Feasibility Study of Wind Farms: A Case Study For Izmir, Turkey
Feasibility Study of Wind Farms: A Case Study For Izmir, Turkey
Feasibility Study of Wind Farms: A Case Study For Izmir, Turkey
Abstract
Wind is one of the world’s fastest growing renewable energy sources. The rapid growth in wind
power is a result of improvements accomplished in technology. This paper presents the technical and
economical feasibility of wind farms. The method is applied to a potential wind farm site located in
Izmir, Turkey. The site is considered on technical and economical parameters for the complete plant
and its running costs. For technical consideration wind speed, prevailing wind direction, and
temperature measurements are performed. For economical consideration, three different scenarios
namely, autoproducer, autoproducer group, and independent power producer (IPP) cases, are
investigated and compared with respect to net present value (NPV), internal rate of return (IRR), and
pay back period (PBP) criteria. The study indicates the costs of generated energy by wind turbines
with different characteristics as a function of the installed capacity. It is concluded that, the larger the
installed capacity, the smaller the generating cost per kWh. The generating cost was calculated as low
as 2.68 UScent/kWh for the IPP scenario. The profitability analysis also shows that, larger installed
capacity with larger rated power wind turbines present higher IRR of the investment. The sensitivity
analysis backs up the findings.
r 2006 Elsevier Ltd. All rights reserved.
Corresponding author. Tel.: +90 232 750 78 01; fax: +90 232 750 78 90.
E-mail address: barisozerdem@iyte.edu.tr (B. Ozerdem).
0167-6105/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jweia.2006.02.004
ARTICLE IN PRESS
726 B. Ozerdem et al. / J. Wind Eng. Ind. Aerodyn. 94 (2006) 725–743
Nomenclature
1. Introduction
Technology of the extraction of power from wind with modern turbines is a well-
established industry, at present. In parallel, the wind energy applications are growing
rapidly, either, as off-shore or on-shore wind energy systems throughout the world.
Improvements in high strength fiber composites, power electronics, and generators are the
main contributors to this growth. The results of improved efficiency and reduced unit cost
of wind turbines make wind power generation competitive to that of conventional sources.
Furthermore, the wind power has an additional advantage of being a non-polluting source
of energy.
ARTICLE IN PRESS
B. Ozerdem et al. / J. Wind Eng. Ind. Aerodyn. 94 (2006) 725–743 727
maximum capacity, which can be installed on the campus with consideration of grid
constraint and accessibility. Economic analysis was done for three alternative scenarios.
The methodology used for evaluating the expected annual energy yield, the cost and
profitability of the investment, the cost of produced unit energy on the basis of siting
parameters and electricity sales price, was applied to this case study. In the economic
analysis, fixed and variable costs were compared with the investment cost of wind farm.
Lastly, sensitivity and variance analysis were done for the three scenarios.
The wind measurement results confirm that the investigated site can be classified as
excellent, in terms of wind energy potential. In addition, the financial parameters for the
complete plant and its operational costs indicated that the most cost-effective scenario was
IPP case in terms of the cost of a kWh energy generated by wind turbines with different
characteristics.
World’s primary energy consumption has been increased by 2.6% in 2002, ahead of the
10-year growth trend of 1.4% per annum [9]. Despite their potential, the contribution of
renewable energies to world energy supplies is still modest. Renewable energies are
estimated to supply about 17% of world primary energy at present [10]. But, most of this is
from hydro-power. The other renewable energies such as solar, biomass, and wind
contribute a much smaller proportion of energy needs (e.g. about 3% of electricity and 2%
of primary energy) at present [11].
Since wind energy is currently viewed as one of the most promising renewable energy
resource, the growth is relatively high. Cumulative global wind energy generating capacity
was reached 47,317 MW by addition of worldwide installed new equipment totaling
7976 MW at the end of 2004 [12]. The growth is expected to continue into the next decade,
as well. Although about 50 countries around the world contribute to the global total,
Europe accounts for 72.4% of the current total capacity especially Germany, Spain, and
Denmark represent the fastest growth coming from Europe. The policy for the promotion
of renewable energy resources has been influenced by international obligations. Kyoto
Protocol is a good example for these obligations at the global level.
