A theory of the reflux distillation column.

[Tropic]

A theory of the reflux distillation column.

A longer column produces greater alcohol concentration because it provides a greater number of distillations. - WRONG (well, sort of wrong).
A longer column produces a greater alcohol concentration because it provides a longer time in the column for distillation of the reflux.
Distillation works on the fact that boiling a liquid with a mixture of chemicals creates a gas with a greater concentration of volatile liquids than the original liquid as per the ethanol vapour concentration (red) produced by a wash at its boiling point (blue) graph:

The boiler.

Boiling (i.e. converting liquid to gas) in the boiler creates the first distillation. With 10% ethanol in the wash, the boiler (and bottom of the column) will be at 93°C producing a gas of 53% ethanol. A liquid of 53% ethanol boils at around 82°C. Inversely, that gas will condense (i.e. convert gas to liquid) when it drops below 82°C. Note that alcohol vapour (droplet suspended in air) can go below that point (like steam in the shower at <100°C) but will condense almost immediately because of the massive surface area supplied by the column packing (like steam on your bathroom mirror).

Rising Gas.

The newly created gas from the boiler creates pressure which pushes the gas above it up the column. The gas travels up the column until it finds a surface below its boiling point to condense on. With an insulated column, the first time it encounters this is near the top of the column where it encounters refluxed liquid and packing cooled by that liquid. Some of the gas will condense but most will pass to the head where it encounters the condenser at well below its boiling point so is all converted to liquid and most is refluxed back down the column.

Falling Liquid

As this reflux liquid falls through the packing, the more volatile chemicals boil out, leaving lower concentrations of volatiles in the liquid, increasing the liquid’s boiling point. As it falls it loses more volatiles and eventually the boiling point of the liquid approaches the temperature of the column, it doesn’t boil and falls through to the boiler.

Rising Gas, Falling Liquid.

As liquid falls through the packing, the more volatile chemicals boil first. If the gas at the top of the column is 75% ethanol, the gas created from the initial boiling of condensate at that concentration will be 85%. This will have the effect of increasing the ethanol percentage in the gas traveling up the column to the condenser. Another way of looking at this is that the gases rising from below are diluting the gas formed from the falling condensate.
Lower in the column, most of the volatiles have come out of the liquid over the upper part of the column leaving a liquid with a higher concentration of lower volatiles (aka tails). Much of this liquid will fall back into the boiler taking much of the lower volatiles with it. But with the column temperature at a similar temperature to the boiler, the condensate will continue to boil as it falls through the column, releasing a higher concentration of lower boiling point volatiles in the lower column. These lower volatile gases are pushed back up the column, much of which will reach the condenser.
At the start of the run with an insulated column, the tails are easier to keep away because the temperature of the system is low which prevents much of the higher volatile chemicals from boiling out of the falling condensate. Towards the end of the run, the concentration of heavier volatiles in the condensate is much higher because the wash concentration is higher since lighter volatiles (product) have been removed. The still temperature is much higher which allows heavier volatiles more time at temperature to come out starting nearer to the top as it travels down the column.



So what does this all mean for designing and running a hobby still?

Column temperature gradient – a hobby version of a fractionating still.

If you could dynamically cool the gas (without condensing it) at every place in the column to the boiling point of the gas at that place it would make it much harder for the lower volatiles to boil out of the falling condensate, create passive condensation as the gas rises to a cooler area, and virtually eliminate tails. The bottom of the column would be near the temperature of the first distillation (eg 82°C from our 93°C boiler at 10% ethanol wash) then there would be a gradient all the way to the top which would be near the temperature of azeo.
Alternatively, if you were able to actively force the temperature at each place in a column as the perfect equilibrium gradient, independent of what the still is doing, just the passive condensation could stop the heads from rising at all and active refluxing from the condenser would be almost unnecessary.
I think this could also be used to concentrate and remove fores/heads more effectively. If you were to force the temperature gradient so that the top was below azeo (say 76°C) it would force down ethanol by the same method that it eliminates tails.
I’ve already built a small prototype lower column cooler https://homedistiller.org/forum/viewtop ... &start=390 . It uses coil inside the base of the column with recirculated coolant which is cooled by a controllable fanned radiator so that the entry temperature is low enough to cool the gas entering the column but high enough not to condense it. It eliminates tails right down to the very end of a run getting out almost all of the ethanol. Later, I’ll start a new thread on designs based on expanding this theory. I started out years ago with the flawed idea of using a dephlegmator instead of just cooling the gas to its boiling point. There were some very insightful comments in that thread https://homedistiller.org/forum/viewtopic.php?t=68760

Stripping.

