8 replies to this topic

[tranthenam]

Hello everybody. I'am trying to simulate my 1,3 butadiene plant by extractive distillation using NMP solvent by hysys software. Does anyone has experiences about it. Please help me. I tried to used several fluid packages, but the result didn't correct.

[tranthenam]

please help me..

[latexman]

You have not shown your model, or described it and the thermo. methods you are using.  So, we have nothing to look at.  You said it is not correct, but we don't know what is not correct.  Who do you think we are, the Amazing Kreskin?  Only you, can help us help you.

[StealthProg]

Use A+ NRTL, you will likely have to regress some data to get good parameters for the NRTL

[breizh]

Hi,

latexman is right ! we are not wizard.

 

Did you try your favorite search engine ? 

 

https://link.springe...1814-012-0075-3
https://www.research...olidone_solvent

 

 

Good luck 

 

Breizh 

Edited by breizh, 30 January 2021 - 08:59 PM.

[SilverShaded]

Use A+ NRTL, you will likely have to regress some data to get good parameters for the NRTL

If you don't have any real data then you may be able to estimate the NRTL parameters using UNIFAC.

[tranthenam]

dear sir.

I'am sorry, I can not show the simulation because it's confidential document. I tried to used NRTL-RK method for fluid package and UNIFAC to predict the binary paramter of component. But when I reduce pressure through a valve, the flowrate of vapor and liquid not correct that lead the result not correct.

[Pilesar]

A couple of suggestions:

1) tune your thermodynamic package against published data such as from here: https://www.scienced...378381289800408

2) Contact BASF for modelling guidelines since they developed this technology.

[pmoneysh]

Your question is vague. For extractive distillation, next to the choice of right thermodynamic model, important aspects include the amount the solvent  and the reflux ratio. Further, how did you calculate the required number of trays or packing height and what is the purity and recovery of the components? Binary interaction parameters available for all the pairs?