Optimisation of LESsCOAL for large-scale high-fidelity simulation of coal pyrolysis and combustion
Principal investigator: Jun Xia
Coal is the most abundant fossil fuel on earth, accounting for over half of the fossil fuel reserve. About 25% of the electricity in the UK is produced by coal-fired power stations. In China the figure is 75%. It is clear that proper prediction, control and optimisation of pulverised-coal combustion characteristics are strategically important for both the UK and China. With rapid development of computing capacity, numerical approaches have become an important and effective research tool in this area, especially high fidelity simulation enabled by high-performance computing. Development of high-performance software tools for coal combustion will enable computational study of advanced clean coal technologies such as coal/biomass co-firing and oxy-coal combustion. One main purpose of our Chinese collaborator is to use the developed solver to investigate alkali metal release during coal pyrolysis and combustion, which is an important reason for the corrosion of industrial coal-fired furnaces.
The research work proposed in this project aims to significantly enhance the parallel code performance of LESsCOAL, which is a dedicated numerical solver for pulverised-coal combustion, on massively parallel supercomputers such as ARCHER. With the project objectives achieved, we will be able to perform large-eddy simulation of pulverised-coal combustion in a large domain such as a laboratory-scale or a small-scale industrial furnace, greatly facilitating numerical experiments of coal combustion.
Development and optimisation of such a high-performance software tool can benefit a variety of academic researchers working on turbulent multiphase flow and combustion. We also aim to publish a paper in a top-level journal in high-performance scientific computing as a conclusion of the project.