United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/4-79-001
December 1978
Air
Nonmethane Organic
Emissions from Bread
Producing Operations
-------
EPA-450/4-79-001
Nonmethane Organic Emissions
from Bread Producing Operations
by
Ralph M. Keller
Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
Contract No. 68-02-2524
EPA Project Officer: Thomas F. Lahre
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
December 1978
-------
This report is issued by the U. S. Environmental Protection Agency
to provide technical data of interest to a limited number of readers.
Copies are available free of charge to Federal employees, current
contractors and grantees, and nonprofit organizations in limited
quantities from the Library Services Office (MD-35), U. S. Environmental
Protection Agency, Research Triangle Park, North Carolina 27711; or,
for a fee, from the National Technical Information Service, 5285
Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency
by Midwest Research Institute, 425 Volker Boulevard, Kansas City,
MO 64110, in fulfillment of a contract. The contents of this report
are reproduced herein as received from Midwest Research Institute.
The opinions, findings and conclusions expressed are those of the
author and not necessarily those of the Environmental Protection
Agency. Mention of company or product names is not to be considered
an endorsement by the Environmental Protection Agency.
Publication No. EPA-450/4-79-001
11
-------
PREFACE
This report was prepared for the Environmental Protection Agency
(Mr. Thomas F. Lahre, Project Officer) to summarize results of laboratory ex-
periments carried out under Work Assignment 4 of Contract No. 68-02-2524. The
work was performed in the Environmental and Materials Sciences Division of
Midwest Research Institute under the supervision of Dr. Chatten Cowherd,
Head, Air Quality Assessment Section. Dr. Ralph Keller was the project leader
and the author of this report.
Approved for:
MIDWEST RESEARCH INSTITUTE
L. J^ shannon, Director
Environmental and Materials
Sciences Division
December 22, 1978
ill
-------
SUMMARY
Two experiments were conducted to measure the emissions of total non-
methane organics during the production of bread. One experiment determined
the emissions during the production of bread by the straight dough process.
The other experiment measured emissions during production by the sponge-dough
process. Measurement of the emissions was performed with a gas chromatograph.
Ninety-five percent of the emissions from the processes was found to be
ethanol and the remaining 5% consisted of three other organic compounds. The
straight dough process emitted 0.5 g ethanol per 1 kg of bread produced. Emis-
sions from the sponge-dough process were 4.6 g ethanol per 1 kg of bread pro-
duced. In both processes, emissions released during baking comprised the ma-
jority of the total emissions. Negligible amounts of ethanol remained in the
bread after baking.
These values are low compared to the value of 8 g/kg of bread estimated
through theoretical considerations. One factor affecting this is the difference
in sweetener concentration between the experimental and theoretical cases. In
these experiments, the sweetener concentrations were approximately 5%, while
the theoretical analysis assumed a sweetener concentration of 10%. Adjusting
for the sweetener concentration results in more comparative values between
experimental and theoretical emission rates.
IV
-------
TABLE OF CONTENTS
Preface iii
Summary iv
Figures vi
Tables vi
1. Introduction 1
2. Breadmaking Processes ..... 2
3. Experimental Apparatus and Procedure 4
4. Experimental Data 7
Straight Dough Process 7
Sponge-Dough Process 11
5. Discussion 15
6. Recommendations 16
References 17
Appendices
A. Commercial Bakeries as a Major Source of Reactive
Volatile Organic Gases jo
B. Sample Calculations 37
C. Raw Data For the Two Experiments 42
-------
FIGURES
Number Page
1 Experimental Apparatus 5
2 Ethanol Emission Rate 10
3 Dilution System 12
4 Ethanol Emission Rate 14
TABLES
Number Page
1 Straight Dough and Sponge-Dough Formulations 3
2 Ingredients and Procedure for Straight Dough Process. . . 7
3 Statistics for the Breadmaking Experiment - Straight
Dough Process 9
4 Ingredients and Procedures for the Sponge-Dough Process . 12
5 Statistics for the Breadmaking Experiment - Sponge-Dough
Process 13
VI
-------
SECTION 1
INTRODUCTION
Within the last year, it has been proposed by Mr. David C. Henderson of
the U.S. Environmental Protection Agency (EPA) Region IX, that the production
of certain bakery products entails nonmethane organic emissions. Mr. Henderson
became aware of the production of ethanol by yeast in bakery dough as the
dough was leavened and investigated the possibility of ^he ethanol escaping
from the dough into the atmosphere. He focused his attention primarily on
bread dough although all yeast leavened dough is capable of emitting ethanol.
A paper detailing his investigation is included in Appendix A.
Because Mr. Henderson's data suggested that large bakeries could be poten-
tially significant sources, an experimental program was initiated to measure
the nonmethane organic emissions from the various processing stages of bread-
making. The program was designed to measure emission rates for total nonmethane
organic compounds in the production of bread. Experiments were conducted to
measure emissions from the two common dough processes. On August 4, 1977,
experiments were performed with bread produced by the straight dough process,
and on November 15, 1977, experiments were performed with bread produced by
the sponge-dough process (see Section 2 for a description of these processes).
The following sections of this report discuss the breadmaking processes,
experimental apparatus, data acquired, implications of the data, and recommen-
dation for future investigation.
-------
SECTION 2
BREADMAKING PROCESSES
One essential step in the production of bakery products is the leaven-
ing of dough to a desired volume by the production of carbon dioxide (C02)
within the dough. The production of C02 may be by chemical or biological
means. The bakery processes which use biological leavening produce etha-
nol as a by-product of leavening and are of interest in this report.
The biological production of C02 within bread dough requires the use of
yeast and sucrose (sweetener). The yeast is a fungus which metabolizes the
sucrose through a series of chemical stages to the end products of C02 and
ethanol (ethyl alcohol). The C02 leavens the dough while the ethanol evap-
orates.
A brief discussion of the chemistry of breadmaking may be found in
Mr. Henderson's paper in Appendix A. The chemical reactions and stages are
shown as well as a theoretical estimate of the quantity of ethanol created
in the leavening process.
Bread is generally produced by two methods: one is the straight dough
process, the other is the sponge-dough process. In the straight dough pro-
cess, all ingredients are mixed into the dough in a single—step procedure.
