Site Visits Related to Diacetyl and Flavorings that Contain Diacetyl:

Food Manufacturing Facility D – Baked Snack Food Production

Table of Contents

  TOC \o "1-5" \h \z \u    HYPERLINK \l "_Toc222544670"  1	Introduction	
 PAGEREF _Toc222544670 \h  4  

  HYPERLINK \l "_Toc222544671"  1.1.	Project Overview	  PAGEREF
_Toc222544671 \h  4  

  HYPERLINK \l "_Toc222544672"  1.2.	Food Manufacturing Facility D -
Overview	  PAGEREF _Toc222544672 \h  4  

  HYPERLINK \l "_Toc222544673"  1.3.	Quantifying Diacetyl Usage	 
PAGEREF _Toc222544673 \h  5  

  HYPERLINK \l "_Toc222544674"  1.4.	Facility, Equipment, and Processes	
 PAGEREF _Toc222544674 \h  6  

  HYPERLINK \l "_Toc222544675"  1.4.1.	Baked Snack Food Dough Production
and Handling	  PAGEREF _Toc222544675 \h  6  

  HYPERLINK \l "_Toc222544676"  1.4.2.	Quality Assurance Activities	 
PAGEREF _Toc222544676 \h  7  

  HYPERLINK \l "_Toc222544677"  1.4.3.	Baking the Snack Food	  PAGEREF
_Toc222544677 \h  7  

  HYPERLINK \l "_Toc222544678"  1.4.4.	Packaging Activities	  PAGEREF
_Toc222544678 \h  7  

  HYPERLINK \l "_Toc222544679"  1.4.5.	Other – Spray Application	 
PAGEREF _Toc222544679 \h  7  

  HYPERLINK \l "_Toc222544680"  1.4.5.1.	Enclosed, Ventilated Tumbler
and Drier	  PAGEREF _Toc222544680 \h  8  

  HYPERLINK \l "_Toc222544681"  2	Process Descriptions for Monitored
Employees	  PAGEREF _Toc222544681 \h  8  

  HYPERLINK \l "_Toc222544682"  2.1.	Production Operator A	  PAGEREF
_Toc222544682 \h  8  

  HYPERLINK \l "_Toc222544683"  2.2.	Production Operator B	  PAGEREF
_Toc222544683 \h  9  

  HYPERLINK \l "_Toc222544684"  2.3.	Bystander (Safety and Health
Visitor)	  PAGEREF _Toc222544684 \h  10  

  HYPERLINK \l "_Toc222544685"  3	Exposure Assessment	  PAGEREF
_Toc222544685 \h  11  

  HYPERLINK \l "_Toc222544686"  3.1.	Sampling and Analytical Methods	 
PAGEREF _Toc222544686 \h  11  

  HYPERLINK \l "_Toc222544687"  3.1.1.	Equipment	  PAGEREF _Toc222544687
\h  11  

  HYPERLINK \l "_Toc222544688"  3.1.2.	Limit of Quantification	  PAGEREF
_Toc222544688 \h  12  

  HYPERLINK \l "_Toc222544689"  3.1.3.	Blank Samples	  PAGEREF
_Toc222544689 \h  13  

  HYPERLINK \l "_Toc222544690"  3.2.	Air Monitoring Results	  PAGEREF
_Toc222544690 \h  13  

  HYPERLINK \l "_Toc222544691"  3.2.1.	Diacetyl and Acetoin	  PAGEREF
_Toc222544691 \h  13  

  HYPERLINK \l "_Toc222544692"  3.2.2.	Previous Diacetyl Air Monitoring	
 PAGEREF _Toc222544692 \h  17  

  HYPERLINK \l "_Toc222544693"  4	Examples of Controls in Place	 
PAGEREF _Toc222544693 \h  17  

  HYPERLINK \l "_Toc222544694"  4.1.	Overview	  PAGEREF _Toc222544694 \h
 17  

  HYPERLINK \l "_Toc222544695"  5	Discussion	  PAGEREF _Toc222544695 \h 
18  

  HYPERLINK \l "_Toc222544696"  5.1.	Relationship between Results,
Sources of Exposure, and Controls in Use	  PAGEREF _Toc222544696 \h  18 


 Butter Flavor and Food Flavors Containing Diacetyl

Site visit report: 

Food Manufacturing Facility D – Baked Snack Food Production 

Executive Summary

Some snack foods include flavorings that contain diacetyl as an
ingredient. ERG evaluated potential worker exposure to diacetyl and
acetoin while employees controlled dough production, extrusion,
conveying, baking, and packaging operations during production of a snack
item that includes diacetyl-containing flavor in the dough. All steps
were highly automated and controlled from central consoles. 

