Modeling of carbon dioxide absorption using aqueous monoethanolamine, piperazine and promoted potassium carbonate

Abstract

Rigorous CO₂ absorption models were developed for aqueous 4.5 m K+/4.5 m PZ, monoethanolamine (7m - 9m), and piperazine (8m) in Aspen Plus® RateSepTM. The 4.5 m K+/4.5 m PZ model uses the Hilliard thermodynamic representation and kinetics based on work by Chen. The MEA (Phoenix) and PZ (5deMayo) models incorporate new data for partial pressure of CO₂ vs. loading and kinetics from wetted wall column data. They use reduced reaction sets based on the more relevant species present at the expected operating loading. Kinetics were regressed to match reported carbon dioxide flux data using a wetted wall column (WWC). Density and viscosity were satisfactorily regressed to match newly obtained experimental data. The activity coefficient of CO₂ was also regressed to include newly obtained CO₂ solvent solubility data. The models were reconciled and validated using pilot plant data obtained from five campaigns conducted at the Pickle Research Center. Performance was matched within 10% of NTU for most runs. Temperature profiles are adequately represented in all campaigns. The calculated temperature profiles showed the effect of the L/G on the location and magnitude of the temperature bulge. As the L/G is increased the temperature bulge moves from near the top of the column towards the bottom and its magnitude decreases. Performance improvement due to intercooling was validated across the campaigns that evaluated this process option. Absorber intercooling was studied using various solvent rates (Lmin, 1.1 Lmin and 1.2 Lmin). It is most effective at the critical L/G where the temperature bulge without intercooling is in the middle of the column. In this case it will allow for higher absorption by reducing the magnitude of the bulge temperature. The volume of packing to get 90% removal with L/Lmin =1.1 at the critical L/G is reduced by 30% for 8m PZ. For MEA and a solvent flow rate of 1.1 Lmin packing volume is increased with intercooling at constant L/G. This increase is compensated by higher solvent loadings that suggest lower stripping energy requirements. The critical L/G is 4.3 for 8m PZ, 6.9 for 9m MEA and 4.1 for K+/PZ.