This method is used most often in home baking and to some degree in commer-
cial baking. The sponge-dough process requires that part of the ingredients
be mixed and allowed to ferment for 3.5 to 5 hr, after which the rest of
the ingredients are added. This process is used most often by commercial
bakeries. A listing of the ingredients and their portions for both dough pro-
cesses are shown in Table 1.1/
In the straight dough process, the ingredients are mixed until a smooth
dough is obtained. After the dough ferments for 2 to 4 hr at 78 to 82°F, it
is kneaded and allowed to ferment for 20 to 30 min. Next, the dough is di-
vided, panned, and allowed to ferment at 95 to 105°F for up to 60 min.
Finally, the dough is baked at 460°F. The bread is allowed to cool before
packaging.
-------
TABLE 1. STRAIGHT DOUGH AND SPONGE-DOUGH FORMULATIONS
Sponge-dough
Straight Sponge Dough
FlourB/
Water (variable)
Yeast
Yeast food
Salt
Sweetener (solid)
Fat
Nonfat milk solids
Softener
Rope and mold inhibitor
Dough improver
Enrichment
100.0
65.00
3.0
0.2-0.5
2.25
8-10
3.0
3.0
0.2-0.5
0.125
0-0.5
as needed
65.0 35.0
40.0 25.0
2.5
0.2-0.5
2.25
8-10
3.0
3.0
0.2-0.
0.125
0-0.5
5
as needed
a/ Ingredients based on 100 parts flour.
In the sponge-dough process, a portion of the flour and water are mixed
with yeast and yeast food. This sponge ferments for 3.5 to 5 hr at 80°F.
The rest of the ingredients are then added and mixed, and the dough is fer-
mented for another 20 to 30 min. Then the dough is divided, rounded and
panned, and allowed to ferment for 60 min. Finally, the dough is baked at
approximately 460°F for up to 20 min. The bread is cooled before packaging.
A detailed description of these processes is presented in Reference 1.
-------
SECTION 3
EXPERIMENTAL APPARATUS AND PROCEDURE
In order to measure the emissions of organics during each stage of the
breadmaking process, the dough was enclosed in a plastic tent. Air was with-
drawn continuously from the top of the tent by the sampling instrument, a gas
chromatograph. Air entered at the bottom of the tent to replace the air with-
drawn at the top. A schematic of the laboratory apparatus is shown in Figure 1.
The dough was baked in an electric oven which had three exhaust ports on
its top side (see Figure 1). Proper operation of the oven shown in Figure 1
required the ports in the top to be left open. The flow rate of the air from
these ports was measured to determine the quantity of air moving through the
oven, and thus the dilution of the emissions by the air flowing through the
oven. Samples for organics analysis were taken from the exhaust stream just
inside the oven.
The concentrations of organics in the sampling stream were determined with
a Beckman Model 6800 Air Quality Chromatograph (total hydrocarbon analyzer) op-
erating under its normal conditions (no special adaptations). A stainless steel
bellows pump inside the chromatograph pulled the sampling stream from the tent
or oven into the chromatograph at a continuous rate of 800 cu cm/min. At 5-min
intervals, the chromatograph measured the organic concentration in the sampling
stream. The chromatograph was calibrated with gas containing 5.1 ppm total hy-
drocarbons consisting of methane.
Because of the high organic concentrations found in the experiment involv-
ing the straight dough process (August 4, 1977), a dilution system was con-
structed to reduce concentrations in the subsequent experiment involving the
sponge-dough process (November 15, 1977). The dilution system, which was used
to reduce the concentration of organics by as much as a factor of 100, involved
the use of two gas flow meters to determine the flow rates of the dilution air
and the undiluted sampling stream. The chromatograph sampled from the diluted
stream.
The ambient air was sampled periodically to determine the background or-
ganics concentration in the air entering the tent and oven. Since the air was
drawn into the tent and oven to replace the air removed by the sampling stream,
no losses of emitted organics were expected from the bottom of the tent or oven.
-------
Ambient
Air
Dilution
System
Beckman
Chromatograph
I
Exhaust
Clean
Air
Figure 1. Experimental Apparatus
-------
The sampling stream exhaust port shown in Figure 1 was directed into an ex-
haust hood which removed the air from the room.
The experiments began with the measuring of the weights of each ingredi-
ent. The ingredients were mixed and the dough was placed inside the tent for
the appropriate time. The chromatograph pump pulled air from the tent continu-
ously.
Next, the dough was kneaded (inside the tent) in the straight dough pro-
cess, or mixed with the rest of the ingredients in the sponge-dough process.
The dough went through its second fermentation inside the tent. The dough was
then panned, proofed, and placed in the oven.
For the sponge-dough process experiment, the loaves were placed in the
tent after baking and allowed to cool. Emission estimates were obtained for
the 40-min cooling period.
Throughout both breadmaking processes, 20-g samples of the dough and
bread were taken to determine the inverted sugar or ethanol concentrations.
The chemical method for determing the sweetener concentration in the dough
involved the conversion of the sweetener (sucrose sugars) to inverted sugars
(fructose and glucose sugars) and the measurement of the quantity of inverted
sugars present, as detailed in Methods of Analysis of the Association of Of-
ficial Analytical Chemists^- The direct measurement of sucrose is very dif-
ficult and is not generally used.
The quantity of inverted sugar expected from the quantity of sweetener
initially present can be stoichiometrically calculated. Ideally, the differ-
ence between the quantity measured in the dough and the calculated quantity
is the amount of sweetener converted to CC>2 and ethanol by the yeast. Un-
fortunately, in these experiments it was not possible to obtain an accurate
measure of the sweetener converted because of starch conversions to inverted
sugar.
The ethanol present in the bread was determined by absorbing the ethanol
into water and measuring the concentration in the water. The concentrations
were measured by standard gas chromatograph techniques.
Both the sweetener and ethanol concentrations were related to the quan-
tity of dough or bread in the sample to give a concentration in the dough or
bread at the time the sample was taken.
