EPA
United States
Environmental Protection
Agency
EPA-600/7-84-094
September 1984
Research and
Development
CHARACTERIZATION OF EMISSIONS
FROM THE COMBUSTION OF WOOD
AND ALTERNATIVE FUELS IN
A RESIDENTIAL WOODSTOVE
Prepared for
Office of Air Quality Planning and Standards
Prepared by
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
-------
Abstract
Overall, oak wood is the best fuel tested, considering both emissions and
stove operation. Compressed wood logs with binders and bituminous coal produce
the highest emissions of S02, particulate, and NO . Compressed wood logs
without binders and treated lumber produce the highest PNA emissions. Important
parameters affecting CO emission levels are fuel structure and, to a lesser
degree, combustion air flow. S02 emission levels are related directly to fuel
sulfur content. NO emissions are controlled by fuel nitrogen content and
combustion air flow rate. Organic emissions are affected by fuel consumption
rate, fuel structure, and amount of air through the stove. PNA formation is
affected by combustion air flow, firebox temperature, and fuel structure.
Bioassay results indicate the presence of both mutagens and promutagens in the
organic extracts from flue gas samples from both wood and coal combustion
tests.
-------
TABLE OF CONTENTS
Section Page
Abstract ii
List of Figures iv
List of Tables v
1 INTRODUCTION 1
2 FUELS 5
3 EQUIPMENT 8
4 SAMPLING PROCEDURES—ORGANICS 12
5 ANALYTICAL PROCEDURES 16
5.1 GAS ANALYSIS 16
5.2 ORGANICS ANALYSIS 16
5.2.1 Sample Preparation 16
5.2.2 Total Organics 17
5.2.3 PNA Analysis 17
5.2.3.1 Gas Chromatography 17
5.2.3.2 Spot Tests 20
6 EXPERIMENTAL 23
7 RESULTS 25
7.1 EMISSIONS COMPARISON 25
7.1.1 Pollutant Levels 25
7.
7.
7.
7.
7.
7.1.1.6 Organics 45
7.1.1.7 Polynuclear Aromatic
Hydrocarbons (PAH) 49
7.2 FUELS COMPARISON 55
7.3 BIOASSAY RESULTS 73
8 CONCLUSIONS 75
REFERENCES 78
APPENDIX -- Final Report, Ames Testing of Stove 82
Combustion Products
i i i
1.1 Particulate 25
1.2 Sulfur Dioxide (S02) 30
1.3 Nitrogen Oxides (NO ) 36
1.4 Carbon Monoxide (C07 42
1.5 Carbon Dioxide (C02) 45
-------
LIST OF FIGURES
Page
1. Stove dimensions and thermocouple placement 9
2. Emission factors: Participate 28
3. Emission factors: S02 32
4. Emission factors: NO 38
P\
5. Emission factors: CO 44
6. Emission factors: Benz(a)pyrene 53
-------
LIST OF TABLES
Number
1 Key: Fuel Nomenclature 6
2 Sample System 13
3 Chromatographic Conditions 19
4 Total PNAs 22
5 Mass Balances—Sulfur and Nitrogen 26
6 Emission Factors—Particulates 27
7 Comparison of Particulate Emissions From Commercial and
Residential Combustion Units Burning Wood and Coal . . 31
8 Emission Factors—Sulfur Dioxide (S02) 33
9 Comparison of S02 Emissions for Commercial and Resi-
dential Combustion Units Burning Wood and Coal 35
10 Emission Factors—Nitrogen Oxides (NO ) 37
11 Percent NO in NO Emissions 40
/\
12 Comparison of NO Emission Factors for Commercial and
Residential Combustion Units Burning Wood and Coal ... 41
13 Emission Factors—CO 43
14 Comparison of CO Emissions from Commercial and Resi-
dential Combustion Units Burning Wood and Coal 46
15 Emission Factors—C02 47
16 Emission Factors—Total Hydrocarbons 48
17 Gas Bulb Organic Carbon Concentration 50
18 PAH Emission Factors 51
19 Comparison of Emission Factors of PNAs 54
20 Fuel Analysis 56
21 Operating Parameters—Sampling Period Averages .... 57
22 Ash Analysis 58
23 Emission Maxima—Wood Tests 60
-------
1. INTRODUCTION
The use of woodstoves for domestic heating has become popular in recent
years. Woodstoves have been installed at the rate of one million per year
since 1977,1 and the trend is expected to continue because of the rising cost
of oil and gas and the increased awareness of the need to find an alternative
to these nonrenewable fuels. However, ambient air studies have revealed that
a significant decline in air quality in certain areas is attributable to
residential combustion of wood.2 3456 Emissions from woodstoves and other
domestic heating heating fuels have not been regulated because of the lack of
Federal regulation of combustion units that produce less than 100 million
Btu's per hour.7 Although wood will probably continue to be the dominant
woodstove fuel, increased utilization of woodstoves for home heating has
caused the price of wood to rise in many areas of the country while its avail-
»
ability has decreased. This trend may encourage a greater degree of use of
coal and other alternative fuels in woodstoves.
Significant emissions from the residential combustion of carbonaceous
fuels include particulates, volatile organics, CO, NO , and S02. Trace ele-
ments are also found in ash from residential combustion of coal and wood.8 9
There is little or no data on emissions from residential combustion of fuels
other than wood and coal.
Emissions from the combustion of solid fuels for home space heating that
are of primary concern are particulates, volatile hydrocarbons, and carbon
monoxide, which are products of incomplete combustion. Coal and wood combus-
tion emissions are generally higher from residential combustion sources than
from other combustion units (such as industrial and utility boilers) because
of the relatively low combustion efficiency of most residential units and
typical operating conditions which are conducive to incomplete combustion.8 9
In addition, industrial combustion units are often equipped with emission
control equipment, such as multiple cyclones for particulate emission reduc-
tion.10 Residential combustion of gas and oil also produces lower levels of
these emissions than the use of coal or wood, which is attributable to the
differences in the nature of the fuels8 9 and to differences in combustion
conditions.
-------
Particulates from the combustion of wood in domestic heating units con-
sist of inorganic ash, carbon chars, and condensable organics.7 Particulates
from residential coal-fired units consist of unburned coal and condensed
hydrocarbons, in contrast to the primarily inorganic particulate matter pro-
duced during the industrial combustion of coal.9 Residential combustion of
wood is considered to be a major source of ambient particulates in several
localities.2 3 4 5 6 A large portion of particulate from residential combus-
tion of wood and coal are inhalable (less than 2.5 pro in diameter).7 9 In
Missoula, Montana, ambient particulate levels have been linked to health
problems of grade school students and persons with pulmonary diseases.5
Levels of particulate emissions are highly variable and depend upon
factors such as fuel and equipment types, firing rate, and burn rate.7 1:L 12
Organic particulate emissions from combustion of wood in woodstoves have been
found to decrease with increasing burn rate, increasing log size, increasing
heat release rate, and increasing stack gas temperature.12 13 14 The quantity
of airborne particulates produced by residential coal combustion appears to be
highly dependent upon coal type.9 10 ll
A variety of condensable organic species may be produced as a result of
incomplete combustion or pyrolysis of organic materials in fuels. Polycyclic
organic matter (POM), aldehydes, aliphatic hydrocarbons, and phenolic compounds
are some of the organic species identified in emissions from residential
combustion of wood and coal.8 9 POM emissions are a major concern because
many POM compounds (polynuclear aromatics) have proven to be potent animal
carcinogens.
In 1976, POM from residential combustion of wood counted for 80 percent
of total POM from all sources, and residential combustion of coal was the
biggest contributor of POM from all coal combustion sources with the exception
of coke manufacturing. POM emissions from residential coal-fired units are
greater than emissions from the combustion of wood in woodstoves,10 15 but
woodstove POM emissions have been found to be higher than those from fire-
places.8 One reason for the relatively high levels of POM from residential
stoves is that maximum thermal efficiency usually requires that air flow
through the stove is reduced to a minimum. These oxygen-starved combustion
conditions are favorable to the formation of POM compounds.
-------
It must be emphasized that total POM emissions are not necessarily indi-
cative of carcinogenic risks, since not all POM compounds are carcinogenic.16
However, several suspect carcinogens have been identified in emissions from
residential combustion of wood and coal.9 17 1S 19
Carbon monoxide (CO) is also a byproduct of incomplete combustion. The
effects of low levels of CO on the central nervous and cardiovascular systems
are well-known.20 CO is of particular concern because large amounts of CO
enter the atmosphere from many combustion processes, and CO emissions exceed
those of all other atmospheric pollutants except C02.20 Although motor vehicles
are the major source of CO, CO emissions from woodstoves are not insignificant.5
Emissions from woodstoves are far in excess of CO emissions from all other
residential heating units,8 including fireplaces, as well as from industrial
wood or bark boilers.21
Formation of nitrogen oxides (NO ) is dependent upon fuel nitrogen content,
x\
the amount of air introduced into the stove, combustion temperature, and type
of combustion equipment.8 N02 is thought to be the most toxic of the nitrogen
oxides, producing local irritation at low levels and lung injury at higher
levels.22 NO is also of environmental concern because nitric acid precipitates
formed from atmospheric NO have had adverse effects on terrestrial and aquatic
ecosystems.23 NO is also known to react with environmentally prevalent amines
to form N-nitroso compounds, many of which are potent animal carcinogens.22
Total NO emissions from residential combustion of wood and coal are
substantially lower than total emissions from domestic heating units fueled
with oil and gas,8 reflecting the greater use of the more conventional fuels.
Levels of NO appear to increase with increasing combustion temperature, so
X
that the high levels of NO associated with combustion of oil and gas may also
J\
reflect the more complete combustion of organics.8
S02 emissions are a result of the oxidation of fuel sulfur. S02 is
primarily an irritant but may also have bronchial and pulmonary effects.24
Levels of S02 less than 1 ppm have caused damage to plant foliage, and aqueous
systems have been detrimentally affected by acid precipitates of atmospheric
S02.24
S02 emissions from residential combustion of wood are low due to the low
sulfur content of woods. Coal, however, has a much higher sulfur content than
wood. Even though only 0.66 percent of all homes were heated by coal in 1976,
-------
S02 emissions from residential coal combustion were second only to S02 emission
from residential combustion of oil.8 Thus, use of coal in domestic heating is
already a major source of S02, and increases in the use of coal for domestic
heating could have relatively major environmental consequences.
The purpose of this study is to measure the emissions from the residential
combustion of alternative fuels, including coal, in a conventional woodstove.
Fuels tested include compressed wood, treated wood, newspapers, commercially
available paper logs, and peat, in addition to untreated oak wood and bituminous
coal. The pollutants discussed above were measured during the course of this
study for the alternative fuels tested, and their emission levels are compared
to those from wood combustion. The effects of the stove operation parameters
on emission levels of these pollutants is also considered. It is hoped that
this information will be useful in estimating the overall effect of these
emissions from residential solid fuel units on ambient air quality.
-------
2. FUELS
During the planning phase of this project, eight fuels were chosen as
likely alternatives to wood for use in residential combustion units. The
fuels chosen were coal, both bituminous and anthracite, peat, newspaper logs,
cardboard logs, compressed wood chip logs, both with and without binders, and
lumber pressure treated with copper salts to retard rot. Alternative fuels
were chosen on the basis of availability and on the likelihood that they would
be used by stove owners as a wood substitute. Except for peat, all fuels
represent commercially available fuels making them likely alternatives to wood
as fuel in residential woodstoves. Table 1 is a list of the fuels screened
during this project and symbols used to represent each fuel.
The coal for the bituminous coal screening tests was obtained at a local
coal supplier. Commonly known as bag coal, this coal is used by most home-
owners with coal space heaters. Sod peat was obtained from Prulean Farms in
eastern North Carolina. This peat was harvested using the sod peat harvesting
technique, which extrudes the peat into cylinders which are three to four
inches in diameter and about 15 inches long. It leaves the peat sods on the
surface of the peat bog to dry in the sun. Although peat is not presently
used in the United States for home heating, it is potentially a great energy
resource in this country, and thus, someday may be employed as home heating
fuel.
Compressed wood logs without binders were obtained from a local stove
outlet. These logs were about 3.5 inches in diameter and about 15 inches long
and were formed by pressing together woodchips and sawdust. The wood in these
logs is over 90 percent hardwood. Compressed logs with binders were obtained
at a local grocery store. These logs consisted of cedar shavings and sawdust
held together with a paraffin binder. Copper compounds have been added to
these logs to produce colorful flames.
Cardboard logs formed of compressed corrugated cardboard chips were
obtained through the U.S. Environmental Protection Agency (EPA) from a manu-
facturer in Oregon. Newspaper logs were made in-house using a commercially
available newspaper log roller. Anthracite coal was obtained from a local
stove supplier. Treated lumber was obtained at a local lumber yard. Treated
lumber is yellow pine pressure treated using the Wolman process.
-------
TABLE 1. KEY: FUEL NOMENCLATURE
W = Oak logs
CW = Compressed wood logs
CWB = Compressed wood logs with binder
C = Cardboard logs
N = Newspaper logs
TW = Treated lumber (pine)
P = Peat
BC = Bituminous coal
AC = Anthracite coal
-------
Two successful runs were performed for each fuel except for anthracite
coal, which was not successfully burned in the stove chosen for this study.
Two tests were also carried out using split and round dry oak which was obtained
from a local firewood supplier. These tests were used as a baseline for
comparison to tests with other fuels. Fuel analyses and information on stove
operation characteristics for each fuel may be found in the discussion section
of this report.
-------
3. EQUIPMENT
The woodstove used in this study was a free standing air-jacketed design
with simple open firebox. This stove represents a type of stove which is
increasing in popularity. Originally designed as a fireplace insert, this
type of stove is being installed in increasing numbers of new homes. Firebox
dimensions are given in Figure 1. The stove flue was 8-inch, double wall,
fiberglass insulated stainless steel stove pipe. The stove pipe stood unsup-
ported, the entire stove assembly rested on a digital balance, which measured
weight change (for fuel consumption) during each test.
An optional shaker coal grate was purchased with the stove along with a
firebrick lining. This was necessary to burn coal, one of the alternative
fuels tested. The coal grate was also used to burn peat, compressed wood
logs, and cardboard logs. A second coal grate was obtained from another
manufacturer because the anthracite coal did not burn well on the first grate.
The second grate performed better than the first with anthracite, but it was
still not possible to successfully burn anthracite coal. The stove design is
such that air flow through the coal grate is not restricted enough to insure a
high velocity of air through the grate, a condition which is necessary in
order to successfully burn anthracite coal.
The stove firebox is enclosed on five sides by an air jacket. During
stove operation air is blown through the jacket, where it is heated, and out
into the room. Three blower speeds are automatically selected by a monitor of
temperature in the air jacket. It was found that the blowers tended to operate
at the medium speed during steady-state operation. An air jacketed stove was
chosen because by monitoring temperature at the blower outlet it was possible
to estimate the stove's heat output. This calculation was performed by the
computer and printed out at 30 second intervals. By monitoring the heat
output value test personnel could tell when conditions of steady-state stove
operations were reached and could maintain these conditions throughout the
tests by adjusting drafts and damper.
Air flow through the stove was controlled with sliding door drafts in the
front of the stove at the bottom of the double doors and by an adjustable
damper mounted at the entrance to the stove flue. The damper was closed as
much as possible once steady-state conditions were reached, and most adjustments
8
-------
Firebox—50.5 cm (back W) x 66.5 cm (front W)
x 55 cm (H) x 39 cm (D)
Thermocouples
R1 - R4
Blower outlets
Blower inlet
Firebox
Stack
I
I
I/
I/
I/ X
I/ X
I/ R4
I/ X
I/ X
I/
I/
I
I
III
IT T — -
I
I
t
I
i
I
I
:c
cue
i
i
•
i
•
i
i
i
33 R2
R7
333 R:
R9
(AMBIENT)
I
I
/I
/I
X /I
X R5
Rl /I
X /I
X /I-
/I
/I
I
_I
5 III
IT T
Figure 1. Stove dimensions and thermocouple placement.
-------
to maintain these conditions were made using the sliding door drafts. The
stove was not air tight, so that even with drafts and damper closed, there was
still enough air flow through the stove for combustion to continue. Tempera-
ture was measured by several thermocouples which were placed in the firebox,
stack, at the blower inputs and outputs, and in the test room (ambient).
Thermocouple placement and numbering scheme are shown in Figure 1.
Temperature data and certain gas data were automatically recorded by the
online DEC POP-1100 computer. The computer took and retained stove operation
parameter measurements at 30 second intervals. In addition, test personnel
could log in comments on a computer terminal in the laboratory. The stove
balance readings, stack flow turbine counts, and stove draft and damper settings
were entered in the comments along with pertinent information about the test
operation.
The start and stop of the Method 5 sampling train were also entered into
the computer. This enabled the computer to calculate sampling period averages
for the various parameters automatically recorded by the computer. At the end
of each test run, the computer printed out a test run summary including these
averages, sampling time, fuel type, and comments entered into the computer
during the test.
Accurate measurement of flue gas velocity in wood stove flues was found
to be quite problematical. Flue gas velocity was far too low to get any gas
velocity measurements using conventional s-shape pitot tubes and water manom-
eters. In another wood stove emission study funded by the U.S. EPA, Monsanto
researchers utilized a micromanometer capable of being read down to 0.0002
millimeters of mercury.25 An equivalent instrument was purchased, but it was
found that it was also unsuitable for the gas velocity measurements. Even at
maximum stack gas flows the micromanometer readings were consistently close to
0. It was felt that using such readings for the flue gas velocity measurement
would be pushing the limits of the instrument.
In order to solve the problem of flue gas velocity measurements, RTI
designed a turbine meter for measuring flue gas flow rates. This instrument
consisted of a metal turbine mounted on a spindle in the stack above the
sampling probe. The turbine rotated freely as flue gas passed through it.
Two small magnets were mounted on the turbine hub; these magnets passed over a
coil mounted on the spindle base, producing signals which were counted by
10
-------
frequency counter. These "counts" were recorded in the comments during each
run. The flow meter was calibrated by adding known)amounts of C02 to the flue
gas. By measuring the amount of dilution of the C02 added to the gas, it was
possible to accurately estimate stack flow rate. The flow turbine was cali-
brated by plotting flow rates against frequency counts. The relationship
between stack flow and turbine frequency counts was found to be best described
by an exponential function.
