Carbon capture and storage using chemical absorption is a viable method for reducing CO2 emissions from the industrial sector. Thermodynamic analysis of MDEA-PZ aqueous alkanolamine solution is of paramount importance to better simulate and improve the efficiency of CO2 capture processes. In this study, we use the Aspen Plus simulator (V12.1) to compare the theoretical differences and the regression results of the two-liquid activity coefficient models ELECNRTL and ENRTL-RK in terms of CO2 solubility. The quaternary system CO2-MDEA-PZ-H2O vapor-liquid equilibrium (VLE) data found in the open literature are first collected. After excluding anomalous points according to some criteria, 521 confirmed data ranging from 40 °C to 120 °C are then used to regress the binary interaction parameters of molecule-molecule and molecule-electrolyte pairs for both models using the maximum likelihood method. The fitting results show an accuracy enhancement of 34 % for ELECNRTL and 68 % for ENRTL-RK compared to the Aspen default models. The regressed models can be used to improve the accuracy of absorber and stripper simulations in CO2 capture processes.