-------
SECTION 4
EXPERIMENTAL DATA
STRAIGHT DOUGH PROCESS
On August 4, 1977, an experiment was conducted to determine the emis-
sion rates of nonmethane organics from bread dough mixed by the straight
dough process. The weight of the mixed dough was l,48o g (3.3 Ib) with a
sugar concentration of 3% (45 g (0.1 Ib)). A listing of the ingredient weights
is given in Table 2. The yeast and other ingredients were obtained from a
local grocery store. The yeast was not of a type acclimated specifically to
bread dough.
TABLE 2. INGREDIENTS AND PROCEDURE FOR STRAIGHT DOUGH PROCESS
* Ingredients: 895 g flour, 360 g water, 102.5 g
milk, 57 g margarine, 45 g sweetener,
14 g salt, 11 g yeast
* Mix well (10 rain)
* Let rise (3 hr)
* Knead (15 min)
* Let rise (1 hr)
* Pan (10 min)
* Bake (40 min)
After mixing, a 20-g sample was taken for analysis to determine the sugar
concentration in the dough. The analysis indicated a concentration of 5.3%, by
weight, of sugar in the dough, which is higher than would be expected from the
quantity of sweetener added to the dough. This may be due to conversion of
some of the flour to inverted sugar by the yeast. The error in the quantita-
tive analysis was less than 5%.
The dough was placed in the tent for its first 3-hr fermenting period.
Then the dough was kneaded for 15 min inside the tent and allowed to rise for
another 60 min. Finally, the dough was formed into two loaves, approximately
equal in weight, and then baked for 40 min at 360°F.
-------
Throughout the experiment, two chromatographs were used to follow the
organics emitted from the dough. The Beckman chromatograph measured the total
nonmethane organic concentration in the sampling stream. The other chromato-
graph, a Varian 2400 Gas Chromatograph, simultaneously measured the ethanol
concentration in the sampling stream. The Varian chromatograph used a Chromo-
sorb 102 column at 120°C and an FID detector. Examples of the output of the
Beckman and Varian chromatographs are shown in Appendix B.
After the bread was removed from the oven, it was weighed. Also, samples
were taken for analysis to determine the inverted sugar content and ethanol
concentration in the bread. The inverted sugar concentration was 4.9%, and
the ethanol concentration in the bread was less than 0.2%. A summary of the
times, weights, and the inverted sugar and ethanol concentrations are shown
in Table 3. The data from the experiment is included in Appendix C»
The chromatographs measured concentrations of organics in parts per mil-
lion by volume in the sampling stream. These concentrations were converted to
mass emission rates by using the sampling stream flow rate, molar weight of
ethanol and the molar volume of ethanol. The concentration of the sampling
stream was adjusted by the dilution ratio. An example of the calculations is
shown in Appendix B.
The ethanol emission rates measured during the experiment are shown
plotted in Figure 2. Graphical integration of this curve results in the total
emissions of ethanol during the time period. The value obtained when integra-
tion was carried out using 10-min interval values was 0,71 g.
When the concentration of ethanol was compared with the concentration of
total nonmethane organics, it was found that 95% of the nonmethane organics
emitted by the dough was ethanol. Three organic compounds constituted the
other 5% of the nonmethane organics emitted. These three organics were of
low molecular weight and may have been aldehydes or alcohols.
The total ethanol emissions of 0.71 g from 1,282 g bread when converted
to the emissions per kilogram of bread produced the following emission fac-
tor for the straight dough process:
Ethanol emissions: 0.5 g ethanol/kg bread produced
(0.5 lb/1,000 Ib bread)
The emissions during baking made up 77% of the total emissions measured, while
the other steps accounted for 23% of the emissions.
The emissions measured were approximately 6% of the factor predicted
by Mr. Henderson in his paper. The discrepancy between the values may be
due to the initial sugar concentration and the type of yeast used. It is
-------
TABLE 3. STATISTICS FOR THE BREADMAKING EXPERIMENT - STRAIGHT DOUGH PROCESS
Mixed
Time elapsed (min) 0
Dough weights (g) 1,480
Inverted sugar con- 5»3% (dough)
concentration
Cumulative Emissions 0
(g ethanol)
End of End of End of End of
first rise kneading second rise baking
210 240 320 370
a/ 1,377 1,373 1,282
a./ _al al 4.9% (bread )^
0.04 0.05 0.09 0.71
_a/ Not measured.
b/ By weight, as determined by chemical analysis. This includes sucrose added as sweetener as well
as any natural sugars present.
-------
20,000
. 10,000
c
I
STRAIGHT DOUGH PROCESS
8-4-77
o
o
_C
600
400
200
0
9:00
Second Rise
Bake
10:00
11:00 12:00 1:00
Time - Hour of the Day
2:00
3:00
4:00
Figure 2. Ethanol Emission Rate
-------
reasonable to assume that a larger concentration of sugar in the dough would
result in greater production of ethanol and CC^* The yeast used was an off-
the-shelf variety rather than a special yeast acclimated to the conditions
of commercial bread producing. It would be expected that the emission factor
would increase with increased sugar content and the use of the proper yeast.
However, the degree of increase cannot be accurately predicted without knowing
the specific relationship of emission rate to sugar content and yeast type.
Initially, it was believed that a mass balance could be constructed
using the quantity of (a) sugar consumed (inverted sugar concentrations);
(b) ethanol emitted; and (c) the quantity of ethanol in the bread after
baking. However, this was not the case since it is possible for the yeast
to break down the flour to form inverted sugars and thus ethanol. The con-
centrations of inverted sugars and ethanol have been included in this re-
port for the reader's reference.
SPONGE-DOUGH PROCESS
On November 15, an experiment was conducted to measure the emission of
ethanol during bread production using the sponge-dough process. The experi-
ment was designed to measure ethanol emissions with the Beckman total hydro-
carbon analyzer during rising, proofing, baking, and cooling. It was assumed
that ethanol was the major organic component of the emissions.
Because of the high emissions expected, a dilution system was devised
that could dilute the emission by a factor of 100 to 1. Two gas flow meters
were used to measure the flow rates of the emissions sampling stream and dilu-
tion air. The chromatograph sampled from this diluted stream to measure the
ethanol concentration.
The dough was mixed according to the directions of Reference 1. The pro-
cedure and ingredients were as shown in Table 4. In the total dough mix
(after samples) of 1,126.5 g, the sugar concentration was 6.2%.