Fouling of turbine by condensed organics or particulate matter during the
tests was not found to be a problem except during runs with bituminous coal
and compressed wood logs with paraffin binders. During these tests, it was
necessary to remove and clean the stack flow turbine periodically. The stack
flow turbine enabled us to get continuous flow measures of the stack during
each test, which were used to flow average all gas concentrations.
11
-------
4. SAMPLING PROCEDURES--ORGANICS
A whole test Integrated sample of polycyclie organic matter and other
organic emissions was collected by a modified Method 5 sampling train similar
to the one described in the Battelle protocol.26 The RTI sampling train was
an all-glass/teflon construction which consisted of the following: a glass
probe, a heated filter box (>204° C) containing a fiberglass filter, a series
of gas condensers, a water-cooled XAD-2 resin chamber, a series of ice-cooled
impingers, a drying cartridge filled with silica gel, and a suitable air
moving and air monitoring device. The basic method and equipment were selected
in accordance with the protocol; however, the physical arrangement of the site
and problems which arose during the testing necessitated some departure from
the basic protocol. The main departure was in the use of a series of two
condensers and a condensate trap located between the filter box and resin
chamber. This section was configured in the following manner. A 90 degree
ground glass joint mounted with the plane of the bend parallel to the roof
brought the gas flow out of the filter box and into an All inn condenser (300
mm). The gas then flowed through a 180 degree ground glass joint mounted with
its bend perpendicular to the roof. The bottom of this glass joint was tapped
to provide a connection for a 250 ml amber condensate flask. The gas was
further cooled by being passed through a Graham condenser (250 mm) prior to
entry into the XAD resin chamber. Any condensate was trapped ahead of the
XAD. This reduced the possibility of flushing of the XAD with large quantities
of water and subsequent channelling in the XAD bed. Other changes of the
protocol included the use of an all-glass probe and the elimination of the
particulate filter between the silica gel impinger and the dry impinger of the
system. This filter was eliminated because of anticipated low flow and low
particulate loading which were thought to be trapable in the front part of the
system. The impinger system consisted of four impingers mounted in series.
The first two impingers were filled with 100 ml of water. The second impinger
was a modified Nuburg. The third impinger was dry, and the fourth contained
300 to 400 g of silica gel.
Altogether the system consisted of five subsections (See Table 2). These
subsections were acid cleaned and washed with methylene chloride prior to
their assembly in the laboratory. The joints were taped with teflon and
clamped to facilitate transport.
12
-------
TABLE 2. SAMPLE SYSTEM
1. Section 1. Probe Section
A. All glass probe
B. Flexible teflon connector
C. Ground glass joint (female)
D. Electrical heating tape, thermocouple, controller
2. Section 2. Particulate Filter
A. Ground glass joint (male)
B. Glass filter holder (2 halves jointed with Bezel ring)
C. Sintered glass filter support with sealing gasket
D. Glass fiber filter
E. Heated box, thermocouple, and remote controller
3. Section 3. Condenser Section
A. 1-300 mm Allihn condenser
B. 1-250 mm Liebig condenser
C. 2-90° ground glass T/S joint
D. 1-250 mL infrared shielded flask
E. 1-180 degree ground glass joint with condensate tap
F. Coolant system
4. Section-4. XAD Resin Column
A. Jacket resin column with sintered glass plug (interior 2" wide,
50 g capacity)
B. XAD resin
C. Spring loaded perforated end plate
5. Section 5. Impinger
A. 4 impingers
1st, 100 mL water
2nd, modified Nuburg with 100 mL water
3rd, empty
4th, silica gel (200-300 g)
B. Water bath cooled with ice water and portable condenser
D. Gas moving and monitoring attachment
13
-------
The train was transported to the roof in sections and assembled. A
standard 8-inch insulated flue was used to exhaust gas from the stove to the
ambient air. The glass probe was inserted in the flue in accordance with the
standard Method 5 protocol leaving 8-duct diameters ahead of the probe free of
obstructions and 2 behind it before exiting to the air. The stack was cleaned
prior to each test to prevent the volatilization of deposited residue which
might bias the sample. Assembly and checkout were conducted according to a
test protocol developed to meet RTI's situation as well as to incorporate the
Battelle protocol. The filter box was heated to 205° C, the resin trap was
cooled to less than 20° C, and the system was leak checked per standard proce-
dures. The taping of the joints facilitated the leak testing and reduced the
time that would have been spent in chasing down leaks. The flow meter was
wired in and all the necessary cooling lines were connected to the condensers
and the XAD column. The cooling water was provided by a portable condenser.
The system was wrapped with insulating materials to protect it from sunlight
and to maintain coolness.
Operation of the sampling equipment consisted primarily of maintaining
the flow of the coolant, monitoring all temperatures, gas flow rates, and
making the necessary adjustments. Backup filters and glassware were cleaned
and at hand and could be rapidly installed with a minimum of downtime.
Mass emissions were collected over a 45 to 120 minute interval depending
on the volume of sample required for analysis. At the conclusion of the
sampling run, the train was shut down, disassembled, and transported to the
laboratory for cleanup and sample recovery. Samples taken were as follows:
Probe wash, filter particulate, XAD resin, XAD module and condenser wash,
condensate, and impinger water. All samples were stored in amber glass bottles
under refrigeration prior to analyses.
Generally, the samples were recovered immediately after the run. However,
if this was not possible, the sampling train subsections were sealed and
stored in a refrigerator for no more than 24 hours. The solvent system found
to be most effective for sample recovery from the front half of the system
(probe and condenser) was a 9:1 methylene chloride/methanol solution. If
necessary, a stiff bristle, chemically inert nylon brush was used in the probe
to remove any adhered particulate. Methylene chloride alone was used to rinse
the XAD column and in the impinger system. The condensate was measured and
14
-------
collected. Contents of the water impingers were weighed at the end of the
test to determine the quantity of condensed water and were bottled for analysis
The silica gel impinger was weighed and the material discarded. All samples
were labeled and refrigerated except for the filter, which was desicated for
24 hours prior to weighing for filter particulate weight.
15
-------
5. ANALYTICAL PROCEDURES
5.1 GAS ANALYSIS
Stack gas composition was continuously monitored during the tests. A
separate 1/2 inch stainless steel probe was used to sample the flue gas for
gas composition analysis. The probe was packed with glass wool, and a water
cooled condenser was placed in the sample line to prevent fouling of the
instruments. Carbon monoxide, carbon dioxide, and methane were analyzed using
infrared detectors manufactured by Horiba Instruments, Incorporated. These
data were recorded by the computer. Nitrogen oxides (NO ) were measured using
/\
a photoluminescent detector manufactured by Thermo Electron Corporation.
Sulfur dioxide (S02) was measured using a photometric detector manufactured by
DuPont Instruments. Both NO and S02 concentrations were continuously recorded
s\
by a conventional stripchart recorder.
In addition to continuous gas analysis, gas bulb samples were also taken
and analyzed by gas chromatography for total organic carbon. Two gas bulb
samples were usually taken for each test. Orsat method measurements of oxygen
were also made at intervals during the later tests using a Fyrite oxygen
analyzer. Both Fyrite and gas bulb samples were taken through the same sample
line as the continuous gas analysis sample. All continuous gas analyzers were
calibrated per EPA standard procedure using gas mixtures of known concentra-
tions. The instruments were calibrated prior to each test.
5.2 ORGANICS ANALYSIS
5.2.1 Sample Preparation
Modified Method 5 samples which were analyzed for organics include:
probe wash (CH2C12 + CH2OH), filter, condenser and XAD module wash (CH2C12),
XAD-2 adsorbent, condensate catch, and impinger water. The probe wash was
filtered to remove particulates captured by the probe during sampling. These
particulates were weighed, and their weight was added to the weight of particu-
late caught on the Method 5 filter to give total particulate sampled during
the test. The sampling train filter was then placed in a glass soxhlet thimble
along with the probe wash filters and the XAD-2 adsorbent. The filtered probe
wash and the condenser/XAD-2 module wash were poured into the soxhlet flask
along with enough fresh CH2C12 to bring the volume to 500 to 700 cc. The
16
-------
soxhlet apparatus was then covered with foil and allowed to reflux for 24
hours. Following extraction, the filtrate was placed into a Kuderna-Danish
evaporator and concentrated to 10 to 25 cc. The resulting concentrate was
stored at 4° C until analysis.
The aqueous portion in the sampling train, impingers, and condensates
were added together and extracted with a separatory funnel at pH of 2.0 ± 0.5
and 12.0 ± 0.5 as described elsewhere.26 Two extractions were done at each
pH. The CH2C12 from each set of extractions was added together and concen-
trated to ~10 to 40 ml_. The resulting sample was stored at 4° C until analysis.
5.2.2 Total Organics
Organic analysis were performed separately on the two types of concentrated
samples described in Section 5.2.1. Total organics with a boiling point of
100 to 300° C were determined by total chromatographable organics (TCO) and
gravimetric analyses described elsewhere.26 Chromatograms of a hydrocarbon
standard that were obtained on the same day as sample chromatograms gave
retention times corresponding to the boiling point range 100 to 300° C.
Calculation of TCO in each sample was based on the integrated peak area between
these retention times and the area and the concentration of an internal stand-
ard (triphenyl ethylene).
Organics with a boiling point above 300° C were determined by weighing
0.5 mL samples that were evaporated to dryness. A complete description of
this technique can be found elsewhere.26 Results from TCO and gravimetric
analyses were totaled to give total organics for the sampling period.
5.2.3 PNA Analysis
5.2.3.1 Gas Chromatography--
RTI has amassed considerable experience in the analysis of PAH in various
process streams, including the aqueous condensate and tar effluents from coal
gasification.27 28 29 During these studies, RTI encountered difficulty in
characterizing these highly complex mixtures by gas chromatography-mass spectro-
scopy (GC/MS) in a time- and cost-effective manner. Time consuming sample
fractionation procedures were necessary for good GC/MS quantitation of PNAs in
these samples. Fractionation of the PNAs by these methods was not complete,4
and the complexity of the fractionation techniques gave rise to some sample
17
-------
loss. These considerations motivated the development of a direct technique
for the analysis of PNAs in complex mixtures utilizing glass capillary gas
chromatography (GC2). This technique has been presented in detail in two
papers.2 3
Gas chromatographs of the sample extracts indicated that the samples
obtained form the residential combustion unit were indeed very complex. This
complexity, in conjunction with project time and cost constraints, led to the
selection of GC2 as the analytical method for determining PNA concentrations
in the modified Method 5 sample extracts. A Varian 3700 GC2 system with a
flame ionization detector (FID) was used for these analyses. The system was
all-glass from the injector to the detector. A wall coated OV-101 capillary
column was used in the system. All samplewetted parts were made of glass.
Helium was used as the carrier gas as well as the makeup gas. Chromatographic
conditions are listed in Table 3. Samples were injected using the Grob "split-
less" technique,30 and sample volumes ranging from 3 to 4 uL were used. The
splitless technique consists of injecting the sample and then 30 seconds
later, opening the splitter to remove the excess solvent. This prevents a
long solvent tail. The advantages of using this technique for polycyclic
materials are well documented.31
Prior to GC2 injection, concentrated sample extracts were internally
standardized. A problem was to find a suitable internal standard for GC
analysis since the extracts were substantially complex. Triphenyl ethylene
(TPE) was found to be most suitable of the many internal standards tested
since it was present at negligible concentrations in the extracts, was similar
in nature to other aromatic compounds, and was well separated from the peaks
of the compounds of interest.
PNA analyses were performed separately on the two samples described in
Section 5.2.1. Aliquots were internally standardized with triphenyl ethylene.
Chromatographic peaks from GC2 analysis of these aliquots were compared to
peaks in duplicate aliquots spiked with standard solution of 26 PNAs and TPE
in order to identify the PNAs in question. Once PNA peaks were identified,
compound quantisation was accomplished using the known concentration of TPE in
the samples spiked with TPE only and response factors (RF) calculated from a
standard composed of 25 PAHs and TPE at known concentrations. The standard
was periodically run on the GC2 to give response factors for each PAH. Response
18
-------
TABLE 3. CHROMATOGRAPHIC CONDITIONS
Instruments
Detector
Column
Carrier
Temperature
Injection
Detector gas flows
Makeup gas
Varian 3700 with "all glass" capillary systems
FID 28° C
25 m x 0.25 mm I.D. WCOT 0V 101, Borosilicate
capillary
Helium, 16 psi
50° C for 1 min, 4° C/min to 265° C, hold 20 minutes
Splitless, 1-minute delay
Optimum
Helium (~29 mL/min through detector)
19
-------
factors account for decreasing area counts for equally concentrated PAHs as
the GC2 approaches its limits for eluting high boiling point compounds and
were determined from calibration curves. In addition, five samples were
analyzed for PAH compounds by GC2/MS for verification of the GC2 results.
Overall, these results compared favorably.
5.2.3.2 Spot Tests—
The PNA sensitized fluorescence spot test32 was utilized to screen the
XAD extract, condenser and probe wash, and in the methyl en chloride extract
from the aqueous impingers and condensate samples for the presence of PNAs.
Both original and concentrated extracts were tested.
A micro!Her amount of sample was applied to a Whatman filter, and in
another spot a micro!Her of naphthalene (30 mg/mL) was applied. Between
these two spots, a microliter of naphthalene was superimposed on a micro!Her
of the sample. Immediately following application of the spots, the filter
©
paper was exposed to ultraviolet radiation in a Chromato-Vue Model CC-20
ultraviolet cabinet. Differences in intensity between the sensitized spot and
the sample spot relative to naphthalene were noted and recorded as one of four
categories: . strong (sample spot self-fluorescent), moderate (sensitized spot
considerably more intense than naphthalene), weak (sensitized spot slightly
more intense than naphthalene), and none (fluoroescence of sensitized spot
equal in intensity to fluorescence of naphthalene).
In most cases, the qualitative spot test results compared well with the
total PNAs determined by GC2. Most of the aqueous (impingers and condensate)
samples exhibited weak sensitized fluorescence, which is consistent with the
low water solubility of PNAs. GC2 analysis indicated that only the low molecu-
lar weight PNAs were present in these samples. The majority of PNAs detected
by GC2 were in the filter extract, XAD extract, condenser wash and probe wash
samples, and these samples were also strongly or moderately fluorescent.
Several of the extracts from the aqueous samples did not exhibit sensitized
fluorescence, although PNAs were identified by GC analysis. Since the spot
test is 10-100 times more sensitive to PNAs than gas chromatography, the lack
of fluorescence may be due to interfering species. Two of the samples, in
fact, inhibited fluorescence of naphthalene, which is a clear indication of
the presence of species that are capable of absorbing fluorescence. Phthalates
20
-------
have been found to negatively interfere in spot tests,33 and this type of
compound has been identified in the gaseous effluent from the combustion of
wood in a woodstove.25 However, in a study of several different types of
industrial combustion effluents, in no cases were PNAs detected by GC/MS in
samples that did not also give positive spot test results.34
Only one sample exhibited weak fluorescence but was not found to contain
PNAs by gas chromatography. This is probably due to the sensitivity of the
spot test to low levels of PNAs. A positive interference is possible, but
standard samples containing a wide variety of organics, many of which would be
expected to be present in these combustion samples, have not exhibited sensitized
fluorescence in the PNA spot test.32 35
Detection limits of individual PNAs vary with experimental conditions.
For example, detection limits for B(a)P have been found to range from 1-100
pg/uL.32 34 This variation has been attributed to differences in purity of
the stock materials, intensity of the ultraviolet lamp, and the subjective
judgment involved in determining detection limits.33 34
In this study, detection limits of B(a)P and phenanthrene were found to
be 10 pg/uL and 10,000 pg/|jL, respectively. The detection limits of most
other PNAs are expected to be the same as B(a)P or phenanthrene or within this
range.33 34
Total PNAs in two of the samples were determined by the method of multiple
dilutions,32 assuming an average detection limit of 100 pg/uL. Results are in
Table 4. Although there is a good deal of uncertainty with this detection
limit, it is likely that GC2 and spot test results agree within an order of
magnitude.33
21
-------
TABLE 4. TOTAL PNAs
Test number Total PNAsa Emission factor
CW2 41 mg 129.14 mg/kg
C2 22 mg 265.39 mg/kg
of sample) = 100 pg/uL x 10(n~i;)
where n = number of 1:10 dilutions required before no sensitized fluores-
cence.
22
-------
6. EXPERIMENTAL
To properly compare the emissions of the alternative fuels tested, it was
necessary to operate the woodstove as similarly as possible from test to test.
Sampling only at steady-state stove operating conditions is the best way to
assure reproducibility. Using considerable in-house experience in residential
woodstove operation, steady-state conditions were defined to approximate
conditions a typical stove owner would achieve for most of the stove's opera-
tion. The sampling period was confined to this steady-state period. However,
it should be stressed that woodstove operation can vary greatly from owner to
owner and that it is very difficult to determine an actual average steady-
state operating conditions. In addition, startup and shutdown conditions can
be very significant from an emissions standpoint when overall woodstove emission
factors are considered. The major purpose of this study is to compare alterna-
tive woodstove fuels on an equal basis, and steady-state conditions were
chosen for sampling periods to assure as much reproducibility as possible from
test to test. Startup and shutdown conditions were judged to be too variable
for reproducible testing. However, continuous gas monitoring was carried out
for startup period and some of the shutdown period following sampling. It
should be stressed that by sampling only at steady-state lower total emissions
can be expected than if the stove was sampled from a cold start.
At the start of the test, the stove was loaded with paper and kindling
(if necessary) and the fuel to be burned. Except for the compressed wood logs
with binders (CWB) and the coal runs, enough fuel was added to almost fill the
firebox. Two CWBs were used for each CWB test. The coal was slowly added to
an existing wood fire per stove manufacturer's instructions. The fire was
started with drafts and dampers open. Once the stack temperature started to
rise significantly (at about 275° C), the damper was closed. The stove was
allowed to reach normal operating conditions before sampling was begun.
Normal steady-state conditions were with blower speeds at medium and heat
output values above 50,000 Btu's/min. Once these conditions were reached,
sampling was started. Sample volume varied from 0.5 to 1.7 m3 depending on
the duration of the sampling period. Initially sampling periods were rela-
tively short to avoid possible overloading the XAD adsorbent module with
organics. However, it was found that longer sampling periods did not result
in XAD overloading, so later tests were longer.
23
-------
Draft and damper settings and fuel addition rate were adjusted for each
fuel to maintain steady-state operating conditions for the entire sampling
period. Descriptions of these parameters for each fuel, along with problems
encountered in maintaining steady-state conditions during each test are dis-
cussed in the results.
