Author: Marshall Schott
Is it die-uhh-see-tuhl or die-ASS-ih-till? I’ve heard both pronunciations are accurate and find myself using them interchangeably. Regardless, I recall first learning years ago about diacetyl– an organic compound (2,3-butanedione) formed during fermentation with an artificial butter character that can be dealt with by pitching proper amounts of healthy yeast and raising the temperature of the beer toward the end of fermentation. Like many homebrewers I know, I adopted these methods and continue to employ them today, 
convinced they’re helping me to avoid the unpleasant butter flavors I so detest… despite the fact what I “know” about diacetyl and how it presents in beer is based purely on hearsay, things I’ve heard or read from others.
Similar to dimethyl sulfide (DMS), diacetyl has become one of the more commonly cited off-flavors in beer evaluation, one of the few go-to issues people focus on when a beer is otherwise decent. I’m not sure if this has to do with the fact diacetyl truly is present in a lot of beer, “butterscotch” is easy to remember, or it just feels good to think we’re able to detect and call out off-flavors. Over the years, I’ll admit to developing quite a bit of skepticism when people report diacetyl, a sentiment motivated not by my concern for a taster’s integrity or honesty, but rather my belief most people share my off-flavor training experience, which ultimately amounts to not-very-much. Moreover, I’ve heard from a surprising number of brewers who entered the same beer in multiple competitions and received feedback from one set of judges docking them for the presence of diacetyl while judges from the other competition made no such comment. Who was right?
For this edition of our off-flavor series, FlavorActiV provided us with concentrated diacetyl flavor standards to serve to blind participants with the goal of learning more about how we perceive this compound.
| PURPOSE |
To evaluate the ability of blind tasters to distinguish between a pale lager dosed with diacetyl flavor standard from an unadulterated sample of the same beer.
| METHODS |
I decided to use a notably clean commercial beer, Bitburger German Pilsner, for this xBmt since all I had on tap at the time was a hoppy APA and hard cider. I considered using a beer I’d brewed, but despite my personal confidence my process allows me to produce beer without noticeable levels of diacetyl, 
since I don’t regularly send it in for lab testing, it seemed a safer bet to use a beer made by a rather large professional brewery.
According to existing literature, the approximate flavor threshold for diacetyl is between 10 and 40 parts per billion (ppb), though I’ve found some references claiming it’s upwards of 100 ppb. I couldn’t find any information regarding typical levels of diacetyl present in Bitburger, in my experience it is a good example of a clean fermented German Pilsner, but it’s certainly possible the compound is naturally present.
FlavorActiV provided me with 5 diacetyl flavor standard capsules, each one said to impart a concentration 3 times the flavor threshold when added to 1 liter of beer. Given reports that some popular commercial beers have upwards of 600 ppb, I settled on dosing a 1 liter volume of beer with 2 capsules, producing a concentration 6 times above the approximate flavor threshold. Assuming a low natural concentration in Bitburger, the dosed beer ostensibly contained between 60 to 300 ppb diacetyl.
To prepare the dosed sample, I followed the instructions provided by FlavorActiv by first gently pouring about 200 mL of beer into a sanitized Nalgene bottle, adding the contents of 2 flavor standard capsules, then adding an additional 800 mL of beer to the vessel. Pouring gently, I was able to do this with minimal foaming. The non-dosed samples were poured in their own sanitized Nalgene bottles to the same volume. Using an inexpensive aftermarket cap that I’ve found holds an airtight seal as good as any growler, I gently rotated the dosed beer to ensure the diacetyl crystals were completely dissolved and homogenized; in an effort to keep the beers as equal as possible, I did the same to the non-dosed samples. The 3 beers were then placed in a small cooler with ice packs where they remained for approximately 30 minutes before data collection began.
| RESULTS |
A total of 19 people with varying levels of experience participated in this xBmt. Each blind taster was served 1 sample of the beer intentionally dosed with diacetyl flavor standard and 2 samples of the clean beer in differently colored opaque cups then instructed to select the unique sample. At this sample size, 11 tasters (p<0.05) would have had to accurately select the dosed sample to achieve statistical significance. Ultimately, 13 tasters (p=0.002) accurately selected the unique sample, suggesting participants were able to reliably distinguish a diacetyl dosed beer from a clean sample of the same beer.
