4 replies to this topic

[mrbabu]

When simulating the gas plant our client adds a so called "gas saturation unit" which is available in some simulators, as far as I understand it gives the possibility to saturate the gas with water, our practice with Prode Properties is to add a certain (little) amount of water, do a isothermal flash and then get the vapor phase as saturated, results are very close to those calculated with the "gas saturation unit", I believe this is a common practice but I would appreciate opinions and comments from experts..

[marchem]

the procedure is correct if you manage to keep the amount of condensate very low (by reducing the amount of water injected), with usual EOS condesnate will not be pure water but it will contain a certain amount of dissolved gas and that will modify the final composition in gas phase.
Altough it is a common practice to include a gas saturation unit you should investigate the validity of the selected EOS with water...
In Prode Properties depending from composition and operating conditions you can select ISO 18453 (Gerg 2004), GERG 2008 or a std. EOS with complex mixing rules.

[mrbabu]

Marchem,
in a test case with a natural gas mixture I see low solubilities of vapor fractions in water phase,
for CO2 about 0.00001 and lower for hydrocarbons as C3, C4 ...
Perhaps the quatity of condensate (as water) should not be so critical, do you agree ?

[marchem]

you can easily verify that, the different component's fractions in condensate should be very low with exception of water.

[mrbabu]

hi marchem,
we use GERG model (which is said to be accurate for water with hydrocarbons) while our client does utilize the "extended" PR model (which is that available in their software) , then there is a little difference, however the method gives equivalent results (when correcting for uncertainty due to different models)
.