Author: Malcolm Frazer
Ever drink a beer, either commercial or homebrew, that tasted of baby vomit or rancid cheese? If so, then you’re likely acquainted with the off-flavor of focus for this xBmt. Butyric acid is produced by various types of bacteria as a metabolic byproduct and is perceptible by most people in rather low amounts, though some is present in nearly all fermentations, just not enough to be noticed. When it comes to clean beer styles, butyric acid is ubiquitously accepted to be an offensive off-
flavor and indicates a contamination occurred. However, there’s some evidence that certain strains of brettanomyces have the ability to convert small amounts of butyric acid into ethyl butyrate, which is lauded for imparting tropical fruit and ripe pineapple character (check out the Milk The Funk wiki).
Since this is an off-flavor series, we were interested in the abilities of tasters to detect butyric acid when used to intentionally dose a clean beer. Is it really as easy to detect as we’ve been led to believe, or will blind tasters even notice their beer has been dosed with a presumably vomit tasting chemical? I was pretty excited to test this one out!
Cheers to FlavorActiV for providing us with the concentrated butyric acid flavor standards for this xBmt!
| PURPOSE |
To evaluate the ability of blind tasters to distinguish between a pale lager dosed with butyric acid flavor standard from an unadulterated sample of the same beer.
| METHODS |
In order to keep me blind to the off-flavor for my own trials, all identifying labels were removed by Marshall from the flavor standard packaging. For this go-round, I used the champagne of beers, Miller High Life, in hopes its delicate character would allow the off-flavor to take center stage.
According to FlavorActiV, the approximate flavor threshold for butyric acid is 2-3 parts per million (ppm) with most commercial beers clocking in at 1.5 ppm or less. Given Miller’s arguably precise and clean process, High Life presumably falls on the lower end of the butyric acid scale, though we were unable to find any lab data confirming this hunch.
Each butyric acid capsule provided by FlavorActiV is designed to impart 3 times the flavor threshold when added to 1 liter of beer, the recommended concentration for off-flavor training. To prepare the dosed sample, I followed the instructions provided by FlavorActiv by first gently pouring about 200 mL of beer into a standard 4 liter pitcher, adding the contents of 2 flavor standard capsules, gently swirling, then adding an additional 1800 mL of beer to the vessel. The non-dosed samples were treated similarly to ensure no differences caused by anything other than the off-flavor.
| RESULTS |
Data for this xBmt was collected during a TRUB Homebrewers meeting being held at Voodoo Brewery. In total, 16 people with varying levels of experience, all blind to the variable, were served 1 sample of the beer intentionally dosed with butyric acid flavor standard and 2 samples of the clean beer in different colored opaque cups then instructed to select the unique sample. In order to reach statistical significance, 9 tasters would have had to select the dosed sample, which is exactly how many made the correct selection (p<0.05; p=0.049), suggesting participants were able to reliably distinguish a beer intentionally dosed with butyric acid from a clean sample of the same beer.
Those participants who correctly selected the dosed beer as being different were instructed to complete a brief set of additional questions comparing only the two different beers, still blind to the nature of the xBmt. When the 9 tasters who correctly identified the unique sample on the triangle test were asked about preference, 8 chose the non-dosed sample while 1 person apparently enjoyed the cheesy character of the adulterated beer.
It was then revealed to participants that one of the samples had been dosed with an off-flavor, though they were not informed which off-flavor was used. When asked to identify the beer they believed was dosed with an off-flavor, all 9 tasters made the correct selection. Next, tasters were provided a list of common off-flavors that included brief descriptions of how they typically manifest in beer and asked to select the one they believed the beer was dosed with– 3 tasters correctly identified the off-flavor as being butyric acid, 2 thought it was isovaleric acid, which is commonly mistaken for butyric, 2 felt it was DMS, 1 chose diacetyl, and 1 believed it was lightstruck (the same person who reported preferring the beer dosed with butyric).
My Impressions: Since I was unaware of the particular off-flavor being used for this xBmt, my trials were more blind than usual, and I’m pretty sure having a newborn in the house gave me a bit of an upper hand on this one. Over the course of multiple triangle tests, I was able to consistently identify the unique sample, as the dosed beer had what us judges are trained to parrot as baby vomit or stomach acid. When I let Marshall know I was finished collecting data, he asked if I had an idea of what the off-flavor was and I responded confidently that it was likely butyric acid. Thanks, Marshall…
| DISCUSSION |
Butyric acid is one of those off-flavors that smells like what it makes me want to do, and the fact tasters in this xBmt not only were able to reliably distinguish a beer intentionally dosed with it from a clean sample but strongly preferred the non-dosed beer is admittedly somewhat promising. I’m not really sure what to think about the 1 correct taster who reported preferring the butyric beer, I’m compelled to believe it was an error, but who knows, preference is subjective and they were presumably able to tell a difference.
