[This post is part of a series of background tests that work in conjunction with the birectifier]
[Kervegant referred to this idea as “quantitative tasting”.]
The Germans developed five useful tests for evaluating spirits (besides the classic titrations) that eventually made their way to Rafael Arroyo in Puerto Rico for rum development. The first major test is Micko distillation invented by professor Karl Micko. This test requires a special laboratory glass still that can operate with very high reflux and repeatably cut spirits up into 8 fractions by volatility. The fractions can be evaluated chemically or organoleptically and lots of actionable information can be gathered. It is sort of like a stethoscope and a microscope for looking at spirits. I will say a lot more about it soon (I finally have the still!).
The next major test is Wüstenfeld’s exhaustive test which we are going to focus on today. Spirits are tested for how far they can be stretched. This was important to the Germans who bought rum concentrates designed to be stretched with cheap fairly neutral domestic spirits. Market price for these rums was tied to their stretchability. This becomes relevant to our fine era because persistence is tied to the rarest most noble congeners and is worth money (the difference between a cru and a grand cru is nuance and intensity). To perform the test, spirits are systematically diluted with water until they can no longer be detected in the solution. There are some extra details here and there.
A related test, which at this point is hard to say how useful it will be, is the surface tension test. It harnesses Traube’s Rule from physics to imply how much long chain carbon stuff is in a solution. As carbon chain length increases, there is a significant effect on surface tension. Depending on equipment used, standardized measures can be created such as dynes/cm² or proprietary in house measures created with cheap equipment like uncalibrated capillary tubes. They can provide potentially actionable comparative measurements within the house, but cannot be duplicated on other equipment. Surface tension may prove most relevant for evaluating fractions of a Micko distillation or for botanical concentrates for gin and liqueur production. A lot of these tests overlap and simply add weight to decisions.
The last two tests are sort of minor. For the first, concentrated sulfuric acid is added to a rum sample (10 ml rum / 4 ml acid) which has the effect of subtracting (through oxidation) all the volatile aromas besides those responsible for rum oil. THE SAMPLE IS RENDERED UNDRINKABLE. If a spirit has rum oil, which is associated with persistence and quality, it will be more easily detectable, BY SMELL. This can help evaluate competitors or help evaluate spirits during product development when parameters are systematically altered such as major fermentation variables. The second related test pertains to letting a small amount of spirit, such as the film of spirit that coats a nosing glass, evaporate over night then organoleptically evaluating the aroma that is left (an oily film at the bottom can often also be felt). Low volatility bass notes, if present, will persist and this simple test will contribute to a notion of quality.
A lot more can be said about these tests on quite a few levels, and, of course, I will eventually expand upon them all. A metaphor I keep using over and over is that they are like being a vinyl DJ relative to only using a Mac Book. They are romantic and organoleptic. They force you to grapple with a creation and build the best kind of involvement. Tourists want to see them because they are more than just numbers, they are experiences. The fine wine business has already been here and chosen many of these tests over the Mac Book / Hewlett Packard alternatives. But think about them for yourselves, because they will all become more useful when we collectively explore them and refine their contemporary protocols.
The Exhaustive Test
The exhaustive test starts by diluting a spirit in an established series of ratios until it can be detected no more. Its exhaustion number is a ratio such as 1:200 which is the last point it could be detected.
[Youtube really compresses the video and makes it look like shit. As I improve the videos, I’ll host some higher resolution versions.]
The best way to set this up is with an automatic pipette and an array of 100 ml volumetric flasks. If a spirit is diluted 1:200, it will require 0.5 ml of spirit and 99.5 ml of water. If it is diluted 1:225, it will require 0.444 ml of spirit and being topped to 100 ml of water (instead of doing that water math!).
What is being glossed over when I say spirit is that I mean a 40% ABV spirit which is chosen because when competitors from the market are evaluated, it is how most will be found. If the sample had 50% it would require 0.8 as much as if it were 40% so we can scale it. We do not need to cut this spirit down to 40% to proceed, we can do some algebra and let our pipette do all the work. For 1:200, instead of 0.5 ml we’d use 0.4 ml and for 1:225, instead of 0.444 we’d use 0.355. A spreadsheet or print out could be generated to hasten the math.
Ounce you vat the test, the samples should shit around for 1-2 hours to settle down.
The Germans had classic numbers, but they may not help us because they were for concentrates or fractions of Micko tests and not the suave (an Arroyo word I re-purpose) stand alone spirits we see and create today. Our collective challenge is to generate numbers as starting point best bets for what we find today.
