Has Anyone Tested Shorter But Larger Diameter Packed Column?

[Salt Must Flow]

I know people say that you go tall for purity and larger diameter for higher take-off rate. People also say that the rule of thumb is the height of the column should be approx 20x the column diameter.

I'm curious for instance if anyone has tested a 4" column with 3' of packing (sized for a 3" column) rather than the traditional 3" column with 5'-6' of appropriately sized packing?

A 4" column has twice the volume of a 3" column. So a 4" column approx half the height contains the same volume of packing as a 3" column that's twice the height. I'd like to test filling a 3' tall segment of 4" column filled with packing that's sized for a 3" column, see if it will operate at the same power input and same take-off rate as a 3" column that's 5'-6' tall.

What would be the potential benefit of doing such a thing? A lot of people just don't have the ceiling height for a 3" reflux column. In the future, if additional height becomes available, it would be very economical and easy to add another 3' segment of column and up-size the packing to upgrade to a 4" reflux column with optimal 6' of packing.

I set my 4" VM up on top of a 15.5 keg boiler with a 3' tall column and the total height is 7' 6". That's relatively short and most any situation could accommodate this height.

[Bolverk]

This is the premises of a creole column. Although, they operated with bubble caps that had much better turn down ratios so they could be operated at lower vapor velocity than say a column with perfs or packing.

The theory is sound as long as the plates/packing can deal with the lower vapor velocity.

[Salt Must Flow]

This is the premises of a creole column. Although, they operated with bubble caps that had much better turn down ratios so they could be operated at lower vapor velocity than say a column with perfs or packing.

The theory is sound as long as the plates/packing can deal with the lower vapor velocity.

You are absolutely correct, the rising vapor would be half the speed in comparison to a column that has half of the cross-sectional area. The test would use the same power input, the same size packing, same amount of (or more) packing and the same take-off rate of a 3" column.

[Yummyrum]

A few years ago I had convinced myself that if a 2” 1m packed column running 2kw could make 95% @1 l/h then if I had 4 identical boilers and reflux stills , I would be producing 4l/h and using 8kw .

So I figured , if I had one boiler with 8kw element and replaced the 4 x 2” columns with one 4” then in theory it would have the same vapour speed as each single 2” would .I theorised that I should be able to achieve the same 95% at 4l/h

Well it didn’t and I could get barely much more than about 92% regardless of what packing I used . In the end , I doubled the height and got my 95% .

I can’t explain why , but the 20x diameter seems to be a thing .

The only thing I can think of was that I was using gas and maybe I wasn’t pushing as much power as I thought

[Salt Must Flow]

A few years ago I had convinced myself that if a 2” 1m packed column running 2kw could make 95% @1 l/h then if I had 4 identical boilers and reflux stills , I would be producing 4l/h and using 8kw .

So I figured , if I had one boiler with 8kw element and replaced the 4 x 2” columns with one 4” then in theory it would have the same vapour speed as each single 2” would .I theorised that I should be able to achieve the same 95% at 4l/h

Well it didn’t and I could get barely much more than about 92% regardless of what packing I used . In the end , I doubled the height and got my 95% .

I can’t explain why , but the 20x diameter seems to be a thing .

The only thing I can think of was that I was using gas and maybe I wasn’t pushing as much power as I thought

Well I should have a couple of sugar washes stripped by this weekend. After that I will have plenty of low wines and time to do a lot of different tests that I have in mind. This will be one of them.

Regardless of how it turns out, I think it will be interesting.

[Steve Broady]

I have nothing useful to add, but I will be following with interest. I’ve wondered about the 20x ration thing for a while. I accept it as having been empirically derived, but I don’t understand why it works. It would be very interesting to have a better understanding of what is actually going on at each point in the column, to better understand what various changes in height, input, packing, etc make to it.

[LWTCS]

A few years ago I had convinced myself that if a 2” 1m packed column running 2kw could make 95% @1 l/h then if I had 4 identical boilers and reflux stills , I would be producing 4l/h and using 8kw .

