The O&M cost of offshore wind turbines accounts for about 25% of the total lifecycle cost– Leading Edge Erosion (LEE) being one of the most critical damage types. O&M repair strategies are often reactive, meaning that repairs are only initiated after the blade damage reaches a particular threshold. However, is this O&M strategy the most cost-efficient?

Lopez and Kolios (2024) suggest a predictive O&M strategy where a Reinforcement Learning agent, trained to optimise O&M costs based on modelled data, makes the maintenance decision. Using two case studies, the authors compare the O&M costs for the reactive and predictive strategies. They find that the predictive method reduces the average O&M costs by 21% and 13%. The cost reduction is partly due to a reduced maintenance cadence towards the end of the turbines’ lifetime.

Lopez, Javier Contreras, and Athanasios Kolios. 2024. “An Autonomous Decision-Making Agent for Offshore Wind Turbine Blades Under Leading Edge Erosion.” Renewable Energy 227 (June): 120525. https://www.sciencedirect.com/science/article/pii/S0960148124005901?via%3Dihub