Bayesian network based probabilistic model for optimized inspection intervals for offshore wind turbine support structures

M.P. Nicoreac, W.M.G. Courage, P.L. Kempker

Netherlands Institute of Applied Scientific Research (TNO), Delft, the Netherlands

The support structures of offshore wind turbines (OWTs) are difficult to inspect due to the environmental conditions and difficult access to structural parts. The inspection and maintenance cost are therefore high, providing a good motivation for further optimizing the planning and costs for these actions. In the context of the FeLoSeFI project, a risk based concept is used in which the uncertainties in fatigue life prediction models are considered. The probability of failure versus the costs of maintenance provide a good picture of the consequences regarding the maintenance actions to be applied.

Key words: Risk based inspection, maintenance, optimization, OWT, Bayesian networks