Phase Envelope (vapor-liquid-liquid) with Prode 3

[mlazard]

I am comparing phase diagrams generated by different models (Peng-Robinson, Soave, Lee Kesler Ploecker, GERG 2008 etc.) for mixtures with hydrocarbons + water,
I would like to use GERG 2008 as reference but it seems that this model doesn't work with light hydrocarbons + water fractions > 0.01 while Peng Robinson (Extended) and models with complex mixing rules are able to generate both three phase vapor-liquid-liquid line and water incipient condensation line (see attached file generated for mixture C1 0.78 C2, C3 ... C8 + CO2, N2, water 0.03 with PRX model) , Lee Kesler Ploecker (with multiphase option) generates a similar diagram.
I know that water generates a solid phase below 0 Celsius but I have not the vapor-liquid-solid option for phase diagrams,
which would be my best option ?

 

[apetri]

you can upgrade to include vapor-liquid-solid phase diagrams however for solid phases the models are derived from Peng-Robinson-Extended, CPA, Regular, Wax and hydrates, I do not think there is a model derived from GERG 2008 for solid phase...
I use mainly GERG 2008 for Vapor-liquid (also including water) but I know there are limits and I doubt it is very accurate with high water fractions,
in the GERG 2004 monograph there is an example of phase envelope of a natural gas plus water, but water is not included in the mixture,
if you wish to adopt GERG maybe a similar procedure (calculate water separately) can produce comparable results.

 

[mlazard]

thanks apetri,
many streams in my projects have hydrocarbons, CO2, N2 and water,
I fear that solving water separately (as pure fluid) I will lose many information about multiphase equilibria.

 

[PaoloPemi]

mlazard,
light hydrocarbons condense at very low temperatures and it is not possible to calculate fugacities for a separate water phase with GERG at those temperatures, then,
instead of the vapor-liquid-liquid (predicted by Peng-Robinson-Extended or similar alpha functions), you may obtain a vapor-liquid or no points,
the obvious solution is a vapor-liquid-solid phase diagram,
to generate a vapor-liquid-solid phase diagram you can adopt the different models available in Prode,
hoping this helps,
Paolo

 

[mlazard]

Paolo,
thanks, I understand the point, with complex models such as GERG 2008 it is not possible to generate "virtual" values for water fugacity at low temperatures (below freezing point), while the alpha function in Peng Robinson allows such extrapolations,
that makes sense and probably the vapor-liquid-solid phase diagram is the only possible option.

 

[rotw]

I fear that solving water separately (as pure fluid) I will lose many information about multiphase equilibria.

What information exactly?
If water is the first compound to condense then solving for water within mixture model or treating it as separate compound should lead to the same, if am correct.

Waiting to be corrected otherwise.

 

[mlazard]

rotw, thanks for the contribute,
I think your objection is correct,
if one knows (a priori knowledge about the process) that all liquid water forms a separate phase probably the results are not too different,
however I see that some models predict some miscibility for heavy hydrocarbons or other components in the mixture, keeping water separated one loses this information.

 

[apetri]

for a mixture of hydrocarbons plus water, simplified models assume that all the condensed water forms a separate (pure) phase, a simplification which, in some cases, can introduce errors,
another case if you have alcohools (methanol, ethanol.. etc.) or other miscible fluids, even if water has the highest vapor pressure (first to condense) you need to include all fluids for the correct phase equilibria.

 

[apetri]

sorry, even if water has the highest vapor pressure (first to condense)
should be lowest vapor pressure ...