In this study, a systematic numerical study on design optimization was conducted for a tension leg platform (TLP)-type floating offshore wind turbine (FOWT), aiming to improve hydrodynamic performance, tendon behavior, and cost-effectiveness. Six design variables associated with hull geometry and tendon properties—pontoon length (PL), pontoon width (PW), platform draft (PD), main column diameter (CD), tendon pre-tension (Pre), and axial stiffness (EA)—were considered. For the global performance analysis, hydrodynamic coefficients were first obtained in the frequency domain, and motion and tendon tension responses were subsequently evaluated in the time-domain under a survival condition representative of a Southeast Asian site. A surrogate model based on the response surface method (RSM) was developed to predict platform responses across the design space. Multi-objective optimization was then performed using the non-dominated sorting genetic algorithm II (NSGA-II), yielding Pareto-optimal solutions that reveal trade-offs among competing performance metrics. The proposed framework is intended to provide Pareto-optimal design candidates for preliminary TLP-FOWT design, while the selection of a final design requires project-specific criteria and is beyond the scope of the present conceptual study. The optimization results show that the tendon tension can be effectively reduced while maintaining cost efficiency by increasing the pontoon length and slightly decreasing the tendon axial stiffness. For the tension–surge motion optimization, the Pareto-optimal solutions provide a balanced trade-off, where tendon tension is clearly reduced with only a slight increase in surge motion. In addition, the cost–nacelle acceleration optimization shows that nacelle acceleration can be further reduced by increasing the platform draft and pontoon length, although this is accompanied by a slight increase in the cost index. These findings provide practical insights for balancing global performance and cost in TLP-type FOWT design.
Song et al. (Sat,) studied this question.