Samples were taken throughout the process for determination of the in-
verted sugar concentration. Samples were taken after the initial mixing, at
the end of the first rise, after the final mixing, before baking, and after
baking. These concentrations are shown in Table 5 along with the time sched-
ule and dough weights. The dough was divided into 2 loaves when it was
panned. The total dough weights are shown.
The ethanol concentration in the bread after it was taken from the
oven was found to be 1 ppm by weight. The low value indicates that very
little ethanol remains in the bread.
The experiment measured the emissions of ethanol during the first rise
period, second rise period, while baking, and as the bread was cooling. The
emission rates for the experiment are shown in Figure 3. These were obtained
11
-------
TABLE 4. INGREDIENTS AND PROCEDURES FOR THE SPONGE-DOUGH PROCESS
* Ingredients: 490 g flour, 455 g water, 21 g acclimated
yeast, 3.5 g yeast food
* Combine ingredients, mix well
* Allow to rise for 3.5-4.5 hr
* Add: 210 g flour, 70 g sweetener, 21 g fat, 21 g milk,
16 g salt
* Mix
* Allow to rise for 20-30 min
* Divide and pan
* Let rise for 60 min
* Bake
* Cool'
in a similar manner as described for the August 4 experiment. The experimen-
tal data are included in Appendix C. Graphical integration of these rates
yields a factor of 4.6 g of ethanol per kilogram of bread produced. The
emissions during baking constitute 99% of the total emissions. These emissions
of 4.6 g/kg of bread are approximately half of those estimated by Mr. Henderson.
The inverted sugar concentrations and ethanol concentrations were not
useful in a mass balance because of the conversion of flour to inverted
sugars. They are presented for the reader's reference.
12
-------
TABLE 5. STATISTICS FOR THE BREADMAKING EXPERIMENT - SPONGE-DOUGH PROCESS
Time elapsed (min)
Dough weights (g)
Inverted sugar
Mixed
0
969.5
1.2%
Total mix
240
1,175
5.3%
Before baking
330
1,126.5
6.7%
After baking
515
947.8
5.4%
concentrations
Cumulative emissions
(g of ethanol)
0
0.003
0.004
4.4
u>
-------
I 60
SPONGE DOUGH PROCESS
11-15-77
§
o *0
V
a
<* 20
c
o
'a
J5 n
—
—
_
l I l i i i
/ \
/ \
/ \
/
. /
V/
^
1 1
i
7 30
i
— 20
o
.2 10
vt
^vt
UJ
0
I
Baking
(Oven)
Cooling
(Tent)
8:30 9:00 10:00 11:00 12:00 1:00
Time - Hour of the Day
2:00
3:00
4:00
5:00
Figure 3. Ethanol Emission Rate
-------
SECTION 5
DISCUSSION
In attempting to apply the experimental data presented in this report to
commercial bakeries, at least three factors affect the application. First, the
production processes in bakeries use machines to mix, knead, and pan, as well
as for most of the other operations. Furthermore, the processes take place in
temperature and humidity controlled rooms. The machines and environmental
conditions will have some effect on the production of ethanol in the nonbaking
steps of the processes.
Secondly, the sweetener content is generally 8 to 10% of the dough
weight. This work has assumed a linear relationship between emissions and
sweetener concentration to compare expected and experimental emissions.
This relationship has not been tested and may introduce error.
Finally, another factor not considered in this work is the effect of
the natural gas burners on the emissions. If the ethanol emissions reach
the flames or are heated to a sufficient temperature, they will be burned.
This will result in lower stack emissions from bread baking than predicted by
this work.
In discussing this possibility with one baking expert,—' it was found
that most bread baking is done in ovens which fire at the bottom with hot
air passing up and around the dough with the temperature of the dough not
exceeding 500°F. There is a possibility that in certain regions of the
oven, flames may reach the ethanol vapors, but this is not likely.
For cakes and some pastry items, the oven has flames above and below
the dough. In these ovens, the possibility of flames reaching the ethanol
vapor is very high. Therefore, in these types of ovens, ethanol emissions
would be expected to be reduced.
The effect of the above variables on emissions can best be determined by
field tests at a commercial bakery.
15
-------
SECTION 6
RECOMMENDATIONS
The experimental work reported here indicates that nonmethane organic
emissions occur during the production of bread. These emissions were found to
consist primarily of ethanol, mostly generated during baking. A more represen-
tative estimate of the emissions from bread production would be obtained if
the sponge-dough process experiment would be repeated using 10% sweetener con-
centration.
The next step in determining the organic emissions from bakeries would be
field experimentation at a bakery site. Since the majority of the emissions oc-
cur during baking, stack sampling of the oven exhaust would have highest prior-
ity. The emissions measured on-site would be representative of the true bakery
conditions.
16
-------
REFERENCES
1. Pomeranz, Y. Editor. Wheat/Chemistry and Technology. Published by the
American Association of Cereal Chemists, St. Paul, Minnesota, 1971.
p. 676.
2. Harwitz, W. Editor. Official Methods of Analysis of the Association
of Official Analytical Chemists. Published by the Association of Of-
ficial Analytical Chemists, Washington, B.C., 1975.
3. Personal communication with representative of Interstate Brands,
Kansas City, Missouri, July 1975.
17
-------
APPENDIX A
COMMERCIAL BAKERIES AS A MAJOR SOURCE OF REACTIVE
VOLATILE ORGANIC GASES
18
-------
ENVIRONMENTAL PROTECTION AGENCY
REGION IX
SURVEILLANCE & ANALYSIS DIVISION
Title: Commercial Bakeries as a Major Source of
Reactive Volatile Organic Gases
Prepared By: David C. Henderson
Environmental Engineer
Air Section, Air & Hazardous Materials Branch
Surveillance & Analysis Division, EPA Region IX
215 Fremont Street
San Francisco CA 94105
(415) 556-8047
Date: December, 1977
19
-------
Summary:
The baking industry appears to represent a major source of
photochemically reactive volatile organic gases in the form of
ethyl alcohol and other gases. Yeast fermentation of bread
baking doughs produces pyruvic acid and acetaldehyde as inter-
mediate products and about equal molar amounts of ethyl alcohol
and carbon dioxide gas (C02) as final products. Recent source
tests performed by an EPA contractor have validated theoretical
estimates of the magnitude of the emissions. Emission factors
are presented in this paper on a per capita and production rate
basis. Large bakeries can emit up to 168 tons/year of ethyl
alcohol.
