As mentioned earlier, underperformances were simulated in
order to validate the proposed methodology as a function of variation
level. Itwas assumed that all bins of the power curve between
the cut-in and the nominal wind speed would undergo suffer the
same relative shift. Two types of variations were considered: a step
change and a ramp change. The step reached its nominal value
immediately (e.g. a one-shot 5% step reducing the entire power
curve) while the ramp was denominated on an annual basis (e.g. a
5% variation over one year, meaning that after one year the power
curve would be reduce by 5%). A ramp shift is the best way to
represent of consequence of problems such as blade erosion, while
step shift give the best representation of sudden changes in
components.