Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser

Feasibility study and sensitivity analysis of a stand-alone

photovoltaic–diesel–battery hybrid energy system
in the north of Algeria
H. Rezzouk a,b, A. Mellit b,n,1
a
Unité de Développement des Equipements Solaires (UDES)/EPST CDER, Bou-Ismail, 42000 Tipaza, Algeria
b
Faculty of Sciences and Technology, Renewable Energy Laboratory, Jijel University, Ouled-Aissa, Box 0.98, Jijel 18000, Algeria

art ic l e i nf o a b s t r a c t

Article history: This paper aims to study the techno-economical feasibility of a photovoltaic–diesel–battery hybrid
Received 7 May 2014 energy system (HES) destined to electrify a research unit (UDES) located in the north of Algeria. For this
Received in revised form aim several scenarios have been studied for a photovoltaic penetration varying from 0% to 100%
13 November 2014
including a stand-alone diesel system and a stand-alone photovoltaic system. For each scenario, the
Accepted 25 November 2014
power system has been designed and optimized to get a maximum output power at a low cost. The
Available online 18 December 2014
performance of these systems has been analysed based on some determinant criteria such as net present
Keywords: cost, the cost of energy, energy excess, load satisfaction, fuel consumption savings, maintenance and
Photovoltaic operation costs of diesel generators, CO2 and pollutants saving rates. The impact of the storage battery
Diesel generator
bank size on the total cost of the power system has also been studied; it has been found that this
Hybrid energy system
parameter is a decisive factor that determines the optimum share of the solar resource in the hybrid
Techno-economic analysis
Sensitivity analysis system. Results showed that 25% PV–diesel–battery HES is the optimal configuration, which presents the
Carbon emissions best compromise between PV penetration efficiency, energy cost and the system stability. A sensitivity
analysis has been performed on the optimal HES in order to study the effect of some parameters'
variation (global solar radiation, diesel price, real interest rate and load consumption) on its total cost,
the cost of energy, photovoltaic resource penetration and fuel intake.
& 2014 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1135
2. Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1136
3. Hybrid energy system description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1138
3.1. Photovoltaic modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1138
3.2. Battery bank. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1138
3.3. Power converter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1138
3.4. Diesel generators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1138
4. HOMER software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1138
4.1. Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1138
4.2. Economic assessment criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
4.2.1. Net present cost. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
4.2.2. Total annualized cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
4.2.3. Capital recovery factor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
4.2.4. Annual real interest rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
4.2.5. Levelized cost of energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
5. Homer input parameters description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
5.1. Load profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139

n
Corresponding author.
E-mail address: [email protected] (A. Mellit).
1
Abdus Salam International Centre for Theoretical Physics, Strada Costiera, 11-34151 Trieste, Italy.

http://dx.doi.org/10.1016/j.rser.2014.11.103
1364-0321/& 2014 Elsevier Ltd. All rights reserved.
 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150 1135

5.2. Solar radiation data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139

5.3. Diesel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
6. System dispatch strategy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
7. Simulation results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
7.1. Stand-alone diesel system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1141
7.2. Hybrid energy system scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1142
7.2.1. Technical and electrical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1142
7.2.2. Economic analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1143
7.2.3. Environmental impacts of hybridization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1145
8. Sensitivity analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1146
8.1. Global solar radiation and diesel price . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1146
8.2. Load consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1147
8.3. Real interest rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1147
9. Concluding remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1148
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149

1. Introduction Renewable energies (REs) are considered as a good alternative

to fossil fuels that would help in mitigating their hazardous effects
Global energy consumption is experiencing a drastic increase in on the atmosphere and compensate for their depleting nature as
both developed and developing countries (1% and 5% per year, they use clean and regularly regenerating flows of energy (sun,
respectively) [1] because they need permanently more energy for wind, water, etc.) [17]. Currently, there is a growing recognition of
their industrial and technological development and to provide the importance of integrating REs in the energy sector of countries
residential comfort for the growing populations in general [2]. The that aim to achieve the goals and objectives of a sustainable
global yearly consumption has been estimated to be an averaged development.
primary power of 16.9 TW [3] and it is expected that the produc- The potential of REs is enough to cover the entire of the
tion capacity should be doubled in the next 40 yr to cover our universal energy demand [2] and it is well distributed across the
energy needs [4]. globe unlike fossil resources. As indicated in Ref. [18], the global
Modern civilization is heavily dependent on energy which is wind resource potential estimated at about 72TW is much greater
met throughout conventional sources, such as nuclear energy and than the global daily electrical consumption; this capacity allows
fossil fuels (coal, oil and gas) [5]. As reported in Ref. [6], about 80% the production of over five times of the global energy needs and
of total energy is produced by burning fossil fuels; the combustion over 40 times of the global electrical consumption of the year 2011.
of these substances leads to the emission of tons of pollutants and Similarly, solar energy (SE) received at the ground on an annual
greenhouse gases (GHG) (carbons (CO2), sulfur dioxide, nitrogen average is estimated to be about 75,000 GToe, which is equivalent
oxides, and unburnt hydrocarbons) that harm the atmosphere, the to 0.9 billion of TWh of the electricity. This energy is 6000 times
environment and induce adverse effects on the health of the the current annual global energy consumption [19,20].
populations [7]. High concentrations of CO2 in the atmosphere Currently, the generation of electricity from REs is approxi-
have contributed to the global climate warming phenomenon [8] mately estimated at 17% of the total electricity mix [21]. The major
and an increase of 3–6 1C of the earth's temperature is expected by problem responsible for slowing the rate of investments in this
the end of the century [6]. As a response to climate change, the field is the difference between the cost of electricity generated
KYOTO protocol was adopted in December 1997 where 38 of the from REs and that generated from alternative resources.
industrialized countries committed to reduce the emission of six The stochastic nature of REs (diurnal and seasonal cycles) is
greenhouse gases (CO2, CH4, N2O, HFCs, PFCs and SF6) by 5.2% on responsible for a mismatch between the demand and supply of
an average between 2008 and 2012 compared to the 1990 levels electrical power [21]. In this case solar and wind power systems
[9,10]. need to be backed up by other sources of energy, generally a
Nowadays, about 1.4 billion people around the world in remote combination with either a DG or a reversible electric storage
communities and rural areas are still not connected to the grid and (batteries) or with both is used in one system such as a hybrid
are totally deprived of a steady electrification network [11]. Owing energy system (HES) [15,22].
to geographical, economical and technical obstacles, the grid is no The use of batteries as a backup system increases the initial
more considered as an appropriate means of electrification for investments of solar and wind stand-alone systems especially
these sites, of low population rates, due to high costs of invest- when it comes to high capacities [4], in this case the use of a DG
ments needed for setting up substations and extending electrical is beneficial mainly economically by supplying electrical energy at
lines, high costs of maintenance and lines losses [4]. a low cost [23].
Decentralized production systems based on diesel power gen- Generally, the combination of renewable and conventional
erators are the most widespread means for rural electrification sources of energy allows the optimization of power generation
since they exhibit reliability and availability and for their low systems both from technical, economical and environmental view-
investment costs compared to the grid [12,13,14]. However, these points by providing continuous and stable power, reducing opera-
systems have some drawbacks such as high costs of operation due tion and maintenance (O&M) costs of DGs, minimizing the
to high fuel consumption, high costs of maintenance, emission of dependence on fuel and curtailing harmful emissions and pollu-
GHGs and pollutants in the local atmosphere [14]. Furthermore, tants [24–26]; for these reasons HESs are progressively used as
for low load levels, 40–50% of the rated power of the diesel distributed systems to produce electricity locally in remote areas
generator (DG), the oil is not burnt in a fit manner in the diesel and communities.
engine which causes carbon build-up; in this case the DG becomes The sizing of HESs is complicated as their performances are
inefficient which increases its maintenance costs [15,16]. strongly affected by the design and the type of the components,
1136 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150

