Maintenance optimization of floating offshore wind turbines: digital twins and stochastic modeling for uncertainty management

Offshore wind turbines (OWTs) face high maintenance costs due to harsh and uncertain environmental conditions.Traditional maintenance strategies-whether fixed-schedule, failure-based, or risk-based-often fail to capture real-time weather impacts or gradual degradation under normal conditions.This study proposes a risk-informed, condition-based predictive maintenance strategy that incorporates environmental uncertainty to reduce costs and improve decision-making.The approach transforms the probability of failure of critical components into a condition factor used to trigger maintenance actions.Two key aspects are evaluated: (1) the cost-effectiveness of collectively maintaining spatially proximate turbines, and (2) the performance of the proposed predictive strategy versus traditional fixed-interval maintenance.Results show that grouped maintenance significantly reduces logistical costs, and the predictive strategy lowers the number of interventions while minimizing downtime, outperforming fixed-schedule approaches.