Energy Conversion and Management 88 (2014) 284–295

Contents lists available at ScienceDirect

Energy Conversion and Management

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

Analysis of hybrid energy systems for application in southern Ghana

Muyiwa S. Adaramola a,⇑, Martin Agelin-Chaab b, Samuel S. Paul c
a
Department of Ecology and Natural Resource Management, Faculty of Environmental Science and Technology, Norwegian University of Life Sciences, Ås, Norway
b
Department of Automotive, Mechanical and Manufacturing Engineering, University of Ontario Institute of Technology, Oshawa, ON, Canada
c
Sunrit Engineering and Consulting Services Inc., Winnipeg, Manitoba, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Due to advances in renewable energy technologies and increase in oil price, hybrid renewable energy
Received 12 June 2014 systems are becoming increasingly attractive for power generation applications in remote areas. This
Accepted 13 August 2014 paper presents an economic analysis of the feasibility of utilizing a hybrid energy system consisting of
Available online 7 September 2014
solar, wind and diesel generators for application in remote areas of southern Ghana using levelized cost
of electricity (LCOE) and net present cost of the system. The annual daily average solar global radiation at
Keywords: the selected site is 5.4 kW h/m2/day and the annual mean wind speed is 5.11 m/s. The National Renew-
Hybrid energy
able Energy Laboratory’s Hybrid Optimization Model for Electric Renewable (HOMER) software was
Energy cost
PV array
employed to carry out the present study. Both wind data and the actual load data have been used in
Wind turbine the simulation model. It was found that a PV array of 80 kW, a 100 kW wind turbine, two generators with
Ghana combined capacity of 100 kW, a 60 kW converter/inverter and a 60 Surrette 4KS25P battery produced a
mix of 791.1 MW h of electricity annually. The cost of electricity for this hybrid system is found to be
$0.281/kW h. Sensitivity analysis on the effect of changes in wind speed, solar global radiation and diesel
price on the optimal energy was investigated and the impact of solar PV price on the LCOE for a selected
hybrid energy system was also presented.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction has a per capita energy consumption of 4532 kW h [4]. Ghana’s

strategic goal is to dramatically increase its electricity generation
It is well known that access to energy is an essential pre-requi- from the 12,024 GW h in 2012 up to 32,915 GW h in 2020 [5].
site for development of any kind: human development; economic A convergence of factors such as global decline in fossil fuel
development, and social development. In fact, access to electricity reserves, damaging effects of global warming, and rising energy
is fundamental in advancing the well-being of any society and pro- demand due to increasing population are forcing a shift to low-car-
moting economic growth as well as employment opportunities for bon sources of energy. As a tropical country Ghana has abundant
young people [1]. solar energy resources. However, until recently the country did
Like most countries in sub-Saharan Africa, access to electricity not produce significant electricity from solar energy. In 2013, it
in Ghana is low compared to more advanced countries. The coun- completed a 2 MW solar power plant, which is the largest in main-
try’s electricity supply is primarily from hydro and thermal power land West Africa [6]. Although, a conservative estimate suggests
plants. It has a total installed electricity generation capacity of that over 1000 km2 of land area exists with moderate-to-excellent
2280 MW in 2012 [2]. This is made up of 1180 MW from hydro wind resource potential in Ghana [7], the country is not considered
power (52%) and 1100 MW from thermal power (48%). In the a player in the wind energy sector. In fact, it does not currently pro-
same year (i.e., 2012) the electricity consumed was 9258 GW h, duce any significant amount of wind power. Nonetheless, there are
out of which 2931 GW h constituted residential use (32%). Unfor- plans to generate additional 150 MW from wind power and 14 MW
tunately, there still remains close to 28% of households that do not from solar energy by 2015 [6].
have access to electricity in Ghana [3]. With its current per capita The problem with solar and wind energy sources is that they are
electricity consumption of only 357.5 kW h [3], Ghana is consid- unpredictable and can be unreliable. A stand-alone solar energy
ered energy poor. To put that figure in perspective, South Africa system cannot provide electricity around the clock throughout
the year if there are cloudy days when there is no sunlight. Similarly
a stand-alone wind energy system may not produce usable energy
⇑ Corresponding author. Tel.: +47 6496 5793; fax: +47 6496 5801. for considerable portion of time during the year due to relatively
E-mail address: muyiwa.adaramola@nmbu.no (M.S. Adaramola). high cut-in wind speed [8]. One way to remove or minimize the

http://dx.doi.org/10.1016/j.enconman.2014.08.029
0196-8904/Ó 2014 Elsevier Ltd. All rights reserved.
 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295 285

weaknesses of these renewable energy systems is through the use 400

Energy consumption/capita
of hybrid energy systems, which employ two or more complemen- 350
tary sources of energy. For example, a diesel conventional generator 300
can be combined with a wind energy system or a solar energy 250

(kWh)
system or both. Feasibility, reliability and economic analyses con- 200
ducted in a number of studies showed that hybrid power systems 150
are more reliable and cheaper than single source energy systems 100
[9–12]. In fact, a number of studies on renewable hybrid energy sys- 50
tems have been performed in different parts of the world. For exam- 0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
ple, a hybrid power system based on wind, diesel and battery was
modelled by Nfah and Ngundam [13] for power supply to remote Fig. 1. Electricity consumption per capita in Ghana from 2001 to 2012 [3].
areas in Cameroon using 4 years wind data for validation. The
results indicate that hybrid systems are a better approach to
increase access to electricity in the remote areas of Cameroon with- is 4.4 [19] and hence, the average electricity consumption per
out expensive grid extension. Later, a method for optimal design of household can be determined as 4.31 kW h per day. For the analy-
small grid-connected hybrid energy systems consisting of solar and sis carried out in this study, each household is therefore assumed
wind sources in Chile was reported by Caballero and Yanine [14]. to require electrical load of 5 kW h/day.
The objective was to minimize the life cycle cost of the system with- Technical and economic assessment of a hybrid energy plant for
out compromising reliability using simulation. The results indicate a mini-grid system is simulated for a small community that con-
that the hybrid system minimizes the cost of electrical power and sists of 400 households in this study. This community would
size of grid. Yanine et al. [15] also reported a technique for the con- require a power plant that can produce 2000 kW h of electricity
trol of grid-connected hybrid micro-generation systems without per day with a peak load of about 83 kW. It is expected that there
energy storage in Chile. Their theoretical framework based on will be daily, monthly, seasonal variation in electricity consump-
homeostatic control principles was assessed via a simulation anal- tion pattern in this community. To account for these variations,
ysis. The results confirmed the study’s hypothesis that sustainable hourly and daily variations in the electrical loads are assumed to
energy systems are completely achievable. In addition, Ismail be 10%.
et al. [16] performed a techno-economic analysis and designed a
hybrid system consisting of photovoltaic panels, a battery system
and a diesel generator for a typical Malaysian village household. 2.2. Solar radiation data
The results showed that the hybrid system developed was optimal.
Furthermore, an economic feasibility analysis of investments in The solar radiation data for the selected location were obtained
hybrid systems in different climatic zones of Tamil Nadu (India) from the Ghana Energy Commission. The monthly daily averaged
based on net present cost was conducted by Kumar and Manoharan global solar radiation is presented in Fig. 2. As expected, monthly
[17]. It was observed that government subsidies hold the key to and seasonal variations in global solar radiation and the monthly
attracting investments in renewable energy utilization in these energy output from solar energy conversion systems would vary
zones because renewable energy systems are initially capital inten- from one month to another as well as from one season to another.
sive in nature. More recently, Adaramola et al. [18] reported an The annual averaged daily global solar radiation is 5.41 kW h/m2/
assessment of the wind energy potential in the coastal regions of day. The monthly clearness index, which is defined as the fraction
Ghana. However, there are no analyses of hybrid energy systems of solar radiation at the top of the atmosphere that reaches a par-
for Ghana in the open literature. ticular location on the earth surface varied between 0.495 (in
The objective of this article is to study an economic analysis of a August) and 0.586 (in October) with an annual average of 0.543.
hybrid energy system consisting of solar, wind and conventional The clearness index provides information on the level of availabil-
diesel generators for application in rural areas of southern Ghana. ity of solar radiation at the surface of the earth as well as changes
It is believed that this information will broaden the scope of in atmospheric conditions [20,21] and the weather condition at a
options available to policy makers and all stakeholders in the site. Based on the range of values of the clearness index, the
energy sector as the country seeks to make critical investments weather conditions in Adafoah can be classified as partly overcast.
in energy. A range of values of global solar radiation are used to investigate
the effect of changes in the solar radiation on the energy system

