Analyzing wind turbine directional behavior: SCADA data mining techniques for efficiency and power assessment

SCADA control systems are the keystone for reliable performance optimization of wind farms. Processing into knowledge the amount of information they spread is a challenging task, involving engineering, physics, statistics and computer science skills. This work deals with SCADA data analysis methods for assessing the importance of how wind turbines align in patterns to the wind direction. In particular it deals with the most common collective phenomenon causing clusters of turbines behaving as a whole, rather than as a collection of individuality: wake effects. The approach is based on the discretization of nacelle position measurements and subsequent post-processing through simple statistical methods. A cluster, severely affected by wakes, from an onshore wind farm, is selected as test case. The dominant alignment patterns of the cluster are identified and analyzed by the point of view of power output and efficiency. It is shown that non-trivial alignments with respect to the wind direction arise and important performance deviations occur among the most frequent configurations.