Research Article

Energy Exploration & Exploitation

2020, Vol. 38(1) 175–200
Assessment of wind energy ! The Author(s) 2019
DOI: 10.1177/0144598719875276
potential across varying journals.sagepub.com/home/eea

topographical features
of Tamil Nadu, India

Shafiqur Rehman1, Narayanan Natarajan2,

Mangottiri Vasudevan3 and Luai M Alhems4

Abstract
Wind energy is one of the abundant, cheap and fast-growing renewable energy sources whose
intensive extraction potential is still in immature stage in India. This study aims at the determi-
nation and evaluation of wind energy potential of three cities located at different elevations in the
state of Tamil Nadu, India. The historical records of wind speed, direction, temperature and
pressure were collected for three South Indian cities, namely Chennai, Erode and Coimbatore
over a period of 38 years (1980-2017). The mean wind power density was observed to be highest
at Chennai (129 W/m2) and lowest at Erode (76 W/m2) and the corresponding mean energy
content was highest for Chennai (1129 kWh/m2/year) and lowest at Erode (666 kWh/m2/year).
Considering the events of high energy-carrying winds at Chennai, Erode and Coimbatore, max-
imum wind power density were estimated to be 185 W/m2, 190 W/m2 and 234 W/m2, respec-
tively. The annual average net energy yield and annual average net capacity factor were selected as
the representative parameters for expressing strategic wind energy potential at geographically
distinct locations having significant variation in wind speed distribution. Based on the analysis,
Chennai is found to be the most suitable site for wind energy production followed by
Coimbatore and Erode.

1
King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia
2
Dr. Mahalingam College of Engineering and Technology (MCET), Pollachi, India
3
Bannari Amman Institute of Technology, BIT Sathyamangalam, Erode, India
4
King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia
Corresponding author:
Mangottiri Vasudevan, Bannari Amman Institute of Technology, BIT Sathyamangalam, Erode, Tamil Nadu 638401, India.
Email: devamv@gmail.com

Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution
4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and
distribution of the work without further permission provided the original work is attributed as specified on the SAGE and
Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).
176 Energy Exploration & Exploitation 38(1)

Keywords
Wind speed, wind power density, renewable energy potential, Weibull distribution, onshore
windmill, topography

Introduction
India has witnessed tremendous challenges in managing the ever-increasing energy demands
from various sectors despite being blessed with plentiful availability of large number of
alternate energy resources. The challenges are primarily accounted from population explo-
ration, environmental degradation, energy pricing, energy security and various socio-
economic crises. As a wake-up call for the nation, recently government has initiated various
policies and programmes strongly in support of renewable energy aiming at significantly
reducing its carbon emissions. Nonetheless, depletion of conventional energy sources and
associated environmental sustainability issues has put up new strategies for researchers and
policy makers to explore and discover renewable energy sources such as solar, wind, geo-
thermal, tidal, biogas, biofuels, etc. Among these, the usage of wind energy has become a
promising resource due to its competitive nature as a clean, abundant, easily harvestable,
inexhaustible and affordable resource (Baker, 2003; Rehman, 2004). It is certain that irre-
spective of rampant irregularities existing in global renewable energy market, overall
installed capacity for wind power has rapidly increased over 104% during the last decade
which is a clear indicative of an increasing role of wind power in fulfilling the future energy
demands (GWEC, 2018). Recently, China and USA upgraded their installed capacity within
one year (2015–2016) by 23.328 GW and 8.203 GW, respectively, meanwhile the global wind
power installed capacity shown a remarkable rise by 12.5% during the same period.
Modern technological developments have led to significant advancement in harvesting
wind power by adopting new techniques for the design of wind machines (Akpinar, 2013).
However, sustainability and efficiency in harvesting wind power in complex terrains largely
depends on the suitability of selected locations for setting up large scale wind farms. For the
development of wind power at any location, the statistical analysis of wind data serves as a
key tool for assessing wind energy potential (Rehman and Ahmad, 2004). Extensive studies
have been conducted on wind characteristics and wind power potential in many countries
around the world (Al-Abbadi, 2005; Ahmed Shata and Hanitsch, 2006; Baseer et al., 2015;
Chang et al., 2017; Elamouri and Ben, 2008; Keyhani et al., 2010; Kose, 2004; Oyedepo
et al., 2012; Rafique et al., 2018a, 2018b; Rehman et al., 2003, 2007, 2017; Sohoni et al.,
2017). Moreover, wind energy studies are being conducted even on buildings from the
perspective of thermal comforts and ventilation design to enhance energy saving (Gao
et al., 2018a, 2018b).
To accentuate the importance of evaluation of wind energy potential across the scales, a
few selected studies have been briefly reviewed here. Alaydi (2011) analysed the collected
wind data at two sites in the Gaza strip, namely, Gaza city and Gaza international airport in
Rafah city and these cities were recommended for wind energy exploitation in the Gaza
strip. Ajayi et al. (2011) assessed the potential viability of wind resource potential in Jos,
Plateau State, Nigeria for power generation based on the monthly mean wing speeds of
1987–2007 and it was concluded that Jos was a suitable site for wind farm projects of
varying sizes and that MW-h to GW-h of electricity was likely to be produced per period
Rehman et al. 177

