Uncertainty modeling in reliability analysis of floating wind turbine support structures

Accurate structural reliability assessment of floating wind turbine (FWT) systems is a desideratum for achieving consistent optimal reliability levels and cost-effective design. Such reliability assessment should consider relevant system uncertainties—a nontrivial task. Formulation of the reliability problem requires structural demand in form of load and load effect. Support structure loads are predicted with time-domain dynamic simulations. This represents a challenge when thousands of such simulations are required to capture the uncertainty associated with design variables. Finite element analysis (FEA) is commonly used to evaluate load effects such as stresses, strains etc. This can be computationally expensive if not prohibitive when such evaluation is carried out for every time step. To tackle these issues, a framework for expeditious load effect computation and robust reliability analysis of FWT support structures under ultimate limit state design is presented. The framework employs linear elastic FEA and Kriging surrogate models. The adequacy of Kriging as applied in this study is investigated using high fidelity simulation data. The results highlight the importance of incorporating the Kriging uncertainty in the formulation of the limit state function. With the framework presented, FWT support structures can be designed at consistent reliability levels leading to cost reductions.