13 The Wind Resource

Chapter 13
The Wind Resource
In the past, wind data were measured and evaluated almost exclusively from a meteoro-
logical point of view. However, these data are not sufficient when one is considering the
commercial exploitation of wind resources by means of wind turbines. The earlier me-
teorological data do not provide much detailed information about the inner case in wind
speed with height up to 150 m or the local wind conditions of a particular terrain. It is
only in the past three decades that extensive wind measurements have been carried out
with consideration of the particular aspects relating to the use of wind turbines. In the
meantime, full-coverage wind data are available in the countries in which wind energy
utilisation is widespread. A reliable database is provided mainly by the long-term
evaluation of the energy supply from existing wind turbines.
Nevertheless, the determination of the wind conditions at the intended site of the
wind turbines remains an important task that cannot be solved by means of the available
large-scale wind maps alone. Providing reliable wind data must, therefore, be the first
step in any planning for the use of wind turbines. Since it is impossible to carry out new,
long-term measurements in either case, critical apriori verification of existing data is of
particular importance. After that, a wind appraisal is obtained with the aid of the avail-
able semi-empirical methods.
Moreover, the value of verbal information provided by the local population and of
natural indicators should not be underestimated. Trees clearly growing at an angle, for
example, are a reliable indicator of high mean wind speeds. A knowledge of the charac-
teristics of the wind and of some of the laws governing its behaviour with respect to its
utilisation is indispensable for a successfully planning of wind energy projects.
13.1 Causes of the Wind and of the Power in the Wind

The movement of air masses in the atmosphere is perceived as wind and has various
causes. The first and most important of these is the heating of the earth by the sun. Wind
energy utilisation is, therefore, an indirect form of solar energy utilisation. The radiation
from the sun is absorbed by the earth's surface and then returned to the atmosphere
above. Since the earth's surface is not homogeneous (land, water, desert, forest, etc.),
506 CHAPTER 13 THE WIND RESOURCE
the absorption of the solar energy varies both with respect to geographic distribution
and with respect to the time of the day and the annual distribution. This nonuniform
heat absorption produces great differences in the atmosphere with respect to tempera-
ture, density and pressure so that the resultant forces will move the air masses from one
place to another. Above all, the tropical regions on the earth absorb much more solar
energy throughout the year than the polar regions. Since, as a result, the tropical regions
become warmer and warmer and the polar regions become increasingly colder, there is a
strong convection current flowing between these regions.
Coriolis forces produced by the rotation will deflect the air masses to the right (seen
in the direction of flow) in the northern hemisphere and to the left in the southern hemi-
sphere. This process causes the familiar spiral movements of air equalisation known
from the cloud pictures of the low pressure regions.
The second effect of earth rotation becomes effective at a medium altitude. Each air
particle has an angular momentum that is directed from west to east. If the particle is mov-
ing in the direction of the poles, it will approach the axis of rotation of the earth more
closely. The law of conservation of momentum causes an increase in the velocity compo-
nent from west to east as compensation for the increasingly closer approach to the pole.
This effect is less in the vicinity of the equator and causes the so-called “west drift” that is
opposite to the global wind direction (Fig. 13.1).
Apart from these global movements of air equalisation in the atmosphere, the wind
flows are also influenced by small-scale topography situations. For example, mountain
slopes facing the sun are heated more quickly. The heating and cooling of large con-
tiguous forest regions differ from that of water surfaces close by. Specially shaped
valley cuttings which follow the main wind direction can cause jet-like effects which
locally accelerate the wind velocity. These effects are certainly of significance for the
local wind conditions and must be taken into consideration when selecting a site for
wind turbines, and can also be utilised to advantage.
Close to the ground, the surface friction produces a decrease in wind velocity, which
also reduces the effect of the Coriolis forces. For this reason, the wind direction close to
the ground is deflected by approx. 30° less than the geostrophic wind in European lati-
tudes. Above the sea where the friction is less because of the relatively smooth surface,
the difference in direction with respect to the geostrophic wind is only about 10°. Since
the equalisation between the different pressure regions takes place mainly by means of
the deflected wind close to the surface, the low-pressure regions, for example, will per-
sist longer over the sea and are also accompanied by higher wind velocities.
At greater altitude, the air moves along lines of equal pressure (isobars). This move-
ment of air masses at an altitude of more than about 600 m is called geostrophic wind,
where the airflow can be considered free of surface influences. At lower altitudes the
influences of the earth's surface can be felt. This part of the atmosphere is known as the
boundary layer. The principal effects governing the properties of the boundary layer are
the strength of the geostrophic wind and the surface roughness, Coriolis effects and
thermal effects.
The influences of the thermal effects are graded into three categories: stable, unstable
and neutral stratification. Unstable stratification occurs when there is a lot of surface
heating. The warm air rises and the result is a thick boundary layer with heavy turbu-
lence in the air. If the adiabatic cooling effect causes the rising air to become colder
than its surroundings, its vertical motion will be suppressed. This is called a “stable
13.1 CAUSES OF THE WIND AND
A OF THE POWER IN THE WIND 507
Fig. 13.1. Global wind flows [1]]
stratification”. It occurs at co
old nights when the ground surface is cold. In this situattion,
the movement of the air is dominated by friction with the earth's surface and thee in-
crease in mean wind speed d with height is large. In a neutral atmosphere, adiabbatic
cooling of the air as it rises is such that it remains in thermal equilibrium with it´s sur-
roundings. This is often the case
c with strong winds. For wind energy utilisation, neuutral
stability is the most importan nt situation to consider at least when calculating the tuurbu-
lent wind loads on the turbine, but, as mentioned before, unstable atmospheric situattions
can influence the vertical win nd shear exponent (s. Chapt. 13.5.3).
Only about 2 % of the so olar energy of the 1.5 x 1018 kWh captured annually byy the
earth's atmosphere is convertted into energy of motion of the air envelope. Neverthelless,
this results in a calculated power of the wind of about 4 x 1012 kWh. This is one hhun-
dred times more than all of the t power station output installed on this globe. Of couurse,
such numerical values proviide virtually no information about the potential that cann be
commercially utilised, but nevertheless
n it is worthwhile to be mentioned. For the pur-
poses of wind energy use an nd wind turbine design, the wind vector is considered to be
composed of a steady wind plus p fluctuations about the steady wind. Whereas for dessign-
ing wind turbines, the steady y wind and the wind fluctuations have to be considered, the
power and energy obtained frrom wind can be based only on the steady wind speed.
The power available in the wind varies with the cube of the wind speed. A common
unit of measurement is the wind power density, or the power per unit of area normal to
the wind direction from the wind is blowing:
1
⁄
2
where:
ϱ = air density at standard atmosphere (kg/m³)

= wind velocity (m/s)
Due to the fact that the main intention is to extract the energy from the wind and not just
producing power, the most important parameter is the mean annual wind energy den-
sity:
1 1
2 8760
With respect to the annual wind frequency distribution it can be obtained:

1
2
where f (υW) is the wind frequency distribution, a Weibull function, as a rule. Instead of
integrating the mathematical Weibull function, the mean value of the third power of the
wind speeds in appropriate time intervals can also be used. (Not the cube of the mean
annual wind speed!)
Wind resource maps often estimate the potential of the wind resources in terms of
wind power classes referring to the annual wind power density.
13.2 Global Distribution of Wind Resources

The most striking characteristic of the wind resource is its variability. The wind is
highly variable, both geographically and temporally. This is amplified due to the cubic
relationship of the wind speed to available energy. On a large scale the global variability
describes the fact that there are many different climatic regions in the world. Within one
climatic region, there is a great deal of variation on a smaller scale influenced by the
physical geography, the size of land and sea, the topography and vegetation on land.
Above the open seas, the wind velocities are highest, whereas they rapidly decrease
above the land surfaces.
One of the most recent, comprehensive studies about the global distribution of the
mean wind velocities and about the usable wind energy potential was published in
2005 by the Stanford University in the US [2]. In this study, the data from 7753
ground stations and 446 radar measuring stations were evaluated. Using the usual
approaches for the increase in wind velocity with height, especially the exponential
13.2 GLOBAL DISTRIBUTION OF WIND RESOURCES 509
law and the logarithmic height formula in various modifications and refinements, a
worldwide overview of the average wind velocities at a height of 80 m was created
(Fig. 13.3). The evaluation period extends over the years from 1998 to 2002. The data
essentially represent the year 2000 which was estimated as the average wind year
worldwide.
The wind velocities were arranged in accordance with the wind power density classes
normally used in the US (Table 13.2). The technical wind energy utilisation is consid-
ered to be economically viable up to class 3. In class 2 only some hot spots are suitable
and class-1 areas are generally unsuitable.
Table 13.2. Wind Classification based on average wind power density according to [1]
Annual average Equivalent

