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:

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!

Share this: Facebook

Twitter

Pinterest

Tumblr

Email



Like this: Like Loading...