How I Run My VM/LM Still

[wv_cooker]

Or how about a small thumper under the column such as Larry's instead of a bubble plate. I believe I remember seeing where lester stated the he could see the line in his. So place the overflow back to the boiler at that level. Wouldn't this allow them to continually be returned to the boiler while also keeping the thumper from overflowing?

[flyingdutchman]

So Looking at Edwins first graph "Relative Volatility of Cogeners and Ethanol" , to avoid confusion here is what I believe am reading .....
1) The X axis is the percent of ethanol (EtOH)
2) The Y axis is the volatility
3) The actual line at Y=1 is the relative volatility of EtOH
4) The other lines represent the relative volatility for each individual cogener
5)The difference between the Y value for EtOH and a given cogener is the relative difference in the volatility between the two. If we call that dy and we use the absolute value than the smaller the value the closer it is to EtOH and I think the harder to separate from EtOH

If that's true then yes it appears that below around 40% EtOH some of them are indeed more volatile and it alters what I typically perceive to happen in our distillations.
But then what about volatility at 60% 80%, 90% ? How close are they to the volatility of ethanol, what is dy?
60% propanol<methanol~= i-butanol<<i-amyl
80% Acrolien (v close to ethanol)<propanol~=methanol<i-butanol<<i-amyl
90% methanol ~=propanol<Acrolein<i-butanol<<i-amyl

Which brings me to DAD's (not my dad LOL) statement

I still ask, rather than bleed the fusels off, what about trapping them in a bubble plate at the bottom? Or does that just continue to smear them across the distillation?

We don't build special takeoffs for methanol so I would ask what do we gain for building one for fusel oils that for the most have a volatility with dy equal to or greater than methanol once we get above ~2 plates of separation in our columns. What if we just run the column in such a way as we establish a good equilibrium and maintain that equilibrium by not pulling product off at a rate greater than the column can support.

Now if we want to remove them how do we pull impurities off the lower plates (or packing) early in a run without taking all the ethanol with it. Is there some physical property such as vapor density we can exploit to do this? A 9% wash with 1 plate (and our boiler is ~ 1 plate) jumps to ~52% already the ethanol is coming off first and very close to the fusel alcohols. Seems to me that to exploit the difference in volatility we would prefer to be doing so in a range of less than 20% where dy is the largest.

ps here is a link to one of the books Edwin mentionshttps://archive.org/details/distillationprin00younrich
and if I remember right he mentions salting out fusel oils

[Odin]

Some use a heads trap. For instance a bubble ball above take-off. Same principle can be used for a tails trap. Only at the bottom. And with a drain. Biggest question will probably be when to start draining ... Too soon would be counter productive.

Odin.

[flyingdutchman]

Hi Odin,
Good Idea. I don't use a heads trap and am not 100% on how it works so it would be cool if someone who does could tell us..
1) How does a heads trap work vs regular column (beyond just it accumulates heads) and what does it bring to the table
2) Why does a heads trap work (with respect to volatility)
Then what is the same and what is different about the fusel oils vs ethyl acetate.

That brings a different question to mind on the graph ethyl acetate has a dy greater than methanol so by that reasoning should it would think it comes off before methanol?

ps that book is pretty cool it has some great info!

[Odin]

Heads are lighter and accumulate as high in the column as possible. If you can trap them in a heads trap, you can bleed them off. That's the idea. The criticism, and rightfully so, I feel is that those heads, wherever you collect them (trap or top of normal column) will pass the take-off anyhow and so by definition cause some smearing ...

That's what I know about it.

Odin.

[flyingdutchman]

I don't know but I would like to ask if a heads trap was intend for a continuous distillation setup with the idea of removing some percentage of the heads(some being better than none). Only because it always seemed to me that ( for batch distillations) we could just forego the trap properly equilibrate the column. Let them all collect at the top then just pull them off and discard? But again I don't really understand them which is why I asked.

Now the fusel oils seem different because below 20% they seem to be more volatile but above say 40% they come are less volatile. Infact they seem to sort of be trapped on the lower plates. Could we some how early on cause them to separate in the liquid of a plate (like let them collect than dilute the plate with water) and pull them off in a modified florentine flask/apparatus? thus taking advantage of their physical property of solubility? Personally I am not inclined to try this just thinking out loud.

