Wind resource and site assessment - Wind Power Generation and Wind Turbine Design - page 61

Figure 6: Iso-speed plot of local fl ow speed around a triangular lattice mast with a
C T of 0.5 normalised by free-fi eld wind speed (from left); analysis by two-
dimensional Navier-Stokes computations and actuator disk theory [3].

wind speed are much more signifi cant, reaching - depending on the local climate -
variations of 50%. Thus it is of crucial importance to measure complete years

since the monthly variations are too large to be corrected and, as a consequence,
partial year data suffers from a seasonal bias.
Wind data is normally sampled for 10 min, the logger then calculates the aver-
ages and the standard deviations, the latter being necessary for determining the
turbulence intensity. The averaging period of 10 min allows the direct use of the
data for load calculations as they also refer to 10-min averages. Great care has to
be taken when the averaging period is shorter than 10 min as the standard devia-
tion and thus the turbulence intensity will be lower in comparison to 10-min
averaging period. This can lead to under-estimated turbulence-induced loads.

±

3 Data analysis

3.1 Long-term correction

3.1.1 Introduction
Variability is an intrinsic feature of climate. Precautions are necessary as the
weather changes from year to year and between consecutive decades. The data
which form the basis of any wind resource study cover a limited period of time,
which in many cases is less than 5 years. The question therefore arises: to what
extent is that period representative for the longer-term climate. A study of climatic