the architecture and the choice of the operating strategy that PV/wind/diesel system with battery storage for rural electrification
should effectively manage different sources of energy [27]. A HES at the lowest cost; the study has been performed for six sites of
should meet the load requirements with minimum investments Algeria. A discrete chaotic harmony search-based simulated
avoiding over-sized components. For this aim, the load profile annealing efficient algorithm has been developed in Ref. [39] for
should be accurately defined and the availability of the REs the aim of finding the optimum design of a PV/wind HES. Rehman
potential for the implantation site should be precisely estimated et al [40] have studied the feasibility of a PV/wind/diesel system to
as long as the output power of the RE system depends on the power a village, in a remote area, electrified with a set of eight
energy flow intensity [28]. diesel generators; a hybrid PV/wind/diesel system with a con-
Massive studies and researches have been carried out on HESs tribution of RE sources of 35% was found to be optimal with an
in the literature since these systems are very interesting and optimal COE of 0.212 $/kWh.
suitable as far as energetic, economical and ecological aspects This paper aims to study the economic feasibility of a hybrid
are concerned. As reported in the literature, selection of the PV–diesel–battery energy system destined to supply electric
optimal configuration of a HES is based on technical, economical power for a research unit (Unité de Développement des Equipements
and environmental criteria. Some parameters to be considered Solaires, UDES) [41] located in Tipaza, north of Algeria. The
when evaluating the performance of a HES are the cost of kWh of investigation has been focused on the analysis of several scenarios
energy (COE) produced by the system, the net present cost (NPC) of energy systems for a PV penetration varying from 0% (stand-
of the system, the specific consumption of DGs, their number and alone diesel system) to 100% (stand-alone PV system). The Hybrid
duration of failure (O&M) costs and the excess of energy that Optimization Model for Electric Renewables (HOMER) has been
should be minimized, and GHGs emission rates [29,15]. Based on used to determine the optimal design for each system. In fact, for
the (COE) and CO2 emission rate parameters, the study carried out each case study, the technical, economical, energetic and environ-
by Shaahid and El-Amin [30] for a remote residence in a tropical mental aspects will be carefully analysed, then a comparative
climate has shown that the hybrid PV–diesel–battery energy study will be conducted in order to determine the different factors
system scenario was economically effective over the two other that influence the total cost of the system and to define an optimal
scenarios consisting of a stand-alone diesel system and a stand- range, in which the choice of the hybrid system will be justified in
alone PV system. In Ref. [11], a techno-economic feasibility study terms of an effective PV penetration for an optimal cost of
of a hybrid energy PV–diesel power generation system was kilowatt-hour of electricity. A sensitivity analysis will also be
performed for the Khavar-E-Bala, a remote rural village situated performed according to some parameters (global solar radiation,
in the eastern part of IRAN, and a PV–diesel system without energy diesel price, load consumption and real interest rate) to study the
storage was found to be cost effective. The feasibility study of a effect of their variation on the design, the size, the total cost of the
hybrid PV–diesel system [31] has shown that the COE is very HES and COE. The main benefit of the HES is that the drawback of
sensitive to the cost of fuel and the fuel intake. In Ref. [32], a multi- one source is covered by the other source.
objective optimization model has been developed, in C- This paper is structured as follows: the next section provides a
programming language, to define the optimum sizing of a hybrid background of the energy sector status in Algeria. Then, the
PV–diesel–battery power generation system with respect to two structure of the HES is described in Section 3. Section 4 provides
objectives which are: a minimized NPC CO2 emission rates. Dekker a description of HOMER software and the assessment criteria.
et al. [33] performed an economic analysis of a hybrid PV–diesel Input parameters are described in Section 5. The energy dispatch
energy system on the basis of the NPC concept, using HOMER strategy is depicted in Section 6. Simulation results are presented
software, various climatic regions were the subject of this study. and discussed in Section 7. In Section 8, a sensitivity analysis of the
Adaramola et al. [34] studied the economic feasibility of exploiting system is performed. Finally, a conclusion of the results obtained
solar energy in parallel with DGs in the northern part of Nigeria in this study is given in Section 9.
instead of using only DGs to provide electrical energy; the study
has revealed that hybrid PV–diesel systems with battery storage
are more effective from an economical viewpoint with a COE 2. Background
varying between 0.348 $/kWh and 0.378 $/kWh (depending on the
interest rate value) compared to systems using only DGs without Algeria is the biggest country in Africa with an area of
storage with a COE that varies between 0.417 $/kWh and 0.423 2 381 741 km2, bordered in the north by the Mediterranean Sea,
$/kWh. in the east by Tunisia and Libya, in the south by Nigeria and Mali,
High investments of autonomous solar- and wind-energy-based in the southwest by Mauritania and western Sahara and to the
installations are the main reason for curbing their integration in power west by Morocco [42].
plants. Thus, their hybridization allows exploitation of their benefits At its independence, the country's electrification system indicates
from both technical and economical viewpoints. As a result, hybrid PV/ an extreme disparity between different regions of the country with
wind power systems have aroused a great interest in view of the coastline and cities electrified and the rest of the country,
numerous studies performed in this field. Li et al. [35] studied the including rural areas, without any form of energy. The national
feasibility of a hybrid wind/PV/battery power system destined to electrification rate in 1969 was 40% including a strong regional
electrify a household in the region of Urumqi in China, according to dispersion, reflecting the regional socio-economic imbalances. Thus
the NPC of the system and the COE, they found that the hybrid wind/ the government very soon perceived the importance of electrification
PV/battery system is more cost effective compared to the hybrid PV/ as a vector of carrying economic and social progress; hence Algeria
battery and wind/battery systems; a combination of both the PV and launched an ambitious national electrification program which has
wind energy resources has allowed a reduction in the NPC of the allowed it to raise the rate of electricity production from 53% at the
system with approximately 9% and 11% compared to a PV/battery and end of 1975 to 61% in 1979 and 83.8% in 1989 to reach 98.6%
the wind/battery, respectively. in 2010; it is notable that from 2001–2011, the total installed
In Refs. [36,37], the authors have developed optimization sizing capacity in Algeria has increased from 5600 MW to 11,389 MW
models for PV/wind power systems based on the loss of power [43–45].
supply probability (LPSP) and the levelized cost of energy (COE). This boost in electricity production had affected the socio-
Saheb-Koussa et al. [38] have presented a model developed, economic development of the country; this could be observed by a
in Matlab Simulink, to perform the optimal sizing of a hybrid significant increase in the per capita electricity consumption, as
 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150 1137

shown in Fig. 1, in black color, the electricity consumed per capita In this perspective, the production of electricity using
has increased significantly in the period from 1971 (131.87 kWh) RE sources, that Algeria is planning to develop, would be 650
to 2010 (982.17 kWh) with an average annual growth rate of 2.21% megawatts by 2015, 2600 by 2020 and 22,000 MW by 2030, or
(21.80 kWh). Equally the per capita energy consumption (esti- 40% of total electricity generation (37% from solar and 3% from
mated in equivalent fuel kilograms) has reached 1089.27 kg fuel wind resource) [37,46]. Algeria aims to export 10,000 MW of the
compared to 229.64 kg fuel in 1971 with a peak of 1119.91 kg fuel 22,000 MW planned for the next two decades, while the remain-
in 2009; on an average period of 39yr, from 1971 to 2010, an ing 12,000 MW are intended to cover domestic demand [46] (see
annual growth rate of 2.03% has been observed as mentioned by Fig. 2). Once completed, this program will save nearly 600 billion
the graph of Fig. 1 in red [43–45]. cubic meters of gas over a period of 25 yr [47].
The production of electricity in Algeria is based on fossil fuels The evolution of the national electrification rate highlights the
(98% on natural gas). This dependence has increased drastically efforts made by the state; nevertheless, there still exist lots of
the carbon dioxide emission rate. The graph of Fig. 1, in blue, isolated sites in Algeria which are not yet connected to the grid
illustrates that the per capita CO2 emissions have increased from especially in the deep south where electricity is provided only by
0.55 T in 1960 to 3.33 T in 2009 [43–45]. diesel generators. In 2006 the penetration of diesel in the energy
Environmental concerns required the development of the so- sector had been estimated at about 6.61% of the total national
called clean and REs mainly solar and wind energy, which are production and 0.9% in 2010 and 2011. Fig. 3 represents the
available and abundant especially in the south (Solar) and south- evolution of fuel contribution in the energy sector over the period
west (wind) of Algeria. The contribution of wind and solar from 2000 to 2006 [48].
energies in electricity production is still low; it represents only The use of diesel plants in remote areas, especially in the deep
3% of the national production [46]. south, is extremely restrictive given the lack of road infrastructures to
In fact, with the aim of protecting the environment, a national deliver fuel, high costs of fuel transportation depending on the site
renewable energy program has been drawn for a gradual intro- location, and frequent maintenance required. Despite all these dis-
duction of alternative sources including solar energy with its two advantages DGs are still used as the main means of power generation
thermal and photovoltaic branches in power generation over the for the isolated sites. A combination of REs with DGs for power
next 20 yr in the period (2011–2030) [46]. generation in HES is suitable as the combination is more advantageous