I would still love to hear any ideas about the volatility of methanol vs ethyl acetate in Edwins chart

[Odin]

The practice of professional distilling and scientific research by Holstein shows that Methanol, even though it has a low bp, comes over in relatively constant quantities during the whole run ... That's very contrary to what we think here on the forum. Now, FDM, you are better in understanding that chart than I am ... does what I write (and Holstein says) makes sense, if you put the information from the chart next to it?

Odin.

PS: I think in general heads and tails bleed offs at higher and lower plates is something suited for continuous distillation. Not sure it has a practical benefit in our batch like approach ... but discussing the theory is mighty interesting!

[Edwin Croissant]

I will try to comment on the all the posts where appropriate and answer any questions but that will take some time

Well Hats off to Edwin he had exactly the image I was looking for, here is his post,http://homedistiller.org/forum/viewtopi ... =1&t=42227

Thank you

Edwin,
From your research, what would be the best TEMPERATURE (hello Dad300!) to bleed off the Amyl Alcohol/EtOH/water mixture? There's no other indicator for when and where to bleed off so we have to do it by temperature, and possibly smell. This time we use temperature as an indicator for impurity, not purity. (No way I'm gonna taste that nasty stuff.)

Ernest Sorel mentions a temperature of 80 to 81 degrees Celsius in his book “La rectification de l'alcool”. In this post (use the Google Chrome browser for an automatic translation) a reference is made to the book "Technologies of alcohol" by Jarovenko stating a temperature of 81 to 82 * C. I think that the whole topic is interesting as Igor Dibrov is trying to automate the fusel oil take-off. He posted several movies on YouTube showing the construction of his device. Unfortunately I do not speak Russian. Are there any members on this forum who speak or can understand Russian? I sure like to know what he is saying.

Now, I've read Edwin's post and I still ask, rather than bleed the fusels off, what about trapping them in a bubble plate at the bottom? Or does that just continue to smear them across the distillation?

A bubble ball seems to be very effective in doing this. And it don't have to be a bubble ball. There is a discussion about this on the MD forum. Last week jokkman reported that after mounting his Gin basket with a diameter of 90 mm and a height of 290 mm and filled with scrubbers under his 35 mm column, the take-off rate almost doubled and his tails where almost gone.

We don't build special takeoffs for methanol so I would ask what do we gain for building one for fusel oils that for the most have a volatility with dy equal to or greater than methanol once we get above ~2 plates of separation in our columns.

During normal fermentation only a very small amount of methanol is produced, so there is no need. A few drops of fusel oil can ruin a 600 liter pipe of good brandy. (page 468, A Complete Treatise on the Art of Distillation published in 1830).

What if we just run the column in such a way as we establish a good equilibrium and maintain that equilibrium by not pulling product off at a rate greater than the column can support.

Drip ... long pause ... drip ... long pause ... drip ... wife complaining, kids complaining, dog needs a walk, thrashcan needs to be put on the street, house needs maintenance, i think the dog has a better life then me

I would still love to hear any ideas about the volatility of methanol vs ethyl acetate in Edwins chart

I understand that ethyl acetate is concentrated in the fores due to the high volitility. After the fores it is gone from the boiler. That is the difference with continuous distillation where the is a constant supply, hence you need a extra column for removal. With a boiler methanol will build up in the top of the column during full reflux. Methanol is present in the vapor but not in the quantity we expect. When it is in the column it will get trapped at the top. So if you put the still long enough in full reflux most (? let's be careful it is poison) of the methanol will be transfered to the top of the column. I think that is the reason that there is a peak at the beginning in a graph showing the methanol concentration during the run. Some methanol accumulate during the stabilizing phase at the start of the run.

[flyingdutchman]

Hi Edwin

Drip ... long pause ... drip ... long pause ... drip ...

yes I have played that game and its like watching grass grow (or golf)
So I do advocate faster collection rates
I looked at the Russian posts (no don't speak Russian) Its very interesting and they don't seem to have any form of collection plate/bal/cup, just T tapped straight into the column. I would love to see a GCMS analysis of what they did collect. Perhaps some but not all fusel oils?*** In their glass can you tell how they separated out? Is it the organic layer on top with the aqueous layer on the bottom (do I see a distinct brown layer there?)
I definitely prefer Jokkman's approach. I believe from his post (if I read it right) he is also adding some ~11 inches of packing to a previously ~ 31" column which could account for some extra separation. Again he may be catching distillate that otherwise would have been returned to the boiler increasing his efficiency (which I also like) Now if it separated out with the organic layer on the bottom a small tap at the bottom would allow for easy draining of just the fusel oils/alcohols. My guess is its the aqueous on bottom but a Florentine like flask drain would let you just pull off the top layer.
What if during equilibration and heads collection the "cup" was allowed to fill. Then halt collection and drain the cup. Or add water to the cup forcing the fusel oils out of solution and collect?

