Environmental Engineering Reference
In-Depth Information
(transformed probability of non-exceedence) is plotted versus wind speed.
The reduced variant (probability) can be expressed in different ways leading to
different plotting positions. Furthermore, a variety of fi tting options are avail-
able leading to a vast number of different results [27]. However, all methods
have one problem in common. The resulting 50-year estimate is highly corre-
lated to the highest measured wind speed event of the time series used for the
analysis [ 28 ].
The European Wind Turbine Standard, EWTS [29] offers an option to estimate
the extreme wind based on the wind speed distribution on site rather than a mea-
sured time series. It suggests a link between the shape of the wind distribution and
the extreme wind. The EWTS relates a ratio determined by the Weibull k factor to
the annual average wind speed to estimate the 50-year extreme wind speed. This
factor is 5 for k = 1.75, <5 for higher values of k and >5 for lower k values (i.e. high
values for distributions with long tails, and low values for distributions with lower
frequency of high wind speeds). The extreme wind is calculated as this factor
times the yearly averaged wind speed.
The extreme 3-s gust or also called design wind speed of the wind turbine ( U e50 )
is a function of the U ref and the turbulence intensity. The extreme gust is estimated
using the relationship, where the turbulence intensity I ext is a high wind speed
turbulence value estimated from the measured data:
=
ref (1
+
2.8
)
UU
I
(13)
e50
ext
This relationship is based on experimental data as well as theoretical work (Fig. 25).
As mentioned in Section 5.2.1 care has to be taken in offshore and forest situa-
tions as the turbulence intensity does not show asymptotic behaviour but increases
with increasing wind speed. It is thus much more diffi cult to estimate I ext under
these conditions.
Figure 25: Measured gust as a function of wind speed [ 16 ] .
Search WWH ::




Custom Search