Applied Energy 154 (2015) 290–297

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Economic performance indicators of wind energy based on wind speed

stochastic modeling
Guglielmo D’Amico a, Filippo Petroni b,⇑, Flavio Prattico c
a
Dipartimento di Farmacia, Università G. D’Annunzio, 66013 Chieti, Italy
b
Dipartimento di Scienze Economiche ed Aziendali, Università di Cagliari, 09123 Cagliari, Italy
c
Dipartimento di Metodi e Modelli per l’Economia, il Territorio e la Finanza, Università degli Studi di Roma La Sapienza, 00161 Roma, Italy

h i g h l i g h t s

 We propose a new and different wind energy production indicator.

 We compute financial profitability of potential wind power sites.
 The wind speed process is modeled as an indexed semi-Markov chain.
 We check if the wind energy is a good investment with and without incentives.

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

Article history: We propose the computation of different wind energy production indicators and financial profitability of
Received 3 February 2015 potential wind power sites. The computation is performed by modeling the wind speed process as an
Received in revised form 29 April 2015 indexed semi-Markov chain to predict and simulate the wind speed dynamics.
Accepted 30 April 2015
We demonstrate that the indexed semi-Markov chain approach enables reproducing the indicators cal-
Available online 19 May 2015
culated on real data. Two different time horizons of 15 and 30 years are analyzed. In the first case we con-
sider the government incentives on the energy price now present in Italy, while in the second case the
Keywords:
incentives have not been taken into account.
Wind speed
Indexed semi-Markov chains
Ó 2015 Elsevier Ltd. All rights reserved.
Monte Carlo simulation
Financial indicators

1. Introduction reviews by [9,10]). The influence of site specific parameters on

Renewable energy is becoming a big player in the world energy
productions. Already in 2009, the world renewable energy generat-
ing capacity was 26% compared to 66% deriving from fossil fuel [1].
In Europe, on account of the 120 GW of installed wind turbines,
284 TW h of electricity are produced per year. Even though the
installed capacity in Europe grows with a mean of 3% per year [2],
producers have to deal with many problems related to renewable
energy. In the case of wind farms, one of the most important hurdle,
is the intrinsic volatility of wind speed that requires accurate eval-
uation of wind speed data before installing wind turbines. Due to its
random behavior [3] and the intrinsic difficulty of storage [4], many
researchers are still working on the quantification of available
energy and performances estimation of wind projects [5–8].
Frequently, in the literature, indicators closely related to the
method of life cycle assessment (LCA) are used (see for example
⇑ Corresponding author.
E-mail address: fpetroni@unica.it (F. Petroni).
the environmental performance of wind energy converters has
been addressed in particular by [11], the issue of site specific
LCAs is further discussed in [12]. These methods are very impor-
tant when assessing large-scale wind projects, where marginal fac-
tors can cause relevant economic variations.
We focus our work on the micro-generation of energy from the
wind speed. In this scenario, the main important aspects are essen-
tially the satisfaction of the electrical power demand and the
return on investments for owners. Wind speed modeling plays an
important role in the micro-generation of electrical power from
wind speed. In fact, it can accurately describe the process in terms
of its intrinsic volatility and fluctuations, which can cause the not
fulfilment of the electrical power demand. Furthermore, good
stochastic modeling can give information on the production of
energy over the short-term, reproduce time lag autocorrelation
and it can also be used to forecast the produced energy for short
time scale.
Accurate probabilistic models permit the forecast of future
wind speed and energy production (see for example [13–15]). At