The stripping run is usually done by pot still to concentrate the ethanol with no consideration of heads and tails. It would be best to consider eliminating as much of the tails as possible in the stripping run because they are constantly trying to make it to the condenser in the spirit run.

Column dimensions.

There is a point beyond which the length of the column becomes redundant and possibly detrimental. That would be the length at which the gas produced from the condensate at the base of the column (which has had the ethanol stripped in the upper column) has a lower ratio of ethanol to tails than the gas created from the boiler. This could be around the point at which the concentration of ethanol in the condensate equals the concentration in the boiler. Beyond this length, the lower portion of the still is increasing the proportion of tails in the rising gas. Of course, this changes with ethanol concentration in the wash and the reflux ratio.
There is likely an optimal gas flow rate up the column. The thinner the column or more powerful the boiler, the greater that flow velocity. Too slow and the gas will just mix, instead of “pushing” that above up the column. Too fast and the “dirty” gasses from the boiler and lower column will reach the top of the column.

[Smog]

I will certainly follow your experimentations very closely !

[Hügelwilli]

This is interesting, because on another forum there is a discussion currently about one detail of the process:

Boiling (i.e. converting liquid to gas) in the boiler creates the first distillation. With 10% ethanol in the wash, the boiler (and bottom of the column) will be at 93°C producing a gas of 53% ethanol. A liquid of 53% ethanol boils at around 82°C. Inversely, that gas will condense (i.e. convert gas to liquid) when it drops below 82°C.

It seems uncertain, if the reflux really has a lower temperature than the vapor.

For example at The Compleat Distiller from Nixon and McCaw page 109 (If you don't have the book, I can post a link):

The temperature of the reflux will be the same as the temperature of the vapor in the top of the column,

But on page 25 (about methanol distillation):

The top left dot shows that the boiling point of a solution containing X% methanol is TxºC. The vapor produced contains a higher percentage of methanol, because it has a higher vapor pressure than water (shown by the top right dot at TxºC) and this vapor will condense at TyºC.

Here the book claims there are two different temperatures.

Do you have something to read, what proves the two temperatures?

[jward]

Sounds like one needs to distill through a packed column jacketed and cooled like a liebig.

[Tropic]

It seems uncertain, if the reflux really has a lower temperature than the vapor.
For example at The Compleat Distiller from Nixon and McCaw page 109 (If you don't have the book, I can post a link):
The temperature of the reflux will be the same as the temperature of the vapor in the top of the column

I haven't read that but at the top of the column, the reflux will be at the temperature it is cooled to by the condenser. Further down the column, I assume it will then be just slightly lower than the gas temperature since it takes time to heat up.

Sounds like one needs to distill through a packed column jacketed and cooled like a liebig.

Exactly what I am going to try! My current base column cooler does a fantastic job of eliminating tails by cooling the gas to one distillation below the boiler so surely I can get even better results by progressively cooling it to a few distillations.

The initial gas cooling from 93°C in the boiler to 82°C (with 10% ethanol in the wash) will take a bit more than the capability of a jacket so it will be done using my existing base cooling coil. Then the gas will then rise through about 600mm of column surrounded by a cooling jacket with an internal spiral circulation to get the right gradient. I'll throw in a few centering rings in case I overdo the cooling and condense some of the gas. Then the bulk of the column refluxing will be done in a narrow range of just a few degrees.
The coolant will first go through the jacket then through the base coil. I might need to drop the temperature of the coil input a bit to get the job done.

Please excuse the janky drawing but here's the jist of what I'm thinking I could build:

And here's an early prototype of my base column cooler. It'll probably need more cooling power but this currently does the job on very low fan speed.

[Demy]

I follow your experiment, it seems interesting to me. From the images it seems that a radiator in which water flows with the fans is cooling down? I think (based on what I understand) it is important to check the temperature at that point, am I right?

[jward]

I realize I am out of my depth but my intuition says you would want the temperature gradient over a longer portion of the column. Please keep sharing.

[Smog]

I realize I am out of my depth but my intuition...

Yup, intuition leads to experiments.. that leads to new ways of doing things or new finds.

I am also looking forward to this.

[Tropic]

It seems uncertain, if the reflux really has a lower temperature than the vapor.
For example at The Compleat Distiller from Nixon and McCaw page 109 (If you don't have the book, I can post a link):
The temperature of the reflux will be the same as the temperature of the vapor in the top of the column

I haven't read that but at the top of the column, the reflux will be at the temperature it is cooled to by the condenser. Further down the column, I assume it will then be just slightly lower than the gas temperature since it takes time to heat up.

