Optimizing maintenance plans of offshore wind farms by calculating the likelihood of future turbine failures

Abstract

Although offshore wind power shows promising energy potentials, high cost of operating and maintaining offshore wind farms concerns investors. Different maintenance strategies are applied by wind farm operators to overcome this drawback. A mixed integer optimization model is developed to find the optimal maintenance plan for an offshore wind farm. The proposed model include probabilistic failure times, multiple components per wind turbine, route decisions and imperfect maintenance. That is, aspects usually studied individually in the literature. Maintenance actions are scheduled based on the calculated likelihood of future turbine failures. Results from numerical experiments show that applying an imperfect preventive maintenance strategy, as opposed to a preventive replacement strategy, is preferable in most scenarios. An additional heuristic algorithm is presented. Close to optimal solutions with optimality gaps between 1% and 3% prove that the heuristic algorithm yields good solutions.