A New PC-SAFT Model for Pure Water, Water–Hydrocarbons, and Water–Oxygenates Systems and Subsequent Modeling of VLE, VLLE, and LLE

Abstract

Accurate analytic thermodynamic modeling of water and its mixtures with hydrocarbon and oxygenates is difficult even with new and advanced equations of state such as the perturbed-chain statistical associating fluid theory (PC-SAFT). Several attempts have been made in the past by various authors to solve this issue. However, current models generally fail to describe simultaneously and accurately pure water properties (especially its liquid density) and liquid–liquid equilibria for mixtures involving water, hydrocarbons, and oxygenates. In the current work, this problem is dealt with by modification in the fundamental structure of the model. It was established that the temperature dependent diameter d(T) does not behave in the same way for water as it inscribed in the original model. Hence, a modification was proposed for d(T) of water in order to correctly represent the phase behavior of pure water and its mixtures with hydrocarbons and oxygenates. The deviations in saturated liquid densities and vapor pressure for pure water were reduced to 0.6% and 2.2%, respectively, in a large temperature range. The results for liquid–liquid equilibrium (LLE), vapor–liquid equilibrium (VLE) and vapor–liquid–liquid equilibrium (VLLE) of various water–hydrocarbons and oxygenates show the accuracy of this new model and its predictive capability when coupled with a group contribution approach. For certain oxygenated mixtures such as water with aldehydes, ketones, ethers, and esters a new contribution to the Helmholtz energy, known as the “non-additive hard sphere” contribution, was used. The cross-interaction parameters obtained for mixtures were validated qualitatively by calculating octanol/water partition coefficients and the Gibbs free energy of hydrogen bonding (ΔG HB). Results are found in good agreement with experimental data.

Predicting phase behavior in aqueous systems without fitting binary parameters II: Gases and non‐aromatic hydrocarbons

“…Later on, this approach, with one temperature‐dependent and another temperature‐independent binary adjustable parameter, has been implemented for estimating the data in water solutions of nitrogen and hydrogen. Some additional references presenting the modeling of phase behavior in aqueous systems should be acknowledged as well. A particular importance for this research present the studies of Economou and Donohue and Economou and Tsonopoulos .…”

Modeling of mixed-solvent electrolyte systems

“…Moreover, improvements in the current work is significant in correlating densities at low salinities and infinite dilution condition which is essential for an improved mixed solvent electrolyte model [25,68]. It can be seen that the densities of alkali halides conform fairly well at low salinities which is the capability of the model developed in our previous work [131]. The parameters for the ions that are finally used in the model are listed in table 6 (set 1).…”

“…Since we have seen that the model is able to describe the salting-out phenomena in alkane or acid gas + brine systems, it can now be used to study vapor-liquid equilibrium of various mixed solvent salt systems. Pure water and several pure solvents such as methanol/ethanol/1-propanol have been already parameterized in a previous work [131] and parameters are recalled in table 3. Binary mixture of methanol-water were parameterized using cross interaction parameters (‰ o , ‹ o and ' o ), here three parameter were chosen to have better accuracy in the description of binary phase equilbria (without fitting ' o , deviations were close to 10% in AADP and AADY, compared to 1.8% when including this additional correction).…”

On Molecular-Based Equations of State: Perturbation Theories, Simple Models, and SAFT Modeling

“…The molecular size and energy parameters of the listed versions vary within the range 1.91-3.59[Å] for σ and 839-2,932[K] for ǫ/k B pointing again to the fact that these parameters may hardly have anything common with physical reality and are pure numbers. As an interesting example we may mention conclusions of the very recent paper by Ahmed et al [82]. They report excellent results for the VLE of water and other compounds with one interesting feature: to obtain these results, the size of the monomer unit has to swell with increasing temperature, which contradicts both common sense and the observed (and obtained by theory) behavior; with increasing temperature, the molecules attain kinetic energy and get closer to each other which means that, effectively, their excluded volume shrinks.…”

“…In their paper [33], Muller and Gubbins conclude that these interactions are more important than it was previously thought and this would also fully agree with the results shown in Table 1. An extension of Equation (33) by including explicitly the dipoledipole contribution was considered by Karakatsani et al [81] to deal with strongly dipolar fluids and recently also by Ahem et al [82]. An extension along the same path was made by Liu et al [83] who, in addition to incorporating the dipole-dipole interaction, employed the hard-core Yukawa for the monomer-monomer interaction instead of the LJ.…”