Spatial Variation In Wind Extremes Under The Context Of Climate Change And Its Impact On Safety Of Wind Turbine

As a highly erected structure, wind turbines are sensitive to wind load and wind load is a key parameter in its design and modeling. Traditionally, climate pattern are supposed to be stationary. The wind turbines are designed according to wind extremes of 50-year recurrence interval calculated with observed historical wind data. Since wind turbines are of a life cycle of 20-25 years and wind extremes varies under the context of climate change, variation in wind extremes could negatively impact reliability and safety of a wind turbine. The risk arising from variation in wind extremes deserved in-depth investigation.In order to find out whether there are spatial variations in wind extremes, wind data in six meteorological stations in Inner Mongolia, Central China and South East coastal China are investigated in this thesis. The wind data from 1973 to 2012 are divided into two parts from 1973 to 1992 and from 1992 to 2012. Since there are series of models could be used for extreme value analysis, the quantile-quantile plot, KS test, and AD test are utilized to evaluate applicability of Generalized Extreme Value(GEV), generalized pareto distribution(GPD) based Peak over Threshold(POT) and Method of Independent Storm(MIS). It is uncovered that the GPD based POT and MIS could outperform GEV notably. While the performance of MIS is even better than POT since dependence among sampled extremes are eliminated. Selection of threshold plays a key role in accurate parameter estimation. Traditionally, it is determined based on residual life function. However, it is not straightforward and maybe misleading occasionally. The residual life function and parameter estimation graph are proposed to be used in combination to determine the appropriate threshold in a more straightforward way. Numerical simulation indicates the performance of the proposed approach.In order to investigate spatial features of climate change induced variation in wind extremes, the wind extremes of 50-year and 1-year recurrence interval in the investigated sites based on observed wind data in 1973 to 1992 and 1993 to 2012 are estimated with GPD based MIS. It is uncovered that there are notable spatial variations in wind extremes induced by changes in climate pattern. The wind extremes of 50-year recurrence interval declines notably in Inner Mongolia, enhanced slightly in coastal east China, and remain stationary in Central China. There are notable declines in wind extremes of 1-year recurrence interval while the spatial variation is not observed. The type of wind turbines are determined based on wind extremes of 50-year recurrence interval estimated with wind data of 1973-1992. Thereafter, its safety is evaluated and the wind extremes of 50-year recurrence interval estimated with wind data of 1993-2012. It is uncovered that the safety in Inner Mongolia and Central China escalated slightly while the safety in Coastal East China decreased notably.