of months, seasons and years. Amoo (2012) examined the wind energy potential of
Abeokuta and Ijebu-Ode, two southwest sites in Nigeria, and Ijebu-Ode was found to be
better in terms of annual variation of mean wind speed. Ilkilic and Aydin (2015) studied the
potential of wind power in the coastal areas of Turkey, namely, Black Sea, Marmara,
Aegean and Mediterranean regions. They found that Istanbul, Izmir, Datca, Bandirma,
Antakya, Canakkale and Hatay are the most promising areas for wind energy systems.
However, Kaplan (2015) showed that although Turkey is having significant wind energy
potential even more than many European countries (both onshore and offshore), the
expected efficiency is not being achieved due to lack of national renewable energy policy.
Dabbaghiyan et al. (2016) evaluated the wind energy potential in the province of Bushehr,
Iran. They concluded that the annual mean wind power density for that location was found
to be around 265 W/m2 for winds at a height of 40 m. Allouhi et al. (2017) evaluated the
potential of wind energy in six coastal locations in the Kingdom of Morocco, namely Al
Hoceima, Tetouane, Assila, Essouira, Laayoune and Dakhla. They concluded that Dakhla
is the most suitable location for harnessing the wind power while Laayoune has been iden-
tified as the second most suitable site. Chen et al. (2018) have analysed the wind power
potential in Saudi Arabia based on the MENA CORDEX (Middle East North Africa
Coordinated regional climate downscaling experiment). They have identified that the high
wind energy potential exists over a vast area of western Saudi Arabia. Salam et al. (2018)
evaluated the wind characteristics, velocity and directions were studied using Weibull dis-
tribution based on the measurement of wind speed at two different locations in Brunei
Darussalam. It was concluded that the annual en energy production at a wind speed of
5 m/s was found to be in the range between 1000 and 1500 kWh for both the locations in
Brunei Darussalam.
So far, technical evaluation and estimation of wind energy potential from small
to medium scale wind farms observed to be limited to a few island countries. Luong
(2015) provided a critical review on potential and current status of wind energy
utilization in Vietnam identifying the scope of large number of small scale wind farms.
Ko et al. (2015) conducted wind power resource assessment by installing a meteorological
mast in Weno Island, Chuuk State, in order to introduce wind energy in Chuuk
State, Micronesia. They have reported that annual energy production in the range of
36 MWh can be derived from a 20 kW rated wind turbine. Becerra et al. (2017) presented
a techno-financial evaluation of two Chilean locations with promising wind energy poten-
tial: Laguna Verde placed in the central region of the country and Porvenir in the
southern region. Albani and Ibrahim (2017) investigated the wind energy potential by
analysing a certain amount of gathered 10 minute measured data at four stations located
at the coastal sites in Malaysia, namely, Kudat, Mersing, Kijal and Langkawi. They found
that only Kudat and Mersing had a promising potential to develop a medium capacity of
wind turbine power, while the other sites were suitable only for wind turbines at a
smaller scale.
Although several studies have been carried out in various parts of the globe, scientific
evaluation of the wind energy potential of sites in India are very few. Indian economy is in a
major shift for implementing ‘low carbon strategies for inclusive growth’ where major gov-
ernment policies and programmes inevitably support renewable energy for combating the
adverse impacts of conventional energy production technologies. As such, wind energy
development in India has gained momentum only over the last few decades.
Ramachandra et al. (1997) evaluated the wind energy availability and its characteristics at
178 Energy Exploration & Exploitation 38(1)

Kumta and Sirsi in Uttara Kannada district of Karnataka based on the primary data col-
lected at these sites for a period of 24 months and concluded that coastal taluks such as
Karwar and Kumta have good wind energy potential. Ganesan and Ahmed (2008) used
wind speed and direction data measured at 10 m and 25 m above ground level for the
assessment of wind power potential for Bhopal, India. They reported that the annual
energy yields of 3.712 GWh and 4.431 GWh corresponding to hub heights of 50 m and
70 m. Chandel et al. (2014) assessed the wind resource potential of the Western
Himalayan region of Himachal Pradesh, India using Weibull distribution function. The
seasonal variations in daily mean wind speeds were used to estimate wind power density
during summer and winter seasons. In another interesting study, Singh and Prakash (2018)
evaluated the availability of wind energy for electricity production at different locations, i.e.
Ranchi, Jamshedpur, Devghar, Lohardaga and Chaibasa in Jharkhand, India and reported
that the sites studied were unsuitable for wind to electricity generation on a large scale at
10 m above the ground but small-scale wind turbines can be installed to extract energy from
low-speed wind. Phadke et al. (2011) have reported Tamil Nadu as one of the potential
states for harvesting 95% of the nation’s wind energy potential. But they have not men-
tioned the detailed intricacies of possible variations in considering the topographical differ-
ences at various locations within the state.
In this paper, wind characteristics and energy potential of three cities, namely Chennai,
Erode and Coimbatore located at different elevations from the mean sea level, with different
topographical features and different distances from the shore has been studied.
Furthermore, the study provides the diurnal variation of the mean wind speed, wind direc-
tional analysis, wind direction frequency distribution, wind speed probability distribution
function, monthly mean variation of the meteorological parameters, annual mean variation
of the wind speed, annual mean variation of the Weibull parameters, wind energy yield and
capacity factors, monthly net energy yield and monthly percent net capacity factor for the
entire data set using five different wind turbines of 2.0 MW rated capacities. Although
several studies have been reported on the potentiality of the wind speed as a direct indicator
of wind power density, studies considering the combined effect of topography i.e., distance
from the shore and the elevation are very limited. This study is the first attempt of its kind,
where the combined effect of the above factors has been considered in estimating the wind
energy potential of a location.

Study area and data description

The study focuses on analysing and determining the wind power potential of cities located in
Tamil Nadu, India. The cities considered for the study are Chennai, Coimbatore and Erode
located at 6 m, 183 m and 411 m above mean sea level (MSL). The map showing the loca-
tions of these cities (in dark boxes) is provided in Figure 1.
Chennai is located on the southeastern coast of India in the northeastern part of Tamil
Nadu on a flat coastal plain known as the Eastern Coastal Plains. Chennai experiences
tropical wet and dry climate and its average elevation is around 6.7 m above mean sea level
(MSL) while the highest point is 60 m. The meteorological station in Chennai is surrounded
by a fence about 3 m high, trees, and three storeys high two buildings. The terrain of this site
is covered with grass and is almost open land on the northern side of the station 25 m away
from the station. Erode District is landlocked and is comprised of a long undulating plain,
sloping gently towards the Kaveri River in the south-east. The meteorological data
Rehman et al. 179

Figure 1. Map showing the geographical location of the cities in Tamil Nadu, India.

collection station in Erode is an inland site 405 km south west of Chennai. The station
surroundings are grassland and few small buildings around. Coimbatore is in the western
part of Tamil Nadu and is surrounded by the Western Ghats mountain range on the west
and north, with forests and the Nilgiri Biosphere reserves on the northern side.
180 Energy Exploration & Exploitation 38(1)

Table 1. Site dependent data summary.

Location

Variable Chennai Erode Coimbatore

Latitude N 13.000 N 11.5000 N 11.000

Longitude E 80.000 E 77.5000 E 76.875
Elevation (m) 6 183 411
Start date 1/1/1980 1/1/1980 1/1/1980
Distance from sea shore (km) 10 227 318
End date 6/1/2018 6/1/2018 6/1/2018
Mean temperature (
C) 27.9 26.2 24.4
Mean pressure (kPa) 100.3 96.23 94.93
Mean air density (kg/m3) 1.152 1.102 kg/m3 1.104 kg/m3
Power density at 50m (W/m2) 129 76 W/m2 97 W/m2
Wind power class 1 (Poor) 1 (Poor) 1 (Poor)

The meteorological station is located 100 km southwest of Erode at Tamil Nadu Agriculture
University campus.
The summary of the site specific details (geographical coordinates, data collection period,
long-term mean ambient temperature and pressure and air density) is provided in Table 1.
Wind speed data were recorded/scanned every three seconds at 50 m above ground level
(AGL) and were saved as hourly averages. The wind speed and other parameters were
measured for the period of January 1980 to June 2018 (38 years). There are no missing
observations in the recorded data. The calm threshold was identified as 0 m/s for all the
stations. The mean temperature varies from 24.4
C to 27.9
C corresponding to Chennai and
Coimbatore and while the respective air densities vary from 1.152 kg/m3 to 1.104 kg/m3
(Table 1). The mean pressure varies from 94.93 kPa to 100.3 kPa. The power density at
50 m height is found to be 129 W/m2 at Chennai, 76 W/m2 at Erode and 97 W/m2 at
Coimbatore. The wind power class was found to be 1 for all the cities.

