Planning of O&M for Offfshore Wind Turbines using Bayesian Graphical Models

The costs to operation and maintenance (O&M) for offshore wind turbines are large, and riskbased planning of O&M has the potential of reducing these costs. This paper presents how Bayesian graphical models can be used to establish a probabilistic damage model and include data from imperfect inspections and monitoring. The method offers efficient updating of the failure probability, which is necessary for risk-based decision making. An application example is presented to demonstrate the capabilities of the method.The costs to operation and maintenance (O&M) for offshore wind turbines are large, and riskbased planning of O&M has the potential of reducing these costs. This paper presents how Bayesian graphical models can be used to establish a probabilistic damage model and include data from imperfect inspections and monitoring. The method offers efficient updating of the failure probability, which is necessary for risk-based decision making. An application example is presented to demonstrate the capabilities of the method