2. Load and energy resources

The renewable resource data used in this study are for Adafoah Global Horizontal Radiation
6 1.0
in the Greater Accra region of Ghana. This site is located on latitude
Daily Radiation (kWh/m²/d)

5°470 N and longitude 0°380 W and at an elevation of about 2 m 5

0.8
above sea level. The solar energy and wind energy resources at
Clearness Index

4
the selected site as well as the cost of diesel (to fuel the generator) 0.6
and electrical loads are presented in this section.
3
0.4
2
2.1. Electrical load
0.2
1
The electricity consumed per capita in Ghana from 2001 to 2012
is presented in Fig. 1. The figure shows that there is general 0 0.0
increase in annual electricity consumption per capita between Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Daily Radiation Clearness Index
2007 (288.8 kW h/capita) and 2012 (357.5 kW h/capita). The
amount of electricity consumed per day per capita in 2012 can Fig. 2. Monthly daily averaged global solar radiation and clearness index for
be estimated as 0.9795 kW h. The average household size in Ghana Adafoah [22].
286 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295

Wind Resource turbine located in this site would also vary from one month to
6
another. Detailed analysis of the wind energy resource in Adafoah
Wind Speed (m/s)

5
(Ghana) can be found in [18,22].
4
3
The daily hourly (diurnal) variation of the wind speed in the site
2
is shown in Fig. 4. It can be observed that the hour of peak wind
1
speed (the windiest hour) is 16.00 h when the wind speed is
0 6.4 m/s. The autocorrelation factor (randomness in wind speed)
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec in this site is taken as 0.85 (typical values are within the range of
0.80–0.95 [23,24] and the diurnal pattern strength (wind speed
Fig. 3. Monthly mean wind speed at Adafoah measured at a height of 12 m a.g.l.
[22]. variation over a day) is estimated as 0.25 (typical values are within
0.0–0.40 [23,24]. A range of values of mean wind speed are used to
investigate the effect of changes in the wind speed on the energy
system configuration and performance.
7.0
6.0
Wind speed (m/s)

5.0
2.4. Diesel fuel
4.0
The diesel price in Ghana in January 2014 (at the time of carry-
3.0
ing out this study) was about GH¢2.49 (US$0.95) per liter [25].
2.0
Since Ghana is a net importer of petroleum/crude oil products,
1.0 the price of diesel fuel in Ghana is vulnerable to the world prices.
0.0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
In order to investigate the effect of changes in fuel price on the
Hours of Day hybrid energy system and cost of electricity, the diesel price was
varied between US$0.8 and US$1.3/L, at an interval of US$0.10.
Fig. 4. Hourly variations of wind speed at Adafoah measured at 12 m a.g.l. [22]. The highest price of US$1.30/L may represent the possible price
of diesel in remote areas (without government subsidies) or a
as well as on the technical and economic aspects of the hybrid sys- sudden spike in fuel prices.
tem considered in this study.

3. Power plant components

2.3. Wind speed
The hybrid solar PV–wind–diesel generator energy system (also
The wind speed data for Adafoah was also obtained from the called PV–wind–Gen hybrid) consists of two parts: (1) power plant
Ghana Energy Commission. The wind speed was measured at a which is made up of a PV module, a wind turbine, a diesel genera-
height of 12 m (above ground level) on hourly basis over a period tor, a battery and a power converter; and (2) a mini-grid transmis-
for 13 months (July 1999 to August 2000). From these hourly wind sion and distribution system. The schematic diagram of the power
data, the monthly wind speed and other wind speed parameters plant system is shown in Fig. 5.
were determined. The Weibull shape and scale factors were deter- The hybrid energy system is designed and analyzed using
mined as 2.0 and 5.98 m/s, respectively. The monthly mean wind National Renewable Energy Laboratory software, Hybrid Optimiza-
speed in Adafoah is presented in Fig. 3. Similar to the solar data, tion Model for Electric Renewable (HOMER). Among the available
variations in the monthly average wind speed can be observed softwares, HOMER is the most-used optimization software for
from this figure and the monthly energy output from a wind hybrid systems [26]. HOMER performs three basic calculations:

Fig. 5. Schematic diagram of proposed PV–wind–Gen–Battery hybrid power plant.

