Follow along: IG @birectifier
Recently, I translated the famed text, Le Cognac, and thought it may be worthwhile comparing some of the distillation ideas to a few from Arroyo as well as my own experiences performing countless case studies with the birectifier.
Cognac producers have grand practices that would be the envy of any distiller. They have spacious direct fire pot stills that are given very long durations under heat with high acid distilling material. When you get into the details, they have a few concerns and hesitations that may inform the rum distiller. For starters, they are concerned with congeners that bounce, and they operate in a way that best lines up all congeners single file such as we see with the birectifier. Congeners, especially at the end of the distilling run, as ethanol is being exhausted, tend to become very chaotic in their sequence as undulations in energy applied to the boiler ripple through the system (or changes to coolant in a reflux device create ripples). You cannot easily sort or separate or catch (or any number of metaphors) congeners whose sequence becomes scrambled. Intentions do not get realized.
The birectifier exemplifies a real world ideal system for lining up congeners in sequence by volatility (which is a function of boiling point and relative solubility in ethanol and water). Energy is very carefully applied to the boiler at an incredibly slow rate keeping a constant output of 25ml per 15 minutes. This system distills very close to the azeotrope using only natural reflux from ambient air (before being condensed by water). There is no water cooled dephlegmator to add a second variable to the system. The way this neatly sorts congeners is uncanny and when you use a standard system of 100 ml of absolute alcohol and eight 25ml fractions, you can learn a lot about a spirit.
Now consider Cognac. They do nearly all the same things we do with the birectifier and they sort of treat the pot still like a reflux still, driving a boil very slowly so that they do not overly challenge the natural reflux of their still head. As seen in Le Cognac, they were very concerned about managing heat at critical times because they did not want any surges of energy rippling through the system. If the door of the furnace was opened, that would draw in oxygen, increase energy, challenge the natural reflux, and everything could be messed up. Congeners would go from single file like the birectifier to bouncing.
Once reason they want single file, is that even though they like to collect deep into the distilling run, they want to sort high value aroma like carotene derived congeners and esters from overly fatty free volatile acidity. These acids were the tell of seconds and detract from the olfactory-clarity of the spirit. When examining birectifier fractions, we see everything high value existing in a narrow band that comes across in fraction 5 while the aroma of seconds follows very quickly in fractions 6-8. Any bouncing would pull a lot of the fatty fraction 6-8 stuff forward in the distilling run spoiling the clarity that the high value fraction 5 stuff deserve.
When you smell your way through birectifier fractions, all these congener classes, their sensory impact, and their location in a well sorted distilling run become less abstract and far easier to understand. There really isn’t a lot of chemistry in play at this point. It is basically stuff in sequence or out of sequence which compromises the position of higher value stuff.
Arroyo enters the comparison at this point when in 1948 he presented the idea of Simultaneous production of light and heavy rum. What Arroyo does is take a stripped distillate analogous to the Cognac brouillis and submit it to extremely careful fractional distillation in a batch column still to mimic the structure of birectifier fractioning. Arroyo’s five fractions were collected from narrow bands of temperature that were drawn from his birectifier observations. The first fraction is set aside for future processing, but the next four fractions become two blends with each being a different ratio of the parts with most of fraction 5 entering the heavy rum.
The fifth and last fraction is collected within the temperature range 85-90°C and will contain 20-30% of alcohol by volume. Its physical appearance will be opalescent or turbid, due to the presence of natural essential oils and some of the highest boiling point esters and aldehydes, all of which are more soluble in alcohol than in water. It is, however, one of the most important fractions from the standpoint of rum making, for here will be found those congenerics of rum fermentation such as those mentioned above, to which good rums of quality in either class owe their more cherished characteristics. The content of higher alcohols will be almost negligible in this last fraction, but there will be a sharp rise in the content of organic acids of the saturated, aliphatic, monobasic series.
What isn’t collected by Arroyo’s distilling run parallels the marc of the bon chauffe and gets recycled back to the stripping run. It’s most dominant characteristic is free volatile fatty acids that Cognac distillers would consider seconds.
Arroyo, no doubt, arrived at these ideas by taking birectifier fractions and making miniature blends with them. I’ve done the same and even made daiquiris from those! What you can learn about what defines a spirit’s personality or obstructs quality is simply amazing. GCMS isn’t that intimate and can’t develop that much intuition for the distiller. Performing birectifier case studies and then creating what I call sketches from the product is among the best educations you can get. When you put a face to the name of the congener classes, re-reading all the classic texts on distillation becomes a richer experience.
