See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/261348660

Wind turbine performance index

Conference Paper · November 2012

DOI: 10.1109/EEEI.2012.6377028

CITATIONS READS

3 1,409

5 authors, including:

Yuri Ditkovich Alon Kuperman

Ariel University Ben-Gurion University of the Negev
6 PUBLICATIONS 59 CITATIONS 244 PUBLICATIONS 4,172 CITATIONS

SEE PROFILE SEE PROFILE

Asher Yahalom Michael Byalsky

Ariel University Hebrew University of Jerusalem
415 PUBLICATIONS 2,097 CITATIONS 13 PUBLICATIONS 65 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Asher Yahalom on 13 January 2017.

The user has requested enhancement of the downloaded file.

2012 IEEE 27-th Convention of Electrical and Electronics Engineers in Israel
Wind Turbine Performance Index
Yuri Ditkovich, Alon Kuperman, Asher Yahalom
Ariel University Center of Samaria
Ariel, Israel
alonku@ariel.ac.il
Abstract—In this paper, two approaches to calculating
Capacity Factor of fixed speed wind turbines are reviewed and
then compared using a case study. The quasi-exact approach
utilizes discrete wind raw data (in the histogram form) and
manufacturer provided turbine power curve (also in discrete
form) to numerically calculate the capacity factor. On the other
hand, the approximate approach employs a continuous
probability distribution function, fitted to the wind data as well
as continuous turbine power curve, resulting from double-
polynomial fitting of manufacturer-provided power curve data.
The latter approach, while being an approximation, can be solved
analytically thus providing a valuable insight into the factors,
affecting the Capacity Factor. Moreover, several other merits of
wind turbine performance are based on the analytically derived
equations and hence are an approximation as well. The note
shows that the results obtained by employing both approaches
are very close, thus the analytically derived approximations are
valid and may be used for wind turbine performance evaluation.
Index Terms— Wind turbine, Capacity Factor, Data fitting,
Probability Distribution Function
I. INTRODUCTION
Since the wind blows inconsistently, wind turbines barely
operate at their rated power. Therefore the concept of capacity
factor (CF) is usually engaged to assess the expected wind
turbine energy delivery [1]. In order to calculate the CF, wind
probability distribution function (PDF) and turbine power
performance curve are required. The capacity factor, given by
the ration between the average and rated turbine power, is
usually formulated as
E  P

 P(v) f (v)dv ,
1 6
CF   (1)
PR PR
0
where E[∙] is the mean value operator, P(v) and PR are the
power curve and rated power of the turbine, respectively, and
f(v) is the wind PDF. In reality, the power curve of the turbine
is a discrete series rather than a continuous function, and the
wind PDF in (1) is the result of fitting the discrete raw wind
data to an a priori assumed PDF. Hence, (1) is actually an
approximation. Nevertheless, several analytic solutions of (1)
have been presented in the literature [2], [3], allowing thorough
understanding of the factors, affecting the CF. The more
accurate (referred to as quasi-exact thereafter) solution, taking
into account the original rather than processed wind speed and
power curve data, exists in a spreadsheet form only [4] and its
Yael Ditkovich, Saad Tapuchi
Shamoon College of Engineering
Ashdod, Israel
bassy@sce.ac.il
outcome is given by a numerical value without any insight into
the CF formation.
In this note, the quasi-exact and approximate approaches to
calculating the CF are reviewed and then compared using a
case study. It is shown that the obtained results are similar,
justifying the use of analytically derived relations in spite the
fact that they describe an approximate solution only. Hence,
analytical derivations of additional merits of wind turbine
performances which rely on capacity factor (such as Turbine
Performance Index [5], Turbine-Site Pairing Performance [6],
etc.) are justified and the results obtained using these relations
are reliable.
II. THE WIND DATA
The wind speed data is usually provided by meteorological
stations as raw matrix of wind speed versus time at 10m height,
while sample times vary from 10 minutes to 1 day. In reality,
the sample time is much higher than stated and the available
data sample is actually an average of tens to thousands of faster
samples. An example of monthly wind speed raw data
represented by 10 minutes samples is shown in Fig. 1.
18
16
14
12
wind speed [m/s]
10
8
4
2
0
31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
time [days]
Fig. 1. A typical monthly wind speed raw data
The raw vector can be either transformed into a histogram
(discrete PDF) or fitted to a known continuous PDF, typically
of Weibull type, as shown in Fig. 2. When creating a
histogram, the bins are typically chosen to be 1m∙s-1 wide to
match the resolution of the manufacturer provided turbine
power curve data (explained in Section 3), resulting in the
following discrete PDF,
 f HST (v)  f (vi ), vi  0.5  v  vi  0.5 , (2) turbine is stall regulated; therefore the output power in the
rated region reduces with the increase of wind speed, as shown
where f (vi ) is the magnitude of the histogram bin, centered in Fig. 3.
at vi .
TABLE I. NEG MICON 1000/60 FIXED SPEED TURBINE POWER CURVE
As to Weibull fitting, several methods of deriving Weibull DATA
parameters from the raw data were compared in [7]. The
maximum likelihood estimates of the Weibull distribution
parameters are typically employed in dedicated software
packages, e.g. MATLAB.
0.16
1000 Manufacturer Data
Polynomial Fitting
0.14 900 ZOH Fitting