Material and methodology

The mean wind speed of the known wind speed data has been calculated using the follow-
ing expression
1
XN 
Vm ¼ V i (1)
N i¼1

where Vm refers to mean wind speed, N is the number of observation and V is the hourly
mean wind speed. The most probable wind speed (Vmp) denotes the most frequent wind
speed for a given wind speed probability distribution. From the scale and shape parameter
of Weibull distribution function, the most probable wind speed can be easily obtained from
the following equation (Islam et al., 2011)
 1=j
j1
Vmp ¼c (2)
j
Rehman et al. 181

The maximum wind energy (VmaxE ) carried by the wind speed can be
calculated from Weibull scale and shape parameters and can be obtained from the following
equation
 1=j
jþ2
VmaxE ¼c (3)
j

The wind resource potentiality of a site can be determined by calculating the wind power
density. The wind power per unit area, P/A or wind power density (W/m2) is (Manwell
et al., 2002)
 
P 1 3 1 3 jþ3
¼ qV ¼ qc C (4)
A 2 2 j

where P, q, A, V are power in Watts, average density of air in kg/m3 (1.23 kg/m3
for atmospheric pressure at sea level at 15
C), area perpendicular to the wind speed
vector in m2 and wind speed in m/s, respectively. By knowing the wind power density,
the wind energy density can be calculated using the following expression (Keyhani
et al., 2010)
 
E P 1 3 jþ3
¼ T ¼ qc C T (5)
A A 2 j

The knowledge of wind speed frequency distribution is very important to evaluate the
wind potentiality of a particular location. The helps in explaining the behaviour of the
wide range of the wind speed data. The simplest method to perform this is to use a
probability distribution function. There are many distribution functions to describe the
wind speed frequency curve such as Weibull, Rayleigh, Lognormal, etc. but the most com-
monly used distribution is the Weibull distribution (Ucar and Balo, 2011). The two-
parameter Weibull distribution is the most appropriate and widely accepted as well as
recommended distribution for wind speed data analysis. This is because it provides a
better probability distribution than other functions (Akpinar and Akpinar, 2005). Hence,
the same has been adopted for this study. The Weibull probability density function is
given as
 j1
j V V j
fðVÞ ¼ eð c Þ (6)
c c

where f(V), is the probability of observing wind speed (V), j and c are the Weibull shape and
scale parameter (m/s). The scale factor can be related to the mean wind speed through the
shape factor, which determines the uniformity of the wind speed in a given site (Oyedepo
et al., 2012).
The Weibull parameters are calculated using maximum likelihood method, least square
algorithm and WaSP algorithm (Windographer software). The mathematical equations of
these methodologies are explained as follows:
182 Energy Exploration & Exploitation 38(1)

Maximum likelihood method

This method was proposed by Stevens and Smulders (1979) for fitting a Weibull distribution
to a set of measured wind speeds. This method uses the following equation to calculate the
Weibull j parameter in an iterative method.
0X XN 11
N
B i¼1 i V k
lnðV i Þ lnðV Þ
i C
j¼@ X N
 i¼1
A (7)
V k N
i¼1 i

where Vi is the wind speed in time step i and N is the number of total time steps. Once the
shape parameter j has been found, the following equation can be used to obtain the value of
the scale parameter c
XN !1=k
k
V
i¼1 i
c¼ (8)
N

Least squares algorithm for Weibull parameter fitting

The definition of the cumulative distribution function of the Weibull distribution is given
as follows
"  k #
V
FðVÞ ¼ 1  exp  (9)
c

This can be rearranged to give the following equation (10) which can be further simplified
as equation (11)
 k
V
ln½1  FðVÞ ¼ (10)
c
( " #)
1
ln ln ¼ klnV  klnc (11)
1  FðVÞ

This equation (11) is of the form y ¼ mx þ c. Therefore, a plot of ln(V) on the x-axis and
n  o
ln ln 1FðVÞ
1
on the y-axis will result in a straight line with a slope equal to k and intercept
equal to k ln c. Therefore, to obtain the best fit Weibull distribution, the values of ln(V)
n  o
and ln ln 1FðVÞ
1
has to be calculated for every data point, then these values are put into a
linear least square solver for calculation of slope and intercept of the line of best fit. The
value of j is set as the slope of that line and c as equal to exp(-intercept/slope).

WAsP algorithm for Weibull fitting

The WAsP algorithm is based on the two principles: (i) the mean power density of the fitted
Weibull distribution is equal to that of the observed distribution and (ii) the proportion of
values above the mean observed wind speed is the same for the fitted Weibull distribution as
Rehman et al. 183

for the observed distribution. Considering the first condition, the following equation (12)
gives the mean power density (WPD) of the Weibull distribution, assuming constant air
density while the mean power density of the observed wind speeds, assuming constant air
density is obtained from equation (13).
 
1 3 3
WPD ¼ qc C þ 1 (12)
2 k

1 X 3
WPD ¼ q N Vi (13)
2N

Applying principle (i) in equations (12) and (13) and solving, the expression to calculate
the values of parameter c so obtained is given by equation (14).
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
u XN
u V 3
t i¼1 i

3
c¼ (14)
NC k3 þ 1

In order to fulfilment of second condition, symbol X is used to represent the proportion

of the observed wind speeds that exceed the mean observed wind speed. The cumulative
distribution function F(V) gives the proportion of the values that are less than V and 1-F(V)
is the proportion of values that exceed V. In the WAsP algorithm, the value of X is first
calculated and then the above equation (15) is solved iteratively, using the Brent method, to
find the shape parameter j. Then, the equation (14) is solved to obtain the value of scale
parameter c.
0 1k
U
s ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
B XN C ¼ lnðXÞ (15)
B3 C
@ U3 A
i¼1 i
NCðkþ1Þ3

Results and discussion

The wind speed characteristics (site dependent, annual, monthly and diurnal scales) over a
long period of almost 38 years at three locations are analysed and described in the following
sub-sections. These characteristics will include the wind speed, direction, wind power den-
sity, Weibull parameters, energy yield and plant capacity factor using five different wind
turbines each of 2.0 MW rated power.

