Posters

Abstract

Typically, in large scale wind farms, there is plenty of room for optimization in maintenance planning. Selecting the ideal times to start maintenance, in order to minimize revenue loss, is not an easy task and is commonly neglected by operators or manually done without proper optimization. This work proposes a methodology to automatically generate a maintenance plan seeking to maximize revenue. In order to achieve this, the following steps were developed: (1) first it is necessary to calculate each wind turbine energy availability in the last days (or after the last maintenance) and suppose this is its current performance; (2) the second step is to gather the hourly capacity factor forecast and the hourly energy price forecast; (3) then, use these information to perform simulations varying maintenance start times, for each wind turbine, calculating revenue losses until the beginning of the maintenance, due to performance, and during the downtime; (4) finally, it is possible to choose, inside local team working hours, the ideal moments to stop each machine and perform maintenance. This process is applied to 667MW of operational assets, generating an objective maintenance plan to guide the service team according to the company's strategies.