| dc.description.abstract | Metallurgical-grade silicon is a critical material for modern society. It is used as an alloying agent in aluminium alloys, feedstock for silicones, photovoltaics and electronics. Silicon is produced by carbothermal reduction of quartz to silicon, using carbonaceous reductants, which will be a source of emissions of CO2. Although there have been initiatives for finding alternative carbon-neutral production processes for producing silicon, it is still the carbothermal production route that dominates the industry, and hence, the process must be modified to meet the need for emission reduction of greenhouse gases. Out of the several proposed alterations of the conventional production process, retrofitting post-combustion carbon capture (PCC) to silicon smelters has gotten increasingly more attention. Flue gas recirculation (FGR) could in this sense be a tool to increase the CO2 concentration in the off-gas from silicon production. Higher CO2 concentration in the off-gas reduces both OPEX and CAPEX of carbon capture and could also enable additional CC technologies if increased sufficiently. FGR for silicon production has previously been theoretically investigated for NOX reduction, but there is a lack of experimental research on the subject.
In the work performed in this thesis, flue gas recirculation for the silicon process was experimentally and theoretically studied to evaluate its effects on the process, including CO2 concentration in the off-gas, possible future carbon capture technologies and reduced NOX emissions. This was done through mass and energy modelling of the silicon process and a coupled carbon capture process. Small-scale experiments were conducted to investigate the effect of different gas atmospheres on silica fume formation. A larger pilot-scale experiment of FGR for silicon production was also conducted, demonstrating the concept in close to industrial conditions. Combustion of the process gases SiO and CO in the tapping gas was also studied on an industrial scale during a measurement campaign on a large silicon furnace.
Modelling the silicon process showed that FGR is an efficient way of increasing the CO2 concentration in the off-gas, with the benefit of not increasing off-gas temperature. The cost of carbon capture in a conventional scrubber stripper setup was found to be reduced with increasing FGR as higher CO2 off-gas concentrations allowed for more efficient capture with smaller components and a lower solvent flow, leading to lower specific energy consumption for capture. Small-scale experiments of SiO combustion in different atmospheres indicated that replacing O2 in the combustion air with CO2 had an insignificant impact on silica fume morphology. However, expanding the combustion of SiO gas from a small-scale to a pilot-scale test resulted in indications that the specific surface area of the silica fume increases with increasing FGR. The pilot-scale experiment was also a successful demonstration of how CO2 concentrations could be increased to over 20 vol % and with reduced specific NOX emissions, although not as significantly as previously predicted by modelling. During the tapping gas measurement campaign, inert gas injection in the tapping gas intended to suppress NOX formation was performed but was not successful. NOX generation was however found to be stronger correlated to the energy added to the tapping gas stream than by the amount of SiO combustion, as previously assumed. It was observed that particulate matter (PM) formed from the oxidation of liquid Si did not generate NOX to the same extent as the combustion of SiO gas directly from the taphole. Process crater temperatures were estimated to be in the range of 1890 to 2200 °C, using the concentration of PM and CO2 in the tapping gas. | en_US |
