5 replies to this topic

[Low]

Dear all,

For our final-year project, my group and I are supposed to design a benzene production plant, to the amount of 500,000 tonnes per year, using toluene and hydrogen as feedstock. We are using a adiabatic reactor without any catalyst.

After the reaction we would like to purify the hydrogen before recycling it back to the reactor. My question is which method would be better, cryogenic separation (not using a distillation column, but would it be better?) or membrane separation? Considering the following conditions as well:

i) we started our preliminary design using cryogenic separation
ii) we know the utility cost would be high for cryogenic systems but how would the cost of a cryogenic system compare to that of a membrane separation system?
iii) either system the recycled hydrogen would have to be recompressed as we are operating at 38 bars at the reactor inlet.

I would really appreciate the help and advice.

Edited by Low, 22 February 2011 - 03:15 AM.

[Art Montemayor]

Low:

Allow me to give you some practical engineering advice for what is clearly an engineering problem:

Before doing anything else, use practical common sense in approaching your problem by simply defining it in simple detail to yourself and to others:

• You have a Unit Process to design. This involves a reaction and products, some of which will be recycled back to the reactor as feed (like Hydrogen).
• The recycle stream is originally a mixture of reactor products. It must be treated and/or purified prior to being introduced back into the reactor.
• It is common sense that in order to determine how to bring about the treating and purification of the recycle stream YOU MUST IDENTIFY WHAT IT IS THAT YOU ARE TREATING OR PURIFYING. This is a common requirement. Otherwise, you can’t discuss any purification process and much less evaluate it.
• Therefore, in order to reach this decision point, you need to have a complete heat and mass balance before identifying what recycle purification process you need or can use. You must identify the composition and conditions of what you intend to purify. You must also clearly identify the minimum purity of the Hydrogen that you can employ in the recycle stream. Don’t you agree?

You haven’t done any of the above or, if you have, you haven’t revealed it to us. Please tell us the IMPORTANT AND REQUIRED BASIC DATA. We don’t care anything about your annual production rate, your feedstock, or whether your reactor is adiabatic or has any catalyst. As a student, you should realize and accept the fact that if you require an accurate and helpful answer to a problem, you need to communicate with required and useful basic data and information. You should know that what we need from you is the following basic data (as a minimum!):

• The composition and conditions of your product stream;
• The required minimum purity of your recycle hydrogen gas AND its conditions prior to being injected into the feed stock;
• A PFD diagram showing all the process details as flows, compositions, conditions.

You should have generated a preliminary PFD with a heat and mass balance already.

[Low]

Thank you for your reply Art. I apologize for the lack of info. I did not want to clutter the thread initially, but I guess I ended up giving close to zero information about the process.

I couldn't attached a PFD, as I have it in Visio format, as well as the HYSYS simulation file (Would that help?).

Stream 11 would be the one where the hydrogen is separated from the rest.

Stream 11: 30C, 36 bar, 5991 kgmol/h
0.854 H2
0.141 CH4
0.004 Benzene
0.0005 Toluene
trace amount of Diphenyl

To be honest I'm not sure how pure the hydrogen should be but for the simulation attached the results were obtained using about 98.3% purity hydrogen recycled with 100% pure hydrogen make-up(assumed). The recycled hydrogen would have to be at 38 bar and 37.7C.

Once again thanks!

p.s. I could email the HYSYS and PFD to you as I couldn't use the Excel convertor/dump to work on HYSYS 2006.

[Low]

I used 'Print screen' to get a shot of the PFD. I hope it's large enough. Please let me know.

This is relevant to the mass balance sheet I uploaded.

[pavanayi]

Low
Looking purely at the flow rate and hydrogen concentration of the inlet stream, I would recommend neither membrane separation nor cryogenic process, but a more suitable Pressure Swing Adsorption process. Benefits include
1. Pure hydrogen stream will be nearly at the same pressure as the feed stream, so you end up saving on re-compression costs.
2. The impurities (Methane, BTX) are best adsorbed in a PSA bed.
3. PSA is a fairly well established process than membrane separation.

I have attached a file I found from UOP website (which, as per google, have been cited 21 times, so should be reliable ). All credits to the original authors Miller and Stocker.
You can find a comparitive review between the three processes.

Attached Files

•  SelOfHydroSepProc.pdf   1.26MB   74 downloads

[Low]

Pavanayi,

Thanks for your advice, Pavanayi. I read that article before and I think our group ended up selecting cryogenic anways. One of the problem with PSA for me is I do not really know how to simulate it in the simulator except for using the splitter with hand calculations. And we would have to change all the mass balance again as well.

Perhaps we might have to look into PSA, but the detailed design for the cryogenic system has to start very soon.

Appreciate your advice.

- Low