Performance monitoring makes use of existing data to evaluate the performance of a turbine and hence a wind farm. Moreover, it does this without the need for intrusive installation of equipment. In that respect it differs from condition monitoring, which can incur significant capital cost while invalidating the turbine's warranty. Performance monitoring can be used to continually evaluate the performance of a wind turbine in a number of ways; for instance correlation of lost yield with event logs can help target finite O&M resources to return the maximum benefit in improved yield. Analysis of trends in time series data is leading to the development of metrics that can give early diagnosis of turbulent fatigue, blade fouling and other indicators of lost yield and predictors of mechanical failure. Pattern recognition can detect incipient departures from normal behaviour, leading to early intervention and prevention of both gradual performance degradation and catastrophic failure.
展开▼
