Measured product production in VM head

[DestructoMutt]

Reflux Ratio: Column Diameter vs. Product Take-Off Port Diameter
---the theory: initial take-off port size controls the reflux ratio, not what the vapor path is reduced to after the initial take-off port size.

Wash: 3 - 1.5L bottles of 95%ABV, with enough water added to make 10 gallons.
Equilibrium Period: 2 Hours
Input Heat: 1125 Watts (4500W at 240VAC element used at 120VAC)
Column Diameter: 2 inch nominal
Column Height: 46 inches, packed height
Column Packing: 15 copper scrubbers, choreboy brand

Product volume was measured for the first hour of production after the equilibrium period. The take-off valve was fully opened and product was collected for 60 minutes, then cooled to 68F and then the volume was measured. ABV of the product was not measured.


Product condensor and valve match the take-off port size.
Take-Off Port Diameter:......... 2"....................... 1"....................... ½"
RR (theoretical*):................. 1:1..................... 4:1..................... 16:1
Product output:................... 2894ml................1139ml.................. 335ml
RR, observed**.................... 0.93:1................3.26:1....................15.89:1


Valve matches the take-off port size, product condensor is ½" diameter, reduction occurs after 90deg. elbow.
Take-Off Port Diameter:......... 2"....................... 1"....................... ½"
RR (theoretical*):................. 1:1..................... 4:1..................... 16:1
Product output:.................... 345ml.................. 334ml.................. 340ml
RR, observed**.................... 15.37:1................ 15.98:1................ 15.65:1


Valve matches condensor size (½" diamter), reduction occurs immediately after take-off port.
Take-Off Port Diameter:......... 2"......................1"........................ ½"
RR (theoretical*):................. 1:1..................... 4:1..................... 16:1
Product output:....................344ml.................. 337ml.................. 336ml
RR, observed**.....................15.97:1................ 15.77:1................ 15.84:1

Conclusion: it appears that the size of the pathway that the vapor "sees" on it's way to the product condensor is what controls the reflux ratio.

Additional information: The column was allowed to drain overnight, with the reflux and product condensors running. The column packing was not changed throughout the test period. After measuring the volume, all product was returned to the boiler, and the volume was checked for losses. Total losses for the entire test period were less than 225ml. A recirculating cooling system was used; 55 gallon barrel and a pond pump, 150 gal/hr. Cooling water was allowed to return to 72F between tests. Reflux condensor is 1/4"ID copper tube wrapped around a 1" mandrel and 10" overall in length. The bottom of the reflux was held at 2" above the top of the take-off port. The packing stopped 1" below the bottom of the take-off port. The product condensors all used 3/8" ID inlets and outlets for the cooling water, and were 18" long. They were followed up with a coiled liebig of ½" ID copper pipe inside 3/4" copper pipe at 9 feet long. There were no other controls on the cooling water flow. Product take-off valves were positioned as close to the column as possible and in a horizontal orientation. The 2" diameter product condensor was of the graham design (3/8" ID copper coil inside a 2" pipe).

* = Theoretical reflux ratio is based on cross sectional area of column and of the take-off port (nominal dimensions).

** = Observed reflux ratio is calculated from theoretical total of vapor production and actual product take-off.

I started on this test in September of '08, initial planning, the first shakedown run was in October '08. initially i was going to use a sugar wash and monitor the %ABV of the product for the run. but the shakedown run showed some problems. bleeding off the foreshots - for example - i wanted a consistent run each time for each head variation, so i switched to store bought everclear and enough to hopefully eliminate any drop in production due to diminishing alcohol supplies in the wash. the next run, using everclear, was much easier, and i was able to work on a procedure to enable consistent and repeatable runs for each VM variation head. i was then able to do one run every four weeks.....i do have other hobbies and a family...

[Husker]

Good information DM. This was your testing from last year? Seems like a while to hold on to it, but i guess you have reasons.

I have read from others that they imply the output rate (hence the RR), was based upon the port ratio. Now, I have not run VM, so had nothing to base my thoughts upon, but that premise of the RR being wholly controlled by port size, did not seem to pass the smell test. I would have though the RR would be controlled by the 'smallest' pathway, since that is where you are going to limit the flow of vapor prior to condensing it.

Your testing seems to show that this (the smallest constriction prior to the condenser) is what the is actually in control of the RR in a VM.

H.

[rad14701]

Thanks, DestructoMutt...