*** check out the fusel oil lab analysis by GCMS herehttp://homedistiller.org/forum/viewtopi ... =Fusel+oil note the propanol, i-amyl alcohol,i-amyl acetate & i-butanol in the foreshots.

ps I heard some where that the biggest problem with methanol is its metabolite.

[flyingdutchman]

So I do advocate faster collection rates

However I would like to put that in context.
It has been demonstrated that RR has a definite effect on the HETP of any packing.http://homedistiller.org/theory/refluxdesign/rr
It has also been worked out the importance of RR ( The Compleat Distiller , Nixon & McCaw)

Reflux Ratio
In this book, we define the reflux ratio as the ratio of the final product returned to the column to the
total delivered to the condenser, and product ratio as the ratio of the product collected to the total delivered.
So the reflux ratio (F), equals L/V, and the product ratio (1-F), is D/V.
If you collect all the condensate, returning nothing, then F = 0. If you return everything (operating under "full reflux") then F = 1.
The balance equation becomes Y0 = F * X1 + (F-1) * XD
Since our aim is to get as much of the volatile component as possible, let’s set XD = 1.
The equation reduces to Y0 = F * X1 + (F-1)
In this equation Y0 is the mol fraction of volatile component in the vapor leaving the boiler and X1 is the mol fraction in the liquid being returned to the boiler.
Our aim is to get all of the volatile component, so X1 = 0.
We end up after all this with the simple relationship:
F = 1 -Y0
This equation shows that when the concentration of the volatile component in the boiler is very low,
producing vapor with a very small Y0, F needs to be very close to 1 in order to achieve our aim (which
means that you'd get hardly any product D). The higher the concentration in the boiler, the smaller F
can be, meaning you need less reflux.
This is the reason you need reflux in a column in order to get a pure product. A reflux column on its
own will provide some of the total reflux required, but not enough to achieve complete separation. In
practice, a reflux ratios in the range of 8/10 to 9/10 give good results with a column containing around
80cm of packing.

My point is just that I see a column as device used to separate various components in a mixture based upon their physical properties (volatility). The wash we produce has multiple components. Its more than just separating water and ethanol. As we have discussed there possibly exists acetone, methanol ethyl acetate, aldehydes, and possible isomers of propanol, butanol and pentanol. As evidence I point to the various GCMS analysis I have posted links to here.
If we push our collection rates too high we do so at the cost of effectively separating these components. I don't mean we need to collect at 1 drop every second. BUT I believe shooting for reflux ratio's that are exceeding low does not allow a fractional distillation column to do what it was designed to do.
Honestly I don't think we can quantify distillate purity buy using %ABV as a bench mark.

I am not knocking peoples efforts to improve their collection rates. I do applaud efforts to increase output without the quality of our product suffering. I love some of the new designs
I believe we can afford to take a hit on HETP (by dropping RR) if we use packing with a greater efficiency or just a longer column (within reason).
I really like the idea of catching distillate before it drops back down to the boiler
If we can efficiently trap some of the fusel oils and funnel them off I am all for it
We can use a large diameter column
But I do believe there is more to our separation than those fusel oils that we can trap. Personally I have doubts about how effectively we can separate multiple different volatile components with a RR where we are collecting way more distillate than we return down the column.
Perhaps there exists a middle ground somewhere greater than 1 drop a second where we can get the best of both.

[Lester]

Honestly I don't think we can quantify distillate purity buy using %ABV as a bench mark.

I agree. What do you suggest we use as a CHEAP bench mark that is readily available worldwide?

[Edwin Croissant]

I believe from his post (if I read it right) he is also adding some ~11 inches of packing to a previously ~ 31" column which could account for some extra separation.

I like this kind of interaction . Your comment made me reread his post and I noticed that he filled his gin basket with 7 scrubbers, side by side about 15 cm high. If the scrubbers are on the bottom of his gin basket there is an open space of 140 mm between the top of the scrubbers in the gin basket and the bottom of his column. That is 0,89 liter. I have asked him to confirm this.
Amyl alcohol demixes above a concentration of 3%, if I recall correctly, and as amyl alcohol raises the boiling point, the sudden drop in boiling point of the remaining ethanol water mixture causes a column hiccup. In the industry we use hydraulic accumulators at the end of a pipeline to prevent water hammer. I think that the open space between the two sections of packing isolate the column from the violent action below in a similar manner.