So I figured , if I had one boiler with 8kw element and replaced the 4 x 2” columns with one 4” then in theory it would have the same vapour speed as each single 2” would .I theorised that I should be able to achieve the same 95% at 4l/h

Well it didn’t and I could get barely much more than about 92% regardless of what packing I used . In the end , I doubled the height and got my 95% .

I can’t explain why , but the 20x diameter seems to be a thing .

The only thing I can think of was that I was using gas and maybe I wasn’t pushing as much power as I thought

Well I should have a couple of sugar washes stripped by this weekend. After that I will have plenty of low wines and time to do a lot of different tests that I have in mind. This will be one of them.

Regardless of how it turns out, I think it will be interesting.

I theorize that the 2" diameter is a more favorable environment (at the hobby scale) to exploit temp gradient?

[Salt Must Flow]

A few years ago I had convinced myself that if a 2” 1m packed column running 2kw could make 95% @1 l/h then if I had 4 identical boilers and reflux stills , I would be producing 4l/h and using 8kw .

So I figured , if I had one boiler with 8kw element and replaced the 4 x 2” columns with one 4” then in theory it would have the same vapour speed as each single 2” would .I theorised that I should be able to achieve the same 95% at 4l/h

Well it didn’t and I could get barely much more than about 92% regardless of what packing I used . In the end , I doubled the height and got my 95% .

I can’t explain why , but the 20x diameter seems to be a thing .

The only thing I can think of was that I was using gas and maybe I wasn’t pushing as much power as I thought

Well I should have a couple of sugar washes stripped by this weekend. After that I will have plenty of low wines and time to do a lot of different tests that I have in mind. This will be one of them.

Regardless of how it turns out, I think it will be interesting.

I theorize that the 2" diameter is a more favorable environment (at the hobby scale) to exploit temp gradient?

Temp gradient? Are you referring to the vapor temp gradient incrementally up the column? If so, I'm not sure.

[LWTCS]

That's what I'm thinking salt. Just thinking out loud is all.

But the narrow diameter reminds me of how well a more narrow cylindical vessel allows for more precise stratification when racking/clarifying.
The more narrow diameter does a better job of piling / stacking within the context of hobby sized volumes.

So in full reflux the lighter constituents must occupy the highest available space. Takes less effort to fully enrich the more narrow diameter is how I see it .

[airhill]

I have a feeling that the 20x was just a general thing to tell newbies, at the time packing was usually scubbies down to marbles.
You can shorten the column and get the same purity if you use a different packing eg spp.
As far as the diameter of the column is concerned there was a long and sometimes quite heated discussion on this about 14 years ago. From memory and simplifying this there became a problem with preferred pathing in columns of 4" width and greater and so the ability to hold the same HETP was reduced.

[OtisT]

I don’t have a theory on why, but my practical experience is the same as yummyrum’s. My 4” column packed 52” will only get me 92-93%. I can get 95% out of 36” of 2”.

[quadra]

I am interested to see your results as well. I read a paper discussing the impact of plate height based on mass transfer and how the taller the column the taller the plate needed to be. That had me wondering if a shorter packed column run at increased pressure could decrease the theoretical plate height and increase efficiency... that led me to another paper that not surprisingly was using SPP in a pressurized column and claiming better efficiency and seperation when running in a partially (...80% rings a bell ) flooded state. The idea of exploring or promoting that type of system for hobby distilling did not sit well with me but given the results the SPP promoters claim with flooded state columns allowing height reduction I am guessing that might be a path forward. In terms of function, the reduced free air column volume of a packed column agrees with your thinking Larry. The additional thermal mass/ stability is possibly the reason for the claim of improved separation...?

[MooseMan]

That's what I'm thinking salt. Just thinking out loud is all.

But the narrow diameter reminds me of how well a more narrow cylindical vessel allows for more precise stratification when racking/clarifying.
The more narrow diameter does a better job of piling / stacking within the context of hobby sized volumes.

