TOTM, 3/07: Column Stills

[jeff]

The first installment of Straightbourbon.com's "Topic of the Month" series will deal with column still design and function. Though the theory behind distillation is relatively simple to understand, in practice it is a complex process to ensure quality, safety and maximum yield. Let's use this month to discuss how column stills work, how they can be manipulated, their general construction and maintenance. So, start typing and tell us what you know about column stills.

:Clever:Sound off:Clever:

[TNbourbon]

...So, start typing and tell us what you know about column stills.

:Clever:Sound off:Clever:

They're built in a column (I think:skep: ). Okay, take it away, guys (and gals).

[Gillman]

On the plate of each column, a distillation occurs and reoccurs as the distillate condenses and falls back through the column. Steam pushing up through holes in the plates vaporises the alcohol. The more plates and higher the column, the more complete the separation of the ethanol will be. In effect, each chamber of a column still operates like a pot still. Originally the column stills had only a few plates and were primitive, e.g. the three-chamber beer still (still used to distill Armagnac and some other foreign liquors). Later they became sophisticated and larger. Chemical engineers work out the computations to design column stills and it is a question of how much throughput you want at what energy cost at what purity.

These calculations can be very complex.

Generally one run through a modern "beer" still is not enough to cleanse the liquor of undesirable congeners (impurities which can flavor the liquor). So they run it through another still, a thumper still or doubler still in U.S. practice. This increases proof by about 5-10 points and achieves the required congener level.

Column stills also can be accompanied by a rectification tower in which purity is increased through condensing the vaporised liquid from the column still and drawing off (or "purging") the undesirable secondary elements such as fusels and aldehydes.

An extraction still sometimes is used to achieve this.

Remember this: most of the undesirable congeners come off above the boiling point of alcohol. So the more often you boil the alcohol off first, the more the bad fusels stay back with the water. However, some bad fusels including methanol lift off before the alcohol. So you must separate that part out and re-distill it. That is the foreshots, the former are the aftershots. You want broadly a "middle cut", where the ethanol is.

You could link 20 pot stills to achieve what a column still does but the loss of ethanol and enormous energy output make that inefficient.

In a different life I want to be a chemical engineer.

Gary

[T47]

Are all Bourbons Column Still products? I recall from some of the Irish whiskey threads the topic of Column vs. Pot Still comes up again and again.

When I tried to find out some information searching the net, it seemed that the prevailing opinion that I could find was that the Column still was created as a less maintenance intensive process and that the Pot Still offered a fuller flavored product? I recall there was some educated disagreement with that theory.

Is there some pairing of Column vs. Pot Still bottles that one could try that might best reflect the process's influence on the final product or in reality does the process not effect the end product significantly?

I don't want to get off topic and start a Pot vs. Column Still debate, just curious about the influence of the process in general.

[TNbourbon]

...Is there some pairing of Column vs. Pot Still bottles that one could try that might best reflect the process's influence on the final product..?..

Woodford Reserve Four Grain (100% pot still) and early-batch standard Woodford Reserve (100% column still, as Old Forester 'honey barrels').

[jeff]

You want broadly a "middle cut", where the ethanol is.

I understand why you would want to remove the "heads" and "tails", but how much is that as a percent of the whole run? And I am really curious as to how this is accomplished when the still is being run continuously?

[Gillman]

I am not sure about the percentages of material held back as heads and tails. But heads and tails there are definitely, even in a column still. It does not really run continuously. When mash goes through the first vapours condensed are re-distilled as being too high in methanol. Ditto for the last vapours before shut-down and cleaning. The cycle is longer than for pot stills but column stills are cleaned out too.

Gary

[barturtle]

I understand why you would want to remove the "heads" and "tails", but how much is that as a percent of the whole run? And I am really curious as to how this is accomplished when the still is being run continuously?

I think what Gary is saying (and it seems to work in my head, though that is no guarantee of accuracy) that by distilling to the proper temperature, the tails are always left behind (being too heavy to be distilled at this low temp). Then by running it through the doubler at a much lower temperatures the heads are taken off (i.e. they don't actually redistil the whiskey, but run the still-a doubler is a still, basically-at just high enough of a temp to lift off the heads.)

I had never thought about how they take the heads and tails off in a continuous process, but this seems to be an appropriate way to get it done...whether this is how it is actually achieved, someone else will have to confirm.