Diacetyl was detected repeatedly in the breathing zone of one employee
who monitored the dough mixing and extruding equipment and performed
basic quality tests of raw dough. The highest result was 0.164 ppm for a
partial-shift sample obtained during a period in which this employee
(Production Operator A) poured the flavoring, handled raw dough, and
cleaned production areas. This is one of three results that are above
the limit of quantification (LOQ) (out of 16 total results obtained).
Diacetyl was detected at levels of 0.111 ppm in the flavor tank area and
0.011 ppm along side the conveyer (the latter value is below the LOQ). 

Other general area test results indicate that no diacetyl was detected
near the baking oven, in the packaging room, and in the Q/A laboratory.
Nor was diacetyl detected in the breathing zones of a bystander and a
second production operator who spent a substantial portion of the shift
assisting with equipment repair work in another part of the plant. 

Acetoin was not detected in any of the samples collected at this
facility.

Butter Flavor and Food Flavors Containing Diacetyl

Site visit report: 

Food Manufacturing Facility D – Baked Snack Food Production 

Introduction

Project Overview

Eastern Research Group, Inc. (ERG) conducted an industrial hygiene site
visit on June 18 and 19, 2008 at Food Manufacturing Facility D, a
producer of baked snack food (hereafter referred to as Facility D). This
is one of a series of site visits conducted for the purpose of obtaining
information on (1) the use of flavor products containing diacetyl and
commonly associated flavor ingredients (e.g., acetoin) in the
manufacture of food and beverage products; (2) the exposures of
employees at the facility to diacetyl and acetoin; and (3) the controls
in place to reduce exposure. In particular, the site visit was intended
to evaluate the exposures of workers who produce and bake an extruded
dough product in the snack food industry.

 Three ERG industrial hygienists met with representatives of Facility D
on June 18 to walk through the facility, discuss the processes
performed, the materials used, and the job categories with potential for
exposure to food flavorings that contain diacetyl at this plant. The
next day the ERG team conducted short-term, partial-shift and full-shift
personal breathing-zone (PBZ) air monitoring and observed the work
practices of employees identified as having potential for exposure to
diacetyl. ERG also evaluated airflow patterns and exposure controls in
use at the facility. 

Food Manufacturing Facility D - Overview

Facility D uses highly automated processes to produce a baked snack food
that includes butter flavoring as an ingredient in the dough, which is
mixed on site. Briefly, an employee manually measures and adds flavor to
a tank of other liquid ingredients, which are automatically incorporated
into a dough. Machinery extrudes the dough as raw snack product onto a
conveyer system that carries the product through a large oven where it
is baked. Finally, the product is air cooled and conveyed to machinery
that packages it in typical plastic-film snack food bags. 

The plant staffs three daily shifts, typically five days per week. Day
and evening shift personnel primarily make product, while most
sanitation activities occur on the night shift. Most of the equipment in
the plant is modern, some installed within the last 3 years, and the
overall impression is of a spacious, clean establishment. 

This production plant is affiliated with a company that has more than
one production site. A plant-based safety manager coordinates with the
company’s main office when specialized health and safety services are
needed. Through a consultant, the company had conducted screening air
monitoring for diacetyl in the flavor-handling area in late 2006 (see
Section 3.2.2 of this report).

Quantifying Diacetyl Usage

According to the product material safety data sheet (MSDS) inspected by
ERG (dated mid-2007) the butter flavor used in this baked snack food
contains 1 to 5 percent diacetyl. Facility D understands that the
formulation diacetyl content is likely at the bottom of this range. No
information on acetoin content was available from either the MSDS or
Facility D. ERG was not permitted to collect a bulk sample of the
flavoring. 

Approximately twice per shift an employee adds several quarts of a
purchased flavor product to a mixing tank containing other liquid
ingredients with which the flavor readily combines to form a solution of
0.05 to 0.5 percent diacetyl (approximate). If the diacetyl content of
the flavor is roughly 1 percent (the low end of the range presented on
the MSDS), ERG estimates that diacetyl usage could be in the range of
0.1 to 0.2 pounds of pure diacetyl per shift (always as a premixed
component of a purchased flavor emulsion). In the less likely event that
the diacetyl content is at the high end of the range published on the
MSDS, the estimated pure diacetyl usage could approach 10 pounds per
shift.

