Validation of Actuator Line against Geometrically Resolved Blades for Wake Steering with a 22MW Turbine

Wake steering through intentional yaw misalignment of upstream turbines is a potentially promising method to increase the power output of wind farms, by mitigating wake effects under certain wind directions and inflow conditions [1]. Considering computational modelling, the most high-fidelity tool feasible for studying wind farm flow control is large eddy simulations (LES), using the actuator line (AL) [2] method for turbine modelling. The AL significantly reduces the computational requirements in comparison to geometrically resolving the rotor blades, by representing the turbine blades through applying body forces in the computational domain. LES with the AL allows time-dependent, turbulence-resolving studies which can accurately represent turbine wake flow, including the transition between near and far wake, wake dynamics, and tip vortices. However, the AL still relies on simplifying assumptions. For example, the use of 2D aerofoil data in load calculation, the impact of the grid resolution (which is significantly coarser than for geometrically resolving the blades) and the use of a smearing parameter. Therefore, it is important to verify and compare the AL against results from geometrically resolved blades, in order to test the sensitivity of conclusions to the choice of computational method.
Previously, Hodgson et al. [3] compared an AL coupled to the aeroelastic solver Flex5 with geometrically resolved blades coupled to Hawc2 for the NM80 turbine, which has a radius of 40m. In a simple inflow case with uniform laminar inflow, load distributions were very similar and thrust predictions agreed to within 0.2%. For more complex inflow cases including shear and yaw misalignment, there was good agreement in the trends of loading and deflection, but the coupled AL displayed an underestimation of the loading and hence a lower thrust. The comparison performed in the current work has a number of differences to the aforementioned study. Firstly, through the Dynamiks coupling framework, it is now possible to couple the AL with Hawc2 rather than Flex5, therefore including the effect of torsion in the AL and using an identical aeroelastic solver between the two models. Secondly, in this study we use the IEA22MW turbine, which has a radius of 142m. This significantly larger and more modern turbine design will present a more challenging case, because of the much larger deflections and more flexible blade design. And lastly, as this study is specifically interested in wake steering through intentional yaw misalignment, a more in-depth comparison of the wake will be undertaken in addition to turbine loading.