Analyzing the wake of a wind turbine actuated with Dynamic Induction Control

Abstract

To meet current renewable energy demands, there is a major increase in global wind power capacity. Wind turbines are commonly clustered in wind farms. When operating, wind turbines extract energy from the wind, creating a region characterized by a low velocity, called the wake. The wake negatively impacts the power production of downstream wind turbines. This power loss can be alleviated with wind farm control, where an upstream wind turbine deviates from optimal individual operation to improve the power production of the total wind farm. A novel wind farm control concept is dynamic induction control. While the results of dynamic control look promising, the working principles behind the increase in energy are not yet well understood. This thesis provides a framework to evaluate the transport of kinetic energy into a cylindrical control volume in a wind turbine wake using the instantaneous Navier Stokes energy equation. This approach can provide insight into the time behaviour of kinetic energy transport and can be used in conjunction with the Reynolds Averaged Navier Stokes energy equation. The wake is analyzed using Proper Orthogonal Decomposition to identify the dominant phenomena. The working principle of dynamic induction control is qualitatively described by dividing the wake into three regions: region I, dominated by pressure, region II, dominated by a vortex ring and region III, dominated by turbulence. The peak of transport of kinetic energy occurs in region II, as is confirmed by analysis through the instantaneous Navier Stokes energy equations. Finally, this thesis shows that for a simulation with uniform inflow and where the effect of the nacelle is not modelled, the transition from region II to region III is triggered by the interaction of the vortex ring with an inner vortex ring that forms around the nacelle wake. The nacelle wake is a result of not modelling the nacelle and deemed non-physical, so the ring vortex breakdown location that arises from this simulation does not have much physical meaning. While there remains a lot of work to be done, this thesis confirms the potential of dynamic induction control shown in previous research. A particularly interesting aspect of dynamic induction control shown in this thesis is that a high amount of wake recovery is compressed into a short distance in the order of $3D$. This property makes dynamic induction control a potential way to decrease turbine spacing in wind farms, allowing the placement of more wind turbines on the same area.