Purpose:
The purpose of this report is to provide information on
volatile organic gas emissions from the baking industry in order
that emission estimates can be included in emission inventories
currently being developed in EPA, Region IX.
Background:
The art of bread baking has changed little in the 2,000
years since the Egyptians discoverd the leavening of bread.
The basic ingredients of bread are flour, water, salt, sugar,
and yeast. Other ingredients are added to enhance the flavor or
texture of the desired product.
20
-------
The role of the yeast in bread baking is to produce carbon
dioxide gas. The evolving CCL raises or "leavens" the bread
dough to a desired volume. Yeast produces the COg by anaero-
bically decomposing the sugar in the natural metabolic process
known as alcoholic fermentation. Alcoholic fermentation of
sugar by yeast produces equal, amounts of C02 gas and ethyl
alcohol, with pyruvic acid and acetaldehyde also produced as
intermediaries.
In a commercial bakery, bread dough is allowed to ferment
from two to four hours prior to baking at an oven temperature of
450°F. The temperature inside the bread does not exceed 212°F.
The ovens used in commercial bread bakeries are predominately
fired by natural gas and are direct fired. In direct fired
ovens, any vapors driven off the bread and any combustion product
gases are removed throught the same exhaust vent. The aroma
associated with fresh baked bread, in the locale of a bakery, is
actually fermentation of alcohols, aldehydes, and possibly other
organics being emitted to the atmosphere.
It is believed that alcohol is produced as a liquid within
the bread dough during the fermentation period. Part of the
alcohol is driven off the bread during oven baking. Since the
oven is operating at 450°F, it may be possible that the alcohol
is undergoing a chemical reaction, such as dehydrogenation to
form aldehydes or esters, before it is exhausted from the oven.
21
-------
Also since a carboxylic acid (pyruvic) and acetaldehyde are
produced as intermediaries, these substances too could be under-
going some type of chemical reaction prior to being exhausted
from the oven.
Bakery products can be divided into two groups; products
that are yeast leavened, and products which are chemically
leavened by baking powder. This review is only concerned with
yeast leavened bakery products. Yeast leavened bakery products
include most breads, sweet rolls, sweet yeast goods, ordinary
crackers, pretzels, and doughnuts excepting cake doughnuts.
Chemically leavened bakery products include cakes, cookies, cake
doughnuts, and quick breads such as corn bread or baking powder
(?\
bisquits. '
STOICHIOMETRY AND MECHANISM OF ALCOHOLIC FERMENTATION BY YEAST:
The following chemical reactions occur during yeast fermentation
of sucrose:
HOH
CH° + CH°
AC. *C. J.J.
(sucrose)
2C6H12°6
4CH.COCOOH
yeast
enzymes
yeast
enzymes
f U J.C. V U Xt U
(fructose)+( glucose)
. 4CH7COCOOH
r 3
(pyruvic acid)
• 4CH.CHO + 4CO?
(acetaldehyde)
22
-------
yeast
4CH3CHO enzymes y 4CH3CH2OH
(ethyl alcohol)
CALCULATION OF ETHYL ALCOHOL PRODUCED PER TON OF SUCROSE
CONSUMED:
For each mole of sucrose consumed four moles of ethyl
alcohol are produced.
/
Molecular weight of sucrose =
C12 = 12 X 12 = 144
H22 = 22 X 1 = 22
Ou = 16 X 11 = 176
342 Ib./lb. mole
The number of pound moles of sucrose in 1 ton =
Ib. moles _ 2000 1b. 1 Ib. mole _ 5.8
ton sucrose " ton 342 Ib.
Since for each mole of sucrose consumed four moles of
alcohol are produced, therefore, for each ton of sucrose con-
sumed there are 23.4 Ib. moles of alcohol produced.
23.4 Ib. moles
5.8 Ib. moles sucrose 4 Ib. moles alcohol _ alcohol
ton sucrose
Molecular weight of ethyl alcohol =
C2 = 12 X 2 = 24
H6 = 6X1= 6
0 = 16 X 1 = 16
46 Ib./lb. mole
23
-------
Pounds of ethyl alcohol produced per ton of sucrose con-
sumed i s =
Ib. ethyl alcohol _ 23.4 Ib. moles alcohol 46 Ib.
ton sucrose ton sucrose Ib. mole alcohol
= 1076 Ib. ethyl alcohol produced
ton sucrose consumed
or, _ .54 Ib. ethyl alcohol produced
Ib. sucrose consumed
However, it is not necessary for sucrose or any other
sweetener to be added to the bread for alcohol to be evolved.
In the absence of a carbohydrate sweetener, the yeast will
reduce carbohydrates in the wheat to maltose and ultimately to
ethyl alcohol and carbon dioxide. A good example of a bread
made without an added sugar is some of the San Francisco sour
dough breads.
Therefore, the amount of ethyl alcohol could be greater
than the value calculated for each pound of sucrose consumed.
ESTIMATION OF BAKERY EMISSIONS BASED ON BREAD PRODUCTION RATES:
The nation's largest bread baker, Continental ITT and
the major manufacturer of commercial baking ovens, Baker-Perkin*
were contacted to determine if any studies had been conducted or
tests performed to establish volatile organic gas emissions from
the fermentation process in commercial bakeries. It was learned
that there is no reliable information available to estimate
emissions from the baking industry. A source test was performed
24
-------
by Baker-Perkin1s contractor, Clayton Environmental, during
August, 1974, on a direct fired commercial bakery oven. The
hydrocarbon emission rate from this test was determined to be
1 Ib. HC/1000 Ib. bread. However, according to Continental ITT,
this test only measured hydrocarbons (i.e. compounds containing
only hydrogen and carbon). Ethyl alcohol, acetaldehyde, and
pyruvic acid are not strictly hydrocarbons as they contain oxygen
in addition to hydrogen and carbon. Also, the test was considered
to be unreliable ana not reproducible by Continental ITT observers
on the scene.
