to four zones in Chamarajanagar district of Karnataka state of India. TNPC and COE of the IRE system were determined by considering peak load shifting based DSM strategy.
It is observed that system cost, system size and load density are
different for all considered scenarios (HILR, MIMR and LIHR). Load
density has been estimated by considering different appliances and
is found to be increased from HILR to LIHR scenarios which results
in the increase of system size and system cost.
Cost of energy has been computed as per total net present cost
of the system (including system cost, appliances cost and saving
cost) divided by total annual energy generation.

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Fig.12. Convergence curve of GA and PSO for HILR scenarios for zone 1.

Fig.13. Convergence curve of GA & PSO (combined) for HILR, MIMR and LIHR scenarios for zone 1.

1179

1180

S. Rajanna, R.P. Saini / Renewable Energy 99 (2016) 1161e1180

Optimal sizes and cost of system components pertaining to the

developed IRE models are found to be different for each zone. For
zone 1 and zone 2, HILR was found to be the most suitable scenario
as the optimal size involved least system cost. Similarly, considering least system cost pertaining to optimal system sizes for zone 3
and zone 4, MIMR and LIHR are found most suitable. Further, GA
results of TNPC of all the three scenarios for zone 1 are validated
with PSO Technique. It is found that the TNPC is almost same for all
the three scenarios.
Further, it has also been observed that DSM with high investment with low power rating appliances is most feasible for the
values of load density up to 3.360 MWh. However, medium investment with moderate power rating appliances is feasible for a
load density range from 3.369 MWh to 4.016 MWh. But low investment with high power rating appliances is found to be feasible
for a load density value beyond 4.016 MWh.
Acknowledgements
The rst author express sincere thanks to Alternate Hydro Energy Centre and Quality Improvement program (QIP) Centre, IIT,
Roorkee, India for providing research facilities and All India council
for Technical Education (AICTE), Government of India for providing
nancial assistance in the form of research fellowship. He also express heartedly gratitude to Malnad College of Engineering (Govt
aided), Hassan, Karnataka, India for nancially sponsoring him to
pursue doctoral study.
References
[1] S. Semaoui, A. Hadj Arab, S. Bacha, B. Azoui, The new strategy of energy
management for a photovoltaic system without extra intended for remotehousing, Sol. Energy 94 (2013) 71e85.
[2] S.M. Shaahid, M.A. Elhadidy, Economic analysis of hybrid photovoltaic diesel
battery power systems for residential loads in hot regions e a step to clean
future, Renew. Sustain Energy Rev. 12 (2) (2008) 488e503.
[3] R. Kallel, G. Boukettaya, L. Krichen, Control management strategy of standalone hybrid power micro-system using super-capacitor, Int. J. Renew. Energy Res. 4 (1) (2014) 210e223.
[4] S. Rajanna, R.P. Saini, A review on planning, congurations, modeling and
optimization techniques of hybrid renewable energy systems for off grid
applications, Renew. Sustain Energy Rev. 58 (2016) 376e396.
[5] R. Ramakumar, Energizing rural areas of developing countries using IRES, in:
Proc. IECEC Conf., 1996, pp. 1536e1541.
[6] Antonio Cano, J. Francisco, S. Higinio, M. Fernandez, Optimal sizing of standalone hybrid systems based on PV/WT/FC by using several methodologies,
J. Energy Inst. 87 (4) (2014) 330e340.
[7] Tareq Alnejaili, Said Drid, Advanced strategy of demand-side management for
photovoltaic wind energy system, in: Proc. STA Conf., 2014.
[8] Ali Al-Alawi, S.M. Islam, Demand side management for remote area power
supply systems incorporating solar irradiance model, Renew. Energy 29
(2004) 2027e2036.
[9] S. Abedi, A. Alimardani, G.B. Gharehpetian, G.H. Riahy, S.H. Hosseinian,
A comprehensive method for optimal power management and design of
hybrid RES-based autonomous energy systems, Renew. Sustain Energy Rev. 16
16 (2012) 1577e1587.
[10] Alireza Askarzadesh, S.C. Leandro dos, A novel framework for optimization of
a grid independent hybrid renewable energy system A case study of Iran, Sol.
Energy 112 (2015) 383e396.
[11] B.B. Alagoz, A. Kaygusuz, A. Karabiber, A user-mode distributed energy
management architecture for smart grid applications, Energy 44 (2012)
167e177.
[12] Linqi Song, Yuanzhang Xiao, Mihaela van der Schaar, Demand side management in smart grids using a repeated game framework, IEEE J. Sel. Areas

Commun. 32 (7) (2014) 1412e1424.

