Hydrocarbon Vapour/Liquid Equilibrium - Impact of Water

[ChemEngSquirrel]

I’m using a process simulator to estimate the volume of hydrocarbon gas produced per barrel of oil produced from a two phase separator. I thought the calculated Gas Oil Ratio (GOR) would be the same whether the feed stream is water saturated or not, but this does not appear to be correct. To explain:

I carried out a flash on a two phase hydrocarbon stream and calculated GOR to be 193 scf/bbl. Repeating this calculation with the same stream, now saturated with water, the dry GOR increases to 206 scf/bbl (I remove the water from this stream after carrying the flash).

All my research into vapour liquid equilibrium shows that water is not taken into account in the flash calculation i.e. phase envelopes are dry basis – so why does this simulation programme suggest that a different separation occurs if water saturated.

 

[zdas04]

If you have a question about why a particular simulator gives you unpredicted results you either need to share the name of the software (and the version) or you need to talk to the people that support that software.

Whether water vapor is taken into accout in a program is a program-design decision, not an industry decision. It sounds like your program doesn't follow the "convention" that you are expecting. It happens. A lot of the conventions in process simulators came from the days of the mainframe when memory was very expensive, and computing cycles were even more expensive. Any shortcut that we could take back then and still provide a "reasonable" match to actual data was ok. Reasonable was pretty loosly defined back then. As the computing power has increased and gotten less expensive, "reasonable" has gotten progressively tighter. Maybe a modern version of your program decided to eliminate the "dry" assumption. There isn't a general answer.

David

 

[ChemEngSquirrel]

Simulation software is Hysys. It is a very well validated piece of software & it is likely that the software is correct. I think the weakness is my understanding of the basic principle.

I want to find out why, hence my original post.

 

[zdas04]

The first time I ran HySys was on the CICS operating system on a mainframe with Hollerith Cards. I'll bet anything you want to bet that if I ran the same deck into Aspen's incarnation of it I would get different answers.

No simulator is ever "correct", but many of them are "right enough". Every version of a program processes the data in a different way (otherwise why would you pay for the upgrade) which is why I asked you in my original post "what software (and the version)" are you using.

David

 

[PaoloPemi]

the flash (multiphase) itself could converge but the results depend from the selcted model, for hydrocarbons-water mixtures see

http://www.eng-tips.com/viewthread.cfm?qid=263493&page=1

hoping this helps a bit

 

[mbeychok]

I seem to remember that in the distillation of light hydrocarbon-water mixtures, water behaves as if it were ethane.

That might give you a clue as to why you are getting different answers.

Milton Beychok
(Visit me at www.air-dispersion.com)
.

 

[Jason5000]

My gut feel here is that your simulation results are "right". What you are likely seeing the impact of the presence of water vapor on the partial pressure of the hydrocarbon components in the vapor phase. Diluting the light hydrocarbons with another vapor, water in this case, causes more of the light ends to flash. Much like the use of stripping steam in an oil fractionator.