Predicting Wind Power of an onshore wind farm in Complex Terrain with Large-Eddy Simulations (LES)

Onshore wind farms are significantly influenced by terrain topography, especially in hilly regions where flow can accelerate due to favorable pressure gradients, direction changes, high shear, and turbulence from flow separation. Accurately predicting the unsteady flow dynamics of atmosphere-to-wake and wake-to-wake interactions is crucial for determining turbine power output in these areas. Lower-fidelity models often miss these complex interactions, leading to inaccurate power predictions. In contrast, Large-Eddy Simulations (LES) are essential for capturing non-linear effects and providing the necessary fidelity for detailed flow analysis. Using the open-source framework Xcompact3d, we perform high-order LES of the atmospheric boundary layer (ABL) with an immersed boundary method and a wall model to replicate terrain features, along with actuator discs to parameterize the wind turbines. This simulation environment has been validated on a steep hill using previously published experimental and numerical data. In this talk, we present the application of our methodology to a real complex terrain for forecasting power production and validating results with SCADA data. The inlet ABL is reproduced in a precursor simulation by replicating the profile from a mast on the farm. The average power and flow statistics for the predominant wind direction are assessed along with a 30-degree window average.