[jeff]

Here is a link

to a PDF file on the distillation and denaturing of alcohol. On Page 63 begins a discussion of modern distilling apparatus. A few pages in you will find several diagrams of distilling plates and column still design. Page 73 has an interior diagram that is helpful.

The page numbers refer to those printed on the document, not the actual PDF page number

[jeff]

Gary made me think about something I thought was relatively obvious, but now I'm not sure:

Is a continuous still really continuous? I assumed this thing could run for hours on end with new beer being added all along. Is this the case, or is it simply run in batches and cleaned in between? If so, how much is a batch and how long does it take to run?

[ILLfarmboy]

I always thought a continuous column still was only continuous through each Bach of fermented mash. In other words with a pot still you have to do a stripping run and then a spirit run then possibly a third run ala Irish whiskey cleaning out the still between runs. But not so with a column still.

Also aren't column stills easer to "distill on the grain" as opposed to a wash (is that the right word?)

[cowdery]

The still is continuous in the sense that, technically, or I should say, in theory, it can run indefinitely. As a practical matter, they run it for a few days, the better part of a week, then shut it down and start over. Basically, you need to clean everything before you start up again. It's also a personnel thing. Remember, it's continuous, which means 24 hours a day. It's quite a logistics job to keep the whole process going.

Still throughput is essentially a function of column diameter.

The beer that enters the still contains grain solids. There is no wort or wash, as there is with malt whiskey.

I think it's inevitable that a discussion of column stills will include comparisons with pot stills and comparisons between American whiskey and scotch whiskey.

Claims for the superiority of pot stills over column still primarily stem from a misunderstanding of how the column still is used in American whiskey production, and how it is used in combination with the doubler. People whose framework for understanding whiskey-making is scotch assume American straight whiskey is analogous to Scottish grain whiskey. The two processes actually differ more than most people making the comparison realize.

The column still is simply an improved pot still, in that it is more controllable and more versatile. A column still can strip spirit out of the mash very efficiently and in Scotland they're used that way. In the United States, their output parameters are closer to those of Scottish pot stills.

Most, but not all, bourbon stills have rectification plates above the stripping section.

[jeff]

Most, but not all, bourbon stills have rectification plates above the stripping section.

So then, do the heads effectively get removed at this point? What I'm really wanting to know is, is it a rather automatic process by design, or are the operators "manually" removing parts of the run, so to speak?

[Gillman]

Jeff, in the American practice, as far as I know, when a still is started up again the first condensate may be rejected if the methanol content is too high (it vaporises at about 150 degrees F.). Ditto for when it is stopped, the tails may be rejected if too high in undesireable elements.

However during the multi-day running process, no purges (sometimes called draws) are done in the American system. I was told that specifically by a distiller during one of the SB distillery tours. In other words, vapours on a specific plate in the column or rectification tower are not removed to exclude from the distillate the elements which vaporised at the temperature represented on that plate. This is because it is not traditional to do so and in this respect one can see continuity (ahem) from the pot-stilling days.

Extra plates on top or a separate rectifying tower are all simply extensions of a column still. They just keep the process going.

In Canadian practice, purges are done at different plate levels to ensure that certain congeners are not in the final product. This is also done through a stage called extraction distillation (diluting the low wines to change the volatility of certain congeners and then heating to draw them off).

I have previously indicated that at www.scocia.com, extracts of a book on distilling are included (under My Research). The Canadian and American chapters, but also those on vodka and gin production, exhaustively explain how column distilling works.

Gary

[jeff]

I guess what I don't understand is, if you're trying to remove methanol, wouldn't you have to be continuously removing it as new beer enters the still? If so, what is the exact mechanism for removing it? Is there a pipe somewhere that somehow diverts the methanol while allowing the ethyl alcohol to continue up?

[Gillman]

It is a matter of concentration. True, a given amount of beer has x% methanol. But with the still running continuously at a given temperature, its concentration is minimal. But if you start up a still and want to barrel the first runnings, they may be too high in methanol and sometimes they have a characteristic bluish or muddy look to them. I think this can happen at the end of the cycle too as the still is wound down (or if it goes too hot, but it is adjusted not to to ensure proof does not exceed 160)). As to how you remove them, they are a condensate and either thrown out (as Michter's used to do apparently) or redistilled (more commonly). But I am fairly sure during a multi-day continuous run it is not necessary to remove parts of the runs for this purpose.

Maybe ethochem, the engineer at Trace, can comment.

