Characteristics of the Wind - Wind Turbine Technology: Fundamental Concepts of Wind Turbine Engineering - page 515

of vertical and lateral gradients in wind speed, between two points separated by distances
the size of a rotor, treating these gradients as random variables. Correlation of test data
on the basis of the steady wind speed at mid-elevation on the rotor, U , surface roughness
length, z 0 , and rotor diameter, D , yields the empirical models shown in Figure 8-25 for the
standard deviations of the average vertical and lateral gradients.

Figure 8-25. Empirical models for estimating standard deviations of the vertical and
lateral wind gradients across the swept area of a rotor of diameter D. [Ramsdell 1978]

Elliott [1984] analyzed wind shear data
from three sites in the U.S. with tall towers
supporting anemometers at various eleva-
tions, under day, night, summer, and winter
conditions. A two-slope system was devel-
oped for categorizing the observed wind
shear profiles (Fig. 8-26), and 30 of the
most common were analyzed statistically
for frequency of occurrence, mean duration
( i.e. persistence), and magnitude of fluctua-
tions in steady wind speed associated with
each profile. Of particular interest here is
the relatively short persistence of many of
the wind shear profiles observed. For ex-
ample, during a winter test at one site, there
was a 71 percent probability that the wind
shear profile would change to a different
pattern in less than one minute. In the
summer this probability rose to 96 percent.
Much additional information on wind shear
as a dynamic phenomenon can be obtained
from this study.

Figure 8-26. Two-slope system for cate-
gorizing vertical wind shear profiles
across a HAWT rotor. [Elliott 1984]