Wind wind power density (W/m²) mean wind speed (m/s)
power 10 m 50 m 10 m 50 m 80 m
class elevation elevation elevation elevation elevation
1 0 - 100 0 - 200 0.0 - 4.4 0.0 - 5.6 < 5.9
2 100 - 150 200 - 300 4.4 - 5.1 5.6 - 6.4 5.9 - 6.9
3 150 - 200 300 - 400 5.1 - 5.6 6.4 - 7.0 6.9 - 7.5
4 200 - 250 400 - 500 5.6 - 6.0 7.0 - 7.5 7.5 - 8.1
5 250 - 300 500 - 600 6.0 - 6.4 7.5 - 8.0 8.1 - 8.6
6 300 - 400 600 - 800 6.4 - 7.0 8.0 - 8.8 8.6 - 9.4
7 400 - 1000 800 - 2000 7.0 - 9.4 8.8 - 11.9 > 9.4
The worldwide distribution of average wind speeds can only be given in a rough
way. The Figures 13.4 to 13.9 present an overview over the continents respecting the
important regions of the world. The data are based on data from sounding stations and
tower measurements from different sources [2].
Global averages and wind energy potential

The study quoted reaches the conclusion that the global average of the wind velocities is
about 4.6 m/s at 80 m altitude. The authors of the study link the global averages with a
deliberation on the usable wind energy potential. Accordingly, all regions which are
above wind velocities of class 3 and higher are combined and from this a wind energy
potential of 72 TW is calculated. If only 20 % of this were used, it would be sufficient
for covering the entire energy demand of the world. With reference to power generation
this would even be seven times as much (s.a. Chapt. 12.1).
North America
In North America, the best wind regions are concentrated on the North-Western and
North-Eastern coast. However, it is noteworthy that there are also extended regions
with relatively high wind velocities (classes 3 to 5) in the interior of the American
subcontinent from Texas in the South to Nebraska in the North. In Canada, the area
around Vancouver on the East coast and the coast from Labrador in the West having
above-average high wind velocities up to class 7 can be mentioned.
Fig. 13.3. Worldwide map of wind speed extrapolated to an altitude of 80 m and averaged over
all days of the year 2000 [2]
South America
The South American subcontinent has high wind velocities only in the extreme South
and at the coast of Chile and in the Caribbean area. In the interior, there are only some
regions in the South of Brazil and in the North of Argentina, respectively, which
have wind velocities of class 3 and can, therefore, also be considered for wind energy
utilisation.
Europe
Due to the great diversity of the continent the wind regimes are also very different.
There are two areas with completely different meteorological conditions, the northern
area with the migrating lows and highs and the Mediterranean area, where the wind is
driven by thermal effects between the Atlantic and the Mediterranean sea.
Very high wind velocities occur primarily in the North-Westerly coastal region of
France in the South up to Denmark in the North and along the British and Irish coasts.
There are numerous regions in wind power density class 7, with an annual average of
more than 9.4 m/s at a height of 80 m. In the Mediterranean zone, relatively high wind
speeds are found on the islands, for example on Sardinia and the islands of the Eastern
Mediterranean.
But there are also relatively large areas in the European interior which approach a
wind class of only 2 or 3 rating where the wind energy will be used. The availability of
large wind turbines in the megawatt power range with a tower height up to 150 m en-
able an economic utilisation.
Africa
Like South America, Africa is not a world region with high wind velocities. There are
areas in the extreme South and on the island of Madagascar which exhibit high wind
velocities. Furthermore the north–west coast of Morocco is considered as a promising
region for wind energy use.
Asia
There are virtually no large regions with very high wind speeds in the entire Asiatic area
from Iran all the way to China and including the Indian subcontinent. It is only at the
South-Western coast of China and on the Japanese isles that the wind velocities reach
class 7 in some cases.
Australia
In Australia, the regions with high wind velocities are concentrated along the Northern,
Eastern and South-Eastern coast and in the South of the Western coast. The wind speeds
in the interior are scarcely adequate for wind energy utilisation.
Fig. 13.4. North America [2]
Fig. 13.5. South America [2]

Fig. 13.6. Europe [2]
Fig. 13.7. Africa [2]

Fig. 13.8. Asia [2]
Fig. 13.9. Australia [2]

Offshore
In contrast to the average of 4.6 m/s on land the offshore average is 8.6 m/s. This means
that, on average, the offshore wind speed is about 90 % higher than the inland one. How-
ever, this information is generally only of limited significance to individual cases where
actual land-offshore reference site figures are of importance.
Figure 13.10 shows the global offshore areas of the great oceans. It is noteworthy
that there are considerable differences from summer to winter. In winter, high wind
speeds occur mainly in the Northern hemisphere, particularly in the North Atlantic, the
North Sea, but also in the Northern Pacific. These regions are, therefore, of particular
significance for offshore wind energy utilisation. In the summer months, the region with
high wind speeds shifts to the Southern hemisphere where wind energy utilisation can
scarcely be considered under current conditions.
Fig. 13.10. Offshore wind energy potential, winter (above) and summer (below) [3]
13.3 Wind Resources in Europe

For Europe, very detailed wind data, contained in the European Wind Atlas, have been
available for more than twenty years. The European Wind Atlas has been developed by
the Danish Research Centre of Risø with support by the European Commission. It is
based on the evaluation of the data from more than 200 measuring stations and a spe-
cially developed mathematical method by means of which the regional wind climate is
determined in a particular region. From the regional wind data, local "peculiarities"
(wind obstacles like buildings, etc., surface roughness and orology) are removed and
then the data are specified as wind conditions prevailing in the "region". From these
regionally valid wind data, the local wind data at a particular site can be determined (see
Chapt. 13.5.2).
Wind resources at 50 m elevation (mean wind speed m/s, wind power density W/m²)
Forest or urban areas Flat land Sea share Open sea Mountains
m/s W/m2 m/s W/m2 m/s W/m2 m/s W/m2 m/s W/m2
> 6.0 > 250 > 7.5 > 500 > 8.5 > 700 > 9.0 > 800 > 11.5 > 1800
5.0-6.0 150 -250 6.5-7.5 300-500 7.0-8.5 400-700 8.0-9.0 600-800 10.0 -11.5 1200 -1800
4.5-5.5 100-150 5.5-6.5 200-300 6.0-7.0 250-400 7.0-8.0 400-600 8.5-10.0 700-1200
3.5-4.5 50-100 4.5-5.5 100-200 5.0-6.0 150-250 5.5-7.0 200-400 7.0-8.5 400-700
< 3.5 < 50 < 4.5 < 100 < 5.0 < 150 < 5.5 < 200 < 7.0 < 400
Fig. 13.11. European Wind Atlas: Distribution of wind velocities at 50 m altitude [4]
13.4 CHARACTERISTIC PARAMETERS OF THE WIND 517
The map from the European Wind Atlas (Fig. 13.11) shows an overview of the re-
gional wind conditions in Europe. The dominance of regions with high wind speeds
along the northern European coasts is obvious. However, there are also limited regions
with high mean wind velocities in the Mediterranean area, for example in Spain, in
Southern France and, above all, on the Greek Isles.
The wind resource maps of the European Wind Atlas show the mean annual wind
speed at 50 m height, divided into five zones. In addition, the data for five roughness
classes from "open sea" to "protected terrain" are specified. Moreover, the wind atlas
specifies the mean specific annual energy content in watts per square meter rotor-swept
area.
Like the general climate, the wind conditions in Europe vary from a maritime climate
in Northern Europe and the British Isles to the continental climate in Central and East-
ern Europe to the Mediterranean climate. With special reference to the wind conditions,
there are two different areas:
- the northern area with distinct maritime low-pressure regions migrating from west to
east,
- the area in Southern Europe that is only partly reached by the migrating low pres-
sures and is greatly influenced by the thermal wind streams of the Mediterranean
area.
Further inland, topography situations and particularly the elevation play a decisive role.
13.4 Characteristic Parameters of the Wind