ERG estimates that the diacetyl content of the finished dough is not
greater than 0.001 percent and is probably less. 

Facility, Equipment, and Processes

Baked Snack Food Dough Production and Handling

The baked snack food manufacturing process involves: 1) a dough
production and handling room with a semi-enclosed flavor area used for
combining flavor into other liquid ingredients (walls, but no doors
separate the flavor area from the adjacent production room); 2) an oven
room containing more than one room-length, enclosed and exhausted oven;
and 3) a packaging room containing multiple form/fill/seal packaging
machines. The facility also has an office area with a conference room
and an employee break room (accessible though the office area). These
non-production spaces are separated from the production areas by closed
doors and a corridor. ERG observed that the dividing corridor appears to
be under positive air pressure compared to the production area,
preventing production area air from entering the office area and break
room.

As noted previously, a production operator (one designated per shift)
manually adds flavoring to other liquid ingredients in a tank, loosely
covered by a sheet metal flap, in the flavor area. An area air sample
was obtained adjacent to the row of three tanks, one of which contained
a mixture including butter flavor produced with diacetyl. 

Monitored by an employee at a remote console, the flavored liquid mix is
pumped into an enclosed batch mixer where it is combined with flour and
other dry ingredients. The resulting dough is automatically discharged
without human contact and processed in batches (with relatively small
surface area for the volume of dough) through a step that prepares the
dough to enter an extruder. 

Employees watch from open control stations along the production line
while the dough is extruded as raw snack product onto an open conveyer,
then proceeds through additional brief process steps (e.g., non-diacetyl
surface treatments) to reach the baking oven. In this room, the total
exposed dough surface area associated with shaping and conveying this
product is estimated at approximately 50 to 200 ft2. 

Monitoring was performed for Product Operator A, one of the three
production operators in the dough production area this day. This
employee was also responsible for measuring and pouring the flavoring
that contained diacetyl. Additionally, an area air sample was collected
at an unused work station immediately adjacent to the midpoint of the
length of the conveyer transporting raw product made with the diacetyl
flavor. 

Quality Assurance Activities

Employees enter a Q/A laboratory through a door in one wall of the
production area. Production operators handle raw dough to perform some
Q/A evaluations in the production area, while in the laboratory Q/A
specialists perform other tests on raw dough (supplied by the production
operators, who obtain samples by hand) and on finished product. The
tests sometimes involve pulverizing the product in a small, enclosed
blender. Sensory testing (smell, taste, appearance) involves an employee
smelling and tasting a small amount of product taken from a finished
package. Q/A personnel perform this testing for a moment every few
hours, during which time the employee’s contact with the product is no
greater (and probably significantly less) than a typical consumer
snacking from a bag.

Baking the Snack Food

 To bake the snack product, Facility D passes extruded dough through the
length of an oven via conveyer. The product travels from one end of the
oven to the other and then transfers to a second conveyer that returns
the product to the front of the oven as it continues baking.  ERG
obtained an area air sample within two feet of the point where the
product transfers from one conveyer to the next for the return trip
through the oven. The oven room is unattended (no employees present,
except during cleaning/sweeping when necessary – perhaps once per
shift).

Packaging Activities

From the oven, baked snack product is automatically conveyed to a
separate packaging room where a metering system feeds product into
form/fill/seal-style packaging machines. ERG collected an area sample in
the packaging room.

Other – Spray Application

Certain products receive an additional coating of a flavoring agent
after baking and before packaging. Certain flavoring agents are applied
using a large, fully enclosed tumbler and an associated drier. ERG
observed the tumbler in action, although diacetyl-containing flavor was
not applied at this location the monitoring date.

Enclosed, Ventilated Tumbler and Drier

When used, the enclosed tumbler includes a large diameter barrel that
rotates on its horizontal axis. Product is dropped from the conveyer
into the tumbler at one end of the barrel and eventually exits from the
other end, guided by internal fins that keep the product advancing down
the length of the barrel as it rotates. Nozzles spray the tumbling
product with an aqueous flavor slurry. The tumbler is fully enclosed in
a sheet metal housing with interlocking fan and an exhaust take-off at
the top that is ducted outdoors through the ceiling. Product enters and
exits the tumbler through chutes that pass through the housing. Although
it was not possible to obtain accurate air velocity readings at all
openings, ERG measured an average face velocity of approximately 670
feet per minute (fpm) at the product intake port. Ventilation design
specifications were not available.

