CFD modelling of wind flow over terrain
The aim of this project is to show the capabilities of a RANS based numerical model in accurately analysing wind flow over real terrain regions, and assess its usage for wind energy applications. The main reasons this type of model is not widely used in the wind energy industry are due to the computational cost and the expertise required to operate such a model. These factors are assessed and various setups of the model are examined to consider the accuracy attained. The modelling process is also automated to reduce necessary user input in the process. The models performance is tested over a number of terrain types: Flat terrain (with surface roughness), an axisymmetric hill and a real terrain region (the Askervein hill). Primary consideration is given to velocity speed-up predictions which are paramount when considering the energy availability in the wind. A number of turbulence models have been tested for each terrain region to assess the improved accuracy obtained by using a more complicated CFD setup. The mesh discretisation has also been analysed for sensitivity to change, providing a comprehensive analysis of wind flow over Askervein. The CFD setup process is automated to reduce the time taken in setting up a model and increase the speed of providing a full wind field assessment for all wind directions, and allowing determination of average yearly values of velocity. This improves the access to such models for non-expert users and improves the availability of the model to wind energy developers siting farms in complex terrain regions. The model is shown to perform well for all terrain and roughness types. The turbulence properties are not well modelled, and that is a known limitation of this model type. The project demonstrates the advantages of CFD models for wind energy applications through the presented results and successful automation of the process.