6. Conclusion
The proposed method for monitoring WTG using power curve and control charts has been validated. Using simulated data, EWMA and GWMA control charts have both shown excellent results with the detection of a 1% annual ramp shift in approximately 230 days. However it does not provides neither the source of the wear or failure nor the moment of behavior change. It is part of a multi-tool solution which, when used along with CMS and regular inspections creates a reliable and optimized O&M. The detection of variations of around 1% per year may appear to be of little use, but in terms of energy, such variations can lead to substantial drops in production. Also, such underperformances are amplified each year: for example, a 1% annual shift the first year will become a 2% shift the second year. The proposed method is not only useful for detecting of underperformance but also shows a low false-alarm rate, a feature that is interesting for industrial application.