Low flue gas velocities made isokinetic sampling unfeasible because to
achieve isokinetic sampling would require extremely long sampling periods or a
very large sampling probe. In addition, flow rates were variable enough
during a test that achieving isokinetic sampling would be difficult if not
impossible. A steady sampling rate of 0.5 cfm was used during each test. The
particle size distribution from woodstove combustion was fine enough that the
deviation from isokinetic sampling probably did not bias results. Personal
communication and literature from other researchers in the area also implied
that isokinetic conditions were not achieved during their programs. Difficulty
in accurate flow measurement was the biggest reason for this problem. Although
we had more accurate flow measurement than most, isokinetic sampling was still
judged to be too difficult a goal to achieve. It should be noted that a
steady sampling rate does weight sampling results with a bias towards periods
of low flow rates. However, for most tests flow rates were consistent enough
during the steady-state period to keep this bias minimal.
24
-------
7. RESULTS
This section is presented in two parts. First is a discussion by pollutant
describing how the fuels compared in emissions factors for each pollutant.
The second part is a description of fuel burning characteristics and general
emission levels of each fuel screened. Factors affecting pollutant emission
levels are also discussed for each fuel.
7.1 EMISSION COMPARISON
Comparisons of the emission factors of the fuels tested are given in a
table and a figure for each pollutant. Emission rates in grams per hour and
emission factors in grams per kilogram of fuel consumed are graphed in each
figure. The same emission results are presented in table form with the addition
of the average concentration of the pollutant in the flue gas for each test.
In the figures, the emissions from duplicate tests were averaged for each
fuel. Sulfur and nitrogen balances are given in Table 5. This table shows
that most of the sulfur and nitrogen was volatilized during combustion except
for fuels with little or no sulfur or nitrogen content.
7.1.1 Pollutant Levels
7.1.1.1 Particulate--
Wood stoves are significant emitters of particulate matter. In some
areas up to 73 percent of total suspended particulates has been attributable
to residential wood combustion at certain times.36 Also residential coal
burning emits more particulate than residential wood burning.
Particulate emission results for the eight fuels successfully tested are
given in Table 6 and Figure 2. Examination of these tables shows that the
fuels may be ranked by particulate emissions as follows (highest to lowest):
1. Compressed wood logs with binders
2. Bituminous coal
3. Newspaper logs
4. Treated lumber
5. Peat
5. Compresssed wood logs
25
-------
TABLE 5. MASS BALANCES—SULFUR AND NITROGEN
Sulfur Nitrogen
Percent volatile Percent volatile
Wl & 2 89.44 95.67
CW1 99.15 97.15
CW2 97.83 97.55
CWB1 89.40 99.40
CWB2 89.62 98.61
Cl 81.40 98.37
C2 63.04 95.89
Nl 88.34 <0
N2 86.48 <0
TW1 -31.40 87.82
TW2 -5.61 89.58
PI 82.60 85.18
P2 86.36 93.03
BC1 93.34 90.15
BC2 - 93.02 90.92
26
-------
TABLE 6. EMISSION FACTORS—PARTICULATES
rsj
Fuel
Oak logs
Compressed wood logs
Compressed wood logs with binder
Cardboard logs
Newspaper logs
Treated lumber
Peat
Bituminous coal
Test number
Wl
W2
CW1
CW2
CWB1
CWB2
Cl
C2
Nl
N2
TW1
TW2
PI
P2
BC1
BC2
g/m3
0.045
0.052
0.182
0.057
0.481
0.296
0.109
0.071
0.227
0.058
0.154
0.529
0.107
0.156
0.343
0.310
g/hr
2.37
3.16
5.14
5.50
51.18
32.00
3.77
2.69
17.68
5.35
3.55
10.01
5.25
5.35
19.10
13.43
g/kg
0.52
0.79
0.93
0.58
16.27
18.20
0.63
0.72
5.53
1.52
0.63
2.26
1.17
1.27
9.98
5.21
-------
ro
CO
15-
g/kg
Emission Factors
PARTICULATE
5-
Figure 2. Emission factors: participate.
-------
7. Cardboard logs
8. Wood.
Wood and cardboard logs were lowest in particulate emissions and had
similar particulate emission factors, although cardboard logs tended to have
higher concentrations of particulate in the flue gas due to lower flow rates
than the wood tests. Peat and compressed wood logs also had similar though
slightly higher particulate levels, probably due to the fact that these fuels
are basically compressed particulate matter and tended to disintegrate upon
burning. Treated lumber was slightly higher still in particulate emissions,
due to the increased amount of condensable organics in the flue gas. The
treated lumber is yellow pine, which has significantly higher resin content
than the aforementioned fuels. In addition, flow rates through the stove were
very low during the treated lumber tests, which probably increased organic
emissions. Newspaper logs were the next highest in particulate emissions. As
the newspaper logs burned, burnt pieces of newspaper spalled off the rolls,
contributing to the higher particulate levels. Organic emissions were very
high during the newspaper tests, contributing significantly to the total
particulate. Bituminous coal particulate emissions were significantly higher
than the previously mentioned fuels. Most of this particulate matter was the
infamous coal soot, a sticky mixture of condensed organics and carbon char.
Coal particulate emitted per unit mass of fuel burned (g/kg) was over 5 times
higher than for wood. However, particulate emissions rates (g/hr) were over
an order of magnitude higher during coal tests than during wood tests. This
discrepancy in difference is because of the higher heating value of coal which
results in a lower fuel consumption rate.
The pressure drop across the orifice pressure meter in the Method 5
sampling station box was noted to decrease more rapidly immediately after coal
addition than at other times during the run. This implies that the filter was
collecting more particulate matter during these periods and that much of the
particulate catch from the coal runs represents organic material volatilized
during the pyrolysis of freshly added coal.
Compressed wood logs with paraffin binders (CWB) produced the highest
particulate emissions of any fuels tested. CWB emission factors were over
twice as high as coal and over 20 times higher than those for wood. The
particulate catch for the CWB was largely organics, probably representing
paraffin volatilization products.
29
-------
Wood, compressed wood, cardboard logs and newspaper logs, and treated
lumber, produced similar particulates in appearance, leaving a dark brown
filter cake. Treated lumber particulate appeared to contain more organics and
had a greenish tinge from copper compounds used in the lumber treatment.
Fragments of burned newspapers could be seen in the particulate catch from the
newspaper log tests. Peat particulate was interesting in that small sand!ike
particles could be seen in the filter catch. Bituminous coal and CWB produced
a sooty, sticky particulate. The smoke during tests with these two fuels and
tests with peat was brown to yellow, in comparison with the white to gray
smoke from the other fuels. CWB particulate was especially gummy and waxy,
because of the volatilization of the paraffin binders.
Table 7 is a comparison of particulate emission factors from commercial
combustion units with those from residential combustion units burning wood and
coal. For coal, residential stove particulate emissions are on the same order
as stoker boilers, but lower than coal furnaces. It should be emphasized,
however, that residential stove particulate differs qualitatively than particu-
late from commercial units. Commercial units emit predominantly inorganic
particulate whereas residential units emit particulate composed largely of
condensed hydrocarbons and unburned carbon char. Residential particulate,
because of a small particle size range and trace carcinogenic hydrocarbons
adsorbed on the particles, probably represents the greater health hazard.
For wood, residential stove particulate emissions are about an order of
magnitude less than emissions from commercial units. This is probably due to
the higher flue gas velocities in commercial units, which results in a larger
amount of entrained particulate in the flue gas stream.
However, it should be noted that wood particulate emission factors have
ranged to over 20 g/kg in other wood combustion studies11 12 13 with lower
burn rates and higher organic emissions. The high particulate emission rates
in these studies are largely attributable to increased total organics in the
flue gas, rather than to high flue gas velocities as with commercial units.
7.1.1.2 Sulfur Dioxide (S02)—
Figure 3 and Table 8 gives S02 emission factors for the fuels tested.
S02 emission factors vary directly with fuel sulfur content. Although S02
emissions are low for wood due to its low sulfur content, S02 emissions from
30
-------
TABLE 7. COMPARISON OF PARTICULATE EMISSIONS FROM COMMERCIAL
AND RESIDENTIAL COMBUSTION UNITS BURNING WOOD AND COAL
Particulate emission factor
(g/kg fuel consumed)
Coal
Stoker boilers37 1-6.5
Furnaces38 10-22
Residential stoves 7.603
Wood
Stoker boiler37 2.3-6.8
Residential stoves 0.66
HThis study, average of two tests.
31
-------
8
g/kg
4-
CO
ro
Emission Factors
SO2
g/kg fuel consumed 17.30
g/hr
W CW* CWB
* Factors based on single test
BC
Figures. Emission factors: S0?.
-------
TABLE 8. EMISSION FACTORS—SULFUR DIOXIDE (S02)
to
CO
Fuel
Oak logs
Compressed wood logs
Compressed wood logs with binder
Cardboard logs
Newspaper logs
Treated lumber
Peat,
Bituminous coal
Test
number
Wl
W2
CW1
CW2
CWB1
CWB2
Cl
C2
Nl
N2
TW1
TW2
PI
P2
BC1
BC2
Avg. concentration
(jg/m3 (ppm)
4.4E4
4.3E2
2.9E4
1.8E5
1.7E5
1.3E5
3.4E4
<2.9E4
1.5E3
<2.9E4
1.1 E5
2.2E5
1.1 E5
5.4E5
1.1 E6
(15.52)
(15.11)
(10)
(64.20)
(60.65)
(44.70)
(12.04)
(<10)
(0.54)
(<10)
(39.98)
(77.46)
(39.91)
(190.20)
(389.61)
g/hr
2.32
2.63
0.80
19.53
18.86
4.42
1.31
0.14
<0.66
2.16
10.86
3.91
30.30
48.33
g/kg
0.51
0.66
0.08
6.21
10.72
0.74
0.35
0.04
<0.12
0.25
2.42
0.92
15.83
18.76
-------
combustion of coal and other fuels with high sulfur content can be significant
and of concern because sulfur emission control devices for residential combus-
tion units are not economical or available.
The eight fuels ranked as follows with regard to sulfur emissions (highest
to lowest):
1. Bituminous coal
2. Compressed wood logs with binders (CWB)
3. Peat
4. Cardboard logs
5. Wood
6. Compressed wood logs without binders (CW)
7. Treated lumber
8. Newspaper logs
S02 emissions from treated lumber, newspaper logs, and CW were negligible.
S02 emissions from wood and cardboard logs were higher but still quite low.
Peat S02 emission were about twice as high as those from wood, reflecting a
similar difference in sulfur content. Bituminous coal S02 emissions were the
highest by far, with emission factors 20 to 30 times higher than those of
wood, again reflecting a like difference in sulfur content. CWB had the
second highest S02 emission factors, about an order of magnitude higher than
those of wood. The reason for this high S02 emission is not immediately
apparent from the CWB fuel sulfur analyses (0.25 percent). The sulfur content
of CWB is lower than that of CW or cardboard logs, both of which had low S02
emissions. One reason for this discrepancy could be that standard coal analyses
were performed to get sulfur content, and because these fuels are not coal,
this could result in spurious fuel analysis results. Other reasons are not
immediately apparent.
S02 concentrations generally rose slowly to a maximum as the test pro-
gressed, and then declined. Subsequent fuel additions caused additional
maxima, again with S02 concentrations slowly rising and slowly falling off.
Table 9 is a list of S02 emission factors for commercial combustion
units, compared with emission factors for wood and coal from this study.
Examination of this table shows that the emission factors for S02 are similar
for commercial and residential combustion units.
34
-------
TABLE 9. COMPARISON OF S02 EMISSIONS FOR COMMERCIAL AND
RESIDENTIAL COMBUSTION UNITS BURNING WOOD AND COAL
S02 emission factor Percent S
(g/kg fuel consumed) in fuel
Coal
Stoker boiler37
Furnace38
Residential stove
19
6.3-15
17. 3a
1.92
0.58-1.5
1.87
Wood
Stoker boiler37
Residential stove
0.7
0.58C
NO
0.09
This study, average of two tests.
ND = No Data.
35
-------
7.1.1.3 Nitrogen Oxides CNO)—
/\
Nitrogen oxides (NO ) emission factors are presented in Table 10 and
Figure 4. Two rankings of fuel by NOX emissions are possible. First, con-
sidering NO emission rates (g/hr), the fuels may be ranked as follows (highest
A
to lowest):
1. Peat
2. Compressed wood logs with binder
3. Bituminous coal
4. Wood
5. Compressed wood logs (no binder)
6. Cardboard logs
7. Newspaper logs
8. Treated lumber
Considering NO emission factors (g/kg fuel consumed), the fuels may be
/\
ranked as follows (highest to lowest):
1. Compressed wood logs with binders
2. Bituminous coal
3. Peat
4. Wood
5. Newspaper logs
6. Cardboard logs
7a. Treated wood (same level as 7b)
7b. Compressed wood logs (no binders)
The difference in ranking between g/hr and g/kg emission factors is due
to difference fuel consumption rates. Higher heating value fuels (coal and
CWB) have low fuel consumption rates because less fuel has to be burned to
produce a unit heat output. Hence, peat is the highest ranked NO emitter on
/\
a g/hr basis, but is replaced by coal and CWB on a g/kg basis because more
peat is burned per unit time than coal or CWB. Similar reasoning can be used
to explain changes in rank between g/hr and g/kg emission factors for the
other fuels.
Two factors were found to be important in determining NO emission magni-
/\
tude: fuel nitrogen content and stack gas flow rate. Two of the three highest
N0x emitters, peat and coal, had the highest fuel nitrogen content, 0.93
percent and 1.54 percent, respectively. This was judged to be the most important
36
-------
TABLE 10.
EMISSION FACTORS—NITROGEN OXIDES (NO )
/\
u>
Fuel
Oak logs
Compressed wood logs
Compressed wood logs with binder
Cardboard logs
Newspaper logs
Treated lumber
Peat
Bituminous coal
Test
number
Wl
W2
CW1
CW2
CWB1
CWB2
Cl
C2
Nl
N2
TW1
TW2
PI
P2
BC1
BC2
Avg. concentration
/ ^ yo/ \
|jg/nr (%)
5.0E4 (42.79)
5.4E4 (35.29)
2.3E4 (15.43)
5.7E4 (39.20)
6.1E4 (45.31)
1.2E5 (88.97)
5.5E4 (38.66)
4.4E4 (29.39)
2.1E4 (14.50)
1.5E4 (10.44)
2.8E4 (20.00)
4.2E4 (28.58)
2.9E5 (192.04)
3.1E5 (206.60)
1.8E5 (126.58)
1.4E5 (98.92)
g/hr
3.12
3.26
0.64
1.61
6.46
13.44
1.90
1.67
1.65
1.40
0.65
0.80
14.11
10.58
10.24
6.27
g/kg
0.69
0.82
0.12
0.17
2.05
7.64
0.32
0.45
0.52
0.40
0.11
0.18
3.14
2.50
5.35
2.43
-------
OJ
oo
Emission Factors
NO.
4
g/kg
2
g/kg fuel consumed
W CW CWB C N TW P
Figure 4. Emission factos: NO .
A
BC
12
-10
-------
cause of high NO emissions from the combustion of these fuels. The other
/\
fuel with exceptionally high NOX emissions, CWB, had the highest stack flow
rates of all the fuels tested. Omitting the fuels with high nitrogen content,
a correlation of N0x emission rate (g/hr) and stack flow rate was apparent.
The linear correlation coefficient for stack flow rates versus NO emission
rate for tests using fuels was 0.717. Increasing stack flow rate, which is
indicative of increased air flow through the stove, thus increases NO emissions
s\
This agrees with conclusions drawn by other researchers. No correlation was
found between firebox temperature and NO emissions, however, which disagrees
J\
with conclusions drawn by other researchers. It could be that firebox tempera-
ture did not vary enough during these tests to affect NO emissions.
/\
Maximum NO concentrations generally occurred a few minutes after fire
start. Other maxima occurred following fuel additions. Air flow changes
through the stove could be detected on the NO chart. Following door opening
to poke the fire, NO readings dropped immediately and then rose to levels
/\
above the level prior to opening the doors. The initial decrease reflects
dilution of the NO in the stack gas from the large amount of air introduced
into the stove following door openings. The subsequent increase in NO levels
reflects a higher emission of NO due to increased air flow through the stove
s\
through the open doors. NO concentrations also increased with increased
draft opening and damper opening, again reflecting increased NO formation
because of increased air flow through the stove.
During several tests, the NO detector was switched to the NO detection
mode to determine the percentage of N02 versus NO in the flue gas. N02 is
generally considered to be more of a problem than NO from environmental and
health standpoint. NO proportions are given in Table 11. It may be seen that
NO accounted for most of the total NO measured during these tests, accounting
/\
for an average of 76.09 percent of the total N0x_
Table 12 is a comparison of NO emission factors (g/kg) from the coal and
/\
wood tests in this study with emission levels from coal and wood combustion in
commercial units. Coal stoker units produce N0x emission factors similar to
those from residential stoves. Coal furnaces are slightly higher in N0x
emission factors. Commercial wood units are significantly higher in emission
factors than residential wood stoves, however. This difference can be attrib-
uted to differences in amounts of combustion air. Commercial wood burning
39
-------
TABLE 11. PERCENT NO IN N0x EMISSIONS
Test number
W1
W2
CWB2
C2
N2
TW2
NO level
?ppm)
116.74
101.52
53.80
50.25
39.59
40.61
45.68
46.19
63.96
46.19
43.14
34.52
36.04
33.50
50.18
42.73
33.29
30.81
49.74
30.96
34.01
26.39
25.38
21.50
16.90
13.31
12.80
24.50
26.00
19.00
19.00
NO level
(ppm)
51.77
43.65
41.11
38.02
26.39
27.41
35.02
31.98
45.68
31.47
33.50
28.93
26.90
27.92
48.20
38.26
27.33
25.84
40.61
20.81
21.83
20.81
20.81
14.85
13.82
11.26
10.75
19.50
29.50
14.50
14.50
Percent N0a
44.35
43.00
76.42
75.76
66.67
67.50
76.67
69.23
71.43
68.13
77.65
83.82
74.65
83.33
96.05
89.53
82.09
83.87
81.63
76.21
64.18
78.85
82.00
69.05
81.82
84.62
84.00
79.59
75.00
76.32
76.32
Average percent NO is 76.09.
