A Comparison of sector-scan and dual Doppler wind measurements at Høvsøre Test Station – one lidar or two?

摘要

Long range scanning lidars have the ability to be deployed along the coastline to measure the near shore wind resource. Within the wind energy scope, this is most applicable to assessing the potential energy production (and thus revenue) of a prospected near shore wind farm (here defined as 3-12km from the coast). Ground based remote sensing has numerous advantages over traditional in-situ (offshore met mast) and buoy based installations, mainly in terms or cost, complexity, and failure/delay risk. Sinceeach lidar can only measure a portion of the wind vector, it is necessary to either deploy two devices in tandem (dual Doppler) or employ a single Doppler scanning strategy such as PPI (plan position indicator, or sector scan) which allows for estimation of the two component horizontal wind vector. In preparation for a six month long measurement campaign along the Danish North Sea, a one week experiment was performed at DTU’s test centre for large wind turbines (Høvsøre), which lies 1.8km inland and consists offlat terrain with predominate winds from offshore. Two lidars in staring dual Doppler mode and one lidar performing 60 degree sector scans had their beams collocated atop a 116.5m met-mast, which provided reference wind speed and direction values. The 10 minute reconstructed lidar measurements were in excellent agreement with the reference instrumentation. The dual Doppler results matched within 0.1% of the reference wind speed, with very low levels of unbiased scatter. Sector scan results also indicate verygood agreement with the met-mast, corresponding within 0.2% for wind speed, with an R2 of 0.998. The sector scan results for wind speed exhibit larger amounts of scatter than with dual Doppler, however the bias is centred around the regression line which gives good indication that wind measurements taken using the sector scan method are valid and acceptable for use in performing wind resource studies in simple terrain and in offshore conditions. Further, we show that when measuring in these cases, a sector size of 38 degrees still measures within 0.6% of the reference data for wind speed.