The demand for energy is still increasing, together with the desire to create more durable solutions. Wind energy is currently one of the most promising ways to provide both. A shortage of suitable land sites was the main reason for considering the location of the wind farms offshore, which also offers higher wind speeds. A problem however is formed by the low accessibility during periods of high wind speeds. This low accessibility could lead to a decrease in availability of the system and high O&M cost, both unfavourable for the levelised production cost and thus the profitability of the system. The levelised production costs have long been used as the main criterion in the evaluation of a farm design. In a private energy market, the system however should be optimised to maximise profit. At present energy prices the difference this
gives for the optimisation of the farm and its O&M can be neglected. The goal of this thesis is the optimisation of operations and maintenance on Offshore Wind Energy Conversion Systems with a probabilistic model. A number of previously executed studies were used to determine what input parameters should playa role. This showed that all subsystems contribute significantly to the initial capital cost, O&M cost and the unavailability of the system. The model should take into account that periods of low accessibility coincide with periods of high energy potential. The model is based on a state-space model of each component. A component can find itself in the states "Failed", "Corrective Maintenance", "Preventive Maintenance" or "Available". Each component in the system is part of the chain connection that transports the produced energy to shore. If this chain is broken by a failure the loss in production capacity is determined by the place of the component in the chain. The failure of a component is determined by a Monte Carlo simulation. The reliability of components decreases in time according to a certain decline rate. After the execution of PM or CM the initial reliability will be restored. The program, written in visual basic, shows results according to predictions by hand calculations and results from other studies. It can be used to optimise the system. A simulation of the base case showed that the turbines are the main cause of unavailability and cost. The main cost drivers of ICC are turbines and their support structure. As expected the main O&M costs are
caused by the turbine and in particular the rotors. The turbines are also the main cause of unavailability: 7% compared to around 2% for the shore- and grid connection. There is a relationship between unavailability of the system and the occurrence of high wind speeds. It shows a decline in availability of about 5% for wind speeds higher than the wind limit
of used O&M tools. This results in a loss of revenues of about 2% per year. The relationship between PM interval, failure- and decline rate has been investigated. The first results indicate a trend towards a decreased need for PM than the intervals used in the base case. The influence of the decline in reliability can be neglected for decline rates up to 25%. The program can be further improved by accounting for the availability of spare parts and crew or enabling more decision rules for the execution of O&M.