0.12
Histogram 800
Weibull fit
700
0.1

power, [kW]
600
probability

0.08 500

400
0.06
300
0.04
200

0.02 100

0
nonrated region rated region
0
0 2 4 6 8 10 12 14 16 v vR vF
0 C
5 10 15 20 27
wind speed, [m/s]
wind speed, [m/s]
Fig. 2. Histogram and Weilbull PDF fit of wind speed raw data of Fig. 1. Fig. 3. Power curve of NEG Micon 1000/60 fixed speed wind turbine

Weibull PDF is defined as The power curve data, provided by the manufacturer may be
k v  ( v )k extrapolated using either Zero Order Hold (ZOH) method [4]
fWBL (v)  ( )k 1 e c (3)
or polynomial fitting [3], as shown in Fig. 3. In the former
c c
with parameters c and k are related to the site wind speed mean case, each data point P(vi) is replaced by a 1m∙s-1 wide discrete
and standard deviation as bin, having the same magnitude, thus creating the following
staircase approximation of the power curve,
 1
v  c 1   (4) PZOH (v)  P(vi ), vi  0.5  v  vi  0.5 , (7)
 k
for i = 1…N.
and
In the latter case, the data is divided into nonrated and rated
 2  1 subsets and each subset is fitted to a distinct polynomial,
 v  c  1     2 1   (5)
 k  k   creating the following polynomial approximation of the power
curve,
respectively, where
 0, v  vC

( x)  t x 1et dt
0
(6)  n
 a1i v i , vC  v  vR
is the complete Gamma function. In case the wind raw data of  i 0
a site is absent, but the mean and standard deviation of the
PPOLY (v)  PR  n . (8)
 a vi ,

wind speed are known, Weibull PDF is usually assumed and its 2i vR  v  vF
parameters are calculated using (4) and (5). i 0

III. TURBINE POWER CURVE 0, v  vF
The power production of a wind turbine is associated with one IV. CAPACITY FACTOR CALCULATION
of the two nonzero regions of the power curve: the non-rated
region for wind speeds between the cut-in speed vC and the In case the raw wind data of the site at hub height is available,
rated speed vR or the rated region for wind speeds between the a discrete PDF in the histogram form, described by (2), may be
rated speed and the furling (or cut-out) speed vF. The turbine constructed. Hence, combining (2) and (7) as
power curve is usually supplied by the manufacturer as N-point PZOH (v) f HST (v)  P(vi ) f (vi ), vi  0.5  v  vi  0.5 (9)
discrete series {vi, P(vi)}, i = 1…N, with vi – vi-1 = 1, i.e. the and substituting into (1), the quasi-exact solution for capacity
data resolution is 1m∙s-1. For example, consider a NEG Micon factor is given by
1000/60 fixed speed turbine power curve, given in Table I. The
 1 N TABLE IV. 2008 MONTHLY QUASI-EXACT AND APPROXIMATE CAPACITY
CFQE 
PR
 P(vi ) f (vi ) .
i 1
(10) FACTORS