Those participants who correctly selected the diacetyl 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.
The first question asked of the 13 tasters who correctly identified the unique sample on the triangle test had to do with preference. The non-dosed beer was preferred by a majority of 9 tasters while 3 said they liked the dosed sample more and 1 person said they had no preference despite noticing a difference.
At this point, the general nature of the xBmt was revealed to participants, namely that one of the samples had been dosed with an off-flavor, though they were not informed of the particular off-flavor. From a list of common off-flavors that included brief descriptions of how they typically present in beer, tasters were asked to select the one they believed the beer was dosed with. Out of the 13 participants who completed this section of the survey, 4 accurately identified the off-flavor as being diacetyl, 3 thought it was cholorphenol, 2 believed it was oxidation, and 1 felt it was metallic.
My Impressions: I like to think I’m relatively aware of my tasting strengths and weaknesses. Although I’m pretty sure I’ve detected diacetyl in homebrewed beer before, I rarely note it as an off-flavor when judging, mostly because it seems what most perceive as diacetyl I experience as something else, usually crystal malt character. I was pretty stoked to have a go at this one and asked a friend to serve me a series of blind triangles, changing up which was the odd-beer-out. In 6 trials, I accurately identified the unique sample a total of 2 times, exactly what you’d expect from random guessing. And I’ll be honest, that’s pretty much what it felt like, I simply couldn’t detect with any confidence the buttery character I was expecting.
| DISCUSSION |
This one got me. It got me good. After my very consistently accurate performance on the previous DMS xBmt, I developed a slight confidence in my ability to detect off-flavors, which I subsequently used to privately validate my assumptions about people mistaking completely normal beer flavors for off-flavors. In fact, as far back as 1993, George Fix noted in his great article, Diacetyl: Formation, Reduction, and Control, “in fresh beer the flavor [of diacetyl] can be confused with that of caramel malts.” Confirmation feels so good!
Then I went and performed terribly on multiple triangle tests when participants were capable of reliably distinguishing the beer dosed with diacetyl from a clean sample. My knee-jerk reaction as I witnessed taster after taster selecting the right cup was to abandon my assumptions and humbly accept my own inability to detect diacetyl, at least at 6 times the normal flavor threshold. However, the data from the follow-up questions had me scratching my head again, particularly the fact 9 out of 13 tasters who were able to tell the beers apart did not accurately identify the off-flavor as diacetyl. Initially, I figured this was no big deal, at least they had what it takes to tell the beers apart, but then I considered the implications as it relates to judging.
Judges aren’t served 3 of the same beer and asked to pick the one that’s unique, but rather a flight of beers that are generally all different. Going in a set order, each beer is judged independently of the others, usually by a pair of judges who end up consulting with each other about their perceptions (even if they’re not supposed to). For the sake of argument, if less than a third of people are capable of distinguishing a diacetyl dosed beer from the same beer without it, yet they are incapable of identifying that the off-flavor is diacetyl, even when provided common descriptors (i.e., buttery, butterscotch), how confident can we be that a judge’s reported detection of said off-flavor is accurate?
Of course, this xBmt doesn’t speak at all to the prevalence of diacetyl in beer, but rather its ability to change a beers flavor enough to make it reliably distinguishable. In this case, at 6 times the standard flavor threshold, there’s fairly strong evidence to suggest it does, which alone is enough to cause me to reevaluate my perspectives on the matter.
If you’ve ever experienced a diacetyl dosed beer or if you think you’ve had a beer with noticeable amounts of diacetyl, please share your thoughts 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!
22 thoughts on “exBEERiment | FlavorActiV Off-Flavor Series: Diacetyl In A Pale Lager”
Thank You and Thank You again. Once again Exbeeriments reveals what I believe to be the single most ignored or explained away issue in the beer judging arena. That is until we accept that all palates are not equal and start classifying judges based on their perception capabilities, and drop the idea that experience enables detection; all beers will not be judged equal. The beer is the control and the judge is the variable which is exactly backwards. As the Grandmaster Gordon Strong said “entering beers in competition is a crap shoot.”