A knowledge of certain parameters and physical laws is of particular importance if the
energy of the wind is to be exploited. While the short-term behaviour of the wind, the
turbulence, is of significance with regard to the structural strength and the control func-
tion of a wind turbine, the long-term characteristics of the wind have relevance with
regard to the energy yield. The long-term characteristics of the wind can only be deter-
mined by means of statistical surveys over many years. These data are then used for
determining the energy yield of a wind turbine (Chapt. 14). At the same time, the dis-
cussion of these parameters presents a guide for procuring the required wind data for
wind turbine siting.
13.4.1 Mean Annual Wind Speed and Wind Speed Frequency Distribution
The mean annual wind speed, understood to be the "invariable" long-term mean value
of the wind speed at one location can only be determined on the basis of measurements
taken over decades. Since there are not many reliable measurements available for peri-
ods longer than 30 years, the measurements are limited to this period. It is, therefore,
quite possible that as the evaluating periods increase, the concepts of the long-term
mean have to be revised (s. Chapt. 13.3.3). Furthermore it cannot be completely ex-
cluded, that the change of the climate will influence the long-term mean value. However
518 CHAPTER 13 THE WIND RESOU
URCE
up to now there is no eviden nce that this will be the case according to serious scienntific
studies, regardless there aree often popular presumptions, that the annual wind sppeed
would become lower.
Knowing the mean annual wind speed is not enough to provide a precise energy cal-
culation. It also requires info
ormation on how frequently the individual wind speedds of
the spectrum can be statistiically expected. The frequency distribution of the annnual
wind speeds can be derived from data measured at a given elevation. The mean vaalues
for ten minutes are commonlly evaluated over one year and are then compiled in deffined
wind speed classes. To ach hieve a sufficiently reliable statistical basis, an evaluaation
period of at least several years, up to ten years according to meteorologists, is necesssary
to find the shape of the distrribution. A reliable information about the long-term annnual
wind speed needs a much lon nger statistical basis (see Chapt. 13.3.3).
The frequency distribution n is generally specified as relative frequency distributionn, or
as cumulative frequency (Fig. 13.12). The relative frequency distribution immediaately
indicates the occurrence of thet most frequent wind speeds. The cumulative frequeency
indicates as a percentage thee period within a year in which the wind speed falls beelow
the value of a certain point on
o the curve. By using the cumulative frequency, the m mean
annual wind speed can be accurately defined and represented geometriccally
(Fig. 13.13). Occasionally, the
t so-called median wind speed is used in literature. It is
defined as the wind speed wiith a cumulative frequency of 50 % and, as a rule, it is 0.3 to
0.5 m/s lower than the mean wind speed.
In practice, the problem is
i frequently that insufficient data about the frequency diistri-
bution of the wind speeds at a particular location are available. In such a case, there iis no
alternative but to use a matheematical approximation for the distribution curve. In norrmal
wind regimes, a Weibull functtion will provide a good approximation (Fig. 13.14).
Fig. 13.12. Wind speed frequenccy distribution for the island of Sylt (Germany), k = 2.0, meassured
at an elevation of 10 m
13.4 CHARACTERISTIC PARA
AMETERS OF THE WIND 519
Fig. 13.13. Definition of mean annual

a wind speed and median wind speed
The Weibull function is defin

ned by the following parameters:
1
where:
Φ = distribution function
n
e = logarithmic base (noormally the natural log, e = 2.781)
A = scaling factor
k = shape parameter
If nothing besides from the mean

m wind velocity is known and a "usual" frequency dis-
tribution can be assumed, thiis is characterised by a form factor of k = 2. In this speecial
case, the Weibull distribution n is called a Rayleigh distribution. The relative frequenccy is
obtained mathematically from m the cumulative frequency by differentiating with resspect
to the wind speed υW.
On real sites the shape paarameter k varies from about 1.5 to 2.5. A value aroundd 1.5
is typical for offshore sites orr on small islands. Over land the factor reaches values uup to
2.5 or somewhat above on veery specific sites. The annual energy yield of a wind turrbine
will be influenced by some percent
p (s. Chapt. 15.2). Furthermore it has to be consideered,
that the shape factor dependss on the increasing mean wind speed with height.
URCE
Fig. 13.14. Approximation of th he measured wind frequency distribution of List on the islannd of
Sylt by a mathematical distributtion function according to Weibull
13.4.2 Increase in Wind

d Speed with Altitude
One of the most important phenomena
p with respect to the utilisation of wind energgy is
the increase in wind speed with
w altitude. The friction of the moving air masses agaainst
the earth's surface slows dow wn the wind speed from an undisturbed value at great alti-
tude (geostrophic wind) to zero
z directly at ground level. Depending on the time off day
and atmospheric conditions,, the range up to where the wind speed is undisturbeed is
between 600 and 2000 m above ground and is called the “atmospheric boundary laayer”
(Chapt. 13.1).
The area of the boundarry layer close to the ground is called the Prandtl laayer
(Fig. 13.15). The flow cond ditions in this area are dominated by the friction of thee air
flow against the earth's surfaace. In addition, the vertical transportation of heat and im-
pulse in this area plays an im
mportant role.
The area lying above the Prandtl
P layer is called Ekman layer in meteorology. Thee in-
fluence of ground friction iss already less dominant here, whilst the wind directioon is
influenced by the Coriolis forces, which generate the so-called Ekman spiral.
The height of the Prandtl layer varies with the meteorological conditions. Duringg the
nocturnal hours, it is only 200 to 50 m thick whereas during the day, the vertical exxtent
is, between 50 and 150 m. Investigations have shown that, for example, a rotor hub
height of 60 m is within th he Prandtl layer for only about 30 % of the annual hhours
whereas it is only about 7 % at a hub height of 100 m. The wind conditions of llarge
turbines with 100 m and more hub height are, therefore, already extensively influennced
by the characteristics of the Ekman layer. This fact was scarcely considered at the be-
ginning of wind energy utilisation and is one reason why the simple, older mathematical
approaches like the logarithmic height formula for the increase in wind velocity with
height were too inaccurate. As a rule, they underestimated the wind velocity above 100
m altitude, an effect which occurred mainly at inland sites.
Fig. 13.15. Atmospheric boundary layer with Prandtl layer and Ekman layer [5]
The instantaneous increase in wind speed with elevation in the lower atmosphere de-
pends on a number of meteorological factors, e.g., temperature layering and humidity.
These largely determine the so-called atmospheric stability. However, the mean value to
be expected statistically over a long term at a particular height is largely determined by
the roughness of the earth's surface (Chapt. 14.6.6). The roughness of the earth's surface
is defined by the so-called roughness length z0 that is specified in metres (Table 13.16).
The increase in wind speed with height can be described as the statistical mean of an
assumed steady-state speed distribution. This simplification is adequate with respect to
problems based on the long-term statistical mean of the wind speed, that is to say the
calculation of the energy yield of a wind turbine. Naturally, instantaneous fluctuations that
can be of significance for certain problems of the strength calculation, for example of the
rotor blades, are superimposed on this mean (Chapt. 6).
Table 13.16. Roughness lengths and roughness classes for various surface characteristics [6]
A common approach used for describing the increase in wind speed with height is the
above-mentioned logarithmic height formula:
ln
·
ln
where:
H = mean wind velocity at elevation H (m/s)

= mean wind speed at reference elevation Href (m/s)
H = height (m)
= reference elevation (measuring elevation) (m)
ln = natural logarithm (base e = 2.7183)
Strictly speaking, the validity of the logarithmic height formula is restricted to the
Prandtl layer close to the ground. There are numerous approaches to improving the
accuracy of this formula and taking into consideration, for example, the influence of
atmospheric stability or taking into account the fact that the increase in wind speed with
altitude also indicates a dependence on the wind speed itself. To be able to handle the
more detailed method in practice, however, several parameters, unknown as a rule, must
be estimated so that it is questionable whether better results are achieved in every case.
A comparatively simple description of the increasing wind speed with altitude is
the power law approximately according to Hellman. This formula is sufficient for many
engineering tasks:
where:
H = mean wind velocity at elevation H (m/s)

= mean wind speed at reference elevation Href (m/s)
H = height (m)
= reference elevation (measuring elevation) (m)
α = Hellmann's exponent (-)
The correlation between Hellman's exponent and the logarithmic formula can be calcu-
lated in approximation by using the formula:
1
ln
With respect to the calculation of the energy delivered by a wind turbine, attention
must be paid to the fact that the logarithmic height formula and, even more so, Hell-
man's exponent, often supply inaccurate values for greater rotor heights (more than 80
m). As a rule, the mean wind speed is underestimated at greater hub heights. This is
particularly important for the assessment of the wind speed on inland sites with higher
values of the roughness length (Chapt. 13.5.3).
13.4.3 Steadiness of the Wind

