Have any Questions?


Jun 08, 2022 View:

Wash Dilution

Hi All,

I am in the planning stages of a new distillery and I am interested in the effect of dilution (with quality water) on the flavor profile of the final spirit. I ask this question in reference to the sizing of a still. I want to purchase a still with a large enough capacity that I am not forced to purchase a second still, too soon, as I expand production. This creates the problem that I may not be able to run smaller experimental batches. I see two ways to get around this either dilute a small batch to the minimum volume charge of the still, or produce large experimental batches ( I am leaving out the obvious approach of a small experimental still for capital cost reasons). The larger batch approach seems to be a very costly and wasteful approach. Dilution seems to be a good approach, however I do not know if the distillate flavor profile from the dilute wash will be representative of the corresponding concentrated wash. Any thoughts or experiences will be very helpful to me.

Thanks, Ian


I have been wondering what the difficulty would be in having a small still, say 5 gallons or so, to run experimental batches. Can TTB live with that approach, as long as they know of it?


TTB doesn't care how big your still is, as long as anything over a gallon capacity is registered. Delaware Phoenix was using only a 30 liter for her whole operation until she got that shiny new Christian Karl. At Artisan Spirits they had a 5 liter experimental still. I have a 10 liter and so does Pacific Distillery.

And they don't cost much either. You can get a 35 liter still for under $300.


That's a good question. Sort of like trying to expand a 3 person recipe out to a 300 person feast.

However, from a purely practical/monetary standpoint, if the small cost (less than $500) of a 'test batch' pot still is the holdback of making/buying one, should a person be starting a distillery?

Not downing you on this, just wondering if you've looked at the bigger picture.


A smallish standard alembic still is currently quite reasonable with the exchange rates. And a small still is quite useful for a distillery.

In the big picture, the extra cost is not that much. But also, you can always redistill something if it didn't come out like you want. Plus, the typical whiskey grains are relatively cheap. With a 50 gallon still, you might be able to get three or four 25 gallon barrels of whiskey a month; maybe a little more, depending on your industriousness and availability of free labor. And that still could be your spirit still if you later expand with a larger spirit still.


In my opinion an experimental still is a must tool.

We have a 30Ltr and 100 Ltr still for that purpose but mostly use the 100 Ltr as receipts can vary too much in small scale as 30 Ltrs.

I have done those (the big ones as well) by myself, it’s equipped with a pot and 2 different reflux heads, depending what you wanna make.

Our government don’t care how many stills you have in operation.



Thanks for the feed back. It sounds like the consensus is to operate a small experimental still in addition to the production still. I have considered a small inexpensive setup, however I have the same question as with dilution; Can the flavor profiles be matched between a small experimental pot-still type setup (as mentioned) and a larger (150-250gal) 4-5 plate Kothe or Karl unit? My interest in understanding the dilution affect is that it would allow me to use the same still I intent to use for scale up. Theoretically, it seems that dilution should allow me to get the same flavor profile as a more concentrated wash, at a fraction of the raw materials cost. With the thought being that the majority of the parameters (plate number, copper surface area, reflux, etc...) will remain the same.

Porter: My comment on prohibitive capital cost was assuming a still nearly identical to a production still (150-250gal) but smaller in size (10-50gal) (i.e. $20-$30K), which would hopefully allow me to easily scale up. No worries on getting down on me, I appreciate the criticism. The more this forum can tell me where I am doing things wrong the more successful I'll be.



When we use an experimental still versus the large production still there is always some change in the profile of the spirit, it does however give you enough to judge how the final product will compare/ turn out. I have diluted before- on the small still after the stripping run I have diluted a batch to have appropriate liquid in the still for the final run.


Ian, I think that if you have similar plates and reflux ratio, then the flavor between a little still and a big still will be similar. Maybe not identical, but similar.

But consider what you're doing when you dilute a batch to fill a bigger pot. You're lowering the alcohol concentration in the pot - and hence changing the solubility and hydrogen bonding with cogeners. It's like starting the run half or two thirds of the way through hearts (depending on how much you're diluting). Sure, you can fiddle with reflux and plates to get the same proof out - but I don't think it's a simple substitution, at all.

Take a look at a Water-Alcohol molar XY plot. For a given piece of equipment and settings, your spirit moves a certain distance along that curve. (Or perhaps a certain diagonal across it.) If the only change is moving to lower alcohol in the pot, then the distance your still moves your spirit up the curve is the same, but you start at a lower point. And that curve is really steep at low alcohol. I think flavor changes a lot along that curve, too.

On my equipment (Col. Wilson 20 gal modified with 1 plate a new lyne arm and condenser) and cutting to tails by taste (but tends to be similar proof, ~130) then I end with nearly the same proof gallons of tails no matter how many proof gallons I put in the pot. If I start with higher alcohol, I get more hearts before cutting to tails. If I start low, I get less hearts - but the same tails. Once I tried running a perry from asian pears - very low ABV (lots of sorbitol in pears, rather than sugar) - and I might as well not have bothered running. It essentially started with the hearts tasting like tails.



