HYSYS MDEA SIMULATION 1

[Mondolfi]

Hi,

I`m working on a MDEA Reg. Unit. and I'm using HYSYS 3.0.1 (amine package)but I'm having a little problem with this.
I have a surge drum before regenerator tower;According to H&MB, the drum is adiabatic and DP=0. Vapor flow in this separator should be near zero @T,P, but for some reason I'm venting almost 90% of the SH2 in this vapor , and should be near zero. Can someone tell me why is this happening?

Thanks in advance.

 

[PHP78]

hi
can you explain in detail? where is surge drum placed exactly and whats its inlet stream etc? also..whats the necessity of it ?

Prashant

 

[Mondolfi]

Hi Prashant,

The surge drums is upstream an amine regenerator tower. The purpose of this equipment is to vent any vapor to a fuel gas absorber,and send the rich amine solution to the tower.

The feed to this s. drum is a stream almost 60% water, 38% MDEA and 2% SH2 (mass basis). Flashed vapor should be near zero. In theory, thermodynamics model of amine solution in HYSYS are: Kent-Eisenberg and Li-Mather. The first one, and old correlation, wasn't prepared to work with MDEA, but apparently HYSYS modified it and now you can. Li-Mather is an electrolitic model that can work with MDEA also. No matter which model I use, I'm "venting" all the SH2 in this surge drum, before going into the tower. This is very wrong, because SH2 should be "dissolved" in the amine, and the model is not "seeing" this.

Do you know if HYSYS Amine Package is reliable?

 

[MortenA]

what is the vapour ration of the feed strem? Perhaps the feed is not seperated at an earlier stage in simulation due to an incosistancy between your model and the physical world? Is the stream heated or cooled prior to entering the surge drum?

Best regards

Morten

 

[Mondolfi]

MortenA,

Feed stream is not cooled nor heated before surge drum. Is in liquid phase du to SH2 is "retained" by MDEA. Surge drum exists in real plant.

Regards,

 

[PHP78]

Dear Mondolfi
i am also working on gas sweetening process and using amine package in HYSYS. I think it is quite reliable. The Kent and Eisenberg model is fitted model to the experiment data. Though in original paper Kent and Eisenberg has not fitted model for MDEA (they fitted it for MEA and DEA), still i think HYSYS has used same anology to MDEA. Also if you see basic books in this area by others like Astarita, there is theoretical basis for these models and then they are fitted to exptal data. So i can suggest you that you can rely on these models. Only problem i can see with this model is there may not be enough data for regenerator case where temp is quite high. I haven't seen much published data for MDEA or other amines for equilibrium at very high temp (very few data). So most of the times these models rely on extrapolation of the existing data.
For gas sweetening case, HYSYS has given some sample examples where they have validated the model for plant data. So lets rely on the model.

Now, this is quite strange that when you are not cooling or heating the stream and pressure drop is also zero, still you see that all H2S is vented off. Are you using flash in HYSYS for this? I will suggest that check which unit are you using and is it isothermal? If there is slight change in pressure or temperature, then equilibria changes and you can see some changes. I guess that when you are saying the upstream of the amine regenerator the it is a rich amine stream living the bottom of the tower.

Do following checks: this will help you
1. The stream you are sending to the surge stream: for this stream check the vapour fraction of the stream. If it is stream from bottom of the regenerator, then vapour fraction is zero. Also double check for its bubble point, it should be same as the stream temperature.

2. If it is vapour stream leaving the regenerator, then its vap fraction is one. and dew point temp is same as stream temp.

Still whole picture is not clear in my mind about the process you describing, but still i think that the tips i gave will be useful. Please write if you even solve the problem. This could be additional knowledge to me.

Regards
Prashant

 

[Mondolfi]

Dear Prashant,

Thanks for your interest. As you mentioned, HYSYS Amine Package is very temperature limited (25-125ºC) and for the bottom temp. of the regenerator (128.5ºC) is out of range.

The stream I'm talking about is the feed stream to the Amine Regenerator. It is a rich amine coming from the absorber tower with lean amine.