What if during equilibration and heads collection the "cup" was allowed to fill. Then halt collection and drain the cup.

That is what i was thinking too, when you dump the content of the column in another container during equilibration you take the fusel oil out of the process.

What I like to know is if the pressure in the boiler rise when the column is saturating with fusel oil. Savalle automated the steam supply to his stills after an explosion of his experimental still in 1846. The automation was based on the pressure in the column. Maybe a similar approach could be used to control the fusel oil take-off.

[Odin]

But I do believe there is more to our separation than those fusel oils that we can trap. Personally I have doubts about how effectively we can separate multiple different volatile components with a RR where we are collecting way more distillate than we return down the column.
Perhaps there exists a middle ground somewhere greater than 1 drop a second where we can get the best of both.

That sums up very nicely how I feel about it as well, FDM.

Odin.

[flyingdutchman]

Hi Lester
I am not an analytical chemist. However from what I see in different articles it seems like GCMS is the standard tool. The thing is I suspect we are looking for mg's per liter or perhaps even less. One of the links I posted about fusel oil analysis was from a wine maker on HD. He implied he has done this before for his wine also. I believe Manu also posted a GCMS analysis of his vodka and it may have been less pure than was expected.
Honestly I think we can forget C-18 reversed phase chromatography methods (like HPLC, or LC-MS) because everything is just too polar. Perhaps an SFC-MS.
This really leaves GCMS by some local testing lab if your lucky enough to have one (not the answer you wanted I know ). Perhaps in the range of $30-$50 would be a guess? I wouldn't do it unless I had a good reason to, like a new design or just easy access. However I would guess that a distillery probably has to, to prove purity on some sort of basis?

Awhile back people here we're trying to build 2 inch plate stills. They worked well at separating water and ethanol but the products all had a really bad taste . Which just shows that percent alcohol does not always correlate with purity. Smell and taste can be exceedingly good and no doubt some industries employ specialized people just for that purpose. Yet working in a lab some things do have a LD50 below our threshold of perception. Other times some people just have lower levels of perception. My friend can't smell trifluoroacetic acid which to me has a very strong smell.

I do use thermo well temps and alcometer to asses my distillate composition with respect to alcohol and water (I just don't assume its a purity measure). Now could I be off a little on my percent alcohol sure. However my guess is that the impurities in there are of a low enough concentration that my readings are not too far off.

[flyingdutchman]

Hi Edwin,
His gin head interests me. I see plenty of pictures from the outside but none on the inside. He has a pipe passing through the bottom(from the boiler) but is it flush on the inside or does it extend upwards? If it extends upwards I wonder how high? (ie how much distillate will it hold?)
In his post you gave the link to, he calls it an alternative bubble cap. How is it rigged to behave like a bubble cap.
Also given that the tube leaving (up to column) is aligned with the tube entering (from boiler) how does he prevent the distillate from just dripping straight through?


ps What about some form of manometer on the gin basket to detect a bump in pressure due to a sudden drop in boiling point?

[flyingdutchman]

Hi Edwin,
Maybe something like this? The inner tube is 2" and extend upward for 1/2" into the chamber to retain some fixed volume of distillate . It could even have a small conical cap like a cap on a stove pipe to prevent distillate from dripping down from above back to the boiler. The outer wall is 4 inch tubing (but could also be 3") it should probably be say 6 to 8" tall (not as drawn)?. The drain is 3/8 tube.

front.jpg (5.63 KiB) Viewed 3694 times

top.jpg (11.55 KiB) Viewed 3694 times

[DAD300]

A. Does anyone feel that the fusel oils can be passed out the normal takeoff and some better ethanol collected afterwards?

Why I ask...Once I start getting tails, I used to stop. But a few times I have collected in 250ml cups into the deep tails...at about 70% I get a cup of obviously oily blue tinted crap and then it clears up. The tails that follow are strong, but not as bad as that 250ml of crap.

If I catch the oily crap I'm talking about I save the rest for feints.

B. Wouldn't a properly cut or ended stripping run trap these components?

[Edwin Croissant]

His gin head interests me. I see plenty of pictures from the outside but none on the inside. He has a pipe passing through the bottom(from the boiler) but is it flush on the inside or does it extend upwards?

