Stefano Alessandrini* Goffredo Decimi*, Enrico Ferrero** Seminario divulgativo presso il centro ECMWF a Reading Reading (UK), 9 Giugno 2009 PRESENTAZIONE POWER POINT * ERSE SpA ** UNIVERSITA’ DEL PIEMONTE ORIENTALE Most of the wind farms installed in Italy are located in the southern regions in complex topography conditions. In these conditions the wind power forecast, strictly dependent on the accuracy of wind predictions, represents a more difficult task than in flat terrain situations. A wind power forecast system suitable for complex topography situations has been developed. It is based on a high-resolution meteorological model and a set of post processors that reduce the systematic errors and compute the predicted power of the wind farm. The forecast system has been applied at two wind farms data-sets. One is located in the south Italy, with 10 Mw of nominal power in a complex topographic situation, while the other is the KLIM wind farm, located on a flat area with 21Mw of nominal power, whose data are supplied in the frame on the ViLab exercise promoted by the POW’WOW project. The main scope is to understand the performances in predicting the power production. A particular attention is given to the comparison between the accuracy obtained in flat terrain and complex topographic situations. Another objective is to understand the influence of the meteorological model’s resolution on the whole system performances. Finally the errors introduced in the two steps of the forecast process are estimated. The forecast system has slightly better performances in flat terrain conditions with a RMSE of 15% of nominal power against a 16% of the Italian case. In both cases the step of the power computation introduces an error around 8%.