http://dx.doi.org/10.1016/j.apenergy.2015.04.124
0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.
 G. D’Amico et al. / Applied Energy 154 (2015) 290–297
the basis of this approach is the evidence that past values of the
wind speed keep informations about present value and future val-
ues. Many scholars have proposed new models that can allow the
prediction of wind speed minutes, hours or days ahead. Many of
these models are based on neural networks [16], autoregressive
models [17,18], Markov chains [19–24], hybrid models where the
previous mentioned models are combined [25–29] and other mod-
els [30–34]. All of these models try to catch, from the past, infor-
mation on the future. Often, these models are either focused on
specific time scale forecasting, or synthetic time series generation.
In a previous series of papers [35–37], we applied different
semi-Markov models to wind speed modeling and we demon-
strated that the semi-Markov framework out performs the
Markov models and should therefore be preferred in the modeling
of wind speed.
The approach we propose here is based on an indexed
semi-Markov chain (ISMC) model that we advanced in [38] and
applied to the generation of synthetic wind speed time series. In
[38] we demonstrated that the ISMC model is able to accurately
reproduce the statistical behavior of wind speed. The ISMC model
is a non-parametric model because it does not require any assump-
tion on the form of the distribution function of wind speed.
Furthermore, we showed that it is able to forecast wind speed at
different time scales without loosing the goodness of forecasting
which is almost independent from the time horizon [39].
The main purpose of this paper is to present a profitability anal-
ysis of wind energy production through economical and financial
indexes.
We beginning by presenting a comparison between the energy
produced by real wind speed and that obtained by implementing
the ISMC model through Monte Carlo simulation. Then, we
advance a new indicator that quantifies the utilization degree of
a wind turbine. This indicator is defined as the Satisfying Power
Demand Percentage (SPDP) and measures the fraction of time in
which the wind turbine satisfies the electrical power demand. It
can be interpreted as a degree of utilization of the wind turbine.
We compute these indicators for real and synthetic data. Finally,
we consider two financial indicators of the cash flow, namely,
Semi-Elasticity and Relative Convexity. We compute these indica-
tors on the cash flow generated by selling the electricity produced
by a wind turbine. The cash flow sequences are unknown and ran-
dom because they depend on the future values of wind speed.
Furthermore, the cash flows are subject to the interest rate risk.
This is an important component that affects the operation risk of
Fig. 1. Scheme of the research methodology adopted.
291
producing wind energy. The computed financial indicators allow
managing such risks.
The paper is presented as follow: First, we show our research
methodology, the database used and the technical details of the
commercial wind turbine. Second, the ISMC model is shortly
described as the theoretical support of the empirical application.
Third, a real data application is executed involving the proposal
of a new index and the computation of classical financial indicators
in our stochastic framework.
2. Research methodology and materials
The research methodology is summarized in Fig. 1. We use real
wind speed data, described here below, to set the stochastic model
that is described in the next section. Then, through Monte Carlo
simulations, we generate synthetic time series of wind speed.
Both time series, real data and synthetic data, are converted into
energy by means of the power curve of the chosen wind turbine.
We then suppose that the produced energy is sold in the market
at two different prices (with and without government incentives).
The sale produces cash flows on which we compute two important
financial indicators, namely, Semi-Elasticity and Relative
Convexity. To the best of our knowledge, we are the first to use
and apply these financial indicators to wind energy investments.
For the analysis conducted in this paper we use a free database
of a weather station situated in Italy at N45°280 14,900 –E9°220 19,900
and at 107 m of altitude. The data of the L.S.I. -Lastem station are
downloadable at www.lsi-lastem.it/meteo/page/dwnldata.aspx.
The database is composed of more than 230,000 data with a sam-
pling frequency of 10 min. The period of sampling starts from
25/10/2006 until 28/06/2011, almost five years. The station uses
a combined speed-direction anemometer at 22 m above the
ground. Since we need to test the model for the energy production
of a commercial wind turbine, we transpose our database at the
turbine rotor altitude using a well known relation, see e.g. [40]:
 a
h 1
v h ¼ v rif a¼ ð1Þ
hrif ln zh0
where v h is the wind speed at the height h of the wind turbine hub
and v rif is the value of the wind speed at the height hrif of the instru-
ment. In our application h = 50 m and hrif ¼ 22 m. The symbol z0
denotes a parameter that takes into account the morphology of
the area near the wind turbine. For a region without buildings or
292 G. D’Amico et al. / Applied Energy 154 (2015) 290–297

20 trees it varies from 0.01 to 0.001, instead for offshore applications it

is equal to 0.0001. For our analyzes we consider a mean value for an
Wind speed [m/s]

15 onshore application, therefore we chose z0 ¼ 0:005. In Fig. 2 we

show the real database and its probability density function.
10 The transformation between wind speed and produced energy
is made through a commercial 10 kW Aircon HAWT wind turbine.
5 The power curve of this turbine is plotted in Fig. 3 and its discrete
numerical values are shown in Table 1. The power curve of a speci-
0
fic wind turbine represents the energy production as a function of
0 200 400 600 800 1000 1200 1400 1600 the wind speed. The continuous values of the power curve are
Days obtained with linear interpolation of the discretized values.

