Chartering a vessel fleet to support maintenance operations at offshore wind farms (OWF's) constitutes one of the major costs of maintaining this type of installations. Literature describes deterministic optimization models based on complete information within scenarios to schedule the maintenance and support decisions on the vessel fleet composition. The operations to be carried out can be classified as preventive and corrective tasks. The first type aims at reducing the likelihood of breakdowns and to prolong the life of turbine components. Corrective tasks are needed to repair breakdowns in turbines when they occur. Our research question is how to generate a vessel fleet composition, where the evaluation is based on scheduling without
complete information. Such a model is a bi-level decision problem. On the first (tactical) level, decisions are made on the fleet composition for a certain time horizon. On the second (operational) level, the fleet is used to schedule the operations needed at the OWF, given random events of failures and weather conditions. A scenario based approach allows evaluation by parallel operational scheduling for each scenario..