The product drier is a separate unit that circulates warm air as the
product moves through the drier. The equipment consists of an outer
sheet metal case and inner conveying apparatus. Like the tumbler, the
drier is exhausted from the top through the ceiling. This equipment is
fitted with a clean-in-place (CIP) system.

Process Descriptions for Monitored Employees

ERG monitored two production operators performing different activities
in different parts of the facility. 

Production Operator A

Three consecutive partial-shift samples totaling 461 minutes were
associated with Production Operator A (working in the dough production
area). Two short-term task samples (15 to 25 minutes) were obtained
concurrently. During this time the employee primarily stood at a control
station approximately 15 feet away from the conveyer carrying dough that
contained diacetyl). Several times per hour the operator walked the
production line to make a visual inspection or to pick up samples of the
diacetyl-flavored dough (manually manipulating the volume of a small
handful). This employee also checked the levels of liquid ingredients in
the flavor room tanks by standing on the mezzanine (so the tank openings
were hip height), lifting the lid, and leaning over the tank to look
inside (5 seconds). Toward the end of the day (late in the second
sample) the employee spent several minutes sweeping the oven room and
cleaning other process areas. 

In the final sampling period, and during the second short-term sample,
Production Operator A measured and poured a sample of butter flavor
using the following process:

Lift lid to check fill-level of the tank;

Gather measuring container and adjust tilt angle of 50 gallon barrel of
butter flavor (spigot is between knee and hip height when barrel is
tilted on portable stand);

Position measuring container below spigot;

Bending over barrel end, open spigot and fill container to the
prescribed level (several quarts);

Close spigot and [typically] carry the open container 15 feet across the
room and up several steps to the mezzanine in front of the liquid
ingredient tanks; 

Lift lid and pour flavor into the appropriate tank (2 minutes for entire
measuring and pouring process);

Carry measuring container around the corner to a sink and rinse with hot
water. (1 minute); and

Adjust flavor barrel upright in its stand.

The entire process takes less than 5 minutes. Two short-term samples
were obtained while this employee entered the flavor room to prepare
batches of liquid ingredient. During the first, the employee and
assistant added a powdered non-diacetyl flavor by placing the bag on the
edge of the tank, slit open the bag, and dumping the ingredients into
the tank. The second short term sample was collected while the employee
measured and poured liquid diacetyl-containing flavor to produce the of
liquid ingredient blend used in the baked snack food.

This employee wore ear plugs, safety glasses and protective footwear.

Production Operator B

Because the intended activity of Production Operator B was anticipated
all day, but eventually cancelled, the employee spent most of the shift
assisting in areas where diacetyl-containing ingredients either were not
present or were present only in unopened bags. This employee had a brief
period of potential exposure to diacetyl-containing ingredients while
using hot water in a sink to cleaning a mixer paddle coated with cheese
flavored seasoning. Production Operator B also helped other production
operators in the dough production area for less than an hour in the
afternoon. A short-term sample was obtained while this employee swept
the area of the enclosed tumbler area; however, the primary debris were
not associated with diacetyl flavors.

Operator B wore ear plugs, safety glasses and protective footwear.

Bystander (Safety and Health Visitor)

This visitor circulated to various locations in the facility where air
monitoring was performed or anticipated. The visitor did not handle or
approach raw dough or diacetyl-containing flavoring.

This visitor wore ear plugs, safety glasses and protective footwear.

 

Table 1.

Job Categories and Activities with Potential for Exposure to Flavors
That Contain Diacetyl 

Food Manufacturing Facility D – Baked Snack Food Production

June, 2008

Job Title	Activities/Potential Sources of Exposure (percent of shift)

Production Operator A 	Monitors dough production and extruding lines
from computer console (80-85%). Walks the lines making visual
inspection. Picks up dough for evaluation (manual handling) (2%). Visits
Q/A lab to deliver samples (3%). Cleans/sweeps (5-10 %). Checks
liquid/flavor tanks (1%). Adds non-diacetyl flavor to tanks (1%). Adds
diacetyl-containing flavor to tanks (1%).