Marshall, thank you so much for this exbeeriment. Many people describe diacetyl perception like an important beer tasting milestone and it’s discouraging that I don’t have that skill “nailed” yet.
I’m still uncertain exactly how diacetyl tastes, but I’m pretty sure I tasted it this week in an Irish Red 15A left over from a BJCP beer judging even. Irish Reds actually allow low levels of diacetyl. I actually thought it enhanced the butterscotch flavor of this beer, I did notice something “different” which I attribute to excessive diacetyl.
So your exbeeriment was encouraging to me in that diacetyl may be more in the overall perception of the beer and not so much an off-taste that immediately makes the beer off-putting.
Happy New Year and keep up the good exbeeriments!
Taste Pilsner Urquell and you’ll get to know diacetyl 😉
I love PU… don’t get diacetyl.
Pilsner Urquell is the only commercial beer where I can routinely taste diacetyl. From what I understand, it’s an intended part of the flavor profile.
I diverge from Marshall here, as I, more often than not detect what I perceive to be diacetyl in Pilsner Urquell. I find it interesting that I do not always detect it! Is that me? The beer? My imagination? I truly don’t know! But it hits me very hard when I think it’s there. FWIW, I enjoy the beer regardless. Is that because I’ve been conditioned to think diacetyl is usually an unwelcomed fault in most cases, but because is “ok or expected” in PU that I accept it? Akin to the smell of kimchee? Give me a sour fish smelling dish without prefacing it and I may recoil.
I thought it was DMS (corn) that I could taste in PU, as in many lagers, not Diacetyl.
My local brewery put together a quality control panel and had us go through all the Off flavors and infection training, Great Experience!! Anyway in doing the training I was one that couldn’t taste or smell Diacetyl. The brewer tried different dosage levels over the course of a year, I never detected it. I couldn’t detect the “slick” mouthfeel that supposed to be an indicator either.
In the class, there were some that could detect in across the room, so there are extremes on both ends. I could smell fresh popped butter popcorn from the dosing vial though.
Now if you’re wondering, I can smell popped butter popcorn, I can smell butter aroma from actual butter, but the compound diacetyl itself eludes me, haha. I guess that’s why I like all my homebrew 😉
I’m aware of having tasted diacetyl in a beer twice. The first time was my own homebrew, an IPA brewed with WLP007. I had made the exact same beer once before, when the weather was warmer, and loved it. I brewed again in the winter, and right at the end, when I normally raise the temp for a rest, we had a cold snap (I’m in San Diego, so a cold snap means daytime temps in the 50s). temp in the fermentation chamber was 60 ish, about six degrees lower than the main primary temp. When the beer was finished, it had a noticeably sweet, slightly buttery flavor (no crystal malt in the recipe), and had lost all notion of crispness. Even though the final gravity was the same, it seemed much heavier than the previous beer.
The second time was when I drank a Shipyard Monkey Fist at a bar. I don’t know if the flavor is always there, or it varies from batch to batch. I didn’t know much about Shipyard at the time, but as soon as I took a sip, the flavor was so distinctly buttery that diacetyl was the first thought in my head. I later learned that diacetyl is an extremely common observation with this brewery.
I have never consumed a test beer dosed with diacetyl, so I guess you can take my comments with a grain of salt, but those two beers exhibited all of the descriptors for it.
Just to add to the complexity, my local home brew club has recently been discovering a lot about the diacetyl problems caused by poor beer handling post-fermentation….usually as a result of oxidation during kegging. If oxygen is introduced at this stage it will often take a while to show up in the beer, especially if it’s kept cold as most kegged beers would be. But eventually (it could be 6 weeks or more) the beer will start to darken as the malts oxidise, and the diacetyl will come through as a rich sweetness, often confused with the overuse of crystal malts.