From the beginning, the unsteadiness of the wind has been a frequent argument against
the utilisation of wind energy. Indeed, the available wind is less steady than, for exam-
ple, direct solar irradiation but this has the considerable disadvantage of disappearing
virtually completely at night. The variation in wind flow is essentially determined by
two factors: the latitude of the site on the globe and the surrounding distribution of land
and water.
In medium continental altitudes, the wind fluctuates greatly as the low-pressure re-
gions move through. In these regions, the mean wind speed is higher in winter than in
the summer months. The proximity of water and of land areas also has a considerable
influence. For example, higher wind speeds can occur in summer in mountain passes or
in river valleys close to the coast because the cool sea air flows into the warmer land
regions due to thermal effects. A particularly spectacular example are the regions of the
passes in the coastal mountains in California through to the lower lying desert-like hot
land areas in California and Arizona (see Chapt. 2.5).
Having regard to the reliability of energy yield of a wind turbine, it is primarily the
long-term fluctuations of wind speed which are of interest as already mentioned. They
are expressed as lines of diurnal variations, as seasonal variations and as fluctuations of
the mean annual wind speed over relatively long periods.
Diurnal Variation
A distinct, periodic change in wind speed over the diurnal period of 24 hours occurs
when thermal effects play a role. This is very known at the sea shores, where the direc-
tion and the intensity of the wind changes from day to night. Thus, for example, higher
wind speeds only occur around noon in the aforementioned regions in California. It
takes until the noon hours before the land behind the coastal mountains is heated up.
Then the cool air is drawn from the Pacific over the mountain passes. This wind contin-
ues until late at night.
This diurnal characteristic is of significance if the demand curve follows the power
consumption or if it has the opposite characteristic. In California, the prevailing wind
matches the characteristic of the power demand quite well which is dominated by room
air-conditioning in this area. Under these conditions, the utility companies offer tariffs
which vary with the time of day and are advantageous at noon in this case. In Northern
and Central Europe, on the other hand, there are no such distinct fluctuations in the
diurnal wind speed (Fig. 13.17).
13.4 CHARACTERISTIC PARA
AMETERS OF THE WIND 525
Seasonal Fluctuations
The seasonal change in wind d speed is a fact which is generally known. As already m
men-
tioned, it is highly dependen
nt on the geographic location on the globe. The line off an-
nual variation of the mean monthly
m values for List/Sylt shows that the winter monthhs to
March have the highest wind d speeds whereas the second peak in spring is less disttinct
(Fig. 13.18).
Fig. 13.17. Line of diurnal variiation (mean hourly values) of wind speed in List/Sylt meassured
at 12 m height (1971-80) [7]
Fig. 13.18. Monthly mean vallues of the wind speed for List/Sylt measured at 12 m heeight
(1971-80), mean annual wind sppeed 7-l m/s [7]
Long-term variation
The diurnal and seasonal fluctuations in wind speed are of little significance for grid-
coupled wind turbines. The fluctuations in the long-term mean annual wind speed from
year to year, on the other hand, are quite important. Wind turbines are investments
which must be financed from their income over relatively long periods (up to 20 years)
(Chapt. 20).
The investor, therefore, has to know the long-term mean annual wind speed as accu-
rately as possible as a basis for his calculation of economic viability. But the fluctua-
tions in energy yields from year to year can also cause considerable problems in the
financing if liquidity is at a premium. The problem becomes even more acute due to the
fact that the delivery of energy by a wind turbine, roughly speaking, varies with
the cube of the mean annual wind speed. Changes in wind speed appearing to be rela-
tively small, therefore, cause a drastic change in energy yield, and thus in the revenues.
This problem had been completely underestimated in the initial period of commercial
wind energy utilisation and entered properly into the consciousness of the investors only
with the succession of years of low winds from about 2000 to 2006. Since then, wind
forecasts are rightly being looked at much more critically and extend to speculations
about the change in wind conditions as a consequence of the general climate change.
All wind forecasts are based on the prediction of a long-term mean value which is
based on experience or, even better, on past measurements. Meteorologists point out
that the observation period should be at "least 30 years" in order to be able to make a
reliable statement about the long-term mean value. The problem with commercial wind
energy utilisation is that there are practically never any reliable wind measurements,
taken over such a long period, available for the actual sites. The wind forecast, there-
fore, has to use "wind models" which are partly theoretical and partly empirical (s.a.
Chapt. 13.5.2).
Fig. 13.19. Relative energy content of the wind in Denmark (5-year averages) over a period of
one hundred years [4]
There are only few reliable data about the mean wind speeds of a site or of a region over
periods of decades available. In Figure 13.19, the relative energy content of the mean wind
speed is plotted over one hundred years for Denmark. The curve is based partly on meas-
ured data for the last few decades, but partly also on an estimate of the wind speed from
other climatic data. Independently of the lacking long-time annual means of the wind
speed, the derivation of which was obviously considered to be too unreliable, the curve
shows the range of fluctuation of the energy content. This is ± 20-25 % around the 100 %
value of the period shown.
Fig. 13.20. IWET operator database according to regions [8]
In recent years, the problem of annual and regional fluctuations in Germany has been
described by the Wind Index. Various organisations and private companies keep statis-
tics about the energy delivered by the existing wind turbines and publish the fluctua-
tions around the 100 % value of the observation period on this basis. The wind index is
a "production index" and not a "wind index" in the strictest sense. The most well known
one is the "IWET Wind Index" [8]. This is based on the monthly averages of the wind
turbines evaluated in 25 regions in Germany (Fig. 13.20). The Index is published for
these regions and shows the deviation from the 100 % value. The statistics have been
kept systematically since about 1989 and thus cover a period of about 20 years. In the
interpretation of the results, two factors are of importance. The recorded energy yield of
the existing wind turbines depends not only on the wind conditions. The result is also
influenced by the efficiency and availability of the turbines. The persons processing the
database point out that they introduced numerous corrections for purging the data of
such influences. The second, grave objection is that the period covered by the statistics
is not long enough in the meteorological sense. The reference level (100 % value) had to
be corrected several times, therefore, as the wind index progressed step by step. As long
as the observation period does not approach the meteorological demand of "at least 30
years", the reference value cannot be interpreted as a long-term, meteorologically sub-
stantiated mean value. In other words: A deviation of the reference value for a period of
15 years from the assumed long-time annual mean, for example from a wind forecast, is
no proof for the incorrectness of the meteorologically substantiated long-term value.
Nonetheless, the wind index is a valuable aid for planning tasks and for the assessment
of economic viability - provided one knows its shortcomings (Fig. 13.21).
Fig. 13.21. IWET Wind Index for Region 6 up to 2006
When considering the IWET statistics from 1989 to 2009, it is noticeable that there
are fluctuations of ± 20-24 % around the reference value even for this relatively small
period. Naturally, this small basis does not provide proof for the range of fluctuations
also derived from Figure 13.18, but it is at least an indication. It is also noticeable here
that there appears to be a cycle of succession of above-average and below-average wind
years. The low-wind years after 1996 are contrasted by the above-average years from
1987 till 1995. Statistically, periods of five to seven years with a sequence of above- or
below-average wind years can apparently be expected. Wind energy utilisation must be
able to cope with the economic aspects of this characteristic of the wind and the resultant
delivery of energy.
Apart from the wind indices normally used in Germany, there have for some years
been attempts of also developing meteorologically supported statistics about the annual,
regional and global fluctuations in the wind speed for larger areas, i.e. not about produc-
tion data of wind turbines. One basis used for this purpose are global weather data
which are continuously acquired by the NCER (National Center of Environmental Re-
search) and the NCAP (National Center for Atmospheric Prediction) in the US. These
are partly data from satellite measurements. Weather models are applied to data already
observed and are calibrated using a "reanalysis method". Using the mathematical mod-
els thus developed, wind forecasts are then created. These elaborate methods are applied
where the forecast cannot be based on long-term empirical values from the existence of
wind turbines and the database for determining the wind speed according to the Euro-
pean wind atlas is also lacking [6].
In connection with the observations on the steadiness of the wind, it must also be
mentioned that the distribution of wind direction to be expected statistically is also sub-
ject to fluctuations [7]. For example, east wind occurs more frequently in one year than
in another one. However, apart from exceptional cases, the fluctuations have little influ-
ence on the energy yields so that they are not analysed in detail at this point.
13.4.4 Wind Turbulence and Gusts

Whereas the long-term fluctuations in wind speed are of significance for the power
output and energy yield of a wind turbine, the loads are marked by the short-term fluc-
tuations in wind speed, the wind turbulence and the wind gusts. The ever-present turbu-
lence provides a significant contribution to material fatigue, especially for the rotor
blades. The extreme wind speeds, which occur more rarely, must also be taken into
consideration for the fatigue strength and, moreover, can increase the loads up to the
breaking point (s.a. Chapt. 6.7).
The study of a wind speed time history measured with sufficiently high resolution
enables its most important parameters to be defined (Fig. 13.22). Ignoring short-term
fluctuations, the level of the prevailing wind speed determines the mean wind speed υW .
It is generally averaged over a period of 10 minutes. Using this steady mean wind
speed, the instantaneous wind speed at a point in time t can be specified as follows:
The superimposed fluctuating part of the wind speed (t) is caused by the turbulence of
the wind. Thus, turbulence is the instantaneous, random deviation from the mean wind
speed. The extent and characteristics of the turbulence are dependent on a variety of
meteorological and geographic factors which are described in the relevant meteorologi-
cal literature.
Since a time segment of wind speed, having the corresponding resolution, is always
limited, the wind turbulence can be covered completely statistically only by using a
spectral representation. These representations contain in most cases the energy content
of the wind speed fluctuations in dependence on the frequency of their probability of
URCE
occurrence. The Kaimal Spectrum, the Von Kármán Spectrum and the spectrum
m by
Davenport are frequently useed representations.
Fig. 13.22. Measured time histo

ory of wind speed [5]
Fig. 13.23. Energy spectrum of wind speeds (turbulence spectrum)

