Abstract
In the presented study a comparison of the difference between 10-minute wind speed data and MERRA data based on calculations of annual power losses is performed to determine if MERRA data are sufficient for that purpose and to detect the possible uncertainties of the applied procedures. Further comparisons are based on real production and wind speeds of nacelle anemometers from existing turbines. Furthermore, possible alternative calculation methods are compared in order to find a realistic procedure with reduced uncertainties and reliable data set to predict the losses of bat caused shut down scenarios. The study shows that in general the course of the wind speeds of MERRA data accompany the course of the wind speeds of 10-minute data from nacelle anemometers, but with an all in all higher level of wind speed values. With the use of scaling factors it is possible to level the bat shut down losses estimated via MERRA data with the estimated losses using 10-minute wind speed values from nacelle anemometers. However the uncertainties must be estimated as high because of the large amount of influencing factors. Furthermore, the difference of wind speed values of a single month can be quite high, which is mainly due to the aspect of the established high amount of peaks in the MERRA wind speed values.