At a commercial scale tidal stream turbines are likely to be installed in multi-device arrays to maximise power output from a tidal site. This paper demonstrates a computationally efficient method for predicting the performance of an array. The approach parameterises the turbine by deriving momentum source terms using blade element (BE) theory. The source terms are added to Reynolds-averaged Navier-Stokes (RANS) momentum equations. The RANS+BE method is similar to blade element momentum (BEM) theory but can predict the flow field as well as the performance of the rotor. This investigation first verifies the RANS+BE method against BEM theory and shows good agreement, although tip-losses still need to be included. Secondly, it compares the RANS+BE approach to a parameterisation based on a uniform resistance coefficient, over a range of ambient turbulence values and thrust coefficients. The RANS+BE method generates increased azimuthal velocities in the near wake compared to the uniform approach. Finally the investigation compares results of a model of an infinitely wide array of turbines with five rows. The RANS+BE model can predict the performance of each turbine. and shows more rapid wake velocity recovery within the array. The investigation provides a detailed insight into the applicability of wake modelling techniques for the configuration of tidal stream turbine arrays.
展开▼
