On the energy output estimation of wind turbines

Abstract

During the operation of the German test field for small Wind Energy Conversion Systems (WECS) on the island of Pellworm five wind turbines were tested following recommendations of the International Energy Agency (IEA) expert group. Possible errors in the estimation of a tested wind turbine's total energy output at a potential installation site are investigated. Different wind speed frequency distributions (the measured one, the Rayleigh and the two-parameter Weibull distribution) are used to calculate the total energy output. The differences between the various distributions are mostly below 10 per cent. An improvement of the energy output estimate by a Weibull-instead of a Rayleigh distribution was not found. It is also shown that the use of the recommended 10 min averages or any other average overestimates the WECS' efficiency, up to 14 per cent on average depending on turbulence intensity. Wind power instead of wind speed is the appropriate parameter for power performance testing. Spectra of wind power and electrical power output show three areas of different correlation. A resistance length for wind turbines is shown to be dependent on the WECS operation status.