 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295 287

simulation, optimization, and sensitivity analysis [27]. In the sim- Using the monthly average global solar radiation and the lati-
ulation process, HOMER models the performance of a particular tude of the site, HOMER used an algorithm developed by Graham
hybrid system configuration each hour of the year to determine and Hollands [30] to generate hourly global solar radiation on
its technical feasibility and life-cycle cost. A feasible system is the horizontal surface data. Since the PV array is inclined at angle
defined as a hybrid system configuration that is capable of satisfy- b (in this study, b = 5.78°, which is the latitude of the site) relative
ing the required load and meeting other constraints imposed on to the horizontal, HOMER determine the hourly global solar radia-
the system. In the optimization process, several different system tion (GT ) incident on the PV array using the Hay, Davies Klucher,
configurations are simulated and HOMER searches for the system Reindl (HDKR) model (Duffie and Beckmann [28]). This model
that satisfies the technical and economic constraints at the lowest takes into account the global solar radiation incident on a horizon-
life-cycle cost. In the sensitivity analysis process, HOMER performs tal surface, the orientation of the PV array, the location of the site
multiple optimizations under a range of input values in order to on the Earth’s surface, the time of year, and the time of day [27].
investigate the effects of uncertainty or changes in the model The lifetime of the PV array is assumed to be 25 years. The cost
inputs (such as changes in wind speed or diesel price). of the PV module in the Ghanaian market is GH¢3.5/W or approx-
For each component of the energy system, the HOMER software imately US$1500/kW [32] at the time of this study. For this simu-
requires information about the cost (capital, replacement, opera- lation analysis, the total initial cost for the PV installation is taken
tion and maintenance), number (or size) of units to be used, oper- as $3000/kW. The derating factor and the ground reflectance are
ating hours and lifetime, and other specific component properties. taken as 0.80% and 20%, respectively. There is no tracking system
In addition, the economic information (such as applicable real included in the PV system. This case represents the worst perfor-
interest rate at a desire location and the overall system fixed, oper- mance of the PV system. Due to possible changes in the cost of
ating and maintenance costs) is required. In this analysis, the sys- the PV array and its implication on economic performance of the
tem’s fixed cost will mainly come from the mini-grid system. hybrid energy system the cost of electricity was calculated for
Detailed descriptions of each component with the required input the following PV cost: $2250/kW, $3000/kW, $3750/kW, $4500/
data are provided and discussed in the following sub-sections. kW, $5250/kW, and $6000/kW.

3.1. Solar PV modules 3.2. Wind turbine

The solar PV array with overall capacity of 80 kW is used in this In this study, the Furhlander FL 100 model wind turbine is
study. The solar PV module is a 72-cell (6  12) poly-crystalline selected. This model is designed to be used at hub height of up
(model number CS6X-300M, manufactured by Canadian Solar) to 38 m high. The detailed technical specifications and power curve
with a rated power of 300 Wp and can produce a maximum of for this wind turbine are presented in Table 2 and Fig. 6, respec-
600 V DC. The detailed technical specifications of this model are tively. The capital cost of this wind turbine model is taken as
presented in Table 1. The output power (Poutput) of a PV array can US$380,000 [33]. The operating and maintenance costs are
be calculated from the following equation and the PV specifications assumed to be 10% of the annual cost of the initial capital cost of
[28,29]. the system (system cost/lifetime). The lifetime of the wind turbine
! is assumed to be 25 years. In addition, the replacement cost of the
GT
Poutput ¼ Y PV f PV ½1 þ ap ðT c  T c;STC Þ; ð1Þ wind turbine is taken as $0 in this study, since the project life is
GT;STC
assumed to be 25 years.
where YPV is the rated capacity of the PV array or the power output In most cases, the available wind data are measured at a height
under standard test conditions in kW, fPV is the PV derating factor that is different from the wind turbine hub-height and since the
(%), GT is the solar radiation incident on the PV array (kW/m2), wind speed at the hub height is of interest for wind power applica-
GT,STC is the incident radiation at standard test conditions (1 kW/ tions, the available wind speeds are adjusted to the wind turbine
m2), ap is the temperature coefficient of power (%/°C), Tc is the PV hub height using the following power law expression:
cell temperature (°C) and Tc,STC is the PV cell temperature under  a
U h
standard test conditions (25 °C). In a case where the effect of ¼ ; ð3Þ
UO h0
temperature on the PV array performance is neglected, ap can be
assumed to be zero and Eq. (1) reduces to: where Uo is the wind speed at a reference height ho, U is the wind
! speed at power height h and a is the surface roughness (in most
GT
Poutput ¼ Y PV f PV ð2Þ cases, a is taken as 0.14 [18]).
GT;STC The power output from a wind turbine is calculated by HOMER
as follows: (i) determine the average wind speed for a given hour
Table 1 at the anemometer a height the monthly wind speed; (ii) calculate
Technical specification of MaxPower CS6X-300P under nominal operating cell
temperature [31].
the corresponding wind speed at the turbine’s hub height using the
power law (Eq. (3)); (iii) use the power curve (see Fig. 6) to calcu-
Parameter Specification late wind turbine power output at the determined wind speed
Nominal maximum power 300 Wp assuming standard air density; and (iv) multiply the calculated
Optimum operating voltage 32.9 V
Optimum operating current 6.61 A
Open circuit voltage 41.0 V Table 2
Short circuit current 7.19 A Technical specifications for Fuhrländer FL100 wind turbine model [34].
Efficiency 14.71%
Operating temperature 40 °C  +85 °C Rated power 100 kW
Temperature coefficient: Rotor diameter 21 m
Pmax 0.43%/°C Hub height 38 m
Voc 0.34%/°C Cut-in wind speed 2.5 m/s
Isc 0.065%/°C Rated wind speed 13 m/s
Normal operating cell temperature 45 ± 2 °C Cut-off wind speed 25 m/s
Dimension 1954 mm  982 mm  40 mm Maximum design wind speed 67.0 m/s
288 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295

140

120

100
Power (kW)

80

60

40

20

0
0 5 10 15 20 25
Wind speed (m/s)

Fig. 6. Power curve for the Fuhrländer wind turbine model FL 100 [34].

power output value by the air density ratio. The density ratio is
given as [29]:
 RBg  
q Bz To
¼ 1 ; ð4Þ
qo To T o  Bz
where qo is the air density at standard temperature and pressure
conditions (1.225 kg/m3), To is the standard temperature
(288.16 K), B is the lapse rate (0.00650 K/m) and R is the gas con-
stant (287 J/kg K), z is the elevation (m), g is the acceleration due
to gravity (9.81 m/s2) and q is the air density at the wind turbine
hub height.