So in full reflux the lighter constituents must occupy the highest available space. Takes less effort to fully enrich the more narrow diameter is how I see it .

My brain also makes sense of it in the same way as you Larry, but I suspect there are several factors at play and they all contribute.
It's a complex thing, that's for sure.

[Swedish Pride]

Not that abv% really means cleaner taste.

But I was able to get 96 from my 2.5 ft scrubbie packed section.
It does in fairness sit on top of 3 plates and I ran it fairly slow.

Was coming out just over 10c, temp corrected that is 96

[Salt Must Flow]

This is an interesting topic and it's not a long one either. Height for purity, diameter for speed, and HETP Calcs.

[OtisT]

Not that abv% really means cleaner taste.

But I was able to get 96 from my 2.5 ft scrubbie packed section.
It does in fairness sit on top of 3 plates and I ran it fairly slow.

IMG20241012155213~2.jpgIMG20241012171346.jpg
Was coming out just over 10c, temp corrected that is 96

I agree. High ABV does necessarily mean a clean product. You need to start with a clean tasting ferment.

Those plates below obviously have an impact on the output ABV by feeding a higher purity (read as ABV) product to the base of the packed section. Same thing as increasing the height of the packed section.

I did a few packed column runs with thermometers placed every 6” or so from bottom to top. I saw that the temp drop from the first to the second then third sensor was significant, and the delta in temp between each subsequent sensor was smaller and smaller. For me those were very informative tests. Showing the temp gradient helped me understand why so much extra height is needed to go from 93 to 95%, and I also saw how long it took for the column to reach equilibrium or react to power changes with a high mass packing and it was not as fast as I originally thought.

[LWTCS]

Just a side thought on the "purity" word.

I feel using the purity word to characterize high abv can be deceiving or not a completely accurate descriptor.

As mentioned by Odin and others in the link / other thread posted here, the shorter column needs a higher reflux ratio in order to maintain proof during collection, as opposed to a column with more plates that doesn't need the same type of RR to maintain proof.

The result is that the shorter column run with a higher RR will drag up more flavor constituents that get trapped on actual plates. The condition in packed columns is not as bad but contaminating flavors still get drug up with the higher reflux ratios needed to maintain proof.

95 with 10 or 15 plates is not the same level of cleanliness as 95 with 30 plates.

[Hügelwilli]

Only regarding this sentence:

The result is that the shorter column run with a higher RR will drag up more flavor constituents that get trapped on actual plates.

Sure? Is it from experience or do you have something more to read?

[LWTCS]

Only regarding this sentence:

The result is that the shorter column run with a higher RR will drag up more flavor constituents that get trapped on actual plates.

Sure? Is it from experience or do you have something more to read?

It is my own experience that high abv rendered from a short plated column is not at all neutral. Far from it.
This is true for hobby sized equipment as well as all sizes of commercial equipment.

As far as something to read ? Well, this forum and perhaps one or two others is littered with examples of short plated columns not making neutral spirits including the link on this very thread.

Beyond that, I'm not sure I understand your question?

[Hügelwilli]

I more hoped that you perhaps have a scientific paper.
Obviously you need a different RR to get the same abv, but then if you get it, in theory the congener composition should be the same of the distillates of both columns.
The forum may be littered with examples, that you need an awful high RR with a short column, but not that this hard earned spirit with 95%abv is less neutral than the easy one with the long column.
And in practice it is hard to compare. Firstly because those two spirits can differ by 0.1%abv and you will not be able to measure it. And secondly because the different energy you have to use for columns with different diameters. One column will have more entrainement than the other.

[shadylane]

My 2 cents worth on refluxing a Rum wash with a 4" x 18" packed column compared to 3 bubble cap plates.
Packing could handle more power and had a much wider range of operation.
The bubble caps were slower but kept the tails in the boiler and out of the bottle better.

