Thermodynamic modeling of saturated liquid compositions and densities for asymmetric binary systems composed of carbon dioxide, alkanes and alkanols

• Phase behavior of the binary systems containing largely different components.
• Equation of state modeling of binary polar and non-polar systems by utilizing different mixing rules.
• Three different mixing rules (one-parameter, two-parameters and Wong–Sandler) coupled with Peng–Robinson equation of state.
• Two-parameter mixing rule shows promoting results compared to one-parameter mixing rule.
• Wong–Sandler mixing rule is unable to predict saturated liquid densities with sufficient accuracy.

The present study mainly focuses on the phase behavior modeling of asymmetric binary mixtures. Capability of different mixing rules and volume shift in the prediction of solubility and saturated liquid density has been investigated. Different binary systems of (alkane + alkanol), (alkane + alkane), (carbon dioxide + alkanol), and (carbon dioxide + alkane) are considered. The composition and the density of saturated liquid phase at equilibrium condition are the properties of interest. Considering composition and saturated liquid density of different binary systems, three main objectives are investigated. First, three different mixing rules (one-parameter, two parameters and Wong–Sandler) coupled with Peng–Robinson equation of state were used to predict the equilibrium properties. The Wong–Sandler mixing rule was utilized with the non-random two-liquid (NRTL) model. Binary interaction coefficients and NRTL model parameters were optimized using the Levenberg–Marquardt algorithm. Second, to improve the density prediction, the volume translation technique was applied. Finally, Two different approaches were considered to tune the equation of state; regression of experimental equilibrium compositions and densities separately and spontaneously. The modeling results show that there is no superior mixing rule which can predict the equilibrium properties for different systems. Two-parameter and Wong–Sandler mixing rule show promoting results compared to one-parameter mixing rule. Wong–Sandler mixing rule in spite of its improvement in the prediction of saturated liquid compositions is unable to predict the liquid densities with sufficient accuracy.