0.4
3. The indexed semi-Markov chain model
0.3
The general formulation of the ISMC model has been developed
Frequency

by the same authors in [38], here we only discuss it briefly.

0.2
Semi-Markov processes are based on a similar idea as those that
generate Markov processes. The processes are both described by a
0.1 set of finite states v n whose transitions are ruled by a transition
probability matrix. The semi-Markov process differs from the
0 Markov process because the transition times T n are generated
0 2 4 6 8 10
Wind speed [m/s] according to random variables. Indeed, the time between transi-
tions T nþ1  T n is random and may be modeled by means of any
Fig. 2. Database and its probability density distribution. type of distribution functions. In studies concerning wind speed
modeling the states v n indicates discretized wind speed at the
nth transition and T n the time in which the nth change of wind
speed occurs.
In order to better represent the statistical characteristics of
10 wind speed, the ISMC model idea was advanced in the field of wind
speed (see [38]). The novelty, with respect to the semi-Markov
case, consists in the introduction of a third random variable
8 defined as follows:
X
m
T nk  T n1k
Um
n ¼ v n1k  : ð2Þ
T n  T n1m
Power (kW)

6 k¼0

This variable can be interpreted as a moving average of order m þ 1

executed on the series of the past wind speed values ðv n1k Þ with
4
weights given by the fractions of sojourn times in that wind speed
ðT nk  T n1k Þ with respect to the interval time on which the aver-
2
age is executed ðT n  T n1m Þ.
The ISMC model also considers this new variable when deter-
mining the probability of changes in wind speed. In a simple
0 semi-Markov model the probability of change depends only on
0 2 4 6 8 10 12 14 the present wind speed and on the transition time. The ISMC model
wind speed (m/s) can be considered as an m order semi-Markov chain, where the
Fig. 3. Power curve of the 10 kW Aircon wind turbine. dependence is given only by the averaged last m states.
The parameter m must be optimized as a function of the specific
database. The optimization is made by finding the value of m that
realizes the minimum of the root mean square error (RMSE)
between the autocorrelation functions (ACF) of real and simulated
data (see [38]). In our analysis m ¼ 7. The ISMC model revealed to
Table 1
Discrete values of the power curve of the 10 kW
be particularly efficient in reproducing, both, the probability density
Aircon wind turbine. function of wind speed and the autocorrelation function (see [38]).
The one step transition probability matrix can be evaluated by
Wind speed (m/s) Power (kW)
considering the counting transition between the three random
1 0 variables previously considered. Then, the probability pi;j ðt; uÞ rep-
2 0
2.5 0.4 resents the transition probability from the actual wind speed state
4 0.9 i, to the wind speed state j, given that the sojourn time spent in the
5 2.3 state i is equal to t and the value of the process U m is u. These prob-
6 3.3
abilities can be computed as:
7 4.9
8 6.7 ni;j ðt; uÞ
9 8.1 pi;j ðt; uÞ ¼ P ; ð3Þ
10 9.3 j ni;j ðt; uÞ
11 9.8
where nij ðt; uÞ is the total number of transitions observed in the
11.5–25 9.8
database from state i to state j in the next period having a sojourn
 G. D’Amico et al. / Applied Energy 154 (2015) 290–297 293