Whoops, here's another in-depth postulation. I suppose this is the reason why this is in "research and theory".
It seems like I will need to modify my thinking a bit. Please read and give me comments.

____________________________________________________________________________________________________

On second thought, I think the reflux temperature MUST be lower than the gas temperature. As we know, a liquid boils at a certain temperature. E.g. a wash of 10% boils at 93°C. Azeo is 97.2% ethanol (and ideally 2.8% water) and boils at 78.2°C. So if you are collecting azeo, you are refluxing azeo which will be 78.2°C as it falls down the column. A normal column without my device is much hotter than this for most of its length. As it falls, the ethanol will be stripped from it and eventually, when it has lost 96.9% of its volume, it will be 10% ethanol (0.3 of the 3.1 remaining) and will boil at the same temperature as the boiler at 82°C. But, I'd say that occurs waaay down the column.

Now, lets say that we have stripped the heads but the boiler has 1% tails compounds (I'm guessing here, I have no idea what a wash really has but this will work with any value). Excluding the water the tails are 9% of the remainder ( 1 divided by 11 (1 part tails + 10 parts ethanol) ). When the wash boils, it concentrates the ethanol but also concentrates all the other compounds lighter than water (aka tails) but probably to a lesser extent. Let's say that the azeo at the top of the column is actually has 97.0% ethanol, 0.2% tails and 2.8% water (again, I'm guessing but let's take this as a very conservative example). That means that as the azeo reflux falls and 96.7% of the ethanol evaporates, we are left with ethanol at 0.3% of the original volume, tails at 0.2% of the original volume and water at 2.6% of the original volume. That gives a ratio of the stripped reflux of 0.3 : 0.2 : 2.8. Divide these by the total (3.3) they leave a reflux of 9% ethanol, 6% tails and 85% water. Excluding the water the tails are 40% of the remainder. That means that by the time the reflux reaches the similar ethanol concentration and temperature as the boiler, it has waaay higher tails concentration and the last thing you want is for this to evaporate. Which is why I postulate in my theory above that beyond a certain length, a longer column can be detrimental to distillation of a neutral spirit.

So, how does my column cooler eliminate tails?
You can effectively set the concentration that the reflux will end up when it exits the cooled section. Looking at the graph above, 85% ethanol solution boils at 79°C. If the surrounding gas is 79°C, the 97% reflux liquid can get to 85% then boil no further.

Starting with 97% ethanol reflux and running the column at:
79°C will leave 85% ethanol in the reflux, re-evaporating 12%.
80°C will leave 75% ethanol in the reflux, re-evaporating 22%.
81°C will leave 65% ethanol in the reflux, re-evaporating 32%.
82°C will leave 55% ethanol in the reflux, re-evaporating 42%.
83°C will leave 47% ethanol in the reflux, re-evaporating 50%.
84°C will leave 39% ethanol in the reflux, re-evaporating 58%.

Now that I look at it, I think I don't want to drop my temperature to 79°C. The only way to make that viable would be to increase the power to about 15,000 watts since only 12% of the reflux would re-evaporate.

I'm currently running my coil to drop the gas temperature to 84°C which actually seems ideal. I think my current coil condenses quite a bit while it cools the gas. I just want it to cool it. Maybe I just need a more efficient cooler. Maybe a shotgun? I think I'll still continue with the jacketed cooler but will be less ambitious with how much I drop the temperature. I'd like to keep this modular so a longer jacket with a more gradual gradient would mean running it across several pipes. I'm not prepared to go to that level quite yet.

Or maybe I'm on the right track by lowering the temperature in the top of the column (for "polishing") but maybe need a longer at the bottom with a higher temperature (for brute force re-evaporating).

Help?

[Hügelwilli]

I don't know if I understand you correctly. In your theory the reflux of something near the azeotrope has a very low abv. Is that right?
But then a LM wouldn't work at all. A LM collects reflux near the azeotrope. So the reflux at the top of the column must be near the azeotrope.

What I write now are not my ideas but more ideas quoted from other distillers:
This is a segment of a column:

%abv(fl)	°C	%abv(vap)
83.8	79.5	89.4
82.8	79.6	89.0
81.7	79.7	88.6
80.6	79.8	88.2
79.5	79.9	87.9
78.4	80.0	87.6

(Comment to Admin: The tables look really ugly here. There are simple BBCodes to make look it better (paddings/margins in the cells, perhaps centered values) which are easily implemented to the board. Unbelievable that at a such an old board noone ever fixed this.)