Long-term variation at three sites

The long-term average wind speed at 50 m above ground level over entire data collection
period at Chennai, Erode and Coimbatore is compared in Table 2. As one moves away from
the coastal site (Chennai) towards in land (Erode and Coimbatore) the mean wind speed
decreases. Also, the mean maximum wind speed of 21.42 m/s is found at Chennai over the
184 Energy Exploration & Exploitation 38(1)

Table 2. Wind speed sensor summary.

Location

Variable Chennai Erode Coimbatore

Mean wind speed (m/s) 5.40 4.34 4.70

Median wind speed (m/s) 5.49 4.15 4.55
Max wind speed (m/s) 21.42 13.97 15.50
Weibull k 2.909 2.285 2.275
Weibull c (m/s) 6.044 4.886 5.300
Mean power density (W/m2) 129 76 97
Mean energy content (kWh/m2/y) 1127 666 849
Energy pattern factor 1.418 1.705 1.694
Data recovery rate (%) 100 100 100

data collection period and a mean minimum of 13.97 m/s at Erode. The parameters obtained
by fitting a two-parameter Weibull distribution are also provided. Maximum wind power
density of 129 W/m2 is obtained for Chennai and a minimum of 76 W/m2 at Erode. The
mean energy content for the three cities are 1127 kWh/m2/year for Chennai, 666 kWh/m2/
year for Erode and 849 kWh/m2/year for Coimbatore. The energy pattern factor varies from
1.418 to 1.705. The higher values of mean wind speed, wind power density and mean energy
content at Chennai may be due to its being the coastal site which is open to sea on the
eastern side and is also relatively flat land. The above parameter values are the least at Erode
measurement site, may be due to inland location though having gentle topographic features.
However, Coimbatore station is relatively better as far as wind characteristics is concerned
though locked by mountains in the north and south and to certain extent in west as well.
The topographical features in this area are expected to be contributing to relatively higher
values of wind speed and other parameters.
The Weibull parameters are calculated using three methods, viz., maximum likelihood,
least squares and WAsP and then the mean wind speed and wind power density (WPD)
values are calculated using these parameters and the measured data. Further to compare the
performance of these methods, the mean wind speed and WPD values, so calculated, are
compared with measured values using coefficient of determination (R2) values. For Chennai
station measured data, the maximum value of R2 is obtained using WAsP algorithm while
almost same performance is obtained from maximum likelihood method (Table 3, column 6)
with R2 value of 0.977. At Erode and Coimbatore, maximum likelihood and WAsP algo-
rithms produced the best values of R2, respectively. With this analysis, it is evident that both
maximum likelihood and WAsP algorithms could be used for the estimation of Weibull
parameters with acceptable accuracy. However, keeping in view the ease of the method,
maximum likelihood method can be preferred.
Higher magnitudes of 6.0 m/s and more are found to be more prevalent from northeast
and southwest directions at Chennai data collection station as shown in Figure 2(a). This is
also reflected in Figure 3(a) that more than 50% times the wind blows from south-west
direction and around 30% of the times from north-east at Chennai. At Erode station, the
higher magnitude winds between 5.0 m/s and 6.0 m/s blow from west-south and west-north
directions as can been observed from Figure 2(b). This is also reflected in Figure 3(b) that
Rehman et al. 185

Table 3. Comparison of Weibull parameters calculated using three methods (1980 -2017).
Chennai Erode Coimbatore

Mean Power Mean Power Mean Power

Weibull WS density Weibull WS density Weibull WS density

Algorithm k c (m/s) (m/s) (W/m2) R2 k c (m/s) (m/s) (W/m2) R2 k c (m/s) (m/s) (W/m2) R2

Maximum 2.91 6.04 5.39 137.1 0.977 2.29 4.89 4.33 84.0 0.973 2.28 5.30 4.69 107.6 0.977
likelihood
Least squares 2.65 6.13 5.45 150.1 0.957 2.33 4.89 4.34 83.0 0.956 2.21 5.33 4.72 112.2 0.971
WAsP 3.26 6.14 5.50 137.0 0.988 2.20 4.85 4.30 84.8 0.964 2.28 5.30 4.70 107.7 0.977
Actual data 5.40 137.0 4.33 84.4 4.70 107.7

Figure 2. Mean wind speed distribution in different wind directions (1980-2017).

Figure 3. Wind direction rose diagram (1980 -2017).

186 Energy Exploration & Exploitation 38(1)

Figure 4. Hourly mean wind speed frequency distribution (1980–2017).

around 45% times the wind blows from west-south and west-north directions.
At Coimbatore, wind speeds of 5.0 m/s to 6.0 m/s blow from the west and west south
directions, Figure 2(c) which is also reflected in Figure 3(c) that 45% to 50% times of the
wind comes from these directions. The frequency distribution of wind speed in different
speed bins at three locations is shown in Figure 4(a) to (c) along with the Weibull fit and is
indicative of excellent representation of wind speed by Weibull fit. It is evident from these
figures that wind speed remained around 70%, 60% and 70% between 4.0 m/s and 8.0 m/s at
Chennai, Erode and Coimbatore; respectively. This means that wind power can be gener-
ated at these sites above these percentages of times.

Long-term annual variation

The annual variation of wind speed over a longer period and other parameters is important to
have confidence on the annual availability of wind for power generation. At Chennai, the
annual mean wind speed varied between a minimum of 5.1 m/s in 2008 and a maximum of
5.9 m/s in 1983 while remaining around 5.2 m/s during most of the years (Figure 5). However,
larger variations are noticed in annual mean wind speed values at Chennai with an overall
decreasing trend from 1980. These variations can be due to being an open side to the sea on
east. At Erode meteorological station (Figure 6), the annual mean wind speed showed a
consistent (4.2 m/s) behaviour with minimum and maximum wind speeds of 4.0 m/s and
4.5 m/s. This consistent behaviour may be accounted for being the locked site from all the
direction, the gentle topographical features and being away from the sea. At Coimbatore
(Figure 7), the wind speed always remained more than 4.4 m/s with a minimum of 4.3 m/s
in 2015 and a maximum of 5.1 in 1990. Overall, larger variations are observed at this site.
These variations may be accounted for rough topographical features in Coimbatore.
The annual mean values of Weibull shape parameter (k) varied from 2.5 to 3.0 during
entire data reporting period with an exception of k being > 3.5 in year 1983. In case of Erode
station, it varied between 2.0 and 2.3, a consistent pattern, with an exception in year 2003
when it reached to more than 2.5. Relatively, larger variations (2.0 to 2.5) are observed in
annual mean k values at Coimbatore peaks of more than 2.5 in years 2003 and 2012.

