Enrtl Fluid Package For CO2 Removal In Aspen Plus 2

[processendbd]

I am a little perplexed seeing how my column is getting/not getting converged.



I am working with CO2 separation by K2CO3. Aspentech has one reference example file. But when I try to increase BOTH (the lean solvent flowrates and K2CO3 concentration) to serve my purpose (different CO2 removal requirements, higher CO2 concentration, and flue gas flow rates). It fails to converge most of the time. On the other hand when if change only one of them (for example either Leans solvent flow rates or leans solvent K2CO3 concentration. It converges most of the time, Note: I have used wide reneges of column internals for al the cases still results are the same.



Do I need to do some hand electrolyte balance to find lean solvent flowrates? As the reactions are in equilibrium I am a little hesitant about how should I go about it. Will the solvent flowrates requirements controlled by rate-controlled reactions?



Note:

I tried few things

1. Advance convergence option

2. Tolerance

 

[pmoneysh]

"But when I try to increase BOTH (the lean solvent flowrates and K2CO3 concentration)"
What is the ratio of Flue gas CO2 flow rate and solvent flowrate?

 

[processendbd]

Thank you for your comment.

Original Aspentech reference file which is solved

CO2 Mole Flow rate 0.0494Kmol/hr
Mole Fractions of CO2 0.02334

TOTAL Solvent Mole Flow Rate: 3.89772550348923 Kmol/hr
K+ mole fraction 0.009148
HCO3- Mole fraction 0.0010929
CO3-- Mole fraction 0.003958



My current requirement is
CO2 Mole Flow rate 296.403 Kmol/hr
Mole Fractions of CO2 0.06260

Solvent Mole Flow Rate
Literature suggests me to use 30% K2Co3 concentration, and I need to determine the flow rates but any concentration other than the Aspentech referance file.

Appreciate it

 

[pmoneysh]

I did some projects regarding CO2 capture but used aspen hysys as suggested by the client (and using a different solvent). We don't need to change the chemistry as it is very difficult to obtain our own reaction kinetics.
Aspen plus has a very useful tool "design spec". Therefore, after fixing the solvent concentration, we can obtain the required solvent flowrate to meet the CO2 recovery rate (In my case ratio flowrate of solvent to feed CO2 was obtained at around 2-2.5). Further, the flowrate of water in the wash section also affects the convergence. Increase the make-up water flowrate (later you can decrease it to minimum once you obtain the converged solution).
Ah! and one more thing how did you deal with the recycle stream?
Also, initially do not use packing type as it increases the convergence time.

 

[processendbd]

Thank you so much for your suggestion.

Actually, we could not decide on the solvent concentration yet as we could not converge the column at our desired solvent concentration. So our simulation is yet to have recycle stream.

"packing type as it increases the convergence time" Thank you again we will try that.

 

[processendbd]

I just found If I don't do "reinitialization" the simulation cases are solving just fine is that okay, necessary?

 

[pmoneysh]

To me it is not okay..... usually such results are not reliable. What is the error msg?

 

[Latexman]

If it truly converged, then increase the number of iterations to the total of all runs and run it from the reinitialized position.

Good Luck,
Latexman

 

[processendbd]

@pmoneysh But As i was having a problem with ENRTL column

Error
" MATERIAL AND ENERGY BALANCES FAILED TO CONVERGE: CHECK COL-SPECS
OR SUPPLY BETTER TEMPERATURE AND COMPOSITION ESTIMATES."
"
THE FOLLOWING STAGES DRIED UP:
10
A LIMIT OF 0.10000E-04 * SUM OF FEEDS WAS IMPOSED ON THE FLOW RATES."


One friend suggested

"You need to slowly walk the solvent concentration to your new concentration. The model is really brittle for concentration, but less so for other operating conditions (T, P, flow).

Aspen can't model that in a rate-based Radfrac column. It has to do with the actual equations of eNRTL."

So I incrementally increased concentration 1% without reinitialization it could solve the column without any error.