The DOWA was developed to improve hourly correlation and to better resolve the vertical profile of wind speed without incorporating a uniform shear-correction factor. Comparison to LiDAR measurements at eight sites and to instrumented meteorological mast measurements at two locations demonstrates that the hourly correlation in the DOWA is better than in the KNW-atlas. In concrete numbers, considering all LiDAR measurement locations and heights, the mean linear regression slope increased from 0.94 (KNW) to 0.97 (DOWA) and the mean R2 value increased from 0.87 (KNW) to 0.91 (DOWA).
Related to the above, the diurnal cycle is better represented. This, without losing significant ability to denote monthly mean and annual average wind speeds. Also, the DOWA performs similar to the KNW-atlas in its depiction of vertical changes in wind speed with height (i.e. vertical wind shear). A better representation of the diurnal cycle is demonstrated from the fact that considering all LiDAR measurement locations, the DOWA shows a reduction of hourly wind speed bias from a mean value of 1.53m/s within the KNW-atlas to a mean value of 1.27m/s within the DOWA. Also, the scatter around the mean wind speed bias is reduced; on average, the 𝜎 value of the hourly wind speed bias was reduced by 0.26m/s (16.99%).
Comparison to EMD-ConWx
The performance of the DOWA in representing North Sea wind conditions is compared to the EMD-ConWx European mesoscale dataset that were modeled in-house in collaboration between EMD and ConWx. The performance of the two datasets is defined relative to light-detection and ranging (LiDAR) measurements at three offshore measurement sites (IJmuiden, Lichteiland Goeree, and the Europlatform). The most significant conclusion from this report is that the DOWA provides improved hourly wind speed correlation compared to the EMD-ConWx European mesoscale dataset.
For more information: https://www.dutchoffshorewindatlas.nl/publications/reports/2019/06/07/tno-report—representation-of-north-sea-wind-conditions-in-the-dowa-and-the-emd-conwx-european-mesoscale-dataset
Improved wind resource assessment
An additional validation study compares observed yields from the Offshore Wind farm Egmond aan Zee (OWEZ) with simulated production yields. The simulations are performed with TNO’s wind farm design tool FarmFlow, using time series of both DOWA and KNW wind atlases. For both wind field sources the results show very good correlations between the simulated and observed power productions per wind turbine. After calibration with a measure-correlate-predict method (MCP), both wind atlases perform similar. Without calibration, DOWA benefits from a more accurate prediction of the vertical wind shear, resulting in better predictions of offshore wind farm yields in the North Sea.
For more information: https://repository.tudelft.nl/view/tno/uuid:511e01b5-352a-4b36-8448-548e5d230219