Long-term monthly variation

For quality of power and grid stability management, the knowledge of monthly variation of
mean wind speed over extended period is important. Furthermore, to match the load
Rehman et al. 187

Figure 5. Annual mean wind speed variation at Chennai (1980–2017).

Figure 6. Annual mean wind speed variation at Erode (1980–2017).

demand, monthly mean wind speed and thereof wind power availability is required. The
long-term monthly mean wind speed is found to increase from January towards mid of the
year, reaching a maximum of 6.23 m/s in June, and then decreased towards end of the year
at Chennai (Table 4). The overall mean wind speed of 5.4 m/s is observed over the entire
data collection period. At Erode (Table 5) and Coimbatore (Table 6), the maximum wind
speeds were observed in the months of July (6.35 m/s) and June (6.94 m/s); respectively. At
both of these stations, like Chennai, an increasing trend of wind speed was noticed from
January to mid of the year and a decreasing trend towards the end of the year. Highest
ambient temperatures were found in Chennai and lowest at Coimbatore. However, higher
values of air density are found during winter time and lower during summer time, as can be
observed from Tables 4 to 6. Similarly, highest values of wind power density of 185 W/m2,
188 Energy Exploration & Exploitation 38(1)

Figure 7. Annual mean wind speed variation at Coimbatore (1980–2017).

Table 4. Monthly mean values of all the meteorological parameters at Chennai (1980 -2017).

Data column Units Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec All

Speed_50m m/s 4.95 4.57 4.85 5.48 5.85 6.23 6.06 5.72 4.63 4.43 5.92 6.14 5.4


Direction_50m 60.4 97.9 140 161 193 242 246 246 228 61.3 41.3 43.9 164


Temperature_2m C 23.9 25.6 28.3 30.6 31.6 30.4 29.5 28.9 28.3 27.3 25.6 24.2 27.8
Pressure_0m kPa 101 101 101 100 100 99.9 99.9 100 100 100 101 101 100
Air Density kg/m3 1.17 1.17 1.15 1.14 1.13 1.14 1.14 1.14 1.15 1.15 1.16 1.17 1.15
Speed_50m WPD W/m2 95 78 87 117 159 185 172 143 85 82 165 173 129

Table 5. Monthly mean values of all the meteorological parameters at Erode (1980-2017).

Data column Units Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec All

Speed_50m m/s 3.75 3.62 3.50 3.29 4.20 6.22 6.35 5.81 4.40 3.34 3.52 3.99 4.33


Direction_50m 74 95 115 158 261 266 272 274 270 288 52 59 276


Temperature_2m C 23.2 25.8 28.9 30.6 29.5 27.1 26.6 26.5 26.4 25.3 23.6 22.6 26.4
Pressure_0m kPa 96.6 96.5 96.3 96.1 96 96 96 96 96.1 96.2 96.4 96.5 96.2
Air Density kg/m3 1.12 1.11 1.09 1.08 1.09 1.1 1.1 1.1 1.1 1.1 1.11 1.12 1.1
Speed_50m WPD W/m2 40 42 41 34 64 177 190 148 68 31 34 45 76

190 W/m2 and 234 W/m2 were observed at Chennai (June), Erode (July) and Coimbatore
(June), respectively. The long-term monthly prevailing directions are also summarized in
Tables 4 to 6 at Chennai (164
), Erode (276
) and Coimbatore (261
), respectively.
The wind duration curves with wind speed bins for monthly and complete data sets are
provided in Figures 8 to 10 for Chennai, Erode and Coimbatore. At Chennai, the peak
frequency of 10% to 13% occurs between 5.0 m/s to 6.0 m/s (Figure 8). Furthermore, the
Rehman et al. 189

Table 6. Monthly mean values of all the meteorological parameters at Coimbatore (1980 -2017).

Data column Units Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec All

Speed_50m m/s 4.44 3.83 3.38 3.15 4.57 6.94 6.93 6.30 4.94 3.65 3.71 4.59 4.7


Direction_50m 80 86 102 249 258 258 262 263 260 256 67 72 261


Temperature_2m C 22.5 25.1 27.6 28.3 26.7 24.2 23.4 23.4 23.7 23.6 22.7 21.8 24.4
Pressure_0m kPa 95.2 95.1 95 94.9 94.8 94.7 94.8 94.8 94.8 94.9 95 95.2 94.9
Air Density kg/m3 1.11 1.1 1.09 1.09 1.09 1.1 1.11 1.11 1.11 1.11 1.11 1.12 1.1
Speed_50m WPD W/m2 65 50 38 31 78 234 233 180 93 44 46 74 97

Figure 8. Percent wind speed frequency distribution in different bins at Chennai (1980 -2017).

wind duration curve, based on entire data set, also follows the same trend and peak fre-
quency of 10% occurs at 6.0 m/s wind speed bin. The peak frequencies at Erode are seen to
be spread over a larger wind speed base (2.0 m/s to 5.0 m/s) and for overall data the peak of
about 5.8% occurs between 4.0 m/s and 5.0 m/s (Figure 9). Almost similar larger spread of
frequencies over wider wind speed bins is observed at Coimbatore (Figure 10). However, the
peak frequency of about 5%, based on entire data set, appeared at 4.5 m/s wind speed. This
frequency duration curve analysis showed that as one moves away from the coastal site in to
inland, though higher altitudes, the peak frequency and wind speed range decreases.

Long-term diurnal variation

Knowledge of diurnal changes in mean wind speed over extended periods is critical to match
the peak load on the grid. The perfect match is the higher wind speeds at peak load times of
the day. At Chennai, the diurnal mean wind speed shows two peaks, one in the early
morning at 4:00 hours and the other at 14:00 hours in the early afternoon (Figure 11).
So, if the local peaks occur at these timings, then this diurnal wind profile can be best
utilized to supplement the peak load through wind power. At Erode, there is a clean cut
peak wind availability at around 15:00 hours which could also best utilized for peak load
saving through wind power if the peak load also occurs at this time. However, there is a flat
190 Energy Exploration & Exploitation 38(1)

Figure 9. Percent wind speed frequency distribution in different bins at Erode (1980–2018).

Figure 10. Percent wind speed frequency distribution in different bins at Coimbatore (1980–2017).

peak between 5:00 and 12:00 hours at Coimbatore which could be best utilized for flat load
demands during this specific duration. Of the three sites, Chennai seems to be the best
situation for utilizing wind power.
The long-term diurnal variation of ambient temperature also dictates the peak load
demand periods during the day because the cooling demand goes high when the temperature
is low and vice versa. Also, the temperature directly affects the air density which is directly
proportional to the wind power density. Higher the temperature, lower the air density and
hence the WPD. Similarly, lower the temperature, higher the air density and then higher the
WPD. The diurnal profiles of ambient temperatures at all the stations show a peak at
around 10:00 hours (Figure 12). This fact is also reflected in air density values (Figure 13)
being lower during this time. As the ambient temperature increases, the air density
Rehman et al. 191

Figure 11. Diurnal variation of hourly mean wind speed (1980–2017).