Continental ITT is the largest baker in the country, pro-
ducing Wonder Bread, Hostess Cupcakes, Hostess Twinkies, and
many other name brand baked goods. At the request of the EPA
Regional Office, Continental ITT's Research Department developed
an estimated emission factor of:
8.0 Ib. ethyl alchohol emitted
1000 Ib. bread baked
The emission factor estimated by Continental ITT is based
on the following assumptions:
(1). 8-10$ of bread's dough weight is sugar
(2). 10-15% of sweet rolls' dough weight is sugar
(3). Only 2-3% of the bread or seeet rolls' dough weight is
consumed and attributable to alcoholic fermentation
25
-------
(4). 50% of the dough weight loss is converted to alcohol
(Note - this correlates well with the weight of alcohol
produced per pound of sugar consumed, which was
calculated as 54% earlier in this report.)
(5). Some of the alcohol remains in the bread
(Note - Continental ITT is estimating that between 25%
and 53% of the alcohol produced remains in the
bread.)
Based on the information provided by ITT Continental, it
was calculated that a large commercial bakery could emit over
100 tons/year of volatile organic gases.
In order to develop an accurate estimate of emissions from
this source, EPA contracted with Midwest Research Institute
(MRI), Kansas City, Missouri, to perform source tests. The
first test was conducted during August, 1977. Two loaves of
bread, each weighing approximately 1.6 pounds were baked in the
laboratory. A straight dough mix was used with a sugar concen-
tration of 5.3% by dough weight. Plastic tents were constructed
over the areas where the bread was mixed, kneaded, and allowed
to rise. Sampling probes were placed in the oven during baking.
A gas chromatograph (G.C.) sampling pump continuously with-
drew samples for analysis from within the plastic tent. The
emission rates of the hydrocarbons were obtained by knowing the
26
-------
flow rate of the G.C. sampling pump and obtaining the concentra-
tion on the flow stream from the G.C. An overall emission
factor of about .8 Ib. ethyl alcohol per pound of bread produced
was calculated from this test data. However, this test was not
considered valid for the following reasons:
1. The bread mix was straight dough and not a commercial
sponge dough process. Home baked breads are normally
straight dough mixes while commercial breads are
normally sponge dough mixes.
2. The sugar concentration in the straight dough mix was
only 5.3%. Sponge dough mixes would contain approxi-
mately 10% sugar.
3. The testing was discontinued when the bread was removed
from the oven which is at the peak of its ethyl alcohol
emissions.
4. The yeast used was not a commercial grade yeast.
A second laboratory test was conducted in November, 1977,
by MRI. In this test, a sponge dough mix was used and the yeast
was a commercial grade obtained from the local Wonderbread
bakery. However, the sweetener concentration in the dough was
only 5%.
In this test, the cumulative emission factor for the entire
baking process was approximately 3.0 Ib. of ethyl alcohol emis-
27
-------
sions per 1000 pounds of bread produced. Almost the entire
amount of this emission is evolved during the baking phase.
Based on these two tests, there appears to be a linear
relationship between sweetner concentration and emissions. If
this assumption is correct, then ethyl alcohol emissions of
approximately 8.0 Ib./lOOO Ib. of bread would be expected from
a commercial dough mix with a sweetener concentration of 10%.
During January, 1978, it is expected that M.R.I, will com-
plete a third experiment, using a sweetner concentration of 10%.
A final report, including all test data, should be available from
M.R.I, after this test is completed.
CALCULATION OF EMISSIONS FOR A LARGE COMMERCIAL BAKERY
Using this emission factor, a calculation was made for a
large commercial bakery.
Assuming that a large commercial bakery:
(1) produces 12,000 Ib. bread/hr.;
(2) operates 14 hr./day;
(3) operates 250 day/year,
the calculated emissions for this bakery using the Continental
ITT emission factor would be:
i H i 1 ton 12,000 Ib. bread 8.0 Ib. alcohol 14 hr. 250 days
aicon01 " 2000 Ib.hour"1000 Ib. bread3ayyear
ethyl alcohol emitted = 168 tons
year year
28
-------
This estimate represents the worst case situation, as it is an
example of a very large commerical bakery. A small commercial
bread bakery may produce only 2,000 Ib./hr and only operate 8
hours per day.
ESTIMATION OF BAKERY EMISSIONS ON A PER CAPITA BASIS:
The following table lists annual yeast leavened baked goods
production, excluding retail single-shop bakeries:
PRODUCTS THOUSAND POUNDS/YEAR
White bread 8,861,343
White hearth bread 426,998
Whole wheat and other dark wheat breads 643,216
Rye breads 509,545
Raisin and other speciality breads 419,506
Rolls-bread type 2,063,124
Sweet yeast goods 875,053
Crackers 1,369,194
Pretzels 139,380
Total 15,307,359
The source of this information is the U.S. Census Bureau for the
year 1966. However, it is reported that these figures have not
(2)
changed appreciably in recent years. Although single retail
bake shops are not included in this listing, they are not con-
sidered to be a major factor in the baking industry.
29
-------
The population of the United States for 1976 was 213.6
million.(6)
1b. yeast leavened bake goods consumed
person-year
15,307,359,000 Ib. bake goods
213,600,000 person-year
_ 71 -, 1K yeast leavened bake goods consumed
" /1%/ ID' person-year
Using the emission factor of 8.0 Ib. ethyl alcohol emitted/
1000 Ib. baked goods, the ethyl alcohol per person per year
would be:
_ 8.0 Ib* fthyl alcohol emitted 71.7 Ib. bake goods
1000 Ib. bake goodsperson-year
.57 Ib. ethyl alcohol emitted
person-year
It should be noted that this estimate includes emissions from
cracker and pretzel baking which comprise less than 10% of the
total production of yeast leavened bake goods. The emission
factor of 8.0 Ib. alcohol emitted per 1000 Ib. bread baked,
estimated by Continental ITT, does not apply to crackers and
pretzels, as these products are not manufactured by this
company. However, lacking any additional information, the
Continental ITT emission factor was-also applied to cracker and
pretzel baking.