3.3. Diesel generator Fig. 7. The generator curves: (a) fuel consumption and (b) efficiency [36].

The generator system is used to supplement the power produc-

turbine or diesel generator). In this study, the Surrette 4KS25P
tion by the renewable energy conversion systems especially when
model with a nominal capacity at 100-h rate of 1904 ampere hours
the required electrical load is not fully provided by these systems.
(A h) and cell voltage of 4 V is selected. The characteristics of Sur-
In general, the diesel generator is sized to meet the peak electrical
rette 4KS25P are presented in Table 3. The capital and replacement
load demand. In this study, the peak load is 83 kW. Therefore, two
cost of one battery is taken as $1230 [37]. In order for the battery to
diesel Perkins generators with combined rated power of 100 kW
produce more power, batteries are generally connected in series to
(50 kW each) are selected. The fuel consumption and efficiency
form a battery string that consists of 12 batteries. Each battery
curves for these generators are shown in Fig. 7. The fuel consump-
string can produce about 91.2 kW h of electricity.
tion from these generators can be estimated from the following
The HOMER can model two battery dispatch strategies which
expression [29]:
are load-following and cycle-charging. Under the load-following
F c ¼ aPrated þ bP gen ; ð5Þ strategy, renewable power sources charge the battery while the
generators produce enough power to meet the load demand. Under
where Fc = generator fuel consumption (L), a = generator fuel curve
the cycle-charging strategy, whenever the generators operate, they
intercept coefficient (L/h/kW), b = generator fuel curve slope (L/h/
produce more power than required to serve the load so that the
kW), Prated = generator rated capacity (kW) and Pgen = generator
surplus electricity charges the battery bank. In this analysis,
power output (kW). From Fig. 7a, the values of a and b are deter-
cycle-charging strategy is selected because this strategy can pre-
mined as 0.00626 and 0.2831, respectively.
serve the lifetime of both the diesel generator and battery thereby
The excess power of 17 kW can be used to cover the spinning
reducing the risk of over-discharging the battery [24].
and/or reserve power as a result of changes in electrical load. In
addition, this may ensure constant electricity supply when there
is a sudden increase in electrical load or sudden decrease in power 3.5. Converter/inverter
generated by the PV and wind turbine. The cost of a generator in
the Ghanaian market depends on its size and model. The costs of Since the hybrid energy system comprises both AC and DC sys-
common medium to large size generators found in Ghana are: tems, a power converter system is required. A power converter
US$15,500 for 20 kW, US$28,000 for 100 kW and US$54,000 for maintains the flow of energy between the AC electrical load and
200 kW [35]. For this analysis, the combined cost of a 100 kW gen- DC components of the hybrid energy system. For simulation analysis
erator is used. The replacement cost is assumed to be the same as
the initial capital cost for the generators while the operational and Table 3
maintenance costs are assumed to be $0.01/h for these generators. Technical specifications for Surrette 4KS25P battery [36].
The operating lifetime of each generator is taken as 15,000 h. To
Nominal capacity @ 100-h rate 1904 A h
reduce the capacity of the inverter, these generators are operated Nominal voltage 4V
as AC power systems. Round trip efficiency 80%
Minimum state of charge 40%
Float life 12 years
3.4. Battery
Lifetime throughput 10,569 kW h
Maximum charge rate 1 A/A h
The battery is used to meet the electrical load during the non- Maximum charge current 67.5 A
availability of power from the energy generating systems (PV, wind
 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295 289

therefore, different sizes of converter capacity (60 kW, 80 kW, 3.8. Economic parameters and calculations
100 kW, 120 kW, and 140 kW) are considered. The lifetime of a unit
is taken as 15 years and efficiency of 90%. In Ghana, the cost of 1 kW The lifetime of the project is taken to be 25 years. According to
converter is GH¢2.1/W [32] or approximately US$900/kW and the the available information, the current interest rate and inflation
replacement cost is taken as US$800/kW. rate in Ghana are 16% and 13.5% respectively [42] and from these
rates, the annual real interest rate is determined as 2.2% using
Fisher expression [43]. In addition to the cost of mini-grid installa-
3.6. Mini-grid system tion, other initial fixed capital costs which can be used to prepare
the site for the system and other related activities are taken as
The cost of a mini-grid depends on the electricity distribution $50,000 and the overall hybrid power system operation and main-
and voltage levels as well as spatial distribution and number of tenance cost of $1500 per year is assumed.
the households within the mini-grid area. In this analysis, the Based on all the cost data for each system component, the dis-
mini-grid system consists of: (1) medium-voltage lines (33 kV), count rate and the project economic lifetime, the optimized system
which represent the main ‘transmission line’ and originated from configurations are ranked based on the minimum value of total net
the hybrid power plant; (2) low-voltage lines (240/415 V), which present cost. The revenues from the system include income from
serve as distribution lines (linking the medium-voltage lines to selling power to the grid and any salvage value that occurs at the
the households) and, (3) the households which represent the load end of the project lifetime. The annualized investment and associ-
points. From a recent study [38], the capital cost for low-voltage ated cost of each component is determined using:
lines in Ghana is given as US$17.0 per meter, and that of medium C a;comp ¼ C comp  CRFði; nÞ; ð6Þ
voltage lines is US$25.0 per meter. These amounts represent the
overall cost for three-phase lines, poles, transformers and labor. where Ccomp is the total investment cost, i is the interest rate (in
From the data provided by Kemausuor et al., [38] the connection fraction) and n is the payment period (in years). The capital recov-
cost (equipment, installation and operation and maintenance cost) ery factor (CRF) is given as:
per household is estimated as US$323. The total length of the med- ið1 þ iÞ
n

ium-voltage lines and low-voltage lines are taken as 1600 m and CRFði; nÞ ¼ n ð7Þ
ð1 þ iÞ  1
10,000 m (25 m per household), respectively. Therefore, the overall
capital cost of the mini-grid is estimated as US$338,000 and the The NPC is then determined from [29]:
operation and maintenance cost of the grid is taken as 3% of the
C a;total
capital cost of the mini-grid. C NPC ¼ ; ð8Þ
CRFði; nÞ
where CNPC is the total NPC and Ca,total is the total annualized cost.
3.7. System simulation: operating reserve
4. Results and discussion
During the simulation process, HOMER ensures that the sys-
tem’s operating capacity is sufficient to provide for both the pri- In this section, the results of the technical and economic perfor-
mary load and operating reserve. Operating reserve is an excess mance of the mini-grid hybrid power plant are presented and dis-
operating capacity that should be able to respond to a sudden cussed. The results are grouped into two sub-sections: (1) the
increase in electrical load and variability in the renewable power overall performance of the energy system (Section 4.1) and, (2)
output from the wind turbine and PV system. In HOMER, three performance of the renewable energy conversion components of
operating reserves are required: one relates to the variability of the hybrid power plant (Section 4.2).
electrical load and the other two relate to variability of wind speed
and solar radiation. In this study, operating reserve taken as a per- 4.1. Overall energy performance
centage of hourly load is 10% as recommended by Cotrell [39] and
Dalton et al. [40]. The total required operating reserve for a given Based on the energy and economic parameters as well as
time is then determined as the sum of the product of each assumed assumptions used in this study, the simulation analysis produced
operating reserve inputs and the system power output at this given six sets of power plant systems. The categorized optimization solu-
time. But if the simulation cannot generate sufficient power to tions for the selected location with annual daily averaged global
cover both the determined operating reserve and electrical load, solar radiation of 5.4 kW h/m2/day, an annual mean wind speed
the HOMER shows the shortfall as ‘‘capacity shortage’. In this of 5.11 m/s, and a diesel fuel price of $0.95/L are presented in
study, the maximum allowable annual capacity shortage is taken Table 4. Based on the NPC and cost of electricity (COE), it can be
as 3% which is within the range of 0.5–5% suggested by Givler observed from this table that PV–wind–Gen hybrid (with and
and Lilienthal [41]. Any simulated system whose capacity shortage without battery) power plants are the most economically suitable
fraction exceeds this value is considered as infeasible system. solution while PV–Gen power plant is found to be the least suitable

Table 4
Categorized optimized energy systems for global solar radiation of 5.4 kW h/m2/day, wind speed of 5.11 m/s, PV cost of $3000/kW and diesel price of $0.95/L.