time spent in the wind speed i equal to t and the value of the index through Monte Carlo simulation, 1000 possible scenarios of wind
process equal to u. speed time series of length 15 years for the first case and 30 years
for the second case. The sampling period is 10 min.
4. Economic application
4.1. Energy production
Mathematical models of wind speed can be used to assess the
economic validity of an investment in a wind plant. In this section, According to the power curve described in Fig. 3 and Table 1,
we show how the ISMC model can be used to evaluate useful eco- wind speed has been transformed into power and then into pro-
nomic indicators like the Internal Rate of Return, the duced energy in a specific time window. Table 4 displays the
Semi-Elasticity and Relative Convexity of the cash flow generated
by an investment in a wind plant at a specific site. We conduct
Table 5
two separate analyzes: in the first, we consider an investment for Mean value, standard deviation, Skewness, Kurtosis and JB test for real and synthetic
a period of 15 years, which corresponds to the period in which gov- energy produced.
ernment incentives on the energy price in Italy are still present. In
Mean STD Skewness Kurtosis JB jB stat p-value
the second case we consider an investment without government
Annual energy
incentives and in this case, being the price of the energy lower than
Real 147.4 42.90 0.331 2.70 Rejected 29 0.001
in the previous case, the period of the investment is settled to be Synthetic 150.5 37.82 0.283 2.65 Rejected 24 0.001
greater than or equal to 30 years. The time period is defined such
that the initial investment is at least repaid by the selling of energy
and the pay back period is reached.
Since the ISMC model is based on discrete wind speeds, in order
to apply it to real data, we selected eight wind speed values (see 0.05

Synthetic Energy Frequency [kWh]

Table 2), to cover the entire wind speed distribution. Table 2 shows
the wind speed states with their related wind speed ranges. Also
the process U m has been discretized, Table 3 shows the states of 0.04
the process and their values.
Since the wind speed database is composed of only 5 years of 0.03
wind speed measures, and we need 15 or 30 years of data for the
analyzes, we apply a random sampling with replacement (boot-
strapping procedure). In our case, the bootstrap method allows 0.02
generating many longer time series, starting from the original
one, keeping exactly the same statistical properties of the original 0.01
data.
The database has been adapted to the blade height by using Eq.
(1). To compare real data with the ISMC model we generate, 0
0 0.01 0.02 0.03 0.04 0.05
Real Energy Frequency [kWh]
Table 2
Wind speed discretization. Fig. 4. Synthetic annual energy distribution as a function of real annual energy
distribution.
Sate Wind speed range (m/s)
1 0–3
2 3–4 10
3 4–5
4 5–6
5 6–7
Power Demand [kW]

6 7–8 9
7 8–9
8 >9

Table 3
U m processes discretization. 7

Sate U m range (m/s)

1 0–2.1 6
2 2.1–2.6 5 10 15 20
3 2.6–3.4 Hours [h]
4 3.4–6
5 >6 Fig. 5. Power demand scaled on the commercial wind turbine rated power. The
original values are taken in reference [42] and here only scaled.

Table 4
Frequencies values of the electrical energy produced in 10 min. Comparison between real and simulated data.

Energy (kW h) 0.0817 0.4083 0.5717 0.7350 1.0617 1.3883 1.5517

Real frequency 0.1767 0.0770 0.0825 0.0858 0.0827 0.0782 0.4171
Synthetic frequency 0.1718 0.0758 0.0818 0.0866 0.0838 0.0791 0.4211
294 G. D’Amico et al. / Applied Energy 154 (2015) 290–297

comparison between the distribution of the produced energy in a
10 min interval for real and synthetic data. Although the transfor-
mation from wind speed into energy is nonlinear, the ISMC model
is able to accurately reproduce a real case of energy production. In
fact, the values obtained in the two cases are almost identical. The
particular form of these distributions is due to the existence of a
cut-in wind speed, under which no energy is produced, and a part
of the power curve that remains at a constant value, from 10 m/s
17
Synthetic Data
Real Data
16
15
SPDP [%]
until 32 m/s, in which the wind turbine produces energy at its
rated power. This explains the existence of the two modes.
In Table 5 we give some statistics of the annual energy distribu-
tion for real and synthetic data. We included the results for the
JarqueBera test (JB), which is a goodness-of-fit test that determines
whether sample data have the skewness and kurtosis matching a
normal distribution. The test rejects the normality hypothesis for
the annual energy. This is mainly due to the absence of a
Gaussian tail of the distribution on its right side (see the distribu-
tion figure in [41]).
In Fig. 4, we show the synthetic annual energy distribution
against the real annual energy distribution. Dots close to the
straight line demonstrate the appropriateness of the ISMC model.
The coefficient of determination for this set of data is r2 ¼ 0:89.

