The actuator disk as a wind turbine model: An experimental assessment of the fluid dynamics

Abstract

Climate change is one of the major challenges of our time. With ambitions to keep the average global temperature below 1:5ý C above pre-industrial levels, significant efforts to transform the global energy production from fossil-based sources to renewable sources are required. There is, for example, steady growth in installed wind energy capacity. Typically, wind turbines are used as wind energy converters and are often grouped in wind farms. From a fluid mechanics perspective, it is interesting to study the interaction between the wakes of different turbines and the interaction between the wind farm flow and the atmospheric flow. Much of the research on the topic is done numerically with actuator disks as wind turbine models. Actuator disks, in the form of porous disks, have also become a popular tool to simplify lab-scale experiments. However, there is no general agreement on the design of these devices, with very different designs found in the literature. Although the wakes of actuator disks and wind turbines have been compared for single wake generating objects, there is still a knowledge gap in understanding how actuator disks replicate the flow in a wind farm. The first two projects in this thesis compare the flow around two different actuator disks, one with a uniform blockage and one with a non-uniform blockage. The wakes are measured through hot-wire anemometry at six streamwise positions, ranging from 3 to 30 diameters downstream of the objects. Neither disk can reproduce the asymmetry seen in a reference wind turbinewake in a laminar inflow. Moreover, the wakes of the two actuator disks are different. Firstly, the wake profiles differ in both velocity deficit and turbulence intensity. Secondly, the structures of the wake are different, with vortex shedding being observed only in the non-uniform disk wake, even though the total blockage ratio of the disks is the same. Increased ambient turbulence levels somewhat reduce these differences, but they are not eliminated. Lastly, there is a ring of highly intermittent flow around the actuator disk wakes, which has only been documented for a wind turbine wake earlier. The third part of this thesis details a design process of an actuator disk. Several actuator disks are compared to a rotating wind turbine model. The thrust, or drag, of the wake generating objects is recorded with a force plate, and particle image velocimetry data of the near-wake is collected. The best design shows acceptable agreement in the mean quantities of the flow profiles between the wind turbine model and the actuator disk. Nevertheless, proper orthogonal decomposition and analysis of vortex structures show that instantaneous features of the flow are not matched. For the last part of the thesis, the flow in the induction and entrance regions of wind farms is compared between actuator disks and rotating wind turbine models. Flow fields are captured through particle image velocimetry. It is shown that there are no measurable differences in the induction between farms consisting of rotating models and actuator disks, nor between different farm layouts and incoming flow angles. In the entrance region, the impact of instantaneous features in the wind turbine wakes is significantly reduced downstream of the second row of wind turbines, such that the flow downstream of the second row of actuator disks more closely resembles the flow downstream of the second row of turbine models. In summary, the results show that it is possible to use actuator disks as static wind turbine models in lab-scale experiments, in particular when evaluating global features of the wind farm flow. Nevertheless, it is essential to consider the actuator disk design to replicate the wind turbine flow optimally. Furthermore, it is important to know that some features of the flow, in particular instantaneous vortex structures and asymmetries in the wind turbine flow, cannot easily be reproduced by actuator disks.