PV FL100 Gen Battery Converter Initial capital Operating cost (1000$/ Total NPC COE ($/ RFa Diesel
(kW) (unit) (kW) (unit) (kW) (1000$) year) (1000$) kW h) (–) (1000 L)
80 1 100 50 1081.0 144.7 3840.5 0.276 0.47 124.0
80 1 100 60 60 1163.8 143.8 3905.6 0.281 0.47 121.2
1 100 796.0 168.2 4004.8 0.288 0.33 148.1
1 100 60 20 887.8 168.9 4109.4 0.295 0.33 146.3
80 100 60 710.0 200.1 4525.4 0.325 0.17 179.6
80 100 60 60 783.8 202.3 4642.5 0.333 0.17 179.4
a
RF: Renewable fraction.
290 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295

Table 5 reported that in general (especially in remote communities), a

Electricity generated and consumed by the power system components. renewable energy conversion system can contribute more than
Production Consumption 75% of the total electricity produced by renewable energy-diesel
MW h/ % MW h/ % generator hybrid power system [44]. However, this high proportion
year year may make the system challenging to control while maintaining sta-
PV array 126.92 16.04 Electrical load 729.98 92.27
ble voltage and frequency [45]. Therefore, it is common to limit the
Wind turbine 248.24 31.38 Excess 49.60 6.27 contribution of the renewable energy system to about 11–25% [46].
electricity On the other hand, the relative large contribution of renewable
Generator 1 307.31 38.84 Losses 27.63 1.46 resource can reduce the dependence of the hybrid power system
Generator 2 108.68 13.74
on diesel price, lower the operating and maintenance as well as fuel
Total 791.14 100.0 Total 791.14 100.0 costs [24]. Therefore, the 47.4% contribution of the PV system to the
total energy generated by the hybrid power system observed in this
study is a good compromise for a community that possesses good
renewable energy resources but has high diesel prices.
option. The renewable energy contribution to the respective
The monthly average of electricity produced by each power
energy system varies from 47% (for PV–wind–Gen system, the
generating unit of the hybrid power plant is presented in Fig. 8.
most viable system) to 17% (in the case of PV–Gen system). As
This figure indicates that electricity produced by each component
can be observed in Table 4, the performance of the PV–wind–Gen
varies from one component to another and from month to month.
system and PV–wind–Gen–Battery system is similar. The LCOE
This is due mainly to the variability in monthly global solar radia-
for PV–wind–Gen system is $0.276/kW h and that of PV–wind–
tion and mean wind speed. The minimum combined average
Gen–Battery system is $0.281/kW h. However, the rest of the dis-
monthly power generated by the PV and wind turbine is
cussion of the study will be focused on the PV–wind–Gen–Battery
35.87 kW (in January) while the maximum power output is
system for the following reasons: (1) the battery can serve as back-
51.70 kW (in October). The combined power generated by the gen-
up system to keep the essential and critical load running during an
erators varies between 42.06 kW (in September, and 42.17 kW in
emergency and (2) using the battery can also reduce the operating
October) and 52.81 kW (in January), when the renewable energy
period of the generators and hence, less operating and mainte-
conversion systems produced the least amount of power.
nance cost for the system. Most importantly, there is less emission
of carbon dioxide and other pollutants from the power plant
system. 4.1.2. Economic analysis
The economic analysis of the hybrid energy system is assessed
4.1.1. Electricity generation by the LCOE and NPC of the system. The breakdown of the cost
The electricity generated by the PV–wind–Gen–Battery hybrid analysis for the PV–wind–Gen–Battery energy system with a wind
power plant (for the base case with a wind speed of 5.11 m/s and speed of 5.11 m/s, global solar radiation of 5.4 kW h/m2/day, diesel
global solar radiation of 5.4 kW h/m2/day) is presented in Table 5. fuel price of $0.95/L and PV price of $3000/kW are shown in Table 6.
It can be observed that the total electricity produced by the power The total initial capital of the power plant and the mini-grid system
plant is 791.14 MW h/year which comprises 126.92 MW h/year is estimated to be $1,163,800 while the total NPC is $3,905,581. For
(16.04%) from the solar PV; 248.24 MW h/year (31.38%) from the a community with existing grid system but lacks regular electricity
wind turbine and; 415.99 MW h/year (52.58%) from the two diesel as a result of damage to exiting power plant, the initial capital can
generators. The generated electricity is utilized as follows: 92.27% be significantly lower than this value. For instance, if mini-grid is
(729.98 MW h/year) is used to satisfy the electrical loads of the modeled in this analysis, the initial capital will then be $743,000.
households, 6.27% (49.60 MW h/year) represents the excess gener- This can have positive impact on the cost of electricity produced
ated electricity and 1.46% accounted for losses in the power con- by the hybrid energy system.
version system (converter/inverter losses and losses in battery Using the information presented in Table 6, the distribution of
system). Hence, this hybrid power plant which consists of 80 kW the initial capital and NPC for each component is shown in Fig. 9.
solar PV, 100 kW wind turbine, 100 kW generator(s), 60 kW con- The PV array accounted for 20.62% of the initial capital while the
verter/inverter and 60 Surrette 4KS25P batteries can produce more wind turbine accounted for 32.65%. The mini-grid system and
than enough electricity to meet the required electrical load. In other initial costs are made up 33.34% with the battery, converter
addition, the excess electricity can be used to take care of any and generators accounting for about 13.39%. Therefore, for this
deferrable load (like water pumping and community center build- hybrid system, the renewable energy conversion systems (RECS)
ing) and any other unexpected additional electrical load. comprise half of the total capital while the generators only made
It can further be noted from Table 5 that the combined contribu- up 2.40%. This indicates that the cost of the RECS is the most
tion of electricity produced by the wind turbine and PV is 47.4% of significant component of the hybrid power system. This figure
the total electricity produced by the hybrid power system. It was also shows that almost 86% of the replacement, operating and

Fig. 8. Monthly distribution of the electricity produced by the hybrid energy system.
 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295 291

Table 6
Cost summary for PV–wind–Gen–Battery system for a case of $3000/kW PV.