Although the spectra by Davenport and Von Kármán reflect the conditions in the free
atmosphere with less accuracy, they are used more frequently because they are easier to
handle (Fig. 13.23). The calculation of the fatigue strength of the structure is mainly
based on the spectral representation of turbulence (see Chapt. 6.7.2).
To characterise the turbulence, the term of turbulence intensity is used which is oca-
sionally also called the “degree of turbulence”. The turbulence intensity σ0 is defined as
the ratio of the standard deviation σv, of the wind speed to the mean wind speed in a
certain averaging time and is specified in percent:
The turbulence intensity changes with the mean wind speed, with the surface roughness,
with the atmospheric stability and with the topographic features. The lowest values are
measured over the open sea (5 % and less) whereas the highest values (20 % and more)
occur over densely settled areas or forest areas. In the load assumptions for wind tur-
bines, values of between 16 and 18 % are assumed depending on wind turbine classes
(see Chapt. 6.3.1).
In the open air, the wind turbulence does not occur in the "one-dimensional" form as
idealised here. Instead, the wind fluctuations are distributed spatially in all directions.
For this reason, meteorology has very complex models for the spatial multi-dimensional
turbulence of wind [1]. However, these multi-dimensional turbulence models are of no
great significance in wind power technology.
Apart from the relatively very frequent fluctuations, "considerable" deviations from
the mean wind speed within a range of a few up to some ten seconds and occurring
occasionally can be observed. These peaks are called gusts. There is no generally rec-
ognized definition but it is largely accepted to classify the gusts by means of a gust
factor. According to Frost, a gust is the increased wind velocity, averaged over a certain
time, referred to the mean wind velocity. In wind energy technology it is normal prac-
tice also to call a sudden decrease in wind speed from the mean value a negative gust. In
the load assumptions for wind turbines, idealised forms of gusts are assumed which are
specified as a load with a defined probability of occurrence in the structural design (s.a.
Chapt. 6.7.2). The relevant literature contains notes on the probability of occurrence, the
period and the spatial extent of the gusts [9].
In this connection, the question also arises for the highest wind speeds occurring,
which are gust-like in their extreme values and are often called “gust of the century”.
The meteorological literature contains reports about the extreme values of the maximum
wind velocities measured. According to these reports, extreme values of up to 47 m/s
were registered over a period of 20 to 30 years in the North German coastal region. This
value is exceeded above the open sea and reaches maximum values of over 60 m/s. In
exposed geographic locations, including the Antarctic, values of 95 m/s are said to have
been observed.
Against this background, the extreme wind velocities have been specified for four
different wind turbine classes and one specific class for offshore siting according to IEC
in the Load Assumptions for Wind Turbines (see Chapt. 6.3). In extreme locations indi-
vidual assumptions have to be made.
URCE
13.5 Topography and

d Local Wind Flow
When considering global wind w resource maps and fundamental laws, it must noot be
forgotten that at any locationn, the prevailing wind regime will exhibit local peculiarrities
which can be of decisive sig gnificance to the siting of a wind turbine. The smallerr the
wind turbine and the more the t local topography, including trees and buildings (obbsta-
cles), differs from the ideal "Hat and obstacle-free terrain", the greater the importancce of
the local orologic relief of th
he immediate surroundings. It is, therefore, useful to geet an
idea of whether the local terrrain can be considered to be "obstacle-free". Accordinng to
Frost, the terrain in the surro
oundings of a wind turbine can be considered to be "flaat" if
(Fig. 13.24):
- differences in elevation do not exceed 60 m within a radius of 11.5 km,

- the ratio of the maximum
m level difference hc to the horizontal distance betweeen
these two marked points is
i less than 0.032 for a distance of 4 km upwind and 0.8
km downwind,
- the height of the rotor relatiive to the lowest point within a distance of 4 km upwindd is
at least three times greater than
t the largest existing level difference hc.
Naturally, the existence of leevel differences can also be utilised in a positive sense w
when
a wind turbine is to be locateed in the area. Wind speeds on mountain ridges with a par-
ticularly advantageous shapee (slopes of 1:3 to 1:4) can increase to double the vvalue
prevailing far away from the ridge [9].
Fig. 13.24. Definition of "flat co

ountryside" in the environment of a wind turbine [9]
13.5 TOPOGRAPHY AND LOC
CAL WIND FLOW 533
Apart from differences in elevation in the terrain, the topographical environment of a

wind turbine is characterized d by the "roughness" of the earth's surface and by the ppres-
ence of "obstacles". The rou ughness of the earth's surface is brought about by evenlly or
randomly distributed surface characteristics, primarily by the type of vegetation (forests,
meadows etc.) or by the diffferences caused by land and water. The European W Wind
Atlas contains extensive nottes and computational methods for determining the rouugh-
ness length for certain types of
o surface [4].
While roughness determin nes the characteristics of the wind regime in a larger aarea,
obstacles only have a limited d local significance. Buildings, trees or groups of trees cre-
ate turbulences which can haave highly undesirable consequences for the operation and
lifetime of a wind turbine. The flow wake caused by a building is shown in Figgure
13.25. The airflow on the leee-side of such an obstacle is separated to approximaately
twice the obstacle's height and is more or less turbulent (separation bubble). On
the down-wind side the distu urbed air flow extends to a distance of up to twenty tiimes
the height of the obstacle. Iff a negative influence on the wind turbine is to be avoiided,
the rotor should be placed ata three times the height of the obstacle and sufficientlyy far
away down-wind.
The relationship between the orology of the environment and the local wind reggime
is extremely complicated. Numerous
N attempts have been made to develop theoreetical
modelling concepts and calcculation methods [11] (see Chapt. 13.5.3). Practical expperi-
ence and observation also plaay a not inconsiderable role, especially in this field.
Fig. 13.25. Flow wake behind an

n obstacle [9]
13.6 Measuring and Determining the Wind Speed

The potential user or planner of a wind turbine is frequently faced with the situation of
not having any data on the wind regime at the intended location. There are basically only
two possibilities for procuring wind data. Either a theoretical method as described, for
instance, in the European Wind Atlas will be applied, or measurements will be taken at
the site. It almost goes without saying that, whenever possible, both methods should be
used in parallel so that the results can be compared.
13.6.1 Measuring Techniques

