Optimal maintenance policies for wind turbines under time-varying costs

In this thesis we introduce a new modelling approach for maintenance optimisation under time-varying costs. We extend the standard age replacement policy (ARP), block replacement policy (BRP) and modified block replacement policy (MBRP) to be able to include timevarying costs. We find that, via a discretisation of time, the optimal ARP can be found using a linear programming (LP) formulation. The optimal BRP and MBRP can be found using a mixed integer programming (MIP) formulation. For the BRP we introduce a sequential optimisation algorithm, a genetic algorithm and a memetic algorithm to decrease computation time for multi-component problems. We compare the performance of the heuristics to the performance of the MIP formulation. We apply the heuristics to a wind farm with 10 wind turbines with a turbine capacity of 9.5 MW and show that we can make significant cost savings using our extended policies.