To complete this section we show another comparison between

14 real and synthetic energies produced. We propose a new indicator
that quantifies the utilization degree of a wind plant. We define the
13 satisfying power demand percentage (SPDP) as the span of time in
which the wind turbine is satisfies the electrical power demand
12 relative to the specified, considered period. Formally it can be
defined as follows:
11
0 20 40 60 80 100 TðPa  Pr P 0Þ
SPDP ¼ ð4Þ
Overdimensioning [%] T

Fig. 6. Variation of the SPDP for a 5 years time period as a function of the rated where the numerator represents the period in which the electrical
power over dimensioning. power available (Pa ) satisfies the electrical power required (P r )

Bootstrap ISMC
0.12 0.12

0.08 0.08
Frequency

Frequency

0.04 0.04

0 0
−6.805 −6.795 −6.785 −6.9 −6.8 −6.7
Semi−Elasticity Semi−Elasticity
(a)
Bootstrap ISMC
0.12 0.12

0.08 0.08
Frequency

Frequency

0.04 0.04

0 0
−7.03 −7.02 −7.01 −7.1 −7.05 −7
Semi−Elasticity Semi−Elasticity
(b)
Fig. 7. Histograms of the Semi-Elasticity for real and simulated data. (a) Case with government incentives and cash flow of 15 years. (b) Case without government incentives
and cash flow of 30 years.
 G. D’Amico et al. / Applied Energy 154 (2015) 290–297 295

and T is the number of unit of time where the inequality Pa  Pr P 0
is satisfied in the considered time horizon. To compute this indica-
tor we evaluate the power available (practically the power available
on the wind turbine) and we compare it against a scaled electrical
power demand. We use the data of [42] and scale its maximum
value. Particularly in [42], the greatest value (during the day) of
the power demand is 14.8 MW. We scaled the entire curve to set
the maximum value equal to the rated power of the chosen com-
mercial wind turbine, which is 9.8 kW. The result of the scaling pro-
cedure is showed in Fig. 5. This represents the power demand in a
day, we extend it for the considered period (five years) without con-
sidering any seasonality.
In Fig. 6 we show the comparison of the SPDP evaluated on real
and synthetic data for a period of five years. Moreover, we compute
it by varying the over-dimensioning of the plant. In other words,
we fix the maximum power required during the day at 9.8 kW
and we assume to extend the rated power of the wind turbine from
0% to 100% of its value. For example, if we have a wind turbine with
a rated power of 10 kW and we want to improve the plant, we are
interested in knowing how the SPDP will change. If we assume a
40% of over-dimension (then we will have a turbine of
10 + 4 kW), the SPDP will be 13%, with a percentage variation of
about 10%.
It is interesting to note that the variation of the SPDP is very
important, it increases almost of 50% with a variation of 100% of
the over dimensioning. This behavior is well kept by the model.
4.2. Financial indicators
In this subsection we will compare important financial indica-
tors computed on real time series, generated through the boot-
strapping procedure, and simulated data, generated through the

Table 6
Statistics of Semi-Elasticity distributions evaluated for real and synthetic data and for the two cases: with government incentives (15 years) and without (30 years).

Mean STD Skewness Kurtosis JB jB stat p-value

Semi-Elasticity
Real 15 6.79 0.0024 0.0758 2.92 Accepted 1.21 0.50
Synthetic 15 6.77 0.0246 0.0658 3.01 Accepted 0.72 0.50
Real 30 7.03 0.0022 0.0183 3.03 Accepted 0.11 0.50
Synthetic 30 7.05 0.0118 0.0110 3.10 Accepted 0.43 0.50

Bootstrap ISMC
0.12 0.12

0.08 0.08
Frequency

Frequency

0.04 0.04

0 0
−10.53 −10.52 −10.51 −10.6 −10.4 −10.2

Relative Convexity Relative Convexity

(a)
Bootstrap ISMC
0.12 0.12

0.08 0.08
Frequency

Frequency

0.04 0.04

0 0
−17.98 −17.97 −17.96 −18.05 −18 −17.95
Relative Convexity Relative Convexity
(b)
Fig. 8. Histograms of the Relative Convexity for real and simulated data. (a) Case with government incentives and cash flow of 15 years. (b) Case without government
incentives and cash flow of 30 years.
296 G. D’Amico et al. / Applied Energy 154 (2015) 290–297

Table 7
Statistics of Relative Convexity distributions evaluated for real and synthetic data and for the two cases: with government incentives (15 years) and without (30 years).

