No replies to this topic

[Jesper]

I am trying to model absorption of CO2 in aqueous ammonia using a RadFrac absorber as part of my master thesis. I used the electrolyte wizard to set up the system and reactions using H3O+ with a true component approach with possible precipitation of salts.

However, my system becomes very unstable with regard to changes in mass flow. The feed stream is 92 kg/s of a mixture of N2, O2, H2O and CO2 and the makeup stream is aqueous ammonia with 80/20 water/ammonia with mass flow in the range of 20-80 kg/s. The system is very hard to get to converge and will only do so for some specific makeup stream mass flows, i.e. 28 kg/s works fine but 29 kg/s does'nt converge.

Convergence improved some with addition of two internal pumparounds with cooling to avoid stage dry up, but it still only works with some very specific numbers (numbers and spans vary with ammonia concentration in the makeup stream). In addition, convergence at flows of about 40 kg/s and above disappeared alltogether with this addition.

It is very common that the err/tol value is 0 or very near 0 (in the order of 1E-9 or less), but the solution never converges anyhow, any ideas as to why that could be and how to get it to converge in those cases?

I am new to both chemistry and to Aspen and would very much appreciate any clues on how to solve this! :)
The absorber will be part of a larger CO2-capture system so the stability really needs to improve in order for that to work..

You can download the .apw file here: Link

Thanks!
// Jesper

Edited by Jesper, 22 February 2011 - 09:13 AM.