1 reply to this topic

[kemisage]

Hey readers,

As the title might have hinted, I am trying to design a very simple sour gas sweetening column in Aspen Plus. It's not something for a class or an assignment or anything as such. I just wanted to try it out. So I started very simple, but somehow I can't seem to figure it out. So any help is appreciated. Sorry for the long post.

 

Feed: 95% methane, 5% hydrogen sulfide by mass.

Sweet gas composition desired: less than 0.3% H₂S.

Solvent: 30% MDEA

 

First, let me put my design decisions in points.

1. Since the absorption is better at higher pressures and lower temperatures, I chose the column pressure to be 25 bar (think around 350 psia).

2. The dew point of the sour feed is very low, so I chose the lean amine temperature to be 10 C (50 F).

3. There was a source where they defined a formula to approximate the lean amine flowrate for a given sour feed rate. Based on that, I chose the sour feed rate to be about 10 MMSCFD and the lean amine rate to be 45 gpm. Amine flowrate is too low?

4. There really doesn't seem to be a basis for estimating the number of stages. I chose 20 initially in an idea to perhaps tinker with it later. I also believe that these might be too many stages but I could be wrong.

I used the RadFrac column with no condenser or reboiler. I wanted ion concentrations, so I chose the Electrolyte NRTL model. For the reactions in the aqueous phase, I defined them using the electrolyte wizard in trial 1, used EMDEA package in trial 2, used KMDEA package in trial 3 and used KEMDEA package in trial 3. I could definitely see that the packages offered parameters for the MDEA-H₂S pair that weren't originally available. I also used the absorber option in the convergence method and also increased the iterations to 200 (although it proved unnecessary).

Now: Trial 1 gave pretty bad results (converges with the above values but doesn't converge when I make certain specific changes in the decision variables I mentioned above). Trials 2, 3 and 4 pretty much converge all the time and produce very (very) similar results. But none of them work in terms of absorption. Nearly all ​H₂Sgoes into the sweet gas stream.

I varied:

1. Column pressure: increased it by bits to about 70 bar (about 1000 psia) but no real improvement in absorption.

2. Lean amine flow: from 30 to about 170 GPM but no real improvement.

3. Equilibrium stages: varied between 5 and 25. Nothing except changes in temperatures of the exit streams.

Am I missing something here? This should be a fairly simply column design but it's making me pull my hairs out. Do you guys see something?

Thank you for reading all the way through and for any input you all have.

[Nikolay_]

Hello,

 

You have answered the questions yourself. There are no parameters for MDEA-H₂S in RadFrac Aspen Plus, so you won't get any desiable results.

 

Regards,

Nikolai

Edited by Nikolai T, 27 September 2018 - 01:50 AM.