Modeling alternatives for the fluid-solid equilibrium of binary asymmetric mixtures.

S. B. RODRIGUEZ-REARTES; G.O. PISONI; J. I. RAMELLO; M. CISMONDI; M. S. ZABALOY

UBERLANDIA

Congreso; VII Congresso Brasileiro de Termodinâmica Aplicada (CBTermo 2013); 2013

Universidade Federal de Uberlandia

An available modeling approach (Rodriguez-Reartes et al, 2009) for the solid-fluid (SF) equilibrium consists of representing the thermodynamic properties of the fluid phases by using an equation of state (EOS), such as the Peng-Robinson EOS (PR-EOS), and of describing the solid phase through an expression for the fugacity of the pure solid heavy component. The parameterization problem can be carried out in a sequential way: first, the EOS interaction parameters are fitted so as to reproduce available fluid-fluid equilibrium experimental data; next, a parameter of the pure solid fugacity expression is adjusted from available experimental solid-fluid equilibrium data. In previous works, we found some difficulties when modeling the SF equilibrium behavior. For example, for the system propane + n-eicosane, we observed that a single set of parameter values would not provide a satisfactory reproduction of the available experimental information (Rodriguez-Reartes et al, 2009). In such work, we used the PR-EOS coupled to classical quadratic mixing rules (QMRs). Besides, the slope of the Solid-Liquid-Liquid equilibrium line for carbon dioxide + n-eicosane did not seem to be properly predicted after imposing the exact reproduction of the known quadruple point temperature. Another example of the problems we found is the overprediction of the pressure of the second critical end point for the system methane + neicosane (Rodriguez-Reartes, 2010). In this work we explore a number of alternatives for improving the representation of the mentioned solid-fluid equilibria. Such alternatives are: (a) changing the choice of the fitted experimental fluid-fuid equilibrium data set, (b) the use of the PR-EOS and QMRs with temperature dependent interaction parameters, and, (c) the use of cubic mixing rules (CMRs), coupled to the RK-PR EOS. Items (b) and (c) correspond to a more flexible representation of the fluid phases. We present modeling results for the three mentioned binary systems.