Reduced-Order Modeling of Wind Turbine Wakes with the Generalized Quasilinear Approximation

Wind farms consist of tightly spaced turbines whose wakes interact in complex ways, affecting power output and structural loading. Accurate modeling of these multiscale turbulent interactions remains computationally challenging. This work investigates the use of the Generalized Quasilinear (GQL) approximation for developing reduced-order models capable of resolving turbine wake dynamics efficiently. We compare GQL models at different streamwise spectral cutoffs against Direct Numerical Simulation (DNS) and Quasilinear (QL) models for a single turbine wake. The turbine forces are applied to the flow field using an actuator disk model. These simulations help identify which nonlinear interactions are important for accurately predicting downstream wake behavior. Based on the results, we conduct systematic statistical analyses to quantify model discrepancies. Time-averaged velocity profiles at several downstream locations show that QL overpredicts the near-wake velocity deficit and underestimates shear at the wake boundaries. The insights gained through this research aim to inform the design and operation of next-generation wind farms for improved performance, longevity, and energy capture.