Definitely good information and inline with my unsubstantiated theory that maintaining take off branch diameter all the way to the condenser would be beneficial... This data proves that out with solid numbers...

I just got a request last night for a VM column build and had already started designing it as 2"x2"x2" with 2" going all the way to either 1/2" or 3/4" inner OD for the liebig... One of the partners is a plumbing contractor so we may even go with 3" if he has some on hand...

[DestructoMutt]

actually i didn't hold on to it so much as waited until i had finished all nine runs. also i didn't want to make noise about it until i had finished all nine runs....nothing like someone saying they are going to do something or are in the middle of doing something but never producing results to get a persons expectations up and then dashed.

regarding the smell test

did not seem to pass the smell test

, yeah, looking at air handling units and the ductwork leaving and returning to them started me thinking about the claims made by others also.



Edit - add the following
p.s. i have listed my procedures and methods so that the test can be replicated. it would be nice to have independent confirmation or ....

[DestructoMutt]

The theory is that on a VM head, an increase in the vertical distance from the take-off port to the product condensor (hereafter called the “drop section”) will create a “reverse chimney affect” in the “drop section" and thus create a “drawing” affect, essentially pulling vapor from the main column into the product condensor.

easy enough to prove - add a 12" piece of “drop section” between my elbow and product condensor, or a 24" section.

Using the same equipment (1" take-off port, 1" gate valve and 1" ID product condensor), wash and procedures from before, but modified - i added a 24" piece of 1" pipe between the elbow and the product condensor.

Measured output was 1143ml over sixty minutes, a 4 ml increase.





Again, i have listed my methods and procedures for review and independent confirmation or......ridicule.

[HookLine]

Okay, DM, I have had a look at your experiment, and I can't accept those numbers you are getting. They are too low, and do not match other (multiple & independent) reports.

I can get about 2.5 litres/hour from my 2" (in reflux mode), with a 1" take-off port, that reduces straight into a 3/4" gate valve, and then into a Liebig with a 3/4" vapour tube.

http://homedistiller.org/forum/viewtopi ... 0#p6768120

(I now have a 1/2" Liebig, so I will test that during my next reflux run. Which will not be for a while. But if your theory is true, then I should see a big reduction in maximum output, seeing as 3/4 down to 1/2" is a reduction of about 55% in cross-sectional-area, CSA.)


Using an identical design column (and still) to mine, except with a 3/4" take off port, Punkin only got 1.5 litres/h. So it seems the take-off port size itself is important, besides any other variable.

http://homedistiller.org/forum/viewtopi ... 0#p6760980
http://homedistiller.org/forum/viewtopi ... 2#p6764582


Minime found no difference between a 1" and 1/2" Liebig, on his 3" column with a 1" take off port, or for different size valves. He also hit 5.5 litres/hour out of a 3" column, using a 2x2x2 head on top, and a 1/2" valve and Liebig.

http://homedistiller.org/forum/viewtopi ... 7#p6767827
http://homedistiller.org/forum/viewtopi ... 5#p6769005

(Incidentally, it is true that Mini used gas for his heat source, but it does not follow that the heat input level was significantly variable and inconsistent, especially in the hands of an experienced operator. Grid electricity is not always nice and stable either, standard grid voltages here can legally vary ± 5%, from 228-252v, which for a 2400 w element equals 2280-2520 w, a significant difference. Did you control for that?)

Maoule reports getting around 2 litres/h for the normal run, from a 2" column, into 2" take off and elbow, and a 3/4" valve and Liebig. Well above what you say he should be able to get.

http://homedistiller.org/forum/viewtopi ... 1#p6769871

(His results from removing the mesh inside the Liebig do indicate that the vapour flow inside the take off arm is finely balanced, and it doesn't take much to disrupt it.)


These numbers and direct experience, from a variety of sources, contradict your claims about maximum output rates. Indeed, I will bet there are dozens of threads/posts on HD, AD, Yahoo, etc, about people getting VM output rates well above what you are claiming, including during normal operation, not even at maximum output. If they did not get those rates, they would be doing something wrong.

What you is arguing is that it does not matter where in the take off arm the smallest restriction is, it is the sole or at least primary determinant of RR. The results from others indicate instead to me that the take-off port has a disproportionally greater effect than other restrictions in the take off arm. (Presumably due to it being where the vapour has to split and bend 90º to get into the take off arm.) Other important factors besides just the simple CSA of the restrictions/reductions are probably in play, such as how much turbulence there is at the take-off port.

