Continuous stripping column - old style

[snuffy]

Google Books is a great reference source. FWIW, the search term I've found very useful is "Pistorius" - the ancient ancestor of the modern eau de vie still.

While rummaging around, I found this little gem: Fractional Distillation by Sydney Young.

http://books.google.com/books?id=fCVIAA ... =firefox-a" onclick="window.open(this.href);return false;" rel="nofollow

In Chapter VII, Continuous Distillation (page 201 in the PDF reader), there are two lovely little stripping columns. And I do mean little. The first, by Carveth, had problems but worked after a fashion. The principle was improved by Lord Rayleigh (yup, that Lord Rayleigh of Rayleigh distillation fame) and worked nicely. The illustration in the book shows a Rayleigh column designed for separating three components. Remove the middle tank/coil and it would do for stripping.

Since Lord Rayleigh lived in a world lit only by fire, the apparatus has burners to heat it. Now that we have things like heating tape and temperature controllers, it could be possible to build a continuous stripper without a boiler that uses an electrically heated column. The top half of the column (above the entry point) would be at the boiling point of alcohol and the lower part would be slightly below the boiling point of water. If the feedstock was used first in the alcohol condenser and then a heat exchanger on the hot water outlet, it could be reasonably efficient.

Take a look at the reference. Has anybody seen a rig like this? It's almost the same as a steam column stripper, but the manner of heating is quite different.

[snuffy]

Upper and lower sections would need separate thermostatically controlled heaters. Controlling the flow would tune the system. The plumbing might look like...

stripper.gif (4.01 KiB) Viewed 5906 times

Obviously not to scale.

[manu de hanoi]

I couldnt get the pdf from google library, how did u get it ?

About your picture:
The wash feed should be on top , just before the elbow. But kudos for the simplicity.

[snuffy]

Linking to Google books doesn't seem to work. Heres a different HTML page, same content

http://chestofbooks.com/science/chemist ... ation.html" onclick="window.open(this.href);return false;" rel="nofollow

If you search Google books for the title and author, you'll find it. I think my bookmark works only on the computer I saved it on.

At any rate, the feed is in the middle of the column and the column is held at two different temperatures on each side of the feed.

[manu de hanoi]

the stripper in the link provided is different than the picture above.
Anyways, feeding on the middle is ok on the link pic because there is reflux above. But feeding on the middle in the pic above is not ok (cause there is no reflux)

[snuffy]

The upper and lower columns are wrapped with thermostatically controlled heaters. So it's more like the first one, but without the overheating problem.

Heaters like these (it would probably take a total of ten) 1490 HEATING ROPE

http://www.sciplus.com/category.cfm/category/155" onclick="window.open(this.href);return false;" rel="nofollow

Omega.com has them, but they are way spendy - +$400.00US for cable with tstat.

[manu de hanoi]

it's cooling you need on the upper part , not heating, the easiest way to achieve this according to me would be to use liquid 95% ethanol to keep the max temp at 78.2 C.


Cheers

[snuffy]

I think the column works as an externally heated flash boiler, not as a reflux column. The only condenser in Carveth's design is off the picture past point D and the condenser in Lord Rayleigh's design is clearly a product condenser.

The upper part of the column is at the boiling point for alcohol, so none will condense in the column but instead will leave as vapor. Any water in the upper column will condense. The lower column is at the boiling point of water (or maybe just below that.) Any alcohol that gets into the lower part gets flash evaporated and stays as vapor until it leaves.

The column (or Rayleigh's coils) slow down the liquid so it doesn't leave the column too fast.

The feed has to be in the middle of the column, otherwise the alcohol vapor would carry too much water with it.

The mixture was introduced at such a rate that it fell in a rapid but visible succession of drops and, when rich in alcohol, it was previously heated. Mixtures containin 20, 40, 60 and 75 percent of alcohol, were distilled and in all cases the water, collected in the lower receiver, was nearly pure, never containing more than 0.5 per cent. of alcohol. The distillate varied but little in strength and contained from 89 to 90.3 per cent of alcohol.