[where I may have gone wrong here is that the German numbers, which seemed really high, may have been in terms of absolute alcohol instead of a 40% unit I describe above. To get from my units to the German units you would just divide by 0.40. We can work backwards from round total dilution numbers with absolute alcohol to be in their ball park. Still, I’m only getting in the 800’s and not the 2000’s like they were with some rums that were possibly Grand Arômes. Arroyo revised the test a bit and pre-diluted rums with ethanol then diluted them again with water. It may have helped him stay precise with the instruments he had. He then scaled his final numbers to the absolute alcohol from the original rum.]
Another consideration is whether water is the best choice. The alternatives would likely be vodka as an off the shelf neutral spirit or a bland reference rum like Bacardi. Looking for exhaustion in context makes sense on paper. Water can be quite distracting to a booze bag and you may think you need a firmament to rest on as you sweep your gaze across the flavourscape. This is what I briefly thought. However, when I tried it with both vodka and Bacardi, I quickly went back to water. This is something everyone will need to briefly try so that we can settle back on water.
As you close in on the dilution ratio where judgement is dicey, the test elaborates to a triangle test. This a UC Davis concept as far as I can tell. Randomly, two samples will be either diluted rum or only water while the third will be what is missing (flip a coin). These will be shuffled so the evaluator loses track and has to determine which two are the same and which is the odd one. If done successfully, the test can continue to the next dilution stage or it can stop, being evaluated as undetectable.
The entire test with four possible evaluators can be performed with less than 10 ml of a 40% ABV sample. This can make it practical for even the rarest and scantest of samples. The tools are also very affordable. A $50 automatic pipette can facilitate the dosing. 100 ml volumetric flasks are $5 each and I recommend at least 8. For extra tools, I like working with a 25 ml glass volumetric transfer pipette (high school lab supply places often have the cheapest versions) which requires a 3-way rubber bulb filler ($10). I start by topping up the volumetric flask under the sink faucet then I switch to doing it with the transfer pipette to quickly and accurately reach the correct level. An entire series of flasks can be generated to perform the exhaustive test in mere minutes. A critical final tool is a set of tasting or nosing glasses.
For an initial test, I compared Mezan XO Jamaica (40%) to the 24 year old Golden Devil Hampden Estates (50%). The Mezan exhausted at 1:200 while the Golden Devil exhausted at 1:275. I started the test at 1:100 and increased it by increments of 25 parts.
Historically, some of the concentrates could be exhausted at 1:2000, but these were diluted often from 70% and not standardized as a 40% unit. Some concentrated fractions of the Micko test could also hit such numbers. There is likely no historical information that can tell us the exhaustive point of the most extraordinary stand alone suave type spirits.
This test of persistence may be able to tell us if there are merits to producing rum concentrates to be skillfully diluted as opposed to producing only suave distillates. Is the concentrate method a way of capturing more of the rarest most prized congeners? We can start to ask lots of questions.
The test takes practice and you will become increasingly good at it as time is spent so be patient. I find the need to only taste a few then get up and return to make more judgements not tainted by ghosts lingering in the mind. Aroma perception has both passive and active modes. Often the passive mode is the most acute at noticing thresholds. Walk up to it…, taste, wham! It should be done too quick to summon a perceptual ghost. The more active you taste, the more you will summon recollections that in other contexts add to the richness of flavor perception, but the poorer your judgement will be in this utilitarian context.
Using two matching nosing glasses can also expedite the process. One should be water while the other is your sample. Switching back and force reinforces the search for contrast. It is important that these glasses match so no difference in haptic sensation from the rim dominates your attention.
1:150 → 0.667 ml
1:175 → 0.571 ml
1:200 → 0.500 ml
1:225 → 0.444 ml
1:250 → 0.400 ml
1:275 → 0.363 ml
1:300 → 0.333 ml
1:325 → 0.307 ml
pipettes are often listed in micro liters → μl so 1.0 ml is 1000 μl
Mezan 2005 Guyana reached 1:175
Mezan XO Jamaica reached 1:200
Mezan 2000 Jamaica reach 1:325 (possibly higher, but I need to make another micro blend)
If you are a new producer exploring these tests, I highly recommend practicing on all the Mezans. They are a mixed bag and incredibly affordable. A few are easily the greatest widely available rum products on the market for the price. A few, however, are misses that I do not feel were blended/selected correctly, but they is still a great opportunity to learn a ton.