A review and analysis of optimisation techniques applied to floating offshore wind platforms

Abstract

The deployment of offshore wind in the UK has seen a rapid increase in the past decade and will continue to increase with the securement of the recent Scotwind sites. Floating platforms will be utilised for 60% of these new sites, creating opportunities to try new platform typologies and further solidify the validity of existing concepts. Since there is no consensus on the platform typology, the cost will vary; however, it is predicted to be double the price of traditional fixed platforms. Finding the most optimal solution in terms of cost and performance is key to keeping cost low, allowing the technology to be more competitive. A technique which has been used in other industries is multi-objective optimisation, searching a large design space much more quickly than traditional methods. By carrying out a multi-objective approach, the optimal platform geometry can be identified over the Pareto Frontier, considering conflicting objectives such as cost and performance. The aim of this work is to review the existing literature on multi-objective optimisation of floating offshore wind (FOW) platforms, highlighting the gaps and shortfalls in the current literature. This review highlights the majority of work has been carried out for the 5 MW NREL turbine on a SPAR platform, utilising a genetic algorithm. Cost reduction has been noted as the main objective, however, the models found within the literature are simplistic, with a number of assumptions. The overall findings of this work highlight future work that could be improved: cost models, the inclusion of an energy production model linked to the platform motion, the requirement for analysis of larger turbines and the potential for a concept selection tool to reduce computational time.