2 replies to this topic

[mrbabu]

I am calculating freezing point depression of binary systems as water-methanol, water-ethanol etc. in Prode Properties
the procedure gives for water-methanol errors of 3-8 K against experimental data at different compositions with Wilson model, with NRTL errors are about the same, I would guess that a selection of different BIPs could improve the accuracy of these predictions, how can I regress from a experimental values (i.e. I have a few freezing point values as function of volume concentration of solvent in water and I wish to use a model to interpolate the whole range) ?
Thank you for assistance.

[PaoloPemi]

with the solid model in Properties you need a reliable method to calculate liquid fugacities at operating conditions, Wilson (or NRTL) do not work reliably for water at temperatures well below 0.0 C, you may try different models as CPA or SAFT but I am not sure about the results for this specific mixture.
As alternative you may model solid-liquid equilibria based on solid saturation pressure correlation which is available in Prode Properties databank.

[mrbabu]

thanks,
I have not SAFT installed in my version, however I was able to correct values and results are acceptable (for my purposes).