4 replies to this topic

[CS Kang]

Hi,

I want to do a short-cut distillation on hysys to separate ethanol from a butanol-ethanol-water mixture with 01425, 0.8197, 0.0379 feed composition respectively. My target for the purity of ethanol is 99.5% so I put the light key in bottom to be (1-0.995)=0.005 composition and heavy key in distillate which is butanol to be as low as possible (I put 0.0001) because I want my bottom product to be high in butanol since the aim is to produce butanol. However, Hysys indicated that the minimum number of trays cannot be determined after 20 iterations. Can anyone help me? Could it be azeotropes that is thwarting the separation? I used NRTL as my fluid package as recommended by some research. 

Thanks in advance!

[PaoloPemi]

for distillation I have a different software (PRODE PROPERTIES) so I am not familiar with the features of "short cut" procedure in your simulator,

as far as I see there are no problems with PRODE (and I expect your simulator) to solve a column with similar feed composition by fixing a number of stages and include a purity spec.,

you can repeat the procedure untill you'll find the optimal design,

by the way Excel can do the work for you, if required,

which are the advantages of that short cut (which seems difficult to converge) ?

Edited by PaoloPemi, 06 February 2014 - 02:09 PM.

[CS Kang]

for distillation I have a different software (PRODE PROPERTIES) so I am not familiar with the features of "short cut" procedure in your simulator,

as far as I see there are no problems with PRODE (and I expect your simulator) to solve a column with similar feed composition by fixing a number of stages and include a purity spec.,

you can repeat the procedure untill you'll find the optimal design,

by the way Excel can do the work for you, if required,

which are the advantages of that short cut (which seems difficult to converge) ?

Thanks PaoloPemi for your prompt reply. I actually thought the problem lies in the possible azeotropic mixture that could be formed by ethanol-water or butanol-water where at certain molar fraction the components could not be separated. But after specifying the top product composition, where we want as much ethanol as possible at the distillate, the simulation gives us 90 over column trays. Is it ridiculous to have so many trays?

[PaoloPemi]

given a composition the number of theoretical stages depends from required degree of separation (specs) and reflux,

have you solved the problem in traditional way  (I mean a rigorous simulation with specified efficiency) ?

An advantage of traditional way (i.e. different simulations with different number of stages) is that for each case you can estimate the cost of the column (i.e. number of stages, and size, related to flows) and energy (i.e. reboiler duty) so at end you get an optimized design.

Edited by PaoloPemi, 06 February 2014 - 02:56 PM.

[Bobby Strain]

I'm pretty sure that with your system, the ethanol-water azeotrope will be the lightest boiling. And the azeotrope is about 95% wt ethanol. So you should probably be looking at a two-column system. Or use molecular sieves to remove water from ethanol. So my advice is to follow Paolo's advice and use rigorous calculation for the separations.

 

Bobby