12 replies to this topic

[Andy Craig]

Hey,

I'm currently working on designing an Ammonia Plant, and I'm now working on designing the CO2 removal and recovery section of my system.

The problem I have en-counted is that when designing the absorber, I can't complete the mass balance. I know the mass flow rate of my feed gas and the mole fraction of CO2 in that gas. But I don't know how to calculate the mass flow and/or the mole fraction of amine solution (MEA of MDEA) required to remove as much CO2 as possible.

Also I'm still undecided on which amine to use either MEA or MDEA as both seem to have their pros and cons. but I was wondering which would produce the most energy efficient process whilst still managing to recover a stream of 99.5% CO2, from the Solvent regenerator to be used in a downstream Urea plant.


any help would be appreciated.


Andy

[Zauberberg]

I would look for a formulated solvent (e.g. activated MDEA) rather than MEA or secondary amines. They have advantages over aMDEA (kinetics) but on the other hand they present serious corrosion and reliability issues.

As far as circulation rate is concerned, consider something like 0.35 - 0.50 mol amine/mol acid gas, with 0.35 being more realistic if you operate absorber at lower pressures (<10 bar). Consider 0.50 mol/mol at P>50 bar. If you can get a copy of GPSA Databook, that's probably the best starting point for you.

[Andy Craig]

Using MDEA I have managed to get my absorber working. But now it is my regenerator which is causing the problems. I need to obtain 99.5% purity for use in a down stream UREA plant. I'm using ASPEN HYSYS but I can't seem to find any operating conditions that give me this purity in my recovery. The highest I seem to be getting is around 40%.

I'm using 35% MDEA with a CO2 loading of only around 2% as any higher and I cant seem to remove enough CO2 from my feed gas.

[Zauberberg]

Are you speaking about the purity of regenerator overhead (acid) gas? In one part of your post you are describing the regenerator performance issues, while in the other part you say that you can't remove more CO2 from the Feed gas. Which is the case here?

For regenerator overhead stream, please bare in mind that the acid gas is saturated with water. I don't know what are the conditions in overhead receiver but certainly you can't get that high purity of CO2 as you would like. Dehydration step is required. Some additional adsorption or membrane process could be mandatory as well, if there is significant quantity of hydrocarbons present in the gas.

[Andy Craig]

Sorry, after re-reading my last post I realise I was rather unclear.

In my absorber, I can remove 99.9% of my CO2 from my feed gas.

Stream No. Sour Vapour into absorber
Temperature K 50.00
Pressure kPa 2000.00
kg-mol hr-1 8618.06
Phase 1.00
Mol Fraction kg-mol hr-1
Methane 0.00117 10.08463
CO2 0.17843 1537.68589
CO 0.00084 7.20385
N2 0.19645 1692.99712
H2O 0.00673 57.98847
H2 0.60753 5235.73939
O2 0.00000 0.00000
MDEA 0.00000 0.00000


and this is my stream out of the absorber.

Stream No. Rich MDEA to stripper
Temperature K 50.52
Pressure kPa 3900.00
kg-mol hr-1 4.79428E+06
Phase 1.00
Mol Fraction kg-mol hr-1
Methane 0.00000 4.30142
CO2 0.00032 1535.88595
CO 0.00000 2.13340
N2 0.00007 344.91449
H2O 0.92626 4440759.24189
H2 0.00034 1615.28521
O2 0.00000 0.00000
MDEA 0.07300 349999.99871


So I have removed almost all the CO2 from the feed. But my problem is I have been asked to recover this saturated CO2 at a purity of 99.5% for use in a downstream urea plant. Are you saying this is impossible to achieve?

[Zauberberg]

Assuming that your overhead receiver operates at 0.5 barg (quite common), the composition of overhead gas - even assuming there are no other components (hydrocarbons) in the system - is:

CO2 = 93.6 mole %
H2O = 6.4 mole %

(Peng-Robinson, Sour PR, and UNIQAC, all give similar results)

If you need the acid gas containing more than 99.5% mole CO2, you would need to either:

1. Compress the gas up to ~50 barg (3 compression stages required, cooling to 45C after each compression stage followed by liquid water knockout from the gas), or
2. Another option is to employ solid desiccant for dehydrating CO2-rich gas from the Regenerator, but again you would need to compress it up to certain pressure (something like min. 7-10 barg). The first option is more common for purity requirements that you have.

See attached flowsheet.

Attached Files

•  HYSYS-Print.pdf   47.86KB   178 downloads

[Andy Craig]

Thank you very much Zauberberg, that flow sheet is very helpful.

I have decided that I am going to restart my design though as at the moment my (mol CO2)/(mol MDEA)is at 0.02 (which although I have seen loading this low in literature i'm sure it is not correct). The result of this such low loading is very high flowrates of water and MDEA in my system, which i am sure is not goign to be energy efficent or indeed a viable option.

