Comparison of the dynamic wake meandering model against large eddy simulation for horizontal and vertical steering of wind turbine wakes

Abstract

Accurately predicting the evolution of wake is crucial for power output and structural load estimation in wind farms. This study aims to validate the dynamic wake meandering (DWM) model, an efficient mid-fidelity wake model, against large eddy simulation (LES). The predictive capabilities of the DWM model for various wake properties, namely time-averaged wake deficit, mean wake center deflection, amplitude, and frequency spectrum of wake meandering, are comprehensively analyzed using the IEA 15MW reference wind turbine under different yaw and tilt misalignment angles. Two turbulent inflows with varying shear and turbulence intensity levels are considered. The comparison highlights the significance of the filter size (Cmeand) in DWM as a key parameter determining simultaneously the time-averaged wake deflection and meandering amplitude, with optimal values differing for horizontal and vertical wake displacements. When the appropriate Cmeand values are selected, the implementation of the DWM model in FAST.Farm demonstrates good agreement with LES data, particularly concerning time-averaged wake deficit, wake centerline deflection, and wake meandering amplitude at eight rotor diameters downstream. However, the DWM model tends to overestimate the energy in the lower frequency region with Strouhal’s number less than 0.1 and underestimate the wake oscillation induced by the shear-layer at higher frequencies, even though the wake motion standard deviation is accurately reproduced if the polar grid size is properly adjusted. Furthermore, the influence of the ground effect on downward wake deflection through tilt control is revealed. These findings clearly demonstrate the strengths and weaknesses of the current DWM model and can serve as a reference for the development of advanced wake models.