Annals of the „Constantin Brancusi” University of Targu Jiu, Engineering Series , No.

1/2023

WEIBULL / RAYLEIGH DISTRIBUTION USED TO MODEL WIND

SPEEDS IN ORDER TO PREDICT WIND ENERGY PRODUCTION

Rodica David „Constantin Brâncuși University of Târgu-Jiu, Romania

ABSTRACT: The Weibull distribution is a popular probability distribution used to model wind
speeds in order to predict wind energy production. It is a continuous probability distribution with two
parameters, the shape parameter k and the scale parameter c.
The Weibull distribution is often used in wind energy applications because it has a flexible shape that
can fit a wide range of wind speed data. The shape parameter k controls the shape of the distribution, with
smaller values of k indicating a more peaked distribution and larger values indicating a more spread-out
distribution. The scale parameter c determines the location of the distribution along the wind speed axis.
This article will focus on the analytical expressions used for wind data assessment and on how to fit
the parameters needed for them deriving analytical expressions to fit the probability density function is way
to reduce the number of parameters to characterize the probability density function and the estimate wind
power resource.
The Weibull and Rayleigh distributions are the most commonly used statistical functions used for
wind power studies and this article have a closer look at them and reveals some commonly used method to fit
their parameters.

KEY WORDS: Weibull, Rayleigh, distributions, probability, wind energy, density function.

1. INTRODUCTION

The Weibull distribution that has

become a widely used standard in wind
energy application due to its simplicity. It
is described by the following equation

(1)
• c is known as the scale parameter
where is the probability of the wind it is also positive and has velocity
being at speed v. dimension.
It depends on two easily estimated Looking at the example we can see
parameters why c is called the scale parameter,
• k is known as the shape parameter multiplying c by 2 stretches the scale two
and is positive and dimensionless. times but keeps the area under the curve
The higher the value of k the to 1 as it is a normalized distribution.
higher the median wind speed. It
settles the shape of the function.
Location with lots of low wind
speeds as well as some very
strong winds would have a shape
value k below 2, whereas
locations with fairly consistent,
wind speeds around the median,
would have a shape value of 3.

20
 Annals of the „Constantin Brancusi” University of Targu Jiu, Engineering Series , No. 1/2023

The Weibull distribution has become c parameters we will need to evaluate

a widely used standard in wind energy some of the function’s moments.
application due to its simplicity and there For example, the first moment which
are simple analytical expressions for the corresponds to the mean wind speed is
moments as will be shown later. It is a the integrand from 0 to infinity of the
reference used in wind energy software product of the wind speed v by the
such as Wind Atlas Analysis and probability density function .
Application program WASP and it is We can obtain an analytical
included in regulations such as that on expression of the mean wind speed
wind turbine power performance testing. expressed as the product of the scale
The cumulative distribution function parameter by the gamma function of one
(CDF) of the wind speed v evaluated at a plus one of the shape parameters, and the
given value v0 is the probability that the scale parameter can be obtained from this
wind speed will take a value less than or expression.
equal to the given value v0.
Cumulative distribution function:
(5)
x = v/c
(2) dv = c dx
thus:
(3)

It is expressed as the integrand from (6)

zero to v0, in other words it gives the area It can be either computed numerically
under the probability density function or tabulated to facilitate its calculation,
from 0 to v0. In the case of the Weibull the properties of the variable distribution
distribution, the cumulative distribution and the gamma function lead to a simple
function also takes an analytical expression for higher moments as well
expression. expressed by this formula.
Gamma function:

(4)
(7)
The cumulative distribution function
is often used to quantify the goodness of (8)
fit of the Weibull distribution with
respect to the observed probability Results that: (9)
density function.
The Rayleigh distribution is a
particular case of Weibull distribution (10)
with shape parameter k equals 2. This is Other methods are used depending of the
convenient as in most locations around available information we might have.
the word the value of k is approximately In all cases, it is important to keep in
2. This reduces the expression of the mind that the wind speed measurements
probability density function to one single should be taken under similar conditions
parameter. to those used to estimate the parameters
Weibull distribution with its two of the distribution. Additionally, the
parameters leads to very simple analytical accuracy of the wind speed estimate will
expressions for the moments which are depend on the quality of the data and the
key to easily estimate the shape and scale appropriateness of the chosen
parameters k and c. So, to adjust the k and distribution.