My guess is that this type of apparatus was forgotten - maybe because of the difficulty of temperature regulation. A continuous (as opposed to switching) temperature controller would solve that problem.

To help with the preheating, I added a product condenser at the alcohol exit and a liquid/liquid heat exchanger at the water exit. This will improve thermal efficiency some and lessen the temperature drop when the feed enters the column. The one point to watch out for would be to de-tune or undersize the lower heat exchanger so the feedstock does not get heated above its boiling point before it enters the column. If it did boil, it could get vapor-locked.

I posted the reference because I've never seen something like this. The closest I've seen was a propane vaporizer that was used for feeding large (did I say LARGE?) quantities of propane into a generator (the generator was about the size of a railroad locomotive) at a refinery. But it was just a flash boiler, no separation took place.

I just ordered some heating cord in case I can actually figure out what the dimensions on this critter might be.

[manu de hanoi]

The upper part of the column is at the boiling point for alcohol, so none will condense in the column but instead will leave as vapor.

Wrong.

[snuffy]

Carveth suggests that by maintaining two parts of a system at different temperatures, corresponding to the boiling points of the two components of a mixture, it should be possible to effect continuous separation, especially if use were made of dephlegmating intercepts. He describes an apparatus used by Derby, in 1900, for mixtures of alcohol and water. It consists (Fig. 56) of a long block-tin tube surrounded at its lower end, a, by the vapour of boiling water, at its upper end, b, by vapour from boiling alcohol, and filled with intercepts. The mixture of alcohol and water to be distilled was slowly dropped in at c, the vapour passing to the condenser at D, and the residue through the trap at e.

carveth.gif (8.71 KiB) Viewed 5653 times

Lord Rayleigh describes an apparatus similar in principle to Carveth's except that the temperatures of both parts of the still-head are regulated by liquids and not by vapours.

The apparatus consists of a long length (12 metres) of copper tubing, 15 mm. in diameter, arranged in two spirals which are mounted in separate pails. For the distillation of mixtures of alcohol and water the lower and longer spiral was heated by boiling water, the upper one by water maintained at a suitable temperature, usually 77°. The spirals were connected by a straight glass or brass tube of somewhat greater bore, provided with a lateral junction through which the mixture could be introduced. With the exception of the two extremities, the whole length of tubing sloped gently and uniformly upwards from near the bottom of the lower pail to the top of the upper pail, where it turned downwards and was connected with an ordinary Liebig's condenser. The lower end of the tubing was, if necessary, connected with an air-tight receiver heated to 100°.

rayleigh.gif (22.45 KiB) Viewed 5646 times

[manu de hanoi]

let's take a closer look at this important sentence:

But the use of the vapour of alcohol for the upper part of the still-head is wrong in principle, and it is a mistake to suppose that this part would thus be maintained "at a temperature corresponding to the boiling point of alcohol." What is here required is to keep the temperature as far as possible from rising above the boiling point of alcohol, in order that the vapour of water may be condensed as completely as possible, while that of the alcohol passes on to the condenser.

This sentence explains why there is indeed some condensation (of water) happening at 77C. But even this sentence is wrong, because you cant selectively condense water as demonstrated empirically by Emile Barbet around the 1890's. This book is from 1922 and it really show that this concept was hard to accept by many pple in the business.

If you want some good old school designs check the masters first :
Coffey still 1831
Brevet de la colonne à distiller Savalle (1868)
As a book I cant recommend better than the Barbet manual (in french). Most of the things you'll find on the free online distillers manuals (until the 1920's)is just some bad copypasta from Barbet's manual.

[vajravarahi]

The book can also be obtained free from the online French National Library:

http://gallica2.bnf.fr/" onclick="window.open(this.href);return false;" rel="nofollow

Simply search "Emile Barbet" (without the quotes). Can also download for offline reading. Many interesting books on distillation and fermentation from the 19th and early 20th centuries. Some even going back to the late 18th century.

You see, there was a reason you should have studied French in high school. Not only would you have been a hit with the girls, you'd be able to read these important books on distilling. (Of course, I took German because I wanted to learn the language of my great-grandparents. But they didn't speak it anymore except to swear, so it didn't really do me any good.)

