in previous posts (in this forum) I mentioned my preference for PR-NRTL-HV or SRK-NRTL-HV models,
the modified versions available in Prode Properties do a decent job with hydrocarbons plus water, chemicals, hydrates etc.
For water (or other compounds which have the capability of forming hydrogen bounds) CPA may be more accurate but slower (compared to PR-NRTL-HV or similar models), CPA is a combination of a EOS (SRK or PR) with the association term of Wertheim, used in SAFT,
CPA requires a separate solver (with a few derivatives solved numerically) , as result some procedures (for example critical points) may not work reliably in some cases.
SAFT (with association) is even more complex (it requires to solve volume plus association), of course there are specific fields of application for these models,
but, for hydrocarbons, where possible, I would prefer the modified PR or SRK models.
Edited by PaoloPemi, 07 October 2016 - 08:25 AM.