Turkey is a rapidly growing country, both in economic and population sense, with 8.5%
economic growth rate and 1.5% population growth rate. In parallel, it is one of the fastest
growing energy markets in the world. In 2004, the total installed and production capacity
of power plants increased to 37,381.6 MW and 149,608.3 GWh, respectively [13]. Turkey
has a considerable potential for electricity generation from wind. A study carried out in
2002 concluded that Turkey has a theoretical wind energy potential of nearly 90,000 MW
and an economical wind energy potential of about 10,000 MW. The most promising region
is in northwest Turkey [14]. Although Turkey has large amount of wind potential, the
installed wind energy capacity is only 19.2 MW at present. The first wind farm was
commissioned in autoproducer status in 1998, following the privatization activities, which
allowed the domestic and foreign private sectors to invest in the energy sector in 1984. The
ARTICLE IN PRESS
B. Ozerdem et al. / J. Wind Eng. Ind. Aerodyn. 94 (2006) 725–743 729
other two wind farms, which commissioned between 1998 and 2000 were established under
Build–Operate–Transfer (BOT) model. The total annual wind energy produced from these
three wind farms is about 65,000,000 kWh.
Developer should focus on the selected site’s further suitability and availability in
feasibility phase of the wind energy project. Site characteristics, in particular, detailed
technical assessment including on-site wind monitoring, natural constraints and grid
connection, and a detailed economic assessment are the main elements considered and
investigated during a feasibility study.
Actual site measurements such as wind speed, wind direction, and temperature are vital
for the wind project [15]. Multiple wind speed measurement heights are encouraged for
determining wind shear characteristics of a site. Since the energy contained in the wind is
directly proportional to the cubic wind speed, the measurements as close as possible to the
hub height of a wind turbine is crucial. But, typical wind measurement heights are 10 and
30 m.
Wind power density is a good indication of a site’s wind energy potential. This value
combines the effect of wind speed distribution and its dependence on wind speed and air
density. The wind power available per unit area swept by the turbine blades is given as the
following equation:
X
Pt ¼ 0:5 r C p ðV 3i ti Þ (1)
where Vi is the mean wind speed for the ith time interval (e.g. 10-min or hourly basis), ti the
number of hours corresponding to the time interval divided by the total number of hours,
Cp is a power coefficient provided by the wind turbine manufacturer, and r the air density
and can be calculated from the following equation:
r ¼ p=RT, (2)
where p is the air pressure (Pa), R the specific gas constant for air (287 J/kg K), and T the
air temperature (K).
A wind turbine generates power between cut-in wind speed, V ci , and cut-out wind speed,
V co . Wind turbine output between rated wind speed, V r , and the cut-out wind speed is
usually constant.
After recording on-site data for at least 1 year and using topographical data, wind
characteristics and power extraction can be predicted by using special modeling softwares
mapping potential energy yield over the site. Energy map helps developer to find the area,
which would give maximum output.
A geotechnical study conducted to determine soil stability, foundation requirements,
drainage and potential erosion problems and a network analysis conducted to determine
the impacts of embedding the wind farm onto the grid should also be considered in
technical assessment duration.
There are some advanced mesoscale wind resource-mapping techniques that combines
numerical weather and wind flow models with a variety of weather data and terrain
ARTICLE IN PRESS
730 B. Ozerdem et al. / J. Wind Eng. Ind. Aerodyn. 94 (2006) 725–743
features such as height variation, roughness and sheltering obstacles [16]. A number of well
known commercially available softwares such as WindPro [17], WAsP [18] and MesoMap
[19] use these techniques to develop wind resource maps in order to produce statistical
results and wind resource map. WindPro software is used as an analysis tool for this study.
A digital map of the site, weather data and terrain features are the main inputs of this
software. The statistical results consist of wind speed, prevailing wind direction and wind
energy distributions.