The approximate solution for capacity factor is obtained by

substituting (3) and (8) into (1) as [3]
)11(

i  i 
n vF k n
( i v i v i v
CFAP   a v e   ci
)
i
2i F
c
( ) a1i (( C )k , )  (a1i  a2i ) (( R ) k , )  a2i (( F ) k , ) ,
i 0 i 1 k k  c k c k c k 
where
y 30

wind speed [m/s]


1
 ( y, x)  t x 1et dt (12) January
( x ) 20
0
10
is the incomplete Gamma function.
0
V. CASE STUDY 1 5 10 15 20 25 30
time [days]
Consider a NEG Micon 1000/60 fixed speed stall-controlled
wind turbine planned to operate under the wind conditions of Histogram

probability
0.1 Weibull fit
Ariel, Israel. The turbine parameters are summarized in Table
II and its power curve in given in Table I. Coefficients of 0.05

fitting the turbine power curve to 6th order polynomials are 0

0 5 10 15 20 25 30
shown in Table III. wind speed [m/s]

TABLE II. NEG MICON 1000/60 FIXED SPEED WIND TURBINE DATA
30
wind speed [m/s]

February
20

10

TABLE III. TH
6 ORDER FITTING COEFFICIENTS OF NEG MICON 1000/60 0
POWER CURVE 1 5 10 15 20 25 29
time [days]

0.1 Histogram
probability

Weibull fit
0.05

0
0 5 10 15 20 25
The 2008 monthly wind raw data in Ariel (10 min resolution) wind speed [m/s]
at 70m height is shown in Fig. 4 along with the appropriate
histograms and fitted Weibull PDFs. The 70m height wind
wind speed [m/s]

speed data was extrapolated from the 10m height data, 20 March
provided by the Israeli Meteorological Service, using the
following relation [4], 10
0.3
v70  70 
 
0
(13) 1 5 10 15 20 25 30
v10  10  time [days]

The resulting quasi-exact and approximate capacity factors, 0.1 Histogram

probability

calculated using (10) and (12), accordingly, are plotted in Fig. Weibull fit
5 along with the percentage difference between the results. 0.05

According to Table IV and Fig. 5, the results are very alike; the
0
worst case error is 2.5% in October, reducing to less than 0 5 10 15 20
wind speed [m/s]
0.25% in January. Two interesting point arise from the case
study. First, the quasi-exact capacity factor is higher than the Fig. 4. Monthly 2008 wind raw data in Ariel and appropriate PDFs
approximate throughout the year. Second, the higher the
capacity factor is, the lower is the percentage difference
between the results, i.e. the October capacity factor is the year
lowest while the January capacity factor is the year highest.
 30 20
wind speed [m/s]

wind speed [m/s]

April August
20
10
10

0 0
1 5 10 15 20 25 30 1 5 10 15 20 25 30
time [days] time [days]

0.1 0.15
Histogram Histogram
probability

probability
Weibull fit 0.1 Weibull fit
0.05
0.05

0 0
0 5 10 15 20 0 5 10 15
wind speed [m/s] wind speed [m/s]

20 30

wind speed [m/s]

wind speed [m/s]

