7 replies to this topic

[jmsuekam]

Hi,

I am a student in my final year of Chemical Engineering and as part of my final design project I have been tasked with the design of a distillation column to take a very dilute feed stream (0.2%wt ethanol with the balance water) to produce a distillate stream around 90%wt ethanol.

The company I am working with specifically suggested a vacuum operation, and I am wondering what advantage this would offer over conventional distillation. Is it perhaps related to the dilute feed stream?

Additionally, I am curious as to what advantages using a thermosiphon partial reboiler might offer over a kettle reboiler in such an operation.

If anyone can offer any insight, it would be greatly appreciated.

[riven]

First off, this separation is not worth it. Your water stream cxontaining 0.2 wt EtOH should be sent to waste water treatment. I cannot see how any separation process could be economical in this instance. My advice: treat it as waste.

As for vacuum distillation versus conventional. Vacuum distillation will affect your VLE curve. If you draw your VLE curve at atmospheric pressure and compare it to one under vacuum, the two curves will be further apart for the vacuum case. This has the effect of increasing your separation potential/makes the separation easier. Vacuum columns should therefore be less tall than their atmospheric counterparts. However due to the increased volume due to vacuum conditions, vacuum columns will also tend to be wider and there is a trade in cost as to which wins out.

Cannot help you on reboiler.

[siretb]

Riven is correct: the economics are not good enough. Do you need to do it because you want clear, ethanol free water in the bottoms?

Operating under vacuum with make the column much larger in diameter, will complicate the condenser operation, and the separation at atmospheric pressure is not too difficult, if you want only 90% Ethanol in the distillate.
Separation will not be much easier, so, should I do it, I'd operate atmosphéric.

A thermosiphon will have less footprint but the turndown ratio is limited to 2:1 or 3:1. I'd tend to prefer a thermosiphon, should I have to design this kind of separation column

[jmsuekam]

Siretb: You say vacuum conditions will complicate condenser operation, why is this so?

Thanks to both you and riven for your insight.

[riven]

Siretb: You say vacuum conditions will complicate condenser operation, why is this so?

Thanks to both you and riven for your insight.

Condenser design under vacuum conditions:

In vacuum operation it is almost always required to reduce your pressure drop across all items of equipment. Reducing your pressure drop in your condenser requires some interesting designs and may even require you to use verticle condensers rather than horizontal. Indded one could alomst say that the pressure drop issue is more important than the heat transfer (within reason).

[Art Montemayor]

Riven and Siretb have hit on the major points affecting this proposal.

As Siretb has indicated, your scope definition fails to state what you are basically trying to achieve: 1) a pure, overhead product; or,
2) a pure, bottoms product.

Like anything included in the Scope of Work, these factors play an important role in making sound engineering process decisions - especially those related to major equipment.

The vacuum specification will greatly affect the size - and more importantly, the capital cost of the column and all equipment related to it - expecially the overheads total condenser, as siretb has indicated. The cost effect almost spirals upward because the higher the vacuum, the larger the diameter of column, piping, and condenser. And the larger the diameter, the more structural reinforcement required for the vacuum operation. The fabrication (and installation) costs start to climb. However, the separation should go easier (although, in many cases this positive effect is not enough to cancel out the increased capital - and operating headaches.

The choice of reboiler is dependent - again - on the Scope of Work. If your column feed contains impurities or solids that tend to deposit, then the kettle reboiler is certainly a loser since this type of exchanger is a natural and efficient trap for all solids and impurities that tend to settle out. Kettles are good - especially for turn-down and turn-up service - but they are only applicable on relatively clean fluids. The kettle would also be very expensive when submitted to a full-vacuum mechanical design.

A thermosyphon reboiler is usally the instrument of choice in this type of operation - although, you also (as always) inherit a trade-off: the thermosyphon has a terrible turn-down or turn-up capability. It is only effective and efficient when the design rate is maintained steady, and constant. Process variances - especially flowrates - give a thermosyphon a lot of trouble and the results are usually off-spec product. So, a lot depends on the quality, condition, and type of flow that you feed the column and the process variances expected.

I hope this additional experience is of some help.

[katmar]

This is all excellent advice and good comments above, but just one more point on the use of condensers under vacuum conditions.

Don't forget that lowering the pressure lowers the boiling point and this means you need a lower temperature coolant to do the condensation. If your cooling water is not cold enough then you might have to install a refrigeration plant. At the very least, you will lose some of your temperature driving force in the condenser, resulting in more surface area being required.

And while I am writing, just to emphasize Art's point of whether you want a pure bottom product or top product being an important (crucial!) factor.

[jmsuekam]

Thanks again for all your help, especially regarding the overhead condenser.

The objective of the column is to obtain a "pure" (>90 wt% to be sent to molecular sieves for further concentration) overhead stream of ethanol. The reason the feed is so dilute is due to the fact that it is produced in a bioreactor and the bugs cannot handle or produce higher concentration ethanol (the highest they go is about 10 g/L). There are no solids in the feed stream, however, there is a slight amount of acetic acid (about one tenth the concentration of ethanol), but this shouldn't have much of an impact on distillate composition. My apologies for not giving this information earlier.

I modeled both a vacuum column and an atmospheric column in Aspen Plus, and I believe we will be going with an atmospheric design instead due to the operational headaches and increased column diameter from a vacuum operation.