Figure 12. Diurnal variation of hourly mean ambient temperature (1980–2017).

decreases, as can be seen in Figure 13. The diurnal WPD profiles match with the long-term
hourly mean wind speed values and also have embedded effect of air density (Figure 14).

Variation of most probable and maximum energy carrying wind speed

In the analysis of wind speed data, the most probable wind speed (Vmp) and maximum
energy carrying wind speed (VmaxE) are of particular interest to the wind power system
planners. In simple words, the most probable wind speed provides the most frequently
occurring wind speed for a given wind probability distribution. In other words, the
192 Energy Exploration & Exploitation 38(1)

Figure 13. Diurnal variation of hourly mean air density (1980–2017).

Figure 14. Diurnal variation of hourly mean wind power density (1980–2017).

maximum energy carrying wind speed represents the wind speed that generates the maxi-
mum amount of wind energy. Wind speed carrying the maximum energy is also used to
estimate the rated wind speed (Oyedepo et al., 2012). On the other hand, the most probable
wind speed represents the peak of the probability density function (Oyedepo et al., 2012).
Since the Weibull distribution accurately fits the wind speed data, these two wind speeds are
calculated using equations (2) and (3).
Rehman et al. 193

Table 7. Long-term values of most probable and maximum energy carrying wind speeds at all the locations
(1980–2017).

Vmp (m/s) VmaxE (m/s)

City Mean Max Min Mean Max Min

Chennai 5.23 5.98 4.76 7.24 7.56 6.88

Erode 3.80 4.29 3.36 6.43 6.93 5.95
Coimbatore 4.11 4.68 3.59 6.99 7.61 6.41

Figure 15. Monthly mean values of most probable wind speed (1980–2017).

The most probable (Vmp) and maximum energy carrying (VmaxE) wind speed values are
calculated using equations (2) and (3), respectively. The Weibull scale and shape parameters
appearing in these equations were obtained from expression given above by equations (7)
and (8). Highest mean (5.23 m/s), maximum (5.98 m/s) and minimum (4.76 m/s) values of
Vmp are found at Chennai while the corresponding minimum values of 3.80 m/s, 4.29 m/s
and 3.36 m/s at Erode, as observed from Table 7. Similarly, the highest of the mean, max-
imum, minimum values of the VmaxE are found at Chennai and the corresponding lowest
values at Erode. It is evident from this discussion that 5.23 m/s mean wind speed is the most
probable speed having maximum frequency in the entire data set at Chennai. At Erode and
Coimbatore, the smaller values of Vmp are indicative of larger presence of comparatively
lower values of wind speed than in Chennai. The mean maximum energy carrying wind
speeds 7.24 m/s, 6.43 m/s and 6.99 m/s are observed at Chennai, Erode and Coimbatore,
respectively.
Higher values (5.0 m/s to 6.0 m/s) of Vmp are observed throughout the year at Chennai
with exceptions in February and October (Figure 15) while June to August at Erode and
June to September at Coimbatore. This means that higher frequencies of higher winds are
found throughout year at Chennai and only for few months at Erode and Coimbatore.
194 Energy Exploration & Exploitation 38(1)

Figure 16. Monthly mean values of maximum energy carrying wind speed (1980–2017).

Table 8. Technical specifications of pitch controlled 2.0 MW rated capacity wind turbines.

Wind turbine Generic names Rotor diameter (m) Cut-in-speed (m/s) Rated speed (m/s)

WT1 82 4 15
WT2 80 3 13
WT3 90 3 14
WT4 87 3 13
WT5 93.3 3 11.2

The monthly mean maximum energy carrying wind speeds of 6.0 m/s to 8.0 m/s, 5.0 m/s to
8.0 m/s and5.0 m/s to 9.0 m/s are observed at Chennai, Erode and Coimbatore. However,
VmaxE remained always more than 6.0 m/s at all the sites based on entire data set (Figure 16).

Wind power generation and capacity factor analysis

In order to study the power generation and capacity factor, five wind turbines, each of
2.0 MW rated capacity and pitch controlled type, are chosen. The technical specifications of
the chosen turbines are summarized in Table 8. It is to be noted that largest rotor diameter
of 93.3 m and lowest rated power of 11.2 m/s corresponds to turbine WT5. The wind power
curves of the chosen wind turbines are provided in Figure 17. For calculating the net energy
generation and net capacity factor calculations, various equivalent losses of 11.4527%
(availability 2%, wake effect 4%, turbine performance 2%, electrical 0%, environmental
2%, curtailment 0% and others 2%) are considered.
Annual energy yields and capacity factors are calculated using the site specific wind speed
and wind turbines technical data (Tables 9 and 10). Wind turbines WT3 and WT5 produced
Rehman et al. 195

Figure 17. Wind power curves of the 2.0 MW rated power capacity used in this study.

Table 9. Monthly net energy yield (MWh) at 80 m hub height at Erode (1980–2017).

Month WT1 WT2 WT3 WT4 WT5

Jan 7315.41 5654.02 8001.53 7311.15 7790.31

Feb 5383.11 4105.94 5873.69 5286.80 5703.40
Mar 6673.59 5126.02 7303.56 6622.08 7102.13
Apr 8886.94 6946.55 9697.37 9015.74 9504.85
May 12,250.86 9753.22 13,065.12 12,503.64 12,998.56
Jun 13,502.72 10,838.29 14,311.89 13,820.31 14,297.50
Jul 12,936.15 10,364.71 13,764.46 13,216.98 13,717.54
Aug 10,936.31 8643.17 11,754.71 11,153.12 11,640.24
Sep 6102.54 4694.04 6619.81 6043.56 6469.58
Oct 5932.64 4560.83 6381.35 5826.03 6244.07
Nov 11,622.31 9306.20 12,457.39 11,869.55 12,349.08
Dec 12,975.88 10,369.28 13,949.90 13,314.64 13,820.56

highest energy while WT1 and WT4 produced lesser energy at Chennai (Figure 18). Wind
turbine WT2 performed the worst compared to other four turbines. Almost same perfor-
mance is observed at Erode and Coimbatore (figures are not included). Highest capacity
factors are obtained in 1983 and lowest in 2008 (Figure 19), irrespective of the turbine type.
The relative performance of all the wind turbines at Chennai, Erode and Coimbatore are
depicted in Figures 20 and 21 based on annual energy yield and capacity factors. The
average annual energy yield of > 3000 MWh over entire data reporting period is obtained
at Chennai based on turbines WT1, WT3, WT4 and WT5 (Figure 20). Similarly, the higher
196 Energy Exploration & Exploitation 38(1)

Table 10. Monthly percent net capacity factor at 80 m hub height at Chennai (1980–2017).

