I'm using a nice clean malted/unmalted barley wash to make vodka, by three distillations - IE - Strip to 30%, combine low wines and run through 3 plates doing a basic heads/tails cut resulting in 160 proof hearts, then running the hearts through my smaller 20 plate still. I do the hearts cut because it is easier to then run the final cuts through the 20 plate without doing any real cuts, since it is all really clean distillate. It amounts to really expensive vodka, but it is really nice.
When I then take the heads cut and run it through the 20 plate reflux, I would think that I would be able to really separate out any off flavors/chemicals from the heads, but alas, I can't. After the ethyl acetate smelling initial fores, I get a good portion (20% of the total batch?) of distillate that has a nice smell (vaguely sweet smelling, almost caramel/butterscotch aroma) but leaves a numbing burning sensation on the palate. That is essentially how I'm making the cuts - IE - once the numbing burn is gone, I proceed to the hearts cut. It seems like I should be able to recover a large portion of this ethanol in the heads using 20 plates and somewhat diluted second run low wines.
What products are remaining in this ethanol mixture that cause the numbing feeling/sweet aroma and why are they so hard to separate out by refluxing? Will carbon filter some of those out?
When you are rerunning these heads, what is the abv charge of the still? Depending on the efficiency of your 3 plate, 20% of LPA for heads of heads doesn't seem outrageous. When we would do similar of running heads through the vodka column we probably took 10-12% as heads of heads to destroy.
I'm re-running the heads thru the 20 plate pretty concentrated - around 50%. And most of the original 20% cut comes out unusable - IE, it all has the sweet, butterscotchy, numbing mouthfeel. I suspect that it is mostly ethanol, but essentially unusable I suspect... Seems like a lot of heads to throw out! Perhaps running it at a lower concentration would help?
Reply:On 12/29/2016 at 10:11 AM, fa20driver said:
Sorry, I was misunderstanding what you said in your original post. It would be helpful if you posted specifications for your equipment and process. So 20% of your 2nd distillation is heads, and when you redistill that none of it is usable to you?
It is very strange sounding to me that you wouldn't end up making a heads cut when running your 'heart' through your 20 plate column. Even if you take a generous heads cut on the 2nd distillation, there will still be heads components smeared into the hearts when using only 3 plates (unless your still works very differently than ones I've used), and you should be able to separate those out with the 20 plate. That suggests to me when you are rerunning the heads through the 20 plate, the still isn't calibrated correctly and you are not getting very good separation.
Lowering the charge strength to something like 25% when running through the 20-plate still would likely also help with getting better separation.
Thanks guys for the feedback. To be more specific - My basic process is as follows:
"Stripping run" through a 200 gallon, 8 inch column with 3 plates - using minimal reflux - no cuts
Second run through 3 plates at about 30% ABV charge, cutting heads and tails using a good amount of reflux to end up with about 160 proof "hearts" - Saving heads and tails to process later. (Tossing anything that smells like acetone, and saving as heads anything that has the butterscotch, numbing feel that comes before the hearts cut)
Third, I'm taking a small portion of these hearts and running through my 15 gallon, 4 inch, 20 plate column still to achieve 190 proof for Vodka - My process on the 20 plate is to go full reflux for about 30 mins, then slowly adjust dephleg to achieve 190 proof (measured with calibrated instruments.) I then collect and taste to separate.
Yes, you are correct Andy, it is still necessary to make a heads and tails cut on this third run. I didn't think I would need to, but after doing this several times, it appears that they are still present in small quantities. They only make up about .05% of charge yield, but they are present.
Here is where my confusion arises...
When re-processing the second run heads cuts (that are amounting to about 20% of the total collection), I am running them through the 20 plate still at about a 50% ABV charge and not getting much of anything that I would want to include in my vodka.
Without much official background in chemistry and distilling, (other than some basic organic chemistry in college,) I had always assumed that anything that the "heads" were comprised of would easily be separated out with 20 plates, and that almost all of the ethanol would be recoverable from these heads as "pure, usable" alcohol. When people speak of tossing out heads, are they really tossing out 15%-20% of their theoretical wash yield combined with a few non-desireable chemicals, or am I being over-selective/doing something wrong...
Thanks Odin, that is an interesting hypothesis - and thx for the posting the diacetyl reference.
I suppose it could be diacetyl...
I'm pretty much using beer fermentation techniques to produce my wash, including a short pasteurization (boil) prior to chilling/oxygenation, followed by a very healthy pitch of clean house yeast from a beer conical (Americal Ale II). I plate my beer yeast regularly on LMDA to test for acid producing bacteria, so I don't think I am getting diacetyl from spoilers, but I do use an open fermenter for my distillation washes.
