3 replies to this topic

[Oluwole]

How to predict asphalthene precipitation with activity coefficient models.

[Technical Bard]

How to predict asphalthene precipitation with activity coefficient models.

This is a current area of much research, for which there is no commonly accepted method. There are numerous papers published on the subject, using cubic EOS, activity coefficients, SAFT models and polymer solution thermodynamics. Unfortunately, they all require significant tuning with empirical data because we don't understand asphaltenes the way we understand other, simpler species. Look for papers by Yarranton.

[luizmoura]

You should use the UNIFAC model. It is a group-contribution method and the only previous knowledge required is the molecular formulae. I suggest this excelent spreadsheet:

http://www.pvv.org/~randhol/xlunifac/

You should find the seroes of this function F(i)=gamma(I)*x(I)-gamma(2)/x(II) where I and II refer to both phases. This function is in fact a vector of functions for every component i. You can find zeroes with Solver in Microsoft Excel.

[PaoloPemi]

the asphaltene precipitation is difficult to predict cause the complex nature of asphaltenes, for example asphaltene molecules are polar, heavier than wax, they behave similar to polymer molecules etc...
you can try Flory Huggins which is good for polymer solutions but many other models based on SAFT etc. have been proposed (as Technical Bard said) , among these I have tested Chung's model (included in Prode) with a limited set of experimental data.
Basically you need to solve a solid liquid flash operation where the solid phase is modeled with your specific method.