Wind turbine wake: bridging the gap between large eddy simulations and wind tunnel experiments

Abstract

Abstract The wind industry has experienced rapid growth over the past two decades and wind energy is expected to continue leading the way in the global energy transition. Additional wind farms are expected to be installed in the coming years, both offshore and onshore. In view of exploiting any suitable terrain, an increased number of turbines are installed in close proximity to existing structures, thus exposing these to the wake turbulence. For instance, electrical cables are regularly exposed to those higher levels of turbulence compared to design loads, potentially leading to premature failure of the power line. It is hence essential to have a good understanding of these phenomena to be able to mitigate their effects. Scale-resolving investigations using Large Eddy Simulation (LES) can provide insight into site-specific loading conditions when combined with an Actuator Line Model (ALM), which reproduces the reaction force of each blade on the flow while still using a relatively coarse grid. The ALM used in this framework was newly-implemented in the spectral element code Nek5000, a highly efficient code with reduced cost per degree of freedom and exponentially fast convergence. The present work aims to bridge the gap between LES and a well controlled wind tunnel test of a scaled wind turbine, for validation purposes. First, a small scale wind turbine of diameter 80 cm is designed and tested in a well controlled wind tunnel environment. Experimental data are gathered ahead of it and in its wake, and the mean velocity and turbulence intensity profiles are presented. Although blockage effects are kept to a minimum, they will be accounted for in the simulations to ensure a better comparison with the measurements. In this work, we focus on a single inflow velocity field and on a fixed rotation speed. Simulating the produced wake using LES combined with ALM remains challenging as the results are very sensitive to the turbulent inflow conditions and also to the airfoil aerodynamics used in the ALM. In the experiment, the incoming profile is also that of a developing flow and accurately reproducing its main characteristics is required for proper inflow to the LES. A “Recycle and Rescale Method” (R2M) will be hence be used as it is well suited for such case. The method will allow for the relevant turbulent boundary layer structures to be properly reproduced numerically, hence ensuring that the wind turbine is exposed to similar inflow conditions than in the experiment. With the developed framework, the comparison between the wake produced in the experiment and that obtained using LES with ALM will provide the necessary data to further adjust the parameter and grid size used for the LES and the polar model used in the ALM.