Component Capital (1000$) Replacement (1000$) O&M (1000$) Fuel (1000$) Salvage (1000$) Total NPC (1000$)
PV array 240.0 0 0 0 0 240.0
Wind turbine 380.0 0 29.0 0 0 409.0
Generator 1 14.0 67.1 66.3 1615.7 2.0 1761.0
Generator 2 14.0 40.4 39.0 580.7 5.7 668.5
Battery 73.8 90.0 9.2 0 35.1 137.8
Converter 54.0 34.6 0 0 9.3 79.3
Minigrid/other initial costs 388.0 0 222.0 0 0 610.0
System 1163.8 232.1 365.5 2196.4 52.1 3905.6

number of years at which the cumulative cash flow of the differ-

50.0 ence between the current system and base case system switches
from negative to positive. That is, the payback indicates how long
Cost compsotion (%)

40.0 Capital cost it would take to recover the difference in investment costs
30.0
Net present cost (NPC) between the current system and the base case system. The lower
the simple payback period, the more desirable it is to undertake
20.0 the project. The simple payback varies between 6.91 years for the
case when PV–Gen–Battery is the base system and 16.80 years
10.0
when the wind–Gen is the base system.
0.0
PV array Wind turbine Generator 1 Generator 2 Battery Converter Minigrid/other
intial costs 4.2. Renewable energy conversion systems performance
System component

Fig. 9. Relative composition of the capital and net present costs. 4.2.1. Wind turbine
The performance parameters of the model FL100 wind turbine
Table 7
are indicated in Table 8. The capacity factor, which is defined as
Economic performance indicators for PV–wind–Gen–Battery versus selected systems. the ratio of an actual energy output during a given period to the
energy output that would have been generated if the system were
Parameters Wind– Wind–Gen– PV– PV–Gen–
Gen Battery Gen Battery
operated at full capacity for the entire period, is given as 28.3%. As
suggested by Mathew [47], for an investment in wind power to be
Present worth (1000$) 99.26 203.86 618.86 736.96
Internal rate of return 4.41 7.73 11.80 15.00
economically viable, wind turbine with a capacity factor greater
(%) than 0.25 in a site is generally recommended and therefore, the
Simple payback (years) 16.8 10.80 7.89 6.91 selected wind turbine (FL100 model) is a good choice for the site
considered in this study and other similar sites in southern Ghana.
The capacity factor is a useful performance indicator for users,
developers and manufacturers of energy systems. It determines
maintenance, and fuel costs were accounted for by the generators
the economic viability of the system and higher value of the capac-
and also about 63% of the total NPC is accounted for by the two
ity factor is desired. Using life cycle cost analysis, the LCOE gener-
generators. The high NPC for the generators is due to the cost of
ated by the wind turbine was determined as $0.086/kW h.
fuelling, replacing and maintaining them. Therefore, any govern-
ment policy that aid decreasing the cost of diesel fuel can reduce
4.2.2. Solar PV
the operating cost of hybrid power system in remote and semi-
The basic performance parameters of the solar PV are indicated
urban area of the country.
in Table 9. The solar PV capacity factor is estimated as 18.1%. It
The economic performance of the PV–wind-generator–battery
should be mentioned that for the analysis carried out in this study,
power system is compared with other selected optimized power
fixed tilt orientation is used and the determined capacity factor of
systems (see Table 4) using the present worth, internal rate of
the PV system can be improved by using a tracking system. The
return and simple payback period parameters. For the purpose of
comparison, the selected PV–wind–Gen–Battery hybrid power is Table 8
used as the current system while other possible alternative power Performance parameters for the model FL100 wind turbine.
systems are used as base case systems. The results of this compar- Parameters Quantity
ison are shown in Table 7. The present worth is the difference
Rated capacity (kW) 100
between the NPCs of the base case system and the current system. Mean output (kW) 28
The sign of the present worth indicates whether the current system Capacity factor (%) 28.3
compares favourably as an investment option with the base case Total production (MW h/year) 248.24
system: a positive value indicates that the current system saves Levelized cost ($/kW h) 0.086
money over the project lifetime compared to the base case system.
From Table 7, it can be observed that the present worth values are
all positive when the PV–wind–Gen–Battery hybrid system is com- Table 9
pared with other systems. Performance parameters for the PV system.
The internal rate of return (IRR) represents the discount rate at Parameters Quantity
which the base case and current system have the same NPC. The
Rated capacity (kW) 80
IRR can be used to evaluate the desirability of investments or pro-
Mean output (kW) 14.5
jects. The higher the IRR, the more desirable it is to undertake the Mean output (kW h/day) 348
project. The IRR varies between 4.41% when the PV–wind– Capacity factor (%) 18.1
Gen–Battery is the base system and 15.00% for the case when Total production (MW h/year) 126.92
Levelized cost ($/kW h) 0.099
PV–Gen–Battery is the base system. The simple payback is the
292 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295

Fig. 10. Optimal energy system type for PV array cost of $3000/kW and diesel price of $0.95/L.

Fig. 11. Optimal energy system type for PV array cost of $3000/kW and wind speed of 5.11 m/s.

value of a capacity factor for the PV determined in this study is influence composition of the optimal hybrid power system, are
comparable to findings from similar studies (e.g., [48]). It can fur- presented in graphical form in Figs. 10–12 for the solar PV cost
ther be observed that the average daily energy output from the of $3000/kW. An optimal system is defined as a solution for energy
solar PV system is observed to be 348 kW h/day. The LCOE gener- system configuration that is able to meet the specified electrical
ated by the wind turbine is determined as $0.099/kW h. load and other imposed constraints. Wind energy and solar energy
are variable natural resources and hence, hourly, daily, monthly
4.3. Sensitivity analysis and annual variations in wind speed and global solar radiation
are generally expected. Government policies and fluctuation in glo-
The effect of changes in the global solar radiation, wind speed bal oil prices can impact local diesel fuel prices. The graphical rep-
and diesel fuel cost, which are the main parameters that can easily resentation of optimal system type provides visual information
 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295 293

Fig. 12. Optimal energy system type for PV array cost of $3000/kW and global solar radiation of 5.4 kW h/m2/day.

that a particular system will be optimal at any given two sets of Furthermore, it can be noted that no feasible energy system is
variable parameters. The cost of electricity per kW h is superim- found at low wind speed and global solar radiation (lower left cor-
posed on each figure. The most significant advantage of the figure ner region of Fig. 10). The choice of the PV–Gen system is an opti-
is that it can be used in the selection of hybrid energy system at mal solution when the global solar radiation is greater than
other locations in southern Ghana. 4 kW h/m2/day and the wind speed is less than about 3.4 m/s. At
the upper left-hand corner region in Fig. 10 where the wind speed
4.3.1. Global solar radiation and wind speed is more than 4 m/s, wind–Gen system is the most viable solution. It
Fig. 10 shows the optimal system type under different global can further be observed from this figure that the LCOE reduces as
solar radiation and wind speed when the diesel fuel price is fixed the global solar radiation at a given wind speed and similarly, cost
at $0.95/L. Three optimal solutions can be observed from this fig- of electricity reduces as the wind speed increases, for a given global
ure: PV–Gen system, wind–Gen system and wind–PV–Gen system. solar radiation.