At a given height is always the same temperature of vapor and liquid. Although it needs temperature differences for rectification, those differences equalize at once as they occur (and then rectification happens).

What drives the rectification is the pushing up of the vapor that isn't able to condense at the temperature it was evaporated, because it's composition has a lower boiling point than the liquid from that it evaporated. Contrarily to the liquid: The liquid will have condensed the same abv than as vapor. So it is able to evaporate right away after condensing.

As I said, those are opinions not something in a large physics book.

[jward]

I'm attempting to brainstorm along with you and not trying to stillsplain anything as I am not qualified for that. Sorry if any of this reads that way. I am sure you understand distillation better than I do.

The theory sometimes seems very close to the thinking that if I run at the correct temp I get azeo. As I understand it once you achieve azeo then you get the corresponding temp but hitting the azeo temp doesn't necessarily mean you are at azeo since other vapor mixes could also give you that temp. Do I have that right? The jacketed column needs to ensure the separation has fully taken place and not just that the column is operating at desired temp. I guess the test will be running the still at all the correct temps and see if you get azeo product coming out.

I don't think a longer column can hurt the distillation. Once the vapor has reached azeo and keeps pushing up the column it should stay azeo unless you lose purity by pushing less pure vapor up the column.

If I understand your design the column will relux and separate (aka stack the column) as the vapor makes its way up the column in sort of one pass. Each area of the column will reflux at it's specific temp. Product take off can happen as soon as vapor reaches the top of the column as it's already reached full separation by the time it gets there. A traditional relux still starts in full relux to stack the column. Once the column is stacked and stabilized one can start taking off product. Once the columns are stacked both approaches seem to be doing the same thing. I would expect in a traditional top reflux still that is putting out azeo product to also be sending condensed azeo product down the still driving the separation lower within the column. Since impure vapors are not really reaching the relux condenser at the top once it's running at equilibrium the chilled column may not have much advantage since both column types would be running very similarly at this point.

On one hand it kinda looks like it's only changing the reflux condenser position from the top of the column to along a section of cooled column. On the other, maybe it is more efficient to create reflux at specific temps along the column right where the separation occurs. Is it more efficient to have the reflux and separation happen close to each other along the column at specific temps along the way? Is the approach also more controllable and easier to establish and maintain the stable gradient in the column?

And maybe I don't understand....

[Tropic]

I don't know if I understand you correctly. In your theory the reflux of something near the azeotrope has a very low abv. Is that right?
But then a LM wouldn't work at all. A LM collects reflux near the azeotrope. So the reflux at the top of the column must be near the azeotrope.

I think you have misunderstood. All these discussions are based on azeo (97.2% abv) being collected/refluxed at the head.

You can effectively set the concentration that the reflux will end up when it exits the cooled section. Looking at the graph above, 85% ethanol solution boils at 79°C. If the surrounding gas is 79°C, the 97% reflux liquid can get to 85% then boil no further.

Starting with 97% ethanol reflux and running the column at:
79°C will leave 85% ethanol in the reflux, re-evaporating 12%.

Azeo is easy to achieve with a long column running at the proper speed. What I'm discussing here is controling the temperature of the column and the temperature of the gas. in the example above, if azeo reflux (boiling point 78.2°C) is poured down a 79°C column, it will boil and preferentially remove ethanol until it reaches 85% ethanol in the reflux which has a b.p. of 79°C.

But what we call azeo can contain impurities. I've achieved elimination of most of these impurities with my column base cooler but I'd like to make it even better.

One of the problems I've had with the device is that is slowed the process. In thinking about the discussion I wrote above, I realise it is because so much of the ethanol is being held back from the reflux. It may be I need to rethink the device position or temperature range.

[Tropic]

I'm attempting to brainstorm along with you and not trying to stillsplain anything as I am not qualified for that.

Thanks. Every point of view gives me something to think about and a greater understanding of the process.

If I understand your design the column will relux and separate (aka stack the column) as the vapor makes its way up the column in sort of one pass. Each area of the column will reflux at it's specific temp. Product take off can happen as soon as vapor reaches the top of the column as it's already reached full separation by the time it gets there.