"The packing stopped 1" below the bottom of the take-off port."

Did you have a turbolator on the top of the packing? If not the it is possible that your arrangement does not cause enough turbulence, and so makes it harder for the vapour to turn into the take off arm. Try dropping the packing to 2" below, and adding a turbolator on top of the packing.

Also, you do not mention column insulation in your notes, nor what material your column is made from. And is the boiler well insulated?

And, incidentally, how are you centring the reflux?

"Conclusion: it appears that the size of the pathway that the vapor "sees" on it's way to the product condensor is what controls the reflux ratio."

This statement is a bit vague and lacking in specifics. We already know in principle that, all other things being equal, restrictions slow the overall flow rate of vapour. But the devil is in the detail... What defines the degree of restriction in this situation? Is it just the literal CSA changes and ratios in the vapour path. Or are there additional factors?

In particular, I suspect that the requirement in a VM for one stream of the vapour to turn 90º into the take off arm functionally reduces the size of the take off port that this vapour stream sees, so that any subsequent reduction in the size off the take off arm will not have much effect, until the reduction reaches the functionally equivalent size of the take off port. So, for example, a 1" take off port could be acting as equivalent to a 3/4". Generating turbulence just below the take off port should increase the equivalent port area.

And the whole thing is also further complicated by the ethanol density issue, and any reverse chimney effect from the extension drops, etc. Speaking of which...

IIRC, the overall results for reverse chimney drops are mixed. Snuffy found that it only made a small difference, but a much bigger one than you claim, maybe 10%. Minime found it made little difference. And Riku reports it made a substantial difference. All up I would have to say this particular issue is not settled.

"Observed reflux ratio is calculated from theoretical total of vapor production and actual product take-off."

Don't agree with this method. Observed ratio should be measured by actual outputs only, not theoretical ones. Measure the output of the column with the reflux condenser removed, the column top capped & sealed, the valve wide open, and the heat source at its normal setting for a reflux run. Then uncap column and return reflux condenser as for normal VM operation, and repeat output measurement with valve wide open. Dividing the first number by the second gives an empirically derived RR.

••••••••••••••••••••

In summary, I do not believe the results from this particular experiment are reliable. They seriously contradict too many other experimental results and everyday experience with VM columns, and I suspect that you have not properly controlled all the important variables (such as turbulence at the take-off port, etc).

I can already get 1.2 litre/h with my new 1/2" Liebig, in normal output mode (ie not even at maximum output), whereas your results say I should only be getting a maximum possible output rate of around 350 ml/h.

Clearly something does not add up here.

My strong advice to all members is do not use the results of DM's experiment when designing your VM column.


One last thing, you have been asked a number of times by different people to post some pics of your experimental set-up. So where are they?

[DestructoMutt]

Hookline-

for all the information you included, you didn't include what your heat source is - gas or electric?

my heat source was/is electric, 120VAC nominal off the grid (115VAC +/-) during runs.

while it is possible for a person to set-up a monitor to carefully control their gas usage, none of the people that use gas have indicated that that is what they do. Also, unless the gas is being used to heat a medium that is then transferring heat to the interior of the boiler, the use of flame heat is subject to considerable variations due to heat transfer, air movement, separation between source and container, available make-up air, available combustion air, etc....i'm not saying it is not possible, only extremely difficult to tune a gas flame to the same level every time it is used, if not turned to full on.

both Minime and Punkin, although talented craftsmen, both have backed off their claims when prodded for details. They both admitted that they ran on gas and couldn't confirm that they had the same heat input both times. they also did not accurately measure the output at the same time for each run, nor did they use the same wash (volume and %ABV) for each run. and, if i recall correctly, Minime did claim that the 1/2" product condensor seriously decreased product output. minime also ran his gas heat quite high to get that 5.5L/hour (which was actually timed over a 10-15 minuter period and calculated up to an one hour equivalent) which he was doing as a stripping run, i believe.

regarding turbulence in the column, simple fluid dynamics indicate that the flow of vapor in the packed column is turbulent. fluid dynamics also indicates that it takes a certain distance for a fluid flow to obtain laminar flow after encountering an obstacle that causes turbulence, this distance is generally regarded as about 3 times the diameter of the tube the fluid is flowing in. while precise measurements can be made, the rule of thumb is generally adequate, and hence my attempt at reducing this distance to 1" in my tests.