[Northsouth]

I made 2 similar attempts on a single malt whisky, both with malted barley, Both stripped down to around 35 abv total, then spirit run. First time, it was with 2 bubble plates 3", and a 3" piece with about 2" height of cut copper tubes as packing, basically just some scrap, I wanted a little more copper in the path and a little more reflux in lack of owning a 3rd plate.
Second time, same, but on top of the scrap packing, I added 15-20 cm (3-4") of SPP of 3.5x4mm mixes with 6x7 mm.

The second batch is really lacking taste, whereas the first one was very tasty, in my view.

From my now many vodka and also neutral runs (2 very different things), my conviction is that temperature gradient is important. I don't even understand the meaning of insulating the column, even less if you read around here and find the most dominant recommendation is to insulate at least the upper part. If anything, you want to insulate the lower part, no? That way you get a hotter lower part and a colder higher part, hence more temperature gradient.

Think of it as 3 plates. If you put a thermometer on top of the boiler, one on each plate and on top of the column, it will go from hot to less hot. The lower plate will be hot, allowing heavy and light vapor to rise. At the next level, you want it a tiny bit colder so the heaviest fall down (after interacting with reflux liquid) while the ethanol goes up. Next level, same with further purification. I think that is the main mechanism and the "magic" is in how the reflux interacts with rising vapor. This "magic" is also where the confusion and discussion about SPP, lava rock, scrubbers, etc. comes in. Just my way of trying to wrap my head around it since there are so many contradicting views here in this forum about it.

[Yummyrum]

Think of it as 3 plates. If you put a thermometer on top of the boiler, one on each plate and on top of the column, it will go from hot to less hot. The lower plate will be hot, allowing heavy and light vapor to rise. At the next level, you want it a tiny bit colder so the heaviest fall down (after interacting with reflux liquid) while the ethanol goes up. Next level, same with further purification. I think that is the main mechanism and the "magic" is in how the reflux interacts with rising vapor.

This is correct Northsouth , except that each plate will be a bit cooler than the one below because of the phase change that happened on it rather than it being cooler for this to occur .

If you think about it , they start at room temp and get to the temp they are because of what is happening on them .

This diagram shows things . The Blue line is the temp of the mixture , the Red line is the vapour leaving it .

You can see if you start with 10% in the boiler , it will boil at 93°C
The liquid on the first plate is 82.5°C .
The Liquid on the second plate is 79.6°C
The Liquid on the third plate is 79°C

This all occurs naturally .

[airhill]

Northsouth you should insulate your boiler and your reflux column else you are just wasting energy.

[MooseMan]

I made 2 similar attempts on a single malt whisky, both with malted barley, Both stripped down to around 35 abv total, then spirit run. First time, it was with 2 bubble plates 3", and a 3" piece with about 2" height of cut copper tubes as packing, basically just some scrap, I wanted a little more copper in the path and a little more reflux in lack of owning a 3rd plate.
Second time, same, but on top of the scrap packing, I added 15-20 cm (3-4") of SPP of 3.5x4mm mixes with 6x7 mm.

The second batch is really lacking taste, whereas the first one was very tasty, in my view.

From my now many vodka and also neutral runs (2 very different things), my conviction is that temperature gradient is important. I don't even understand the meaning of insulating the column, even less if you read around here and find the most dominant recommendation is to insulate at least the upper part. If anything, you want to insulate the lower part, no? That way you get a hotter lower part and a colder higher part, hence more temperature gradient.

Think of it as 3 plates. If you put a thermometer on top of the boiler, one on each plate and on top of the column, it will go from hot to less hot. The lower plate will be hot, allowing heavy and light vapor to rise. At the next level, you want it a tiny bit colder so the heaviest fall down (after interacting with reflux liquid) while the ethanol goes up. Next level, same with further purification. I think that is the main mechanism and the "magic" is in how the reflux interacts with rising vapor. This "magic" is also where the confusion and discussion about SPP, lava rock, scrubbers, etc. comes in. Just my way of trying to wrap my head around it since there are so many contradicting views here in this forum about it.

