Simulated wind farm wake sensitivity to configuration choices in the Weather Research and Forecasting model version 3.8.1

• Published in: Geoscientific Model Development 2020-06, Vol.13 (6), p.2645-2662
• Main Authors: Tomaszewski, Jessica M, Lundquist, Julie K
• Format: Article
• Language: English
• Subjects: Analysis; Atmospheric boundary layer; Computer simulation; configuration choices; Configurations; Diurnal; Electricity; Energy; Kinetic energy; Mathematical models; Meteorological observations; Meteorological research; Numerical weather forecasting; Numerical weather prediction; Parameterization; Prediction models; Resolution; Sensitivity; simulated wind farm wake sensitivity; Simulation; Surface temperature; Temperature; Turbine engines; Turbines; Turbulence; Turbulent kinetic energy; Wakes; Weather; Weather forecasting; weather research and forecasting; Wind effects; WIND ENERGY; Wind farms; Wind power; Wind power generation; Wind power plants; Wind speed; Wind turbines
• ISSN: 1991-9603; 1991-959X; 1991-962X; 1991-9603; 1991-962X
• DOI: 10.5194/gmd-13-2645-2020

Wakes from wind farms can extend over 50 km downwind in stably stratified conditions. These wakes can undermine power production at downwind turbines, adversely impacting revenue. As such, wind farm wake impacts must be considered in wind resource assessments, especially in regions of dense wind farm development. The open-source Weather Research and Forecasting (WRF) numerical weather prediction model includes a wind farm parameterization to estimate wind farm wake effects, but model configuration choices can influence the resulting predictions of wind farm wakes. These choices include vertical resolution, horizontal resolution, and whether or not to include the addition of turbulent kinetic energy generated by the rotating wind turbines. Despite the sensitivity to model configuration, no clear guidance currently exists for these options. Here we compare simulated wind farm wakes produced by varying model configurations with meteorological observations near a land-based wind farm in flat terrain over several diurnal cycles. A WRF configuration comprised of horizontal resolutions of 3 km or 1 km paired with a vertical resolution of 10 m provides the most accurate representation of wind farm wake effects, such as the correct surface warming and elevated wind speed deficit. The inclusion of turbine-generated turbulence is also critical to produce accurate surface warming and should not be omitted.