Production Operator B	Performed prep work setting up bag dumping hopper,
mixing tank, enclosed tumbler, and product drier - no flavor involved
(10%). Works with engineers and maintenance personnel (35%). In
administrative area (35%). Assists monitoring conveyer line in dough
production area – no dough handling (10%)*. Cleaning production areas
(5%). Cleaning cheese-flavor seasoning from mixer paddle (5%)*.

Bystander 

(Safety and Health Visitor)	Observes other employees’ activities from
a distance. Dough/raw product production area (15%). Flavor room (2%).
Q/A laboratory (15%). Tumbler area (20%). Packaging room (2%).

* For Operator B, only tasks indicated with an asterisk involved work in
areas where diacetyl flavor was present. 



Exposure Assessment

Sampling and Analytical Methods

Equipment

Sampling and analytical methods are summarized in Table 2. ERG used SKC
AirCheck 2000 sampling pumps combined with an SKC constant pressure
controller (CPC, part number 224-26-CPC) and SKC adjustable-flow single
tube holders to draw either 0.10 or 0.05 liters per minute. The pumps
were checked for calibration levels immediately before sampling and
again at the end of the period using Drycal DC-Lite Primary Air Flow
Meters (5 ml/min to 5000 ml/min), model 412360, manufactured by Bios
International Corporation, Butler, New Jersey.



Table 2. Sampling and Analytical Methods

Analyte &

Sample Type	Sampling & Analytical Method	Media	Air Volume	Air Flow Rate
Sample Duration	Notes/Special Handling



Diacetyl/Acetoin

Short-term samples 

(15 to 30 minutes)

	OSHA Methods

1012 and 1013

[publication anticipated Fall 2008]	Silica gel

200/400 mg

SKC 226-10-3

(two tubes in series)	3 L	100 ml/min

	15-30 minutes	None



Diacetyl/Acetoin

Partial-shift samples (30 to 180 minutes)	OSHA Methods

1012 and 1013 [publication anticipated Fall 2008]	Silica gel

200/400 mg

SKC 226-183

(two tubes in series)	9 L	50 ml/min

	3 hours

	None

Air Velocity, Temperature, Humidity

(real-time)	TSI VelociCalc Air Velocity Meter (Model 9545-A thermal
anemometer)	N/A	N/A	N/A	N/A	None

Notes:

L means liters.

ml/min means milliliters per minute.

SKC# indicates the brand and model number of sampling media.

N/A means “not applicable.”





ERG maintained all sampling media enclosed in appropriate size
semi-opaque plastic tube covers (SKC model 222-3XD-1) attached to the
tube holders. Analysis was performed by the OSHA Salt Lake Technical
Center (SLTC) Laboratory, Sandy, Utah. 

All instruments and calibration equipment were factory calibrated within
the previous 12 months. 

Limit of Quantification

The laboratory reported the numeric value that the analytical equipment
provided. ERG reports these results in Table 3 even when they are below
the limit of quatification (LOQ), believing they offer more information
about the approximate airborne concentration than would a blanket
description of “< LOQ.” ERG notes that while the values below the
LOQ area associated with greater uncertainty than results above the LOQ,
they may suggest trends that could be confirmed later using more precise
techniques (if available). For each result in Table 3, ERG indicates
whether the result is above or below the LOQ and reports the LOQ as a
part per million (ppm) value for each sample, calculated from the per
sample reporting limits of 0.609 micrograms (μg) per sample for
diacetyl and 0.667 μg per sample for acetoin.

Results for samples in which no analyte was detected are reported as ND
(“none detected”). In calculating averages, ERG takes ND to be equal
to 0 parts per million (ppm). The laboratory did not provide a minimum
limit of detection (LOD), since the actual measured values were reported
down to the level at which no analyte was detected.

Blank Samples

ERG submitted media and field blanks for each batch of sampling media.
No measurable background levels of diacetyl and acetoin were reported
for blanks. 

Air Monitoring Results

Diacetyl and Acetoin

All samples were analyzed for diacetyl and acetoin; however, OSHA has
not published an exposure limit for either diacetyl or acetoin. For this
reason, all results are presented relative to each other, rather than
relative to a fixed standard. 

Table 3 provides details of all PBZ and area samples obtained at this
site. ERG collected three consecutive partial-shift PBZ samples each for
Production Operators A and B. These samples were of 57 to 223 minutes
duration and analyzed using OSHA’s 2008 diacetyl/acetoin analytical
method (publication anticipated Fall 2008). 

