Multi-objective optimization for an autonomous unmoored offshore wind energy system substructure

Autonomous unmoored floating offshore wind energy systems are an unconventional but promising technological solution to access the greatest wind resource in deep waters far offshore. This study proposes a trimaran substructure for such an autonomous unmoored system producing green hydrogen, named the Wind Trawler, which serves as both a power generation system and an energy transport vessel. Using a multi-objective optimization approach founded on the non-dominated sorting genetic algorithm (NSGA-II), the principal geometric parameters of the trimaran substructure hulls (primary hull and two symmetrically-spaced equivalent outriggers) are optimized with respect to minimization of system steel mass and minimization of the average unit power consumption per unit generation. Three discrete cases of maximum allowable overturning angle in heel (5°, 10°, and 15°) are considered. The results show that outriggers provide the majority of stability while the primary hull is most significant in influencing power performance and steel mass. Constraints of system stability and minimum outrigger draught induce a tradeoff between the outrigger waterplane area and the outrigger spacing from the primary hull. Steel mass minimization correlates with small hull spacing, shallow hull draughts, and greater outrigger proportion of the total mass. Large hull spacing, deep hull draughts, and reduced outrigger size maximizes the proportion of total mass in the streamlined primary hull, and therefore minimizes power consumption relative to generation. The effect of increasing maximum allowable overturning angle is that outrigger draught is much deeper at equal spacing from the primary hull. This induces a much greater outrigger mass and less efficient system overall, which drives up power consumption relative to generation. Ultimately, the optimum Wind Trawler substructures yield steel mass results that promise to compete with that of conventional systems, especially considering the elimination of mooring and electrical transmission subsystems. Metrics of power consumption relative to generation are unique to this type of unmoored system and thus need to be translated to conventional units of net energy production. This can be performed by considering the varying expected duration and trajectories of each operational mode.