Skip to main content

Watchdog Piece Suggests Toxic Compound Diacetyl May Pose Risk to Roasters’ Health

coffee roasting

Roasted coffee in a cooling tray. 2010 Creative Commons photo by Michael Allen Smith

A report from the Milwaukee Journal Sentinal suggests commercial coffee roasteries may be putting the lung health of their employees at risk through exposure to a toxic chemical compound known as diacetyl.

Diacetyl, known for a buttery flavor, is a chemical compound found in many foods and drinks, such as beer, wine and yogurt, and the United States Food and Drug Administration maintains that it is safe to ingest in small amounts. But when diacetyl is inhaled in unsafe amounts, it can cause a potential deadly or debilitating lung disease known as bronchiolitis obliterans, also known as constrictive bronchiolitis, which causes small airways to be constricted by fibrosis.

Unhealthy levels of diacetyl have most commonly been associated with the production of popcorn or other chemically flavored products. The chemical compound has been at the center of a much-publicized case involving the deaths of multiple people who worked at a facility in Tyler, Texas, where coffee was being roasted and flavored.

The Journal-Sentinel piece provides one of the only known studies of the compound in unflavored-coffee roasting environments. And although the sample size is undeniably small, and remarkably few known cases of bronchiolitis obliterans have been associated with commercial coffee roasting, the results may warrant further study.

Journal Sentinel reporter Raquel Rutledge contacted several Wisconsin roasteries for the story, asking if the organization could test for the air in production areas where coffee was being roasted and/or ground. While several prominent roasteries declined, Milwaukee’s Valentine Coffee Roasters and Madison’s Just Coffee Cooperative accepted the invitation.

Both roasteries are known to have modern airflow and ventilation systems designed to mitigate the potential for poor air quality, yet the tests showed that diacetyl levels exceeded what the National Institute of Occupational Health and Safety considers, in a current draft report, the target maximum levels for employee safety. That draft currently suggests 5 parts per billion as the exposure limit during the course of an eight-hour shift. At Just Coffee’s two roasters, the Journal Sentinel testing revealed levels of 6.9 and 8.1 ppb, while Valentine’s roasting area showed a level of 7.44 ppb. The tests showed exponentially higher diacetyl levels where coffee is ground — 19.3 and 19.6 ppb, respectively.

We followed up with Just Coffee Co-Founder Matt Earley, who says the presentation of the data related to grinding may not reflect actual exposure in JCC’s production facility.

“The MJS took the two hour tests and extrapolated them out over an eight hour day,” Earley tells Daily Coffee News. “The grinding numbers being four times higher than recommended exposure levels is a little confusing. They ran those numbers to reflect an eight-hour grinding shift. We generally do not do shifts over two hours, and over three is unheard of, so that number is a higher than any actual exposure here.”

Earley suggests diacetyl is largely unknown throughout the coffee roasting industry, and says JCC “jumped at the chance” to have the facility tested when approached by Rutledge.

“We always want to be on the leading edge of making sure that our business lives up to our principles and our ethics,” Earley says. “If there is even a chance that our people our being exposed to something potentially harmful, we want to get on top of it. I am frankly surprised so many roasters turned her down, but I feel like now that more information will be coming out on this that other roasters will take it seriously and do what they need to do to protect their workforce.”

Read the full Journal-Sentinel piece, which followed interviews with numerous people outside and inside the industry, including SCAA Executive Director Ric Rhinehart, who points out a lack of epidemiology related to bronchiolitis obliterans throughout coffee roasting’s history.

The National Coffee Association, for one, released an official position statement on diacetyl in May of this year, as the U.S. Centers for Disease Control and Prevention is currently in the draft stage of new safety standards related to the compound. The NCA statement reads, in part:

The National Coffee Association (NCA) is aware of isolated instances where environmental exposure to large quantities of synthetic diacetyl in flavorings has been cited as causing a very rare health condition called Bronchiolitis Obliterans (BO). These instances had the commonality of exceedingly high exposure to synthetic diacetyl in flavorings under poor plant hygiene conditions, which included improper ventilation and absence of personal protective equipment.

For naturally occurring diacetyl as a byproduct of roasting coffee, there are well-established safety protocols and procedures designed and deployed to eliminate worker exposure. Commercial coffee roasting ovens are generally “closed systems” that are built to channel smoke, chaff, volatiles like diacetyl, and other natural byproducts of roasting through closed pipes to exhaust stacks. After roasting, coffee is released onto “cooling carts” (basically large circular trays), since the beans must be immediately cooled to arrest further roasting that would impact flavor profiles. While some systems have closed cooling carts, standard practice for open cooling carts involves air and volatiles being sucked down through holes in the bottom of the cooling tray as ambient air necessary for cooling is drawn down by fans through the beans and out below through the holes. The air and volatiles drawn down in the cooling carts is then channeled by pipe into exhaust stacks that vent the air externally. Stemming from these protocols and procedures, there has been no proven connection between the natural roasting byproduct diacetyl and obstructive lung disease.

We’ll have more on this subject as it develops.



Paul Emerick

Sounds like Dunkin’ Donuts might be busy hiring attorneys for worker’s comp.

david P.

I’m curious if there’s a chart/diagram that lays out how much of this is released per pound. Do small batch roasters have as much to worry about as say a 60kilo + roaster? With that, what impact will this have on open air fluid bed roasters like ashe, sono ect?

Matt Earley

Thanks for the article Nick. I do want to make the point that any way we cut it, levels detected from both grinding and roasting were higher than we or the scientist expected and those results warrant further testing and working on mitigation.