Electricity consumption (Kwh/Capita)

Energy consumption (fuel equivalent Kg/Capita)
CO 2 Emission (T/Capita)
1200
1000 3.5

1000 3.0
800
2.5
800
600 2.0
600
1.5
400
400 1.0
200
0.5
200
0 0.0
1960 1970 1980 1990 2000 2010
Year

Fig. 1. Electricity and energy consumption, CO2 emission rates per capita in Algeria. (For interpretation of the references to color in this figure, the reader is referred to the
web version of this article.)

Fig. 2. Structure of the national power generation in MW [47].

1138 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150

440000 Gasoil (TM) Table 1

Technical and economical characteristics of the HES components.
420000
Description Data description
400000
Photovoltaic modules
Gasoil (TM)

Technology REC solar

380000
Capital cost (US$/kW) 5678
Replacement cost (US$/kW) 5678
360000 Operation and maintenance cost ($/yr) 0
working life 25
340000 Derating factor (%) 90
Tracking system Two axis
320000 Ground reflectance (%) 20
Storage battery
300000 Technology Hoppecke-1–10 OPzS
2000 2001 2002 2003 2004 2005 2006 1000
Year Nominal voltage of each battery 2V
Nominal capacity of each battery 500 Ah
Fig. 3. Fuel penetration in the energy sector. Capital cost (US$/kW) 506
Replacement cost (US$/kW) 506
Operation and maintenance cost ($/yr) 4
Min. state of charge 30%
Power converter
PV modules Charge Controller Diesel generator Technology EFFECTA
Capital cost of power converter (US$/kW) 734
Converter Replacement cost of power converter (US$/kW) 734
Operation and maintenance cost of power converter 2
(US$/yr)
Battery Load Considered sizes 50 kW
Diesel generator
DC Bus AC Bus Capital cost of diesel generator (US$/kW) 313
Replacement cost of diesel generator (US$/kW) 313
Fig. 4. Structure of the PV–diesel–battery HES. Operation and maintenance cost of diesel generator 0.02
(US$/h)
compared to those based on single energy systems: compensate rising
fuel prices, reduce operating costs, provide better efficiencies and
reduce carbon emissions.
The technical characteristics, costs and life time of each compo-
nent of the hybrid energy system used by HOMER to perform
3. Hybrid energy system description simulations are summarized in Table 1. For the PV subsystem, the
capital cost of 1 kW PV generator is 5678$ and it includes all the
The HES under study consists of PV modules, batteries for energy costs associated with the PV modules which are the mounting
storage, power converters and diesel generators. Fig. 4 shows the hardware, the tracking system, the maximum power point tracker,
structure of the hybrid energy PV–diesel–battery system. the wiring and installation costs.
An estimated factor of 2.5, which takes into account the
3.1. Photovoltaic modules purchase costs, import taxes and transportation costs to UDES
has been multiplied by the purchase price of each component
A photovoltaic module is an electric direct current generator to compute the final cost that has been used to perform
which consists of a variable number of photovoltaic cells electri- simulations.
cally connected. The mono-crystalline PV modules of REC solar
technology have been used in this study.

3.2. Battery bank

4. HOMER software
The battery bank is used to serve the required load. If the
battery bank is less than 30% charged, the diesel generator starts 4.1. Description
supplying the load. The HOPPECKE battery (2 V, 500 Ah) has been
chosen in this work from the HOMER list. The Hybrid Optimization Model for Electric Renewables (HOMER)
is a software developed by the National Renewable Energy Labora-
3.3. Power converter tory (NREL) [49] in the United States (USA). It is a powerful tool for
designing and analyzing off-grid and grid-connected hybrid electrical
The converter is required in this system for two purposes; to power systems; it offers a large number of energy production and
convert AC power from diesel generator to DC power used to storage systems (wind turbines, PV systems, hydroelectric dams, fuel
charge the batteries, in this case it is called converter or to work as cells, etc.).
an inverter (DC–AC) to serve AC loads. Based on input parameters, which are the load profile, resources
potentials, technical and economic characteristics of the HES
3.4. Diesel generators components, the software simulates all possible configurations
and displays the results as a list of the feasible ones, that meet
DGs are used to meet the load demand when the load is not the load needs in the most cost-effective way, ranked by their NPC
satisfied by the PV power system or when the batteries are in ascending order. HOMER also displays results in the form of
discharged. tables and graphs [50].
 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150 1139

4.2. Economic assessment criteria data have been provided by the accounting service of the UDES by
referring to the Algerian Electricity Company bills. For the area
4.2.1. Net present cost under study, the consumption of electrical energy is high between
The net present cost includes the cost of installing and operat- 8 h and 30 and 16 h and 30 for each month throughout the year,
ing the system throughout its lifetime. The economic outputs of and it coincides with office hours, outside this interval the
the system are calculated for the purpose of finding the net electricity consumption falls down and the little amount of
present cost [50,51]. The net present cost is calculated using the electricity consumed during the night is used to provide lighting.
following equation: The UDES energy consumption has been estimated at about

509 kWh/day (distributed between lighting, heating, air condi-
NPC ¼ TAC=CRF i; Rproj ð1Þ
tioning, laboratories and electronic machines).
where
TAC is the total annualized cost ($); CRF is the capital recovery
5.2. Solar radiation data
factor; i is the interest rate (%) and Rproj is the project life time (yr).
The solar radiation data for the area under study have been
4.2.2. Total annualized cost downloaded from the NASA website [52]. As can be seen from
The total annualized cost (TAC) of energy is the sum of the Fig. 6, the solar radiation data vary in a range of 2.120–6.795 kWh/
annualized costs of each component of the power system includ- m2/day. The scaled annual average value, computed by Homer
ing capital, replacement, operation and maintenance and fuel costs software is 4.512 kWh/m2/day with a clearness index of 0.552.
[50,51]. The high radiation intensity periods are observed from April
(5.198 kWh/m2/day) to September (5.085 kWh/m2/day) with the
4.2.3. Capital recovery factor highest intensities in June and July (6.795 kWh/m2/day and
The capital recovery factor is a ratio used to calculate the 6.740 kWh/m2/day, respectively) and the lowest intensity period
present value of a series of equal annual cash flows [50,51]. is in December (2.120 kWh/m2/day).
CRF ¼ ½in ð1 þiÞN =ðiþ 1ÞN  1  ð2Þ
5.3. Diesel
where
N is the number of years and i is the annual real interest rate.
According to a study prepared by the GIZ, the German firm
specializingin multi-sectorial expertise at the regional and inter-
4.2.4. Annual real interest rate
national levels, Algeria is ranked among the top 10 countries
The annual real interest rate (i) is the discount rate used to
where fuel is cheapest in the world [53]. Currently, the price of
convert between one-time costs and annualized costs [50,51]. The
diesel in Algeria is 13.5 Algerian Dinar (AD) per liter, which is
annual real interest rate is a function of the nominal interest rate
equivalent to $0.1685/L. In fact the cost of fuel varies by region; in
as shown by Eq. 3
remote areas the price would be higher because of the transporta-
i ¼ ði'  f Þ=ð1 þ f Þ ð3Þ tion costs. Another cause responsible for increasing fuel prices is
where fuel shortage, in some cases fuel prices increase up to 50 AD per
i is the real interest rate; i' is the nominal interest rate; and f is liter ($0.625/L). Thus, in this study several simulations were
the annual inflation rate. performed with different diesel prices to see their effect on the
total cost of the system [53].
4.2.5. Levelized cost of energy
The levelized cost of energy (COE) is the average cost per kWh
6. System dispatch strategy
of useful electrical energy produced by the system. It is computed
by dividing the annualized cost of producing electricity by the
HOMER simulates how the components of the HES function
total useful electric energy production [50,51]. The equation for
together as a single system for each hour and takes decisions as to
the COE is as follows:

which generators should operate and at what power level and
COE ¼ TAC= Eprim;AC þEprim;DC ð4Þ whether to charge or discharge the batteries. In this study, the
where following load dispatch strategy has been adopted where the load
Eprim,AC is the AC primary served load and Eprim,DC is the DC is supplied by the PV system and DGs work only when the
primary served load. batteries are discharged to produce enough of energy to meet
the primary load requirements. Under this dispatch strategy
HOMER dispatches the power sources in order to work in a way
5. Homer input parameters description to satisfy the primary load at the least total cost.