So what is the big difference between the slowly heated pot still process of the Cognac producers and Arroyo’s batch column process (I call super-fractioning)?—A lot of well integrated water. Cognac producers build in more water at a critical point while Arroyo’s idea will need to add it later risking aroma breakage. What needs to be said is that Arroyo is only presenting us with a high concept idea that may be useful to a narrow slice of people. This wasn’t his best practice and he is actually the earliest reference to the idea of aroma breakage. So, super-fractioning sort of goes against his own advice, but it does help us better understand Cognac practices which really may be the grandest most fine method.
One of the surprises in both methods is that higher alcohols are never too significant to decision making because ferments that justify these processes have below average levels when you start to make comparisons to 36-72 hour economy ferments (think Bourbon). A hesitation of the Cognac producer is that they didn’t believe much meaningful esterification happened in the still despite all their investment in time under heat. If a little butyric acid was hypothetically thrown in, they did not think you could brute force butyric ester. They also were afraid of esters breaking. If the brouillis was too high ABV, and needed watered back slightly before redistillation to get a desired output ABV, they acknowledged esters breaking that may not reform during the second distillation. This means that to accomplish butyric ester, you may need to form the ester as a bio-transformation in your ferment or have something in your retort to skew esterification kinetics such as a very substantial amount of volatile acidity beyond what Cognac sees.
Building in the water to prevent aroma breakage down the line sets the stage for an idea introduced by Kervegant that is basically “aging to completion“. The spirit will gradually lose ethanol during maturation and will be ready for consumption when it reaches drinking proof. The starting ABV becomes like a timer, but higher does not equal better because that implies a lighter spirit entering the barrel. An ideal emerges with a maturation duration that is probably shorter than what is currently practiced. Longer would not be better and we know that many old Cognacs get moved to glass demijohns. As another ideal, spirits never see more than 10% water to cut them. Cognac may be the only spirit that comes close to employing this ideal but then they ultimately obfuscate it in blends. Aging to completion may be the ultimate way to express terroir and it may also illustrate the potential for a tropical advantage in maturation.
Now that we understand the grandest possible ideas in distillation, what comes next is understanding when not to practice them. Not all ferments are as special as Cognac ferments. There is a vast spectrum of quality in rum ferments. Garbage in, garbage out is the rule of the day. Cognac distillers encounter flawed wines but know how to alter their practices so undo resources are not expended upon them. They also know that maturation highlights many flaws instead of masking them. Demystifying still operation options, will push the onus back on fermentation to send better material to the still.
Demystifying these concepts may also help us better automate stills. I briefly worked on a project with an engineer to automate birectifier distillation, but it was cancelled (postponed?) due to lack of demand. Basically, we were creating a PID (think sous-vide controller) that didn’t regulate constant temperature but rather constant liquid output. The output would be 25ml per 15 minutes but could be changed as well. Output was measured via a high end digital scale. A thermometer would add extra data points for record keeping or decision logic. The device even had a turret to rotate into another collection vessel once the 25ml fraction was collected. The software could scale from the birectifier to a table top pot still processing experimental ferments to other larger commercial units. It would be awesome to complete the project so we could have more structured repeatable experiment scale distillations. Repeatability may offer new avenues for distant collaboration, especially among those chasing terroir.
I’m not sure I did this topic justice so let me know what you think.
2 thoughts on “Cognac vs. Arroyo vs. Birectifier and Congeners that Bounce”
So can you use the Birectifier to analyze what the appropriate amount of ice (water ) to add to say a serving of cognac (or any other spirit) so that you can drink it at its maximum generosity i.e. it shows all the best possible aromas, flavors, etc?
Not really, but the birectifier can give you some hints. To do that you’d do systematic dilutions and hold a tasting panel. Some of the companion organoleptic ideas to birectifier analysis are well worth looking into. One set of ideas is “quantitative tasting” and you use systematic dilutions to create indexes of persistence for a congener class. This idea could give you a numerical index of something like fusel oil that you can make sound decisions off of while being far easier that performing an elaborate fusel oil titration. The same techique can be used for other stuff like ethyl acetate. You can conduct a lot of analysis without ever resorting to GCMS.