Fig. 13. Optimal energy system type with renewable operating reserves.
294 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295

0.310 However, for a given combination of wind speed and solar radia-
tion, the cost of electricity is observed to be higher for a system
0.300 with renewable operating reserves than a system without reserves.
For instance, at the base case of a wind speed of 5.11 m/s and glo-
LCOE ($/kWh)

bal solar radiation of 5.4 kW h/m2/day, the LCOE ($0.282/kW h) of

0.290
the optimal system with RORC is 2.17% more than the LCOE
($0.276/kW h) of the optimal system with RORC.
0.280

4.3.5. Effect of PV price on LCOE

0.270 The effect of solar PV price on the LCOE of the PV–wind–Gen–
Battery hybrid energy system is shown in Fig. 14 when the wind
0.260 speed is 5.11 m/s, diesel fuel price is $0.95/L and the global solar
1000 2000 3000 4000 5000 6000 7000 8000
radiation is 5.4 kW h/m2/day. As expected, a roughly linear rela-
Solar PV cost ($/kW)
tionship exists between the solar PV price and the LCOE. The LCOE
Fig. 14. The effect of solar PV price on the LCOE of the PV–wind–Gen–Battery is observed to decrease with increasing PV price. With a PV price of
hybrid energy system. $3000/kW as a base case, the LCOE for this energy system is esti-
mated to decrease by about 3.2% when the PV price is reduced to
$1500/kW (from $0.281/kW h to $0.272/kW h). However, the LCOE
4.3.2. Global solar radiation and diesel fuel cost
is observed to increase by about 9% (from $0.281/kW h to $.306/
Fig. 11 shows the optimal system type under different global
kW h) when the PV price is increased from $3000/kW to
solar radiation and diesel fuel price when the wind speed is fixed
$7500 kW. Therefore, it is concluded that the impact of PV array
at 5.11 m/s. Two optimal solutions can be observed from this fig-
price on LCOE is insignificant at this location.
ure. They are wind–Gen system and wind–PV–Gen system. The
most viable energy system at the lower region of Fig. 11 is wind–
5. Conclusions
Gen energy system when the global solar radiation is less than
about 3.8 kW h/m2/day and at a high diesel fuel price. In the rest
The present study has investigated the techno-economic feasi-
of the region, the wind–PV–Gen is the most economically viable
bility of utilizing hybrid solar–wind–diesel–battery systems for
system. However, it can be expected that the region covered by
electricity generation in remote areas of southern Ghana. The solar
the wind–Gen system will increase with increasing cost of PV array
and wind energy resource data are collected from the weather sta-
and hence, could make the wind–PV–Gen energy system less eco-
tion of Adafoah in greater Accra region of Ghana. The HOMER soft-
nomically viable. It can further be observed from this figure that
ware was used for the analysis and viability was determined on the
the LCOE reduces as the global solar radiation increases, for a given
basis of net present cost (NPC) and cost of electricity (COE). Based
diesel price. However, the cost of electricity is observed to be
on the assumptions used in this study, six economically feasible
increasing with diesel price at a given solar radiation.
configurations of hybrid power plant systems were identified for
the selected location with annual daily averaged global solar radi-
4.3.3. Global solar radiation and diesel fuel cost
ation of 5.4 kW h/m2/day and annual mean wind speed of 5.11 m/s.
Fig. 12 shows the optimal system type with the most econom-
The total electricity produced by the PV–wind-generator–battery
ically viable energy system under different wind speed and diesel
power plant is 791.1 MW h/year which comprised 126.9 MW h/
fuel price when the global solar radiation is fixed at 5.4 kW h/m2/
year (16.04%) from the solar PV; 248.2 MW h/year (31.38%) from
day. Two main optimal solutions can be observed from this figure
the wind turbine and; 416.0 MW h/year (52.58%) from the two die-
and they are PV–Gen system and wind–PV–Gen system. The most
sel generators. The cost of electricity for this hybrid system is
viable energy system at lower region of Fig. 12 is PV–Gen energy
found to be $0.281/kW h.
system when the wind speed is less than about 3.6 m/s and at high
Moreover, using the sensitivity analysis results, the findings of
diesel fuel price. In the rest of the region, the wind–PV–Gen is the
this study can be applied to all other locations in southern Ghana
most economically viable system, except at a very high wind speed
with global solar radiation and wind speed similar to the site con-
of more than 7 m/s when wind–Gen is the most viable option. It
sidered in this study. The authors hope that the analysis carried out
can further be observed from this figure that the LCOE reduces as
will be useful for policy makers and all stakeholders in the energy
the wind speed increases, for a given diesel price and, the cost of
sector as Ghana pursues her strategic energy goal of doubling its
electricity is observed to be increasing with diesel price at a given
electricity generation in the next decade.
wind speed.
Acknowledgements
4.3.4. Effect of renewable operating reserve
It was previously mentioned that HOMER can operate two oper- The authors would like to thank Kofi Nketsia Tabiri for his help
ating reserves relating to variability in wind speed and solar radi- with additional energy information on Ghana and the Electricity
ation. The optimal energy system and the LCOE of the systems can Company of Ghana Limited for providing the electrical energy con-
be different by implementation of these two operating reserves. sumption data.
Fig. 13 shows the optimal system type with the most economically
viable energy system under different wind speed and global solar References
radiation conditions when the renewable output operating
reserves are taken as 25% and 50% for solar and wind energy con- [1] Karekezi S, McDade S, Boardman B, Kimani J, Lustig N. Global Energy
Assessment: Toward a Sustainable Future. Energy Poverty and
version systems, respectively. For this figure, the diesel fuel price is
Development. International Institute for Applied Systems Analysis; 2012.
fixed at $0.95/L while the PV cost is fixed at $3000 per kW. Com- Chapter 2.
paring this figure with Fig. 10 (with no renewable operating [2] National Energy Statistics, 2000–2012 (2013), Energy Commission, Ghana.
reserves), it can be observed that more energy systems (five sys- [3] Ghana Energy Commission. National Energy Statistics 2000–2012; 2013.
[4] Annual Report of the World Bank 2012. <http://siteresources.worldbank.org/
tems) are feasible when the simulation process optimization is EXTANNREP2012/Resources/8784408-1346247445238/
subjected to renewable operating reserve constraints (RORC). AnnualReport2012_En.pdf>.
 M.S. Adaramola et al. / Energy Conversion and Management 88 (2014) 284–295 295

