Sciencemadness Discussion Board

Distillation problems using different diameter glassware.

SWIM - 1-4-2019 at 08:12

Is there much of a problem with using equipment with varying diameters in a distillation rig?
What I mean is can you use a larger diameter fractionating column with adaptors without running into some sort of problems from the change in vapor velocity from the column to the head?
I'm talking about using a 45/50 column with adaptors along with a 29/42, or even 24/40 size heads and condensers.

I'm thinking the wider column might even be a slight improvement because of the slower vapor transit through the column, but would appreciate hearing any thoughts or experiences on this sort of thing.

Sulaiman - 1-4-2019 at 08:21

over-sized columns allow greater throughput at the cost of greater holdup,
- for a Vigreux column that should be negligible.

It is possible that heat cannot be put into a small rbf holding the boiling liquid
as quickly as an oversized column can remove it
- unless insulated to some degree

Given the choice I'd rather run an (easily insulated) oversized column than an (easily flooded) undersized column.

You can run your heating for rapid boiling
and adjust/reduce the insulation of the column to increase the internal reflux ratio
(giving better fractionation)
so that the smaller condenser is not overwhelmed.
(or more practically, increase the column insulation until the vapour front reaches the condenser and a reasonable product collection rate is achieved).

[Edited on 1-4-2019 by Sulaiman]

SWIM - 1-4-2019 at 13:39

Thanks Sulaiman.
Good to get the opinion of an engineer who has an interest in these matters as well.
It's actually a 600 mm column, but with a vacuum jacket and silvering, so the heat lose is probably pretty low.




markx - 5-4-2019 at 09:46

If you wish to run the system at a capacity that is allowed by the size of the column, then the smaller "bottlenecks" in the vapour path are likely to create a bubblecap effect. The liquid phase can not return through the restricting connectors due to vapour speed and pressure from below and starts to collect above them causing flooding problems. Also it may create sharp jumps in the pressure gradient through the column. If none of these effects shall cause harm to your purpose, then it probably shall work just fine.

Sulaiman - 5-4-2019 at 13:17

Quote: Originally posted by SWIM  
...It's actually a 600 mm column, but with a vacuum jacket and silvering, so the heat lose is probably pretty low.

I have no experience with silvered vacuum jacketed columns,
but if the silvering and vacuum do a good job as I believe they should,
there will be very little heat loss from your column,
so there will be very little internal refluxing.

This implies to me that you need to either go EXTREMELY slowly to have good fractionation,
or, add a reflux condenser and partial take off head above the column.
Other than that I guess that you will need a less efficient column :D

Regarding the vapour flow mentioned by markx above,
to me this translates as 'don't put packing in narrower sections'

BromicAcid - 5-4-2019 at 14:07

Round and round we go. You want the column completely adiabatic if possible. Vacuum jacketed silvered columns help loads. Ideologically I agree with Sulaiman that you want heat loss so that you have refluxing in your column to give theoretical plates. Unfortunately that is completely false. Better minds than myself have looked into this. If you want to dig into it you can start with the SM threads then work into the literature.

http://www.sciencemadness.org/talk/viewthread.php?tid=38826

http://www.sciencemadness.org/talk/viewthread.php?tid=12409

Magpie - 5-4-2019 at 14:51

Adiabatic is good. Note that refinery distillation columns are insulated. What you want is condensing of low vapor pressure components and evaporation of high vapor pressure components. That is the job of the column. This is not refluxing.

Narrow columns are prone to flooding but have less holdup. Longer columns give better separation but you need more heat to drive the components up the column.

I usually work with small volumes so excessive holdup can be a problem for me.

[Edited on 5-4-2019 by Magpie]

[Edited on 6-4-2019 by Magpie]

S.C. Wack - 6-4-2019 at 10:22

It would be simple enough to experimentally test separation ability with whichever smaller and larger size (except for the column) and the same amount of the same mixture at the same rate. It would not be surprising if the results varied by operator, mixture, and hardware.

It would also not be a surprise if columns are typically designed for a vertical rather than horizontal thermal gradient...what works for a Vigreux is not what works for a packed column.

[Edited on 7-4-2019 by S.C. Wack]

markx - 8-4-2019 at 01:52

Quote: Originally posted by Sulaiman  


Regarding the vapour flow mentioned by markx above,
to me this translates as 'don't put packing in narrower sections'


One does not have to put packing there (it sure does not help with improving "floodproofness") and can still have problems if the dimensions are a big enough mismatch. But really, every setup is different and behaves in a different manner. Even between separate runs one can observe remarkable fluctuations on the same setup regarding efficacy, flooding and output.

The discussion about what is the best configuration can be stretched for miles and years, but in the end one just has to put together a system and learn how to operate it in a way that is most suitable for this apparatus and the end purpose.

SWIM - 8-4-2019 at 17:30

Okay, adiabatic with partial take off.
Reckon I'll use one of these magnetic gizmos.
The other picture is with it on the column. I added a coonskin cap to the picture for scale.
That way you can use your own coonskin cap to estimate the size.


DSC04114.JPG - 2.9MB DSC04116.JPG - 3MB

The head's got an 24/40 adaptor stuck in the top that's probably never coming out because the vacuum jacket makes it tough to use a torch, and it can't be pushed or tapped from the bottom because that thing has obstructions inside it (glass tubing, fragile glass tubing), but that puts the restriction between the head and condenser which may be less of a problem than having a restriction right above the column.

It'll be the second shiniest still around this lab.

markx - 8-4-2019 at 22:19

Impressive glassware! :) Be sure to treat it carefully, would be a shame if one breaks a beauty like that one.
I would leave any stuck adaptors or other joints well alone as they are and work around them by adjusting other parameters to make the system work properly.