Two short-term exposure limit (STEL) samples were also obtained for
Production Operator A while this employee made additions to the
non-diacetyl flavor tank and then to the diacetyl-containing flavor tank
[simulated]. To give an indication of the airborne diacetyl in the oven,
packaging, Q/A laboratory and dough/raw product conveyer areas, an area
sample was collected in each location.

Diacetyl was detected throughout the day in the breathing zone of
Production Operator A. Detected levels ranged from 0.006 ppm (below the
LOQ) to 0.164 ppm – associated with a 57-minute sample that included
periods during which the employee was observed handling raw dough (for
quality checks) and performing activities associated with replenishing
the butter flavor/liquid ingredient tank. The overall time-weighted
average (TWA) (461 minutes) for the three consecutive partial-shift
samples collected for this employee is 0.0298 ppm (rounded to 0.03 ppm).
The unsampled portion of the shift was spent in an early morning meeting
in the laboratory (where no diacetyl was detected in an area sample
collected later in the day) and ERG assumes that no exposure occurred
during the unsampled period. Based on this assumption, the 8-hour TWA
for Production Operator A is 0.0286 ppm (rounds to 0.029 ppm). A
15-minute sample obtained while the operator measured and poured the
butter flavor was associated with a result of 0.139 ppm. It is not
immediately clear why this short-term result is slightly lower than the
57 minute sample. 

Results for Production Operator B, who spent a substantial portion of
the shift assisting with activities in other parts of the plant, and the
bystander show that diacetyl was not detected in their breathing zones.

Diacetyl was detected at levels of 0.111 ppm in the flavor room tank
area and 0.011 ppm beside the raw product conveyer (the latter value is
below the LOQ). Other general area test results indicate that no
diacetyl was detected near the baking oven, in the packaging room, and
in the Q/A laboratory. 

Acetoin was not detected in any of the samples collected at Facility D.

Table 3. Individual Air Monitoring Results for Food Manufacturing
Facility D – Baked Snack Food Production



Sample

No.1	

Sample Type	

Sample

Duration

(minutes)	

Sample Volume

(liters)	Diacetyl Results	Acetoin results





Diacetyl 

Conc.2

(ppm)	

LOQ3, 4

(ppm) 	

Result above LOQ?	Acetoin 

Conc. 2

(ppm) 	

LOQ4, 5

(ppm) 	

Result above LOQ?

Production Operator A

D-5	TWA	181	8.72	0.017	0.019	No	ND	0.021	No

D-6	TWA 	223	11.10	0.006	0.016	No	ND	0.017	No

D-11	TWA	57	2.90	0.164	0.059	Yes	ND	0.064	No

--	[Overall TWA for period monitored]	[461]	--	[0.0298]	--	--	[ND]	--	--

D-8	STEL	25	2.27	ND	0.076	No 	ND	0.082	No

D-14	STEL 	15	1.40	0.139	0.124	Yes	ND	0.132	No

Production Operator B

D-1	TWA	195	9.83	ND	0.018	No	ND	0.019	No

D-9	TWA	169	8.70	ND	0.02	No	ND	0.021	No

D-13	TWA	96	5.10	ND	0.034	No	ND	0.036	No

	[Overall TWA for period monitored]	[460]	--	[ND]	--	--	[ND]	--	--

D-2	STEL	30	3.10	ND	0.056	No	ND	0.06	No

By Stander (Health & Safety Observer)

D-4	TWA	180	9.75	ND	0.018	No	ND	0.019	No

D-7	TWA	220	10.14	ND	1.017	No	ND	0.183	No

	TWA for shift	400	--	ND	--	No	ND	--	--

Oven Area

D-3	Area – Between ovens 2 and 3, adjacent to conveyer turn-around –
product visible through oven opening	185	9.56	ND	0.018	No	ND	0.019	No

Packaging Area

C-3	Area - across aisle from package machines	257	13.51	ND	0.013	No	ND
0.014	No

Flavor Room

D-10	Area – wall behind tanks	190	9.60	0.111	0.018	Yes	ND	0.019	No

Q/A Lab

D-12	Area – bench top	194	9.66	ND	0.018	No	ND	0.019	No



Table 3. Individual Air Monitoring Results for Food Manufacturing
Facility D – Baked Snack Food Production



Sample

No.1	

Sample Type	

Sample

Duration

(minutes)	

Sample Volume

(liters)	Diacetyl Results	Acetoin results





Diacetyl 

Conc.2

(ppm)	

LOQ3, 4

(ppm) 	

Result above LOQ?	Acetoin 

Conc. 2

(ppm) 	

LOQ4, 5

(ppm) 	

Result above LOQ?