Mean STD Skewness Kurtosis JB jB stat p-value

Relative Convexity
Real 15 10.52 0.0025 0.0768 3.07 Accepted 1.2277 0.50
Synthetic 15 10.40 0.0677 0.0226 3.24 Accepted 2.6253 0.50
Real 30 17.97 0.0035 0.0024 2.99 Accepted 0.0013 0.50
Synthetic 30 17.98 0.0176 0.0050 3.16 Accepted 1.1506 0.50

ISMC model by means of Monte Carlo simulations. As explained
above, the bootstrapping procedure allows generating model inde-
pendent time series by randomly resampling the original data. We
consider two different cases: In the first case we compute the cash
flow obtained by selling the energy produced by the wind turbine,
taking into account Italian government incentives on the selling
price. This choice is motivated by interest determine whether the
investment can also be justified without the government incen-
tives. In Italy, for a wind plant with a rated power under 1 MW
and until 15 years after its setup, the price of the energy is fixed
(at the time of writing, to 0.291 €/kW h). In this case we consider
a cash flow of 15 years. Just to give a quantitative comparison,
the second case takes into account the selling of energy at the pre-
sent market price in Italy, without government incentives
(0.027 €/kW h). In this case the considered period of the invest-
ment is extended to 30 years.
For each case we generate 1000 possible scenarios and we eval-
uate the financial indicators for both real and synthetic data. The
trajectories are transformed into power through the power curve
of the wind chosen turbine, then into energy by considering the
time between two successive steps and after in Euro by multiply-
ing for the specific price considered. The annual cash flows are
built and, at time 0, the cost of the initial investment is considered.
Commonly for wind turbines with a rated power under 100 kW, a
price of 3000 € for each kW is considered (this value comes from
some quotes obtained from sector companies). This price takes into
account the cost of the turbine, the setup cost and also the annual
maintenance cost. Given that our wind turbine has 10 kW of rated
power, we chose the initial investment of 30,000 €.
The considered indicators are the Semi-Elasticity and Relative
Convexity.
4.2.1. Indexes used for the measurement of interest rate risk
In this subsection we evaluate some measures of the interest
rate risk that are commonly used in financial analyzes and we
apply them to the investment in a wind plant. As we have seen,
the production of energy generates a cash flow sequence. The value
of this stream may change in time due to variations of the interest
rate. The wind energy producer, as any other firm, may suffer the
risk of changes in the interest rate and then it is necessary to con-
trol such an uncertainty.
One of the most important indexes is the Duration (D). It is
defined as a weighted average of time to payment of the cash flow
sequence. From D it is possible to recover the Semi-Elasticity (Se),
which is a relative measure of the variation of the value of the cash
flow stream. Depicted in formula we have:
Pn s
1 s¼1 s  CF s  ð1 þ rÞ D
Se ¼  P n s ¼  : ð5Þ
1þr CF s  ð1 þ rÞ 1 þr
s¼1
Another important index is the Relative Convexity (RC) that
expresses the convexity of the discounted value of the cash flow
stream as units of value variation. In formula we have:
Pn
1  ðs þ 1Þ  CF s  ð1 þ rÞs
s¼1 s C
RC ¼ Pn s ¼ ð6Þ
Se s¼1 CF s  ð1 þ rÞ Se
and C denotes the convexity of the cash flow. The behavior of the
duration and of the convexity were computed, by the same authors,
in [41].
In Fig. 8 we plotted the Relative Convexity in the same cases as
the previous indicators and in Table 7 its statistic values are
summarized.
As can be observed, for each pair of indicators and for each
specific investment period, the mean values of the indicators are
almost the same for real and simulated data.
5. Conclusion
We defined the Satisfying Power Demand Percentage as a new
indicator that quantifies the utilization degree of a wind turbine
and we evaluated it in conjunction with two financial indicators
(Semi-Elasticity and Relative Convexity)) of the cash flow gener-
ated by the sale of electrical energy produced by a given wind tur-
bine. These indicators provide valuable information concerning the
suitability of investment in a wind farm and measure the degree of
risk involved in wind energy production in conjunction with the
interest rate risk.
The computations were performed on simulated data obtained
by an indexed semi-Markov chain (ISMC) model of wind speed and
compared with those evaluated on real data. The results suggest
that the ISMC model is a correct model for wind speed analysis
and applications.
Future work will focus on using the evaluation of the proposed
indicators to evaluate new wind farm sites by applying them to
real case of investments in wind energy production and to extend
the analysis when we assume a stochastic evolution of the interest
rate process.
The importance of the present work is twofold: from one side
we have introduced indicators that can be used during the decision
process of investing in a wind farm in a specific site; on the other
site we have showed that the ISMC model reproduces, almost
exactly, the results obtained with real wind data, meaning that
the ISMC model is well suited for producing synthetic data of wind
speed.
Acknowledgement
The comments and helpful suggestions made by two anony-
mous referees are gratefully acknowledged.