This is the dream. Passive reflux occurring throughout the column creating a pure neutral by the time it gets to the top. But to achieve this, I would need a veeeery long column - ideally long enough to accommodate all the gas the boiler can produce. Probably around five storeys high. Then it needs to be thermoregulated its entire length. So, not exactly feasible.
What I am proposing only gets half way there. Rapidly reduce the temperature at the base of the column then allow the remainder of column to do a more gradual rectification process. Reflux that contains tails will release a greater percentage of tails in the gas as it drops down the column because the ethanol will preferentially evaporate first. So a column one with a temperature gradient will release tails lower in the column than a standard column. But it will also release less ethanol due to the lower temperature. One of the side effects of cooling the base of the column has been a big drop in coolant required at the head.

I would expect in a traditional top reflux still that is putting out azeo product to also be sending condensed azeo product down the still driving the separation lower within the column. Since impure vapors are not really reaching the reflux condenser at the top once it's running at equilibrium the chilled column may not have much advantage since both column types would be running very similarly at this point.

In a standard column, almost none of the gas (with tails) produced from the boiler is condensed in the column so most of it will reach the still head. Reflux that gas, throw it back down the column and the ethanol will preferentially evaporate out of the reflux. This higher purity gas from the reflux joins the boiler gas and is pushed back up to the still head. By only collecting a small portion of the product, we allow the clean refluxed gas to far exceed the boiler gas, diluting it to a tolerable level. But you will always have tails in your product because gas from the boiler will always get to the still head.
Inducing reflux within the column could prevent most of the raw boiler gas making it to the still head.

[Hügelwilli]

In a standard column, almost none of the gas (with tails) produced from the boiler is condensed in the column so most of it will reach the still head.

A long normal reflux column with a normal condenser on top in full reflux has a temperature gradient from the boiler temperature to almost azeotrope temperature. The rising vapor will condense where the temperature is its condensing temperature.
An example:
10%abv in the boiler will produce 54.9%abv steam with 92.4°C. The condensing temperature of 54.9%abv is 82.2°C. So this vapor will climb until the temperature is 82.2°C. This is not at the top of the column but exactly only one theoretical plate higher than before.
The steam will always climb only one theoretical plate higher. Then it will condense and partially reevaporate.

[Demy]

I don't have the technical knowledge to have my say, but from what I understand the challenge here is to have a pre-capacitor with a temperature that is hot enough to pass the lighter alcohol but can condense most of the tails. I think the key is precisely the temperature of the fluid flowing in the coil, just my thought.

[jward]

I would expect in a traditional top reflux still that is putting out azeo product to also be sending condensed azeo product down the still driving the separation lower within the column. Since impure vapors are not really reaching the reflux condenser at the top once it's running at equilibrium the chilled column may not have much advantage since both column types would be running very similarly at this point.

In a standard column, almost none of the gas (with tails) produced from the boiler is condensed in the column so most of it will reach the still head. Reflux that gas, throw it back down the column and the ethanol will preferentially evaporate out of the reflux. This higher purity gas from the reflux joins the boiler gas and is pushed back up to the still head. By only collecting a small portion of the product, we allow the clean refluxed gas to far exceed the boiler gas, diluting it to a tolerable level. But you will always have tails in your product because gas from the boiler will always get to the still head.
Inducing reflux within the column could prevent most of the raw boiler gas making it to the still head.

This seems wrong to me. Don't the gases as they go from the boiler through the column separate as they go up? How can the vapor from the boiler remain intact all the way through the packing and stacked liquid up to the top without getting changed? The lighter parts should tend to separate and move up and the heavier stuff condense and stay or move down as vapor goes through the column. The relux/condensing vapor drives the separation but that doesn't mean the vapor coming off the boiler is the same as the vapor reaching the top. The constituents within the vapor should be getting changed all along the way. If the vapor came straight up with all the tails all the way to the top you probably would never get clean product as long as the boil runs.

It seems the reflux condenser and product condenser have to be condensing essentially the same vapor. The vapor at the top of the column either goes to the relux condenser to condense and drip back down as reflux or escapes over to the PC to become product. On a tradition column still doesn't one run the still to load the column and stack the light stuff on the top and heavier stuff on the bottom before product take off? Then you maintain enough reflux to keep the product as pure and you can (or want). You can't really have tailsy vapor get to the top of a stacked column to become reflux then get purified later.

[Tropic]

A long normal reflux column with a normal condenser on top in full reflux has a temperature gradient from the boiler temperature to almost azeotrope temperature. The rising vapor will condense where the temperature is its condensing temperature.
An example:
10%abv in the boiler will produce 54.9%abv steam with 92.4°C. The condensing temperature of 54.9%abv is 82.2°C. So this vapor will climb until the temperature is 82.2°C. This is not at the top of the column but exactly only one theoretical plate higher than before.
The steam will always climb only one theoretical plate higher. Then it will condense and partially reevaporate.

