5.4. Application to real data
The long-term power time series of a WTG was analyzed using the current method, along with the EWMA control chart. The wind farm had been serviced in 2006, with data available since 2009: results were clearly detecting an underperformance by the summer of 2011, as shown in Fig. 8. The presence of an issue was confirmed by the long-term observations shown on Fig. 9. After August 2011, the WTG remained statistically out of control; the EWMA has been outside the control limits since that time. The difference in Annual Energy Production (AEP) calculated using the 2009 and the 2014 power curve and the observed Weibull distribution of wind speed from 2009 to 2015 is approximately 3.4%. The source of this underperformance was the erosion of the blades' leading edge.Fig. 10 clearly shows significant erosion near the blade tip. Periods without any records in 2009 and 2011 are corresponding to database failures.