Measuring the wind on site is the most reliable method, but frequently the possibilities
and results of such measurements are linked with unjustified expectations, or measure-
ments are taken the results of which are insufficient for providing correct and adequate
answers to the questions which are really of interest. The question therefore arises:
What can the user of a wind turbine achieve by carrying out his own measurements?
The energy yield of a wind turbine can only be predicted with statistically validated
values of mean wind speed, wind speed distribution and the vertical wind profile. Statis-
tically reliable values require long-term measurements, though. Generally, specification
of a reliable value for the long-term mean annual wind speed requires the mean value
taken over at least thirty years. This means, however, that the user must rely on the
literature published by meteorological institutes and organisations. Measurements taken
by oneself over a period of only a few months with simple equipment are not suitable
for this purpose.
On the other hand, the areas covered by the wind maps are usually so large that the
data are not transferable to specific local situations, particularly not in a complex ter-
rain. Mountainous topographies or particular building developments can cause consid-
erable local deviations from large-scale data. These uncertainties can be eliminated by
short-term measurements.
Wind measurement over a relatively short period, for instance one year, offers the possi-
bility of comparing these values with the long-term data measured within the same period
of time at the closest location at which the long-term mean value is known. This permits
verification of whether and to what extent the local value deviates from the value ascer-
tained for the larger area. If, in addition, the wind speed is plotted over time, indications as
to the turbulence intensity can be derived from the wind variations. From the point of view
of the prudently planning operator, wind measurements having this objective make sense
and may even be necessary.
If several wind turbines or even a large wind farm is to be installed, the question of
the prevailing wind direction or even the wind direction distribution arises. Efforts will
always be undertaken to situate the wind turbines according to the prevailing wind
direction (Chapt. 18). However, these measurements are highly time-consuming, so that
frequently qualitative information relating to the prevailing wind direction has to be
relied on. The growth pattern of trees and bushes is an especially reliable indicator in
this respect (Fig. 13.26).
13.6 MEASURING AND DETERMINING THE WIND SPEED 535
Fig. 13.26. Wind-flagged trees as qualitative indicators of the local wind resources
In addition to the planning of the installation of wind turbines, their operation also re-
quires some sort of continuous wind measurement. Although almost all large turbines
have their own operational wind measuring system, additional information about the
wind speed is still frequently of interest. The operator wants to compare the actual energy
yield with the theoretically possible value. This already requires an additional wind
measuring system, which must be sufficiently clear of the turbine to remain unaffected
by the rotor (Chapt. 10.1.1). In some cases, especially in large wind farms, turbine-
independent wind measurement is also used for power output management and supervi-
sion. Thus, for example, it is desirable for various reasons to perform the shut-down and
start-up sequence in groups. For this purpose, the signals from a turbine-independent
measuring point are used.
What instruments and measuring methods are suitable for this type of wind meas-
urement? An instantaneous "measurement" of the wind speed with a conventional cup
anemometer by hand may on occasion be quite illustrative, but its value for technical
purposes is virtually zero. Usable data can only be obtained by means of measurements
over a certain period and recording of the measured values. This requires a stationary
measuring system on a mast with a logging device for the measured data. The sensors of
a wind measuring system suitably consist of a combination of an anemometer and wind
vane or, most recently, of an ultrasonic sensor which has no moving parts (Fig. 13.27).
Time and again, the accuracy of the sensors becomes the subject of discussions about
the quality of the measurement data which is why considerable importance must be ac-
corded to the calibration of the anemometers. The anemometers must be checked and
recalibrated at regular intervals. Today, independent institutions will provide anemome-
ter calibrations and an operator is well advised to pay attention to these.
Fig. 13.27. Ultrasonic anemometer (Ammonit)
The sensors are mounted on top of or on side arms of a mast or a tower. The mast height
depends on the requirements to be met. If only a comparison with the data in the wind maps
is intended, then the standard measuring elevation of 10 m is enough. With larger wind
turbines, the problem then arises of having to extrapolate the wind speed to the rotor's hub
height. If there are too many uncertainties concerning wind shear, nothing remains but to
build the wind measuring mast as high as the rotor's hub height. However, a wind measur-
ing mast with a height of fifty or more meters represents a cost factor and may require a
building permission (Fig. 13.28).
As the size of the wind turbines increases, measuring the wind at rotor hub height by
using an anemometer mast becomes almost impossible. The operator must rely on other
methods. Occasionally, special radar methods are used today for determining the wind
velocity at a greater height, for example SODAR (Sonic Detecting and Ranging).
In the past, mechanical anemographs were used which plotted wind speed and direc-
tion on a paper strip. Mean values were determined with the help of a plotting rule.
Nowadays, electronic recorders are used almost exclusively where the measured data is
stored on tape or in chips. This provides for immediate data evaluation by computer.
The manufacturers offer an almost endless variety of suitable storage devices and ana-
lysers.
In recent years, specific data loggers, sometimes called wind classifiers, have been
developed especially for wind turbines. They count the duration of the wind speed
within certain speed ranges (classes) and thus provide direct information on the wind
speed frequency distribution (Fig. 13.29). Moreover, it is possible to obtain additional
information on instantaneous wind speed or the average wind speed reached in the past
month or the like.
Fig. 13.28. Anemometer mast with sensors (SIGGELKOW)
Fig. 13.29. Data logger for wind speed analysis (AMMONIT)

Logging the wind speed distribution, and thus also the average wind speed, over a
year by using such a data logger, in combination with long-term wind data which may
be available from neighbouring areas, represents a useful basis for obtaining a reliable
assessment of the energy yield of a wind turbine. Wind classifiers can thus supply valu-
able information for a wind turbine operator. However, they are no exclusive substitute
for long-term wind statistics.
The wind speed is indicated in meters per second in the SI system of units. Despite
this, the Beaufort scale of "wind forces" is still widely used in many places, and is per-
haps also more graphical (Table 13.30). Above all, the correlation between wind speeds
and visible effects is a valuable aid to assessing the wind regime, especially for a wind
power technician.
Table 13.30. Classification of wind speed according to the Beaufort Scale

13.6.2 Ascertaining the Wind Data and the Energy Yield from the
European Wind Atlas
In recent years, the European Wind Atlas has become one of the most important tools in
determining a site for wind turbines and predicting the energy yield to be expected. In
European countries, the usual "wind studies" are produced almost exclusively by this
method if they cannot be based on the evaluation of measurements at the site itself. The
calculations of economic viability, and thus the decision regarding investments for
many projects, rely on the information provided by this semi-empirical method. For this
reason, some fundamental notes relating to the methods involved and the reliability are
indispensable. The correct application of this method is assisted by very good documen-
tation including data sources and calculation methods provided on floppy disks which
are commercially available [4]. It is not difficult to handle but the method is not without
risk if it is used in a purely formal way, i.e. without adequate general experience in the
field of wind conditions and site factors.
The European Wind Atlas consists of two parts: the first part describes wind condi-
tions in Europe and the second part contains a mathematical method by means of which
the wind conditions and the energy yield of one or more wind turbines can be predicted
at a particular site from these data.
The first part was originally based on about 220 measuring stations from which
measurement data are available for a relatively long period (essentially from 1970 to
1980). These measurement data, in most cases measured at the standard meteorological
measuring height of 10 m, supply the raw data of the atlas. The measurement data in-
clude the local co-ordinates of the measuring station, the measuring height, the so-called
“roughness rose”, i.e. the information on environmental roughness in directional sectors
and the frequencies of wind speed and wind direction specified in the sectors. In addi-
tion, the database contains the diurnal and annual variation in wind speed. From these
data, the Weibull parameters A and k have been calculated for each directional sector.
From these raw data, the so-called regional wind climatology is determined by using
the geostrophic law of friction. The geostrophic law of friction is a fundamental theoreti-
cal approach to describing the wind conditions in the boundary layer of the earth's atmos-
phere. The forces resulting from the pressure gradients in the atmosphere are brought into
equilibrium with the frictional forces of the earth's surface, removing local influences like
orology, environmental roughness and obstacles from the local raw data of the measuring
stations and calculating the Weibull parameters (A and k) for "regionally" valid wind data.
These data then apply to "flat and even" terrain and "no shading by obstacles" and are
calculated for four different roughness classes. They are also specified in directional sec-
tors and, in this form, represent the wind conditions that can be applied to a region of
about 200 x 200 km. When extrapolating to greater heights, the wind data correspond to
the wind conditions of the geostrophic wind. In the wind atlas, the data thus determined
are represented at a height of 50 m, using maps and tables (s.a. Chapt. 13.2.1).
The second part of the wind atlas, the "Wind Atlas Analysis and Application
Programme"(WASP) contains descriptions on how the wind data for an actual potential
site for a wind turbine can be determined from the regional data. In this method, the
calculation of regional data from the basic local raw data of the measuring stations is re-
versed and the same physical and mathematical models are used. In practice, it is handled
in such a manner that a suitable station located in the vicinity of the site is selected from
regional wind data and then the site is classified in accordance with the criteria of orogra-
phy, surface roughness and shading by obstacles. For this purpose, the landscape is di-
vided into five different landscape types and four surface roughness classes are defined.
The roughness length z0 is then determined from, among other things, the so-called
"roughness elements" (e.g. large trees, houses, etc.). Using correction factors derived from
these, the site-specific data are calculated from the regional wind data, especially the pa-
rameters of the wind frequency distribution (Weibull distribution) at the hub height of the
rotor. Roughly speaking, the WASP program transforms the wind data of a point A via the
geostrophic wind to an arbitrary point B in the vicinity of A (Fig. 13.31).
The WASP computing model is being continuously improved and in its present
form has been very successful for open and flat coastal regions and also the level inland
regions. The results present problems in mountainous terrain since the classification of
the site in accordance with the above criteria becomes very difficult in this case and,
moreover, the wind conditions are influenced by microclimatical orologic situations.
A further inaccuracy is found in the calculation of the mean wind speeds at heights of
more than 80 m. The basic logarithmic height formula is only reliable in the lower area
of the boundary layer of the Prandtl layer which frequently only extends up to a height
of 60 m. The mean wind speed at a height of more than 80 m has often been considera-
bly underestimated when using the calculation according to WASP. In recent WASP
versions, a special correction factor is used for improving accuracy above 80 m height
(see Chapt. 13.4.2)
The deviation in the wind data determined is estimated by the publishers of the wind
atlas to be approx. 5 %, and the possible error of the mean wind power (energy yield)
is approx. 15 %. However, this very cautious statement should not lead to the conclu-
sion that a more comprehensive wind prognosis, that also includes comparisons with
existing measurements in addition to the calculation according to WASP, must have
these error tolerances. In the major regions of wind energy utilisation, for example in
Denmark or in the North German lowlands, there are by now so many reference points
available due to the increasing installation of wind turbines that wind and yield progno-
ses can be created with a much greater accuracy than that mentioned above, at least in
those areas, if this experience and information is correctly included. In the wind studies
it is common practice to use the wind turbines with vicinity, which energy yield is
known as a reference for adjusting the relevant parameters for the WASP calculation
regarding a new intended site. In this way the accuracy of the wind speed prediction can
be improved up to an uncertainty of only few percents.
The Danish Risø research institute offers a comprehensive programme package for
planning wind power projects on the basis of the WASP computer model. The various
software programs are combined under the trade name WINDPRO, not only for
determining the wind speed and the energy yield but also for a number of further
planning tasks like noise emission shadow effects or even more issues up to economic
calculations [10].
Fig. 13.31. Basic procedure for determining the wind speed for a site using the European Wind
Atlas (WASP) [4]
13.6.3 Numeric Models of Three-Dimensional Wind Fields

