Open Access

Tanja von Platen

• The simulation model developed in this work describes the maintenance processes of major components of onshore wind turbines based on remaining useful life (RUL) data. The model was implemented using agent-based modelling and simulation, thereby describing each of the predefined major components, i.e. gearbox, main bearing, main shaftand rotor, as an individual entity with specific behaviour. The wind turbine was also described as an agent, whose behaviour is dependent on the behaviour of its components. In accordance with the objectives of this work, alternative maintenance strategies are defined. These include the full replacement of a turbine using new or used components, single component replacement with new or used components and refurbishment of installed components. Additionally, replacement time thresholds are defined to evaluate the impact of clustering maintenance visits on the total maintenance cost. The results provide an insight into the dominating costs constituting the total maintenance cost. A major cost factor is the component cost, whereby the replacement of a component, even if a used component is considered, outweighs the cost of refurbishment. Moreover, the service cost is an important factor to consider in the reduction of maintenance costs. The upfront crane rental fee has the greatest effect on the service costs, hence, considering the clustering of maintenance visits provided promising results. The greatest cost benefit, compared among all maintenance strategies and replacement time thresholds, was obtained for the refurbishment of single components with a threshold value of 183 days. The advantage of the lower cost of refurbishment versus replacement, and in addition, the reduction of service costs through the clustering of maintenance visits, resulted in a greatly reduced total maintenance cost.