5.1. Load profile

7. Simulation results
The UDES is a research unit located in Bou-Ismail, department
of Tipaza, at about 45 km from the west of the capital of Algeria, Simulations have been performed using the HOMER input
361390 latitude in the north and 21420 longitude in the east. It is parameters described in Section 5. The study consists of compar-
one of the active Algerian organizations of the R&D network in the ing the optimal configurations of a PV–diesel–battery energy
field of RE in Algeria [41] (see Fig. 5a). system with (25%, 50% and 75%) PV penetrations, a stand-alone
The initial data in the implementation of a RE system are the diesel system and a stand-alone PV power generation system. The
energy demand which must be determined relatively to the load simulation has been performed for a project life time of 25 yr and
that may be supplied. an annual real interest rate of 6%. In the case of a system using
The load profile considered in this work is the UDES monthly DGs, HOMER decides by default for each hour of the year whether
average electrical consumption of the year 2012 (see Fig. 5b). The to operate the generator or not according to the load demand.
1140 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150

Fig. 5. (a) The location of the research unit (Unité de Développement des Equipements Solaires, UDES) [41] and (b) UDES daily consumption profiles.

Fig. 6. Solar radiation data.

 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150 1141

In some cases DGs have been scheduled not to operate in 7.1. Stand-alone diesel system
certain times to avoid production of electricity in excess. PV panels
are mounted on a two-axis tracking system so they rotate about The stand-alone diesel system was found to be the most cost-
both vertical and horizontal axes. This makes the solar radiation effective system compared to the other systems analysed in this
perpendicular to the PV panels' surface, then the output power is study since it doesn't comprise any RE resource which has high
extremely maximized. A derating factor of 90% has been applied installation costs. The optimal configuration of the only-diesel
on the output power of PV panels to account for losses when the system has been obtained with the combination of two-diesel
PV module is submitted to real working conditions: higher generators DG1 and DG2 with a rated power of 50 kW and 23 kW,
ambient temperature, different operating voltages and soiling. respectively. The system satisfies the annual load (185,785 kWh)
with only 6017 kWh (3%) of energy produced in excess. The

Table 2
Technical and electrical characteristics of the stand-alone diesel system.

System components Electricity production Excess electricity Unmet electric load Capacity shortage Annual hours of operation Annual fuel consumption

DG1 127,931 6017 (3.15%) 727 (0.39%) 1873 (1.01%) 3233 44,915

DG2 63,143 6842 28,375

Total 191,074 10,075 73,290

Generator 1
Generator 2
Monthly Average Electric Production
30

25
Power (kW)

20

15

10

0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Fig. 7. Monthly average electric power produced by the stand-alone diesel system.

Table 3
Economic characteristics of the diesel system.

Component Sizes (kW) Capital ($) Replacement ($) O&M ($) Fuel ($) Salvage ($) N.P.C ($) L.C.O.E ($/kWh) Operating cost ($/yr)

DG1 50 15,650 37,369 41,329 96,746  2230 188,863 0.142 24,519

DG2 23 7199 39,871 40,233 61,120  1001 147,423
Total – 22,849 77,239 81,562 157,866  3231 336,285

Capital
Replacement Fuel
Cash Flow Summary Operating Salvage

200,000
Net Present Cost ($)

150,000

100,000

50,000

-50,000
G1 G2

Fig. 8. The net present cost of the diesel system as a function of its different components' cash flows.
1142 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150

Generator 1
Cash Flow Summary Generator 2
200,000

150,000

Net Present Cost ($)

100,000

50,000

-50,000
Capital Replacement Operating Fuel Salvage

Fig. 9. Cash flow summary of different costs of the diesel system.

Table 4
Annual greenhouse gases emitted by the stand-alone diesel system.

Pollutant Carbon dioxide Carbon monoxide Unburnt hydrocarbons Particulate matter Sulfur dioxide Nitrogen oxides Total

Emissions (kg/yr) 192.997 476 52.8 35.9 388 4251 198200.7

electrical characteristics of the system are shown in Table 2. Fig. 7 Table 5

illustrates the monthly average electric power produced by the Technical characteristics of the hybrid and stand-alone PV power systems.
diesel system.
PV contribution (%) PV G1 G2 G3 Storage bank Converter
Based on the fuel price of 0.1685$/L, the total NPC of the system (kW) (kW) (kW) (kW) (kWh) (kW)
has been estimated at about 336,285$ with a COE of 0.142 $/kWh.
The details of the diesel system cash flow are reported in Table 3 25 23 75 – – 170 50
and presented in Fig. 8. 50 43 15 11 35 700 75
75 66 15 50 – 10,000 100
The initial capital cost of the stand-alone diesel system has 100 80 – – – 25,000 100
been estimated at about 22,849$ where 15,650$ belongs to the
DG1 7109$ to the DG2. The generators are considered to have a
lifetime of 15,000 h. 7.2. Hybrid energy system scenarios
Fig. 9 shows the cash flow distribution of the system as a
function of different cash flows; it has been noted that the cost of 7.2.1. Technical and electrical analysis
fuel consumption is the most expensive component (157,866$) The HESs studied in this work consisted of a combination of
followed by the (O&M) costs (81,562$) which are considered as one RE source which is the PV resource working in parallel with
important compared to the installation costs. This feature is the DGs used as buck-up subsystems, a battery bank for energy storage
main drawback of the stand-alone diesel systems where a lot of and power converters. A suitable mix of these components has
money is spent on fuel to satisfy the load over the project been explored to match the predefined peak load with a varying
lifetime, the task for which the system is conceived. In fact, the PV penetration in the range of 0% to 100% with an increasing step
cost of fuel is not stable and could increase, this leads to of 25%.
increase in the total cost of the system and therefore increase Table 5 shows the technical characteristics of the optimal
in the COE which could, thereafter, have an impact on the choice configuration for each system. In the case of the 25% PV-based
of the production system. system, the optimal configuration consisted of a PV generator of
As mentioned in Table 2, the DG1 operates 3233 h/yr and 23 kW, a diesel generator of 75 kW, a storage bank of 170 kWh and
contributes 67% in the total production which leads to the a power converter of 50 kW. This system produces 206,852 kWh/
consumption of 44915L of fuel per year with a rate of 0.351 L/ yr of electrical power where 51,025 kWh/yr (25%) comes from the
kWh, concerning DG2 it operates 6842 h/yr to contribute 33% of PV source and the remaining 155,827 kWh/yr (75%) is produced by
the total production, this leads to the consumption of 28,375 L of the DG. The energy produced in excess is estimated at about
diesel per year with a rate of 0.449 L/kWh. The absence of a RE 699 kWh/yr and the load is satisfied at 100%.
resource in this system has increased the quantity of fuel con- The 50% PV-based system is composed of 43 kW aggregated PV
sumed for its operating, consequently it has led to an ejection of panels, a set of three diesel generators of 15 kW, 11 kW, 35 kW
about 192,997 kg/yr of CO2. The combustion of fuel leads to the rated capacities, respectively, a 700 kWh storage bank and a
emmission of not only CO2 but big amounts of other pollutants 75 kW rated power converter. The PV subsystem produces 50%
and GHGs which are harmful for the environment the time they of the annual electrical energy produced by the HES and the
contribute to the greenhouse effect, cause global warming and remaining energy is ensured by the three DGs. In this system the
destruct the ozone layer. three DGs have been scheduled to operate in specific hours of the
According to Refs. [54,55], there is a direct relationship day depending on the load demand in order to minimize the
between the quantity of released pollutant and global warming; excess of electricity produced and not used by the load, which
the emission of a trillion tons of carbon causes a 2 1C peak means reducing their operating hours and therefore minimizing
warming. Also under the effect of nitrogen oxides and sulfur the fuel intake and reducing the rate of released GHGs and
dioxide, snow and rain become more acidic and alter the ecosys- particulate matter.
tem. The total quantity of pollutants emitted by the autonomous For the hybrid system with 75% PV penetration, the recom-
diesel system under study is estimated to be 198200.7 kg/yr; the mended configuration is composed of a 60 kW PV generator, two
details of these emissions are given in Table 4. DGs with a rated power of 50 kW and 15 kW, respectively, a storage
 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150 1143