the reflux is being centered by dripping from the reflux condensor. fluid dynamics suggest that a flow constrictor (a centering collar would be considered a flow restrictor as it has an inside diameter smaller than the column and thus restricts or constrains the vapor flow) would disrupt the vapor flow in the column, compressing and speeding up the vapor flow at that point in the column.

yes my results contradict lots of "direct experience, from a variety of sources", but my results are from controlled runs, using the same wash, using a consistent power input, using real measuring devices, over a set time period. many of your sources, upon being questioned about their protocols, back away from their initial statements. simple reasons such as, "well, i tuned the flame by ear", or "the jug looked fuller this time than last time", or "well i didn't really start measuring until a little bit after the flow got going good". i'm just trying to present reliable data, whether you wish to believe or not is up to you. and the last time i checked, pictures of equipment doesn't prove anything other than i have access to the materials.

i tried to use simple, easily controlled, methods, that anyone could repeat. boiler setup, column setup, and wash used were all the same to minimize differences between runs.

Don't agree with this method. Observed ratio should be measured by actual outputs only, not theoretical ones. Measure the output of the column with the reflux condenser removed, the column top capped & sealed, the valve wide open, and the heat source at its normal setting for a reflux run. Then uncap column and return reflux condenser as for normal VM operation, and repeat output measurement with valve wide open. Dividing the first number by the second gives an empirically derived RR.

that would be nice if it worked that way, but reflux ratio is the ratio of material returned to the column versus the amount removed as product. 10 divided by 1 is not the same as 10:1 or 11:1. it would also be nice if the "variety of sources "used the same method for determining their RR's. or actually were consistent with their measurements, but they have not provided assurances that they are, have done so or have been.

I can already get 1.2 litre/h with my new 1/2" Liebig, in normal output mode (ie not even at maximum output), whereas your results say I should only be getting a maximum possible output rate of around 350 ml/h. One last thing, you have been asked a number of times by different people to post some pics of your experimental set-up. So where are they?

and what is your heat input? 3800W? or are you guessing on your gas heat input? and where are the pictures of your set-up that prove you are running ethanol and not just water?

i've reread your post a couple times now and it appears that you have missed a minor piece of information - the output from each run is based on an input heat of 1125W. if you increase the heat input, the amount of output increases proportionately

[DestructoMutt]

I can get about 2.5 litres/hour from my 2" (in reflux mode), with a 1" take-off port, that reduces straight into a 3/4" gate valve, and then into a Liebig with a 3/4" vapour tube.

and what's your heat input? for my numbers, my heat input was 1125W.

Okay, DM, I have had a look at your experiment, and I can't accept those numbers you are getting. They are too low, and do not match other (multiple & independent) reports.

too low for what? the heat input i'm using is 1125W. yes, you can get more throughput if you use higher heat, there is no question about that. my numbers are not maximum amounts achievable from the various head set-ups. they show the reflux ratio of various head set-ups at one heat input, 1125watts.

if you increase the amount of heat, say to 2400watts, you will get 2.13 times as much product out, or approximately 750ml in the first hour (from a 2" column and 1/2 product condensor).

[myles]

I don't know what to say I do know what I can do.

I am running a 3kW electric, pulse width modulated, propylene glycol, double boiler, on which I am monitoring the bath temperature. I am using 12 litres of glycol to heat 22 litres of wash and the thermal buffer effect is significant. I can stabilise the bath so that it maintains a constant temperature during the run. I have tried it with bath temperatures up to 140 degrees centigrade. It can go higher but I do not see the point.

I currently measure vapour temperature at the base of the column and a few inches below my VM take off.

When I don't have anything better to do, I shall monitor a spirit run, and post my results, but I do not expect them to match what is being claimed here. Could be wrong of course but I doubt it.

You will have to calculate the temperature gradients and heat inputs yourself, because I can't be bothered to work them out. If you really want, I can run identical washes at different bath temperatures. More data for you to ‘discus’

[The Baker]

[quote="DestructoMutt"]Hookline-


the reflux is being centered by dripping from the reflux condensor. fluid dynamics suggest that a flow constrictor (a centering collar would be considered a flow restrictor as it has an inside diameter smaller than the column and thus restricts or constrains the vapor flow) would disrupt the vapor flow in the column, compressing and speeding up the vapor flow at that point in the column.

Just suppose instead of a (flow constricting) centreing collar you used another (non-constricting) method of directing falling droplets of condensate away from the edge of the column.
(You want to do this so that the descending droplets are directly in the path of ascending vapour, right? Not running down the side of the column.)