[13] E. Matallanas, et al., Neural network controller for active demand-side management with PV energy in the residential sector, Appl. Energy 91 (2012)
90e97.
[14] N. Gudi, L. Wang, V. Devabhaktuni, A demand side management based
simulation platform incorporating heuristic optimization for management of
household appliances, Electr. Power Energy Syst. 43 (2012) 185e193.
[15] G. Kyriakarakos, D.D. Piromalis, A.I. Dounis, K.G. Aventis, G. Papadakis, Intelligent demand side energy management system for autonomous polygene
ration microgrids, Appl. Energy 103 (2013) 39e51.
[16] M. Stathopoulos, D. Zarakis, K. Kavadias, J.K. Kaldellis, The role of residential
load-management in the support of RES based power generation in remote
electricity grids, Energy Procedia 46 (2014) 281e286.
[17] S. Upadhyay, M.P. Sharma, Selection of a suitable energy management strategy for a hybrid energy system in a remote rural area of India, Energy 94
(2016) 352e366.
[18] A. Chauhan, R.P. Saini, Techno-economic optimization based approach for
energy management of a stand-alone integrated renewable energy system for
remote areas of India, Energy 94 (2016) 138e156.
[19] A.B. Patil, R.P. Saini, M.P. Sharma, Development of IREOM model based on
seasonally varying load prole for hilly remote areas of Uttarakhand State in
India, Energy 36 (9) (2011) 5690e5702.
[20] R. Ramakumar, P. Sudhakara Shetty, K. Ashenayi, A linear programming
approach to the design of integrated renewable energy systems for developing countries, IEEE Trans. Energy Convers. 4 (1) (1986) 18e24.
[21] Randa Kallel, Ghada Boukettaya, Lot Krichen, Demand side management of
household appliances in stand-alone hybrid photovoltaic system, Renew.
Energy 81 (2015) 123e135.
[22] S. Rajanna, R.P. Saini, Modeling of integrated renewable energy system for
electrication of a remote area in India, Renew. Energy 90 (2016) 175e187.
[23] Ariel Miara, Craig Tarr, Rachel Spellman, Charles J. Vorosmarty, Jordan
E. Macknick, The power of efciency: optimizing environmental and social
benets through demand-side-management, Energy 76 (2014) 502e512.
[24] R. Ramakumar, M. Abouzahr, K. Ashenay, A knowledge-based approach to the
design of integrated renewable energy systems, IEEE Trans. Energy Convers. 7
(4) (1992) 648e659.
[25] S. Diaf, G. Notton, M. Belhamel, M. Haddadi, A. Louche, Design and technoeconomical optimization for hybrid PV/wind system under various meteorological conditions, Appl. Energy 85 (2008) 968e987.
[26] Hongxing Yang, Lin Lu, A. Wei Zhou, Novel optimization sizing model for
hybrid solar-wind power generation system, Sol. Energy 81 (2007) 76e84.
[27] Novel effects of demand side management data on accuracy of electrical energy consumption modeling and long-term forecasting. Energy Convers.
Manage. 78 (2014) 745e752.
[28] A.M. Michael, T.P. Ojha, Principles of Agricultural Engineering, Jain Brothers,
2013.
[29] K. Subramanian, Engineering Hydrology, McGraw-Hill Companies, 2012.
[30] A. Gupta, R.P. Saini, M.P. Sharma, Steady-state modeling of hybrid energy
system for off grid electrication of cluster of villages, Renew. Energy 35 (1)
(2010) 520e535.
[31] M.S. Ismail, M. Moghavvemi, T.M.I. Mahlia, Genetic algorithm based optimization on modeling and design of hybrid renewable energy systems, Energy
Convers. Manage. 85 (1) (2014) 120e130.
[32] A. Chauhan, R.P. Saini, Renewable energy based off-grid rural electrication in
Uttarakhand state of India: technology options, modeling method, barriers
and recommendations, Renew. Sustain Energy Rev. 51 (2015) 662e681.
[33] Subho Upadhyay, M.P. Sharma, Development of hybrid energy system with
cycle charging strategy using particle swarm optimization for a remote area in
India, Renew. Energy 77 (2015) 586e598.
[34] A. Kaabeche, M. Belhamel, A sizing optimization of grid-independent hybrid
photovoltaic/wind power generation system, Energy 36 (1) (2011)
1214e1222.
[35] A.B. Patil, R.P. Saini, M.P. Sharma, Integrated renewable energy systems for off
grid rural electrication of remote area, Renew. Energy 35 (2010) 1342e1349.
[36] MNRE, Development of Renewable Energy, Role of Proper Incentives Mechanisms. http://mnre.gov.in/le-manager/UserFiles/presentations-pwc-IREDA.
pdf (accessed 25.03.16).
[37] A. Chauhan, R.P. Saini, Renewable energy based off-grid rural electrication in
Uttarakhand state of India: technology options, modeling method, barriers
and recommendations, Renew. Sustain Energy Rev. 51 (2015) 662e681.
[38] S. Rajanna, R.P. Saini, Development of optimal integrated renewable energy
model with battery storage for a remote Indian area, Energy 111 (2016)
803e817.