Environmental contours based on kernel density estimation

Abstract

An offshore wind turbine needs to withstand the environmental loads, which can be expected during its life time. Consequently, designers must define loads based on extreme environmental conditions to verify structural integrity. The environmental contour method is an approach to systematically derive these extreme environmental design conditions. The method needs a probability density function as its input. Here we propose the use of constant bandwidth kernel density estimation to derive the joint probability density function of significant wave height and wind speed. We compare kernel density estimation with the currently recommended conditional modeling approach. In comparison, kernel density estimation seems better suited to describe the statistics of environmental conditions of simultaneously high significant wave height and wind speed. Consequently, an environmental contour based on kernel density estimation does include these environmental conditions while an environmental contour based on the conditional modeling approach does not. Since these environmental conditions often lead to the highest structural responses, it is especially important that the used method outputs these conditions as design requirements.