Thermodynamic Modeling of Composition Dependence of Relative Volatility

Separation tasks are omnipresent in process engineering in the oil & gas industry. Product specification, energy efficiency, economic optimization, regulation, process safety, asset utilization, process intensification, market expectation are criteria for making an optimum choice of different separation process options. The leading separation process is distillation due to the ease of separating vapor and liquid phases by gravity. The driving force for splitting a feed stream into two or more product streams of higher value is provided by component distribution between the phases following thermodynamic equilibrium.

Volatility of components in a mixture is a result of two effects: the vapor pressure of pure components and the impact of mixture composition. In systems with close boiling components the composition dependence becomes decisive for relative volatilities and can lead to the formation of azeotropes. Splitting of C8 aromatics components and fractionation of a-olefins are examples of processes where an accurate thermodynamic model is crucial for process simulation. The application of cubic equations of state combined with gE-mixing rules allowed to combine the advantages of local composition based solution models (for example NRTL) with those of equations of state (for example SRK).

Strategies for developing the best set of required pure component and binary interaction parameters will be shown. Parameter fitting to experimental vapor-liquid equilibrium data for a typical slate of at least 20 components is typically limited by missing data. Parameter generalization and correlation of data estimated by group contribution methods are options to complete the set of model parameters according to best available knowledge.