Prediction of Equilibrium Interfacial Tension between CO₂ and Water Based on Mutual Solubility

Abstract

In this paper, an interfacial tension (IFT) correlation (i.e., Model #2) between CO2 and water is generalized as a function of their mutual solubility and the reduced pressure of CO2 in a temperature range of 278.2–469.2 K and a pressure range of 0.10–69.10 MPa. Two sets of correlation coefficients are respectively regressed for pressures lower and higher than the critical pressure of CO2 (i.e., 7.38 MPa). Such a newly developed correlation is calculated to yield an absolute average relative deviation (AARD) of 4.5%, a maximum absolute relative deviation (MARD) of 32.3%, and a maximum absolute deviation (MAD) of 1.65 mN/m, respectively. The newly developed model is found to greatly outperform the four existing correlations as a function of temperature, pressure, or CO2 solubility in water. A higher CO2 solubility in water (i.e., x CO 2 ) leads to a lower IFT for all pressures, while a higher water solubility in the CO2 phase (i.e., y W ) results in a lower IFT mainly at lower pressures. Both pressure and temperature exert effects on IFT mainly through regulating the x CO 2 and y W . Since the addition of inorganic salt can decrease the CO2 solubility in water, a higher salinity leads to a higher IFT. The developed model in this work can be a good estimate for the IFT between CO2 and brine if salinity is not too high.

Comparative Analysis of Four Neural Network Models on the Estimation of CO2–Brine Interfacial Tension

“…In recent years, Niroomand-Toomaj et al, Partovi et al, and Chen and Yang have adopted a multilayer perceptron model, least squares support vector machine, adaptive neurofuzzy inference system, radial basis function network optimized by particle swarm optimization method, adaptive neurofuzzy inference system trained by a hybrid method, and mutual solubility model; the models are established for temperature, pressure, salinity/molality concentration, and gas mole fraction, all of which have achieved high prediction accuracy.…”

A Supervised Learning Approach for Accurate Modeling of CO2–Brine Interfacial Tension with Application in Identifying the Optimum Sequestration Depth in Saline Aquifers

“…However, empirical correlations are demonstrated to exhibit relatively low accuracy and weak generality when used for estimating new data that are blind to model constructions. , The density gradient theory (DGT) based on phase thermodynamics is another alternative method for predicting CO 2 /brine IFT data, whose accuracy however remains questionable when applied for modeling systems with high brine salinities . The correlations based on the mutual solubility are demonstrated to result in noticeable underestimation of the CO 2 /brine IFT at high salinities or when bivalent cations are present in the brine phase.…”