Offshore wind farm managers and schedulers need to manage large numbers of wind turbine visits every day, in order to: repair minor faults; conduct inspections; and perform scheduled service operations. Daily schedules form a choice of which maintenance activities to conduct, taking account of: constraints on weather conditions, shifts, vessel and technician capabilities and availability; and the impact of activities on wind farm profitability. This forms a formidable optimisation challenge that today is solved "by hand" by a scheduler. The work presented here contains three aspects of importance. First, a powerful and flexible meta heuristic optimisation model is developed to solve this problem, where the simulation algorithms and objective can be altered without any change to the optimiser. Second, a practical valuation methodology is developed, where historic wind farm data can be used to identify strengths and weaknesses in any maintenance planning method and estimate financial return on investment from implementation. Finally, the methodology described is implemented and tested, by applying the valuation methodology to data from the Princess Amalia Wind Park in The Netherlands. Even given the limited scope of this case study, automating daily maintenance planning can bring significant financial benefits: 302 k(sic) over 5 months. (C) 2018 Elsevier Ltd. All rights reserved.