We have brewers in our club that were certain they had good handling techniques while kegging, only to find their beers performing poorly in club tastings/comps. We’ve had some focus sessions on it with guest speakers, and almost everyone has moved to ‘closed system’ kegging techniques now to avoid the issue.
your alpha=0.05, you can’t have 2 p-values
Actually, p=0.002 in this xBmt. To reach significance at p<0.05, 11 needed to be right.
i’ve experienced diacetyl dosed beer during my certified cicerone exam, and i have many friends who have done the same. we all have different opinions with regards to diacetyl. I feel like i can smell it in a glass three seats away from me, while my friend is basically blind to it. and that is supported be many professional tasters. some people are just blind to certain flavors/aromas. some people are incredibly sensitive to some flavors/aromas.
my only concern in the xBmt is that highier dosage does not necessarily equal easier perception. some flavor compounds change what they taste like in larger quantities. perhaps try a lighter dosage and see if the buttery aroma is more perceptible to you.
I’ve done the off flavor training several times, (both in studying for bjcp years ago and in training others) diacetylmorphine is one of the flavors I can’t detect at the standard threshold or even triple it when knowing what the heck to look for and which sample is doctored. I’ll pick it up in mouthfeel before I get any aroma or taste from it. My wife can pick it out quite easily though.
Even with that, I’ve had beers commercially and when judging that undoubtedly have diacetyl… Although I usually try to confirm with my judging partner as for me it’s a known weakness. Learning where your strengths and weaknesses are as well as the characteristics of flavors well enough to put terms to them when detected are where experience with judging helps. But no matter how much more I judge I know I’ll never pick out diacetyl at the stated thresholds.
I’ve found myself to be pretty sensitive to diacetyl, so I tend to be overly critical of my beer, as I look for it.
That buttersotch/ buttery taste, can be masked as acrystal malt, over over bitter messiness.
You found yourself to be sensitive to diaceytl when using flavor standards?
I wish I knew more about sensory science. I know that diacetyl is one of those common off-flavors that some people genetically cannot taste. I have friends who cannot taste it, but they can “feel” it as a slickness on their tongue. Unfortunately, sometimes they claim to “taste diacetyl” when I can’t taste it, perhaps confusing it with some other thing that is giving the beer a slickness. The genetic disposition for people to taste things differently is just as fascinating to me as developing the skill and knowledge to taste and describe things, as well as how peoples’ preferences can differ so much for certain flavors.
I tend to be very sensitive to it, but not usually during judging. Usually I will taste diacetyl in a beer after I’ve been sipping on it for a while, and usually after it warms up a little bit. To me it is completely distinguishable to buttery malt character because it lingers heavily on the aftertaste rather than when the beer is actually on my tongue.
Have you actually tasted diacetyl, or is your sensitivity based on when you think you’ve tasted it? I ask because it seems most fall into the latter camp. I always thought I perceived diacetyl more as a slickness than a flavor, but given my performance on this one… I apparently just don’t perceive it 😕
I’ve tasted it from the BJCP off-flavor kit, yeah.
That’s what I figured. Very cool. The notion that different people might have different sensitivities to off-flavors is something I find fascinating.
Interesting test. Diacetyl is one of the off flavors that I can taste…and smell. I’ve tasted it in everything from homebrew beers and breweries too. Within all of the times I’ve tasted it, I’ve only found it to be a consistent problem with one brewery, where nearly all their beers had diacetyl. Heavier grained beers masked it a bit more…but it still shown through…and their lighter beers you can really smell and taste it. Currently, they have worked through that problem…but it was an annoying problem indeed.
This is a fantastic read, for whatever reason I’m obsessed with trying to find off – flavors and understand how they got there in my brews. Looking for that ‘ah ha!’ moment. I made a light pilsner and thought I had diacetyl flavors but it drank beautifully and quickly. It had to be the caramel malts in the recipe that I was enjoying. Every now and then I pick up hint of caramel in some commercial beers (Sam Adams lager maybe) that I wanted in my beers, i think with the light lager I brewed the crystal malts were coming through more than I expected. I don’t feel so guilty now for enjoying the heck out of this brew. Love this site, thanks Marshall.