The cost of generating electricity in a wind farm has three main components: capital
cost, operation and maintenance costs, and financing cost. Absence of fuel cost is a distinct
characteristic of renewable energy sources such as wind. The unit energy generation costs
are slightly affected, once the wind farm is built. Hence, the economics of wind energy is
very sensitive to capital cost. Investors want to maximize the profit. In other words,
minimizing the cost is a method for achieving this goal. Generation cost per kWh energy,
GC, can be determined from the following expression:
A common and simple way to evaluate the economic merit of an investment is to calculate
its PBP or, in other words, break-even time. This value is usually measured in years and
shows amount of time to recover the total investment.
PBP ¼ C=AS; (6)
where AS is the annual savings.
Variance analysis is an analysis of the variables that have contributed to the outcome of
the IRR, NPV and break-even time values. Variance analysis can be carried out for both
costs and revenues.
4. Case Study
Successful wind farm development depends on good location. Location dictates three
key factors: wind energy output, grid availability and construction conditions. The
proposed site is on the campus area of Izmir Institute of Technology, which is located in
Urla county, Izmir. The site selection was done by using Analytical Hierarchy Process
(AHP) method, which takes the sites wind power potential and accessibility to roads and
power transmission lines into consideration as decision criteria [20]. Fig. 1 shows the aerial
view of the site.
Wind data were collected by two masts located on different coordinates on the site
between July 2000 and January 2003. They all had two anemometers at 10 and 30 m
heights and a wind vane at 30 m height of the mast, which was constructed as tabular
tower. Temperature, humidity, and atmospheric pressure data were also obtained during
the monitoring period. These data were collected with a rate of 144 observations per day
using 10-min time interval.
Collected data were evaluated by WindPRO software [17] for wind statistics. Fig. 2
shows mean monthly variation of measured wind speeds. The wind speed variations are
best described by the Weibull probability distribution function with two parameters,
namely, k as shape parameter and A as scale parameter. Table 1 shows the Weibull data
corresponding to 12 sectors. Weibull distributions for measured wind speeds at two
prevailing directions of north (N) and north-northeast (NNE) and total one representing
all sectors are given in Fig. 3.
Wind energy map in Fig. 4 indicates excellent zones corresponding to the Class 6 of
National Renewable Energy Laboratory (NREL) standards [15]. Four site locations were
selected which have mean wind speed of 7.9 m/s at 30 m height to install wind turbines,
virtually. Due to the availability of land and cost-effective grid connection, the wind farm
power capacity was revised. The nearest substation is in Urla county. It is connected to
154/34.5 kV line with 25 MVA transformer power. The short-circuit power at the network
nodes was calculated as 238 MVA. Accordingly, the maximum capacity which could be
connected to 34.5 kV line was calculated as 12 MW at most.
The economic analysis of the project was conducted by using RETScreen software [21].
The analysis was performed for turbines having different rated power capacities. Table 2,
shows the technical data of wind turbines used in the study. The cost-related data were
taken from Ref. [22], as well.
12
30 m 10 m
10
Mean Wind Speed (m/s)
0
n.
b.
ay
ne
ly
p.
.
ov
ar
pr
ug
ct
ec
Fe
Ja
Se
Ju
M
Ju
M
O
A
D
N
A
Months
Table 1
Weibull data corresponding to 12 sectors for 30 m height
Sector A-parameter (m/s) Mean wind speed (m/s) k-parameter Frequency (%) Wind shear
Fig. 3. Weibull distributions for measured wind speeds at two prevailing directions (N and NNE), and total one
representing all sectors for 30 m height.
The study covers three scenarios. First scenario was to build a farm, which provides the
energy of the Institute in autoproducer status. Second scenario was an autoproducer group
project including another facility besides the Institute. Third scenario was an IPP case. The
ARTICLE IN PRESS
734 B. Ozerdem et al. / J. Wind Eng. Ind. Aerodyn. 94 (2006) 725–743
analysis was conducted for project life of 25 years, debt ratio of 70%, debt term of 10
years, discount rate of 15%. Since the Institute owned the land, this cost was not included
in the economic analysis. Tax and insurance related issues, also, were not included in order
to make the analysis less complicated.