[snuffy]

Cool on-line library. My French was never good and now seems worse, but I can puzzle my way into much of it. Exhaustive encyclopaedists seem to find France fertile soil.

The two experimental apparatuses don't vaporize the water (much). The max temperature going into either rig at the bottom is at the boiling point of water so there won't be much vaporization because of the small temperature differential. The water will be right at or below boiling and any vapor will get condensed very quickly, so water isn't moving very far uphill.

The original question was "are there any examples like these?" and the answer is still no. One of the things that prompted me along this path was a minor squabble that arose over the wording of Washington state's new microdistillery law last winter.

The original language tried to sharply regulate the type of stills allowed under the craft distilling license. After testimony, the language was softened. I later found out that the sticking point was testimony and lobbying by some people who wanted to build a new type of still that would use something like a thin-film flash boiler. So the language in the original draft was changed to be more inclusive. The original draft didn't distinguish between rectifying columns and coffey columns so it was flawed to begin with -- it was unclear and could have been interpreted as excluding stills owned by the bill's commercial sponsor Dry Fly. Typical legislative sausage making. But enough of that nonsense.

The hints dropped about this new type of still got me started thinking about stills using flash boilers. They are used in some oil refineries (but not at the ones I worked at.) And I've never seen any reference to flash boilers being used for distilling spirits. So the old examples above were interesting because both use a form of flash vaporization. The only thing comes close is the temperature controlled rectifier in the Pistorius still, but it's more of a plate and it's not used as a boiler.

Here's another draft for a prototype that looks much more like a traditional stripping column, but uses an electrical heating jacket. I just ordered 1KW of rope heaters for $10 from sciplus.com So I'm going to be wrecking some perfectly good tubing making this not work a couple of times.

The fat column is packed and wrapped with a heater. 1KW around 1" tube is about two feet long. I've sacrificed efficiency for safety by running the product condenser on a separate water circuit, so interruptions of the feed won't cause failure of condensation. The liquid trap at the bottom will direct all vapor up towards the outlet.

An efficient condenser on the water outlet at the bottom will recover sufficient heat to warm up the feed, but not enough to cause vapor lock. Scratch calculations of steady-state heat transfer suggest feed will enter the column at about 50C.

The downtube inside the column will boost the temp of the feed to near 78C, so the liquid / vapor line will be near the end of the down tube. It should be thermally stable, though if the column is hot before feed enters, there will be an upset if the initial feed is too fast -- that should settle down as the column equilibrates. There will be some concentration of tails as vapor in the lower third of the column, but if this is a problem cycling the heater should cause the column to dump and purge the tails (will some loss of product if the waste outlet drops below 80C)

Looks like this:

stripper2.gif (5.23 KiB) Viewed 5426 times

Comments, suggestions, raspberrys?

[snuffy]

*/ edit: after looking around, I couldn't find any instances of heating the column, all used a boiler at the bottom and allowed vapor to transfer the heat up the column. endedit*/ I just did some tortuous calculations on the power density needed to get 20 in/sec vapor velocity at the top with a 1" column: 4KW

It looks like roughly 85% of the power goes into bringing the wash up to vaporization temp and 15% goes into latent heat of vaporization. For 1KW heater, that means something like a 5/8" column for approx 1KW. Or 3/4" and use a 1/2" down tube. This is too small to make packing around the down tube worth while - too much wall / volume. So I'm going with 1" column and 3/8" downtube and boosting the power to 2KW. Flow rate is going to be in the range of 5-15 ml / sec for the feed.

It will have adjustable reflux by varying the insertion depth of the downtube, so I can experiment with different settings.

One condenser coil wound and the foot done. Most of the head is complete and I just soldered the joints on the column. I think a couple of more weeks will do it. The longest delay is going to be getting the overheat control built. I'm going with a quick and dirty switching circuit, but it would be nice to get proportional heating control. Initially, the control will be regulating the feed.

[manu de hanoi]

perhaps a search on the continuous columns already built on this forum will bring you some valuable information related to sizing. If you want to save yourself some trial and error that is....

[snuffy]