Some Observations Regarding the Association Kernel of SAFT-VR-Mie. Is the Molecularly Inspired Contribution Always Necessary?

Abstract

This study compares one relatively simple analytical and two unequally sophisticated molecularly inspired associative contributions to the Helmholtz free energy of SAFT-VR-Mie equation of state in predicting the thermodynamic properties of pure water and the global phase behavior of water−hydrocarbon (alkanes, alkenes, and aromatics) systems in the entire thermodynamic phase-space. The classical van der Waals one-fluid mixing rules with zero values of the binary adjustable parameters are implemented. Although this entirely predictive implementation is particularly challenging, the considered approaches can be characterized by a satisfactorily good accuracy. However, despite the expectations, the most sophisticated molecular-inspired association kernel does not exhibit an overall advantage not only comparing with its less complicated version, but also with the substantially simpler analytical approach. Consequently, the current results indicate that the advanced molecular background does not necessarily guarantee an improvement of the overall robustness and reliability in estimating actual thermodynamic data.

Water–Hydrocarbon Phase Equilibria with SAFT-VR Mie Equation of State

“…The endpoint of the curve, which corresponds to the UCEP, is very accurately predicted, although for short-chain alkanes, it is slightly overestimated. In the work of Garrido et al, however, the UCEP seems to be consistently overestimated, regardless of the alkyl-chain length. A plot of the UCEP with CN shown in Figure reveals that the UCEP reaches a plateau at an n -alkane chain length of about 26 carbons.…”

“…The end point of the curve, which corresponds to the UCEP, is very accurately predicted, although for short chain alkanes it is slightly overestimated. In the work of Garrido et al 54 however, the UCEP seems to be consistently overestimated, regardless of the alkyl-chain length.…”

“…From those few publications reporting LLE of hydrocarbons and water with CPA ,, and PC-SAFT, − it is shown that the predictions presented in this work are more accurate to the aforementioned results. Similarly, SAFT-VR Mie predictions are also provided by Garrido et al The comparison between the models however is not entirely fair, as it has been shown that the pure component parameters of water have a significant effect in the ability of EoS to predict the LLE of water–hydrocarbon mixtures . Furthermore, the parameters yielding accurate density, vapor pressure, and hydrogen bond fraction simultaneously are the ones that perform poorly for phase equilibria .…”

“…addressed the mutual solubility by introducing different binary interaction parameters to model each phase, 52 an approach adopted by Waseem to describe hydrate formation between water and light hydrocarbons. 53 Garrido et al 54 discuss the importance of a molecularly inspired association kernel of SAFT-VR Mie, and conclude that sometimes less accurate analytical associative contributions, like the one based on a hard sphere reference fluid, are actually more accurate. The authors have based their results on the prediction of pure component properties, as well as on the global phase behavior of water with hydrocarbons.…”

“…CP-PC-SAFT on the other hand predicts the solubility minima in an entirely predictive way, but this is not generalized for all water–hydrocarbon mixtures. , SAFT-variable range (VR) Mie, a recent SAFT formulation that will be used in this work, has shown overall promising results for water–hydrocarbon mixtures; Dufal et al addressed the mutual solubility by introducing different binary interaction parameters to model each phase, an approach adopted by Waseem et al to describe hydrate formation between water and light hydrocarbons . Garrido et al discuss the importance of a molecularly inspired association kernel of SAFT-VR Mie and conclude that sometimes less rigorous analytical associative contributions, like the one based on a hard sphere reference fluid, are actually more accurate. The authors have based their results on the prediction of pure component properties, as well as on the global phase behavior of water with hydrocarbons.…”