Multi-Scale Offshore Wind Farm Modelling
Dr. Xiaosheng Chen, University of Oxford
The research project is focused on CFD modelling of wind turbine wake evolution and merger in large-scale wind farms, and particularly in addressing the lack of understanding ofwind turbine wake physics. The contemporary engineering models for big wind farms have been shown to neglect key physics, especially the wake evolution and interaction in large wind turbine arrays. This leads to limitations of the engineering models such that the accuracy of power prediction is scenario dependant and the low fidelity in lifetime yield, unsteady loading and fatigue damage rate prediction. For example, the Largest Wake Deficit (LWD) model gives a reasonable prediction downstream the 2nd turbine in a 2-turbine array when the turbines are in line with each other; meanwhile, the LWD shows a large deviation but the Linear Super-Position (LSP) model gives a close prediction in this scenario when the 2nd turbine is partially in the wake of the 1st turbine. Additionally, all those models becomeless accurate when the size of the array increases.Hence this project is aiming to develop a more reliable wake evolution reduced-order modelfor large wind farms. To achieve this, a good understanding of the turbine wake physics, both the evolution of a single wake and the evolution & interaction of multi-turbine wakes, is essential. The project will be supported by the following detailed objectives;
A. The understanding of single wake evolution and turbine performance underdifferent inflow conditions. This will be achieved by conducting blade resolved simulations of a wind turbine in uniform and non-uniform (shear profile or wakeprofile) inflow conditions over a range of turbine operating conditions andturbulence intensities.
B. The understanding of wake merger and interaction of small (2-4) turbine groupwith key array arrangements. This will be supported by CFD simulations with acombination of DES/LES and Actuator Line method.
C. The development of a wake evolution reduced-order analytic model for wakepropagation through a wind farm that conserves momentum across scales, from theindividual turbine scale to the wind farm scale.
D. Demonstration and validation of the reduced-order model through comparison withexisting engineering wake models in predicting the power distribution, unsteadyload, fatigue damage rate and lifetime yield uncertainty.