What I'm going to suggest you is to "see it yourself". The stream characteristics are as follows:

STREAM RICH AMINE
T=61.57ºC ; P=5 kg/cm2_g

Molar Flows (kmol/h)

Water ...... 6728.517
SH2 ........ 244.135
NH3 ........ 0.021
H2 ......... 10.933
C1 ......... 1.501
C2 ......... 1.070
C3 ......... 0.340
iC4 ........ 0.021
nC4 ........ 0.061
iC5 ........ 0.001
nC5 ........ 0.070
nC6 ........ 0.024
Benzene .... 0.002
MDEA ....... 679.302

TOTAL: 7666.0 kmol/h

If you take this stream into a separator DP=0 and adiabatic, you should obtain, according to mass balance, a 14 kmol/h vapor stream with 0.83 kmol/h SH2, but I'm obtaining 233 kmol/h of vapor with 210 Kmol/h SH2, almost 86% of inlet SH2!.

I'm using HYSYS 3.0.1, Amine Package (Kent-Einsenberg or Li Mather makes no difference in results)and no-ideal vapor model.

Check it your self.

Thanks for your interest ans time,

Regards

 

[PHP78]

dear mondolfi
I have HYSYS 2.4 version. and it doesn't have NH3 for amine package. But it doesn't matter as its conc is very very low compared to major components.
I checked for this tream in separator and I am getting correct answer. Vapour stream is 14.56 knol/hr with H2S 0.612 kmol/hr. I am wondering whats wrong in your simulation.

Regards
Prashant

 

[aikmeam]

Be careful of software bugs. It's typical when upgrading from one version to another. To check, I suggest, simulate using a lower version, then open the same simulation at a higher version. Results should be quite the same if not similar. If problem persist... contact ASPEN Tech, let them crack their head thereafter. It's no more an engineering problem.

 

[PHP78]

dear aikmeam and mondolfi
its problem with HYSYS 3.1 version. I just checked the same stream n HYSYS 3.1 version and got the results which mondolfi was stating before .i.e. 210 kmol/hr H2S in vapor phase.
i think we need to write to ASPEN ..

Prashant

 

[SEAGRAVES]

Dear Mondolfi,

I'm not a Hysys user. I've used Aspen and other programs for amine simulations before. Here's an idea....have you checked whether you specified the H2S and your other gas components as a Henry's component? This is something I've encountered in Aspen that makes a difference in the simulation.

Also, the vessel before the regenerator tower is a flash drum. It typically does take a pressure drop across it in order that the hydrocarbon vapors can be vented off. Are you sure it has zero pressure drop?

 

[Mondolfi]

Dear Seagraves,

After several tries, I have found that the inclusion of NH3 as component (even if the mol rate is zero)will "activate" the bug. If you do the flash without the component, results are as expected.

Regards,

 

[SEAGRAVES]

Dear Mondolfi,

Glad you found the bug. I would have never thought that the NH3 would trigger it.

 

[FarshadN]

Hi Mondolfi,

Maybe it is late, but have you traced the problem back to the exit streams of the absorber?

You can easily check validity of your absorber solution if you calculate rich amine loading. It must be something like 0.35. Li-Mather is much stronger in prediction, but K-E is also good for natural gas applications.

I can check your case if you send me the file itself or make a report (Ctrl+R).

Rgds,

 

[FarshadN]

Hi Again,

Sorry, but I doubt NH3 is the reason. NH3 is not a valid component at all for use with Amine package. It cannot handle NH3. HYSYS does not allow one to select it with the package. How did you manage to select NH3 in the simulation basis manager?

Rgds,

 

[Canoman]

I've run a number of MDEA and DEA simulations in HYSYS 3.0.1 and 3.1, and I have encountered the NH3 bug as well. If you're having lots of difficulty, you can always send a copy of the simulation into the tech support at Hyprotech and they will attempt to help you out. A colleague of mine has found MANY bugs with version 3.1 that weren't present in 3.0.1. They're not limited to using the amines package at all - it just seems as if the last revision wasn't examined as well as the previous version before it was released. Pretty frustrating at times. He's lost lots of work hours due to bugs.

Regards,
John

 

[Mondolfi]

Hi CANOMAN,

I think it would be very helpful for all if your friend share encountered bugs within Hysys V3.0.1 and V.3.1.

Regards,

Mondolfi