17 replies to this topic

[gilad]

I did research on topic of separation of ABE products and found that one of the method used is 3 distillation Columns. In the first one acteone ethanol and water are sent to the top stream, and the butanol sent to the bottom stream. The top stream is sent to a distillation column to separate ethanol and acetone. The butanol stream continues to another distillation column to be separated from water.
I attached a diagram of the process.


I was wondering why not separate the components based on volatility?
First acetone, then ethanol and lastly Butanol.
One reason I can think not to do it is there might be more azetrope mixtures: ethanol-water,butanol-water.
------>harder separation---->more expenisve unit operation to perform the separation----->increase in capital cost.



I would love to hear your inputs on the subject.

Attached Files

•  Diagram.jpg   91.97KB   101 downloads

[riven]

I cannot properly comment as there are not enough details on the feed to the ABE separation column or how pure the authors of the article expect their steams to be specifically around ABE separation column. If you can provide that information we can start discussion.

I was wondering why not separate the components based on volatility?

That is what they are proposing. They propose to remove the volitile components in the ABE separation column leaving the 'heavies'; butanol water.

[gilad]

First of all thanks for you respone.

So far, I have not found the data you requested.
However, lets try starting a discussion based on an hypthetotical case.
Feed is diluted with the following Weight Percent
93% water, butanol 4.6 %, ethanol 0.5%, Acetone 1.9%.
Requirement of Butanol/Acetone/Ethanol products are to be 99.5 percent pure.

That is what they are proposing. They propose to remove the volitile components in the ABE separation column leaving the 'heavies'; butanol water.
[/quote]

I didn't formulate my question right in the first post.
Why not seperate each component at a time instead of two at a time?

[riven]

You still need to specify the data I requested. To put it more clearly what is required is the mass balance around the ABE separation column.

You have specified the feed now we need to know the compositions of the outlet streams from this colum i.e. the composition of the feed to the ethanol acetone column and the composition of the feed to the butanol water column.

The reason for not separating each component one by one is due to azeotropes forming, componetns will boil off together. And this is why I want the above data. Ethanol and butanol will form an azeotrope with water as you already specified. Therefore I would expect water to exit the top of the column in the acetone ethanol product stream. This would complicated the final goal to get pure ethanol and pure acetone.

The butanol water separation will require a separation scheme using 2 distillation colums and a phase separator. I will attach some files at a later date.

[gilad]

I am working on simulating the distillation process using ChemCad. I managed to do all the separations using Shortcut column. However, when I tried to change them to regular towers one tower doesn't converge.
Will it help if I attach the chemcad file?
I was leaning toward breaking the azetrope between butanol and water by adding cyclohexane or benzene as entrainer.
I would appreciate if you attach the file regarding the Butanol water separation.

[gilad]

Here is the ChemCad simulation of the distillation process.

Attached Files

•  Distillation_ABE.zip   74.9KB   80 downloads

[riven]

Here is the ChemCad simulation of the distillation process.

I do not have chemcad but I will see if I can get it installed.

I would expect that the butanol water tower would not converge, if it is the simulation is probably producing questionable results. I would need to see the simulation however.

In the meantime which column is not converging. And how many colums are you using to simulate the final 3 steps.

[gilad]

Don't bother looking at the Chemcad, since it is not a true represnattion of the situation.
I looked more carefully at the chemcad simulation using towers. I observed that different results are obtained when using tower as oppose to shortcut. I can't think of a reason for that.
Those different results don't allow me to simulate the tower configration using data from the shortcut.
I used only 3 columns.
Thanks for all your help.

Here is the ChemCad simulation of the distillation process.

I do not have chemcad but I will see if I can get it installed.

I would expect that the butanol water tower would not converge, if it is the simulation is probably producing questionable results. I would need to see the simulation however.

In the meantime which column is not converging. And how many colums are you using to simulate the final 3 steps.

[riven]

3 towers for the complete train from ABE separator to butanol tower and acetone ethanol tower will not be enough for sure. For the butanol water separation see below: gives a good idea for what is required.

I would expect that a minimum of 1+2 (acetone ethanol water separation) +2 (butanol water) distillation towers to be required. However a single tower may be acceptable if there is not much water in the acetone ethanol tower.

Update having trouble attaching the file will try again later.

[Patrician]

Greetings and sorry for little offtopic.

What is the efficiency of gas stripping in this process?
Wouldn't it give us highly concentrated azeotrope of ABE and little amount of water?

[gilad]

The bacteria used for the process can produce ABE to a maximum titer of 33 g/L total solvent (water). A higher product concentration will cause product inhibition. By using gas stripping the ABE products are continuously carried out with the gas, so that inhibitory ABE concentration is never reached.
If you want, I can refer you to articles about gas stripping.

You are right, but the goal is to use minimal amount of water.

Greetings and sorry for little offtopic.

What is the efficiency of gas stripping in this process?
Wouldn't it give us highly concentrated azeotrope of ABE and little amount of water?

[Patrician]

If you want, I can refer you to articles about gas stripping.

I'll be greateful.

You are right, but the goal is to use minimal amount of water.

They're using steam or neutral gas?

[gilad]

They are using CO2 and H2 gases produced in the fermenter.
Look at "production of acetone-butanol-etanol from concentrated substrate using clostirduim acetobutycilum in an integrated fermentation-product removal process" by Maddox.

[riven]

I have been unsucesssful in attaching the file.

Reference
Control of the Heterogeneous Azeotropic n-Butanol/Water Distillation System
William L. Luyben
Energy Fuels, 2008, 22 (6), pages 4249-4258

DOI: 10.1021/ef8004064 • Publication Date (Web): 19 September 2008

[gilad]

Thanks for the link it was really helpful.
I have one question regarding it if you don't mind.
For the Butanol/water separation with 10 mol% butanol the feed is fed staright to the decanter, whereas for a more diluted feed it is fed to the distillation column and then to a decanter. What is the advantage in feeding staright to the decanter?

[riven]

Thanks for the link it was really helpful.
I have one question regarding it if you don't mind.
For the Butanol/water separation with 10 mol% butanol the feed is fed staright to the decanter, whereas for a more diluted feed it is fed to the distillation column and then to a decanter. What is the advantage in feeding staright to the decanter?

The feed to the decanter will split into two phases at the higher feed concentration, 10% butanol. This split produces two streams that can be seperated into two pure products.

If we fed directly into a column first, this column would be handling all of the water butanol mixture. It would produce pure water and the resulting azeotrope stream would be fed to a decanter. The size of this azeotrope stream would be more or less the same size as the stream in figure 4. Therefore the resulting decanter and second column would be similar size to that in figure 4. The first coulmn would be larger and require more energy.

[drumjeff215]

I would expect that a minimum of 1+2 (acetone ethanol water separation) +2 (butanol water) distillation towers to be required. However a single tower may be acceptable if there is not much water in the acetone ethanol tower.

How do you distill the off the acetone and ethanol off from the butanol water? Simulating this setup in ASPEN results in ethanol going into the bottoms with the butanol and only acetone going off the top. My acetone ethanol tower has a lot of water in it.

[gilad]

You are right most of the ethanol goes on the bottom, but it's weight fraction in the bottom stream is quite small (for my case 3.3 %weight). What is your definition of a lot? I get 7 weight percent in the tops.