The penetration of wind energy utilisation into the inland regions with their much more
complex topographical conditions in comparison with the open and flat coastal regions
has made it clear that a more accurate determination of the wind data is necessary. The
widely used method of the European Wind Atlas which is based on the conventional
"wind study" or "wind assessment" shows distinct weaknesses at inland sites. These are
basically attributable to two reasons:
- The influence of the terrain relief (orography) on the local wind field is reproduced
only inadequately.
- The logarithmic height formula for the increase in wind speed with height, used in
the WASP method, loses its validity outside the Prandtl layer.
Against this background, numeric simulation models are increasingly used which, al-
though they are much more complex and, as a result, are also associated with high costs,
do provide more accurate results. Such models are used in many areas of meteorology
such as weather prediction, but also for assessing the spread of pollutants in the atmos-
phere. The basic concept of these simulation calculations is based on a digital three-
dimensional model of the orography including surface features. Onto this model, the
wind field of the geostrophic wind is superimposed, which is not influenced by relief
and surface features. The result is a three-dimensional wind field which reproduces the
influence of the shape of the terrain with its surface characteristics and manages without
extrapolating to a greater height the wind speed of measuring stations close to the
ground. Similar to the WASP method, a certain lack of certainty is given with the cor-
rect estimation of the nature of the surface. For this reason, a certain amount of experi-
ence is required when handling these models [11].
A model that has recently been used successfully for determining wind data, particu-
larly in inland regions, is known by the name "FITNAH". This is a so-called mesoscale
model which can be used with a mesh size (spatial resolution) of between 25 and 50 m.
Figure 13.32 shows the result of a wind field simulation for a region under examination
with a size of 10 x 10 km. The annual mean of the wind speed is represented in a 200 m
grid pattern. Using this information, the choice of site for wind turbines can be adapted
very precisely to the considerable differences in wind conditions in the field under ex-
amination in a complex mountainous terrain, or the differences in energy yield of the
individual turbines can be calculated if their sites have already been established.
The more accurate determination of the mean wind speed at rotor hub heights of
more than 80 m is an important aspect which speaks for the use of such simulation
models (see Chapt. 13.4.2). The models do not use the logarithmic height formula and,
therefore, are not tied to the validity of this approach being restricted to the Prandtl
layer. The economic significance of a precise knowledge of the mean wind speed at
heights of more than 100 m for the utilisation of wind energy in inland regions is not to
be underrated. The usual rotor hub heights are between 80 and 100 m there and will
exceed the 100 m limit with the next generation of large turbines.
13.7 ABOUT THE WIND ENERGY POTENTIAL 543
Fig. 13.32. Simulated spatial distribution of the annual mean wind speed at a height of 98 m
above ground level for a mountainous inland site using the FITNAH model [11]
13.7 About the Wind Energy Potential
The technical feasibility and reliability of power generation by wind energy can no
longer be seriously questioned today. At sites with suitable annual wind speeds the
economics are also no longer an issue. Today, these facts must be acknowledged even
by the critics of wind energy utilisation.
Supporters of a practically unchanged continuation of the traditional energy-supply
structures, therefore, increasingly use the supposed lack of macro-economic perspective
of wind energy, and of renewable energy sources as a whole, as an argument. The re-
newable energy technologies are praised for their usefulness in certain market niches
only to explain in the same breath that, of course, "wind and sun" could under no cir-
cumstances replace the reliable, conventional coal and nuclear power stations. The con-
tribution of the "renewables" would basically remain restricted to "a few percent".
Against this background, the question of the potential of wind energy arises, particu-
larly in the political discussions about future energy sources. This is essentially of no
importance to individual users. They will use a technology which has been proven, and
is economic, regardless of whether its potential is high or low, as long as the overall
economic conditions give them a chance of acting within a useful economic margin. In
power economics in particular, however, structural conditions have been set from politi-
cal points of view and they decide about the economic scope for action by the individual
consumer.
In the physical sense, potential - the ability to perform work - is a state variable
which can be accurately described and thus does not need to be interpreted. In the case
of wind energy it has been pointed out that about 2 % of the solar energy captured by
the earth's atmosphere is converted into motion of the air masses. Mathematically, this
yields a power (power because the energy is quantifiable only over a certain period of
time) of approximately 4 x 1012 kW. This is one hundred times more than the total
power-plant capacity installed on earth (s. Chapt. 13.1). Naturally, the exploitable po-
tential is very much smaller, but all other restrictions are firstly of a technical nature and
then of an economic nature, and are ultimately a question of the social consensus about
the importance to be accorded wind energy utilisation. Apart from the basic physical
value mentioned, all other figures about the potential of wind energy thus always
amount to saying "if it is assumed that ... then a potential of ... is obtained". Controver-
sies are caused less by the results than by the assumptions on which they are based.
Nearly all serious studies about wind-energy potential arrive at the result that under
realistic conditions, the wind energy potential has the order of magnitude of "hundreds
of thousands of megawatts". However, one cannot expect unambiguous and reproduci-
ble results from these investigations. One must be aware that, as has been mentioned
before, the technically usable wind-energy potential is not a physically defined state
variable and hence precise and undisputed figures cannot be expected. The assumptions
on which the numerical values are based allow a wide range of results. Thus, an assess-
ment of these figures can only be carried out by evaluating the assumptions made. By
comparison, gaps still existing in the data on the wind regimes are of much lesser
significance. Regionally, unreliable wind data can quite easily lead to an erroneous
assessment of the wind's potential, but this has no significant influence on the global
resources. The attempt to compile the results of the various estimates in conjunction
with the preconditions on which they were based in an overview produced the follow-
ing categories concerning the potential of wind energy:
The atmospheric wind-energy potential

As mentioned earlier, the atmospheric potential of wind energy exceeds the globally in-
stalled power-plant capacity by two orders of magnitude. It is true that, from an eco-
nomic point of view, this fact is not of interest, nevertheless it should not be completely
ignored. This energy potential does indeed exist and if it were possible to use even less
than one percent of it, all energy problems with respect to the generation of power
would be solved.
Basically, there are neither physical nor technical obstacles precluding the utilisation
of this potential. For example, the exploitation of wind energy at an altitude of one
thousand meters above the open sea is no further away from current technological real-
ity than the utilisation of nuclear fusion by commercial utilities. The decisive difference
as to why vast sums of dollars are being invested in the research into nuclear fusion and
not, e.g., wind power is probably only due to the fact that the established scientific and
technical lobby for nuclear physics and its attendant industry has enough political influ-
ence to force through these immense expenditures.
13.7 ABOUT THE WIND ENERGY POTENTIAL 545
The economically usable wind-energy potential

The currently available technology can only be used for exploiting the atmospheric
wind-energy potential economically on land and in shallow coastal waters of the seas
where the mean annual wind speeds exceed 4 m/s at 10 m height, and where the gener-
ated electricity does not need to be transported to the consumer over continental dis-
tances. Moreover, wind energy utilisation is only feasible up to a rotor hub height of
180 or 200 meters. In other words, technical and economical utilisation is only possible
at sites where wind turbines of currently known designs and dimensions can be set up.
According to this criterion alone, the usable wind-energy potential for example of the
European Union is still immense. The installable power is definitely of the order of
several hundred thousand megawatts, corresponding to a generated energy volume of
the order of one billion kWh. This is more than half the total energy consumption of the
EU in 1991 (approx. 1.8 billion kWh).
It is possible to extract such quantities of energy from the wind from a technical point
of view alone. The economic aspects are by no means utopian, either. In areas with wind
speeds exceeding 4 m/s (at 10 m height), the costs of power generation are of the level of
current production costs of conventional power stations. Economically viable power
costs are still achieved in areas with weaker wind regimes.
Utilising wind energy to such an extent would, of course, not be possible without
profound changes in the structure of the technical energy production of today. The prob-
lem of a firm power supply would have to be solved. Without a sophisticated combina-
tion of the most varied solar generation units and possibly even retaining considerable
sections of conventional generation plants, firm power could only be guaranteed by an
extremely expensive storage technology, given the current state of technology. In the
long run, energy storage via hydrogen generation could represent an economically vi-
able solution to this problem.
From the point of view of the energy politics of the utility world, the scenario de-
scribed is, of course, nothing short of a revolution, but it is far from being technologi-
cally utopian, however frequently this is asserted. What is utopian are not the technical
possibilities, but breaking with outdated traditional priorities in the energy industry and
politically accepted concessions to technological and economic developments.
The politically acceptable wind-energy potential

