Analysis on the relationship between wind veer and the wind turbine wake in the context of wake steering

Abstract

Wake steering is the intentional yaw misalignment of wind turbines to deflect the wakes away from turbines downstream and improve overall farm power production. It is most applicable in the stable atmospheric boundary layer (SBL), which is characterised by low turbulence intensity and high wake losses. Wind veer is common to the SBL and is the change in wind direction with height. This has been shown to affect the skew of the wake and its trajectory at different heights across the rotor. This thesis investigates the influence of wind veer on wake behaviour, in the context of wake steering. The Simulator fOr Wind Farm Applications (SOWFA) was used to model the turbine and SBL. The turbine yaw angle and Coriolis forcing were varied to represent wake steering and generate different inflow veer profiles. The wake shape was observed to skew according to the inflow wind veer profile, with the veer of the wake shown to match the inflow veer near the rotor before fading with downstream distance. The inflow wind veer profile and wake shape skew were found to have a directional impact on the deflection of wakes generated in veered flow. The direction of skew observed in the wake region below hub height appeared to dominate the wake centre location. Skewed wakes were shown to recover faster than wakes produced in non-veered flow. A two turbine simulation in non-veered and veered flow revealed a disparity in the upstream turbine yaw angle that improved the net power, indicating that wake steering controllers must account for the effect of wind veer on wake behaviour to maintain efficacy between veered and non-veered flows. Wake models used in wake steering controllers were compared with the simulation results and found to poorly capture veered wake dynamics. The thesis outcomes highlight implications for wake steering for wind farms in veered wind flow and provide a basis for future development of wake steering controllers, wake models and applying both in realistic conditions.