I judge in a lot of competitions, which means I taste a lot of homebrewed beer, some that have occasionally had what I perceived to be characteristics of butyric acid. While there are times my perceptions don’t always align perfectly with the people I’m judging next to, it’s been my experience that the puke-like aroma of butyric acid tends to be fairly consistent between judges, usually only getting confused with the other cheesy off-flavor of isovaleric acid. Regardless, if you end find yourself in possession of a score sheet with comments indicating butyric acid, it may not be a bad idea to consider the potential vectors for contamination in your setup, as it’s possible a bug of some sort is finding its way into your beer.
If you have any thoughts about this xBmt, don’t hesitate to share them in the comments section below!
Support Brülosophy In Style!
All designs are available in various colors and sizes on Amazon!
Follow Brülosophy on:
FACEBOOK | TWITTER | INSTAGRAM
If you enjoy this stuff and feel compelled to support Brulosophy.com, please check out the Support Us page for details on how you can very easily do so. Thanks!
18 thoughts on “exBEERiment | FlavorActiV Off-Flavor Series: Butyric Acid In A Pale Lager”
Thanks for sharing Malcolm. I found puzzling that only 9 out of 16 homebrewers are able to detect the odd aroma or flavor at 3 times the threshold concentration. May I suggest you ask FlavorActiv how did they decided on the threshold. (Was it done in water or in beer). A common standard is what science calls LD50 (yes, I know LD is lethal dose 🙂 ), and applied to threshold it would be the concentration at which 50% of the tasters detect the odd flavor.
In other words, you “stumbled” upon a very interesting observation, which is that the threshold for butyric in beer may be higher than previously thought (3 times higher…). I would humbly suggest you add this information to your findings as it is quite interesting.
This may also trigger an idea for future experiments with standards in which you do 3-4 concentrations of the standard (sorry to suggest extra work for you).
All the best and hoping to meet you at HomebrewCon.
Cheers,
Oh man, I’m amazed it wasn’t 16 out of 16. some of the guys must have gone straight in for a taste without smelling first. Butyric acid has to be one of the most off putting aromas I’ve ever smelled in beer.
I too was surprised “only 9” correctly ID’d the adulterated sample. You asked it if was beer or water – did you read the article? Haha. It was not only beer, but the champagne of beers!.
I encourage you to contact FlavorActiV or similar companies and report back to us! Thanks.
While I cannot speak or FlavorActiV, the thresholds used by them and other such providers of sensory training aids, are usually based upon a baseline data obtained from years of analysis. Many tasters, over many years, at many locations gives them averages. In fact, some companies (ie a brewery) even track the baseline for their internal sensory panels and individual panelists.
They settle on 3x (typical) threshold for training purposes. To cue people into being able to ID the off-flavor.
For testing or threshold evaluation purposes, there are often serial dilutions done to determine at what concentration a person (or panel) can no longer reliably detect adulterated samples from controls.
Stumble upon an interesting observation? Sure, you can use that wording. Lol. I prefer to see it as we seek out such reactions from our panelists and the readers. That’s the point of this series! To encourage conversation, not only about the off-flavor itself, but also the unanticipated responses.
For example, a few people selecting isovaleric vs butyric, having hosted several sensory training events, that was foreseeable. Someone preferring the spiked sample?… Not impossible as preference is hard to account for, but it’s hard for me to imagine someone liking butyric acid – that raised an eyebrow. People selecting DMS vs butyric….. hmmmm. That I found curious! What is because they were guessing? Was it bc DMS often follows other flaws (infection)? I don’t know.
For future experiments what I most wish to perform are tests at 1.5-2x threshold, and then at threshold.
I believe that the “water or beer” comment was directed at how FlavorActiV tests for their thresholds, not the triangle test in your article
Could be. I am slow.
Indeed it was 🙂
Butyric is one of the most easily idetifiable off flavors, IMO, but during our homebrew clubs sensory training, is was pretty obvious that some of us picked it up more easily than others, and the more sensitive ones tended to like it less. Anecdotal, but consistent with your results here. There do seem to be a few people who don’t mind it much. It’s also why brewery sensory panels keep track of individual differences-it’s worth knowing how the individuals on the panel react to different off flavors, and how sensitive they are to them, as Malcolm mentioned.
This experiment pulls into question the results of every other experiment performed to date. If only 9 people out of 16 could identify a 3x threshold dose of butyric acid in Miller High Life, then I question their ability to isolate any flavour compound in any given beer. If this is indicative of the wider population (and I realise this may be a stretch) then roughly 40% of tasters actually can’t taste much regardless of what you serve up. So my question is for all previous ‘non significant’ experiments, which tasters could not identify a difference in a triangle test because the differences were physically imperceptible to all humans and which tasters could not identify a difference in a triangle test because their palate is so rubbish that they literally couldn’t tell if someone puked in one of the beers (and should never really be part of a tasting panel)??
