Abstract
In this research project, a numerical model and study, using fully transient simulations, of the turbulent flow over a generic 3-blade horizontal axis tidal turbine (HATT) subjected to real tidal flow inlet velocities, as captured in the lower Bay of Fundy (Digby area), was performed in order to determine the actual turbine behaviour under real turbulent tidal flow conditions. This numerical study shed more light on the findings from the Queen’s University group as to the lowering of the turbine Cp and power production in a real tidal flow. The results of the simulations also provide additional insight into the wake dynamics and led to a comparison between steady flow wakes as studied previously and unsteady flow wakes encountered in real tidal flow environments.
The three objectives of this work were: 1)Develop a numerical model, using ANSYS CFX, enabling reliable study of unsteady turbulent flow over a horizontal axis turbine using real-life tidal flow data from the lower Bay of Fundy (Digby area), 2) Characterize the nature of the turbulent flow in the wake (length, zone of impact, strength of turbulence), and 3) Compare the unsteady results to the steady results obtained using the previously developed methodology in order to determine the impact on the turbine power and thrust values, and the wake behaviour.
A numerical model of a three bladed horizontal axis tidal turbine under realistic turbulent tidal flow was created. The results of this investigation have been compared with steady flow numerical model results with good agreement in trends. Prediction of both Cp and Ct are very similar in both cases. These similar trends observed in both Cp and Ct curve are important as they indicate that the appropriate flow physics are being accounted for. Only TSR values related to maximum Cp and Ct changed. For maximum Cp, tip speed ratio is approximately equal to 3.9 in steady flow and 4.1 for transient conditions. In transient flows, this results however in an approximate 4% reduction in performance (for TSR = 3.5), though there is increased uncertainty due to the levels of scatter in the numerical data points.
Velocity deficit plots show the wake is wider in transient simulations than steady ones. The velocity deficit disappears also faster in the constant velocity simulations. Turbulent effects in the wake seem to increase after a distance of 10D downstream of the turbine in this setup. These turbulence effects are higher in the transient simulations.
This comparative analysis of numerical steady and transient simulations shows the impact of the unsteadiness of realistic tidal flows on the performance of tidal turbines. Based on these results, the current use of steady state testing (numerical or experimental) for design stage can be questioned. The observed changes in the wake’s characteristics and the high variations of the loads on the blades (reference to Ct-curve as a function of time) must also be better assessed.