bank of 23,000 kWh and a 100 kW sized power converter. The PV duration of the project lifetime, are given in columns [2–6]. The
subsystem produces about 155,542 kWh/yr of electrical power. COE estimated ($/kWh), the net present cost ($) and the operating
In the case of 100% PV penetration, the optimum configuration cost ($) are equally reported in columns [7–9].
has been obtained using an 80 kW PV generator, a storage bank of As indicated in Table 7, the COE and the NPC increase when the
25,000 kWh and a 100 kW sized power converter. The total required penetration of the PV resource increases in the HES; the COE has
energy is provided by the PV subsystem with 187,619 kWh of energy increased from 0.260 $/kWh to 0.374 $/kWh then to 2.924 $/kWh
produced annually. The details of the electrical characteristics (PV to reach 6.904 $/kWh as well as the NPC was found to be equal to
and diesel energy production, total production, electricity excess, 617,489$, 885,813$, 6,944,177$ and 16,251,990$ for the 25%, 50%,
unmet load, capacity shortage) of each system are reported in 75%, 100% PV-based systems, respectively. These high costs could
Table 6. be justified by the low efficiency of PV modules compared to fuels
and the use of storage devices which have high costs.
7.2.2. Economic analysis Fig. 10 presents the details of the NPC by cost type (capital,
The details of the cash flow summary of each system design are replacement, operating, fuel and salvage) for each optimal hybrid
represented in Table 7. The initial capital cost (ICC), replacement, system. The distribution of each cost type, as a function of the HES
operation, fuel consumption and salvage costs, evaluated on the components (PV modules, DGs, batteries and converters) is

Table 6
Electrical characteristics of the hybrid and stand-alone PV power systems.

PV contribution PV (kWh) G1 (kWh) G2 (kWh) G3 (kWh) Total production Excess electricity Unmet load Capacity shortage
(%) (kWh) (kWh) (kWh) (kWh)

25 51,593 155,301 – – 206,894 474 (0.229%) 0.000135 (0%) 1.38 (0%)

50 100,851 21,213 43,473 36,462 201,999 1228 (0.618%) 59.6 (0%) 105.6 (0%)
(11%) (22%) (18%)
75 155,542 12,915 (6%) 38,600 – 207,057 649 (0.32%) 0.000234 (0%) 0 (0%)
(75%) (19%)
100 187,619 – – – 187,619 0.000365 (0%) 1.25 (0%) 2.78 (0%)

Table 7
Economic characteristics of the hybrid and stand-alone PV power systems.

PV contribution (%) Capital ($) Replacement ($) O&M ($) Fuel ($) Salvage ($) NPC ($) L.C.O.E ($/kWh) Operating cost ($/yr)

25 271,111 185,551 64,761 122,694  26,628 617,489 0.260 27,096

50 672,497 147,141 63,197 75,469  72,490 885,813 0.374 16,687
75 5,597,393 1,750,556 528,468 41,362  973,604 6,944,177 2.924 105,354
100 131,776,40 4,085,969 1,278,336 0  2,289,956 16,251,990 6.904 240,496

Cash Flow Summary Cash Flow Summary

300,000 800,000
PV battery
PV Generator 2 battery
Generator 1 Converter
Generator 1 Generator 3 Converter
250,000
600,000
Net Present Cost ($)

Net Present Cost ($)

200,000
400,000
150,000

100,000 200,000
50,000
0
0

-50,000 -200,000
Capital Replacement Operating Fuel Salvage Capital Replacement Operating Fuel Salvage

Cash Flow Summary Cash Flow Summary

6,000,000 20,000,000
PV Generator 2 Converter PV Converter
Generator 1 battery
5,000,000 battery

15,000,000
Net Present Cost ($)

Net Present Cost ($)

4,000,000

3,000,000 10,000,000

2,000,000
5,000,000
1,000,000
0
0

-1,000,000 -5,000,000
Capital Replacement Operating Fuel Salvage Capital Replacement Operating Fuel Salvage

Fig. 10. NPC details of each power system as a function of different costs.
1144 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150

battery cost ($)

NPC ($)
16000000

14000000

12000000

costs($) 10000000

8000000

6000000

4000000

2000000

0
0 % 25 % 50 % 75 % 100 %
PV Contribution
Fig. 11. Domination of the cost of hybrid systems by the cost of the storage system.

C.O.E ($/KWh)
NPC ($)
PV subsystem initial Cost ($)
0000000 8 Storage Bank Initial Cost ($)
1500000

400000
5000000 6
1000000

0000000 4

200000
5000000
5000000 2

0 0 0 0
0 25% 50% 75% 100%
PVPenetration

Fig. 12. Effect of the PV penetration on the NPC, COE, PV system and battery bank costs for each power system.

represented in different colors. For the 25% PV-based system, the subsystem and the storage bank system costs, the variation of
initial capital cost of the PV subsystem constitutes 46.07% of the the NPC and the COE as a function of the PV penetration. It's
total capital cost of the system where the storage bank constitutes evident that the NPC and COE curves have the same shape as that
31.72%, which means that the cost of the PV subsystem is of the storage bank cost curve; this means the cost of the storage
dominant in this case. For the other scenarios where the PV system affects the COE and NPC of the HES rather than the cost of
penetration varies from 50% to 100%, the capital cost of the PV the PV subsystem, which can be seen more clearly (see Fig. 13). So
subsystem constitutes 36.30%, 6.7% and 3.44% of the initial capital it can be concluded that for this class of power systems, where
cost of the concerned system, respectively. The capital cost of the storage devices are used, a large investment is required to achieve
storage bank varies in its turn from 52.6% to 90.39% to reach high penetration of the PV resource.
95.99%. This means that the cost of batteries dominates the initial An analysis of the replacement costs of the HESs components
cost of HES installations, thereby the high costs of the energy shows that the storage battery bank has the highest percentage
produced by a HES result from the high costs of the storage bank which has been estimated at about 84,696 $, 110,441$, 157,734$,
system which should be provided in such a system to store energy 3,944,334$ for the 25%, 50%, 75% and 100% PV-based HESs,
as a consequence of solar resource intermittency that doesn't respectively, and it's dominant for the last three systems. This
allow the production to be adapted continuously to the load needs. situation could also be interpreted by the high costs of batteries.
Similarly, the NPC of the HES is affected by the size of the The operating and maintenance costs of the DGs and the
storage system, for the 50%, 75% and the 100% PV-based systems quantity of fuel consumed by the HESs have decreased linearly
the size of the storage bank has increased from 700 kWh to when the percentage of the PV penetration increased as mentioned
10,000 kWh, then to 25,000 kWh and as a consequence the NPC in Fig. 14. This is mainly due to the reduction of operating hours of
of each system is dominated by the cost of the storage bank as DGs when the PV penetration increases, which automatically
mentioned in Fig. 11. Fig. 12 shows the variation of the PV implies the reduction of fuel consumption and O&M costs.
 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150 1145

NPC ($)
16000000 LCOE ($)
6
14000000

12000000
4
10000000

8000000
2
6000000

4000000

2000000 0

0
0 (0%) 271111 (25%) 354200 (50%) 5060000 (75%) 11638000 (100%)
storagebankcost ($)
Fig. 13. Effect of the battery bank on the NPC and COE of each power system.

