High-frequency field measurements and time-dependent computational modelling for wind turbine siting

Field measurements by cup anemometer at a sampling frequency of 1 Hz and computational modelling were used for characterisation of the wind conditions at a region of complex orography. Because the wind rose displayed low wind occurrences and high turbulence intensity for specific wind directions, spectral analysis of computational data was performed which showed the existence of preferred time scales for some wind directions, in agreement with spectral analysis of field data. Empirical orthogonal functions and frame by frame analysis of the time-dependent computational data showed that these time scales were related with a pair of vortices in the lee of a mountain. We concluded on the appropriateness of the methodology being used, which enabled the determination of these structures without reverting to the use of more complex and time consuming techniques. This study, initiated as part of a wind resource evaluation and possible installation of a wind farm, also proves the need for more detailed determination of the wind conditions, compared with the standard practice in wind turbine siting.