21
 Annals of the „Constantin Brancusi” University of Targu Jiu, Engineering Series , No. 1/2023

2. MOST COMMONLY USED energy potential but does not ensure the
METHODS TO PREDICT maximum likelihood with the observed
distribution.
WIND SPEED
It is usually used when the mean wind
speed and the wind energy density are
Some of the most commonly used
known.
methods when only the mean wind speed
The WASP method computes these
is known is the Rayleigh distribution. It is
parameters from the first and third
in general accurate enough and often
moments as well as the probability of
used in wind atlas. The procedure is the
exceeding the mean wind speed 1 -
same: we evaluate c from the mean value
which must be estimated from the data.
v and use the gamma function.
The method focuses on the right-hand
When more than one moment are
tail of the Weibull distribution which is
known, as for instance the mean wind
an important part of the distribution in
speed and the standard deviation than the
terms of energy.
methods of Moments can be used.
This is why the WASP method is
To compute the shape parameter k,
preferred amongst the wind energy
this method of estimation of distribution
industry.
parameters was introduced by Karl
In this method k and c are calculated
Pearson in 1894.
by solving this equation:
The power density method is also a
method of moments. It ensures that the
energy content of the fitted Weibull
distribution equals the energy content of (13)
the observed histogram.
Wind energy is the kinetic energy of Popular methods also include the
energy mass. For a constant wind speed v usual maximum likelihood estimate. The
crossing a normal area a during a given principle of the maximum likelihood
period t kinetic energy contained for the originally developed by Ronald Fisher in
air mass will be the one expressed here in 1912 states that the desired probability
the following formula. distribution is the one that makes the
Power Density Method. observed data most likely, which means
that one must seek the value of the
(11) distribution parameters that maximize the
likelihood.
The wind power density for a unit of It is a method which is similar to the
area and per unit of time is defined then graphical method but overcomes its
as the kinetic energy per unit of area and shortfalls such as the absence of
unit of time, which in its integral confidence interval estimate or the
continuous form it it is related to the third ambiguous choice of the initial parameter
statistical moment estimation.
The procedure is simple: we build the
cumulative distribution function f and
isolate the exponential and we then apply
(12) the natural logarithm twice on each side.
Taking the analytical expressions of
the moments and defining the energy (14)
pattern factor we obtain an expression for
the shape parameter k. This method (15)
ensures the best estimation of wind Plotting the values and
approximating them to a straight line we

22
 Annals of the „Constantin Brancusi” University of Targu Jiu, Engineering Series , No. 1/2023

will obtain the Weibull parameters k and BIBLIOGRAPHY

c. The comparison of the wind speed 1. Ram, Mangey, Reliability
distributions and the observational data is engineering: methods and
made on the cumulative density function applications, CRC Press: 2019;
for wind speed. 2. Constantin Dumitrescu, Distribution,
The real wind histograms deviate Laws Used in Reliability,
significantly from the Weibull https://www.utgjiu.ro › rev_mec ›
distribution because of the complex mecanica › pdf;
topography of regions where the wind 3. Using the Weibull Distribution –
flow is weaker and more isotropic, but Reliability, Modeling, and Inference,
yet it is the most commonly used Wiley Publishing House, 2012;
statistical method. 4. Describing Wind Variations: Weibull
Distribution (xn--drmstrre-64ad.dk)
3. CONCLUSIONS 5. Analysis of Wind Energy
To use the Weibull distribution for Conversion System Using Weibull
wind predictions, historical wind speed Distribution - ScienceDirect
data can be analyzed to determine the 6. Weibull Rayleigh distribution -
appropriate values of k and c. Once these WIND RESOURCES | Coursera
parameters are determined, the Weibull
distribution can be used to estimate the
probability of various wind speeds
occurring, which can be useful for
designing wind turbines and predicting
wind energy output.
It is worth noting that the Weibull
distribution is just one of many
probability distributions that can be used
to model wind speeds, and the choice of
distribution should depend on the specific
characteristics of the wind data being
analyzed. Additionally, it is important to
validate any wind predictions using
actual data to ensure accuracy.
Once the wind speed is estimated and
the direction is known historically, can be
predict the wind power energy using the
power curve of the wind turbine. The
power curve represents the relationship
between the wind speed and the power
output of the turbine. Typically, wind
turbines start generating power at a wind
speed of around 3-4 meters per second
and reach maximum power output at
around 12-14 meters per second. The
power curve can be obtained from the
manufacturer or by testing the wind
turbine under different wind conditions.
By using the wind speed data and the
power curve, the expected energy output
of the wind turbine can be predicted.

23