O&M ($)
80000 Fuel ($) 150000

60000
100000

40000
50000

20000

0
0% 25% 50% 75% 100%
PV Penetration

Fig. 14. Impact of increasing PV penetration on the fuel and O&M costs of each power system.

80000 6000
Fuel Consumption (L/Yr)
CO2 (Kg/Yr) 200000
60000 Pollutants (Kg/Yr)

4000
150000
40000

100000 2000
20000

0 50000
0

-20000 0
0% 25% 50% 75% 100%
PV Penetration

Fig. 15. Effect of increasing the PV penetration on fuel consumption and GHG saving.

7.2.3. Environmental impacts of hybridization quantity of released dioxide carbon and other pollutants as a
The hybrid mode has positive effects on the environment; the function of PV penetration, as the penetration of PV resource
hybridization of the only diesel system with the PV resource increases, the fuel consumption decreases due to the reduction of
avoided the emission of tons of GHGs in the atmosphere. Fig. 15 operational hours of DGs. The annual quantity of saved fuel has
represents the variation of the quantity of fuel consumed, the reached 23.68%, 53% and 72.27% for 25%, 50% and 75% PV pene-
1146 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150

Renewable Fraction
Renewable Fraction vs. Global Solar PV Array Capacity
0.34 35

0.32 30

PV Array Capacity (kW)

Renewable Fraction
0.30 25

0.28 20

0.26 15

0.24 10
4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0
Global Solar (kWh/m²/d)

Fig. 16. Effect of varying the GSR on the PV fraction in the HES.

tration, respectively, compared to the only diesel system. Further- cost-effective compared to other systems with the lowest COE of
more, the reduction of operational hours of DGs has avoided the 0.260 $/kWh.
emission of 45,712 kg/yr, 102,251 kg/yr and 147,210 kg/yr of CO2 in
the local atmosphere compared to the only diesel system, also
it has been noticed that the introduction of PV resource contri-
buted to the reduction of other pollutants such as unburned 8. Sensitivity analysis
hydrocarbons, nitrogen oxides, sulfur dioxide and particulate
matters. So the high costs resulting from introducing PV resource In this section, various sensitivity analyses have been per-
with DGs are justified in terms of saving the atmosphere and formed on the 25% PV–diesel–battery HES, which involved chan-
improving the living conditions. Equally, reducing operating hours ging some parameters in order to anticipate possible costs of the
of DGs stretches their lifetime duration and reduces their O&M system, the parameters studied are: global solar radiation, diesel
costs. The 100% PV system that does not contain any conventional price, load consumption and real interest rate.
resource is considered as clean with 0% emission.
From the results obtained in this section, the following key
remarks can be made:
8.1. Global solar radiation and diesel price
 Replacing diesel mini-grids with hybrid PV–diesel mini-grids is
an effective measure to reduce diesel fuel consumption and The efficiency of PV modules depends on the intensity of solar
emissions of GHGs while providing electrical service 24 h per radiation; when the irradiation increases the power produced by
day of small islands and other rural or remote infrastructures. the PV system increases and vice-versa. Thus, the contribution of
 The scheme which has the greater PV penetration has the solar energy in the hybrid power system depends on the global
bigger cost, consumes less fuel, emits less GHGs and vice versa. solar radiation (GSR) received and captured on the surface of the
 The size of the battery bank is a key element that influences PV panels. Fig. 16 depicts the variation of the PV penetration as a
largely the choice of the optimal power system at the basis of function of the global solar radiation for the same PV array
the life cycle cost of the system and the cost of kilowatt hour of capacity (23 kW), the contribution of PV energy has been opti-
electricity produced by the HES over its lifetime. An effective mized (see Fig. 16), as can be seen, for an increase of 0.50 kWh/m2/
management of electrical energy would maximize the lifetime day, 1 kWh/m2/day and 1.5 kWh/m2/day approximately; an
of the batteries because even their replacement cost has a increase of 3%, 6% and 8%, respectively, of the renewable fraction
significant share in the total cost of the system. has been observed.
 Even though DGs play a very important role in the HES to As shown in Fig. 17, the increasing GSR intensity has, equally, a
ensure the quality of service when PV technology produces direct influence on the NPC and the COE; as the renewable fraction
little energy or when the batteries are discharged, the use of increased, the NPC and the COE decreased linearly (always for the
DGs should be minimized due to the high costs of fuel same PV array capacity of 23 kW). Table 8 reports the effect of the
consumed over the project's life span. GSR increase on the renewable fraction, the NPC, the COE and the
 As reported in the literature, the acceptable PV penetration in an annual fuel consumption.
off-grid PV–diesel HES varies in the range of 11–25% [30]. The results of the sensitivity analysis performed on both
A penetration of less than 11% does not allow us to reap the parameters, the price of diesel and the GSR are reported in
benefits of using this type of systems; such as reducing depen- Fig. 18. The diesel price has been varied from 0.168$/L to 0.6$/L
dency on fuel and reducing GHG emissions. On the other hand, for and the GSR from 4.5 kWh/m2/d to 6.00 kWh/m2/day. As clearly
high penetrations of PV resource, it results in technical difficulties shown in Fig. 18, the cost of a kilowatt hour generated by the
to maintain a stable voltage and frequency. hybridized system depends directly on the local solar resource
Taking into account that the objective of this work is to determine (which determines the amount of electricity generated by the PV
the optimal power system in respect to load satisfaction, mini- system for a given investment) and the local cost of diesel. The
mized NPC, optimized COE (around 0.2$ as reported in Ref. [30]), COE generated by the HES is acceptable as long as the diesel price
reduction of CO2 emissions, considering an effective rate of PV is lower than 0.25$/L for the considered range of solar radiation,
penetration that guarantees a stable voltage and frequency of the once the diesel price exceeds this value, the COE is no longer
network, the 25% PV-based HES was found to be optimal and adequate for this study.
 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150 1147

Total Net Present Cost

Total Net Present Cost vs. Global Solar Levelized Cost of Energy
675,000 0.286

Levelized Cost of Energy ($/kWh)

0.284

Total Net Present Cost ($)

670,000

0.282
665,000
0.280
660,000
0.278

655,000
0.276

650,000 0.274
4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0
Global Solar (kWh/m²/d)

Fig. 17. Effect of the GSR on the NPC and the COE of the HES.

Optimal System Type

System Types
0.6
0.393 0.384 0.377 0.373 PV/G1/Battery

Superimposed
Levelized COE ($/kWh)

0.5
0.363 0.355 0.348 0.345
Diesel Price ($/L)

0.4
0.332 0.326 0.320 0.317

0.3
0.302 0.296 0.292 0.290

0.287 0.282 0.278 0.276

0.2
0.262 0.258 0.255 0.253

4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0

Global Solar (kWh/m²/d)

Fig. 18. Effect of varying GSR and diesel price on the COE generated by the HES.

Table 8 155,211$ and increasing of O&M costs from 55,269$ to 66,844$ as

Effect of GSR on the PV contribution, NPC, COE and fuel consumption.
shown in Fig. 20. The COE has decreased because the PV penetra-
GSR (kWh/m2/d) PV contribution (%) NPC ($) COE ($/kWh) Fuel (L/yr) tion has been reduced; in this case it is the local cost of diesel fuel
that influences the COE and not the cost of the PV resource (see
4.51 25 674,549 0.284 55,919 Fig. 20). Table 9 reports an overview of how the diesel and PV
5.00 28 664,937 0.280 53,242 penetrations, the NPC, the COE, the fuel consumption and O&M
5.50 31 657,087 0.277 51,140
6.00 33 653,403 0.275 50,206
costs are influenced by the load consumption variation.

