Henry Korb doctoral thesis defence

Henry Korb’s doctoral thesis investigates how to make simulations of wind farm flows faster and cheaper.
Through seven articles, the thesis builds a software framework based on a novel approach to high-fidelity simulations of wind farm flows utilising graphics processing units to reduce the cost of simulations,  accelerating research and enabling industrial application of high-fidelity simulations. 

Understanding how the wind flows through and interacts with wind farms requires the use of large supercomputers and is very expensive and time consuming. 
The lattice Boltzmann method is a promising approach to lower the cost of these simulations by several orders of magnitude through the efficient use of GPUs.
Such a cost reduction does not only enable researchers to gain a more thorough understanding of wind farm flows faster, but is also an important step towards industrial application of high-fidelity models, ultimately lowering the cost of electricity. The thesis brings the methodology from academic cases closer to simulations of complex, real-world scenarios.

The thesis aims to broaden the understanding of the wind energy community of the new method and establish the hurdles currently facing industrial adoption. A novel software framework is developed and validated against measurements and improved to model wind turbines more accurately and include the effect of temperature on the flow. The thesis also applies the approach to train a machine learning model and examine a novel approach to control wind turbines, improving the efficiency of large wind farms.