40
-------
TABLE 12. COMPARISON OF NO EMISSION FACTORS FOR
COMMERCIAL AND RESIDENTIAL COMBUSTION
UNITS BURNING WOOD AND COAL
NO emission factor Percent N
x (g/kg) in fuel
Coal
Stoker boiler37
Furnace38
Residential stove
3-7.5
6.3-15
3.89a
ND
ND
1.54
Wood
Stoker boiler38
Residential stove
4.5
0.76
ND
0.29
This study, average of two tests
ND = No data.
41
-------
units operate with high excess of air, a condition conducive to NOX formation.
Starved air conditions used in most residential units minimizes NOX formation.
The same effect was not seen for coal because fuel nitrogen appears to have a
greater effect on NO emissions than stack flow for high nitrogen content
J\
fuels such as coal or peat.
7.1.1.4 Carbon Monoxide (CO--
CO emission factors are given in Table 13 and Figure 5. CO emissions for
the various fuels tested did not vary as much as with the previously discussed
pollutants. Ranking fuels according to CO emission factors (g/kg) fuel consumed)
is as follows (highest to lowest):
1. Newspaper logs
2. Compressed wood logs with binders
3. Peat
4. Bituminous coal
5. Cardboard logs
6. Compressed wood logs without binder
7. Treated lumber
8. Wood
Ranking of fuels according to CO emission rates (g/hr) is as follows
(highest to lowest):
1. Newspaper logs
2. Compressed wood logs without binders
3. Peat
4. Cardboard logs
5. Compressed wood logs with binders
6. Treated lumber
7. Wood
8. Bituminous coal
Reasons for the change in ranking between emission factors (g/kg) and
emission rates (g/hr) are related to fuel consumption rates and fuel heating
values as discussed in the previous section on NO emissions.
J\
CO emission levels could not be successfully correlated with stove operat-
ing parameters, including stack flow rates, firebox temperatures, stack tempera-
tures, fuel consumption rate, and heat output. Examination of Figure 5 reveals
42
-------
TABLE 13. EMISSION FACTORS—CO
CO
Fuel
Oak logs
Compressed wood logs
Compressed wood logs with binder
Cardboard logs
Newspaper logs
Treated lumber
Peat
Bituminous coal
Test
number
Wl
W2
CW1
CW2
CWB1
CWB2
Cl
C2
Nl
N2
TW1
TW2
PI
P2
BC1
BC2
Avg. concentration
pg/m3 (%)
2.5E6
2.9E6
1.4E7
8.9E6
2.1E6
2.0E6
9.4E6
5.1E6
4.8E6
3.3E6
9.0E6
5.4E6
5.3E6
8.4E6
2.0E6
3.0E6
(0.20)
(0.23)
(1-12)
(0.71)
(0.17)
(0.16)
(0.75)
(0.41)
(0.38)
(0.36)
(0.72)
(0.43)
(0.43)
(0.67)
(0.16)
(0.24)
g/hr
130.57
175.31
395.93
248.63
226.15
216.37
323.97
195.08
369.49
298.09
208.22
101.71
257.71
286.73
111.41
130.17
g/kg
28.90
44.00
71.51
26.09
71.88
123.02
54.42
52.43
115.49
84.94
37.01
22.97
57.31
67.82
58.22
50.52
-------
100
80
60
g/kg
40
20
Emission Factors
CO
g/kg fuel consumed
300
200
g/hr
100
TW P BC
Figures. Emission factors: CO.
-------
that the fuels with the lowest CO emissions are wood, treated lumber, and
bituminous coal. This is significant because all the other fuels with higher
CO emissions are manmade. They represent either compressed particulate or
chips (CW, peat, cardboard logs) cemented particulate (CWB), or rolled news-
paper. The physical structure of these fuels probably is the major cause of
their higher CO emissions. The tightness of these fuel structures probably
prevents air from reaching the inner portions of the fuel as it is burned,
resulting in incomplete combustion and higher CO emissions. Newspaper logs
have a tightly rolled structure of many sheets and the tightness of these logs
prevents air from reaching the inner layers before the outer layers burn off.
There is probably an oxygen concentration gradient across the log as it burns.
Paraffin impregnated CWB also is a very tight fuel which burns slowly and does
not crack open during burning as does coal or wood. The particulate formed
fuels (peat, CW, and cardboard logs) tended to fall apart when burned, forming
a pile of smoldering particulate matter in the stove. Oxygen starved condi-
tions almost certainly exist in the center of the pile, leading to increased
CO emissions in these fuels as compared to wood.
Maximum CO concentrations were generally a few minutes after fire start.
Other CO maxima occurred following fuel additions during each test.
Table 14 compares CO emission factors from commercial combustion units
with those from residential units (this study) for wood and coal. It may be
seen that CO emissions from residential stoves are significantly higher than
those from commercial combustion units. This can be attributed to the starved
air combustion conditions in residential wood stoves.
7.1.1.5 Carbon Dioxide (C02>~
C02 emission factors for the fuels tested are given in Table 15. C02
emissions were highest for compressed wood logs with binders. C02 emissions
for the other fuels were fairly uniform.
Maximum C02 concentrations generally occurred a few minutes after the
fire start. Other maxima occurred after each fuel addition. C02 maxima
usually lagged behind CO maxima by 2-3 minutes.
7.1.1.6 Organics--
Table 16 is a listing of emission factors for total organics. A compila-
tion of results from total chromatographical organics and gravimetric analyses
45
-------
TABLE 14. COMPARISON OF CO EMISSIONS FROM COMMERCIAL AND RESIDENTIAL
COMBUSTION UNITS BURNING WOOD AND COAL
CO emission factors
g/kg fuel consumed
Coal
Stoker boiler37 1-5
Furnace38 4.4-13
Residential stove 34.45
Wood
Stoker boiler37 0.9-27
Residential stove 54.36a
This study, average of two tests.
46
-------
TABLE 15. EMISSION FACTORS—C02
Fuel
Oak logs
Compressed wood logs
Compressed wood logs with binder
Cardboard logs
Newspaper logs
Treated lumber
Peat
Bituminous coal
Test
number
Wl
W2
CW1
CW2
CWB1
CWB2
Cl
C2
Nl
N2
TW1
TW2
PI
P2
BC1
BC2
hvg. concentration
ug/m3 (%)
1.2E8
8.6E7
2.2E10
1.7E8
1.5E8
1.0E8
1.8E8
1.2E8
5.1E7
4.8E7
1.9E8
1.9E8
1.2E8
1.9E8
7.2E7
7.2E7
(5.90)
(4.39)
(11.13)
(8.46)
(7.44)
(8.46)
(9.34)
(5.95)
(2.58)
(2.44)
(9.92)
(9.74)
(6.32)
(9.52)
(3.66)
(3.69)
g/hr
6.1E3
5.3E3
6.1E3
4.7E3
1.6E4
1.1 E4
6.3E3
4.4E3
3.9E3
4.4E3
4.5E3
3.6E3
6.1E3
6.4E3
4.0E3
3.1E3
g/kg
1.3E3
1.3E3
1.1E3
4.9E2
4.9E3
6.5E3
1.1 E3
1.2E3
1.2E3
1.3E3
8.0E2
8.2E2
1.4E3
1.5E3
2.1E3
1.2E3
-------
TABLE 16. EMISSION FACTORS—TOTAL HYDROCARBONS (+100° C ) (g/kg)
Run
Wl
W2
CW1
CW2
CWB1
CWB2
Cl
C2
Nl
N2
TW1
TW2
PI
P2
BC1
BC2
TCO
(100-300°Ca)
1.35
1.80
3.10
0.85
1.28
1.56
1.40
3.15
4.80
3.40
1.71
1.00
7.00
0.83
0.74
0.18
Gravimetric
(+300°Ca)
3.54
4.67
3.67
1.15
1.45
1.28
3.06
6.13
21.13
13.43
1.28
5.77
17.58
8.52
7.52
0.70
Total
4.89
6.47
6.77
2.00
2.73
2.84
4.46
9.28
25.93
16.83
2.99
6.77
34.58
9.35
8.26
0.88
Boiling point
48
-------
and represents total organics with boiling points greater than 100° C. It may
be seen in this table that total organic emissions in the flue gas were similar
for all fuels except for newspaper logs and peat, which had higher organic
emission factors. Newspaper logs showed a high amount of high boiling point
(+300° C) compounds, contributing to a high level of total organics. This is
somewhat surprising since newspaper logs had the lowest PNA emission factors
(see Fuels Comparison). Peat also had fairly organic emission factors, again
with high boiling organics contributing a large portion of the total organics.
Coal, with organic emissions comparable to most other fuels, had the highest
proportional contribution of heavy organics. The second coal run was sus-
piciously low in organics. Very low emission factors were also seen for PNAs
for this test. The disagreement with the first coal run, along with high
organic concentrations in coal flue gas reported by other researchers, implies
that the organic results for this test should be held in question.
Results of the gas bulb analyses are presented in Table 17.
7.1.1.7 Polynuclear Aromatic Hydrocarbons (PAH)--
Emission factors for the 24 polynuclear aromatic hydrocarbons analyzed in
this study are presented in Table 18. Figure 6 is a comparison of benz(a)pyrene
(B(a)P) emission factors for the eight fuels tested. It may be seen from this
table that B(a)P emission factors are lowest for newspaper logs. Wood and
coal have similar levels, contrary to results of other researchers. This may
be due to fairly low firebox temperatures and heat output during the coal run,
combined with good air circulation through the stove during the coal run.
Stove design may also be a factor. Other fuels had higher B(a)P emission
factors probably due to burning characteristics which reduce air to the fuel
(particulate fuels) or a high resin content combined with very low air flow
through the stove (treated lumber).
Some of the wood and coal PNA emission factors were compared to emission
factors from combustion of coal and wood in other types of residential units
(Table 19). The PNA emissions from wood combustion differ by an order of
magnitude, which reflects the difference in stove type and combustion conditions
Emissions from bituminous coal combustion in the wood stove and in a residential
hot water boiler, however, are in much closer agreement.
49
-------
TABLE 17. GAS BULB ORGANIC CARBON CONCENTRATION
Run
Wl
W2
CW1
CW2
CWB1
CWB2
Cl
C2
Nl
N2
TW1
TW2
PI
P2
BC1
BC2
Concentration
(ppm)
173
336
310
82
1,100
456
158
99
354
773
620
1,140
810
680
270
170
303
2,800
807
16
230
206
100
178
69
50
-------
TABLE 18. PAH EMISSION FACTORS (mg/kg)
Compound
Naphthalene
Biphenyl
Acenapthene
Fluorene
Phenanthrene
Anthracene
Carbazole
1-Methyl phenanthrene
9-Methyl anthracene
Fluoranthene
Pyrene
Benz( a) anthracene
Chrysene
12-Methyl benz(a)anthracene
6-Methyl chrysene
7-Methyl benz(a)anthracene
7,12-Dimethyl benz(a)anthracene
Benzo(b)f 1 uoranthene
Benzo(k)f luoranthene
Benzo(e)pyrene
Benzo(a)pyrene
Perylene
3-Methyl cholanthrene
Dibenz(a,h)anthracene
Benzo(g,h,i)perylene
Coronene
Wl
30.19
13.32
NA
4.75
8.19
4.69
NA
9.50
3.13
2.94
2.44
0.13
2.00
NA
NA
1.13
ND
0.44
0.69
0.44
0.31
ND
ND
NA
W2
43.29
9.46
8.03
4.59
11.32
7.38
ND
19.85
1.29
4.37
2.80
ND
1.15
0.07
0.50
1.79
ND
0.65
0.57
1.22
0.22
NA
0.07
ND
CW1
53.48
7.88
3.13
6.18
23.27
8.72
0.10
6.32
6.04
16.74
11.56
5.66
2.57
ND
1.88
6.04
ND
2.47
2.74
1.18
1.53
NA
0.90
0.90
CW2
27.94
2.29
0.56
1.94
8.16
1.88
ND
1.11
ND
5.38
4.10
1.39
1.28
ND
0.24
2.12
ND
2.08
1.94
0.83
ND
NA
2.05
ND
CWB1
335.01
16.35
1.39
1.65
48.80
ND
ND
15.44
20.05
24.22
14.05
ND
5.06
ND
3.61
8.25
ND
3.08
5.72
14.60
ND
NA
ND
1.24
CWB2
260.52
29.70
1.48
10.53
42.66
7.02
4.45
25.11
ND
17.01
12.01
1.75
1.35
ND
0.41
ND
0.94
0.54
0.67
0.67
0.54
NA
0.013
ND
Cl
70.90
8.21
1.62
3.52
19.82
1.62
0.29
3.11
6.34
15.01
9.25
2.65
1.99
ND
0.37
2.70
ND
2.70
1.82
0.70
ND
NA
1.62
ND
C2
108.34
8.71
4.24
8.76
21.89
2.71
ND
5.37
ND
9.79
3.21
2.08
2.03
0.04
1.17
3.07
ND
1.41
1.17
0.54
0.90
NA
0.14
ND
NA = Not analyzed.
ND = Not detected.
-------
TABLE 18. (continued)
Ul
r\>
Compound
Naphthalene
Biphenyl
Acenapthene
Fluorene
Phenanthrene
Anthracene
Carbazole
1 -Methyl phenanthrene
9-Methyl anthracene
Fluoranthene
Pyrene
Benz( a) anthracene
Chrysene
12-Methyl benz(a)anthracene
6-Methyl chrysene
7-Methyl benz(a)anthracene
7,12-Dimethyl benz(a)anthracene
Benzo(b)f luoranthene
Benzo(k)f 1 uoranthene
Benzo(e)pyrene
Benzo(a)pyrene
Perylene
3-Methyl cholanthrene
D ibenz( a, h) anthracene
Benzo(g,h,i)perylene
Coronene
Nl
64.44
33.22
NA
17.27
23.92
13.12
NA
5.48
NO
9.22
0.08
1.83
1.74
NA
NA
0.50
ND
0.33
0.25
0.17
NA
ND
ND
NA
N2
142.64
21.03
7.60
6.35
21.24
4.16
ND
4.16
ND
9.27
2.60
0.21
3.33
ND
0.10
0.21
0.21
0.42
0.31
ND
0.94
NA
ND
ND
TW1
1.15.87
11.67
2.35
5.96
18.30
4.98
1.18
3.44
ND
7.67
6.35
1.99
1.29
ND
0.34
4.03
ND
1.93
1.54
0.62
5.15
NA
0.20
ND
TW2
297.89
54.40
15.01
31.92
48.57
35.42
1.04
17.97
1.02
17.49
13.58
4.87
1.63
ND
0.39
6.31
ND
3.42
3.20
ND
8.44
NA
0.39
1.13
PI
82.98
18.36
NA
13.53
18.84
13.47
NA
10.68
8.91
12.72
8.22
6.33
6.60
ND
NA
0.07
0.07
3.67
5.78
6.33
0.07
ND
1.43
NA
P2
36.46
8.21
3.55
7.33
7.33
3.77
0.04
7.46
4.74
7.50
3.77
ND
1.97
ND
0.26
ND
0.22
0/IQ
. to
3.16
ND
NA
ND
ND
BC1
67.01
12.08
ND
10.75
24.29
6.56
NA
10.12
ND
8.59
7.05
1.95
2.51
NA
NA
1.75
ND
0.99
0.77
0.27
ND
ND
ND
NA
BC2
26.96
3.59
0.15
1.57
7.822
1.673
0.49
4.43
ND
2.71
1.67
+
0.30
ND
0.15
0.74
ND
0.54
0.20
3.05
0.20
NA
ND
ND
NA = Not analyzed.
ND = Not detected.
+ = May be present; peak interference.
-------
en
CO
5-
4-
mg/kg
n
W CW CWB C N TW P
Figure 6- Emission factors: benz(a)pyrene.
BC
-------
TABLE 19. COMPARISON OF EMISSION FACTORS OF PNAs (g/kg)
Naphthalene
Biphenyl
Fluorene
Anthracene/phenanthrene
Fluoranthene
Pyrene
Chrysene/benz(a)anthracene
7,12-Dimethyl benz(a)-
anthracene
Benzopyrenes and perylene
3-methyl cholanthrene
Oak,
this
study
0.0367
0.0114
0.0047
0.0158
0.0037
0.0026
0.0017
0.0015
0.002
0.0003
Oak,25
Baffled
stove
0.2729
0.0228
0.0224
0.0745
0.0180
0.0156
0.0125
-
0.0083
0.00007
Bituminous
coal, this
study
0.0670
0.0121
0.0108
0.0309
0.0086
0.0071
0.0045
0.0018
0.0020
NO
Coal,38
Hot water
boiler
0.28
0.009
0.048
0.029
0.01
0.009
0.007
0.15
0.006
0.003
NO = Not detected.
54
-------
7.2 FUELS COMPARISON
Table 20 gives results of proximate and ultimate analysis of the nine
fuels tested. Heating values (dry and wet basis) for each fuel are also given
in this table. Table 21 is a listing of important stove operating parameters
for each fuel test run. Note that heat output estimates are only used for
test comparisons and are not accurate estimates of actual heat output. Follow-
ing is a discussion, by fuel, of conditions encountered in each test, the
effects of these conditions on emissions, and any problems encountered in
reaching and maintaining steady-state stove operation. Table 22 is a listing
of ash analyses for all 16 tests.
Wood (Oak)
Wood (oak) was the best fuel tested in this program on the basis of
emissions and ease of stove operation. The stove chosen for this study was
designed to burn wood, and this was reflected by stove operating parameters.
Examination of Table 21 shows that there was the least amount of variation of
these parameters between the duplicate test runs than for any other fuel
tested. This is also reflected in the emission factors, which are very consist-
ent between "the two duplicate tests (Tables 6, 8, 10, and 13).
Because of this similarity between tests, it was not possible to determine
the effect of stove operating parameters on emissions for the wood tests. As
discussed under emission comparisons, wood was among the lowest emitting fuels
for all pollutants except for NO . NO emissions from wood were significantly
}\ }\
lower than the high NO emitting fuels, however, with NO levels more comparable
X A
to the four low NO emitting fuels (Figure 4). Maximum NO , CO, and S02
X X
levels for the wood tests are given in Table 23 along with the time in each
test these maxima occurred. It may be seen from this table that pollutant
emission maxima occurred shortly after the fire was started or after wood
addition to the fire.