8.2. Load consumption 8.3. Real interest rate

The output power of an energy system depends on the load The impact of varying the real interest rate on the COE and the
requirements that should be satisfied at all times. If the load NPC is represented in Fig. 21. The real interest rate has been varied
consumption increases, the energy production should impera- from 5% to 10%. The COE and the real interest rate are linearly
tively be increased; in the case of the HES under study the related, the COE increased from 0.252 $/kWh for a real interest of
production capacity has been increased from 185,785 kWh/yr for 5% to 0.304 $/kWh for an interest rate of 10% which is because the
a load consumption of 509 kWh/d to 233,452 kWh/yr for a con- TAC of the HES increases with the increasing real interest rate (Eq.
sumption of 640 kWh/day (see Fig. 19). Since the size of the PV 4) as mentioned in Table 10. However, the real interest rate doesn't
subsystem has remained the same, the capacity shortage is have a significant impact on the COE as we can see when the real
ensured by the DG. The dependence on the conventional energy interest rate decreases by 50%, the COE decreases only by 2.7%.
to meet the peak load has increased the diesel penetration to over Unlike the COE, the net present cost of the system is inversely
80% and decreased that of the PV subsystem to (20%). The total related to the real interest rate, because according to Eqs. (1) and
cost (NPC) of the HES has increased from 621,498$ to 689,348$ (2) when the real interest rate increases, it will decrease the cost of
due to the increasing costs of diesel consumption from 121,049$ to the present value accumulated over the project's lifetime.
1148 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150

Total Electrical Production

Total Electrical Production vs. udes Renewable Fraction
260,000 0.25

250,000 0.24

Total Electrical Production (kWh/yr)

Renewable Fraction
240,000 0.23

230,000 0.22

220,000 0.21

210,000 0.20

200,000 0.19
520 540 560 580 600 620 640
udes (kWh/d)

Fig.19. Effect of load variation on the electrical production of the HES and the renewable fraction.

Total Net Present Cost

Total Net Present Cost vs. udes Levelized Cost of Energy
690,000 0.265

680,000 0.260

Levelized Cost of Energy ($/kW

h)
Total Net Present Cost ($)

670,000 0.255

660,000 0.250

650,000 0.245

640,000 0.240

630,000 0.235

620,000 0.230
520 540 560 580 600 620 640
udes (kWh/d)

Fig. 20. Impact of load variation on the NPC and the COE of the HES.

Table 9
Effect of varying the load consumption on the diesel and PV contribution, NPC and COE, fuel consumption and O&M costs of the HES.

Load consumption (kWh/d) Diesel production (kWh/yr) (%) PV production (kWh/yr) (%) NPC ($) COE ($) Fuel consumption ($) O&M costs ($)

509 155,621 (75%) 51,025 (25%) 621,498 0.262 121,094 55,262

540 167,180 (77%) 51,025 (23%) 637,193 0.253 129,199 58,120
600 189,530 (79%) 51,025 (21%) 668,131 0.239 144,685 63,239
640 204,565 (80%) 51,025 (20%) 689,347 0.231 155,211 66,844

9. Concluding remarks alone PV system. HOMER has been used to perform the techno-
economic analysis of each system for a projection period of 25 yr,
In this paper, a feasibility study of a PV–diesel–battery HES based on the recommendations cited above concerning the selec-
destined to electrify a research unit (UDES) located in the northern tion of the optimum power generation system on the basis of
part of Algeria has been performed. The PV penetration has been optimized NPC and COE concepts, fuel consumption economy,
varied from 0% to 100% including a stand-alone diesel and a stand- operation time of the DG and the GHGs saving rates.
 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150 1149

Total Net Present Cost

Levelized Cost of Energy
Total Net Present Cost vs. Interest Rate
0.31

650,000 0.30

Levelized Cost of Energy ($/kWh)

Total Net Present Cost ($)
0.29

600,000

0.28

550,000 0.27

0.26

500,000

0.25
5 6 7 8 9 10
Interest Rate (%)

Fig. 21. Impact of increasing real interest rate on the NPC and the COE of the HES.

Table 10 References
Impact of increasing real interest rate on the total annual cost of the HES.
[1] Boudghene Stambouli A, Koinuma H. A primary study on a long-term
Real interest rate (%) 5 6 7 8 9 10 vision and strategy for the realization and the development of the
Sahara Solar Breeder project in Algeria. Renew Sustain Energy Rev 2012;16
T.A.C ($/yr) 46,767 48,618 50,514 52,454 54,440 56,469 (1):591–8.
[2] Luthra S, Kumar S, Kharb R, M.d. Fahim Ansari, Shimmi SL. Adoption of smart
grid technologies: an analysis of interactions among barriers. Renew Sustain
Energy Rev 2014;33:554–65.
The analysis of the power systems recommends the 25% PV– [3] de Castro C, Mediavilla M, Miguel LJ, Frechoso F. Global solar electric potential:
diesel–battery HES obtained with a 23 kW PV system, a 75 kWDG, a a review of their technical and sustainable limits. Renew Sustain Energy Rev
170 kWh storage system and a 50 kW power converter as the 2013;28:824–35.
[4] Akikur RK, Saidur R, Ping HW, Ullah KR. Comparative study of stand-alone and
economically optimal system, compared to the other systems, with
hybrid solar energy systems suitable for off-grid rural electrification: a review.
the lowest NPC and COE (617,489$ and 0.260 $/kWh, respectively) Renew Sustain Energy Rev 2013;27:738–52.
and with lesser GHGs where about 102,251 kg/yr of CO2 has been [5] Himri Y, Boudghene Stambouli A, Draoui B, Himri S. Techno-economical study
avoided to be released in the atmosphere (compared to stand-alone of hybrid power system for a remote village in Algeria. Energy 2008;33
(7):1128–36.
diesel system), which helps to keep a safer and clean environment. [6] Agudeloa A, Valero A, Usónb S. The fossil trace of CO2 emissions in multi-fuel
The sensitivity analysis performed on the 25% PV-based HES energy systems. Energy 2013;58:236–46.
indicated that the global solar radiation has a significant influence [7] Shafiullah GM, Amanullah MT, Jarvis D, Shawkat Ali ABM, Wolfs P. Prospects of
solar energy in Australia. IEEE Int Conf Electr Comput Eng 2010:350–3.
on some parameters of the HES since the efficiency of a PV system [8] Leggett LMW, Ball DA. The implication for climate change and peak fossil fuel
increases with the increasing radiation, the PV contribution in the of the continuation of the current trend in wind and solar energy production.
HES has consequently increased for the same size of PV system and Energy Policy 2012;4:610–7.
[9] Höhne N, Wartmannb S, Heroldc A, Freibauer A. The rules for land use, land
then for the same initial investment which has a direct impact on use change and forestry under the Kyoto Protocol – lessons learned for the
reducing the NPC and COE, fuel consumption and GHG emissions. future climate negotiations. Environ Sci Policy 2007;10(4):353–69.
The study of the impact of both diesel price and global solar radiation [10] Hu Y, Monroy CR. Chinese energy and climate policies after Durban: save the
Kyoto Protocol. Renew Sustain Energy Rev 2012;16(5):3243–50.
on the 25% PV HES allowed us to define a maximum threshold of
[11] Ghasemi A, Asrari A, Zarif A, Abdelwahed S. Techno-economic analysis of
diesel price (0.25$/L) beyond which the price of kWh generated by stand-alone hybrid photovoltaic–diesel–battery systems for rural electrifica-
the HES is not acceptable and the HES is no more profitable. tion in eastern part of Iran – a step toward sustainable rural development.
Increasing the load consumption has a direct impact on reducing Renew Sustain Energy Rev 2013;28:456–62.
[12] Ahlborg H, Hammar L. Drivers and barriers to rural electrification in Tanzania
the NPC, the COE and the PV penetration, on the contrary the
and Mozambique-grid-extension, off-grid, and renewable energy technolo-
consumption of fuel, O&M costs and diesel contributions increase gies. Renew Energy 2014;61:117–24.
remarkably because the supplementary load is satisfied by DGs. The [13] Bezmalinovi D, Barbir F, Tolj I. Techno-economic analysis of PEM fuel cells role
in photovoltaic-based systems for the remote base stations. Int J Hydrogen
sensitivity analysis of the real interest rate has shown that it is
Energy 2013;38(1):417–25.
inversely related to the NPC of the HES and linearly related to the COE. [14] Lahimer AA, Alghoul MA, Yousif F, Razykov TM, Amin N, Sopian K. Research
and development aspects on decentralized electrification options for rural
household. Renew Sustain Energy Rev 2013;24:314–24.
Acknowledgments [15] Panapakidis IP, Sarafianos DN, Alexiadis MC. Comparative analysis of different
grid-independent hybrid power generation systems for a residential load.
The first author would like to thank Dr. M. Hatti (Unité de Renew Sustain Energy Rev 2012;16(1):551–63.
[16] Elamari K., Lopes L.A.C. Frequency based control of electric water heaters
développement des équipements solaires, Bousmail, Tipaza) for in small PV–diesel hybrid mini-grids. In: Proceedings of the Canadian conference
his help and for the valuable comments on the present work. on electrical and computer engineering (CCECE); 29 April–2 May 2012. p. 1–4.
1150 H. Rezzouk, A. Mellit / Renewable and Sustainable Energy Reviews 43 (2015) 1134–1150