Exactly the problem with a non-regulated column. Where does this condensation temperature occur? For the first distillation equivalent within the column it is definitely not one theoretical plate (about 15cm) above the boiler. It would be great if it did drop 10°C in the first 15cm, then down to 79.4°C at 30cm. But it doesn't. In a non-regulated column this is way towards the top. It depends on the amount and temperature of the reflux generated from the head as well as the power and temperature of the boiler. And as the wash gets lower in ethanol, the high temperatures go even higher up the column. Tails in the product. Yuck. Shut off the product collection, 'stack the column', start collecting again. How cumbersome.
With just the cooler I have at the bottom of the column, I don't have these problems. It just just slows down towards the end of the run. The product collected at 98°C wash is just as good as the middle of the run. And regulating the radiator temperature is pretty easy with the variable fan speed.

the column will relux and separate (aka stack the column) as the vapor makes its way up the column in sort of one pass. Each area of the column will reflux at it's specific temp. Product take off can happen as soon as vapor reaches the top of the column as it's already reached full separation by the time it gets there. A traditional relux still starts in full relux to stack the column. Once the column is stacked and stabilized one can start taking off product.

This comment has taken me further towards the idea of temperature gradient regulation. Thanks! This is a completely new type of still. Not lm, cm or vm. And it requires a completely different way of thinking.

Originally, I was just going to use the jacket to cool the gas a bit and allow a bit of passive reflux then reflux at the head as per a normal old column. But why not take it all the way? It solves some of the big problems with an unregulated column. The hard thing to contemplate was the major shift in how the still works. There would be no reflux passed down the column from the head. None! The refluxing would be done throughout the column. Well, probably the top two-thirds would all be generating reflux. The gas will rise until it hits its boiling/condensation point, condense at a temperature just below its boiling point, drop down the column a bit re-boil to a bit more pure gas, rise a bit further. Repeat until it can get past 78.2°C. And hey presto! I collect the purest neutral.

Removing heads.
Now this is where I haven't mastered my current system. I was just running it the same as for product and waiting for the heads to go away. But they stayed in the first quarter to third of the product. From reading a bit here, I realise I should have been stacking the column with heads while refluxing 100%. With my column with a base cooler but no regulation after that, I can see my mistake.
But regulating the column temperature gradient should also be great for compressing and removing heads. Set the gradient similar to the product run but instead of setting the top to 78.2°C, set it to a bit lower than that. The lighter heads will get past to the head to be collected but not much ethanol. Obviously, it won't be that simple and will require a gradual raising of the upper column temperature to get the rest but it should be better than before.

[Hügelwilli]

A long normal reflux column with a normal condenser on top in full reflux has a temperature gradient from the boiler temperature to almost azeotrope temperature. The rising vapor will condense where the temperature is its condensing temperature.
An example:
10%abv in the boiler will produce 54.9%abv steam with 92.4°C. The condensing temperature of 54.9%abv is 82.2°C. So this vapor will climb until the temperature is 82.2°C. This is not at the top of the column but exactly only one theoretical plate higher than before.
The steam will always climb only one theoretical plate higher. Then it will condense and partially reevaporate.

Exactly the problem with a non-regulated column. Where does this condensation temperature occur? For the first distillation equivalent within the column it is definitely not one theoretical plate (about 15cm) above the boiler. It would be great if it did drop 10°C in the first 15cm, then down to 79.4°C at 30cm. But it doesn't. In a non-regulated column this is way towards the top. It depends on the amount and temperature of the reflux generated from the head As you can read what I wrote was meant for full reflux as well as the power and temperature of the boiler. And as the wash gets lower in ethanol, the high temperatures go even higher up the column. This contradicts the McCabe & Thiele -theory, the theory all industrial rectifying columns are calculated with. If one plate needs more than 15cm at full reflux, then the HETP is simply higher than 15cm. Poor packing or wrong wattage for this packing.

[jward]

This comment has taken me further towards the idea of temperature gradient regulation. Thanks! This is a completely new type of still. Not lm, cm or vm. And it requires a completely different way of thinking.

Originally, I was just going to use the jacket to cool the gas a bit and allow a bit of passive reflux then reflux at the head as per a normal old column. But why not take it all the way? It solves some of the big problems with an unregulated column. The hard thing to contemplate was the major shift in how the still works. There would be no reflux passed down the column from the head. None! The refluxing would be done throughout the column. Well, probably the top two-thirds would all be generating reflux. The gas will rise until it hits its boiling/condensation point, condense at a temperature just below its boiling point, drop down the column a bit re-boil to a bit more pure gas, rise a bit further. Repeat until it can get past 78.2°C. And hey presto! I collect the purest neutral.