The simplest way I can see to do this, is to use something like a shortish length of packing; but tease/shape it so that the bottom narrows toward the centre. Think of it as being like a goatee beard.
With a bit of a gap between it and the rest of the packing so the droplets can fall.
The droplets will follow it away from the edge of the column and then fall, but it won't impede the rising vapours any more than the packing which is already in the column.
There would be no need for it to come to a real point, probably better if it doesn't as the droplets would be better distributed across the lower packing but away from the edge.
There may be a benefit in having two of these, one near the top and one in the middle of the 'packing lower down'. (The lower one would just be a shaping of the packing and a bit of a gap.)
I haven't figured the best way to hold them in place but it would be simple. A circlip against the edge of the column; a thin wire on each side (three wires?) to hang it from the condenser; a support wire (three of them?) from the bottom?

[Djurmz]

DM,

Great information. There is an abundance of anecdotal evidence on this board, it is great to see some controlled studies. Thanks for taking the time. Your data has caused me to modify my design on the 2" VM I'm building. I'm wondering what your 2" takeoff condenser looks like because I've only see the much smaller liebigs and such on the site. Do you have a picture?

My 2" VM will be stainless with 2" tri-clamp connections and teflon gaskets. I'll use a SS full port 2" ball valve. I want to make a 2" leibig with a 3" case, but have no idea on what length I'll need and I'm wondering what worked for you.

edit: nevermind. I read the posts again and got all the answers to my questions.

[LWTCS]

In summary, I do not believe the results from this particular experiment are reliable. They seriously contradict too many other experimental results and everyday experience with VM columns, and I suspect that you have not properly controlled all the important variables (such as turbulence at the take-off port, etc).

My strong advice to all members is do not use the results of DM's experiment when designing your VM column.

The implication is that the OP's conclusions were,,,,,,,anecdotal.

[HookLine]

The implication is that the OP's conclusions were,,,,,,,anecdotal.

That is the polite way of saying it.

[Djurmz]

I believe the OP's evidence was anything but anecdotal. He performed nine different runs using the same equipment and wash, changing only the variable pieces. He executed a well planned scientific experiment. I believe you focused on the wrong part of his data. His output rates were low because he had a low heat input, but the point of this experiment focused on reflux ratio:

Product condensor and valve match the take-off port size.
Take-Off Port Diameter:......... 2"....................... 1"....................... ½"
RR (theoretical*):................. 1:1..................... 4:1..................... 16:1
Product output:................... 2894ml................1139ml.................. 335ml
RR, observed**.................... 0.93:1................3.26:1....................15.89:1


Valve matches the take-off port size, product condensor is ½" diameter, reduction occurs after 90deg. elbow.
Take-Off Port Diameter:......... 2"....................... 1"....................... ½"
RR (theoretical*):................. 1:1..................... 4:1..................... 16:1
Product output:.................... 345ml.................. 334ml.................. 340ml
RR, observed**.................... 15.37:1................ 15.98:1................ 15.65:1


Valve matches condensor size (½" diamter), reduction occurs immediately after take-off port.
Take-Off Port Diameter:......... 2"......................1"........................ ½"
RR (theoretical*):................. 1:1..................... 4:1..................... 16:1
Product output:....................344ml.................. 337ml.................. 336ml
RR, observed**.....................15.97:1................ 15.77:1................ 15.84:1

Conclusion: it appears that the size of the pathway that the vapor "sees" on it's way to the product condensor is what controls the reflux ratio.

The outputs that you refute are based on heat input.

I can get about 2.5 litres/hour from my 2" (in reflux mode), with a 1" take-off port, that reduces straight into a 3/4" gate valve, and then into a Liebig with a 3/4" vapour tube.

Again, none of his data refute your output values. His data and yours are not mutually exclusive. You clearly have a higher heat input than he does.

To further prove the point that his conclusion from the data did not concern output rates, maximum or otherwise I used the find function and you were the first in the thread to use the phrase 'maximum output' DM made no such claim. The only reference to output is when he lists the actual output of each setup. You're arguing against statements that DM never made.

What you is arguing is that it does not matter where in the take off arm the smallest restriction is, it is the sole or at least primary determinant of RR.

He is saying ("arguing") nothing other than this statement:
"It appears that the size of the pathway that the vapor "sees" on it's way to the product condensor is what controls the reflux ratio."

Everything else in his post was a description of his experiment and a presentation of objective Data.

