[I wrote a detailed description of operation from new sources here:]
Well warranted excitement about a whiskey made on a three chambered still recently crossed my desk. What caught my eye was talk of oils. There were a few different lines from David Driscoll regarding the Leopold Bros. whiskey:
It ends up extracting two things en masse: flavor and oil.
Oil? Radiant oil?
It is absolutely unreal in its aroma; it’s a complete class of its own. Then you taste it and the oils completely smother your senses, leaving your taste buds saturated in oatmeal cookie dough and sweet rye character. I’ve gone back three or four times with the sample I have and every time I’m more excited than previously. IT’S FUCKING INSANE.
Oil? Sounds like that dude got a dose of Damascenone, a rose ketone.
Driscoll’s excitement left me pondering what may be special about a three chambered still. Hopefully I can build a case that these are basically rum oil stills (and that’s a big deal). They are all about chasing β-carotene derived aroma by investing in time under heat. Pursuing and achieving these compounds will unlock a new era for fine spirits. Something higher value will be found than the cost of the fuel.
I’ve read great bits over the years from David Wondrich describing three chambered still operation:
The three-chamber still usually took the form of a tall, tapered column built out of cypress— or cedar-wood staves and hooped with iron (the Leopold Bros.’ version is made of copper, as some were then). Inside, it’s divided by horizontal copper plates into three compartments. The bottom one has a pipe going into it carrying live steam and an outflow valve for spent wash. There are two or three upside-down “J”-shaped copper pipes going out the top of that chamber into the one above it, and a copper pipe leading out of that middle chamber, through the top one and into a doubler or “thumper keg,” like the one many bourbon distillers still use. That, in turn, leads to the standard condenser.
A critical idea I cannot get a complete consensus on is which chamber distillate is taken from. Does distillate get taken from the bottom chamber or from the middle? If we rely on the logic of this line: “a copper pipe leading out of that middle chamber, through the top one”, the middle chamber is acting like the pot we know and can relate to best. The top chamber is gaining valuable time under heat while it preheats which is easy enough to understand. Cognac stills have a pre-heater. Things get funky in that bottom chamber if it works like I think.
In Circular 106 from 1938, Rafael Arroyo described a concept he called tail waters:
One of the biggest advantages of a discontinuous batch distillation still for rum is the ease with which we can separate from the tail products, those harmful from ones that can be useful to us. Harmful tail products generally end their distillation when the hydrometer marks around 30-40 P. From this point until it reaches zero, the products that distill change their aroma surprisingly, which becomes smooth and pleasant. It is very useful to collect these slightly alcoholic waters in a separate container; but loaded with essential oils, higher alcohols of very high boiling point, and valuable esters; also of high boiling point. The use we can give to these “tail waters” we will discuss when dealing with the curing of raw rum.
The bottom chamber may be all about tail waters. High value congeners (HVCs) are less volatile than fusel oil, as illustrated by the birectifier. Seventy five percent of fusel oil—that other infamous oil (higher alcohols) is found in birectifier fraction 4, while rose ketones—rum oil and higher molecular weight esters are found in birectifier fraction 5. In the second chamber, a heads, hearts, & tails cut can be made and then, as described by Arroyo, there is still a ton of value in the tail waters.
[This was all clarified by both an exchange with Todd Leopold as well as discovering an intimate look at three chambered still operation by IRS chemist A.B. Adams in 1910. Distillate is collected, not from the middle chamber, but at the top. A very large volume from the bottom of the doubler is recycled to the middle chamber and the lower chamber does likely contribute a significant amount of tail waters.]
The bottom chamber would spend its time in a state of subtle reactive distillation (extended time under heat producing even more carotene derived aroma) blowing steam volatile HVCS into the second chamber and then on upwards. Accumulation would force more of them into the hearts fraction. All of this would happen without a care in the world about gnarly fusel oil, because, being less volatile, fusel oil was already removed in the upper chambers. All this investment in time under heat also preps the stillage for increased potential when recycled in the sour mash process. Things don’t go directly from carotene to rose ketone. There is a chain of complicated intermediaries who get broken down during fermentation and then often a staling phase.
