Risk Based Maintenance of Offshore Wind Turbines Using Bayesian Networks

This paper presents how Bayesian networks can be used to make optimal decisions for repairs of offshore wind turbines. The Bayesian network is an efficient tool for updating a deterioration model whenever new information becomes available from inspections/monitoring. The optimal decision is found such that the preventive maintenance effort is balanced against the costs to corrective maintenance including indirect costs to reduced production. The basis for the optimization is the risk based Bayesian decision theory. The method is demonstrated through an application example.