Evaluation of floating wind turbine substructure designs by using long-term dynamic optimization

The substructure design plays an important role in the capital expenditure as well as the dynamic performance of floating wind turbines (FWTs). To determine the most cost-effective hull shape and mooring configuration from various design concepts, this paper proposes an evaluation methodology based on long-term dynamic optimization. To make the long-term dynamic performance assessment computationally affordable in the iterative optimization process, the FWT responses are approximated by a Kriging surrogate model, which is established via regression analysis between representative met-ocean parameters and the corresponding short-term responses. The short-term responses in stochastic wind and waves are simulated by an efficient reduced-order model (ROM) with eight degrees of freedom (DoFs). The Kriging model is employed to derive the long-term performance indicators, e.g. the lifetime accumulative fatigue damage at tower base and fairleads, the peak platform motion and nacelle acceleration. The model effectiveness is confirmed by the verification campaign against OpenFAST. Afterwards, this model is implemented into a multi-objective optimization framework that aims to find out the Pareto optimal designs that have the best trade-off between long-term dynamic performance and manufacturing cost. The NSGA-II algorithm is adopted to explore the substructure design space, with the consideration of constraints related to the inherent properties as well as the dynamic performance of structures. This methodology is applied to two typical semi-submersible FWTs: the Y-shaped and the square-shaped concept. By comparing their Pareto fronts, it is found that the square-shaped semi-submersible is expected to be a more favourable substructure for our case study. Under the same level of manufacturing cost, the tower base fatigue of square-shaped concept is mostly 30%–50% lower than that of the Y-shaped. Hydrodynamics is the main cause of the fatigue difference between the two concepts. This work provides a decision basis on which FWT substructure design is employed as input to the detail design stage.