Conveyer

D-15	Area – control station	176	9.32	0.011	0.019	No	ND	0.02	No

 Notes

1       	Samples are listed by worker. Some workers were sampled more
than one time; as a result some of the sample numbers 	are out of
sequence. 

2	Sample concentrations are based on the period monitored (i.e., the
sample duration). Samples were reported based on 	any amount of analyte
that was found; reporting limits were not considered.

0.609 micrograms (μg) per sample. 

4	The values listed in the LOQ columns are the LOQ concentrations based
on reporting limits and the sample volumes.

5	The reporting limit (limit of quantification) for acetoin is 0.667 μg
per sample.

*An ND result was reported by the laboratory when there was no
indication of the analyte in the sample. 

Temperature and humidity during collection of samples. D-5, -6, -11, -8,
-14, and D-1, -9, -13, -2, and D-10, -15, C-3: T=82˚F, RH = 41%; D-3:
T=94˚F, RH= 28%.	Abbreviations

Conc. =	Concentration.

LOQ = Limit of quantification.

ND = Not detectable. 

TWA = Time weighted average (for period sampled).

STEL = Short term exposure limit.







Previous Diacetyl Air Monitoring

As mentioned earlier, an industrial hygiene consultant evaluated
diacetyl vapor levels in the Facility D flavor room in late 2006. A
30-minute general area screening sample was associated with a result of
less than 2.5 mg/m3 (below the limit of detection). The consultant used
this result to calculate an 8-hour TWA, reported as less than 0.16
mg/m3. No other diacetyl values were reported. Converted to parts per
million, these results correspond to values of less than 0.710 ppm (for
the 30-minute sample) and less than 0.044 ppm (calculated 8-hour TWA).

The table of sampling results from 2006 made available to ERG did not
list activities that occurred during the 30-minute sampling period or
the sampling and analytical method used to obtain the result.

Examples of Controls in Place

Overview

Most controls in place at Facility D are secondary benefits of workplace
conditions developed for other reasons (e.g., substantial heat removed
with the oven exhaust system creates a constant air exchange in the
adjacent spaces, including the dough production area). ERG noted that
the following existing controls (intentionally or unintentionally) may
help reduce the diacetyl and acetoin exposure of production workers at
the plant.

The oven exhaust system also exhausts the adjacent dough production
area. ERG measured an average of 300 fpm air velocity through 78 square
feet of wall opening between the two spaces (openings included man
doors, conveyer passage slots). This exhaust rate results in an
estimated 23,400 cubic feet per minute (cfm) of air exchange in the
dough production area caused by the oven ventilation.

The facility has an active housekeeping/sanitation program. Ingredients,
dough, and product rarely spilled and if it did, the volume was small
and it was cleaned up quickly.

An automated flavor/liquid delivery system meters the blended flavor
solution into the dough mixer from a tank. An employee must measure and
pour the flavor just once or twice per shift, rather than for every
batch of dough.

Other aspects of the dough/raw product lines are controlled remotely
from central computer consoles. This minimizes the amount of time
employees physically handle dough and maximizes the distance between the
employees and the dough.

Loosely covered flavor tanks limit air circulation at the surface of
tank contents.

Slight positive pressure in the Q/A laboratory minimizes the amount of
air/vapor from the dough production area that enters the lab.

A canopy ventilation hood that is fit closely (few inches gap) over a
tote of discarded baked snack food pieces in the oven area might capture
vapors (in addition to heat) if off-gassing occurs from warm baked or
partially baked, but rejected product. Note that an even tighter fit
(less gap), a wider canopy, or a different hood design might offer even
better protection if it is ever determined that vapor from waste baked
product needs to be controlled.

The tumbler and associated drier are both fully enclosed and fitted with
exhaust ventilation, again probably to control heat and possibly product
dust. These ventilation systems have the added advantage of potentially
reducing the amount of diacetyl vapor and dust/mist that could escape if
a diacetyl-containing coating were to be applied. ERG measured an
average air velocity of approximately 670 fpm into the tumbler housing
through the conveyer inlet (an opening of 7 inches by 10 inches, or
0.486 feet2). This translates to 325 cfm. A larger opening on the other
side contained a constant outflow of baked snack product, which
prevented accurate measurement. Nevertheless, ERG estimates that the air
flow into the tumbler housing may be at least 750 cfm and more likely
1,000 cfm.