We are hosting the CDC’s National Institute of Occupational Safety and Health (NIOSH) in July to do that work and will help supply them with the data that could be used to both create standards and identify techniques that help roaster protect their workers in the future. We are looking forward to sharing that with the industry as we get results back.

Thanks again,



Hi Matt… Just reading this post and wanted to follow up on the meeting you had with NIOSH’s CDC. Was there any agreement on standards or figures we could use in our small stores?

Thanks in advance!

Matt Earley

Hi Jose,

We had the NIOSH team out here in July and they spent the better part of a week with us testing all areas in real time. They had expected to have results to us in August, but we still have not received them. As soon as we do we will share with you all. They were very professional and great to work with. If you are interested I can put you in touch with the team that was here. Thanks!


David P.

Hopefully this data becomes available quickly and publicly because clearly with all the industry standards in place and up to date, they are still woefully inadequate. This is now clearly an industry issue, Roaster fabricators and Roasting companies alike need to figure this out NOW. I’m not going to pay another dime for a Roaster until I know its safe.

Raquel Rutledge

Thanks for the interest in our story, Nick, and for sharing with your readers. Just want to correct a few things you mentioned and to expand upon Matt’s comments for those in the business interested in the details of the proposed safety standards and those established by the American Conference of Governmental Industrial Hygienists (ACGIH). First, NIOSH’s proposed 8-hour time weighted average draft recommendation is 5 parts per BILLION, not million. Second, the plant in Tyler, Texas is a coffee roasting facility. It does not flavor popcorn. Lastly, we took long-term samples for two roasters and one grinder at Just Coffee. We sampled for exposure to both diacetyl and 2,3-pentanedione. We sampled the roasters for 3.5 to 4.5 hours and the grinder for 2.5 hours. NIOSH and OSHA have differing methods on calculating time weighted averages (TWA). NIOSH assumes workers will have continued exposures of the same levels measured for an 8-hour shift, unless workers state they are finished and leaving the building, for instance. OSHA assumes no additional exposure beyond the sampled time and factors a zero into the calculation. We did similar tests at Valentine Coffee Company but in addition, we took short-term samples to measure peak exposures that scientists believe may be key to understanding how diacetyl and 2,3-pentanedione destroy the lungs.

NIOSH and ACGIH have short term safety limits of 25 ppb and 20 ppb, respectively (NIOSH’s is a draft proposal and not finalized yet.) Short-term limits are based on 15-minute exposures. Our 13-minute samples at Valentine found levels of nearly 70 ppb of diacetyl (vapor concentration; more than 60 ppb for STEL TWA). Hope that helps. Anyone interested in discussing further is welcome to call me. 414-224-2778.


Raquel Rutledge


Hi Raquel. Just read your post and had some questions that I would thank you a lot if you could answer if you have some time: My coffee company is small. We use a 10 kg (22 lb) per cycle drum roaster, that we use 2 to 3 days of the week, 6 cycles per day. The rest of the time is off. We must be consuming between 5 to 6 kilos a day in the coffee shop, all of them must be ground prior to use. Besides, we send away another 10 kilos to coffee shops and restaurants, and some of them must be ground (perhaps 30% of them)

The coffee roaster and coffee shop is placed in a 250 sqft store. We have windows and two doors. We use closed pipes to funnel fumes out of the roaster, but there are remains that stay in the store, or once the coffee is pulled to the cooling tray. We use ventilation to drive away all these fumes.

With all this in mind, how could I know what is the ppb exposure to diacetyl to all of our 10 employees.

Thanks in advance, God bless!

Raquel Rutledge

Hi Jose,
The only real way to know is to have an industrial hygienist familiar with diacetyl and 2,3-pentanedione testing come in and sample the air in the breathing zones of the workers and in the area. One way you can locate an industrial hygienist in your area is by going through the American Industrial Hygiene Association:
It’s really impossible to guess. There are so many factors at play. Local exhaust should pull air down away from the face, grinding warm beans tends to release higher levels, bins where coffee is stored after roasting to off gas release really high amounts when opened, etc. Every shop and situation is different. You need a professional to tell you. I know some companies have sought the free assistance of NIOSH and even OSHA (which has a consulting side where you can have them test and advise and remain anonymous but you have to agree to make any necessary improvements should they find any.) Otherwise hiring an industrial hygienist could cost anywhere from a few thousand dollars on up, depending on what all you have them do. Levels that experts say you should be concerned about range from 20ppb during a 15-minute span (such as during grinding) to 5 ppb over the 8-hour work day. It sounds like you don’t run the machines steadily all day every day but it looks like you might have concerns about repeated short-term exposures. I am happy to discuss further via phone if you want to call me. I’m at 414-224-2778. Thanks!

Chris Schaefer

I’m particular encouraged to see that Raquel and Matt have left comments regarding this. And, as another has posted, it will be interesting to see the “scale” of this as it relates to small-batch and boutique roasters. The implications of toxicity of this byproduct are alarming but, as with many things, it all must be taken “with a grain of salt.” That is, our full undestanding of the harmful effects and the conditions which produce this are far from being fully understood.


As a ‘home roaster’ I know many who roast in the basement or kitchen without any ventilation – this article has really made an immediate impact on them to re-think their location.

I can’t imagine it would be a big deal for consumers grinding a few ounces of Coffee a day…..or at least I hope not.


Same here. I roast in a garage by the open door, but without any other ventilation. I inhale a lot of smoke some days – more than you would get off a commercial roaster with afterburners and such. Is the concern the smoke from the roasting process, or something in the beans?

Comments are closed.