Probability distributions of wind turbine loads are estimated from lim
ited darn. The impact of different models on fatigue damage is shown.
Common one-parameter probability models, such as the Rayleigh and expo
nential, are found to give significantly different estimates of load d
istributions and damage. Greatest differences occur in materials with
relatively high values of the S-N exponent b, such as composites. In s
uch cases more accurate damage estimates are found by matching at leas
t two moments of the load data (Weibull model), or still higher moment
s as well. For this purpose, a new, four-moment ''generalized Weibull'
' model is introduced, For edgewise loads it appears a notable improve
ment over the basic Weibull model, while it supports the Weibull model
in the flapwise case. Uncertainty in damage estimates is also quantif
ied along with the implied data needs.