10 replies to this topic

[aamerbebo]

Hi all,
I am about to do my master`s defence in CO2 removal process using a mixture of amines.
I used Unisim R400 to model the process and I also used DBR amine package to study the effect of Piperazine into the removal process.
A part of my thesis care of the effect of amine circulation rate Vs. CO2 in treated gas while making alllll other parameters constant. I expected for CO2 in treated gas to decrease while increasing the circulation rate till reaching the equilibrium. but I found the following


Uploaded with ImageShack.us


I monitored the results for 4 different temperatures for the lean solution. Every one of them have a deviation as the CO2 in treated gas increases suddenly and then back to decrease as its normal performance. I do not have a good interpretation for this deviation.
I just suspect of transition from laminar to turbelent flow which decrease absorption efficiency then increase it especially with a mass transfer limited reaction such as this one....

Kindly help

[aamerbebo]

By the way, Unisim is the same like Hysys
Heeeeeeeeelp

[henryleung]

To verify a transient flow effect, you can calculate the Re to see if it is the the transition region.

I notice that you y-axis unit is really small. Could this cyclic behavior due to the numerical calculation in the simulation?
Looks like the rise and fall is between 0.006% and 0.013%. That is 0.00006 and 0.00013. What is the tolerance of the simulation? I am not an expert in Amine CO2 removal, so I do not know the physics or chemistry that may give rise to such an effect. But whenever I see something strange in simulation, I would check whether I am hitting the limit of the numerical calculation or check am I near a critical point in the calculation.

[aamerbebo]

henryleung,
Thank you for ur reply. I do not know how could I calculate Re in columns specially trayed ones. Could you help?
The tolerance is from 0.6% to 1.3% It's my fault.

[sanaullah]

As for as i am concerned with your question, i can justify those things as,
1. Absorption is temp dependent operation...low temp ... good absorption ... plz observe with 65 deg temp curve
2. while discussing on curvature of the plot given, i shall say that pepraize is added to enhance the absorption but increasing the flowrate of solution shows inverse effects on treated gas, which is we are looking at the curve.
Regards

[RockDock]

There is a lot more to it than the circulation rate. The solvent selection, residence time of liquid/vapor contact in the column, temperature pressure and the regeneration of the amine are far more important.

Also, I would recommend using a rate-based simulator, such as ProMax, to do the modeling. I hear they give free licenses to Universities, too. ProMax actually takes into account the reaction kinetics involved in Amine Sweetening. The kinetics are especially important when dealing with CO2 predictions, since CO2 is much more tricky to predict.

As for your question about the peaks of the CO2 content in the sales gas, we really need more information, such as T, P, composition, column internals, reboiler duty, etc..

[aamerbebo]

sanaullah,
Increasing flow rate decreases CO2 in the treated gas which means increasing absorption efficiency which is logic...
and I still have the same problem which is the peak

[aamerbebo]

RockDock,
Thnak you for your reply, I just need to clarify that I am using DBR amine package which is designed to treat with Piperazine and works well with the reaction kinetics.
I think the data you required will not help as allll other parameters are constant and no changes in composition, T, P or even enternals.

[RockDock]

aamerbebo,

I highly doubt any reaction kinetics are taken into account with that model. Most likely, what is modeled are correlations with constant value adjustments being applied. Most amine unit designs I see are done using ProMax for that reason. But, use what works for you.

While the T, P and internals are constant, is your reboiler duty? That should not be.

The reason those other constants are important is because CO2 is a slow reaction. The temperature affects the rate of reaction. All these reactions are exothermic, the the column will heat up as the solvent flows down the column. As the solvent rate increases, you are sending more fluid in the column, at a higher temperature than the inlet gas. That alone could affect the temperature profile throughout the column. Then there are the competing reactions. With more solvent, comes the potential for more reactions, also causing a T increase.

The pressure affects the reactions of H2S and CO2, also. As you increase the flow rate, thus changing the temperature profile and composition (perhaps even phase) per stage, a low pressure system may favor the H2S reaction for a particular temperature range, while a high pressure system may allow CO2 to compete in that same pressure range.

The internals are important, because you may be increasing contact time on the trays, thus having more reaction, or you may be increasing the flow rate past your reaction approach threshold, essentially not helping much. Any study on this needs to report the reaction approach in the column.

Can you post a .pdf of the report from your model with stream T, P, Flow and Compostition?

[Dacs]

IIRC CO2 absorption by amines is largely affected by kinetics than equilibria.

Try to do some sensitivity analysis by modifying your tray parameters (that will affect the amine residence time in particular).

[Erwin APRIANDI]

Just to share what we praticioner know,

Why ProMax is still a better tools to do an absorption analysis is because the ProMax take into account the hydraulics pasrt of the absorption in column, HYSYS or Unisim do improved in term of calculating the reaction kinetics, but they still have a minor issue in the hydraulics part of the column. However for a research I believe this minor issue can be quite major, since you do analyse up to the tiniest details.

This hydraulics part include the flow pattern of the liquid in the column.

Please correct me if I'm wrong, since I do this analysis in FEED stage project which does not required tight margin.