Assessment of the potential carbon footprint of engineered processes for the mineral carbonation of PGM tailings

Abstract

The engineered sequestration of carbon dioxide (CO2) emissions through mineral carbonation typically requires energy intensive conditions and chemical reagents to accelerate the naturally slow reaction processes. In order to be an effective carbon dioxide mitigation strategy, engineered mineral carbonation processes have to result in a net reduction of CO2 emissions. This is particularly the case for feedstock such as platinum group metal (PGM) tailings, which are comprised largely of the relatively inert pyroxene mineral. This study evaluated the viability of using these tailings for carbon sequestration through mineral carbonation on the basis of an assessment of the potential carbon footprint of selected engineered processes. The processes selected include the ammonium salts process, the direct aqueous process, the Åbo Akademi University (ÅAU) multi-stage gas-solid route, a mineral acid pH swing process and Lackner’s multi-stage HCl extraction process. Aspen Plus v8 software was used for mass and energy balance modelling, whilst the Life Cycle Assessment (LCA) software programme SimaPro v7.7.3 was used for carbon emissions accounting. The selected processes all resulted in higher emissions of carbon dioxide than those sequestered. This was particularly the case for Lackner’s multi stage HCl process (18 295 kg-CO2e) and the indirect aqueous ammonium salts (8 798 kg-CO2e) processes. The process having the lowest carbon footprint was the ÅAU process (1 354 kg-CO2e), followed by the direct aqueous process (2 364 kg-CO2e) and the mineral acid pH swing (3 126 kg-CO2e). The unit processes making the most significant contribution to the carbon footprint of the mineral carbonation process systems are heat requirements and chemical reagent make-up. Sensitivity analysis shows that the direct aqueous and ÅAU process emissions can be reduced beyond the CO2 emissions threshold when conversion is increased.