2013

Karlsson, Timo; Turkia, Ville; Wallenius, Tomas

Icing conditions may cause significant wind power production losses, which increase short-term wind power forecasting errors if not taken into account. In this study real measurements and production data were used to evaluate a method to estimate production losses due to icing. The result of this study is called production loss module and it can be used as part of wind power forecasting system. Production loss module defines production losses based on icing time and wind speed inputs from weather prediction model. This publication describes only the production loss module and the procedure how it was created. Weather prediction models and forecasting of icing events are not considered here. Production loss estimation method used by the module is based on icing induced power loss in reference data. Icing periods in real production data were classified based on wind speed and duration of power loss. Statistics of these power losses were then used to estimate the production losses during icing events. Power loss gets increasingly more severe the longer the icing event lasts. The largest power loss occurs during the first icing hour. This would suggest that even short icing events can have significant effects on wind power plant performance and on short term wind power forecasting.