Total organic emissions were similar for the two duplicate tests with
wood (Table 16). Organic emissions were also similar for wood and all other
fuels except peat and newspaper logs, which had significantly higher organic
emissions. Heavy organics (+800° C boiling point) accounted for about 2/3 of
the total organic emissions. PNA emission factors were also comparable for
the duplicate wood tests. Wood PNA emission factors were generally lower than
those from other fuels, especially considering the heavier PNAs (those with
55
-------
TABLE 20. FUEL ANALYSES
en
a\
wa
CW
CWB
C
N
TW
P
BC
AC
Proximate analysis (as received)
Percent moisture
Percent ash
Percent volatile
Percent fixed carbon
Btu's/lb (as received)
(dry basis)
Ultimate analysis (dry basis)
Percent carbon
Percent hydrogen
Percent nitrogen
Percent chlorine
Percent sulfur
Percent ash
Percent oxygen
10.87
0.99
74.14
14.01
7310
8201
50.90
6.20
0.29
0.00
0.09
1.11
41.41
10.45
0.42
76.62
12.51
7946
8873
53.30
6.12
0.44
0.00
0.89
0.47
38.78
2.
0.
89.
6.
70
71
77
82
13925
14311
67.
9.
0.
0.
0.
0.
20.
56
92
34
29
25
73
97
9.
1.
77.
11.
23
17
93
67
7331
8077
49.
6.
0.
0.
0.
1.
42.
96
01
37
00
35
29
02
8.54
0.51
78.68
12.27
7805
8534
53.06
6.04
0.00
0.02
0.16
0.56
40.16
9.16
1.14
77.32
12.38
7890
8686
52.19
6.20
0.43
0.00
0.005
1.25
39.93
12.74
2.20
56.85
28.21
8668
9933
54.89
5.98
0.93
0.05
0.21
2.52
30.42
2.37
10.97
32.61
54.05
12731
13040
73.36
4.72
1.54
0.06
1.87
11.24
7.21
4.08
10.22
4.54
81.16
12178
12696
84.04
2.00
1.14
0.01
0.57
10.65
1.59
Average of two analyses.
By difference.
-------
TABLE 21. OPERATING PARAMETERS—SAMPLING PERIOD AVERAGES
en
•vl
Oak logs
Compressed wood
Compressed wood w/binders
Cardboard logs
Newspaper logs
Treated wood
Peat
Bituminous coal
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
Flow
rate (dry)
(scfm)
30.73
35.88
16.64
16.49
62.62
63.62
20.34
22.40
45.78
53.97
13.62
11.14
28.89
20.15
32.78
25.54
Fuel
consumption
rate (kg/hr)
4.518
3.984
5.490
9.530a
3.128
1.750
5.953
3.711
3.199
3.510
5.648
4.427
4.497
4.227
1.914
2.580
Stack
tempera-
ture (°C)
286
248
197
262
254
250
284
238
219
253
279
188
300
273
288
335
Firebox Sample
tempera- volume
ture (°C) (scm)
515
537
601
570
644
516
633
531
400
430
483
554
565
529
464
597
1.392
1.375
0.787
0.933
1.140
1.302
0.889
0.850
1.155
1.344
1.659
1.688
0.473
1.222
1.343
1.646
Estimated
heat output
Btu's/hr
65477. 1
62053.7
72975.6
59706.2
77406.3
62218.3
61181.6
64320.6
19644.5
53334.4
66197.8
61273.7
64502.9
35937.5
52499.7
Extrapolated.
-------
TABLE 22. ASH ANALYSIS
en
CO
Wl & 2
CW1
CW2
CWB1
CWB2
Cl
C2
Proximate (as received)
Percent
Percent
Percent
Percent
moisture
ash
volatile
fixed carbon
Btu's/lb (as received)
(dry basis)
Ultimate (dry
Percent
Percent
Percent
Percent
Percent
Percent
Percent
basis)
carbon
hydrogen
nitrogen
chlorine
sulfur
ash
oxygen9
1.15
57.52
35.17
6.16
2209
2235
22.82
0.25
0.45
0.04
0.34
58.19
17.91
1.85
54.93
35.15
8.07
2701
2752
25.93
0.65
0.40
0.00
0.24
55.97
16.81
8.
25.
27.
38.
08
50
63
79
0
93
16
-9
7513
8173
55.
1.
0.
0.
1.
27.
13.
07
59
58
05
04
74
93
4
0
0
0
.29
.10
.22
.61
245
346
.64
.15
.45
.84
5.86
93
-5
.37
.31
2
81
27
-11
1
1
11
0
0
0
5
83
-2
.18
.82
.17
.17
170
196
.41
.40
.63
.84
.13
.64
.05
6.17
61.37
27.99
4.47
699
745
9.06
1.63
0.36
0.33
3.89
65.41
19.32
2.26
73.67
24.88
0.81
2066
2114
16.06
0.37
0.45
0.25
3.83
75.38
3.66
By difference.
(Continued)
-------
TABLE 22. (continued)
en
ID
Nl
N2
TW1
TW2
PI
P2
BC1
BC2
Proximate (as received)
Btu1
Percent moisture
Percent ash
Percent volatile
Percent fixed carbon
s/lb (as received)
(dry basis)
10.94
31.11
23.44
34.51
6348
7128
22.72
29.45
22.05
25.78
4961
6420
6.45
33.43
15.67
44.46
8225
8792
8.72
36.42
6.10
48.76
7363
8066
2.37
28.32
13.23
56.08
9163
9385
2.56
40.14
10.77
46.53
7593
7793
1.80
47.71
6.61
43.88
6864
6990
2.23
61.69
5.98
30.10
4892
5004
Ultimate (dry basis)
Percent carbon
Percent hydrogen
Percent nitrogen
Percent chlorine
Percent sulfur
Percent ash
Percent oxygen
48.77
1.43
0.59
0.12
1.00
34.93
13.16
44.60
1.16
0.59
0.13
1.00
38.11
14-41
58.90
0.72
1.77
0.00
0.22
35.72
2.67
53.86
0.89
1.86
0.00
0.22
39.90
3.27
62.64
0.95
1.66
0.07
0.44
29.01
5.23
51,92
0.69
1.29
0.07
0.57
41.19
4.27
47.06
0.88
0.95
0.03
0.78
48.58
1.72
21.06
0.47
0.90
0.02
0.84
63.09
13.62
By difference.
-------
TABLE 23. EMISSION MAXIMA—WOOD TESTS
Test number
NO Wl
X
W2
S02 Wl
W2
Maximum concentration Time of maximum
254.74 ppm 3 min after fire start
68.24 ppm 9 min after 22
addition; 2
draft/damper
None Levels uniform
test
None Levels uniform
test
.6 Ib wood
min after
opening
throughout
throughout
CO
Wl and 2
1.0%
2 min after fire start
60
-------
boiling points greater than that of perylene). Wood PNA emission factors were
similar to those for bituminous coal and higher than those from newspaper
logs, the fuel with the lowest PNA emissions.
The wood used in this study was fairly dry (11 percent moisture) and was
obtained from a local firewood supplier. It is representative of firewood
available in North Carolina. The burn rate (about 4 kg/hr) was higher than
other studies using wetter wood,11 12 13 14 which resulted in lower particulate
emission factprs for this study. Wood with high moisture has been shown to
have higher particulate emissions, with emission factors reaching over 20
g/kg.13 Because of the great variability of reported particulate emission
rates for woodstoves, under certain conditions wood particulate emission
factors could be higher than for the other fuels in this study. However,
during this study the burn rate for each fuel was controlled (when possible)
to provide equal heat output from the stove for each test. Thus, we feel that
the comparison between the alternate fuels' particulate emission factors and
the emission factors for dry oak are valid. Care should be taken when compar-
ing the wood particulate emission factors in this study with those from other
studies with different stoves, wood types, wood moisture content, and burn
rates.
Compressed Wood Logs (without binders)
These logs, formed of compressed wood chips, were determined to be unsuit-
able for use in residential wood stoves on several grounds. First, it was
impossible to control the fire with draft and damper settings. Even with the
stove closed up as completely as possible, the flame remained very high. The
fact that the stove used in this study was not airtight contributed to this
problem, but with an airtight stove, drafts would have to be so restricted
that increased pollutant production including CO, particulate, and condensible
organics, would almost certainly result. In addition, immediately following
ignition these logs swelled and fell apart, resulting in a pile of smoldering
sawdust in the stove. On the first test with these logs, test personnel,
unaware of this problem, opened the doors of the stove to add more fuel and
received a lapful of burning sawdust which had piled up against the doors as
the log disintegrated. The use of the coal grate in the second test alleviated
this problem, but it is still felt that these logs represent a definite safety
hazard for woodstove use.
61
-------
In addition, compressed wood logs (CW) had higher CO emission factors
than wood, related to their tendency to fall apart when burned. NO and S02
A
emissions were low for CW. Participate emissions from CW were higher than
wood or treated lumber but lower than other fuels.
In spite of an attempt to keep conditions constant between duplicate
tests with CW, the two tests differed some in stove operating parameters (see
Table 21). Although average flow rate for the sampling period was similar for
the two tests, other parameters show that the runs were different. These
differences can be used to show some of the effects of changing stove opera-
tion parameters on emissions. CW1 had lower fuel consumption but a higher
relative heat output than CW2. Comparison of stack and firebox temperatures
for the two tests reveals the reason behind this difference. Higher firebox
temperatures and lower stack temperatures for CW1 indicate that flow through
the stove was more restricted during CW1 than during CW2, resulting in lower
stack heat loss, higher amounts of heat radiated into the room by the stove,
and lower fuel consumption for this test compared to CW2. Higher fuel consump-
tion resulted in a lower g/kg particulate emission factor for CW2, in spite of
a similar emission rates (g/hr) for the two tests (Table 6). CO emission
factors (g/kg) and emission rates (g/hr) were both significantly higher for
CW1 than for CW2. The more restricted flow during CW1 is probably responsible
for these high CO emission levels. Although NO emissions were low for both
J\
CW tests, lower NO emission factors and emission rates for CW1 also reflect
P\
the more restricted air flow through the stove during this test.
Total organic emissions for the CW tests are similar to those from tests
with most other fuels (Table 16). Heavy organics (boiling points >300° C)
accounted for a slightly larger portion of total organics than the light
organics (boiling point 100° C-3000 C); but levels of both were fairly similar.
Total organic emissions were higher for CW1 than for CW2, again reflecting
more restricted air flow during this test and have lower combustion efficiency.
PNA emission factors were higher than for wood tests and were similar to
those from tests of other composite fuels (except newspapers) and to tests of
treated lumber. This reflects relatively low combustion air availability
attributable to structural characteristics of the fuel and low air flow through
the stove, as for the treated lumber tests. PNA emissions factors for CW1
were generally higher than those for CW2, although this difference was less
62
-------
apparent for the higher molecular weight PNAs. Again this is probably related
to more restricted air flow through the stove for CW1 compared to CW2. PNA
discharge severities were highest for CW than any other fuel, reflecting
higher PNA production and low flow rates which result in a more concentrated
flue gas effluent.
Compressed Wood Logs (with binders)
It should be noted that package directions on this product specifically
warned against using these logs in a woodstove and burning more than one of
these logs at a time, and that both recommendations were not followed during
the tests with compressed wood logs with binders (CWB). Compressed wood logs
with binders produced a thick, greenish gray smoke when burned. Emission
factors for CWB were, except for total hydrocarbons, were among the highest of
all fuels for all the pollutants measured. Particulate emission factors for
CWB were by far the highest of all fuels (Figure 2). The particulate produced
during the combustion of CWB was sooty and sticky, similar to that from coal,
and probably represents organic volatilization products from the paraffin
binder used in the fuel. Buildup of this particulate resulted in a high
pressure drop across the Method 5 filter during the later state of sampling.
NO emission factors (g/kg) were also highest for CWB and NO emission rates
/\ ./\
(g/hr) were second only to peat. This switch in ranking is due to the lower
fuel consumption rates for CWB. Because CWB fuel nitrogen content was moderate,
high NO emissions from CWB are probably attributable to the high air flow
through the stove during the CWB tests, highest for all fuels, and to the
relatively high average firebox temperatures (644° C and 516° C) during the
tests. S02 emissions from CWB were second only to bituminous coal. The
reason for this is not immediately apparent because the CWB sulfur content was
only 0.25 percent. It is possible that the fuel analysis is in error since
these analyses were designed for coal samples, and CWB is not very similar to
coal.
CO emissions factors for CWB were comparable to those from other fuels,
although once again comparing emissions on a g/kg basis produced a different
ranking (2nd) for CWB than for emissions on a g/hr basis (5th). This is
because of the high heating value and hence low fuel consumption rate for CWB.
63
-------
Total organic emissions were surprisingly low (Table 16) considering the
waxy, organic appearance of the particulate. It is possible that methylene
chloride was not completely effective in dissolving the filter organics. PNA
emissions were fairly high for CWB, however, on a mg/kg basis, being very high
for naphthalene and biphenyl, compared to other fuels, and highest in B(a)P
emissions on a mg/kg basis.
Test conditions were similar for the two duplicate tests with CWB except
that the damper was half open during CWB2 and closed during CWB1. This resulted
in higher firebox temperatures for the first test because of more restricted
air through the stove, also reflected by the lower flow rate for CWB1 and
higher heat output for CWB1. Some interesting conclusions about the combustion
characteristics of CWB may be arrived at from examining the differences in
operation parameters and emissions for the two tests. Fuel consumption was
lower for the second test than for the first in spite of a higher air flow
through the stove during CWB2. This implies that fuel consumption rate is
controlled more by combustion temperature than availability of excess air for
combustion. It follows that the physical structure of CWB is the most important
limiting factor on air supply for combustion. In other words, the availability
of oxygen to the site of combustion of the on the CWB is limited by the diffusion
of air to the combustion area, and the rate of this diffusion controls combustion.
The CO emission rates (g/hr) reflect this, since CO emission rates are remarkably
similar in spite of difference in firebox temperature, combustion air flow,
and fuel consumption rates.
Particulate emission rates (g/hr) are higher for CWB1 than for CWB2,
reflecting the higher fuel consumption rate for CWB1. Indeed, putting particu-
late emissions on a g/kg basis, shows that a similar amount of particulate is
emitted per unit mass of CWB burned, although particulate emissions factors
are slightly higher than for CWB2, probably reflecting the lower firebox
temperatures which encourage less complete combustion. NO emissions were
higher on both a g/hr and a g/kg basis for CWB2, reflecting the increased air
flow through the stove for this test. Since firebox temperatures are lower
for CWB2 than for CWB1, this observation supports the conclusion that combus-
tion air flow has a greater effects on NO emissions than firebox temperature.
.A.
Comparison of PNA emission factors (Table 18) mg/kg for CWB2 shows that,
although the emission factors for the lighter PNAs are similar for two runs,
64
-------
emission factors for the heavier PNA (chrysene on) are about an order of
magnitude lower than CWB2 than for CWB1, in spite of lower fuel consumption
rate for CWB2. The lower average firebox temperature and the higher flows
through the stove during CWB2 seems to have reduced the amount of cyclization
reactions in the firebox during this test.
Cardboard Logs
Except for wood, cardboard logs (C) were about the best fuel tested in
this particular stove. Cardboard logs burned similarly to wood being easy to
start and keep burning during the tests. No problems were encountered keeping
heat outputs constant at sufficiently high levels during the cardboard logs
tests. Emissions from cardboard logs also compared favorably with those of
wood. Particulate emissions were low, comparable but slightly higher than
those from wood combustion. Carbon monoxide emissions were higher for card-
board logs than for wood or coal, but were the lowest among the manmade par-
ti cul ate type fuels. As discussed before, elevated CO levels for cardboard
logs compared to wood were probably due to the compressed particulate nature
of this fuel. NO emission factors were lower for cardboard logs than for
s\
wood. S02 emissions were low, at about the same level as S02 emissions from
wood combustion.
Total organics from the combustion of cardboard logs were comparable to
most other fuels. As with wood, about two thirds of the total organics were
heavy organics with boiling points greater than 300° C. PNA emission factors
for cardboard logs were higher than those for wood combustion, but similar to
those from other formed particulate type fuels. Again this reflects the
burning characteristics of the cardboard logs which results in local oxygen
starved conditions in the combustion zone and higher CO and PNA emissions than
for naturally formed fuels like wood.
Stove operating parameters for the duplicate tests for cardboard logs
were comparable except that a higher firebox temperature was achieved during
Cl, probably because the flow rate through the stove was lower during this
test. This higher firebox temperature was concurrent with a higher fuel
consumption rate during Cl. As with CWB, increased air flow through the stove
did not result in increased fuel consumption because increased flow also
reduced firebox temperature. Emission factors (g/kg) were very similar for
65
-------
all pollutants during Cl and C2. However, slightly lower emission factors for
CO and heavy PNAs for C2 probably reflect the higher air flow through the
stove during C2, as does the slightly higher NO emission factor. The higher
total organic emission factors for total organics for C2 also reflect the
lower firebox temperatures during this test. Pyrolysis products were not
burned as completely during C2 as during Cl because of the lower temperatures
and a decreased residence time in the stove from the higher flow rates. This
reasoning also can explain the lower PNA formation since lower residence time,
lower temperatures, and higher amounts of combustion air can tend to inhibit
PNA formation through cyclization reactions.
Newspaper Logs
As a stove fuel, newspaper logs (N) were not up to par with the other
fuels considering stove operation parameters. Although newspaper logs com-
pared well with wood in heating value per unit mass, it was difficult to
achieve satisfactory heat outputs from the stove during the tests with news-
paper logs. Firebox temperatures remained low during the newspaper log tests.
This was primarily because of the construction of the newspaper logs. These
logs were formed from tightly rolled newspapers, a construction which restricted
air circulation into the logs, resulting in a slow burn rate. In addition,
the outer layers tended to burn and remain on the surface of the newspaper
logs, insulating the remaining logs and further inhibiting its combustion.
Very high air flows had to be maintained through the stove to assure continued
combustion. Rolling the logs less tightly may be a partial solution to this
problem.
Newspaper logs had negligible S02 emissions. NO emissions were also
low, in spite of the high air flows through the stove during the newspaper
tests. Low firebox temperatures are thought to be partially responsible for
the low NO emissions. Firebox temperatures simply were not high enough for
appreciable NO formation. Another reason for low NO emissions is that fuel
X X
nitrogen content for newspapers was zero. Thus, all of the NO emissions
}\
during the newspaper tests are from the oxidation of atmospheric nitrogen.