Food for thought.
I very much agree with this. I have read many of these and I am beginning to strongly question relying on any Xbeeriment for more than just amusement.
Eh I disagree but obviously I’m biased. If you do question results of xBmt I’d really encourage you to try one or two of the non significant ones yourself and have someone serve them to you blind. Triangle testing variables is quite eye opening IMO
I appreciate your post. Due to small sample sizes, these results aren’t even close to generalizable (and brulosophy is always good to acknowledge the limitations of their studies and note that their results don’t necessarily prove anything). Given the lack of statistical confidence and precision associated with the sample sizes, the results are anecdotal at best. But that’s missing the bigger problem with triangle tests: Even if the results were generalizable, all they demonstrate was whether most people could taste a difference or not. It doesn’t mean the difference wasn’t physically there. this is why I’m much more interested in the blogger’s impressions section where we get the perspective from a trained palate that designed and executed the study. Nonetheless, it’s interesting to see if most people at a random homebrew club meeting in California could taste a difference when an element is varied – especially with silly boogeyman stuff like squeezing grain bags. Keep up the good work guys!
By design, the difference was physically there. We have to trust that Malcolm did spike one of the beers 🙂
I beg to disagree that the results are anecdotal at best. Yes, more data is needed, but when the expectation is that 15 or 16 out of 16 detect the off flavor (even if they do not know what it is) and one only gets 9, there is a very high probability that the threshold is wrong. I doubt that whoever calculated the threshold did it with more than 50 tasters, probably less; and my guess is that they did it in water not in beer (but I can be wrong).
Just an FYI this specific xBmt was done by Malcolm in PA, and the BIAB one was done by me in Denver.
To your comment relying on our impressions. I have to say I would be careful with that and take a more holistic view (i.e. Do our impressions match up with the data generally) as we know certain humans have certain thresholds for compounds so we as the writer may not pick up on something that is there whereas the tasters may have a more diverse set of genetic abilities.
Thanks for reading we appreciate the support!
“But that’s missing the bigger problem with triangle tests: Even if the results were generalizable, all they demonstrate was whether most people could taste a difference or not. It doesn’t mean the difference wasn’t physically there.”
Agree!! This is exactly the problem we have with the random nature of the sensory panels used. Especially considering this experiment involved dosing a beer with a taint to a known concentration your variable is the panel itself.
I know we look at the sensory panel as the ‘apparatus’ used to determine any differences in the triangle test, but really the opposite is true. Each experiment merely tests the palate resolution of the sensory panel, which is generally quite poor.
I reached out to FlavorActiv this morning. Hope they respond… I will let everyone know.
BTW – The published thresholds for butyric in water are as follows:
For smell 240 ppb – 0.24mg/L
Fazzalari, F A., editor, Compilation of Odor and Taste Threshold Data, ASTM Data Series DS 48A (1978)
Guadagni, D. G., Am. Perf. Cosm., 82, 43 (1967)
For taste 6200-6800 ppb – 6.2-6.8 mg/L
Fazzalari, F A., editor, Compilation of Odor and Taste Threshold Data, ASTM Data Series DS 48A (1978)
Siek,T. J., I.A. Albin, L. A. Sather and R. C. Lindsay, J. Food Sci., 34, 265 (1969)
Patton, S., J. Food Sci., 29, 679 (1964)
Very interesting to note the difference between flavour and taste thresholds. It’s quite likely that some of the panelists are not engaging their olfactory senses as much as some others. An approximate factor of 27.5x lower threshold for taste vs smell could certainly account for the differences in ability to detect.
I’m curious if there were any instructions to the panelists on tasting properly?
Thanks Lupulus for finding and sharing this!
Another side comment. Malcolm, do redirect as appropriate…
DMS – picked up by only 3 out of 11 tasters
Diacetyl – picked up by only 13 out of 19 tasters
Acetaldehyde – picked up by only 15 out of 22 tasters
Butyric – picked up by only 9 out of 16 tasters
The beers chosen seem reasonable choices as media for the testing.
There is something wrong with the flavor thresholds in my humble view. I have not searched/ researched the percent population blind to these aromas/ flavors but cannot be that high.
This is cool as hell. I recently brewed my first ESB. It has a very subtle taste that reminds me of my college days when I vomited four pictures of beer all over the dorm hallway.
It is not exactly unpleasant, but not quite complimentary. Do you think “dorm vomit” would be an accurate description of the taste of butyric?