Table 2
Technical data of wind turbines used in the study
Wind turbine type Rated Cut-in Cut-out Rated Hub Rotor Expected
output speed (m/s) speed (m/s) speed height diameter (m) life (years)
(kW) (m/s) (m)
Table 3
Capital cost break down for autoproducer scenario (US$)
Equipment 1,349,680 61.74 921,260 54.09 1,437,604 64.02 1,546,240 65.42 2,099,540 71.31
Feasibility 40,600 1.86 40,600 2.38 40,600 1.81 40,600 1.72 40,600 1.38
Development 157,000 7.18 157,000 9.22 157,000 6.99 157,000 6.64 157,000 5.33
Engineering 36,100 1.65 36,100 2.12 36,100 1.61 36,100 1.53 36,100 1.23
Balance of plant 512,000 23.42 476,000 27.95 481,000 21.42 486,000 20.56 491,000 16.68
Miscellaneous 90,815 4.15 72,238 4.24 93,092 4.15 97,638 4.13 119,970 4.07
Table 4
Summary of all alternatives for IZTECH autoproducer scenario
IZTECH autoproducer 1 2 3 4
power capacities were investigated. Wind turbine with 1800 kW rated power capacity was
not taken into the analysis due to over production. Percent capital cost breakdown is
shown in Table 3. The operating and maintenance costs are 186,615 US$ for 0.6 MW,
172,023 US$ for 0.9 MW, 181,819 US$ for 1 MW, 176,384 US$ for 1.5 MW and 207,957
US$ for 1.8 MW capacity wind turbines for this case. Table 4 shows the summary of all
alternatives for this case.
ARTICLE IN PRESS
736 B. Ozerdem et al. / J. Wind Eng. Ind. Aerodyn. 94 (2006) 725–743
40.0
600k W 900k W 1000k W 1500k W
35.0
30.0
25.0
IRR (%)
20.0
15.0
10.0
5.0
0.0
7.5 8.5 9.5 10.5 11.5 12.5 13.5 14.5
Energy Sales Price (UScent/kWh)
Table 5
Variance analysis summary for Enercon E40-600 kW
IRR values were calculated versus different energy sales prices and compared with each
other. For most of the energy projects, the required minimum IRR value is 15%. It is seen
from Fig. 5 that, wind turbine with 600 kW rated power has 11.12 UScents/kWh of energy
sale price corresponding to this rate. After having established base case, it is important to
determine the impact of the major cost components on the project. Therefore, an example
of variance analysis is presented in Table 5. After selecting the proper wind turbine for a
ARTICLE IN PRESS
B. Ozerdem et al. / J. Wind Eng. Ind. Aerodyn. 94 (2006) 725–743 737
particular scenario, variance analysis is needed in order to find out the most effective
parameters on the IRR and NPV values. The parameters, which are taken into account for
this analysis are energy sales price, operating cost, capital cost, inflation rate, debt, and
repayment period. Because those parameters are not certain and change the investment
credibility.
As it is seen from Table 5, the most effective parameters are energy sales price and
capital cost for this scenario. Repayment periods less than 10 years make this investment
unattractive.
Revenue is the only positive component of the cash flow. It is largely determined by sales
price. But the change in production has effect on it. Capital cost is an input at the very
beginning of the project and has negative influence on the discounted cash flow. Operation
cost exerts a strong impact on the cash flow, as well. Fig. 6 shows the sensitivity of these
costs as IRR versus variance.
Discount and interest rates are also important parameters in the project evaluation.
Figs. 7 and 8 presents NPV and IRR relationships versus discount and interest rates,
respectively.
