9 replies to this topic

Dear all,
Recently, I just try to develop NGL recovery using turbo-expander like in the PFD I attach below. All data required are inserted for each stream. When I want to run demethanizer column (the 1st column), the error status is appeared, "column 23 does not have enough information for the column to run". I just insert pressure and temperature in stream 23. is anything I missed?

Besides that, is it possible to run two fractional column simultaneously like in the PFD I attach below? All your help are highly appreciated. Thank you

Attached Files

•  NGL Recovery (turbo-expander).doc   52.5KB   166 downloads

[Zauberberg]

Yes, you certainly have to provide more data since the tower shown in your PFD is a multiple-feed column. You need to define P,T, flow and composition for each feed stream (otherwise, how the software can know what enters the column?). Since some of them are recycle streams, you need to use the "Recycle" unit operation in HYSYS. If you are not familiar about it, read more in the HYSYS tutorials.

Yes, it is possible to run two columns simultaneously, provided that all the feed streams and column parameters (degrees of freedom) are fully defined. Remember to use the "Recycle" unit operation whenever you use a product stream (or product stream fraction) as feed to the unit operations upstream of the unit operation from which the product stream exits.

[Erwin APRIANDI]

Yes Zauberberg is right,

please remenber when you do a simulation,
the principle of simulation is the same as an equation 1+1+1+1+1 is equal to ?
means you need the input stream to converge, before the next stream can calculate

while in you case

you only provide 1+A+B+C+D is equal to ?
even a genius can not solve that.

An yes, you have to provide a recycle on each recycle stream to allow the simulation run
later on, but dont't forget to check the recycle tollerance for how much accuracy that you required?

please see enclosed for further details

Attached Files

•  NGL Recovery _turbo-expander_.pdf   32.97KB   230 downloads

Thanks a lot for the help. Now, I can run the column but new the problem is occur. The column is not converged. Is anything I miss when running the column? To let you know, this simulation is for NGL recovery For the first column, some of the methane will recycle back to the system and some will be the residue at stream 39. How can I do that and make the column converged? Please help me

Edited by Hamidihizami, 27 August 2010 - 08:41 AM.

Can anyone help me?

[Zauberberg]

Zip the Hysys file and post it in the forum, otherwise we need a crystal ball in order to see exactly where the problem is.

I already zip it. Thanks a lot for your help

Attached Files

•  TURBO-EXPANDER 3.zip   127.01KB   153 downloads

[Zauberberg]

Is the first column De-C1, or De-C2 tower? It looks like it is De-C1 but I couldn't figure it out from the specs. If it is a Demethanizer, column pressure is too high (in reality it does not go above 23-24 barg), and critical conditions for Ethane are exceeded in the reboiler, which means it cannot be recovered as liquid product.

Also, the first tower is fed with so many streams which creates difficulties for the model to converge. It would be good if you can give us a little bit more detailed description of what are you trying to separate in these towers, and what are the product specifications.

Thanks,

Is the first column De-C1, or De-C2 tower? It looks like it is De-C1 but I couldn't figure it out from the specs. If it is a Demethanizer, column pressure is too high (in reality it does not go above 23-24 barg), and critical conditions for Ethane are exceeded in the reboiler, which means it cannot be recovered as liquid product.

Also, the first tower is fed with so many streams which creates difficulties for the model to converge. It would be good if you can give us a little bit more detailed description of what are you trying to separate in these towers, and what are the product specifications.

Thanks,

This is the description of the process flow diagram. It start at [0042] but the composition I change a bit in HYSYS simulation

Attached Files

•  NGL Recovery Methods and Configuration (Turboexpander).pdf   253.38KB   228 downloads

Edited by Hamidihizami, 28 August 2010 - 04:26 AM.

[Zauberberg]

I went through your simulation case and compared it with the article attached in your previous post, and there is a lot of conceptual discrepancies, e.g.:

- What kind of multiple draw-offs have you employed in the first tower?
- Why are you feeding the reboiler with an external material stream?

I would say that you need to reconfigure the model, and confirm what kind of separation (and products) you need to achieve. If you are not familiar with cryogenic processing, please go through related chapters in the GPSA Databook and Campbell's "Gas Conditioning and Processing" to get a feeling how these things work in reality.