Wind Energy

326 N. Cosack et al.
with 70 m rotor diameter and hub heights of 85, 100, and 130 m and a 5 MW
machine with 128 m rotor diameter and hub heights of 100 m and 130 m. Both
turbine types are typical pitch controlled variable speed wind turbines.
We applied a common approach for modeling the incoming wind: The
deterministic part and the turbulent part of the wind are modeled separately,
superposition of the two parts gives the full three dimensional turbulent wind
field. The turbulent wind field has been modeled with a method described
by Veers [4], using the Kaimal power spectral density functions as given in
the IEC guidelines [5]. For both profiles, a standard deviation in longitudinal
wind speed of 0.7 m s −1 has been estimated from measurements.
The center of the LLJ has been measured at 180 m, while the top of the
surface layer is located at 50 m above the ground. In Profile A, the mean wind
speed at a reference height of 85 m is 4 m s −1 with an additional 25% increase
at the center of the LLJ, compared to what one would estimate solely from
terrain and stratification. In Profile B, these parameters are 7 m s −1 and 6%.
61.3 Results
Two scenarios have been examined, S1 and S2. In S1, the results of simulations
with the profiles A and B have been compared to results when the
vertical wind profile has been modeled only according to the logarithmic law
and stratification. This corresponds to the error that we would make, if we
had to estimate the turbines performance by just knowing the wind speed
at hub height and some basic environmental parameters, but would not be
aware of the LLJ. The investigated sensors are power production, the standard
deviation of the blades flapwise bending moment and the mean tilt moment
at the nacelle center.
Figure 61.1 shows the relative deviations due to the LLJ. The influence
is visible in all sensors, although the difference in power and blade bending
moment stays below 5%. As expected, profile A gives higher differences compared
to profile B. This is due to the stronger increase in wind speed from
surface layer to the center of the LLJ. Also, the results for the 5 MW turbine
are generally higher than those for the 1.5 W turbine, due to the bigger rotor
plane. The influence of the LLJ is highest in the case of the 5 MW turbine
with a hub height of 130 m and profile A.
To make things worse, we assumed in S2 that the given wind speed has
not been measured at hub height as before, but at some point below. This
is often the case, when the wind resource at a site is evaluated during the
planning of a wind park or when too small measurement masts are used in
a wind park for performance evaluation. In comparison to S1, the difference
between the situations with and without LLJ is not only the wind profile.
The LLJ now results also in an increased mean wind speed at hub height and,
as the standard deviation is kept constant, reduced turbulence intensity. At
reference height of 85 m the wind speed for both situations is equal.