Which type EOS model is fit for Toluene - Water -inert gas at vacuum

[silebilee]

Dear lady and gentleman

Could you please help me to tell which type of EOS model is best fit for toluene - water system?

This is toluene – water will form azeotrope so I will plot the x-y diagram to determine whether the simulator is calculating the azeotrope.
(Note this system consist of 80% O2 and N2 at P=0.103kg/cm2 so I am not sure should I consider to generating ternary bimodal plots or residue curve map.)
Anyway, I will treat it as binary system. (Please see
1.SRK ,2 Wlson 3. SRK-modified model x-y drawing in attached drawing)4.Grayson-streed


Thank you so much.

 

[PaoloPemi]

At low pressures an activity based model (I suggest NRTL or UNIQUAC) should do the work, with a pressure above 3-5 Bar with my software (Prode Properties, see prode.com) I prefer to set Peng Robinson (extended) UNIQUAC or Peng-Robinson-NRTL as mixing rules I set Wong Sandler or Huron Vidal.
A good alternative should be CPA (Cubic Plus Association) Prode has two variants (SRK and PR) but I have not tested this model with water-toluene.

 

[silebilee]

Dear Sir


Thank you very much for your suggestion. It seems that Wilson is best model for toluene-water system. Because only Wilson model can get the pinch point (x=y at the azeotrope )in the x-y diagram (please see my last post attachement), I conclude that the Wilson is best model in toluene-water system
.
But, Wilson model seems only good for vapor-liquid calculation not good for vapor-liquid-liquid calculation in the vapor-liquid-liquid equilibrium stae.

Though, it seems to me that Wilson model in this vapor-liquid-liquid equilibrium case is the same like others, I still don't know how to use aspen or pro II to generate the tertiary plot to prove this model really work.

Could you please give me more suggestion?

Thank you very much, and appreciate for your generous help.

 

[PaoloPemi]

Wilson doesn't work for liquid-liquid equilibria, you should select NRTL , UNIQUAC or equivalent activity based model.
For Water-Toluene NRTL or UNIQUAC should be more accurate than Wilson (these models include additional data), with these models you must verify that the software includes the correct BIPs (three values for NRTL and two for UNIQUAC, more with temperature dependent parameters)
Generally simulators include a large number of BIPS, for example my copy of Prode Properties has about 30000 BIPS including those for Water-Toluene for the different models.
To test a possible Vapor-Liquid-Liquid condition you can simply solve a multiphase isothermal flash operation, you define a operating temperature and pressure and the software finds the number of phases at equilibria and compositions.
I am not familiar with the software you mentioned but I suppose you can solve a multiphase flash, with Prode you can solve a vapor-liquid, liquid-liquid, vapor-liquid-solid (up to 5 different phases at equilibria in my version), an alternative could be a phase diagram.