A case study of wind turbine rotor over-speed fault diagnosis using combination of SCADA data, vibration analyses and field inspection

Reduction of wind turbines down times is paramount in wind energy develop- ment and cost effectiveness. Such reductions require early fault detection and more importantly, a comprehensive Root Cause Analysis (RCA) to completely fix at risk component(s). Given the complexity of wind turbines (and their con- trol systems) and the chaotic environment in which they operate, such a goal is difficult to obtain in many cases. This paper presents a successful case study that combines data analysis approaches with deep knowledge of wind turbine operation. It is shown how occurrence of over speed faults which normally hap- pen during gusty winds (i.e. high turbulent intensity) can be due to a bearing slipping issue in the gearbox. An integrated highfrequency vibration analyses, gearbox oil particle count, and 1 Hz operational and meteorological data min- ing approach, in addition to site inspection create a very suitable path towards establishment of a reliable operational strategy and repairs in a timely manner. A financial and operation strategy that demonstrates the proposed short- and long-term solutions is also presented.