The only other thing I do differently from my beer making is a higher ferm temp (up to 85 F,) and I also add some gluco-amylase after a few days of fermentation to get the last few points of otherwise un-fermentable sugars from the wash. Usually higher ferm temps with my house yeast result in apple estery aromas/flavors rather than butterscotch.
Perhaps what I am separating out as heads may be perfectly drinkable ethanol, but I certainly don't think I want it in my Vodka, especially since it also has something that results in some numbing on the pallet.
Perhaps I need to just send a mid to late heads sample to the lab and see what is actually in there...
I'm pitching live yeast slurry from Beer conical fermenters from our brewery. I'm not counting cells but I can tell you that I am overpitching. On my last batch of 14 bbls of wash at 16 plato (434 gallons) I pitched about 20+ gallons of live yeast slurry. I'm guessing that is 2x or 3x what I pitch for beer. I did a quick search on over pitching yeast, and the one thing it may lead to is off flavors from an over abundance of dead yeast cells (autolysis).
I sent an email to my yeast supplier to see what flavors they think would result from over pitching this yeast strain and fermenting at a higher than optimal beer fermenting temp.
From my limited study, higher than recommended beer fermentation temps = more esters (not so good for beer, but good for whiskey, questionable for vodka), possibly more fusels? Not sure about autolysis, but it's not sitting around long enough for that to happen (5 days ferment, 3-4 days to distill) I taste the resulting wash, and it just tastes like I would expect an unfiltered, unhopped 7% ABV beer would taste - no observable off-flavors.
Quote from my yeast supplier when asked what flavors/aromas I should expect in the beer from fermenting American Ale II at 85F and over-pitching:"I've never tried fermenting that yeast at such a high temp so I couldn't even guess,
Generally, in beer wort, dramatic over pitching can lead to decreased flocculation, rapid fermentations, and a higher proportion of mother cells to daughter cells."
Reply:14 hours ago, fa20driver said:
Just a comment from the peanut gallery.
Concur on the fermentation temperature and yeast suggestions, but I will suspect that it will manifest itself as a slightly wider hearts cut on the 3 plate distillation. So yes, an increased batch yield, but I don't think it is going to significantly improve the ability to recycle the heads and extract additional ethanol using the 20 plate column.
The concept of being able to 'recycle', I think, has it's roots in pot still distillation, where the heads/tails/feints are at a higher alcohol:congener ratio. By the time the heads cut is being made in this example, they've already going through the equivalent of 5 "distillations" worth of compression. How many plates are required to further fractionate that already tight fraction? Clearly the 20 being used in this example isn't enough. We're talking about congeners that have a odor and flavor detection thresholds that are absolutely tiny. If we were talking about redistilling the heads from a single pass pot still distillation, I think the discussion would be very different.
Based on @fa20driver original post, if 20 plates couldn't clean the heads up before, I don't think it's going to do it after a cooler ferment with a cleaner yeast either.
Realistically, if increasing the product yield is the goal, the better approach is to perhaps take a wider/dirty hearts cut on the 3 plate, and use more passes on the smaller 20 plate (with cuts) to yield the finished vodka. Yes, it's more work, but if you were running an additional "head cleaner" distillation, why not run another product distillation instead?
I think @JustAndy suggestion of further dilution might be a good approach, but if that means having to make 2 or 3 more "heads cleaning" passes on that small 20 plate, I'd still argue that you are better off making additional product passes on the little still, and just dumping the heads.
I gotta say though, kudos on your very novel approach. I had to read the thread 3 times before I finally understood it. It seems like it would be an interesting way to make a vodka while still carrying through a considerable amount of character.
Thx silk - My current license limits me to using locally produced ingredients. That combined with the existing brewery equipment currently in use is why I am using barley. I was never much of a Vodka drinker, but this vodka is pretty tasty all by itself. It is also pretty time consuming and expensive to make. I will take Tom's suggestion on a lower ferm temp and see if I can't get a bit less congeners in the mix.
Regarding doing more runs on the 20 plate with a wider heads cut - I get what you are saying regarding yield vs time. The 4 inch column is slow, so I like to do as quick and small of a heads cut as possible so I don't have to babysit it as diligently. I can run the 20 plate while stripping with my larger still and still make beer in the brewery, so as cumbersome and inefficient as it seems, it actually works.
