Analyzing the experimental data of CO2 equilibrium absorption in the aqueous solution of DEA plus MDEA with Random Forest and Leverage method

The continuing production of energy from fossil fuels is responsible for large emissions of CO2 component of greenhouse gases. Employing amine-based solutions is a common approach for removing the produced CO2 in numerous carbon capture systems. In this communication, a novel methodology namely Random Forest (RF) is employed for developing a tree-based predictive tool. The prediction capability of the proposed RF model is compared to the modified Deshmukh-Mather thermodynamic model. The presented RF model shows an average absolute relative deviation percent (AARD%) of 3.74, while the modified Deshmukh-Mather model estimates the CO2 loading capacity of the DEA + MDEA solution with AARD% of 12.10. Furthermore, the reliability and quality of the experimental data for CO2 solubility in DEA+MDEA aqueous solution has been investigated in this study using Leverage algorithm. According to the results, there are two probable doubtful data points in the investigated database. (C) 2017 Elsevier Ltd. All rights reserved.