110th Anniversary: A Case Study on Developing Accurate and Reliable Excess Gibbs Energy Correlations for Industrial Application

Abstract

Excess Gibbs energy (G E) or activity-coefficient models have a strong niche in applied thermodynamics since they provide an effective method to correlate and estimate phase equilibrium for nonideal systems. The history of their development is well represented by many papers in the technical literature, and notably in the five score and ten years of publications in Industrial & Engineering Chemistry. The science and technology has moved from formulating and establishing ideas and concepts to the automation of modern software, where phase-equilibrium correlations are often developed by algorithms with little or no expert human intervention. In addition, there is belief among practitioners that increasingly theoretical models, based upon molecular thermodynamics, will lead to more accurate correlations and with wider applicability. In this paper we use an apparently complex, but actually simple system (the lower alcohols), to study the development of G E correlations for industrial application. The subjects focused on are (1) use of multiproperty data, (2) property trends within the family of compounds, and (3) the need for advanced models.

Simplified Approach to Understanding, Evaluating, and Parameterizing the NRTL Model for the Description of Binary VLE: Tττ-VLE Approach

“…Often used forms of temperature dependence include only a ij (isothermal BIPs), only b ij , or a combination of both a ij and b ij . As such, the form of inverse linear temperature dependence was used throughout the investigations. …”

“…Often used forms of temperature dependence include only a ij (isothermal BIPs), only b ij , or a combination of both a ij and b ij . As such, the form of inverse linear temperature dependence was used throughout the investigations. ln ( γ 1 ) = x 2 2 ( τ 21 G 21 2 ( x 1 + x 2 G 21 ) 2 + τ 12 G 12 ( x 2 + x 1 G 12 ) 2 ) ln ( γ 2 ) = x 1 2 ( τ 12 G 12 2 ( x 2 + x 1 G 12 ) 2 + τ 21…”

“…This insight may significantly simplify more complex correlation procedures in future studies. For example, the simultaneous correlation of both binary VLE data and excess enthalpy data is gaining interest, , where the latter is more sensitive to the temperature dependence incorporated (the TDPs). The Tττ-approach explicitly demonstrates the model’s flexibility and sensitivity to TDP choices.…”

Applications of an Association Activity Coefficient Model, NRTL-PA, to Alcohol-Containing Mixtures

“…The PC-SAFT association term is widely used and provides composition dependence analogous to ESD and CPA . Representation of experimental phase equilibria must be precise for accurate process design as highlighted by Mathias, and the excess enthalpy should be simultaneously regressed for accurate modeling …”

“…6 Representation of experimental phase equilibria must be precise for accurate process design as highlighted by Mathias, and the excess enthalpy should be simultaneously regressed for accurate modeling. 30 The paper is organized as follows. Section 2 provides the key model equations and parameters.…”