Newspaper logs ranked third in particulate emission factors. Much of
this particulate represents pieces of burnt paper which spall off the newspaper
logs as they burn. The shape of these particles (platelike) enhances their
66
-------
entrainment in the flue gas stream. High, heavy organic emissions from news-
paper logs contributed to the high particulate levels.
Newspaper logs produced the highest CO emissions, in spite of high combus-
tion air flows through the stove. High CO emissions for the newspaper logs
thus seem to be because of the newspaper log construction, which, as previously
discussed, prevents sufficient combustion air from reaching the fuel.
Newspaper logs had, on the average, emitted the highest amount of total
organics during combustion. However, PNA emission factors from newspaper
combustion were the lowest, especially for the heavier PNAs. Comparing the
fuels, B(a)P emission factors were lowest for newspaper logs. This discrepancy
points to some interesting conclusions about PNA formation. Firebox tempera-
tures were lowest for newspaper logs than for any other fuels and stack flow
rates were second highest. These conditions seem to have inhibited the cycliza-
tion reactions necessary for forming PNA compound to the extent that not much
of the heavier more condensed PNAs were formed. Most of the organics (75 per-
cent) collected by the modified Method 5 train were heavy organics (boiling
points greater than 300° C). These probably represent pyrolysis and distilla-
tion products evolved from the newspapers. At the combustion site in the
newspaper logs, these products evolved because insufficient combustion air was
present in the combustion zone. These conditions were also responsible for
high CO levels during the newspaper tests. As these pyrolysis products evolved
form the burning logs, a combination of low firebox temperatures and low
firebox residence time (from high flow rates) prevented extensive combustion
or cyclization of these compounds. This led to high total organic concentration
but low concentrations of heavy PNAs in the flue gas stream.
Comparison of the stove operating parameters for the duplicate tests with
newspaper logs reveals that the two tests were similar except for a lower
firebox temperature and much lower heat output during Nl (Table 21). The very
low heat output for Nl reflects the fact that medium blower speed was never
achieved during this test, and blower speed was factored in when calculating
heat outputs. The newspaper logs simply would not burn fast enough during Nl
to heat the stove up to normal operating temperature. Better results were
achieved for the second test by not rolling the newspaper logs as tightly as
for the first test.
67
-------
CO and total hydrocarbon emission factors were lower for N2 than for Nl
(Tables 13 and 16). This reflects higher firebox temperatures which resulted
in more complete combustion of these emissions during N2. PNA emission factors
were slightly higher for N2 than for Nl. Again this reflects the higher
firebox temperatures during N2, which seems to increase PNA formation through
cyclization reactions.
Treated Lumber (TW)
Treated lumber (TW), like CW, burned very readily in the woodstove with a
very high flame spread. This required both draft and damper setting to be
closed as much as possible in order to try to control the burning of TW. Even
with draft and dampers completely closed, flames were still visible in the
firebox during the combustion of TW. However, combustion was not as violent
as with CW, so that flow rates through the stove were lower than the tests
with CW, and were the lowest encountered for any fuel tested.
S02 emissions were negligible for TW. NO emission were lowest for TW as
compared to other fuels. This reflects the low flow rate through the stove
during the treated wood test. In spite of this low flow rate, CO levels were
very similar- to those on the combustion of wood. This supports the conclusion
that fuel structure is important in determining CO emissions from different
fuels, because treated wood and wood (oak) are similar in structure and dissimi-
lar in structure from the fuels formed of compressed particulates all of which
had high CO emissions. Particulate emissions from TW ranked fourth among
fuels when expressed as g/kg. TW ranked higher in particulate emissions
factors than other fuels because of an increased amount of condensed organic
on the filter during the tests with TW. The two tests with TW varied consider-
ably in their particulate emission factors. The first test had emission
factors for particulates similar to those from the wood combustion test.
However, the second test had much higher particulate emission factors. Examina-
tion of total organic emissions for the TW tests (see Table 16) shows that
organic emissions during TW2 were indeed higher than those during TW1. Most
of the organics in TW2 were heavy organics. Higher amounts of heavy organics
for TW2 reflects the lower flow rates for this test than for TW1. Relatively
high ranking of TW in particulate emission factor is largely due to the high
heavy organic loading on the filter during the TW2 test.
68
-------
Examination of those operating parameters for the TW tests shows that
flow rates were higher for TW1 than for TW2. This resulted in a higher fire-
box temperature and lower stack temperature than for TW2 and a slightly lower
fuel consumption rate. Unlike particulate fuels, the fuel consumption for TW
seems to be controlled by the flow rate of air through the stove during stove
operation. As mentioned before, the lower flow rate during TW2 resulted in
higher particulate and hydrocarbon emission factors for this test compared to
TW1. It appears that air flow was restricted enough during TW2 to prevent
combustion of the heavy organics evolved from the wood as pyrolysis products
during combustion. These heavy organics deposited on the filter in modified
Method 5 train resulting in higher particulate emission factors and higher
organic emission factors for TW2.
Overall organic emissions compared favorably with most of the other fuels
for the TW tests. However, the second TW test had a higher proportion of
heavy organics than did most of the other fuels. PNA emission factors were
fairly high for the TW tests. PNA emission factors were lower for TW1 than
for TW2, reflecting higher flow rates through the stove than for TW1. This,
in combination with a lower firebox temperature for TW1, seems to have led to
lower PNA formations for TW1 compared to TW2.
The high flame spread from TW, probably resulting from high resin content
of this fuel, which was treated pine, makes this fuel unsuitable for wood
stove use. In addition, contact with the manufacturer of this product revealed
that the lumber was pressure treated with copper oxides compounds to retard
rot. Trace amounts of arsenic and chromium compounds at ppm levels are also
added to this wood to enhance its resistance to fungus and insect damage.
Examination of the Modified Method 5 sampling train after the TW tests revealed
that the XAD module had turned a rather bright shade of green. This is visual
evidence of a rather significant copper volatilization during the tests with
TW. Since copper is a less volatile element than arsenic, it can be expected
that most or all of the arsenic present in the wood will be volatilized during
the combustion of the wood. Other studies have shown that in the combustion of
chlorophenol-treated wood products polychlorinated dibenzo-p-dioxins (PCDD)
and polychlorinated dibenzofurans (PCDF) are emitted.39 For this reason, it is
recommended that TW not be burned under any circumstances including residential
applications such as woodstoves and fireplaces.
69
-------
Peat
From the stove operating standpoint, peat was a fairly satisfactory fuel.
Although the sod peat burned in this test was a particulate fuel, and thus
fell apart when burned, it did not have much flame spread as did the CW.
Therefore, stove draft and damper settings similar to those used with wood
could be also used with peat. Using the coal grate for the peat combustion
test prevented the problem of particulates falling out of the stove when the
doors were opened. Flames were fairly low off the burning peat and after a
while, all that could be seen in the firebox were red glowing remains of the
peat sod. Normal stove operating temperatures and heat output were easily
achieved using the peat. The smoke from peat combustion was very fragrant.
Opinions differed on whether this fragrance was pleasant or decidedly unpleasant.
Peat combustion in residential installations may therefore have problems from
an aesthetic standpoint.
NO emissions for peat were the highest among all the other fuels tested
on a g/hr basis and were third highest on a g/kg basis (Figure 4). The high
NO emission from peat combustion can be attributed to the high nitrogen
A
content in the peat (0.93 percent). Nitrogen fuel content for peat was lower
than that for bituminous coal but considerably higher than nitrogen contents
for the other fuels. Peat ranked third in S02 emissions. CO emissions were
relatively high for peat combustion, compared to those form other particulate
fuels, and higher than those form the naturally structured fuels, coal, treated,
wood, and wood (oak). High CO emissions factors for peat are probably due to
its particulate structure, as explained for the other particulate fuels.
Particulate emissions from peat were similar to those from CW. Peat ranked
fifth in particulate emission factors.
Examination of the stove operating parameters (Table 21) shows the two
test with peat were similar. Differences in emission factors for the various
pollutants measured during these tests cannot be readily explained by differ-
ence in stove operating parameters during duplicate tests with peat.
Peat ranked second in total organic emissions. Most of the total organics
from peat combustion were heavy organics (boiling points greater than 300° C).
PNA emission factors were fairly high for PI but were low for P2. No explana-
tion is immediately apparent for this discrepancy.
70
-------
Bituminous Coal
In order to burn bituminous coal in the stove used for this study, a
special optional shaker grate was purchased from the stove manufacturer.
Although this grate did make it possible to burn bituminous coal in the stove,
it is felt that the grate design in combination with the stove design prevented
the most efficient combustion of this fuel. The shaker grate sat in the
middle of the stove's firebox. One to two inches of space was around the
grate on all sides. This enabled much of the air entering the stove to flow
up and around the grate, thus bypassing the burning fuel bed. This resulted
in a less than efficient combustion of the bituminous coal.
S02 emissions from combustion of bituminous coal were higher than for any
other fuel tested. This is a reflection of the fact that bituminous coal had
the highest fuel sulfur content (1.87 percent) of all fuels tested. NO
emissions from coal were also very high. Bituminous coal ranked second in NO
){
emissions on a g/kg basis and third in NO on g/hr basis. These relatively
/\
high NO concentrations in flue gas from bituminous coal combustion are largely
due to the high fuel nitrogen content of bituminous coal (1.54 percent). CO
emission factors for bituminous coal were moderate, slightly higher than those
from the combustion of wood or treated wood but lower than emission factors
for CO for the particulate structured fuels. Again this reflects the natural
structure of bituminous coal versus the composited particulate structure of
the particulate fuels. Bituminous coal ranked second in total particulate
emissions. This particulate is the infamous coal soot, a sticky mixture
largely composed of unburned carbon particles with some adsorbed volatile
organics.
Organic emissions form coal were surprisingly low. Total organics were
largely composed of heavier organics (boiling points greater than 300° C).
PNA emission factors were also low, comparable or a bit lower than those for
wood, depending on which bituminous coal test was compared to the wood test.
This may be because of the stove in which the bituminous coal was burned.
Examination of stove operating parameters for bituminous coal tests
(Table 21) shows that there was some difference between the two duplicate
tests. Stove drafts were fairly open during the first test of bituminous coal
per manufacturer's instruction. This resulted in lower burn rate for this
71
-------
test and a fairly low heat output for this test. Firebox temperatures for BC1
were also lower than those for BC2. During BC2, the stove was closed off more
so that it could reach proper operating temperatures. This resulted in a
lower flow rate through the stove at a higher fuel consumption, probably
because of the higher firebox temperature. Dampers were left halfway open
for both tests resulting in stack temperatures considerably higher than those
from the other tests. S02 and CO emission factors were similar in BC1 and
BC2, NO emission factors were higher in BC1 than for BC2, however. This
/\
reflects higher flows through the stove during BC1. Particulate and total
organic emissions were higher for BC1 than for BC2. These are believed to be
related since most of the extra organic matter in BC1 compared to BC2 seems to
come from heavy organics (boiling point greater than 300° C). Thus high
boiling organic compounds contributed to the particulate catch for BC1, result-
ing in a higher emission particulate factor for this test. PNA emission
factors were higher for BC1 than for BC2 (Table 18). Since most of the heavy
organics emitted during BC1 probably represent pyrolysis product from the coal
and since higher firebox temperatures and lower flow rates during BC2 would
seem to favor formation of PNA during combustion of BC2, lower levels of PNA
actually detected in the flue gas samples for BC2 imply that PNAs from bituminous
coal combustion are primarily pyrolysis products from bituminous coal. This
is supported by the fact that bituminous coal has a much more condensed aromatic
structure than the other fuels tested. However, PNA concentrations in the
sample extract for BC2 are admitted suspiciously low. Analytical problems
thus cannot be ruled out. For this reason, in calculating the PNA discharge
severities only the values for BC1 were used.
Anthracite Coal
Anthracite coal could not be burned successfully in the stove chosen for
this study. The coal grate designed for the stove did not perform well with
anthracite. Air flowed around the grate instead of through it, and after the
kindling loaded in the stove died down, the coal smoldered and went out. On a
subsequent test, an oak fire was started and a good bed of coals was established
before adding coal. Although the coal was added to the coal bed in small
amounts, it did not burn with any flame and as soon as the wood coals died, so
did the fire. A second coal grate was obtained form another manufacturer who
72
-------
said it could successfully burn anthracite in a stove very similar to the one
used in this study. This grate was designed so that most of the air coming
into the stove went through the grate. This grate performed better, but a
self-sustaining anthracite fire still could not be established. Air was
forced through the grate using a blower during one test, and a small blue
flame could be established on the coal bed. However, with the removal of the
forced air, this flame quickly died out, and the coal once again did not burn.
It was therefore concluded that it is very difficult to burn anthracite in a
conventional woodstove and that a stove specifically designed to produce high
air velocities through the coal grate is necessary for successful anthracite
combustion. This is due to the high fixed carbon content and low volatile
content of anthracite (see Table 20).
7.3 BIOASSAY RESULTS
Method 5 sample extracts from one wood combustion (Wl) and one coal
combustion test (BC1) were subjected to an Ames Salmonella mutagenicity assay
to measure their mutagenic potential. Although IERL Level I Environmental
Assessment and the protocol of Ames and others40 specify testing with five
Salmonella 'strains, only two strains (TA98 and TA100) were utilized for this
study because of limited sample size and economic considerations.
The results of bioassay analysis suggest the presence of frameshift and
base pair substitution mutagens in both samples. Both samples were highly
mutagenic with TA98 and moderately mutagenic with TA100. Both samples demon-
strated an increase in mutagenic activity with the addition S9, a metabolic
activator of promutagen compounds. Therefore, both samples contain direct-
acting mutagens and promutagens.
The coal combustion sample was more mutagenic than the wood combustion
sample, based on the slope of the dose/response curves in units of revertants/
mg of sample. Putting bioassay results on a revertants/kg of fuel consumed
basis, the coal extract is more mutagenic than the wood extract by a factor of
two. Since emission factors (g/kg) for the PNAs analyzed in this report are
only slightly higher for BC1 than for Wl, this suggests that compounds other
than the 24 PNAs analyzed in this report may be contributing to the mutagenic-
ity of these samples.
73
-------
Complete bioassay methods and test results are presented as a separate
report. In this report, 110201 is the wood sample and 11007 is the coal
sample. These correspond to samples Wl and BC1 discussed in this report.
74
-------
8. CONCLUSIONS
1. Overall oak wood was the best fuel, considering both emissions and stove
operation. Cardboard logs (C) were almost as good as wood. Although
they did emit more CO and PNAs than wood, levels of these pollutants were
lower than for most other fuels, and stove operation was easier with C
than with other fuels.
2. Compressed wood logs with binders (CWB) and bituminous coal (BC) produced
the highest emissions (g/kg fuel consumed) of S02, particulate, and NO .
/\
In addition, CWB emissions were high in CO and PNAs.
3. Compressed wood logs without binders (CW) were determined to be unsuitable
for stove use on safety grounds. CW also emitted large amounts of CO.
4. Treated wood (TW) should not be burned under any circumstances because of
the presence of arsenic compounds which probably volatilize during combus-
tion.
5. Peat (P) emissions had relatively high levels of NO , S02, CO, and PNAs.
A
6. Particulate matter from BC and CWB combustion was sooty and sticky.
These fuels produced the highest particulate emission by far. Composite
fuels (CW, C, P, newspaper (N)) produced particulate emissions higher
than those of wood. High particulate levels for N and TW were largely
attributable to condensed organics.
7. Important parameters affecting CO emission levels were fuel structure
and, to a lesser degree, combustion air flow. Fuels with a manmade,
compressed particulate structure (CW, CWB, C, P) and rolled newspapers
had high CO emissions because their structure inhibited air flow to the
combustion zone. Wood, treated lumber, and coal had the lowest CO emissions
as these fuels would shrink and crack when burned, permitting sufficient
air to reach the burning fuel. Results from duplicate tests for each
75
-------
fuel suggest that air flow through the stove is also a factor affecting
CO emissions, with reduced air flow leading to increased emissions. CO
emissions were significantly high and of concern for all fuels.
8. S02 emission levels generally could be related to fuel sulfur content,
with higher fuel sulfur content causing higher S02 emissions. S02 emis-
sions were at levels of environmental concern only for P, BC, and CWB.
9. NO emissions were controlled by fuel nitrogen content and combustion air
/\
flow rate. High nitrogen content fuels (P and BC) had highest NOX emis-
sions. Increased air flow through the stove also led to increased NO
A
emissions. NO levels were generally low and as a result were not as
much of a concern as other pollutants.
10. Organic emission levels were comparable for all fuels except peat and
newspaper logs, which had high levels of organics in the flue gas effluent
stream. Organic emissions were affected by fuel consumption rate, fuel
structure, and amount of air through the stove. Higher fuel consumption
sometimes led to increased organics. Lowering air flow through the stove
increased organic emissions. Newspaper logs had high organic emissions
because of their physical structure, which inhibited air from reaching
the combustion zone leading to increased pyrolysis products.
11. PNA formation was affected by combustion air flow, firebox temperature,
and fuel structure. Composite structured fuels had higher PNA formation
except for newspaper logs, which, in contrast to high total organic
emissions, had very low emissions of heavier PNAs. It was concluded that
during the tests with newspaper logs firebox temperatures were too low
for extensive cyclization reactions leading to PNA formation. Also flow
rates through the stove were high for newspaper logs decreasing the
pyrolysis products residence time in the combustion zone and hence inhibit-
ing PNA formation. Other composite fuels had relatively high PNA produc-
tion rates. This is attributable to their structure, which limits the
availability of air during combustion and creates starved air conditions
favorable to PNA production. Tests with treated wood also had relatively
76
-------
high levels of PNAs in the flue gas effluent stream, attributable to low
air flow through the stove during these tests. Wood and coal had similar
PNA emissions, with coal emitting less PNAs than wood. PNA emissions
from coal could possibly be pyrolysis products from the coal itself.
12. Bioassays on organic extracts from one wood test and one coal test demon-
strated the presence of both mutagens and promutagens in the sample
extracts. Organics from coal combustion were about twice as mutagenic as
those from wood combustion on a mutagenicity per unit mass of fuel con-
sumed basis.
77
-------
REFERENCES
1. Trefil, J. Sawing Away at OPEC. Science 80, Vol. 1, No. 5, 1980. pp.
16-18.
2. Cooper, J. A., and J. G. Watson. Portland Aerosol Characterization
Study. Oregon Department of Environmental Quality, 1979.
