Comparison of large-eddy simulations of wakes with wind farm wake parametrizations using the Weather Research and Forecasting model

Abstract

Abstract Wind farm parametrizations are nowadays commonly used to predict the output of wind farms in real-time numerical weather prediction mesoscale models. However, their accuracy both to reproduce the wind speed and turbulent kinetic energy fields behind turbines is a matter of debate. Here, and to the authors knowledge for the first time for a single turbine, the in-built wind farm parametrization of the Weather Research and Forecasting model is evaluated using detailed large-eddy simulations of the wake performed with a generalized actuator disk model that was implemented in the same modeling system. Thus, a fairer evaluation is achieved compared to previous works, as we try to set the simulations as similar as possible within the same modeling framework. We find that both types of simulations can be used to provide similar inflow conditions to the turbine. Most importantly, by comparing the detailed-wake results, within the area where the turbine is placed, with the mesoscale results at the analogous position, velocity deficits can differ up to 50%. However, within the same area, the vertical profile of turbine-generated TKE is nearly identical between the two types of simulations. Our findings demonstrate that the generalized actuator disk implementation can provide us with a solid foundation for evaluation of wind farm parametrizations within weather models.