30
-------
ESTIMATE OF BAKERV EMISSIONS IN THE ^TH COAST AIR BASIN:
For illustrative purposes, an estimate of bakery emissions
for the approximately 11 million people residing in California's
South Coast Air Basin (greater metropolitan Los Angeles area)
would be:
_ .57 Ib. alcohol emitted 11.000,000 persons
person-year
_ 6,270,000 Ib. ethyl alcohol emitted
year
n 6,270.000 Ib. 1 ton
• year2,000 Ib.
_ 3,135 tons 1 year
year365 days
_ 8.6 tons
3ay
t
The total stationary source emissions on non-methane
organics in the South Coast Air Basin for 1974 are estimated
at 651 ton/day. The percentage of emissions from station-
ary sources in the South Coast Air Basin due to bakeries is:
8.6 ton 100
day
651 ton
day
= 1.3%
The following table compares emissions from commercial
bakeries in the South Coast Air Basin to other major source
categories; for Non-Methane Hydrocarbons (NMHC) for 1974:
31
-------
% Total
Stationary Source Category NMHC Emissions
Miscellaneous Organic Solvent Usage 26.8
Surface Coating (Painting, etc.) 25.5
Petroleum Marketing 24.0
Petroleum Refining 7.3
Solvent Degreasing 5.8
Dry Cleaning 4.4
Structural Fires 4.2
Utility Equipment (lawn mowers, etc.) 3.1
Wild Fires . 2.3
Pesticides 1.4
Commercial Bakeries 1.3
Power Generating Plants 1.3
Petroleum Refining-Fuel Combustion .8
Orchard Heaters .6
Industrial Fuel Combustion .5
Petroleum Production .4
Metallurgical Processing .4
According to this estimate, commercial bakeries would be
the llth largest NMHC emission category, within the South Coast
Air Basin, according to this system of classifying sources.
32
-------
BIBLIOGRAPHY
(P Private commrr -jication, Mr. Robe Mantsch, Red Star Yeast
Company, Universal roods Corporation, Milwaukee, WI
(414) 271-6755, May, 1977.
(2) McGraw-Hill Encyclopedia of Science and Technology, 1971,
Volume 5, page 445.
(3) Private communication, Mr. Peter Tobiason, ITT Continental
Baking Co., Inc., Rye, New York, (914) 967-4747, May, 1977.
(4) Private communication, Mr. Petri, Baker-Perkin Corp.,
Saginaw, Michigan, (517) 752-4121, May, 1977.
(5) Statistical Attracts of the U.S., 1976, U.S. Bureau of
the Census.
(6) Preliminary Emissions Inventory and Air Quality Forecast
1974-1994, Final Report, Air Quality Maintenance Planning
Task Force, South Coast-Southeast Desert Air Quality
Maintenance Areas, Final Report, May, 1976.
(7) Preliminary Test Results, conducted on August 4, 1977 and
November 15, 1977, Midwest Research Institute, Kansas City,
Missouri. Mr. Ralph Keller, Project Officer, (816)753-7600.
33
-------
Question:
Keller:
(MRI)
Question;
Keller:
Question:
CONDENSED DISCUSSION
What type was the oven and how did you sample
the emissions from it?
Our oven was an electric heating oven. We
had three holes in the top of the oven and
we had glassware coming out of the holes and
hooked into our GC. We had flow rate
measured out the top. We thus knew the con-
centration and the flow rate coming out the
top. We also believe because of our reali-
tively high flow rate coming out of the oven,
we did not have any organics turning back
onto the electric heaters on the bottom
of the oven.
How did you measure flow from the first of
the operation?
The GC unit has a sampling rate of 2.8 cubic
feet per minute. The holes were stuck into
a bag that had an inlet that was just open to
clean air so that the GC pulled out 2.8
and supposedly clean air came in at 2.8.
How do you feel about the statistical sig-
nificance of only using two loaves for de-
veloping emissions factor? You only did two
loaves - you are going to do two more.
34
-------
Keller:
Question:
Henderson:
Keller:
Question:
Y< s, 15GQ grams.
W; y not make 2. or 30 continuously so you
identify the different variances?
Well we are not baking 20 or 30 because of
money constraints. It took six man days to
do the analysis and baking of just two loaves
of bread. This is quite a lot of money.
I believe that the two or four loaves we
make will provide fairly representative
emission factors. We will be using a
process similar to what bakers use and we
are making sure our bread and dough is
homogenious and that it is representative
of the "standard" loaf of bread.
Again, this was a first look to see how close
we c?me to the theorical values. If we
come out the second time and we find we are
far off then we may discuss with the project
officer to see if we should load up our
equipment, contact a baker and go out to
get field samples.
What were the other species that you
measured besides ethyl alcohol? Was there
anything else picked up in the GC?
35
-------
Keller: We found the ethyl alcohol was 95% of the
hydrocarbons coming out. There was no
methane. The other 5% of hydrocarbons had a
boiling point similar to ethanol. As we
pointed out there are a couple other organic
species that could be coming out, but we did
not sample to see what they were. We also
did not sample the water emission rate.
36
-------
APPENDIX B
SAMPLE CALCULATIONS
37
-------
The calculation of the emission rate for organics involves the adjust-
ment of the concentration reading from the chromatographs for the dilution of
the emissions and the conversion of the concentration by volume to an emission
value expressed in terms of mass/time. Examples of the output of the two chro-
matographs are shown in Figure B-l (Varian) and Figure B-2 (Beckman).
A concentration is obtained from these graphs by subtracting the portion
of the peak height due to background organic concentrations and multiplying
by the calibration value (ppm/cm of peak height).
This concentration is then multiplied by the ratio of the total sampling
stream flow rate to the sampling stream flow rate out of the tent or oven.
The molecular weight of methane and the molar volume at room temperature are
then used to produce a mass emissions value.
Next, the response of the chromatograph to ethanol is taken into account.
A value of 0.28 was determined for this chromatograph by injecting known quan-
tities of ethanol.
An example of this procedure follows.
Example; Sponge-Dough Process, November 15* 1977, 3;30 PM, During Bakina
• Calculation of organics concentration (ppm) from Beckman chromatograph
data:
Peak height: 32 cm
Background: 24 cm
Calibration: 0.137 ppm/cm
Oven dilution factor: 1050 = total oven air flow rate -f sampling rate
of chromatograph.
