Stochastic simulation of stresses caused by wind and waves to offshore wind turbines

The tendency to avoid nonrenewable energy resources in recent years has led- among other alternative modes of energy production- to the rapid development of offshore wind farms. With regard to the structural design, international standards suggest using dynamic models, which use short-term simulations, to calculate the forces exerted on offshore wind turbines. This thesis aims to model the behavior of the forces by using stochastic methods, and compare the stochastic simulation method to the method proposed by the international standards. Initially, in terms of data analysis, we use the largest available time series of wind speed, wave height and period which includes recordings in the Gulf of Alaska. By using the appropriate deterministic equations, we calculate the time series of forces due to aerodynamic and hydrodynamic loads. Moreover, we calculate the statistic moments, the periodicity and form the distribution of the time series of the forces. Τhe empirical distributions are fitted to the theoretical distribution Pareto-Burr-Feller and by using the parameters of the fitting we calculate the four central moments. Then, regarding the stochastic analysis, we study the correlation structure of data by using the climacogram. The stochastic models HK, GHK, HHK are fitted to the empirical climacogram with model GHK having the best fit. Finally, in order to conduct the stochastic simulations we use the GHK model in the SMA scheme. As input parameters we use the parameters resulting from the fitting of the GHK model to the empirical climacogram and the first four moments resulting from the fitting of the PBF distribution to the empirical one. In addition, we examine the fitting of the results of the simulation to the empirical climacogram and distribution and we compare the results of the simulation with the results of the proposed methods.