All "serious" estimates of the potential of wind-energy start out by listing areas where
the annual mean wind speed reaches the minimum value considered necessary
economically, and then continue with a long enumeration of restrictions preventing the
siting of wind turbines as, for example:
- Regions of residential settlement

- Landscape conservation areas
- Wildlife reserves
- Traffic routes
- Protected biotopes
- Regions of special tourist interest

- Transmission lines
- Microwave links, transmitters and directing lights
- etc.
According to the presentations of the relevant interest groups, these regions should
remain free of wind turbines and, moreover, large "safety distances" maintained, if
possible. The areas remaining are considered acceptable for the siting of wind turbines
and on this basis, the "wind energy potential" is then calculated, including the techni-
cally required minimum clearances between the turbines. It is amazing that even in
densely populated areas, there is still considerable remaining potential, but this is natu-
rally only a fraction of the technically usable potential. In other words: all existing
claims on land utilization are considered permanent and immutable. The utilization of
wind energy is relegated to what niches remain.
What if these standards had also been applied to the extension of traffic routes, espe-
cially road construction, or to the exploitation of conventional primary fuels? Coal min-
ing did not only spoil entire landscapes but, in the case of open-cut brown coal mining,
completely destroyed them. The resettlement of entire villages is only one of the more
humane side-effects in this context. The quality of life of the people living in the mining
districts has been quite poor for more than a century, a fact which has been and still is
tolerated and accepted by society as the unfortunate but necessary tribute to be paid for
a technical civilisation.
Whosoever demands even the relocation of a microwave radio link for the benefit of
wind energy utilisation is regarded as a starry-eyed dreamer. This illustrates the status
which society is willing to grant an ecologically oriented energy supply system. Verbal
declarations cannot obscure the fact that emission-free and sustainable energy generation
occupies the last place on the "list of priorities". It is only when all other fields of interest
remain untouched, from the optimal position of microwave radio links to the preservation
of a "natural" landscape, that society as represented by the vested interests is prepared to
permit the installation of wind turbines.
The decision about how much wind-energy potential is available is reached virtually
without considering other aspects, only these. The natural preconditions and the techni-
cal and economical feasibility do not impose any real limits on the development of
wind-energy utilisation. Neither do uneconomically high costs present an obstacle to
wind-energy utilisation. It is important to keep these facts in mind. The statement by H.
Scheer on the utilisation of solar energy can be applied equally to wind energy [12]:
"The question, frequently asked, of how high a percentage of the supply of energy could
be contributed by solar energy utilisation really does not make sense: as the potential of
solar energy is more than enough for the energy requirements of humankind, there is
also no limit to the usable proportion of solar energy. The magnitude of the proportion
of solar-energy is a question of input alone: the more political initiatives and economic
investments there are, the higher the proportion."
What positive conclusions can then be drawn from this situation? It would certainly be
wrong to demand a "hard approach" to utilising the wind energy potential. But it should
not be lost from sight as an option for cases of emergency. If crises develop due to the
continuation of today's energy politics, the only solution in the long run will be to
push ahead with the utilisation of wind energy in this way. Wind energy is the only
REFERENCES 547
emission-free source of energy which could come to the rescue with currently available
technology and at economically justifiable costs.
But this situation does not as yet exist. Continuing exploitation of the wind-energy
potential will therefore only be pursued within socially accepted limits. However, a
totally restrictive approach is to be rejected. Intentionally or naively, it is a strategy for
maintaining outdated, traditional structures to the disadvantage of the ecology and for
the benefit of the protagonists of the current line of energy politics. Wind energy must
be granted a certain priority. In the individual case, it must not be subordinated because
lines of microwave radio links will not be relocated or because of the mere suspicion
that certain bird species might avoid the sites of wind turbines.
If the education of the public in these matters is consistently continued, basic legisla-
tion for the benefit of wind-energy utilisation will certainly be understood and accepted.
This would clear the way for supplying quite a considerable proportion of the energy re-
quirement with power provided by wind energy. This proportion will undoubtedly not
be restricted to "a few percent" of the energy requirement. Whether the wind-energy
potential to be exploited in the course of the coming century - for example in the Euro-
pean Union - will amount to one hundred thousand or two hundred thousand megawatts,
can be safely left to the future. The decisive fact is rather that the wind-energy potential
is high enough to cover a significant part of the power generation.
References
1. Frost, W.D., Asphiden, C.: Characteristics of the Wind. In: Spera, D.A. (ed.) Wind Turbine
Technology. ASME Press, New York (1994)
2. Archer, C.L., Jacobson, M.Z.: Evaluation of Global Wind Power. Stanford University, Stan-
ford (2005)
3. Süddeutsche Zeitung: NASA Ocean windpower maps, SZ No 160 (2008)
4. Troen, T., Peterson, E.: The European Wind Atlas. Risø National Laboratory, Denmark
(1989)
5. Steinmann, R.: Potentiale Binnenland, Windconcept, Germany (2008)
6. Deutscher Wetterdienst: Windkarten der BRD, Deutscher Wetterdienst, Dept. Klima- und
Umweltberatung, Offenbach (1995)
7. Molly, J.P.: Windenergie, 2nd edn., Verlag C. F. Müller, Karlsruhe (1990)
8. Häuser, H., Keiler, J., Allgeier, T.: Betreiber-Datenbasis Hamburg (2001)
9. Frost, W., Long, B.H., Turner, R.E.: Engineering Handbook on the Atmospheric Environ-
mental Guidelines for Use in Wind Turbine Generator Development. NASA Technical Paper
1359 (1978)
10. WINDPRO, Risø Test Station, Denmark (2005)
11. Gross, G., Frey, T., Trute, P.: Die Anwendung numerischer Simulationsmodelle zur
Berechnung der lokalen Windverhältnisse in komplexem Gelände, DEWI-Magazin, No. 20,
Wilhelmshaven (February 2002)
12. Scheer, H.: Sonnenstrategie, Piper-Verlag (1993)

13 The Wind Resource

Uploaded by

Copyright:

Available Formats

13 The Wind Resource

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

13 The Wind Resource

Uploaded by

Copyright:

Available Formats

Chapter 13

The Wind Resource

13.1 Causes of the Wind and of the Power in the Wind

Fig. 13.1. Global wind flows [1]]

ϱ = air density at standard atmosphere (kg/m³)

With respect to the annual wind frequency distribution it can be obtained:

13.2 Global Distribution of Wind Resources

Annual average Equivalent

Global averages and wind energy potential

Fig. 13.4. North America [2]

Fig. 13.5. South America [2]

Fig. 13.6. Europe [2]

Fig. 13.7. Africa [2]

Fig. 13.8. Asia [2]

Fig. 13.9. Australia [2]

13.3 Wind Resources in Europe

13.4 Characteristic Parameters of the Wind

Fig. 13.13. Definition of mean annual

The Weibull function is defin

If nothing besides from the mean

13.4.2 Increase in Wind

H = mean wind velocity at elevation H (m/s)

H = mean wind velocity at elevation H (m/s)

13.4.3 Steadiness of the Wind

Fig. 13.20. IWET operator database according to regions [8]

Fig. 13.21. IWET Wind Index for Region 6 up to 2006

13.4.4 Wind Turbulence and Gusts

Fig. 13.22. Measured time histo

Fig. 13.23. Energy spectrum of wind speeds (turbulence spectrum)

13.5 Topography and

- differences in elevation do not exceed 60 m within a radius of 11.5 km,

Fig. 13.24. Definition of "flat co

Apart from differences in elevation in the terrain, the topographical environment of a

Fig. 13.25. Flow wake behind an

13.6 Measuring and Determining the Wind Speed

13.6.1 Measuring Techniques

Fig. 13.27. Ultrasonic anemometer (Ammonit)

Fig. 13.28. Anemometer mast with sensors (SIGGELKOW)

Fig. 13.29. Data logger for wind speed analysis (AMMONIT)

Table 13.30. Classification of wind speed according to the Beaufort Scale

13.6.3 Numeric Models of Three-Dimensional Wind Fields

13.7 About the Wind Energy Potential

The atmospheric wind-energy potential

The economically usable wind-energy potential

The politically acceptable wind-energy potential

- Regions of residential settlement

- Regions of special tourist interest

You might also like