In the numerical application the interest rate r was set equal to 3%

References
and the results of the Se are summarized in Fig. 7. In Table 6, we
give some numeric values of the statistics of the distributions of [1] Fouquet D. Policy instruments for renewable energy-from a European
the Semi-Elasticity estimated for real and simulated data. perspective. Renew Energy 2013;49:15–8.
 G. D’Amico et al. / Applied Energy 154 (2015) 290–297
[2] Suomalainen K, Pritchard G, Sharp B, Yuan Z, Zakeri G. Correlation analysis on
wind and hydro resources with electricity demand and prices in New Zealand.
Appl Energy 2014;137:445–62.
[3] Sahin C, Shahidehpour M, Erkmen I. Generation risk assessment in volatile
conditions with wind, hydro, and natural gas units. Appl Energy 2012;96:4–11.
[4] Johnston L, Díaz-González F, Gomis-Bellmunt O, Corchero-García C, Cruz-
Zambrano M. Methodology for the economic optimisation of energy storage
systems for frequency support in wind power plants. Appl Energy
2015;137:660–9.
[5] Chang TP. Estimation of wind energy potential using different probability
density functions. Appl Energy 2011;88(5):1848–56.
[6] Chang TP. Performance comparison of six numerical methods in estimating
weibull parameters for wind energy application. Appl Energy
2011;88(1):272–82.
[7] Nagai BM, Ameku K, Roy JN. Performance of a 3 kw wind turbine generator
with variable pitch control system. Appl Energy 2009;86(9):1774–82.
[8] Danao LA, Eboibi O, Howell R. An experimental investigation into the influence
of unsteady wind on the performance of a vertical axis wind turbine. Appl
Energy 2013;107:403–11.
[9] Lenzen M. Life cycle energy and greenhouse gas emissions of nuclear energy: a
review. Energy Convers Manage 2008;49(8):2178–99.
[10] Davidsson S, Höök M, Wall G. A review of life cycle assessments on wind
energy systems. Int J Life Cycle Assess 2012;17(6):729–42.
[11] Zimmermann T, Gößling-Reisemann S. Influence of site specific parameters on
environmental performance of wind energy converters. Energy Procedia
2012;20:402–13.
[12] Zimmermann T. Parameterized tool for site specific lcas of wind energy
converters. Int J Life Cycle Assess 2013;18(1):49–60.
[13] Jiang Y, Song Z, Kusiak A. Very short-term wind speed forecasting with
bayesian structural break model. Renew Energy 2013;50:637–47.
[14] Liu H, Tian H-q, Li Y-f. Comparison of two new arima-ann and arima-kalman
hybrid methods for wind speed prediction. Appl Energy 2012;98:415–24.
[15] Cassola F, Burlando M. Wind speed and wind energy forecast through kalman
filtering of numerical weather prediction model output. Appl Energy
2012;99:154–66.
[16] Bivona S, Bonanno G, Burlon R, Gurrera D, Leone C. Stochastic models for wind
speed forecasting. Energy Convers Manage 2011;52(2):1157–65.
[17] Poggi P, Muselli M, Notton G, Cristofari C, Louche A. Forecasting and simulating
wind speed in corsica by using an autoregressive model. Energy Convers
Manage 2003;44(20):3177–96.
[18] Ailliot P, Monbet V, Prevosto M. An autoregressive model with time-varying
coefficients for wind fields. Environmetrics 2006;17(2):107–17.
[19] Shamshad A, Bawadi M, Wan Hussin W, Majid T, Sanusi S. First and second
order markov chain models for synthetic generation of wind speed time series.
Energy 2005;30(5):693–708.
[20] Song Z, Jiang Y, Zhang Z. Short-term wind speed forecasting with markov-
switching model. Appl Energy 2014;130:103–12.