It seems you have drawn your own conclusions from DM's objective data, then attributed them to DM and proceeded to argue with him.

But my main concern is that I'm designing a VM still based on DM's recommendations, and you advise directly against that. I want a column that I can adjust from a reflux ratio of 1:1 and go up from there, as opposed to having a built in RR minimum of 16:1.

So, in light of my above understanding of the situation, can you respond by telling me what would go wrong with a VM still built based on this data?

I don't intend to start an argument, I intend to learn more about this so I can build a good VM column...

[theholymackerel]

I don't intend to start an argument, I intend to learn more about this so I can build a good VM column...

Are you sure? Yer post struck me as an attempt to troll.

DM was asked repeatedly to answer a couple questions and post a pic and diagram of his setup. He not only wouldn't provide this needed back-up info for his "experiments", but he responded with vitrol and nasty PM's.

If he would have simply politely explained a few things his "experiments" would have been taken seriously. But he didn't... now you are showin' up with an attitude and sayin' we are the ones with the attitude... once again I say ya sound like a troll.

If yer not, maybe less attitude will get you better results.

[Djurmz]

If my post struck you as an attempt to troll, I do offer my apologies. I didn't intend to show up with attitude or accuse anyone of having attitude. If that is the way my post came across then I apologize. After reading the entire discussion, it seemed to me that Hookline made was arguing incorrect conclusions about DM's data, conclusions which DM never made.

Per your recommendation THM, this is my attempt to ask the question with less attitude. I don't know of any other way to put it that doesn't sound like it's dripping with sarcasm, but you'll just have to take me at my word that I'm not trolling or being sarcastic. Here goes...

I am planning on building a VM column based on DM's data. That is, my 2" column will have a 2" valve (full port), 2" condensor, and 2" takeoff. Hookline strongly suggested against this. I am curious to know why. Is it that the still will produce less efficiently or at a lower alcohol percentage?

[kiwistiller]

From reading over it I wouldn't say hookline is suggesting strongly against that, but more suggesting that DM is... well... full of it. He is no longer with us.

I don't think you'll have any ill effects from building your column like that, but your wallet will probably notice

Just out of curiosity, I'm going to cap my VM and get a baseline for reflux ratio on the next neutral spirit run I do. Should be interesting. I have a two inch port, reduced to 3/4" after elbow. Actually, you can see it in the link in my sig.

[Djurmz]

Kiwi,

Thanks for the reply and explanation.

After doing much more reading and googling that led to the "other site" DM mentioned, I now realize that this discussion had much more context than I was aware of and I was late to the game. I'll now respectfully bow out of this discussion as it seems enough time has been spent on it already.

Now I suppose I had better get back to work...

[The Baker]

big snip

What you is arguing is that it does not matter where in the take off arm the smallest restriction is, it is the sole or at least primary determinant of RR. The results from others indicate instead to me that the take-off port has a disproportionally greater effect than other restrictions in the take off arm. (Presumably due to it being where the vapour has to split and bend 90º to get into the take off arm.)

snip the rest

This gave me a couple of thoughts, not necessarily valid.

1. I think the behaviour of the vapour is the same as the behaviour of a gas, which expands to fill the whole of its container.
Which could mean that the idea of the vapour having to split and bend (thus being somehow restricted) is not valid. Or only partly valid.

2. But if there is some validity in that notion
(perhaps there are physics rules governing flow, as well as those governing the expansion of gases to fill their container, in place here);

has anyone tried , instead of a 90 degree tee, installing a take-off pipe at an upward (I should think) angle half-way between the right-angle that is standard, and the vertical pipe? With a bend to take it to horizontal after a short distance, and then the usual set-up.

AND MAYBE a tapered horizontal tube to the tap; but that is another story.

Because the upward take-off tube would be cut at a (forty-five degree?) angle,
(and it would have to be rounded to the curve of the column?)
the hole in the column it is fixed (welded/soldered) to would be more or less an oval, allowing lots of vapour path.

(Maybe there fittings available for that angle take-off.)

[airhill]

I wasn't going to say anything however

While gasses in a sealed container expand to fill that container I believe we are more concerned with the rules that govern flow. For even vapour splitting the gasses have to exceed a certain speed the Re factor (named after Osborne Reynolds) or turbulence has to be introduced just below the splitting point. Mineme has done a lot of work on the second.
What it gets down to however are there are a number of other variables at play so no one has yet determined exactly how it works; built things that do work yes but explaining why they work as they do no.

