Advanced nacelle anemometry and SCADA
Transcription
Advanced nacelle anemometry and SCADA
Advanced nacelle anemometry and SCADA-data, analysis techniques and limitations Frank Ormel Chief Specialist in Product Integration Vestas Outline • Introduction • State of the art • Advanced methods • Nacelle Power Curve standard • Conclusion Contributors: Irene Lauret Ducosson, Martin de Mare, Rolf Erik Keck, Carsten Krogh Nielsen 2 EWEA presentation July 2012 Lyon The purpose: performance monitoring and park optimisation 3 EWEA presentation July 2012 Lyon The problem: correlation to free stream wind 4 EWEA presentation July 2012 Lyon The problem: correlation to free stream wind 5 EWEA presentation July 2012 Lyon Apparent AEP & real AEP The apparent effect of a nacelle anemometer measurement error Original power curve The real effect of the same nacelle anemometer measurement error 6 EWEA presentation July 2012 Lyon It seems like the power performance has ben reduced by 10% whereas in reality it has been reduced by 0.5% The problem: seasonal variation 7 EWEA presentation July 2012 Lyon State of the art: detection of icing Scatterplot of AvgPower vs Wind speed (Normalized) 1800 Day 2010-12-24 2010-12-25 2010-12-26 2010-12-27 2010-12-28 1600 1400 A vgPower 1200 1000 800 600 400 200 0 2 4 6 8 10 12 14 16 Wind speed (Normalized) Scatterplot of AvgPower vs Wind speed (Normalized) 1800 Day 2010-12-18 2010-12-19 2010-12-20 2010-12-21 2010-12-22 2010-12-23 2010-12-24 2010-12-25 2010-12-26 2010-12-27 2010-12-28 2010-12-29 2010-12-30 2010-12-31 1600 1400 A vgPower 1200 1000 800 600 400 200 0 8 EWEA presentation July 2012 Lyon 2 4 6 8 10 Wind speed (Normalized) 12 14 16 State of the art: incorrect pitching 9 EWEA presentation July 2012 Lyon State of the art: overview In general, we can confidently detect the following field performance issues: • Icing • Yaw error • Pitch problems • De-rating problem • Curtailment • Incorrect control parameters influencing performance • Incorrect nacelle transfer function • Directional influence 10 EWEA presentation July 2012 Lyon Standardisation: IEC 61400-12-2 standard The project team has submitted the FDIS to the IEC June 2012. When approved by the national committees the standard can be expected around summer 2013 The standard is not easier than -12-1; a lot of focus has been put in to reduce variance. Still, total uncertainty of the test is higher than -12-1 standard (power curve test with met-mast) Useability will largely depend on OEMs to be compliant with this standard through their software – in which case costs are significantly lower than for a -12-1 test Main challenge is to separate terrain influence on wind speed from rotor infuence 11 EWEA presentation July 2012 Lyon Advanced techniques: NTF calculated by CFD WS Correlation graph of Free WS vs Anemometer Sensor WS (CFD & Field) 18.00 Some encouraging results but high variability from site to site and operator to operator. Not ready for field testing or standardisation Anemometer_Right-CFD Anemometer_Left-CFD Anemometer_Right-Field Anemometer_Left-Field 16.00 Anemometer Sensor WindSpeed (m/s) 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 0 2 4 6 8 10 12 14 16 Freestream WindSpeed (m/s) Wind speeds 12 EWEA presentation July 2012 Lyon Actuator disc Actuator line Field data 4 0.998 0.982 1.010 6 1.024 - 1.105 8 1.018 1.025 1.036 10 1.021 1.037 1.025 12 1.024 1.032 14 1.015 - 1.052 - 16 1.009 1.008 - Advanced techniques: Spinner anemometry • • • • • • 13 EWEA presentation July 2012 Lyon Site calibration done Turbine running with spinner anemometer for half a year Availability 51% (100% of the data) per 10-min data set, 97% available (99% of the data per 10-min data set) Correlation to free stream wind has R2 of 0.89, which is less than standard anemometer (R2 = 0.97) Correlation to free stream turbulence is equally bad between standard anemometer and spinner anemometer More data requirement for spinner anemometery to establish itself as a valuable option Limitations and Conclusions In order to get value out of nacelle anemometry data we have to overcome these limitations: • Data quality currently lowers the capability of the analysis – we are looking with ‘foggy glasses’ and results are not clear • Main challenge of understanding the physics through the available data – data must be used to verify hypotheses, not only to postulate problems • Data must be used together with other sources (other turbines, met-masts, other signals, logbooks While addressing these challenges, we can identify and solve a number of problems with park operation and add value by optimising park performance 14 EWEA presentation July 2012 Lyon Thank you for your attention