1 reply to this topic

[adw07]

Hi,

Looking for a little advice for the application I am developing. Without going in to too much detail, it is necessary to model 3-phase (oil/gas/water) separators and I am unsure whether I should consider the aqueous phase completely separately when doing equation of state (P-T flash) calculations.

Would using Peng-Robinson EOS be terribly bad for water (in which case I could use a simple correlation instead) or just a little inaccurate?

Not sure if it would cause convergence problems or similar, as most of the literature I've seen seems to disregard the water.

Any help appreciated.

[Chellani]

I am not sure what you are developing but here are few inputs
1. Your flash should figure that whether third phase (aqueous) exist or not. Peng Robinson's (PR) performs a three phase flash and estimates existence of third phase based on water's solubility in aqueous phase (using API correlation and few more corrections).
2. Yes PR is terrible for pure water. Try estimateing boiling point of water using PR at different pressures and cross-check with steam table. If you are sure that your system will have only water, using steam table (or property packages available in HYSYS based on steam table) would be good idea.