[21] Kantz H, Holstein D, Ragwitz M, Vitanov NK. Markov chain model for turbulent
wind speed data. Physica A: Stat Mechan Appl 2004;342(1):315–21.
[22] Castino F, Festa R, Ratto C. Stochastic modelling of wind velocities time series. J
Wind Eng Ind Aerodynam 1998;74:141–51.
297
[23] Hocaoğlu FO, Gerek ÖN, Kurban M. A novel wind speed modeling approach
using atmospheric pressure observations and hidden markov models. J Wind
Eng Ind Aerodynam 2010;98(8):472–81.
[24] Fawcett L, Walshaw D. Markov chain models for extreme wind speeds.
Environmetrics 2006;17(8):795–809.
[25] Pourmousavi Kani S, Ardehali M. Very short-term wind speed prediction: a
new artificial neural network-markov chain model. Energy Convers Manage
2011;52(1):738–45.
[26] Denison D, Dellaportas P, Mallick B. Wind speed prediction in a complex
terrain. Environmetrics 2001;12(6):499–515.
[27] Pandey M, Van Gelder P, Vrijling J. Bootstrap simulations for evaluating the
uncertainty associated with peaks-over-threshold estimates of extreme wind
velocity. Environmetrics 2003;14(1):27–43.
[28] Liu H, Tian H-Q, Chen C, Li Y-f. A hybrid statistical method to predict wind
speed and wind power. Renew Energy 2010;35(8):1857–61.
[29] Lei M, Shiyan L, Chuanwen J, Hongling L, Yan Z. A review on the forecasting of
wind speed and generated power. Renew Sustain Energy Rev
2009;13(4):915–20.
[30] Salcedo-Sanz S, Perez-Bellido AM, Ortiz-García EG, Portilla-Figueras A, Prieto L,
Paredes D. Hybridizing the fifth generation mesoscale model with artificial
neural networks for short-term wind speed prediction. Renew Energy
2009;34(6):1451–7.
[31] Segura-Heras I, Escrivá-Escrivá G, Alcázar-Ortega M. Wind farm electrical
power production model for load flow analysis. Renew Energy
2011;36(3):1008–13.
[32] Tar K, Szegedi S. A statistical model for estimating electricity produced by
wind energy. Renew Energy 2011;36(2):823–8.
[33] Lazić L, Pejanović G, Živković M. Wind forecasts for wind power generation
using the eta model. Renew Energy 2010;35(6):1236–43.
[34] Abdel-Aal R, Elhadidy M, Shaahid S. Modeling and forecasting the mean hourly
wind speed time series using gmdh-based abductive networks. Renew Energy
2009;34(7):1686–99.
[35] D’Amico G, Petroni F, Prattico F. First and second order semi-markov chains for
wind speed modeling. Physica A: Stat Mechan Appl 2013;392:1194–201.
[36] D’Amico G, Petroni F, Prattico F. Performance analysis of second order semi-
markov chains: an application to wind energy production. Method Comput
Appl Probability 2014:1–14.
[37] D’Amico G, Petroni F, Prattico F. Reliability measures of second-order semi-
markov chain applied to wind energy production. J Renew Energy 2013. Article
ID 368940.
[38] D’Amico G, Petroni F, Prattico F. Wind speed modeled as an indexed semi-
Markov process. Environmetrics 2013;24:367–76.
[39] D’Amico G, Petroni F, Prattico F. Wind speed and energy forecasting at
different time scales: a nonparametric approach. Physica A: Stat Mechan Appl
2014;406:59–66.
[40] Gass V, Strauss F, Schmidt J, Schmid E. Assessing the effect of wind power
uncertainty on profitability. Renew Sustain Energy Rev 2011;15(6):2677–83.
[41] D’Amico G, Petroni F, Prattico F. Forecasting wind speed financial return, arXiv
preprint arXiv:1312.3895; 2013.
[42] Mabel MC, Raj RE, Fernandez E. Analysis on reliability aspects of wind power.
Renew Sustain Energy Rev 2011;15(2):1210–6.