So I will start the design again from a feed of the following composition:

Temperature (K) 50.00
Pressure (kPa) 2000.00
flow rate (kg-mol hr-1) 14408.06

Mol Fraction kg-mol hr-1
Methane 0.00070 10.08564
CO2 0.10720 1544.54403
CO 0.00050 7.20403
N2 0.11750 1692.94705
H2O 0.40530 5839.58672
H2 0.36340 5235.88900
O2 0.00000 0.00000
Argon 0.00070 10.08564
Ethane 0.00370 53.30982
Propane 0.00080 11.52645
Butane 0.00010 1.44081
Pentane 0.00000 0.00000
Hexane 0.00000 0.00000

I am going to redesign my absorber and attempt to get a better loading in the range you suggested earlier 0.35-0.5 (mol MDEA/(mol CO2).

again thankyou for your help, no doubt you'll see another post form me shortly
regards

Andy

[Zauberberg]

0.02 mol/mol (acid gas/MDEA) probably refers to lean amine loading (regenerator bottoms) but still looks a little bit high. However it can be tolerated if acid gas removal requirements in the Absorber are not in the ranges of ppm CO2 remaining in the treated gas. For deep CO2 removal application, I think 0.015 mol/mol is the highest you can go, especially at lower absorber pressures.

0.35 mol/mol rich amine loading (absorber outlet) is easy to achieve but you need to ensure sufficient number of stages in the absorber, and sufficient residence time since CO2 absorption by MDEA is kinetic-driven process.

In any case, you can do it without Hysys/Aspen - just follow the guidelines from GPSA (I recommend this book to everyone) and see at what amine flows you will end up. Remember that your regenerator overhead gas (or the overhead receiver gas, to be more accurate) composition is solely determined by the:

1. Receiver pressure
2. Receiver temperature
3. CO2 in the Rich amine

Amine circulation flow is not important, and it doesn't affect composition of the overhead gas - just the factors I mentioned above.
Amine flow could affect regenerator overhead composition in reality - by the amount of co-absorbed hydrocarbons with CO2 flowing to the Regenerator. More amine flow = more hydrocarbons in the regenerator overhead stream. And that is something you don't need.

Best regards,

[Zauberberg]

I have uploaded a few pages from GPSA Databook for your reference - if nothing else at least for you to see how simple and comprehensive it is. I remember a clear word from one of the Senior gas processing experts: "Whatever piece of equipment you decide to size based on GPSA guidelines, you'll never make a mistake. The unit/equipment may be too big than what you need in reality and thereby non-optimum design solution, but it will certainly work 100% reliable".

You can get it for only 162 USD (non-GPSA members) or 72 USD if you are a student. Check at: http://www.gasproces...book.html#order

Attached Files

•  Chemical Solvent Processes.pdf   245.83KB   180 downloads

[Andy Craig]

Ok, so I have managed to fully simulate my CO2 removal process in HYSYS (couldn't get a copy of GPSA hand book, so couldn't perform the hand calculations).

The problem is I need to size my columns, obviously now I know all my flow rates, pressure drops, number of stages, column diameter etc. But I'm unsure of the appropriate way to calculate the height considering I want to use structured packing?

Any suggestions?

[Zauberberg]

I wouldn't use structured packing for amine towers, if I were at your place. Absorbers are usually random-packed (e.g. IMTP from Koch-Glitsch) or trayed, Regenerators are almost always 100% trayed towers.

Any distillation/absorption textbook is good enough for preliminary design calculations. I believe some spreadsheets are also available at ChE website. Additionaly, you can download KG Tower for preliminary tower sizing at: http://www.koch-glit...re/default.aspx

Keep in mind the overall system/tray eficiency - it can be as low as 25% for Absorbers and 45-50% for Regenerators.

[Che JEDI]

Why would you not reccomend structural packing in an amine tower. from my research i have found that structural packaging would give higher efficiencies. Also for the regenerator i came across examples where they have used packed column in industries e.g. UoP. if you could help or clarify with this that would be really helpful.

Edited by Che JEDI, 25 March 2010 - 09:17 AM.

[Zauberberg]

No doubt that structured packing will yield higher efficiencies, and lower pressure drop as well. However, in reality, towers equipped with structured packing frequently suffer from vapor-liquid maldistribution, accelerated fouling, they are more sensitive to foaming and damage of internals/pressure surges, etc etc. Nowadays, plant reliability and availability are major issues, and everyone tries to design and run the plant in such way so that all possible risks of having unplanned and extended shutdowns are eliminated as much as possible.

The only service where accepting all these risks is justified - based on my experience - is Crude Vacuum Distillation Unit, where low pressure drops are of vital importance. Perhaps there are some other applications in Petrochemical or Pharmaceutical industry which I am not aware of, and which also call for structured packing.

I have never seen an amine Absorber equipped with structured packing in real, existing plants - and I have seen or investigated dozens of them worldwide. Not saying that your research is a bad idea, but it should also take into account all factors which real plants are faced with - and those are the ones listed in the first paragraph of my post. Running the experiment of a tower with structured packing in amine service should account for corrosion, particles, foaming, vapor-liquid (mal)distribution issues, prolonged service (e.g. 3-5 years without cleaning/maintenance), and then some real judgements can be made. I wish you success in your work, and I hope you will come up with some interesting findings and conclusions in future.

Best regards,