3. Sanborn, Cedric, R., Richard L. Poirot, Gregory A. Heil, and Michael A.
Blanchet. Preliminary Analysis of the Ambient Impacts of Residential
Woodburning in Waterbury, Vermont. In: Proceedings of the Wood Heating
Seminar, February 21-24, 1981, New Orleans, LA. Wood Heating Alliance,
1981. pp. 231-241.
4. Imhoff, Robert E., Justice A. Manning, William M. Cook, and Timothy L.
Hayes. Preliminary Report on a Study of the Ambient Impact of Residen-
tial Wood Combustion in Petersville, Alabama. In: Residential Solid
Fuels—Impacts and Solutions. John A. Cooper and Dorothy Maleck, eds.
Oregon Graduate Center, 1982. pp. 520-538.
5. Carlson, James H. Residential Wood Combustion in Missoula, Montana: An
Overview of Its Air Pollution Contributions, Health Effects, and Proposed
Regulatory Solutions. In: Residential Solid Fuels—Impacts and Solu-
tions. John A. Cooper and Dorothy Maleck, eds. Oregon Graduate Center,
1982. pp. 539-550.
6. DeCesar, Richard T. The Quantitative Impact of Residential Wood Combus-
tion and Other Vegetative Burning Sources on the Air Quality in Medford,
Oregon. In: Residential Solid Fuels-Impacts and Solutions. John A.
Cooper and Dorothy Maleck, eds. Oregon Graduate Center, 1982. pp.
551-565.
7. United States Department of Energy. Health Effects of Residential Wood
Combustion: Survey of Knowledge and Research, DOE/EV-0114, September
1980.
8. DeAngelis, D. G., D. S. Ruffin, J. A. Peters, and R. B. Reznik. Source
Assessment: Residential Combustion of Wood. Prepared by Monsanto
Research Corporation for EPA, EPA-600/2-80-042b, NTIS No. PB 81-136160,
March 1980.
9. DeAngelis, D. G., and R. B. Reznik. Source Assessment: Coal Fired
Residential Equipment Field Tests. Prepared by Monsanto Research Corpor-
ation for EPA, EPA-600/2-78-004o, NTIS No. PB 283699, June 1978.
10. Milliken, John 0. Airborne Emissions from Wood Combustion. Presented at
the Wood Energy Institute, Wood Heating Seminar IV, Portland, Oregon,
March 22-24, 1979.
78
-------
REFERENCES (Continued)
11. Butcher, Samuel S. , and Michael J. Ellenbecker. Particulate Emission
Factors for Small Wood and Coal Stoves. In: Residential Solid Fuels--
Impacts and Solutions. John A. Cooper and Dorothy Maleck, eds. Oregon
Graduate Center, 1982.
12. Cooke, Marcus W., John M. Allen, and Robert E. Hall. Characterization of
Emissions from Residential Wood Combustion Sources. In: Residential
Solid Fuels—Impacts and Solutions. John A. Cooper and Dorothy Maleck,
eds. Oregon Graduate Center, 1982. pp. 134-163.
13. Hubble, B. R., J. R. Stetter, E. Gebert, J. B. L Harkness, and R. 0.
Flotard. Experimental Measurements of Emissions from Residential Coal-
Burning Stoves. In: Residential Solid Fuels—Impacts and Solutions.
John A. Cooper and Dorothy Maleck, eds. Oregon Graduate Center, 1982.
pp. 79-138.
14. Barnett, Stockton G., and Damian Shea. Effects of Woodstove Design and
Operation on Condensable Particulate Emissions. In: Residential Solid
Fuels-Impacts and Solutions. John A. Cooper and Dorothy Maleck, eds.
Oregon Graduate Center, 1982. pp. 227-266.
15. Milliken, John 0. Airborne Emissions from Wood Combustion. Presented at
the Wood Energy Institute, Wood Heating Seminar IV, Portland, Oregon,
March 22-24, 1979.
16.
17.
Dipple, Anthony. Polynuclear Aromatic Carcinogens. In: Chemical Car-
cinogens. Charles F. Searle, ed. ACS, Washington, DC, 1976. pp. 245-314
Peters, James A. POM Emissions from Residential Woodburning: An Environ-
mental Assessment. In: Residential Solid Fuels—Impacts and Solutions.
John A. Cooper and Dorothy Maleck, eds. Oregon Graduate Center, 1982.
pp. 267-288.
18. Murphy, Dennis J., Roy M. Buchan, and Douglas G. Fox. Ambient Particu-
late and Benzo(a)pyrene Concentrations from Residential Wood Combustion
in a Mountain Resort Community. In: Residential Solid Fuels—Impacts
and Solutions. John A. Cooper and Dorothy Maleck, eds. Oregon Graduate
Center, 1982. pp. 495-505.
19. Moschandreas, D. J., J. Zabransky, Jr., and H. E. Rector. Residential
Indoor Air Quality and Wood Combustion. In: Proceedings of the Wood
Heating Seminar, February 22-24, 1981, New Orleans, LA. Wood Heating
Alliance, 1981.
20. National Institute for Occupational Safety and Health. Criteria for a
Recommended Standard for Occupational Exposure to CO. Available from
National Technical Information Service, PB 212 629, 1972.
79
-------
REFERENCES (continued)
21. Wainwright, Phyllis. Wood Energy, the North Carolina Effort. In:
Proceedings of Wood Heating Seminars 1980/1981. Wood Heating Alliance,
1981.
22. National Institute for Occupational Safety and Health. Criteria for a
Recommended Standard for Occupational Exposure to Oxides of Nitrogen
(Nitrogen Dioxide and Nitric Oxide). U.S. Department of Health, Educa-
tion, and Welfare. NIOSH 76-149, March 1976.
23. U.S. Environmental Protection Agency. Air Quality Criteria for Oxides of
Nitrogen. Environmental Criteria and Assessment Office. Research Triangle
Park, NC, EPA 600/8-82-026f, December 1982.
24. National Research Council, National Academy of Sciences. Sulfur Oxides.
Available from Office of Publications, National Academy of Science, 2101
Constitution Avenue, NW, Washington, DC 20408.
25. DeAngelis, D. G., D. S. Ruffin, and R. B. Reznik. Preliminary Character-
ization of Emissions from Wood-Fired Residential Combustion Equipment.
EPA-600/7-80-040, NTIS No. PB 80 182066, U.S. Environmental Protection
Agency, 1980.
26. Allen, J. M., W. H. Piispanen, and M. Cooke. Study of the Effectiveness
of a Catalytic Combustion Device on a Wood Burning Appliance, EPA-600/7-
84-046, PB 84 171545, March 1984.
27. Gangwal, S. K., and D. G. Nichols. Chemical Characterization of Tar from
Fixed-Bed Gasification of Eastern and Western Coals. Proceedings of the
20th Hanford Life Sciences Symposium, Rich!and, WA, October 1980.
28. Gangwal, S. K., et al. Pollutants from Synthetic Fuels Production:
Sampling and Analysis Methods for Coal Gasification. U.S. Environmental
Protection Agency, EPA-600/7-79-201. NTIS No. PB 80-104656, 1979.
29. Gangwal, S. K., et al. Evaluation of Relative Environmental Hazards froni
a Coal Gasifier. U.S. Environmental Protection Agency. EPA-600/7-81-100.
NTIS No. PB 81-217648, 1981.
30. Grob, K., and G. Brob. Splitless Injection in Glass Capillary Gas Chrom-
atography. J. Chromatogr. Sc., Vol. 7, 1969. p. 584.
31. Lee, M. L., and B. W. Wright. Capillary Column Chromatography of Poly-
cyclic Aromatic Compounds: A Review. J. Chromatogr. Sc., Vol. 18, 1980.
pp. 345-358.
32. Smith, E. M., and P. L. Levins. Sensitized Fluorescence for the Detection
of Polycyclic Aromatic Hydrocarbons. EPA-600/7-78-182, NTIS PB 287-181,
Arthur D. Little, Inc., Cambridge, Mass., September 1978.
80
-------
REFERENCES (Continued)
33. Merrill, R. G., R. E. Lure, and L D. Johnson. A Spot Test for Polycyclic
Aromatic Hydrocarbons. Validation Studies. In: Chemical Analysis and
Biological Fate: Polynuclear Aromatic Hydrocarbons. M. Cooke, and A. J.
Dennis, eds. Battelle Press, Columbus, OH, 1981.
34. Smith. T. R. Evaluation of Sensitized Fluorescence for Polynuclear
Aromatic Hydrocarbon Detection. EPA-600/7-79-207, NTIS PB 80-108 475,
1979.
35. Smith, E. M., and P. L. Levins. Sensitized Fluorescence Detection of
PAH. In: Polynuclear Aromatic Hydrocarbons: Chemistry and Biological
Effects. A. Bjorseth and A. Dennis, eds. Battelle Press, Columbus OH,
1980. pp. 973-982.
36. Custin, H. W., and D. J. Murphy. Air Quality Emissions Inventory of the
Gore Valley, Vail, Colorado. Town of Vail, Community Development Series.
1976.
37. Surprenant, N. R., P. Huing, R. Li, K. T. McGregor, W. Piispanen, and S.
M. Sandberg. Emissions Assessment of Conventional Stationary Combustion
System: Vol. IV. Commercial/Institutional Combustion Sources. EPA-600/
7-81-003b, NTIS No. PB 81-145187, Table 16 (p. 43), January 1981.
38. Hughes, Thomas W., and Darryl G. DeAngelis. Emissions from Coal-Fired
Residential Combustion Equipment. In: Residential Solid Fuels—Impacts
and Solutions. John A. Cooper and Dorothy Maleck, eds. Oregon Graduate
Center, 1982. pp. 333-348.
39. Jansson, B., G. Sundstrom. Formation of Polychlorinated Dibenzo-p-Dioxins
During Combustion of Chlorophenol Formulations, The Science of the Total
Environment. Vol. 10, 1978, pp. 209-217.
40. Ames, B. N., J. McCann, and E. Yamesaki. Methods for Detecting Carcino-
gens and Mutagens with Salmonella Mammalian Microsome Test, Mutation
Res., Vol. 31, 1975, pp. 364-374.
81
-------
APPENDIX
Final Report
Ames Testing of Stove Combustion Products
82
-------
RTI/1914/39/0001F
Date: February 22, 1982
FINAL REPORT
AMES TESTING OF STOVE COMBUSTION PRODUCTS
Prepared for:
Robert S. Truesdale
Process and Chemical Engineering
Division
Research Triangle Institute
Post Office Box 12194
Research Triangle Park, NC 27709
Prepared by:
Robert S. DeWoskin, M.S.
Toxicologist
Life Sciences and Toxicology
Division
Thomas J. Hugfies, M.!
Senior Genetic Toxicologist
Life Sciences and Toxicology
Division
Debra M. Simmons, B.S.
Microbiologist
Life Sciences and Toxicology
Division
83
RESEARCH TRIANGLE PARK, NORTH CAROLINA 277
-------
CONTENTS
Figures. gtj|
Tables 86^
1.0. Introduction g^r |
2.0 Summary go?"!
3.0 Laboratory Facilities and Quality Control "91-1
3.1 Ames Mutagenesis Laboratories 91 \
3.2 Quality Control 92*5
• . . ••»
4.0 Experimental Procedure 95 j
4.1 Sample Preparation 95 1
4.2 Protocol for Testing ~ 9S *
4.2.1 Toxicity Testing, Plate Incorporation Method 97 ;
4.2.2 Mutagenicity Testing, Plate Incorporation Method . . . 97 J
4.2.3 Positive Mutagen Control Testing, Plate Incorporation
Method 97j
4.2.4 Sterility Testing, Plate Incorporation Method 100^
4.3 Formulation of NADPH Generating System 100
4.4 Salmonella Strain Validation 100;
4.5 Data Handling and Presentation 104
4.6 Data Analysis and Interpretation 104
4.6.1 Toxicity Assay/Data Analysis and Interpretation. . . . 104 :
4.6.2 Mutagenicity Testing/Data Analysis and Interpretation. 105 ,
5.0 Results and Discussion A?7
5.1 Toxicity Pre-Screen 107 _
5.2 Mutagenicity Testing 107
6.0 References , 11°
Appendix
A. Raw Ames/Salmonella Data
B. Stove Combustion Poster Abstracts 132
84
-------
FIGURES
Number Page
1 Ames protocol. ..................... 16
85
-------
TABLES
Number Page
1 General Project Information 3
2 Quality Control Information 7
3 Sample Receipt and Storage .... 10
4 Standard Protocol for Toxicity Determinations in Ames/
Salmonella Bioassay 12
5 Standard Protocol for Mutagenesis Determinations in
Ames/Salmonella Bioassay 13
6 Ames Activation System 15
7 lERL/Level I Ames Assay Evaluation Criteria 20
8 Mutagenicity of 110201 and 11007 TA98 (Detects Frame-
shift Mutagens), Determined at Nontoxic Doses 22
86
-------
1.0 Introduction
The Ames/Salmonella mutagenicity assay measures the mutagenic potential
of chemical compounds (Ames et al., 1975). The molecular basis of this
assay is the interaction of a chemical mutagen with the DNA of a Salmonella
bacterium carrying a mutation in a gene whose normal function is to allow
the de novo synthesis of the essential amino acid histidine. These mutant
bacteria are unable to grow in a medium deficient in this amino acid.
However, when a chemical mutagen interacts with the DNA of this bacterium
in such a way as to introduce a second mutation in the histidine operon
which corrects the original mutation, the bacterium can then survive in a
histidine deficient medium, giving rise to a visible colony on histidine
deficient agar. The frequency of mutational reversion is proportional to
the concentration and potency of the chemical mutagen. Therefore, the
number of colonies produced after addition of mutagen is directly propor-
tional to its mutagenic potential (McCann et al., 1975). This test is
normally conducted with five tester strains of Salmonella bacteria expressing
frameshift (TA98, TA1538, TA1537) and base-pair substitution (TA100, TA1535)
mutations in the histidine operon. The combustion samples in this study
were only tested with TA98 and TA100. Since many substances are not active
mutagens unless metabolically transformed (promutagens), the in vitro assay
has been modified to provide this transformation function by adding a rat
liver microsomal fraction (S9) which is rich in mixed-function oxidases
(P450 and P448), the enzyme systems which are mainly responsible for bio-
transformation of promutagens. The relationship between the potency of a
chemical as a mutagen to its potency as a carcinogen has been the subject
of considerable controversy and study with regard to the utility of the
bacterial assay as a predictor of carcinogenic potential (Rinkus and Legator,
1979). However, this assay has been shown to predict known carcinogens as
mutagens with an accuracy approaching 90% (McCann et al., 1975; Brusick,
1979).
87
-------
This study tested organic compound extracts from the combustion products
of a residential wood stove for mutagenie activity in the Ames Salmonella/
mutagenicity assay. General information about the study is shown in Table
1. The samples were supplied by the Process and Chemical Engineering
Division, Research Triangle Institute and consisted of a Soxhlet (SOX)
extract in dichloromethane and an aqueous extract (AQ) for each of two
fuels (wood and coal). The SOX and AQ extracts were generated from different
stages of the sampling train, but were combined as a composite sample for
testing.
These composite samples were tested for toxicity to determine the dose
levels for the mutgenicity assays. It should be noted that the mutagenicity
protocols of Ames et al. (1975) and IERL-RTP Level I Environmental Assessment
Manual Brusick and Young, 1980) specify testing with five Salmonella strains
(TA98, TA100, TA1535, TA1537 and TA1538). Because of the limited sample
amount and minor emphasis on the bioanalysis of the stove combustion products,
only strains TA98 and TA100 were utilized.
-------
TABLE 1. GENERAL PROJECT INFORMATION
Title of Project
Ames Mutagenicity Testing of Stove Combustion Products
Objective of Project
To assess the mutagenic potential of extracts of the organic compounds
from the combustion of wood and coal in a residential type stove.
Sponsor
Process and Chemical Engineering Division, Research Triangle Institute,
Research Triangle Park, NC 27709.
Project Number
RTI Task No. 47U-1914-39
Test Substance Identification
Test Substance
Wood - DCM Extract
Wood - Aqueous Extract
Coal - DCM Extract
Coal - Aqueous Extract
Sample No.
110201 KD
110201 pH Extract
11007 Sox
11007 pH Extract
State/Purity/Stability/Solvent
Solid/unknown/light-heat sensitive/
DMSO
Solid/unknown/light-heat sensitive/
DMSO
Solid/unknown/light-heat sensitive/
DMSO
Solid/unknown/light-heat sensitive/
DMSO
Storage Location of Results
Original Data: RTI Archives
Final Report: RTI Archives
Storage Location of Samples
Returned to Dr. Robert Truesdale, Process and Chemical Engineering
Division, Research Triangle Institute.
Dates
Samples Received: 01/12/82
Testing Initiated: 01/18/82
Testing Completed: 01/26/82
Preliminary Results Reported: 02/05/82
(via telephone)
Final Report Completed: 02/22/82
89
-------
2.0 Summary
Sample 110201 yielded 56 mg of extracted solid; 42 rag from the soxhlet
extract and 14 mg from the aqueous extract. Sample 11007 yielded 121 mg of
extracted solid; 120 mg from the soxhlet extract and 0.6 mg from the aqueous
extract.
Sample 110201 and sample 11007 demonstrated mutagenic activity with
Salmonella tester strains TA98 and TA100. The MEC value for both samples
was 50 yg/plate with TA98 and 500 jjg/plate for TA100. This result suggest
the presence of frameshift and base pair substitution mutagens. Based upon
the MEC values the samples were categorized as highly mutagenic with TA98
and moderately mutagenic with TA100. Both samples demonstrated an increase
in mutagenic activity with the addition of S9 with strains TA98 and TA100.
Therefore, both samples contain direct-acting mutagens and promutagens.
The presence of promutagens in organic compound extracts from residential
stove combustion products has been previously reported (Austin et al.,
1982, Appendix A).
Sample 11007 was more active than sample 110201 for both strains, with
or without the addition of S9. This comparison was based upon the slope of
the dose-response curves in units of revertants/pg of sample. A more
relevant comparison might be based on revertants/Joule of heat generated.