[upinthehills]

New here and my head is swimming a bit. So apologies if I put this in the wrong place. I'm agonizing a bit about my still design so reading this and all the other great threads here.

In reading the descriptions of performance of these designs and the mysteries there of - I think a missing link is knowing how much air is in the system around the condensers. I have seen comments that they don't need to be very big ( and that might very well be true ), that only 1/3 of the condenser gets warm. Well if that is true, then only one third of the condenser is being exposed to vapor.

For the VM designs, there are two condensers, both of which may be powerful enough to completely absorb the power of the boiler, am I right on this? If that is the case, you cannot really determine where the vapor will go very well. I particularly don't understand how the product output condenser is "primed". Since it faces downward and the vapor is a high temperature, it will tend to hold air. Only the part which is exposed to vapor will draw more vapor. And even that part will also be happy to draw outside air.

The ratio of vapor distribution between the two legs will depend on the effectiveness of the two condensers compared to each more then any other factor. I think it is air in the condensers that is making the results unpredictable. If there where one way valves to purge the still of air, but avoid pressure build up - I think results might start to look different.

I have some knowledge of operating steam engines and also a heating system for a large apartment building many years ago. So I am not pretending to be knowledgable about stills yet. But I learned these things working with steam a bit. In the apartment building I lived in the basement and many large steam pipes ran thru my rooms. On a 4" main distribution, the distance between hot and cold ( 212 and room temp ) was less then two feet as the system started up. You could feel the steam move right down the pipe, I was always amazed at that. That's what makes me think the output condensers may often be full of air.

Perhaps a small restriction on the vent at the top of the reflux condenser would help flush vapor all the way thru the output condenser? Just a ball bearing or a block of wood might be enough. Perhaps a small pressure and also one way function could be provided by plumbing the vent downwards with 1/2" copper into a 3' or 4' length of larger tube filled with water. That might provide 1 or 2 pounds of pressure, but it would not suck water up such a large straw into the the still?

Or maybe I don't have a clue and need to read another thousand pages Thanks for such a great site!

[upinthehills]

Sorry, one other comment on the same line. Steam engines run a vacuum pump on their condenser to make sure it is charged with vapor. it has only to deal with a very small volume of material, at most a few litres per hour for these stills...

[rad14701]

Welcome to the forums, upinthehills...

To be honest, you're thinking too hard... Don't make the process any more complicated than it is...

As the vapor rises up the column it has a slight pressure variance above barometric pressure... As the vapor rises it looks for a way to escape due to path of least resistance... When the vapor makes its way to the take off branch some of it will split off and head in that direction and the rest will continue up to the reflux condenser where it is cooled and returned as reflux... The vapor that heads over into the take off branch is immediately losing heat so it starts collapsing, as condensate, back into a liquid... The cooler product condenser, usually a liebig, causes additional collapse even though it points downward because that is where the partially cooled vapor is already headed...

I recently posted a drawing of a VM column with two coiled condenser heads, one for reflux and the other for product cooling... I am doubtful that this design would perform any better than a liebig because it is the size of the take off branch in relation to the column and reflux head that determines how much vapor heads into the take off branch, along with the gate valve...

Hope this helps...

[kiwistiller]

I thought that ethanol vapour above about 50% was actually heavier than air??? might be wrong and can't be assed looking it up for the sake of DM's lingering bullshit. anyway...

here in a link to an experiment I ran of which the results strongly suggest that DM was just making up numbers: http://homedistiller.org/forum/viewtopi ... 65&start=0

The other thing to note from that experiment is that a bog standard VM can achieve a RR of ~1:1. That's all you'll really want and if you want to get lower, stuffing a bit of mesh under the coil is a hell of a lot easier than 1 way purge vavles.

Anyway welcome along and good luck with your build! be sure to take lots of pretty pictures along the way.

[blind drunk]

rad -
As the vapor rises up the column it has a slight pressure variance above barometric pressure... As the vapor rises it looks for a way to escape due to path of least resistance... When the vapor makes its way to the take off branch some of it will split off and head in that direction and the rest will continue up to the reflux condenser where it is cooled and returned as reflux... The vapor that heads over into the take off branch is immediately losing heat so it starts collapsing, as condensate, back into a liquid... The cooler product condenser, usually a liebig, causes additional collapse even though it points downward because that is where the partially cooled vapor is already headed...

