Mixture thermodynamics

Pressure-temperature phase diagram for a petroleum fluid composition X0.
This area concerns the modeling of phase equilibria and mixing enthalpies in complex systems, particularly those involving petroleum fluids. Oil companies need to know, when operating a well, whether the fluid they are going to extract is oil or gas, which largely conditions the taxes they will have to pay, as well as the duration of the well's operation. Phase equilibrium curves provide access to this kind of information.

It is therefore crucial for these companies to have models for predicting phase equilibria, including light compounds (methane, CO2, N2, H2S...) and heavy hydrocarbons (alkanes, cycloalkanes, naphthenes, aromatics). These models need to be predictive, sufficiently accurate and enable rapid calculations. In addition, since the abandonment of oil platforms, as the seawater is no longer separated from the oil fluid on the platform, the exploitation of offshore oil deposits has raised a new problem: the formation of hydrocarbon hydrate "plugs" preventing the subsea routing of the water-oil mixture; to prevent this formation, an inhibitor, generally methanol or a glycol, is added to the mixture. The phase equilibria to be predicted are then much more complicated, since both liquid-vapor (VLE) and liquid-liquid (LLE) equilibria must be represented simultaneously.
The models developed by the team are based on the cubic Peng-Robinson equation of state, combined with the excess free enthalpy derived from the generalized NRTL (Non Random Two Liquids) equation, which is particularly well-suited to the representation of ELLs. These models retain the simplicity of a cubic equation, while being fully predictive since they require no parameter adjustment on experimental data. The use of a group contribution method based on very few parameters enables thermodynamic properties to be predicted from knowledge of the molecular structure of the mixture's constituents alone. Several models have been successively developed:
- The NRTL-PR model has made it possible, for the first time, to simultaneously and accurately predict not only the ELVs of petroleum mixtures or water and hydrocarbons under pressure, but also ELLs with very high demixing, in water-hydrocarbons and ethylene glycol-hydrocarbons systems.
This model has also provided a satisfactory representation of the solubility of solid polycyclic aromatic hydrocarbons in water and in supercritical fluids (CO2).
- The NRTL-PRA model, which is an extension of the NRTL-PR model for modeling mixtures containing "highly associated" compounds, such as methanol. 
- The NRTL-PRA-salts model for the prediction of hydrocarbon, gas and "highly associated" compound systems in the presence of "salts".

It should be noted that the NRTL-PR model is implemented in ProSim's Simulis Thermodynamics software.

http://www.prosim.net/fr/actus-le-modele-thermodynamique-nrtl-pr--6_5.php et 

and that the NRTL-PRA model will be very soon.
NRTL Models