Large-Eddy Simulations of Wind Turbine Wakes Subject to Different Atmospheric Stabilities

Abstract:

As a byproduct of energy extraction, wind turbines create a low-speed, turbulent wake that propagate downwind. When wind turbines are situated in a group, as in a wind plant, the interactions of these wakes with other turbines are important because wake effects decrease the efficiency of the wind plant, and they increase mechanical loads on individual turbines. Wakes propagate downstream differently depending on the inflow conditions, and these conditions are heavily dominated by atmospheric stability. For example, we know that wakes are more persistent in stable conditions than in unstable conditions. Also, stable conditions often have significant wind veer which skews wakes laterally. Different levels of turbulence intensity are associated with different atmospheric stability levels, and turbulence intensity acts to diffuse wakes and to cause wake meandering.

Wake physics are complex, and to understand them better, a high-resolution representation of the flow is necessary. Measurements are difficult with current sensing equipment because of the sheer size of wakes and the unsteady atmospheric environment in which they are found. Numerical simulations complement measurements and provide a high-resolution representation of the entire three-dimensional, unsteady flow field. In this work, we use large-eddy simulation (LES), the highest fidelity type of computational fluid dynamics (CFD) feasible for high-Reynolds-number wake flow. LES directly resolves the larger, energy-containing turbulent scales and models the effects of the subgrid scales that the computational mesh cannot resolve. Our solver is based on the OpenFOAM open-source CFD toolbox. Turbines are modeled using rotating actuator lines.

Here, we present our LES of the wake behind a modern 1.5 MW turbine subject to different inflow atmospheric stability. We will present results of wakes subject to stable (strongly and weakly stable), neutral, and unstable conditions. We are particularly interested in how stability affects wake recovery, wake skewing, and wake meandering. Figure 1 shows horizontal slices of instantaneous contours of vorticity magnitude in the computed wake of a turbine subject to weakly stable atmospheric inflow. A multi-resolution mesh is used with the finest region of 1.25 m resolution surrounding the turbine and the wake.