That rolled off the tongue nicely. I think I FINALLY get it. God it took a while Cheers, bd.

[rc22]

It is very early in my own study of the subject matter, and I have no first hand experience to draw upon while reading the fascinating material on HD.
Less advanced topics have led me over here, into Research and Theory, chasing references within other threads.
The level of debate and my own desire to learn draws me into comment upon RR as presented by the member who has apparently jumped ship.
It appears to me that DM has not considered a critical factor in his hypothesis.
A still creates vapor which expands in the primary section,then is condensed in volume through the mechanics of the device, if operating properly.
His hypothesis could be correct, in that ANY restriction would be the limiting factor, irregardless of it's position within the process, IF the volumetric properties remained static.
They do NOT. The process is not linear.

It occurs to me this why the classic pot still is configured to first produce expansion, then eventually terminates with output of the process from a small orifice. Variations upon that device are refinements for a more specific aspect, or cost effective implementation of materials at hand.

but I don't know nuthin


My thanks to all who have contributed to such a wonderful resource.

[Seb]

I would like to understand but I can't - the more I read, the more it gets complex and goes in all directions. Please someone enlighten me even with a very short answer or even suggestion:

- what is the optimum dimensions of the take-off port on top of the head of a 6 inches column?

- How is it best reduced until reaching both of my 1 inch condensers coils?

- Where should the output valve be put?

- Does the take off have to be oriented downwards?

Thanks.

[bluefish_dist]

A 6" vm is much bigger than most here. I do run a VM on a 6" column, but due to valve cost, I only run a 2" or 3" valve. This is the limiting factor of scaling a VM. The valve cost gets too high.

I reduce the column from 6" to 4" prior to attaching the valve. I do this by adding a T at the top and having VM off one side and a Nixon stone offset on the other. I also have a 3" condenser, this also changes the reflux ratio by adding restriction on the Rc side.

Finally you have to ask if you really need a 1:1 reflux ratio. I don't for most of what I run, but it's one and done runs. If you strip and finish you may need a lower ratio.

[Seb]

A 6" vm is much bigger than most here. I do run a VM on a 6" column, but due to valve cost, I only run a 2" or 3" valve. This is the limiting factor of scaling a VM. The valve cost gets too high.

I reduce the column from 6" to 4" prior to attaching the valve. I do this by adding a T at the top and having VM off one side and a Nixon stone offset on the other. I also have a 3" condenser, this also changes the reflux ratio by adding restriction on the Rc side.

Finally you have to ask if you really need a 1:1 reflux ratio. I don't for most of what I run, but it's one and done runs. If you strip and finish you may need a lower ratio.

Thanks Bluefish. I already read a lot of your posts about your columns. I will use a 6" column by 98" because I plan to make a continuous still for fuel ethanol. The bottom half of the column (49") will be packed with marbles that I can replace everyday to clean since the marbles will get the crude wash, and the upper half will be packed with SS scrubbers as it only gets clean vapours. I will send you the blueprint in MP if you don't mind.
Both condensers will be identical 1 inch worm condensers because of the ease to build them.
I wish to get azeo from wash and I don't know where to aim in terms of ratio.
I read that "the Russians" don't use equal diameters up there. What do they do? Can one side be smaller?
I read that eventhough we are advised to use the biggest valve possible, the actual opening of the valve while running the still is very narrow. The valve is nearly closed from what I read. So, is it that important to have a 4" or even a 3" valve?
Do you use a ball- or a drawer 4" valve?
Because of my worm condensers my refluxed spirit inlet will be lower in the column than it would be if I used a Nixon-stone condenser like you do. Do you see any problem with that?

Thanks in advance

[bluefish_dist]

I would guess a 2" valve will be big enough, but I would put a 4" port just in case. A ball valve will work as they are a lot cheaper. Especially for a continuous column as I would not expect you to make many if any adjustments once it's running.

You will need a pretty high reflux ratio to make it work with marbles if it will even hit azeo. Marbles have a pretty big hetp compared to other packing although at the 6" diameter scrubbies don't work. I am waiting on structured packing for my 6" vodka column. While scrubbies worked great at 4", they did not work at all for 6". I had flooding and could not come close to azeo.

You need a reflux condenser that can drop liquid back at the top of the column. If you don't use all the column height you won't have any chance of hitting azeo. People who use offset columns find that they need a pump to move liquid from the bottom to the top of the previous stage for best performance. You might want to consider a offset column to get more height and then you can have a clean and dirty side.