May September
20
10
10

0 0
1 5 10 15 20 25 30 1 5 10 15 20 25 30
time [days] time [days]

probability
Histogram Histogram
probability

0.1 0.1
Weibull fit Weibull fit

0.05 0.05

0 0
0 2 4 6 8 10 12 14 16 0 5 10 15
wind speed [m/s] wind speed [m/s]

30
wind speed [m/s]

June 20
wind speed [m/s]

20 October

10 10

0
1 5 10 15 20 25 30
time [days] 0
1 5 10 15 20 25 30
0.15 time [days]
0.15
probability

0.1 Histogram
probability

Weibull fit Histogram

0.1 Weibull fit
0.05
0.05
0
0 5 10 15
wind speed [m/s] 0
0 2 4 6 8 10 12 14
wind speed [m/s]
wind speed [m/s]

15 July 30
wind speed [m/s]

10 November
20
5
10
0
1 5 10 15 20 25 30
time [days] 0
1 5 10 15 20 25 30
time [days]
0.15 Histogram
0.15
probability

Weibull fit
probability

0.1 Histogram
0.1
Weibull fit
0.05
0.05
0
2 4 6 8 10 12 14
wind speed [m/s] 0
0 2 4 6 8 10 12 14 16
wind speed [m/s]
Fig. 4. Monthly 2008 wind raw data in Ariel and appropriate PDFs (cont'd) Fig. 4. Monthly 2008 wind raw data in Ariel and appropriate PDFs (cont'd)
 30
VI. CONCLUSION
wind speed [m/s]

December Two approaches to fixed speed wind turbines Capacity Factor

20
estimation were presented and compared in the paper using a
10 case study. The quasi-exact approach, which utilizes discrete
wind raw and manufacturer provided turbine power curve data,
0
1 5 10 15 20 25 30 is only able to numerically calculate the capacity factor. The
time [days]
approximate approach, while being an approximation (since it
0.1 utilizes continuous approximation of wind and power curve
probability

Histogram
Weibull fit data), results in an analytical solution and hence provides a
0.05 deeper understanding of the factors, affecting the Capacity
Factor and several other merits of wind turbine performance,
0
0 5 10 15 20 25 based on the analytically derived equations. The note has
wind speed [m/s] revealed the close similarity between the results obtained by
Fig. 4. Monthly 2008 wind raw data in Ariel and appropriate PDFs (cont'd)
employing both approaches, enforcing the validity of the
analytically derived approximations.
REFERENCES
0.8
[1] K. Abed and A. El-Mallah, "Capacity factor of wind turbines,"
Quasi-Exact
Energy, vol. 22(5), pp. 487 – 491, 1997.
Capacity Factor

0.6
Approximate
0.4 [2] M. Albadi and E. El-Saadany, "New method for estimating CF
0.2
of pitch-regulated wind turbines," Electr. Power. Syst. Res., vol.
80, pp. 1182 – 1188, 2010.
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec [3] Y. Ditkovich, A. Kuperman, A. Yahalom and M. Byalsky, "A
generalized approach to estimating capacity factor of fixed
2.5 speed wind turbines," IEEE Trans. Sust. Energy, vol. 3(3), pp.
2 607 – 608, 2012.
% Difference

1.5 [4] G. Masters, Renewable and efficient electric power systems,

1 John Wiley & Sons, New York, 2004.
0.5 [5] S. Jangamshetti and V. Rau, "Normalized power curves as a tool
0 for identification of optimum wind turbine generator
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
parameters.," IEEE Trans. Energy Convers., vol. 16(3), pp. 283
– 288, 2001.
Figure 5: 2008 Monthly Capacity Factors and percentage differences
[6] S. Hu and J. Cheng, "Performance evaluation of pairing between
sites and wind turbines," Renew. Energy, vol. 32, pp. 1934 –
1947, 2007.
[7] T. Chang, "Performance comparison of six numerical methods
in estimating Weibull parameters for wind energy application,"
Appl. Energy, vol. 88, pp. 272-282, 2011.

View publication stats