It's a little funny that me misunderstanding you gives you ideas.

[Saltbush Bill]

Are you 100% sure your not just very much over thinking things and building something that will only do what a good reflux still does anyway.
A well run, well built reflux still stacks fractions.....holds tails back till the very end and does an excellent job of separating fores and heads from hearts.
Seems like a whole lot of maybes and trouble for litte if any gain to me.

[LWTCS]

@tropic, are you able to be more specific when you say
"unregulated column"?
Or, familiarize yourself a bit more with distilling terminology perhaps? I'm not exactly clear what that is supposed to mean?
I do apologize if your native tongue is not English.

This discussion is interesting and thus far has been moving along with enthusiastic participation.
But please be mindful not to declare percieved or expected still behavior as a matter of fact. If you don't have first hand experience or a documented example to bring to the table to make your point, then it would be best to frame one's thoughts in the form of a question rather than a practical observation.

[Tropic]

Are you 100% sure your not just very much over thinking things

I'm 100% emphatically sure that I AM very much over thinking things. What I am proposing is overly complicated and could be a huge waste of time. But I have the time to waste so I will go ahead with the experiment, possible just for my amusement and that of people who are following this idea. If it doesn't work, I end up with a column jacketed with pvc pipe which could be a good insulated column.
But then again, it might just work (and still be overly complicated so no-one will follow suit and build another, anyway). I've had great success with the prototype cooler at the base of my column so just maybe...

and building something that will only do what a good reflux still does anyway.
A well run, well built reflux still stacks fractions.....holds tails back till the very end and does an excellent job of separating fores and heads from hearts.
Seems like a whole lot of maybes and trouble for litte if any gain to me.

Yep, you are right again. But isn't that what this site is about? Otherwise we'd all just knock out a pre-designed standard Boka or maybe a CCVM and be done with it. As long as our builds are safe then I think weird and whacky ideas like this should be encouraged. I was posting here to get people's ideas and to see if anyone had tried something similar so I could hear their experiences. Thanks to everyone who has given me things to think about.
I've been running and modifying my current boringly common still for many years. Sitting there, watching it, pondering. Surely we all contemplate other ways of doing it.

[LWTCS]

Ok good on you for going through the exercise of discovery.

The pre-separation HX will work best when you can optimally control ALL of your input values.

Therefore, it is my assertion that when trying to recruit the pre-separation HX for batch distilling, you will find that promoting a naturally occurring gradient will bring your apparatus on line so much faster with better finished product

The forum is rife with examples of reflux condensers that are positioned down low on the column riser not working very well at all with respect to distilled spirits.
Will this type of set make alcohol? Yes. Will it make good alcohol? Meh,,,,depends on your level of exposure. You need more exposure btw.

With batch distilling, you will not at all be able to control kettle temps. Therefore you will constantly have to adjust your pre-separation HX cooling medium flow rate in order to get optimal behavior. Think: chasing your tail.
The pre-separation HX simply works best on a steady state, continuous apparatus where all of the variables are more easily controlled.

Allowing for a naturally occurring gradient will get you across the finish line faster and with better results when batch distilling,,,imo.

[Saltbush Bill]

As long as our builds are safe then I think weird and whacky ideas like this should be encouraged.

Agree 100% ...without experimentation the hobby never goes forward.

[Tropic]

The forum is rife with examples of reflux condensers that are positioned down low on the column riser not working very well at all with respect to distilled spirits.

Yep, but I'm not using a condenser low on the column, not even a dephlegmator. I'm currently using a device to cool the vapour entering the column with a coil using temperature regulated coolant, usually at a minimum 84°C. There should be no condensate if the coil is below the vapour's condensation point, just a cooling action. It means the "naturally occurring gradient" begins at a lower point (say 84°C instead of 93°C), spreading a more gradual gradient throughout the column and therefore (I think) giving better separation.
Have you seen this done before? I've only found it on an innovative commercial continuous whiskey still used to "knock down tails" but can't find the link.

[Zeotropic]

The forum is rife with examples of reflux condensers that are positioned down low on the column riser not working very well at all with respect to distilled spirits.