Insulating the column provides thermal stability.
There is a very delicate balance of gradients, driven by phase changes, going on in the column, that you have created by carefully controlling heat input, cooling water, packing and column dimensions and so on.

If you disturb that by allowing (Uncontrolled and random) environmental conditions to interact with it, you have a much less stable and much less efficient column.

[Yummyrum]

I agree that insulating a packed column is imperative to keep those natural temp gradients stable .

Also a am guilty of discussing plates here , but this topic is about packed columns not plates .

[Stump Lake]

I'm following this thread. This is a subject that I have been interested in for a while. Unfortunately I'm to new to add any sound information. It's cool to see the heavy hitters of this fourm expressing their opinions.

[haggy]

Salt Must Flow,

Your premise is:

You are absolutely correct, the rising vapor would be half the speed in comparison to a column that has half of the cross-sectional area. The test would use the same power input, the same size packing, same amount of (or more) packing and the same take-off rate of a 3" column.

So, with the same power input and the same take-off rate as a 3", the 4" would have about the same reflux ratio as the 3".

The product abv would then mainly be determined by the number of theoretical plates used, which is determined by the packing length and the HETP.

>>>>>> Theo Plates = Packing Length / HETP

Then the main difference in the 4" vs 3" is the length of the packing ( about 1/2 the length ) and the slower vapor speed up the column.

So, the 4" with the same length packing and about the same HETP ( or maybe a little lower ) would have fewer theo plates and give a lower abv product.

3" >> 60" length / 4 HETP = 15 theo plates >> ok product abv 95+%

vs

4" >> 30" length / 4 HETP = 7.5 theo plates >> likely, product abv about 92%

4" >> 30" length / 3 HETP = 10 theo plates >> maybe, product abv about 94%

4" >> 30" length / 2 HETP = 15 theo plates >> not likely, product abv about 95+%

So, to have the same number of theo plates and product abv, the HETP of the 4" would have to be 1/2 that of the 3". And that is not likely to happen.

I have seen a reference where the HETP in chromatography did decrease as the vapor velocity decreased, the Golay Equation. But the effect was not great, maybe HETP lower by 25% to 3, not 50% to 2 in this case. The reference is:

https://h-o-m-e.org/hetp/


And,
The reference also states that the packing length will influence the HETP. The HETP might increase as the packing length decreases. Bad. But the standard deviation ( see below ) might be less. Overall, it is likely bad for the 4" HETP with half the length of packing vs the 3".

From the reference:

A formula given for HETP is HETP = σ 2/L, where σ is the standard deviation of the compound’s retention time and L is the length of the column. The standard deviation measures the degree of variation in retention times for a particular compound. A smaller standard deviation indicates a more efficient separation process, as the compound is consistently retained and released at the same point in the column.

In summary, it is probably all about the HETP and the number of theo plates. The HETP of the 4" at similar operating set points and packing and half the packing length of the 3" is most likely the same or slightly less than the 3", so the 4" theo plates are lower and 4" product abv is lower.

Haggy

[shadylane]

.. the rising vapor would be half the speed in comparison to a column that has half of the cross-sectional area. The test would use the same power input, the same size packing, same amount of (or more) packing and the same take-off rate of a 3" column.

So, with the same power input and the same take-off rate as a 3", the 4" would have about the same reflux ratio as the 3".

Why compare the 4" at the same power as the 3?
Wouldn't it be a better comparison to run the 4" at twice the power of the 3" so both have the same vapor speed?

[haggy]

Shady,

Read the first few posts. I was quoting the premises of Salt Must Flow - to compare the two columns at the same power input.

Yes, it would be better for the 4" to run at higher power and get more reflux flow. But 2 X the 3" watts on a 4" might be too high a power and drag tails, as you and others have stated many times.

And if there are not enough theo plate, the 4" probably would not reach 95+% even with more reflux flow.

Haggy