A three chambered still is far more efficient for collecting tail waters than a typical batch pot described by Arroyo. The alcohol exhausted wash is already hot so you’re adding little extra energy to it and you also save on condensing energy because it is condensed by the second chamber. [Its hard to say how true this efficiency concept is because these things were set up to chug steam and they recycled very large volumes.]
Rose ketone compounds, like damascenone, even have an impact on the sensory matrix. Your attention and even immune system bends around them. What this means is a spirit rich in them can have relatively higher levels of fusel oil and still remain harmonious. Perfumers believe they create a state of relaxation. I hypothesize that at high levels they create a different immune response to ethanol inebriation with possibly less dehydration, but we don’t know much beyond anecdotes because they appear in so few spirits.
Historically, and from my own experience, spirits rich in rose ketones have a startling stretchability. This either leads to absolutely remarkable persistence in a straight spirit (think never ending Chateau Y’quem persistence) or massive economy as a blending stock. The challenge of harnessing all this unbridled epicness becomes, can you juggle the other facets to create a suave straight drinking spirit? or, do you go grand arôme and create a beast that must be blended down to even be palatable?
The three chambered still may not be the end of all the magic going on at Leopold Bros. There are many channels to create these congeners besides what happens in the still. Leopold may be leading with absolutely exemplary grain selections. Next may be their fermentations, but I would also issue caution. The article basically describes a high risk extended duration fermentation, but the active ingredient may be not so much from additional effort by bacteria and yeast, but by beneficial staling.
Effect of pH on the β-Damascenone Level of Aged Beer.
Chevance and co-workers (17) suggest that the increase in β-damascenone during beer aging could be partially due to acidcatalyzed hydrolysis of glycosides present in fresh beer.
The effect of pH on the β-damascenone concentration was investigated in lager beers aged for 5 days at 40 °C (Figure 1). The higher the pH of the fresh beer, the lower the concentration of this compound after aging, especially in beers in which the pH was initially adjusted to 3 and 4.2. When the pH was higher, the β-damascenone concentration tended toward the level found in fresh beer. These results support the view that β-damascenone is produced in the bottle by acid hydrolysis of precursors, especially when the pH is <4.2. Taking into account that most commercial beers have a pH range between 3.8 and 4.5, our results show how crucial could be a slight pH adjustment of the final beer.
Staling (an older brewing industry term) isn’t the end of the story, but I suspect its the most salient. Alt-yeasts and bacteria, all seem to exhibit more enzyme activity that can hydrolyze glycosides than Sacharoymces yeast. Any fermentation complications you can juggle in your ferment without producing off aromas and peppered-whiskey (acrolein), the more damascenone you will unlock. Appreciable acetic acid however, may jeopardize making the spirit suave and push it to being more of a grand arôme concentrate. It really sounds like Leopold is firing on all the cylinders!
Driscoll’s first hand experience has me really excited, but not just for Leopold, I’m also thinking of West Indies Rum Distillery (WIRD) in Barbados which is putting back online a very large three chambered still. After a little reading, I also may have tasted rum from this still which was the most memorable and enigmatic I’ve ever experienced. This was the 1986 Plantation Barbados, probably released in 2003 (I was very young). I remember aroma from this rum being abnormally diffusive and penetrating. Anyone sitting next to you was also drinking it with you. The Single Cask Rum blog calls this a Rockley rum. They give a sensory description:
Many describe its key-components as smoky, medicinal or even disinfectant-like with notes of honey and, depending on which release, a rather strong ester influence. It is a style of rum which is not necessarily everyone’s darling, similar to rum from Caroni or super-high ester rum from Jamaica. You will probably either love or hate it.
I recognize the description in my own experience and I loved it. Part of the description I’d like to attribute to abnormal levels of damascenone, but other aspects are wild card. From the article, WIRD was rumored to use sea water in some ferments (Scotch historically at times used sea water so there is a precedent):
Another topic that did not make it into this article was the use of different fermentation methods (Mr. Gabriel revealed that WIRD used sea water for some fermentations in the past). But there are just too few information out there to dig deeper into that.
This is all a lot of uncharted territory that will be exciting to follow. As always, if you want advanced first hand understanding of spirits, there is no better tool than the birectifier.