30%
25%
20%
IRR
15%
10%
5%
Op Costs Capital Revenue
0%
-20% 0% 20%
Variance Parameter
3500
3000
2500
Project NPV (US$*1000)
2000
1500
1000
500
-500
0% 3% 6% 9% 12% 15% 18% 21% 24% 27% 30%
Discount Rate
20
19
18
17
16
IRR(%)
15
14
13
12
11
10
5% 6% 7% 8% 9% 10% 11%
Interest Rate
energy sales price and capital cost for this scenario too. Repayment periods less than 10
years are also unattractive.
Table 6
Capital cost break down for autoproducer group scenario (US$)
Equipment 2,004,520 68.89 1,802,520 67.57 2,835,208 75.57 3,052,480 76.54 2,099,540 71.31
Feasibility 40,600 1.40 40,600 1.52 40,600 1.08 40,600 1.02 40,600 1.38
Development 157,000 5.40 157,000 5.89 157,000 4.18 157,000 3.94 157,000 5.33
Engineering 36,100 1.24 36,100 1.35 36,100 0.96 36,100 0.91 36,100 1.23
Balance of plant 553,000 19.00 522,000 19.57 532,000 14.18 542,000 13.59 491,000 16.68
Miscellaneous 118,649 4.08 109,329 4.10 151,036 4.03 160,127 4.01 119,970 4.07
Table 7
Summary of all alternatives for autoproducer group scenario
Autoproducer group 1 2 3 4 5
55.0
50.0
45.0
40.0
35.0
IRR (%)
30.0
25.0
20.0
15.0
10.0
5.0
0.0
7.5 8 8.5 9 9.5 10 10.5 11 11.5 12 12.5 13 13.5 14 14.5 15
Energy Sales Price (UScent/kWh)
Table 8
Variance analysis summary for NM52-900 kW.
Table 9
Capital cost break down for IPP scenario (US$)
Equipment 13,146,800 82.34 11,506,380 82.01 16,821,248 85.94 12,099,920 83.50 14,466,780 85.44
Feasibility 59,000 0.37 59,000 0.42 59,000 0.30 59,000 0.41 59,000 0.35
Development 390,500 2.45 390,500 2.78 390,500 2.00 390,500 2.69 390,500 2.31
Engineering 158,000 0.99 158,000 1.13 158,000 0.81 158,000 1.09 158,000 0.93
Balance of Plant 1,590,000 9.96 1,368,000 9.75 1,382,000 7.06 1,218,000 8.40 1,197,000 7.07
Miscellaneous 622,772 3.90 548,275 3.91 761,430 3.89 566,017 3.91 659,851 3.90
Table 10
Summary of all alternatives for IPP scenario
IPP 1 2 3 4 5
50
600kW 900kW 1000kW 1500kW 1800kW
45
40
35
30
IRR (%)
25
20
15
10
0
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
Energy Sales Price (UScent/kWh)
Table 11
Variance analysis summary for NM52-900 kW
It is seen from Fig. 10 that, wind turbine with 900 kW rated power had 6.12 UScents/
kWh of energy sales price corresponding to the base IRR value of 15%. As it is seen from
variance analysis study, which is presented in Table 11 for this scenario, the most effective
parameters are energy sales price and capital cost. Repayment periods less than 10 years
are unattractive, as well.
5. Conclusion
In this article, a technical and economical feasibility study of wind farm alternatives to a
selected site in Izmir, Turkey has been proposed. Because of high wind resource capacity,
ARTICLE IN PRESS
742 B. Ozerdem et al. / J. Wind Eng. Ind. Aerodyn. 94 (2006) 725–743
Table 12
Comparison of three cases studied.
6
Unit Cost (UScent/kWh)
0
0 2 4 6 8 10 12 14
Installed Capacity(MW)
the site has been classified as excellent. The site has been evaluated for the unit cost of
electricity produced by each alternative, as well. The influence of each main cost parameter
has been estimated. The illustration has been done for three scenarios, namely
autoproducer, autoproducer group and independent power producer (IPP) cases.
