correct thermo method for simulation

[saidsaz]

Dear all ,

wonder if anybody out there into simulation who can advice me on correct thermo method to use for a mixutre of components including H2 , N2 , methane , ethylene , butene, isopentane at 30 kg/cm2 g and 95 c fed to S&T exchanger .

the exchanger is cool enough to partially condense this mixure.

i have used one old version of proII , and set up a rigorous exhcanger module . i have tried a number of thermodynamic methods SRK ,t BK10 , GS ...ETC .

the liquid composition in the exchanger outlet oultet keep showing liquid H2 and ETHYLENE , i expect to see only pentane as liquid .

i have doubt the problem is in selecting a correct themo .

i would appreciate very much somebody 's advice .

thnx in advance .

 

[rbcoulter]

I assume that you are cooling at ambient conditions, say about 25 C. In that case, ignore the hydrogen and nitrogen because they are not going to condense.
The other components look similar enough to assume activity coefficients equal to one, at least for a first
pass calculation (could use UNIFAC to check it).
In that case Raoult's law can be used to calculate
the K (partitioning coefficients) values.

 

[grt12]

You could try PR

 

[joerd]

The thermo to use in Pro/II for such mixtures is SRK in principle. SRK is their 'workhorse', so it has been modified to suit a wider range of applications than the classic SRK equation.

You should probably find a very low amount of H2/N2 in your liquid, both for numerical and thermodynamic reasons. In order to correctly predict the heat transfer coefficients when rating/designing the exchanger, you do need to have them in your simulation!

Try first a simple HX model, with fixed outlet temperature, and see what comes out. Then, if you want, use the rigorous method to figure out the performance of the exchanger, or contact a thermal rating engineer who can use HTRI or HTFS to design/rate the exchanger. Regards,

Joerd