Month WT1 WT2 WT3 WT4 WT5

Jan 12.61 9.74 13.79 12.60 13.42

Feb 10.18 7.76 11.10 9.99 10.78
Mar 11.50 8.83 12.59 11.41 12.24
Apr 15.82 12.37 17.27 16.05 16.92
May 21.11 16.81 22.51 21.55 22.40
Jun 24.68 19.81 26.15 25.26 26.13
Jul 22.88 18.33 24.34 23.37 24.26
Aug 19.34 15.29 20.79 19.72 20.59
Sep 11.15 8.58 12.10 11.04 11.82
Oct 10.49 8.07 11.29 10.30 11.04
Nov 21.24 17.01 22.77 21.69 22.57
Dec 22.95 18.34 24.67 23.55 24.44

Figure 18. Annual net energy yield at 80 m hub height at Chennai (1980–2017).

Figure 19. Annual net capacity factor at 80 m hub height at Chennai (1980–2017).
Rehman et al. 197

Figure 20. Annual average net energy yield (1980–2017).

Figure 21. Annual average net capacity factor (1980–2017).

capacity factors between 17 to 19% are obtained at Chennai corresponding to the same
turbine types as above while at Erode the capacity factors remained around 10%
(Figure 21). However, at Coimbatore the capacity factor values remained between 12 and
14%. In terms of turbine performance, WT3 and WT5 produced the maximum energy at all
the locations. In terms of location performance, Chennai is found to be the location,
Coimbatore the second best while Erode is the least performer.

Conclusions
This study investigated the wind power potential and wind characteristics at three cities
located at different elevations and distance from the shore line across the state of Tamil
Nadu, India. For the wind resource characterization, 38 years of wind data were statistically
analysed in detail using two-parameter Weibull distribution function. The diurnal, monthly
and annual wind speed variations have been investigated. Furthermore, the most probable
and maximum energy carrying wind speed, wind power generation and capacity factor have
also been studied. The critical conclusions of this study are summarized as follows:
198 Energy Exploration & Exploitation 38(1)

Mean wind speed for Chennai, Erode and Coimbatore are 5.4 m/s, 4.34 m/s and 4.7 m/s
while the mean maximum are 21.42 m/s, 13.97 m/s and 15.50 m/s. The mean maximum wind
speed is highest in Chennai since it is located near the seashore. The wind speed decreases for
cities located inland. The mean wind power density for Chennai, Erode and Coimbatore are
129 W/m2, 76 W/m2 and 97 W/m2 and the mean energy content 1127 kWh/m2/year,
666 kWh/m2/year and 849 kWh/m2/year. The wind speed remained around 70%, 60%
and 70% between 4 m/s and 8 m/s at Chennai, Erode and Coimbatore. The probable
wind direction is predominantly from north-east and south-west in Chennai, west-south
and west-north in Erode, west and west-south in Coimbatore.
The annual mean wind speed varied between a minimum of 5.1 m/s in 2008 and a max-
imum of 5.9 m/s in 1983 at Chennai. At Erode, the annual mean wind speed varied between
4 m/s and 4.5 m/s. At Coimbatore, the variation was between 4.3 m/s and 5.1 m/s. The
monthly variation of the mean wind speed showed an increasing trend from Jan to mid
of the year and a decreasing trend towards the end of the year at all the locations.
The peak frequency of wind speed is observed between 5 m/s and 6 m/s at Chennai,
between 4 m/s and 5 m/s at Erode and Coimbatore. The highest and lowest of the mean,
maximum and minimum values of Vmp and VmaxE were found Chennai and Erode, respec-
tively. The mean maximum energy carrying wind speeds at Chennai, Erode and Coimbatore
were 7.24 m/s, 6.43 m/s and 6.99 m/s.
Highest capacity factors are obtained in 1983 and the lowest in 2008 irrespective of the
type of turbine. In terms of turbine performance, WT3 and WT5 produced the maximum
energy at all the locations. Chennai is found to the best site for wind energy development
followed by Coimbatore and Erode has the least prospects. Chennai has a higher mean wind
power density since it is located near the shore and possesses the least elevation above the
mean sea level. Although, both Coimbatore and Erode are located inland, Coimbatore has a
higher elevation compared to Erode which has resulted in a higher mean wind power den-
sity. Hence, proximity with the sea shore and the height of the station above mean sea level
plays a role in capturing the higher winds and hence higher energy yields.

Acknowledgements
The authors acknowledge the support provided by King Fahd University of Petroleum and Minerals,
Dhahran-31261, Saudi Arabia in completing this study.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or
publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of
this article.

ORCID iD
Mangottiri Vasudevan https://orcid.org/0000-0003-3503-318X
Rehman et al. 199

