Computational fluid dynamics modelling of semi-dry flue gas desulphurisation in a spray dry scrubber

Abstract

The combustion of fossil fuels releases SO2 to the atmosphere, which has known negative environmental and health impacts. SO2 produces acid rain and has been shown to have a deleterious effect to aquatic life amongst many harmful effects. Companies are thus under pressure to minimise their SO2 emissions and with ever increasing focus on climate change and global warming, South Africa and in particular Eskom, which supplies over 90% of the country’s energy, are mandated to abate SO2 emissions from existing power plants and related infrastructure.

In this work, a comprehensive model for the desulphurisation process in a spray–dry scrubber was developed in the StarCCM+ modelling platform. The work aimed at improving the efficiency of the process as well as understanding the complex hydrodynamics, drying kinetics and SO2 absorption inside the dryer. These three key phenomena were systematically modelled, with the model validated at critical points of the development process using experimental data obtained from a modified BUCHI B-290 laboratory spray dryer. The modelling approach began with singlephase modelling, which saw the implementation of three turbulence models, the Standard k–ε, Realizable k–ε and SST k–ω models. These were investigated for their ability to accurately characterise the system hydrodynamics and the best model was identified as the SST k–ω model based on the root mean square error (RMSE). The developed single phase was then further amended into a two–phase model which incorporated the Euler–Lagrangian framework in describing the spray drying process. The drying models that were investigated are the perfect shrinkage model, the classic d2law model and the mechanistic model. The best quality of fit was obtained with the mechanistic model followed by the classic d2law model. Finally, a submodel describing the SO2 absorption process was then implemented to complete the desulphurisation model. The two drying models were then evaluated for their impact on the SO2 removal efficiency, and ultimately the mechanistic model was found to be the best drying model based on the quality of fit between the model predicted and SO2 absorption measurement data. Sensitivity analysis was conducted on key process variables (inlet flue gas temperature, the Ca/S ratio and the L/G ratio) affecting both the drying and SO2 absorption process. An increase in the inlet temperature of the flue gas was attended to by a decrease in the SO2 removal efficiency. On the other hand, increasing both the Ca/S ratio and L/G ratio results in an increase in the SO2 removal efficiency, although coming at reduced sorbent utilisation and increased solid product moisture content respectively. The work has improved the fundamental understanding of the spray–dry scrubbing process and provided insights for further improvement.