Concentration = (32-24) x 0.137 x 1050 = 1150 ppm as methane.
• Conversion of ppm to
C = concentration expressed as me'thane (ppm by volume)
V = flow rate (cc/min)
16 g = methane molecular weight
24.5 liters = molar volume; 25°C, 1 atm
0.28 = response factor of chromatograph to ethanol
38
-------
Emission rate c i . '/ cc . 16 3 . l mole . i . io6
~~ 106i min mole 24.5 i 1CH cc g 0.28
= C • V • 1.4 x 1CT3 Ug
= 1150 • 2.74 x IO4 • 2.3 x 10'3 = 73500 ug/min
Note: V = 2.74 x 10^ cc/min obtained from raw data table in Appendix C for
sponge-dough process.
39
-------
Figure B-l. Varian Chromatograph Output
40
-------
^(r\
t -
i-i
i- i
i: II
i H
lu
Figure B-2. Beckman Total Hydrocarbon Analyzer Output
-------
APPENDIX C
RAW DATA. FOR THE TWO EXPERIMENTS
42
-------
AUGUST 4, 1977 BREADMAKING EXPERIMENT
STRAIGHT DOUGH PROCESS
Time
9:30 AM
9s50 AM
10:10 AM
10:30 AM
10:40 AM
10:50 AM
llsOO AM
11:10 AM
11:20 AM
11:30 AM
11:40 AM
11:50 AM
12:00 NOON
]2l30 PM
12i40 PM
12:50 PM
1:00 PM
1:15 PM
1:20 PM
1:40 PM
1:50 PM
2:00 PM
2:10 PM
2:30 PM
2«40 PM
3:00 PM
3:10 PM
3:20 PM
3:30 PM
3:40 PM
Ethanol E.R.
d
V
V
Time
elapsed
(min)
0
20
40
60
70
80
90
100
110
120
130
140
150
180
190
200
210
225
230
250
260
270
280
300
310
330
340
350
360
370
= C • V • Ci/0.28
= 16/24,500 = 6.53
= 800 cc/nln (tent
Nonme thane
Concentration
expressed
as methane
(ppm)
0
27
70
83
88
96
84
92
92
120
109
161
150
148
157
157
170
280
282
290
290
270
269
240
240
60
110
119
212
230
(see Appendix B)
x 10-*
)
organics
Accumulated
Ethanol ethanol
emission rate emissions
(iJg/min) (yg) Comments
0 First rise period begins
50
130
150
170
180
160
170
170
2JO
200
300
280
275
290
290
320 40,200 End first rise
520 Kneading
530 49,000 Kneading
540 Second rise period
540
500
500
450
450 88,500 End second rise period
5,800 Baking
10,600
11,400
20,400
22,100 713,000 End baking
= 41,050 cc/«in (ov»n)
-------
NOVEMBER 15, 1977 BREADMAKING EXPERIMENT
SPONGE-DOUGH PROCESS
Nonme thane
Concentration
Time
8:30 AM
10:00 AM
10:10 AM
10:20 AM
11:00 AM
11:10 AM
11:20 AM
11:30 AM
11:40 AM
11:50 AM
12:00 NOON
12:25 PM
12:30 PM
12:40 PM
12:50 PM
1:05 PM
1:20 PM
1:50 PM
2:00 PM
2:10 PM
2:20 PM
2:30 PM
2:40 PM
2:50 PM
3:00 PM
3:10 PM
3:20 PM
3:30 PM
3:40 PM
3:50 PM
4:00 PM
4:10 PM
4:20 PM
4:30 PM
4:40 PM
4:50 PM
5:00 PM
5:10 PM
Ethanol E.R. =
cl
V
V =
Time
elapsed
(min)
0
90
100
110
150
160
170
180
190
200.
210
235
240
250
260
275
290
320
330
340
350
360
370
380
390
400
410
420
430
440
450
460
470
480
490
500
510
520
C • V • Cl/0.28
16/24,500 = 6.53
10 cc/min (tent)
expressed
as methane
(ppm)
0
333
252
252
434
141
292
817
524
888
746
646
1,049
1,341
1,452
1,473
555
424
605
585
575
444
272
212
595
867
1,241
1,150
938
676
535
400
817
1,130
1,120
1,059
938
847
(see Appendix
x 10-4
organics
Accumulated
Ethanol ethanol
emission rate emissions
(yg/min) (yg) Comments
0 First rise
8
7
7
10
3
7
20
10
20
20
15
25
30
35
35 3,200
13 Kneading
10 4,000 Proofing
15
13
13
10 4,500
17,500 In oven
13,500
38,100
55,600 97,400
79,700
73,500
60,200
43,400
34,320
25,700
20 4,423,000 Out of oven
25 In tent
25
25
20
20 4,424,000 Ended
B)
'
2.74 x 10* cc/min (oven)
44
-------
TECHNICAL RtPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/4-79-001
2.
3. RECIPIENT'S ACCESSIOr*NO.
4. TITLE AND SUBTITLE
5. REPORT DATE
Nonmethane Organic Emissions From
Bread Producing Operations
npr.ember 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Ralph M. Keller
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Midwest Research Institute
425 Volker Boulevard
Kansas City, MO 64110
10. PROGRAM ELEMENT NO.
2AA635
11. CONTRACT/GRANT NO.
68-02-2524
12. SPONSORING AGENCY NAME AND ADDRESf
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
EPA Project Officer: Tom F. Lahre
16. ABSTRACT
A laboratory testing program is described wherein ethanol emissions were
measured from the various operations involved in bread production. The approach
involved monitoring, via a flame ionization detector, ethanol given off during the
mixing, rising, and baking steps in the making of bread. Both the straight-dough
and sponge-dough processes were evaluated. Emission factors are developed in
terms of quantity of ethanol emitted per quantity of bread produced. No tests were
made at an actual bakery to confirm the emission factors determined in these
laboratory tests; however, the values determined compare reasonably well with
theoretically determined values.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Bakery
Baking
Bread
Dough
Emissions
Ethanol
Ethyl Alcohol
Evaporative Emissions
Volatile Organic
Emissions
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
Unclassified
_5_L
22. PRICE
EPA Form 2220-1 (9-73)
------- |