Yep, but I'm not using a condenser low on the column, not even a dephlegmator. I'm currently using a device to cool the vapour entering the column with a coil using temperature regulated coolant, usually at a minimum 84°C. There should be no condensate if the coil is below the vapour's condensation point, just a cooling action. It means the "naturally occurring gradient" begins at a lower point (say 84°C instead of 93°C), spreading a more gradual gradient throughout the column and therefore (I think) giving better separation.
Have you seen this done before? I've only found it on an innovative commercial continuous whiskey still used to "knock down tails" but can't find the link.

There is one problem here. The vapor coming from the boiler is precisely at its boiling/condensing temperature so if you cool it at all all you are doing is condensing vapor which would be taking power away from the column which could be accomplished by turning down the power.

In my opinion
If you instead allow the packing in the column to condense some of the liquid and then hold up that liquid and prevent it from dropping back to the boiler immediately then the upcoming vapor will be hot enough to reboil some of it and since that condensed vapor is at a higher alcohol content than the boiler was you will get another distillation and a further increase in ABV. One thing I've read on here in some of the in-depth discussions in the past is that you can't expect theoretical plates to be literal to their dimension because the lower part of the column is much less efficient so the theoretical plates are much taller and then up towards the top the packing gets more efficient and the theoretical plates are much shorter so when we refer to the theoretical plates of a column that is the average. (I think[emoji848])

[Tropic]

There is one problem here. The vapor coming from the boiler is precisely at its boiling/condensing temperature so if you cool it at all all you are doing is condensing vapor which would be taking power away from the column which could be accomplished by turning down the power.

I don't think this is correct. It is my understanding that the vapour from boiling is at the same temperature as the boiler but has already effectively gone through one distillation. As I stated under the "Boiler" heading at the very start:

Boiling (i.e. converting liquid to gas) in the boiler creates the first distillation. With 10% ethanol in the wash, the boiler (and bottom of the column) will be at 93°C producing a gas of 53% ethanol. A liquid of 53% ethanol boils at around 82°C. Inversely, that gas will condense (i.e. convert gas to liquid) when it drops below 82°C.

This logic is correct, isn't it?

One thing I've read on here in some of the in-depth discussions in the past is that you can't expect theoretical plates to be literal to their dimension because the lower part of the column is much less efficient so the theoretical plates are much taller and then up towards the top the packing gets more efficient and the theoretical plates are much shorter so when we refer to the theoretical plates of a column that is the average. (I think[emoji848])

I also think that the high temperatures take a lot of the column length before it really starts to work well.
Which is exactly why I cool the vapour down to just above its boiling point since I think it gets past these lower column inefficiencies. I think my system starts to work efficiently much lower in the column.

Has anyone ever actually mapped out the temperature gradient on a column still?

[Saltbush Bill]

Has anyone ever actually mapped out the temperature gradient on a column still?

Ive seen it done but not here ....Captainshooch did it on a 5 or 6 plate bubble cap column. He had thermo ports at each plate. Temp gradient was greater than I thought it would have been.....not sure I can even find a link to it.....Yummy may know where it is ...his brain seems to work like a search engine when it comes to forums...mine doesnt.

[Yummyrum]

Ive seen it done but not here ....Captainshooch did it on a 5 or 6 plate bubble cap column. He had thermo ports at each plate. Temp gradient was greater than I thought it would have been.....not sure I can even find a link to it.....Yummy may know where it is ...his brain seems to work like a search engine when it comes to forums...mine doesnt.

This one Salty ... BTW ... the forum brain isn’t what it used to be .
http://www.coppercustomstillcomponents. ... 553#p19551

[Tropic]

All this is my opinion only as to what goes on in a packed column:
I would assume that in a packed column the temperature gradient would be entirely governed by the reflux. I would think that if there was no reflux and the column was fully insulated, there would be no condensing and the column would simply be a conduit for the vapour. And as such the temperature would be the same temperature as the boiler all the way to the head.
Azeo reflux produced at the head is probably a fair bit colder than 72°C. So it immediately would cool the top section. But as it drops, it gets heated by the vapour. It evaporates, creating a bit of evaporative cooling. And as it falls, it will encounter vapour hotter than its boiling point but the laws of physics says that the liquid can only be as hot as its boiling point. So it would be cooling the column as it falls, shedding ethanol as it goes until its boiling point approaches the vapour temperature. The lower it goes, the less reflux volume exists and that reflux has less cooling ethanol to evaporate. With only a small volume of low ethanol reflux, the lower column will mostly act as a conduit until it reaches a point where it encounters reflux with a decent amount of ethanol.
And as the boiler drops in ethanol, the temperature rises in the lower column section. There is less reflux to cool the column and the high temperatures get higher and higher.