Discussion

Relationship between Results, Sources of Exposure, and Controls in Use

General area diacetyl concentrations were greater in the flavor room
(where tanks were loosely covered and flavor was poured) than in the
dough production area. Sources of the airborne diacetyl detected in the
production area might have included air currents from the flavor room
and dough surface area.   

Production Operator A, the employee who worked most closely with the
diactetyl-containing flavor, consistently had detectable levels of
diactyl exposure. As might be expected, the greatest exposure occurred
during direct handling (pouring) of the flavor product. The 15-minute
STEL (0.139 ppm) was obtained concurrently with the 57-minute TWA sample
(0.164 ppm), thus exposure from the flavor pouring task was captured by
both samples. During periods when the employee did not pour flavor, the
modest levels detected (below the level of quantification) might have
arisen from entering the flavor room to check tank levels or from the
same sources previously mentioned as potential sources of diactyl in the
dough production room. For example, this employee handled dough for a
brief moment several times during the shift. 

It is interesting to note that the area sample result (0.011 ppm )
obtained at an empty workstation next to the conveyer carrying
diacetyl-containing dough was quite similar to the exposure levels
reported for Production Operator A during periods without flavor pouring
(0.017 and 0.006 ppm). All three of these results are below the limit of
quantification. No diacetyl was detected in the breathing zone of
Production Operator B, who spent less than an hour in the dough
production area and did not handle dough.

Notable air movement throughout the dough production area might have
helped keep exposure levels lower than they would have been if the air
exchange rate were lower, although the air flow direction from the
flavor room toward the oven room likely permitted some vapors to move
down the length of the production room. This flow pattern was driven by
dough production room air (an estimated 23,400 cfm) pulled as makeup air
into the oven room, which contained ovens fitted with large exhaust
ducts. Although a solid wall and closed doors separated the two spaces,
air moved rapidly through conveyer passage slots in the wall.

 Facility D is aware that, in response to customer concerns, where
possible many flavor manufacturers are reducing the diacetyl content of
butter flavors to the extent that they can. A facility representative
described anecdotal information about cases where food manufacturers
suspected or learned that the diacetyl content did not match the
percentage listed on the MSDS and instead was lower than the listed
value or range. 

 If the flavor contains 1 percent diacetyl and flavoring usage is 10
pounds twice per shift, ERG estimates that daily diacetyl usage might be
0.01x20= 0.2 pounds per shift.

 These tunnel ovens are fully enclosed except where product conveyers
enter and exit, and at a “turn around” point at the far end where
product shifting to the return conveyer is visible through an
approximately 6 square foot access slot.  Each oven is exhausted through
the high ceiling by vertical ducts (approximately 24-inch diameter)
rising directly from the oven housing (top). The exhaust rate was not
available; however, the number and diameter of ducts suggests
substantial air rises out of this room. See additional notes in Section
4.1.

 The presence of turbulence and moving product prevented meaningful
evaluation of air flow into the housing through the product output
chute, which measured approximately twice the area of the intake port.

#

‰

è

Œ

é

-



 

!

"

#

$

%

A

B

C

D

#D

J

K

g

h

i

ƒ

„

…

†

‡

ˆ

‰

Š

‹

§

¨

©

ª

°

±

Æ

Ç

È

â

ã

ä

å

æ

ç

è

é

ê

 

9

:

;

<

=

>

j

k

l

†

‡

ˆ

‰

Š

‹

Œ

Ž

ª

j;

#ª

«

¬



±

²

Ç

È

É

ã

ä

å

æ

ç

è

é

ê

ë

j/

j¸

j¬

j

h8

h8

h8

hâ

gdâ

hâ

hâ

 h<

h<

h<

gd<

gd<

hâ

hâ

" híQ

" híQ

Ff

" híQ

" híQ

摧咙)

摧⼍Â

摧咙)

摧柭|

  h

 h<

, usage of diacetyl flavor was half or one-third normal and a flavor
addition was not needed this day. At the end of the shift the facility
arranged a realistic simulation that involved the employee measuring
flavor and pouring it through a funnel into a temporary holding barrel
(for future use) instead of into the tank. The simulated operation took
the same amount of time handling the same product as would have been
involved in the real task.

 PAGE   3 

Site Visits Related to Diacetyl –Food Manufacturing Facility D        
                                               

 

