Bayesian analysis applied to statistical uncertainties of extreme response distributions of offshore wind turbines

Extreme response is an important design variable for wind turbines. The statistical uncertainties concerning the extreme response distribution are simulated here with data concerning physical characteristics obtained from measurements. The extreme responses are the flap moment at the blade root and the overturning moment of the support structure of an offshore wind turbine situated in the North Sea. The statistical uncertainties concern the choice of the distribution model and uncertainties concerning the distribution parameters. The uncertainties are treated with Bayesian analysis. The inclusion of the uncertainties has only marginal effect for the calculated long-term estimates of extreme responses when non-informative priors for the distribution parameters are used. The inclusion of uncertainties may have larger effect if data concerning the moments are obtained from measurements