I was hoping I could get more use out of the heads other than as sanitizer in the brewery or fire starter. I'm thinking that the tails might make an interesting malt whiskey, otherwise I will just carbon filter them and re-distill for a gin base -
I think the bigger concern is going to be a change in the character. If you and your customers are liking what you are making - going to something like EC-1118 fermented in the low 70s is going to be very different from an Ale yeast in the mid 80s.
Given you've already got the yeast from the brewing, I think you are on the right path, try dropping the ferment temps with the ale yeast.
Also, if you are fermenting in your conicals. Try cold crashing for a day before distillation to get all that yeast to floc and fall out. That'll give you a slightly cleaner distillate (fewer volatile esters).
While changes to your fermentation scheme could certainly help, to me the issue sounds like poor performance on the 20 plate still. For vodka, we stripped wheat mash through a 5-plate brandy column with a small forescut (at about 90%), and redistilled that on a 20 plate CARL column. We still needed to take a heads cut during the vodka run, and even after the heads cut it was possible to produce clean 95% and dirty 95% all depending on the configuration and parameters (plate loadings, dephelg input, take-off speed, heat input, still charge). We would redistill the heads through the 20 plate, and after a generous heads cut were able to recover out about 70% of the total alcohol in that heads redistillation as usable (although very different from the core vodka produced, and use primarily for heavily flavored liqueurs).
Who made the 20-plate still you are using? In my experience small stills are extremely touchy to get working consistently, and many are poorly/incorrectly designed. Before investing a lot of time and cycles into learning that equipment, I would suss out how practical it is to make vodka at that scale.
Thx guys - all good suggestions.
JustAndy - The 4 inch 20 plate is a modular stainless 'still dragon' design with copper pro-caps. Probably not the best out there, but I think that they have a proven concept. I'm still pretty new to the distilling game so I will try re-running my heads a bit slower with a lower ABV charge next time to see if I can clean them up. I am getting decent alcohol without any objectionable odors/flavors, but it just isn't the same as what I am getting from the hearts. That, and it leaves a little bit of burn on the pallet. Perhaps that is the best that I am going to get from the heads. You mentioned that you were able to recover a decent amount of alcohol from your wheat heads, but it wasn't the same as what you would put in your main product. I think that is what I needed to hear. I had assumed that any product run through enough plates should taste the same - but I think that I am finding that is not the case. I also need to just take a sample of my re-run heads and let some other experienced distillers taste them to see what they think.
Regarding the economics of running that size still - I can appreciate that concern and that is good advice for anyone investing in this business. I am able to run it at the same time as my stripping still and still brew 600 gallons of beer all at the same time - so time-wise it works. I sell most of the vodka over the counter both in drinks and in bottles to go and locally self distributed - so the numbers work. I certainly wouldn't be sending any to a distributor any time soon though, although even then it is more profitable than sending out tap handles and kegs that might not ever come back...
Silk - when you mention cutting down esters for Vodka by dropping the yeast - would you say that those same esters (from yeast) would be beneficial in a malt whiskey? Under pitching yeast usually enhances esters - I have heard (at least in beers like hefeweizen) so I am wondering if my over pitching yeast for whiskey is counter-productive... I am aging the malt whiskey in 30 gallon medium char barrels for probably 2 years..
Reply:34 minutes ago, fa20driver said:
I think there are two things happening here. First, as has already been mentioned, your fermentation is outside your yeast strain's "comfort zone," and it seems to be expressing more head compounds than necessary. That is typical of heat-stressed adult yeast cells. Most beer strains that are happy at 65-75 will express elevated acetone, acetylcholene, and pentanes at 85. If time isn't a factor, try your ferments in the happy place.
Second, your vodka still is having trouble with lighter fractions. The still dragon is a great piece of equipment, but the column is a bit undersized, as is the volume between plate stages. I suspect that you may be sustaining vapor velocity above the level that your liquid-vapor interface can sustain. If you have a dephlegmator, try using it to choke the vapor velocity down. If no dephleg, you will have to use the heat in the pot to slow down your distillation. An extended equilibrium period may also help to concentrate your light fractions. Dilution *may* help, depending on the exact composition of your head fractions, and it will certainly help slow vapor velocity, and increase your vapor pressure differential.
While Odin is right that diacetyl is likely your butter culprit, your high fermentation temp suggests that is not the case. High levels of methyl ethyl cetone can have an oily feel on the tongue when tasted in conjunction with some alkenes. I have heard some people refer to this as butterscotch, although it's more of a mouthfeel than a taste. While it's more common on stonefruit fermentations, it can certainly present with barley.