I did a little research and I see what your saying. Below ~15-20mol% EtOH in wash the volatility of EtOH changes more dramatically with wash concentration than at concentrations >15-20mol%. I believe this is due to hydrogen bonding in a round about way. At low concentrations (high water conc.) the EtOH-wash mixture will tend to segregate to some degree and the EtOH will accumulate near the surface of the wash making evaporation easier, dilution will only exaggerate this affect. I need to look up similar xy plots for the fusels (amyl alchohol, butanol, propanol, etc...). I would expect the H-bonding to be fairly similar for the lower alcohols <C5 and at these low concentrations any effect on EtOH volatility should be nearly equaled in the other primary alcohols (i.e. all have similar increase in volatility relative to EtOH). The other congeners (esters, aldehydes, etc..), however, are likely to have markedly different H-bonding responses to dilution relative to EtOH. At higher concentrations >15-20mol% the solubility of EtOH is increased in the bulk wash (i.e. EtOH dissolves in EtOH/water mixtures better than in water) and the impact of dilution is minimal. So I think your right the flavor profile of a diluted wash will be dictated by the relative H-bonding responses of the flavor components relative to EtOH. It looks like the only way dilution would work as a proxy for a larger run is to collect the whole distillate profile in a stripping run and then do a spirit run, but that means I would probably have to run a couple batches to collect enough high wines to stay above ~20mol% with dilution, which doesn't make sense if my objective is to minimize operational costs. So it looks like dilution is out. Sorry if this is nonsensical ramble I'm still waking up (only half cup of coffee into the day).

Cheers, Ian


Remember that in normal operation you'll be collecting multiple mash runs into one spirit run anyway.

And with your experimental still being small, you may well be able to make multiple runs quickly.


Hey guys,

Just a few thoughts on this topic.

- A test still should be at least 75-100l to get results that can later on be used. Also, the construction of your still should be similar.

- The maximum fill level of a still is usually 50% of the capacity it is designed for (e.g. 150l on a 300l still).

- Plates on modern potstills are designed to work best under 15-20% alc. by. vol. You have to keep in mind that some fruit mashes have less than 5%, and the distillation also works well at this level.

- Stripping and spirit runs are a double distilling technique. Given the plates most people have available, it may not be necessary to work that way. In distillation theory, plates were designed to speed up the process as well as protect flavor and aroma.

Also, I don't think you will need as many test runs as you are imagining.

Please let me know how things progress.

All the best,


Robert Birnecker

Kothe Distilling Technologies.

Award winning handcrafted German engineered potstills for the production of high quality fruit and grain spirits, as well as bioethanol. "Kothe Destillationstechnik" uses patented technology to specially engineer each still with solid quality and energy saving compounds to meet the particular needs of each distiller. Kothe Distilling Technologies is the sole representative of "Kothe Destillationstechnik" in North America, Canada, and Mexico.


Kothe Distilling Technologies Inc.

5121 N. Ravenswood Ave

Chicago, IL 60640 and

[email protected]

(312) 878 7766



Thanks for the input, very helpful information. Out of curiosity, what phenomena (i.e. wall effects, heat transfer, etc...) dictate the >75-100l range for a test still to be scalable? Or is this just a general rule of thumb?

Thanks again for the post, Ian


The chemistry of there not being a "one for one" ratio for size of the unit, will be experienced by those who will not listen. So, just for the record, because I know there are people who misunderstand this info, you must have a DSP license to distill for beverage ready products. It is not the size of the still that parameters the law. Terms such as "experimental still", "test batch", "under 1 gallon" and the like tend to be taken out of context by those who may want to read in their favor. I've actually started to only one time per individual state this during our distillery tours. You look like an idiot to everyone trying to argue with a guest when he says something like "well in my county in (pick a state) there are no rules against it, I've checked". Or how about "My cousin is an accountant for the Weather bureau (NOAA)and he should know!"

Let us always be as clear as our Spirits on this.

Cheers, Bob



I believe I get the gist of what your saying, but just for clarification, to all the readers and posters of this thread, my intention in posting this thread was to gain some theoretical insight into how diluting a wash affects the spirit profile compared to running a more concentrated wash. With the underlying purpose of determining what equipment, and more specifically the appropriate sizing and quantity of equipment, a new startup will need to include in their budget when developing a business plan. In my opinion this is necessary information to understand upfront, primarily because the experimentation/product development phase can not be done in advance of starting a business and obtaining a DSP, therfor this work will need to be done in parallel with production (i.e. developing a whiskey recipe, while producing vodka), to do this work in serial (i.e. experimental/product development then start production) will require a lot of operational expense without the required income. When I use the term "experimental still," "Test Still" or "Small still," etc... I am referring to a second still operating under a DSP (which I believe in California would also require it's own "Still" license code 06). I believe most of the posters here understood my intent, however for those who did not I hope this clarifies the intent of this thread.



Hey Ian,

To follow up. The 75l - 100l is essentially based on the amount of mash you can put in vs. the amount of alcohol coming out for proper separation and actually having something to work with.

I guess the answer would be, it's a general rule of thumb.

All the best,