According to the results, it can be concluded that the projects were more feasible for
autoproducer, autoproducer group, and IPP cases with 600, 900, and again 900 kW rated
power wind turbines, respectively. All three scenarios are promising in respect to
generating cost per kWh and IRR value. But the most suitable scenario has been found as
IPP case. Table 12 compares the three cases in terms of turbine type, plant capacity, energy
delivered, initial cost, and unit cost of energy generated. Fig. 11 also shows the variation of
generating cost per kWh versus installed power capacity. The results have indicated that,
the larger the installed capacity, the smaller the generating cost per kWh. It has been
shown that, the generating cost could be as low as 2.68 UScent/kWh in the IPP scenario for
11.7 MW installed capacity. The profitability analysis has shown that, larger installed
capacity with larger rated power wind turbines present higher IRR value for the
investment. Variance analysis studies have concluded that, the most effective parameters
are energy sales price and capital cost for the proposed scenarios. Repayment periods less
than 10 years are found unattractive. The sensitivity analysis study has backed up these
findings, as well.
ARTICLE IN PRESS
B. Ozerdem et al. / J. Wind Eng. Ind. Aerodyn. 94 (2006) 725–743 743
Despite encouraging growth forecasts, private investment in the renewable energy sector
is currently very modest and limited. The industry is still heavily dependent on government
and public funding. Therefore, incentive policies are essential for the commercialization of
renewable energy. In general, these incentives can be public funds for research and
development, direct support of investment costs, price support for electricity from
renewable resources, favorable interest rates, and tax incentives.
References
[1] National Wind Energy Coordinating Committee, Permitting Wind Energy Facilities—A Handbook, NWCC
Siting Subcommittee, www.nationalwind.org, March 1998.
[2] A. Pantaleo, A. Pellerano, F. Ruggiero, M. Trovato, Feasibility study of off-shore wind farms: an application
to Puglia region, Sol. Energy 79 (2005) 321–331.
[3] K.M. Stockton, Utility-scale wind on islands: an economic feasibility study of Ilio Point, Hawaii, Renew.
Energy 29 (2004) 949–960.
[4] M.A.H. El-Sayed, Substitution potential of wind energy in Egypt, Energy Policy 30 (2002) 681–687.
[5] A. Marafia, H.A. Ashour, Economics of off-shore/on-shore wind energy systems in Qatar, Renew. Energy 28
(2003) 1953–1963.
[6] W. El-Osta, Y. Kalifa, Prospects of wind power plants in Libya: a case study, Renew. Energy 28 (2003)
363–371.
[7] F. Cavallaro, L. Ciraolo, A multicriteria approach to evaluate wind energy plants on an Italian island,
Energy Policy 33 (2005) 235–244.
[8] J.K. Kaldellis, An integrated time-depending feasibility analysis model of wind energy applications in Greece,
Energy Policy 30 (2002) 267–280.
[9] EWEA and Greenpeace, Wind Force 12: A blueprint to achieve 12% of the World’s electricity from Wind
Power by 2020, www.awea.org, 2002.
[10] R. Gross, M. Leach, A. Bauen, Progress in renewable energy, Environ. Int. 987 (2002) 1–18.
[11] N. Hatziargyriou, A. Zervas, Wind power development in Europe, Proc. IEE 89 (2001) 1765–1782.
[12] Global Wind Energy Council, Global Wind Energy Markets, www.gwec.net, March 2005.
[13] Republic of Turkey, Prime Ministry State Institute of Statistics, www.die.gov.tr, March 2005.
[14] US Department of Energy-Office of Fossil Energy, An Energy Overview of the Republic of Turkey,
www.fossil.energy.gov, February 2005.
[15] National Renewable Energy Laboratory, Wind Resource Assessment Handbook, April 1997.
[16] WAsP User’s Guide, www.risoe.dk.
[17] WindPro Software, Ver. 2.1, www.emd.dk.
[18] WAsP Software, www.wasp.dk.
[19] MesoMap Software, www.awstruewind.com.
[20] D. Scanlin, X. Li, C. Larson, Wind Farm Siting, North Caroline Wind Summit, 2002.
[21] RETScreen Software, www.retscreen.net.
[22] Product Data, www.retscreen.net, March 2004.