References
Ahmed Shata AS and Hanitsch R (2006) Evaluation of wind energy potential and electricity gener-
ation on the coast of Mediterranean Sea in Egypt. Renewable Energy 31: 1183–1202.
Ajayi OO, Fagnenle RO, Katende J, et al. (2011) Availability of wind energy resource potential for
power generation at Jos, Nigeria. Frontiers in Energy 5(4): 376–385.
Akpinar A (2013) Evaluation of wind energy potentiality at coastal locations along the north eastern
coasts of Turkey. Energy 50: 395–405.
Akpinar EK and Akpinar S (2005) An assessment on seasonal analysis of wind energy characteristics
and wind turbine characteristics. Energy Conversion and Management 46: 1848–1867.
Alaydi JY (2011) Assessment of wind energy potential in Gaza strip. Frontiers in Energy 5(3): 297–304.
Al-Abbadi NM (2005) Wind energy resource assessment for five locations in Saudi Arabia. Renewable
Energy 30: 1489–1499.
Albani A and Ibrahim MZ (2017) Wind energy potential and power law indexes assessment for
selected near-coastal sites in Malaysia. Energies 10(3): 307.
Allouhi A, Zamzoum O, Islam MR, et al. (2017) Evaluation of wind energy potential in Morocco’s
coastal regions. Renewable and Sustainable Energy Reviews 72: 311–324.
Amoo AM (2012) Evaluation of the wind energy potential of two south west sites in Nigeria. Frontiers
in Energy 6(3): 237–246.
Baker JR (2003) Features to aid or enable self-starting of fixed pitch low solidity vertical axis wind
turbines. Journal of Wind Engineering and Industrial Aerodynamics 15: 369–380.
Baseer MA, Meyer JP, Rehman S, et al. (2015) Wind power characteristics of seven data collection
sites in Jubail, Saudi Arabia using Weibull parameters. Renewable Energy 102: 35–49.
Becerra M, Moran J, Jerez A, et al. (2017) Wind energy potential in Chile: Assessment of a small scale
wind farm for residential clients. Energy Conversion and Management 140: 71–90.
Chandel SS, Ramasamy P and Murthy KSR (2014) Wind power potential assessment of 12 locations
in western Himalayan region of India. Renewable and Sustainable Energy Reviews 39: 530–545.
Chang T-P, Liu F-J, Ko H-H, et al. (2017) Oscillation characteristic study of wind speed, global solar
radiation and air temperature using wavelet analysis. Applied Energy 15: 650–657.
Chen W, Castruccio S, Genton MG, et al. (2018) Current and future estimates of wind energy poten-
tial over Saudi Arabia. Journal of Geophysical Research: Atmospheres 123: 6443–6459.
Dabbaghiyan A, Fazelpour F, Abnavi MD, et al. (2016) Evaluation of wind energy potential in
province of Bushehr, Iran. Renewable and Sustainable Energy Reviews 55: 455–466.
Elamouri M and Ben AF (2008) Wind energy potential in Tunisia. Renewable Energy 33: 758–768.
Ganesan S and Ahmed S (2008) Assessment of wind energy potential using topographical and mete-
orological data of a site in Central India (Bhopal). International Journal of Sustainable Energy
27(3): 131–142.
Gao R, Liu K, Li A, et al. (2018b) Biomimetic duct tee for reducing the local resistance of a ventilation
and air-conditioning system. Building and Environment 129: 130–141.
Gao R, Wang C, Li A, et al. (2018a) A novel targeted personalized ventilation system based on the
shooting concept. Building and Environment 135: 269–279.
Global Wind Energy Council (GWEC) (2018) Global Wind Report. Available at: www.gwec.net/wp-
content/uploads/vip/GWEC_PRstats2016_EN_WEB.pdf (accessed 20 April 2018).
Ilkilic C and Aydin H (2015) Wind power potential and usage in the coastal regions of Turkey.
Renewable and Sustainable Energy Reviews 44: 78–86.
Islam MR, Saidur R and Rahim NA (2011) Assessment of wind energy potentiality at Kudat and
Labuan, Malaysia using Weibull distribution function. Energy 36: 985–992.
Kaplan YA (2015) Overview of wind energy in the world and assessment of current energy policies in
Turkey. Renewable and Sustainable Energy Reviews 43: 562–568.
Keyhani A, Ghasemi-Varnamkhasti M, Khanali M, et al. (2010) An assessment of wind energy poten-
tial as a power generation source in the capital of Iran, Tehran. Energy 35(1): 188–201.
200 Energy Exploration & Exploitation 38(1)

Ko DH, Jeong ST and Kim YC (2015) Assessment of wind energy for small-scale wind power in
Chuuk State, Micronesia. Renewable and Sustainable Energy Reviews 52: 613–622.
Kose R (2004) An evaluation of wind energy potential as a power generation source in Kutahya,
Turkey. Energy Conversion Management 45: 1631–1644.
Luong ND (2015) A critical review on potential and current status of wind energy in Vietnam.
Renewable and Sustainable Energy Reviews 43: 440–448.
Manwell JF, McGowan JG and Roger AL (2002) Wind Energy Explained: Theory, Design and
Application. Amherst: John Wiley and Sons.
Oyedepo SO, Adaramola MS and Paul SS (2012) Analysis of wind speed data and wind energy
potential in three selected locations in south-east Nigeria. International Journal of Energy
Environmental Engineering 3: 1–23.
Phadke A, Bharvirkar R and Khangura J (2011) Reassessing wind potential estimates for India:
Economic and policy implications. Environmental Energy Technologies Division. Berkley Lab
report, Report No. LBNL-5077E. Available at: https://isswprod.lbl.gov/library/view-docs/public/
output/rpt81160.PDF
Rafique MM, Rehman S, Md Alam M, et al. (2018a) Feasibility of a 100 MW installed capacity wind
farm for different climatic conditions. Energies 11(8): 2147.
Rafique MM, Rehman S and Alhems LM (2018b) Developing zero energy and sustainable villages – A
case study for communities of the future. Renewable Energy 127: 567–574.
Ramachandra TV, Subramanian DK and Joshi NV (1997) Wind energy potential assessment in Uttara
Kannada district of Karnataka, India. Renewable Energy 10(4): 585–611.
Rehman S (2004) Wind energy resources assessment for Yanbo, Saudi Arabia. Energy Conversion and
Management 45: 2019–2032.
Rehman S and Ahmad A (2004) Assessment of wind energy potential for coastal locations of the
Kindgom of Saudi Arabia. Energy 29(8): 1105–1115.
Rehman S, Ahmed MA, Mohamed MH, et al. (2017) Feasibility study of the grid connected 10 MW
installed capacity PV power plants in Saudi Arabia. Renewable and Sustainable Energy Reviews
80: 319–329.
Rehman S, Halawnai TO and Mohandes M (2003) Wind power cost assessment at twenty locations in
the Kingdom of Saudi Arabia. Renewable Energy 28: 573–583.
Rehman S, El-Amin IM, Ahmad F, et al. (2007) Wind power resource assessment for Rafha, Saudi
Arabia. Renewable and Sustainable Energy Reviews 11: 937–950.
Salam MA, Yazdani MG, Rahman QM, et al. (2018) Investigation of wind energy potentials in Brunei
Darussalam. Frontiers in Energy. Available at: https://doi.org/10.1007/s11708-018-0528-4 (accessed
26 August 2019).
Singh R and Prakash O (2018) Wind energy potential evaluation for power generation in selected
districts of Jharkhand. Energy Sources, Part A: Recovery, Utilisation and Environmental Effects
40(6): 673–679.
Sohoni V, Gupta S and Nema R (2017) A comparative analysis of wind speed probability distributions
for wind power assessments of four sites. Turkish Journal of Electrical Engineering & Computer
Sciences 24: 4724–4735.
Stevens MJM and Smulders PT (1979) The estimation of the parameters of the Weibull wind speed
distribution for wind energy utilization purpose. Wind Engineering 3: 132–145.
Ucar A and Balo F (2011) An investigation of wind turbines characteristics and the wind potential
assessment of Ankara, Turkey. Energy Sources, Part A: Recovery, Utilisation and Environmental
Effects 33(13): 1291–1303.