Offshore wind energy is key in the UK’s plan to deliver the legally binding Net Zero 2050 targets, quadrupling the capacity by 2030. First-generation offshore wind monopiles are rapidly approaching their end of designed life. The next-generation of wind turbines are significantly larger, yet still monopile support
structures dominate. Accurate estimation of accumulated monopile fatigue is essential now, to inform decommissioning decisions, and optimise future design and maintenance. Due to unpredictable offshore environments, and the difficulty of taking structural measurements, fatigue predictions are subject to significant error.
This project proposes an industry-compatible step-change advance in accumulated fatigue assessment via novel integration of physical modelling and machine learning. The proposed model provides intuitive prediction of the level of fatigue for any turbine within the farm, at any point of its lifetime from distinct operational and environmental conditions, verifiable against physical models, yet with increased efficiency and fidelity of lifetime fatigue estimation
Dr Nina Dethlefs - Principal Investigator
Dr Dethlefs' research interests lie in computational linguistics, particularly in computational language learning for interactive systems and natural language generation. She investigates machine learning models that can automatically extract linguistic patterns from data, and use these to understand and generate language in new contexts. Dr Dethlefs is a member of the Computational Science Research Group, and a member of the Digital Centre.