[17] Pao H-T, Li Y-Y, Fu H-C. Clean energy, non-clean energy, and economic growth [35] Chong LI, Ge Xinfeng, Zheng Y, Chang X, Ren Y, Song C. Techno-economic
in the MIST countries. Energy Policy 2014;67:932–42. feasibility study of autonomous hybrid wind/PV/battery power system for a
[18] Sahu BK, Hiloidhari M, Baruah DC. Global trend in wind power with special household in Urumqi, China. Energy 2013;55:263–73.
focus on the top five wind power producing countries. Renew Sustain Energy [36] Diaf S, Diaf D, Belhamel M, Haddadi M, Louche A. A methodology for optimal
Rev 2013;19:348–59. sizing of autonomous hybrid PV/wind system. Energy Policy 2007;35
[19] L'énergie solaire: PV and concentré. 〈http://energie.cnrs.fr/IMG/pdf/SOLAIR (11):5708–18.
E_PV-C.pdf〉; 2014 [accessed 2014]. [37] Diaf S, Belhamel M, Haddadi M, Louche A. Technical and economic assessment
[20] Zyadin A, Halder P, Kähkönen T, Puhakka A. Challenges to renewable energy: of hybrid photovoltaic/wind system with battery storage in Corsica island.
bulletin of perceptions from international academic arena. Renew Energy Energy Policy 2008;36(2):743–54.
2014;69:82–8. [38] Saheb-Koussa D, Haddadi M, Belhamel M. Economic and technical study of a
[21] Pazheri FR, Othman MF, Malik NH. A review on global renewable electricity hybrid system (wind–photovoltaic–diesel) for rural electrification in Algeria.
scenario. Renew Sustain Energy Rev 2014;31:835–45. Appl. Energy 2009;86(7–8):1024–30.
[22] Ghoddami H, Delghavi MB, Yazdani A. An Integrated wind-photovoltaic- [39] Askarzadeh A. A discrete chaotic harmony search-based simulated annealing
battery system with reduced power–electronic interface and fast control for algorithm for optimum design of PV/wind hybrid system. Sol. Energy
grid-tied and off-grid applications. Renew Energy 2012;45:128–37. 2013;97:93–101.
[23] Tsai J-H, Chen S-J, Huang K-L, Lin W-Y, Lee W-J, Lin C-C, et al. Emissions from a [40] Rehman S, Md Mahbub Alam, Meyer JP, Al-Hadhrami LM. Feasibility study of a
generator fueled by blends of diesel, biodiesel, acetone, and isopropyl alcohol: wind–PV–diesel hybrid power system for a village. Renew Energy 2012;38
analyses of emitted PM, particulate carbon, and PAHs. Sci Total Environ (1):258–68.
2014;466–467:195–202. [41] 〈www.udes.dz〉.
[24] Rehman S, Al-Hadhrami LM. Study of a solar PV–diesel–battery hybrid power [42] Stambouli AB. Promotion of renewable energies in Algeria: strategies and
system for a remotely located population near Rafha, Saudi Arabia. Energy perspectives. Renew Sustain Energy Rev 2011;15(2):1169–81.
2010;35(12):4986–95. [43] Hatti M., Merzouk N.K., Mahrane A. Energetic hybrid systems for residential
[25] Bajpai P, Prakshan P, Kishore NK. Renewable hybrid stand-alone telecom use. In: Méndez-Villas A, editor. Materials and processes for energy: commu-
power system modeling and analysis. IEEE-TENCON 2009;23–26:1–6.
nicating current research and technological developments. 〈www.formatex.
[26] Lau KY, Yousof MFM, Arshad SNM, Anwari M, Yatim AHM. Performance
info/energymaterialsbook/chapters.html〉; 2013. p. 678–85.
analysis of hybrid photovoltaic/diesel energy system under Malaysian condi-
[44] Banque mondiale. 〈http://www.banquemondiale.org/〉; 2013.
tions. Energy 2010;35(8):3245–55.
[45] Global Energy & CO2 Data. Comprehensive and up-to-date online database for
[27] Deshmukh MK, Deshmukh SS. Modeling of hybrid renewable energy systems.
energy supply and demand and GHG emissions; 〈http://www.enerdata.net〉.
Renew Sustain Energy Rev 2008;12(1):235–49.
[46] Renewable energy and energy efficiency program. Ministry of Energy and
[28] Perera ATD, Attalage RA, Perera KKCK, Dassanayake VPC. Designing standalone
Mines; 〈http://www.mem-algeria.org〉 [accessed March 2011].
hybrid energy systems minimizing initial investment, life cycle cost and
[47] L'Algérie trace un programme ambitieux pour relever le défi des énergies
pollutant emission. Energy 2013;54:220–30.
renouvelables; 〈http://portail.cder.dz/spip.php?article1114/〉 [accessed April
[29] Johnson R.A., Agrawal A.N.,Chubb T.J. Simulink model for economic analysis
and environmental impacts of a PV with diesel–battery system for remote 2011].
villages. In: IEEE, power engineering society general meeting, vol. 2. Denver, [48] Gaz-oil et produits propres: enjeux et défis, ministère de l'énergie et des
CO; June 2004. p. 1898–905. mines, journée sur le gasoil, Algiers; 〈http://www.mem-algeria.org/〉 [accessed
[30] Shaahid SM, El-Amin I. Techno-economic evaluation of off-grid hybrid photo- July 2007].
voltaic–diesel–battery power systems for rural electrification in Saudi Arabia – [49] 〈www.nrel.gov/homer〉.
a way forward for sustainable development. Renew Sustain Energy Rev [50] Lambert T, Gilman P, Lilienthal P. Micropower system modeling with homer.
2009;13(3):625–33. USA: Wiley; http://dx.doi.org/10.1002/0471755621.ch15 (chapter 15).
[31] Khelif A, Talha A, Belhamel M, Hadj Arab A. Feasibility study of hybrid Diesel– [51] Getting Started Guide for HOMER Version 2.1; 〈https://www.homerenergy.
PV power plants in the southern of Algeria: case study on AFRA power plant. com〉 [accessed April 2005].
Electr Power Energy Syst 2012;43(1):546–53. [52] 〈http://power.larc.nasa.gov/〉 [accessed 04.03.13].
[32] Agarwal N, Kumar A, Varun. Optimization of grid independent hybrid PV– [53] German Technical Cooperation (GTZ). The 2012/13 International fuel prices
diesel–battery system for power generation in remote villages of Uttar report. The 2012/13 ed. Germany. 〈http://www.giz.de/expertise/downloads/
Pradesh, India. Energy Sustain Dev 2013;17(3):210–9. Fachexpertise/giz2013-en-ifp2013.pdf〉; April 2013. p. 05–6 [accessed April
[33] Dekker J, Nthontho M, Chowdhury S, Chowdhury SP. Economic analysis of PV/ 2013].
diesel hybrid power systems in different climatic zones of South Africa. Electr [54] Ngan MS, Tan CW. Assessment of economic viability for PV/wind/diesel hybrid
Power Energy Syst 2012;40(1):104–12. energy system in southern Peninsular Malaysia. Renew Sustain Energy Rev
[34] Adaramola MS, Paul SS, Oyewola OM. Assessment of decentralized hybrid PV 2012;16(1):634–47.
solar-diesel power system for applications in Northern part of Nigeria. Energy [55] How much is too much. Estimating greenhouse gas emissions. 〈http://www.
Sustain Dev 2014;19:72–82. scientificamerican.com/〉; 2013.