Offshore Winds - Wind Energy Meteorology: Atmospheric Physics for Wind Power Generation - page 114

Fig. 5.20 Frequency
distributions of turbulence
intensity [see Eq. ( 3.10 ) ] at
90 m height at FINO1 in the
German Bight for 13 wind
speed classes for the period
September 2003 to August
2007
the thermal effects and turbulence intensity increases again (Barthelmie 1999 ).
Maximum values of turbulence intensity at wind speeds below 18 m/s are higher
than 20 % and therefore not visible in the plotted range of the turbulence intensity
values in Fig. 5.19 . They peak at 48.8 % for 13 m/s wind speed and are between
36.3 and 50.0 % for wind speeds between 1 and 13 m/s.
Due to the non-Gaussian frequency distribution median and arithmetic mean of
turbulence intensity differ below wind speed values of about 11 m/s while at
higher winds speeds these two values are nearly equal. The absolute minimum of
turbulence intensity for each 1 m/s wind speed bin is less than 1 % up to wind
speeds of 20 m/s. Inspection of the synoptic conditions suggests that under very
stable atmospheric conditions situations can occur with very low turbulence
intensity even at relative high wind speeds. Above 20 m/s wind speed the influ-
ence of the more and more rough surface and so increasing friction stress can
always break up this very stable layering and so also the absolute minimum of
turbulence intensity begins to increase to values near the 10th percentiles. At
higher wind speeds the spread of turbulence intensity values within one wind
speed class is continuously becoming smaller (Large and Pond 1981 ).
Turbulence in the atmospheric boundary layer is either generated by shear or by
thermal instability. While for lower wind speeds thermal production of turbulence
is dominant it becomes nearly negligible for high wind speeds when compared to
shear production. The shear production is proportional to the surface roughness.
Onshore the surface roughness is a function of the surface characteristics only and
assumed to be independent from the atmospheric conditions. This is different for
offshore conditions. The oldest proposition for a description of this dependence is
by Charnock ( 1955 ) who proposed the relation ( 5.1 ). Garratt ( 1977 ) reviewed the
topic of sea surface roughness and recommended to estimate z 0 by Charnock's
relation [Eq. ( 5.1 )] with a = 0.0144 for j = 0.41 which according to Wu ( 1980 )
corresponds to a = 0.017 for j = 0.4. The IEC standard 61400-3 (IEC 61400-3,
2006) assumes a = 0.011 for offshore conditions. Figure 5.20 shows frequency