Gary

[Gillman]

It is a matter of concentration. True, a given amount of beer has x% methanol. But with the still running continuously at a given temperature, its concentration is minimal. But if you start up a still and want to barrel the first runnings, they may be too high in methanol and sometimes they have a characteristic bluish or muddy look to them. I think this can happen at the end of the cycle too as the still is wound down. As to how you remove them, they are a condensate and either thrown out (as Michter's used to do apparently) or redistilled (more commonly). But I am fairly sure during a multi-day continuous run it is not necessary to remove parts of the runs for this purpose.

By the way redistillation does not build up undesireable amounts of these congeners. You would think it would, but it does not, somehow they are broken down or eliminated by re-distillation.

Maybe ethochem, the engimeer at Trace, can comment.

Gary

[jeff]

Thanks Gary, the concentration angle makes sense to me. So when a still is first fired up, there is a higher concentration of methanol that must be removed, but as the still gets going, the methanol concentration drops to an acceptable level, do I have that right?

[TNbourbon]

Implied all along here, Gary, I think, but not explicitly stated yet (I don't think) is that during startup and stoppage, there are periods of time when the still is operating at a temperature below optimum, and thus 'throwing off' distillate with those higher methanol concentrations.

Is that right?

[doubleblank]

Going on my hazy recollections from several distillery tours, I recall they said that they "burp" out some of the undesireables (methanol, acetones, etc) from the doubler into the atmosphere from time to time. These would tend to remain in the vapor phase once steady state operation is obtained in the column. And from a material balance standpoint ("what goes in has to come out"), there would need to be the occasional removal of underireables between stoppages. Otherwise, my guess is that the concentrations of methanol, etc would increase such that some would begin to appear in the distillate.

Randy

[cowdery]

The traditional idea of "heads and tails" is largely a function of temperature and is another advantage of the column still.

What a pot still does in time, a column still does in space.

At Woodford, you can stand at the box and taste the distillate, and its flavor changes considerably every few minutes, because the still is pulling off different compounds at different times. The output of the column still is consistent in flavor though you would have the same effect if you did pull off samples at each plate.

There are high boilers and low boilers among the things they want out. Some are "burped" out into the atmosphere in some processes. I know Four Roses does this.

[barturtle]

But if you start up a still and want to barrel the first runnings, they may be too high in methanol and sometimes they have a characteristic bluish or muddy look to them. I think this can happen at the end of the cycle too as the still is wound down (or if it goes too hot, but it is adjusted not to to ensure proof does not exceed 160)).

Gary

Actually, I think if the still gets too hot, then the proof will be lower as you are getting more water vapor going out of the still, a lower temperature would yield less water vapor and therefore higher proof-within limits, of course.

[Gillman]

True, Tim, all I meant was, too high and you risk too many aldehydes and acids coming over with the water, too low and too much methanol. The high end is controlled partly by the 160 proof limit; the low end by separating any portion of the distillate which is too methanol-rich. I believe the methanol-rich part comes with a start-up of a (column) still, after that it is absorbed sufficiently by the middle cut.

Gary

[cowdery]

In a typical pot distilling operation (i.e., Scotland), the first distillation is strictly about stripping the alcohol out. It is in the second distillation that a body cut is taken. In a column still, the plates below the beer intake do the stripping while the plates above begin to remove particularly undesirable fusel oils, especially the low boilers. The high boilers can be left behind in the doubler by controlling the temperature. In a pot still operation, the heads are mostly low boilers while the tails are mostly high boilers. Basically, though, the heads and tails concept isn't applicable to column stills so the only way to talk about it is by analogy.

In Scotland, column stills are used to make grain whiskey, mostly for blending, and the distillation proof is around 189, i.e., 94.5 % ABV. At that level you don't have to worry about specific congners because you're basically stripping out all of them. The spirit is nearly neutral. In the USA, where the top distillation proof is 160 (80% ABV) and often lower, down to as little as 110 (55% ABV), it's a very different matter.

[cowdery]

Here is how the venting thing works.

Some of the undesirable congeners are high boilers, but they boil higher than the good congeners by just a couple of degrees difference. By condensing the vapors off the column still back into a liquid, then cooling the liquid further, sub-cooling it, the last vapors to collapse (i.e., condense) are the bad high boilers. At that point, just before those bad congeners condense, the condenser is vented. The bad high boilers go out and are not condensed. Then the condensed liquid can be boiled at a lower temperature in the doubler, too low for the remaining bad high boilers, which are left behind in the liquid.

This was explained to me by John Rea at Four Roses. They use a venting condenser. Some distilleries do not.