Efficient estimation of CO2 solubility in aqueous salt solutions

Industrialization and modernization have led to an increasing trend of greenhouse gas emissions entering the atmosphere. Consequently, many countries are imposing strict environmental regulations due to increasing concern on the climate change, and a significant amount of research is being conducted to tackle this problem. Carbon capture and sequestration are one of the major methods, which can remarkably contribute to this issue since a major greenhouse gas in Earth's atmosphere is carbon dioxide (CO2). The general objective of this work was to investigate the significance of potential use of aqueous salt solutions for capturing CO2 by developing an accurate model for estimation of this phenomenon. In particular, 608 experimental data of CO2 solubility in aqueous solutions of salts as a function of temperature, pressure, salt solution concentration (molality), hydrogen bond acceptor count of salts, molecular weight of salts, and complexity value of the salts were collected from the literature. Four state-of-the-art computational methods (LSSVM, ANN, PSO-ANFIS and GA-ANFIS) were employed to examine their predictive capability for the problem. The results reveal that the LSSVM offers the highest accuracy with R-2 = 0.9927 and %AARD = 537. Although the neural network and fuzzy interface systems are more complex approaches, the superiority of the LSSVM method for the problem under study demonstrates the easy nature of this phenomenon. Thus, the LSSVM model is recommended for estimating the CO2 solubility in aqueous salt solutions because not only it has more accuracy, but also it is easy to apply, has less mathematically complicated and less time-consuming. (C) 2019 Published by Elsevier B.V.