[5] Strategic National Energy Plan: 2006 – 202. Energy Commission, Ghana. [26] Bernal-Agustin JL, Dufo-Lopez R. Simulation and optimization of stand-alone
<http://energycom.gov.gh/Strategic-National-Energy-Plan/>; 2006. hybrid renewable energy systems. Renew Sustain Energy Rev
[6] Volta River Authority, Ghana. VRA completes first solar plant. <http:// 2009;13:2111–8.
www.vra.com>; 2013. [27] Lambert T, Gilman P, Lilenthal P. Micropower system modeling with HOMER.
[7] Park GL, Richards BS, Schäfer AI. Potential of wind-powered renewable energy In: Farret FA, Simões MG, editors. Integration of alternative sources of
membrane systems for Ghana. Desalination 2009;248:169–76. energy. John Wiley and Sons Inc.; 2006.
[8] Renewables Global Status Report, 2009 Update, 2009. <http://www.ren21.net/ [28] Duffie AD, Beckman WA. Solar engineering of thermal processes. 4th
pdf/RE_GSR_2009_Update.pdf> [accessed 15.10.12]. ed. Hoboken, New Jersey: John Wiley & Sons, Inc.; 2013.
[9] Adaramola MS. Feasibility study of off-grid hybrid energy systems for [29] Rohani M, Nour G. Techno-economical analysis of stand-alone hybrid
applications in Ondo state, Nigeria. J Eng Appl Sci 2012;7(1):72–8. renewable power system for Ras Musherib in United Arab Emirates. Energy
[10] Adaramola MS, Oyewola OM, Paul SS. Technical and economic assessment of 2014;64:828–41.
hybrid energy systems in south-west Nigeria. Energy Explor Exploit [30] Graham VA, Hollands KGT. A method to generate synthetic hourly solar
2012;30(4):533–52. radiation globally. Solar Energy 1990;44(6):333–41.
[11] Ajao KR, Oladosu OA, Popoola OT. Using homer power optimization software [31] Canada-solar. <http://www.canadiansolar.com> [accessed 21.01.14].
for cost benefit analysis of hybrid-solar power generation relative to utility [32] http://www.powerworld.com.gh/index.php [accessed 21.01.14].
cost in Nigeria. Int J Res Rev Appl Sci 2011;7:96–102. [33] DuPont H. Fuhrländer Wind Turbine Overview. Wind–diesel 2004 Workshop
[12] Nema P, Nema RK, Rangnekar S. A current and future state of art development 2004, Girwood, Alaska, USA. <http://www.windpoweringamerica.gov/
of hybrid energy system using wind and PV–solar: a review. Renew Sustain winddiesel/pdfs/2004_wind_diesel/company/lorax.pdf.
Energy Rev 2009;13:2096–103. [34] <http://en.wind-turbine-models.com/turbine/279/fuhrl-nder/fl-100-astos>
[13] Nfah EM, Ngundam JM. Modelling of wind/diesel/battery hybrid power [21.01.14).
systems for far North Cameroon. Energy Convers Manage 2004;9(8): [35] Cummins Gh. Ltd. <http://www.ghanayello.com/company/12575/Cummins_
1295–301. Gh_Ltd>.
[14] Caballero F, Sauma E, Yanine F. Business optimal design of a grid-connected [36] Perkins Diesel Generations – Model YXP60RSC data sheet. <http://yorpower.
hybrid PV (photovoltaic)–wind energy system without energy storage for an co.ke/graphics/pdf/1580Diesel-generator-60kVA-Perkins.pdf>.
Easter Island’s block. Energy 2013;61:248–61. [37] http://www.wholesalesolar.com/, 22 Jan 2014.
[15] Yanine FF, Caballero FI, Sauma EE, Córdova FM. Homeostatic control, smart [38] Kemausuor F, Adkins E, Adu-Poku I, Brew-Hammond A, Modi V. Electrification
metering and efficient energy supply and consumption criteria: a means to planning using Network Planner tool: the case of Ghana. Energy Sustain Dev
building more sustainable hybrid micro-generation systems. Renew Sustain 2014;19:92–101.
Energy Rev 2014;38:235–58. [39] Cotrell J. Modeling the feasibility of using fuel cells and hydrogen internal
[16] Ismail MS, Moghavvemi M, Mahlia TMI. Techno-economic analysis of an combustion engines in remote renewable energy systems. Technical, Report;
optimized photovoltaic and diesel generator hybrid power system for remote 2003. [NREL/TP-500–34648 (<http://www.nrel.gov/docs/fy03osti/34648.
houses in a tropical climate. Energy Convers Manage 2013;69:163–73. pdf>)].
[17] Kumar US, Manoharan PS. Economic analysis of hybrid power systems (PV/ [40] Dalton GJ, Lockington DA, Baldock TE. Feasibility analysis of stand-alone
diesel) in different climatic zones of Tamil Nadu. Energy Convers Manage renewable energy options for a large hotel. Renew Energy 2008;33:1475–90.
2014;80:469–76. [41] Givler T, Lilienthal P. Using HOMER software, NREL’s micropower optimization
[18] Adaramola MS, Agelin-Chaab M, Paul SS. Assessment of wind power model, to explore the role of Gen-sets in small solar power systems: Case
generation along the coast of Ghana. Energy Convers Manage 2014;77:61–9. Study: Sri Lanka. Technical Report NREL/TP-710-36774; 2005. [<http://
[19] Statistics Ghana – 2010 Population and Housing Census – Summary report of www.nrel.gov/docs/fy05osti/36774.pdf>].
final results. <http://www.statsghana.gov.gh/docfiles/2010phc/Census2010_ [42] Bank of Ghana. <http://www.bog.gov.gh/> 21 Jan 2014.
Summary_report_of_final_results.pdf>. [43] http://en.wikipedia.org/wiki/Fisher_equation (21 Jan 2014).
[20] Energy commission, Ghana. Solar and wind energy resource assessment [44] Are, 2011 Are. Hybrid mini-grids for rural electrification: lessons learned.
(SWERA). <http://www.energycenter.knust.edu.gh/downloads/5/5598.pdf>; <http://www.ruralelec.org/fileadmin/DATA/Documents/06_Publications/
2003. Position_papers/ARE_Minigrids_-_Full_version.pdf>; 2011.
[21] Jiang Y. Correlation for diffuse radiation from global solar radiation and [45] Shaahid SM, El-Amin I, Rehman S, Al-Shehri A, Ahmad F, Bakashwain J.
sunshine data at Beijing China. J Energy Eng 2009;135:107–11. Dissemination of off-grid hybrid wind–diesel–battery power systems for
[22] Adaramola MS. Estimating global solar radiation using common electrification of isolated settlements of hot regions. Int J Sustain Energy
meteorological data in Akure, Nigeria. Renew Energy 2012;47:38–44. 2007;26(2):91–105.
[23] Askari IB, Ameri M. Techno-economic feasibility analysis of stand-alone [46] Infield DG. Wind diesel systems technology and modelling—a review. Int J
renewable energy systems (PV/bat, wind/bat and hybrid PV/wind/bat) in Renew Energy Eng 1999;1(1):17–27.
Kerman, Iran. Energy Sources Econ Plan Policy 2012;7:45–60. [47] Mathew S. Wind energy: fundamentals, resource analysis and
[24] Adaramola MS, Paul SS, Oyewola OM. Assessment of decentralized hybrid PV economics. Heidelberg: Springer; 2006.
solar–diesel power system for applications in Northern part of Nigeria. Energy [48] Ayompe LM, Duffy A. An assessment of the energy generation potential of
Sustain Dev 2014;19:72–82. photovoltaic systems in Cameroon using satellite-derived solar radiation
[25] http://www.globalpetrolprices.com/diesel_prices/#Ghana. datasets. Sustain Energy Technol Assess 2014;7:257–64.