El Meragawi, Sally (2015-08). Modeling of Thermodynamic Properties and Phase Equilibria of Multicomponent Systems Related to the Oil and Gas Industry using the PC-SAFT Equation of State. Master's Thesis.
Thesis
Equations of state (EoS) have proved to be a reliable tool in chemical engineering thermodynamics for modeling the physical properties of complex systems. Various types of EoS have been developed based on different theories. For various reasons, some have become more popular for use in industry and academia. Of the popular EoS, two were chosen for investigation in this thesis. The first one was the Perturbed Chain- Statistical Associating Fluid Theory (PC-SAFT), an equation derived based on statistical mechanics and the second was the Peng-Robinson (PR) EoS, a cubic EoS commonly used in industry. In this work, the prediction capabilities of these two EoS were compared for several properties. The analysis began with an evaluation of their use in the prediction of the saturation properties of pure components and derivative properties from ambient conditions to the supercritical range. The particular derivative properties studied include the isochoric and isobaric heat capacities, the speed of sound, and the isothermal compressibility. In general, it was concluded that PC-SAFT outperforms PR in all cases. Next, the same primary and derivative properties of several binary and a select ternary mixture were studied. To improve agreement with experimental data, a binary interaction parameter was introduced and fitted to binary mixture vapor - liquid equilibria (VLE) data. This procedure drastically improved the accuracy of the models compared to the case where no binary interaction parameter used for the case of VLE predictions. However, for the case of the derivative properties, the use of the binary interaction parameter to ensure a more accurate representation of the interactions between molecules had only a marginal effect on the prediction of these properties. Finally, phase equilibria of hydrates were studied. As EoS for fluids are not designed to predict the properties of solid phases, the van der Waals-Platteeuw model was incorporated to allow for the prediction of three-phase equilibrium conditions of various hydrate formers. Specifically, this work focused on the equilibrium of a water-rich liquid phase, a hydrate phase and a vapor phase rich in a hydrate former. In all cases, calculations of the solid hydrate phase properties are based on the Kihara potential. This potential requires three parameters to be defined; initial values for which were found through a review of the literature. The accuracy of the predictions of the three-phase equilibrium is highly dependent on the reliability of these parameters. Thus, one of the parameters, the so-called ? parameter, was fitted to hydrate equilibrium data and resulted in a significant improvement in the accuracy of predictions of both PC-SAFT and PR EoS. The new set of parameters was then used to predict the three-phase equilibrium of several binary, ternary and quaternary mixtures of hydrate forming agents. Several conclusions are drawn from this work, including the observation that the accuracy of the models is reduced when the number of components increases.
Equations of state (EoS) have proved to be a reliable tool in chemical engineering thermodynamics for modeling the physical properties of complex systems. Various types of EoS have been developed based on different theories. For various reasons, some have become more popular for use in industry and academia. Of the popular EoS, two were chosen for investigation in this thesis. The first one was the Perturbed Chain- Statistical Associating Fluid Theory (PC-SAFT), an equation derived based on statistical mechanics and the second was the Peng-Robinson (PR) EoS, a cubic EoS commonly used in industry.
In this work, the prediction capabilities of these two EoS were compared for several properties. The analysis began with an evaluation of their use in the prediction of the saturation properties of pure components and derivative properties from ambient conditions to the supercritical range. The particular derivative properties studied include the isochoric and isobaric heat capacities, the speed of sound, and the isothermal compressibility. In general, it was concluded that PC-SAFT outperforms PR in all cases. Next, the same primary and derivative properties of several binary and a select ternary mixture were studied. To improve agreement with experimental data, a binary interaction parameter was introduced and fitted to binary mixture vapor - liquid equilibria (VLE) data. This procedure drastically improved the accuracy of the models compared to the case where no binary interaction parameter used for the case of VLE predictions. However, for the case of the derivative properties, the use of the binary interaction parameter to ensure a more accurate representation of the interactions between molecules had only a marginal effect on the prediction of these properties.
Finally, phase equilibria of hydrates were studied. As EoS for fluids are not designed to predict the properties of solid phases, the van der Waals-Platteeuw model was incorporated to allow for the prediction of three-phase equilibrium conditions of various hydrate formers. Specifically, this work focused on the equilibrium of a water-rich liquid phase, a hydrate phase and a vapor phase rich in a hydrate former. In all cases, calculations of the solid hydrate phase properties are based on the Kihara potential. This potential requires three parameters to be defined; initial values for which were found through a review of the literature. The accuracy of the predictions of the three-phase equilibrium is highly dependent on the reliability of these parameters. Thus, one of the parameters, the so-called ? parameter, was fitted to hydrate equilibrium data and resulted in a significant improvement in the accuracy of predictions of both PC-SAFT and PR EoS. The new set of parameters was then used to predict the three-phase equilibrium of several binary, ternary and quaternary mixtures of hydrate forming agents. Several conclusions are drawn from this work, including the observation that the accuracy of the models is reduced when the number of components increases.
ETD Chair
Economou, Ioannis Senior Associate Dean for Academic Affairs and Graduate Studies, Texas A&M at Qatar