I have studied jokkman's pictures and I see a brass pipe with a kind of flange soldered to the bottom of the container. The flange does not stick through. He reported that the scrubbers are standing like “hot-dogs in a can” on the bottom and that there is indeed a 140 mm open space above them.

Your design looks good and I think it will need a conical cap. I presume the 2” tube contain packing.

Just thinking aloud: Lester observed that the fusel oil is floating on top so there is definitely demixing going on. Mount a small float operated valve that discharge the bottom of the container to the packing underneath at a certain level. Extend the take-off to this level. Then you can decant the fusel oil that is floating on top.

My idea is to have about 10 cm of packing (about one distillation?) at the bottom of the column, then 5 cm open space and then the rest of the packing. Make a small vapor take-off in the 5 cm open space section with a small valve and run the vapor through a small condenser. Maybe connecting a silicon hose to the take-off with the other end hanging in cold water will also work by pulling the hose slowly out of the water until the end start to bubble. I like to keep things simple


What type of 3D software are you using? Your drawings sure look good I did a lot of work with Autocad in the past. I tried Blender but I don't even recognize what I am looking at with this program.

[Lester]

A. Does anyone feel that the fusel oils can be passed out the normal takeoff and some better ethanol collected afterwards?

Why I ask...Once I start getting tails, I used to stop. But a few times I have collected in 250ml cups into the deep tails...at about 70% I get a cup of obviously oily blue tinted crap and then it clears up. The tails that follow are strong, but not as bad as that 250ml of crap.

If I catch the oily crap I'm talking about I save the rest for feints.

B. Wouldn't a properly cut or ended stripping run trap these components?

A. With the way I run my still, I don't have much good alcohol remaining in the boiler when I stop. I use lots of reflux to keep the tails down near the end of the run. If I collect continuously beyond what I know is good stuff then the nasty tails come up. I have collected these and it wasn't much, about 50ml only. Very oily and stinky and beyond that it is mostly water that comes out, boiler temp >99 deg C. Had to clean my column afterwards so I don't do this anymore.

I have no answer for "B".

It looks like we now have 2 topics on this thread.

[DAD300]

The last time we talked about trapping the fusel...the easy answer was a stripping run that left most behind...then the spirit run cut with clean water trapped even more.

I'll take this to another post...

[Edwin Croissant]

Very oily and stinky and beyond that it is mostly water that comes out, boiler temp >99 deg C. Had to clean my column afterwards so I don't do this anymore.

Looking at the vintage stills with those two story high columns during my visit to the National Genever Museum in Hasselt, Belgium, I realized that cleaning such large column is a pita and then I understood why Sorel en Savalle did so much research to find a solution for the infestation of the column by the fusel oil. Stills at the end of the 19th century had a boiler up to 40000 liter and a column several stories high. Sorel investigated the effect of the depletion of the boiler content and found that at the end of the run the fusel oil contaminated the upper plates of the column. He concluded that squeezing the last alcohol out was not economical feasible. The cost for cleaning was more then the value of the alcohol.

Btw, I like your idea of measuring the temperature with bare diodes, I think that a fast response useful. How do you get the wires out, do you use glass or ceramic feed-through capacitors?

[Maritimer]

Hi Edwin,

I'm thinking of using Lester's thermometer, too. I have some 1/2" dia teflon rod that I was thinking of using. Maybe make a bung with two small holes for the wires to go through.

M

[Lester]

Btw, I like your idea of measuring the temperature with bare diodes, I think that a fast response useful. How do you get the wires out, do you use glass or ceramic feed-through capacitors?

I just solder twisted-pair cable onto the diodes. The wires that are exposed to vapor are bare. This assembly is wrapped with PTFE tape (only for sealing, only where the wires come out) then forced-fit into a 5mm copper tube. This forms the sensor assembly which then goes into 6mm thermo wells in my column.

No need for feed-thru caps. The circuit is immune to noise.

[flyingdutchman]

Hi Lester

It looks like we now have 2 topics on this thread.

I have to agree and apologize. I am afraid a lot of what I have posted has drifted pretty far off topic. For what its worth I really like your design and your ideas! I hope to see more of them.

Perhaps a new thread more specific to Fusel Oils, what is known and how we might deal with them is more appropriate.


FD

PS DAD beat me to it...