90
-------
3.0 Laboratory Facilities and Quality Control
3.1 Ames Mutagenesis Laboratories
The bacteriological section of Research Triangle Institute's In Vitro
Environmental Toxicology Laboratory is completely equipped for large-scale
Ames/Salmonella testing, as well as research and development studies. This
facility is equipped with automated plate pouring and colony counting
apparatus and large-capacity plate incubation equipment for high-volume
testing. The four mutagenesis laboratories are equipped with three Baker
NCB-6 laminar flow carcinogen hoods and one Lab Con exhaust hood. The
Class II Type-B (Baker NCB-6) hoods and the exhaust hood are vented through
charcoal filters on the roof of the testing facility. Plate incubation and
counting equipment are vented for personnel protection. The incubators
used in the evaluation of hazardous materials are vented to the roof before
the doors of the incubators are opened. All laboratory personnel are
appropriately garbed with laboratory clothing, safety apparatus and two
pairs of gloves while working with hazardous materials. Gloves and labora-
tory waste are double-bagged (plastic bags) and stored in 55 gallon drums
until disposal. Drums are disposed of at a certified burial site. An OSHA
approved, negative pressure, Toxic Substances Laboratory is employed for
the handling of mutagenic, carcinogenic and radioactive materials. A
central bacteriological media preparation and glassware cleaning facility
. o
is contained adjacent to one of the mutagenesis laboratories. Two 80 ft
walk-in constant temperature incubators are used for large-scale bacterio-
logical cell preparations. The Ames mutagenesis laboratories also include
A
adequate refrigeration and freezing facilities (two 90 ft walk-in cold
rooms; Kelvinator Ultracold Freezer -70°C, two Liquid Nitrogen Freezers)
for separate storage of media, test substances (including known carcinogens),
and Salmonella strain storage. The laboratories are subject to timer
controlled ultra-violet sterilization daily. Each laboratory is equipped
91
-------
with adequate storage and disposal facilities. All wastes are incinerated
(2,500°F) at Duke University Medical Center, an approved carcinogen waste
disposal site. Room for bousing animals associated with the proposed work
is available in the RTI Animal Research Facility. The Animal Research
Facilities have received full accreditation from AALAC, and are located in
a separate (10,000 square foot) building adjacent to the laboratory building.
The Ames facility has computerized the data collection and handling
system in order to subject test results to rigorous statistical assessment.
The data is entered at RTI and analyzed by the TUCC (Triangle University
Computation Center) IBM computer. Analysis of results and interpretation
of the data are then performed by the project leader.
3.2 Quality Control
Quality control is an important part of the Ames assay. Tester strains
are checked for the proper characteristics, positive controls are prepared
fresh for each assay and the equipment is inspected before use. The S9
activation preparation is tested for activation potential with known
promutagens and is compared with RTI's historical data base. The spontane-
ous revertant rate for Salmonella tester strains should be within the
recommended range (deSerres and Shelby, 1979). The quality control informa-
tion, for this assay, along with a listing of equipment and the sources of
chemical controls, are given in Table 2.
92
-------
TABLE 2. QUALITY CONTROL INFORMATION
Sponsor:
Process and Chemical Engineering Division, Research Triangle Institute,
Research Triangle Park, NC.
Project No.
RTI Task No. 47U-1914-39
Testing Facility
Research Triangle Institute, Post Office Box 12194, Research Triangle
Park, NC 27709.
Test Substance
Organic compound extracts (RTI No. 110201 and 11007)
Source: Combustion products from residential stove; wood or coal
was utilized as fuel
Physical State: solid residue
Purity: unknown
Composition: unknown
Stability: assume light and heat sensitivity
Storage conditions: 4°C in closed glass vials protected from light
Bioassay
Ames/Salmonella/mammalian microsome mutagenicity assay (Ames et al.,
1975).
Controls
Solvent: Dimethylsulfoxide (Fisher Scientific)
Positive: TA98: +S9 - 2-Aminoanthracene (IITRI)
-S9 - 2-Nitrofluorene (IITRI)
TA100: +S9 - 2-Aminoanthracene (IITRI)
-S9 - Sodium azide (IITRI)
Activation Mixtures
Source: Fischer 344 male rats, 200 grams, induced with Aroclor 1254
at 500 mg/kg i.p. per Ames et al. (1975)
Lot No.: RLI002
QC Assay Date: 5-14-81
93
-------
TABLE 2. (continued)
Strain Marker Verification
Strain Spont. Rev. Range rfa Mutation
TA98 15-40 +
IA100 150-220 +
Equipment
Description
Laminar flow hoods
Freezer
Incubator
Colony Counter
R-factor
UV
Sensitivity
Manufacturer
Baker NCB-6
Kelvinator Ultracold
Fisher Model 307
Artek Systems Corp.
Date last checked
Feb., 1982
Feb., 1982
Feb., 1982
Feb., 1982
94
-------
4.0 Experimental Procedure
This section describes the sample preparation and the procedure for
the Ames Salmonella/mammalian microsome mutagenesis assay including the
procedure for analyzing, interpreting and presenting the data.
A.I Sample Preparation
Organic compound extracts from residential stove combustion products
were received from PCED in clear glass sample vials at room temperature.
The sampling train for each test condition generated two extracts; a Soxhlet
extract of the particulate matter in dichloromethane (DCM) and an aqueous
extract from an impinger. Both extracts were transferred to preweighed
scintillation vials, the solvent was removed by evaporation with a stream
of nitrogen, residues were weighed, and resuspended in dimethylsulfoxide
(DMSO) for the Ames test.
The residue from extracts of sample 110201 weighed 56 mg: 42 mg from
the soxhlet extraction and 14 mg from the aqueous extraction. These two
residues were combined and serially diluted with DMSO to dose levels of:
1,000, 500, 100, 50 and 10 |jg of extract per plate. The residue from
extracts of sample 11007 weighed 121 mg: 120 mg from the soxhlet extraction
and only 0.6 mg from the aqueous extraction. These two residues were also
combined and serially diluted with DMSO to dose levels of 1,000, 500, 100,
50 and 10 pg of extract per plate.
Prepared samples were protected from light and stored at 4°C prior to
testing. Table 3 lists the sample receipt and storage information for
these samples.
4.2 Protocol for Testing
The procedures for handling the strains and preparing media components
were those of Ames et al. (1975). The S9 microsomal preparation was obtained
from Fischer 344 male rats injected with Aroclor 1254; protein was measured
by the method of Lowry et al. (1951).
95
-------
TABLE 3. SAMPLE RECEIPT AND STORAGE
Sample Receipt;
Sponsor Dr* R°bert Truesdale, Process and Chemical Engineering Div., RTI
Sponsor Code 110201. 10007 _
Method of Fabrication or Source Residential stove _
Date Received 1-12-82 _
Storage Location Ames lab freezer (A°C), Rm 215, Bldg. 3. RTF. NC _
RTI Code 110201, 11007 _
Sample Description;
11007 Sox (13.2 ml): AQ (pH extract 4.9 ml)
Identity 11Q20I Sox (KD 10 ml): AO fpH extract 5.2 ml) _
Lot No. ~
State/Color/Purity s°lid residue/clear to dark yellow/unknown
Density unknown
Stability assume light and heat sensitive
Solubility >1^ in dimethylsulf oxide
*Safety Data potential mutagens
Test Performed:
Bioassay/Date toxicity prescreen 1-18-82; mutagenicity 1-22-82
Date Sample Prepared 1-18-82
Storage Location of Prepared Sample Ames lab (No. 215) feezer A°C
Length of Time in Storage Prior to Assaying Tox-1 day: Mut-4 days
96
-------
The standard assay was divided into four parts as follows.
4.2.1 Toxicity Testing, Plate Incorporation Method (see Table 4)
Approximately 500 cells per dish were plated on nonselective media
(histidine-positive overlay). Tests were performed with and without addition
of Aroclor-induced S9 at five dose levels. Due to a limited amount of
sample some doses were tested without replicate plates. The viability
ratio (VR) was calculated as the fraction of surviving colonies with sample
to surviving colonies without sample. A viability ratio of less than one
indicated toxicity of the sample compound; a value of one or greater indi-
cated no toxicity. A dose that produced a VR of less than .5 was considered
a toxic dose. All toxicity determinations were performed in duplicate
except where sample amount was limited. Bioassay quality control was
accomplished with solvent controls, positive controls and strain controls,
with and without microsomal S9.
4.2.2 Mutagenicity Testing, Plate Incorporation Method (see Table 5)
o
Cells were plated at 10 cells/plate with selective media (histidine-
negative overlays). All mutagenic determinations were performed in tripli-
cate with and without S9 addition.
4.2.3 Positive Mutagen Control Testing, Plate Incorporation Method
o
Approximately 10 cells were plated in each dish on histidine-negative
overlays. Known mutagens were tested to assure that the strains were
active and the S9 preparation was activating promutagens to the desired
levels. If known positive controls did not demonstrate proper mutagenic
activity, the test components (cultures and/or S9) were rejected. Control
compounds were:
Strain Without S9 With S9
TA 98 2-Nitrofluorene 2-Anthramine
10, 5 pg/plate 5, 1 pg/plate
TA 100 Sodium Azide 2-Anthramine
5, 1 Mg/plate 5, 1 pg/plate
Positive control tests were performed in triplicate.
97
-------
TABLE 4. STANDARD PROTOCOL FOR TOXICITY DETERMINATIONS IN AMES/SALMONELLA BIOASSAY
Ub Code ID
«nd
Tube No.
1-1
1-2
1-J
1-*
1-5
1-6
1-1
1-8
0.1N
0.1
0.1
0.1
O.lJ
without
•ctlvitlon
fraction
0.1N
0.1
0.1
0.1
neMtlv*
'controls
0.1'
Lath done In triplicate.
In 0.15 M XC1.
-------
TABLE 5. STANDARD PROTOCOL FOR MUTAGENESIS DETERMINATIONS IN AMES/SALMONELLA BIOASSAY
Added to 1.0 ml »«nr, hlitldlne '(-). before iHyerlnR onto plites*
IO
Ub Cadi ID
• nd
Tube No.*
2-1
2-2
2-J
2-4
2-5
2-6
2-7
2-1
2-9
2-!0
2-11
2-12
2-11
2-1*
2-15
Siople In Vehicle
•1
.1 (1000 UO/pUtc
.1 (500 UK) /plate
.1 (250 ugl/plate
.1 (100 un)/pl«te
.1 (10 ngUpl.te
.1 (1000 ui)/pl«te
.1 (SOO pg)/pl«te
.1 (2SO M)/pUtc
.1 (100 un)/pl«te
.1 (10 u«>/Pl*te
0
0
0
0
0
S-9
Mlcrosomei
per ml of S-9 Ml«
0.1
0.1
0.1
0.1
0.1
0
0
0
0
0
0
0.1
0
0.1
0
S-9 Mix In
UjO ml
0.5
0.5
0.5
0.5
0.5
0
0
0
0
0
0
0.5
0
0.5
0
Act Iviit Ion
Prep. Medlun
Phosphate Buffer
Ml
0
0
0
0
0
0.5
0.5
0.5
0.5
0.5
0.5
0
0.}
0
0
Vehicle
•I
0
0
0
0
0
0
0
0
0
d
0.1
0.1
0
0
0
Bccterli
•1 (10*)
celli/Bl)
O.rNwtth
0.1
0.1
0.1
Induced
kctt*«tlon
fraction
0.1-'
0.^
O.I
O.I
O.I
without
•etlvatlon
fraction
D.!"
O.I-X
0.1 |nt|t(tiv*
0.1 \cont roll
0.1
O.J
*tich done In triplicate.
b|n 0.15 H KC1.
-------
4-2.4 Sterility TestineT Plate Incorporation Method
Sterility tests were conducted with histidine-positive overlay plates,
with the identical components employed in the tests. Components tested were
sample(s), positive controls, solvent(s), water, direct-acting mix, S9 mix,
and agar plates. Sterility testing was performed in duplicate.
4>3 Formulation of NADPH Generating System (see Table 6)
Components M.W. Concentration/ml S9 Mix
l- NADP 765.4 4 praoles
2- G6P 282 5 pmoles
3- M8C12 203.3 8 pinoles
*• KC1 74.5 33 (jmoles
5. Sodium phosphate 100 (Jmoles
buffer pH 7.4
6. Organ homogenate 100 (Jliters
(S9 fraction)
Make-up following stock solutions:
0.2 M NADP Millipore filter
0-2 M G6P Millipore filter
0.4 M MgCl2 Sterilize by autoclaving
1.65 M KC1 Sterilize by autoclaving
0.2 M Na Phosphate buffer pH 7.4 Sterilize by autoclaving
A. 2.0 ml and 0.2 ml aliquots of 0.2 M NADP is dispensed into sterile
vials and kept frozen at -70°C until used.
B. 2.5 ml and 0.5 ml aliquots of 0.2 M G6P is dispensed into sterile
vials and kept frozen at -70°C until used.
C. S9 mix is then produced following the formulas in Table 5 S9 is added to the
S9 mix in a 1:10 part ratio.
Toxicity and Mutagenicity Preparation (see Figure 1)
1. Add 1.8 ml of the appropriate suspension to a dram vial and keep on
ice.
2. Add 0.3 ml of the test material to the dram vial. Shake on vortex.
3. Dispense 0.7 ml/2.0 ml overlay, vortex and plate.
4.4 Salmonella Strain Validation (from Ames' Methods Paper, 1975)
The cells are taken from oxoid broth inoculated and grown overnight.
100
-------
TABLE 6, AMES ACTIVATION SYSTEM
1.
2.
1.
».
5.
6.
7.
Component
HADP*
Clurcmt-6-
phoaphate
dlbnalc
Sodlua
phosphate
HgCL,
KCL
Hooogenatc
H20
H HW Supplier
0.2 765.4 1CN
0.2 2R2 Sld*<
0.2 142 Dl Slftoa
1)8 Hnno
0.4 20).] S1«M
1.65 74.5 SlRU
-
Stork Preparation
76.6 «/500 *n HjO
28.2 d/500 B! HjO
2 liter dlba*lcc
300 til »onod
40.7 d/500 •> HjO
61.5 d/500 .1 H20
Standiird KC1 9,000 « £
aupematant
VoluM of Stnck added/ml
of final HU
20 1
25 1
500 I
20 1
20 i
100 1
315 1
Per 100 ml
2.0
2.)
so.o
2.0
2.0
10.0
H.S
final Concentration
of Component /•! In HU
« ante*
) anlea
100 Bolea
1 «olea
3] M>lea
ApproKlaatelf 25 m*. of
fre«h tlaau* equlvalant
*AH mi* l» 1-7 with the •dditlon of bacterl*.
Sodiua pho»ph«tt U used •• NH mlm flu* bacteria, pH 7.4.
C2 liter dlbaclc - 56.8 KK/2000 ml.
d)00 «l mono • 8.28 g»/300 ml.
Coaponenta 1 and 2 are prepared in sterile dlatllled water and aialntalned at -20*C.
Component* 3, 4, and 5 are prepared In dlatllled water, (terllUed, and Maintained at 4*C.
Component* 6 la prepared and atored at -80*C until used,
Coaponenta 1-5 combined • core reaction •
-------
Figure 1. Ames protocol.
Induced
Konlnduced
S-9 in KC1
NADP + C-6-P
Bacterial Suspension
Buffer*
Bacterial
Suspension
Sa=ple In Vehicle
•* Soft Agar
45'C
With histidine for toxiclty and
sterility testing, Without histidine
for mutagenicity and positive crntrol
testing
Incubate 48 hr at 37*C
and Count
102
-------
1. Histidine Requirement
Base Layer Plates, total needed = 30
5 strains - .1 ml/plate
Overlays - H+ (1A) and H~ (IB)
A. Cells should grow with addition of histidine, heavy cloudy
background.
B. Spontaneous Background - must be within acceptable limits.
2. Crystal Violet/rfa Character
Base Layer Plates, total needed = 15
5 strains - .1 ml/plate
Overlays - H
10 |Jl of crystal violet on filter paper disc
Place disc + cv on plate after pouring overlay with bacteria
A clear zone of inhibition should be present around the disc
indicating that rfa cell wall mutation is present.
3. Ampicillin Resistant R Factor (Qualitative)
Base Layer Plates, total needed = 15
5 strains - .1 ml/plate
H overlays
Pour overlays and cells; using q-tip, streak fresh ampicillin
down middle of plate
(Ampicillin - 100 pi of 8 mg/ml in .02 N NaOH)
Strains 98 and 100 - No zone of inhibition around ampicillin
streak
Strains 37, 35, 38 - Zone of inhibition around ampicillin
streak
A. UV Sensitivity/AuvrB Deletion
Base Layer Plates, total needed = 15
5 strains - .1 ml/plate
Overlays - H+
Pour plates, irradiate plates: with screen with T - cut out
Strains 98 and 100 - 8 sec
Strains 35, 37, 38 - 6 sec
Cells should be present only on non-irradiated area.
103
-------
5- Ampicillin Resistant R. Factor (Quantitative Assay)
Base Layer Plates, total needed = 15
5 strains - 0.1 ml/plate after dilution in NB
5/100 + 1/100 -» 10/100 f 20/100
Overlays - H+
Place 0.1 ml of diluted cells and 100 ml ampicillin in H+
overaly
Strains 98 and 100 - full growth
Strains 35, 36, and 38 - no growth
6. Control for Ampicillin
Same as #5, but no ampicillin is used.
7. Sterility
Plates - 3 ea.ch
Overlays - H_ - 3 each
Overlays - H~ - 3 each
4.5 Date Handling and Presentation
Raw data for both the toxicity and mutagehicity assays were analyzed
by computer. The results are presented in tabular and graphical form in
Appendix A. The information presented includes the counts from each plate,
the average count, the standard deviation, a quality control check for an
acceptable spontaneous revertant count, and either a viability ration (VR)
for the toxicity assay or a mutagenic ratio (MR) for the mutagenicity
assay. The MEC (minimum effective concentration) is also indicated by an
asterisk if the sample was mutagenic.
4.6 Data Analysis and Interpretation
A.6.1 Toxicity Assay/Data Analysis and Interpretation
The results from the toxicity assay for the combustion samples deter-
mined the doses utilized in the mutagenicity assay. The expression utilized
to assess a toxic effect is called the viability ration (VR) and was calcu-
lated as follows:
„.,.,. r, . no. of surviving colonies in sample
Viability Ratio = no. of surviving colonies in solvent
A toxicity assay dose that resulted in a 50% reduction in bacterial growth
(i.e., a viability ratio of 0.50) became the highest dose tested in the
104
-------
mutagenicity assay. If the viability ratio is below 0.50 the sample at
that dose was considered toxic. An asterisk will appear in the table
column headed "TOX" for the lowest concentration that demonstrated a toxic
effect. A viability ratio greater than 1.0 indicated that the sample at
that dose was nontoxic. The interpretation of the VR considers the values
for testing both with and without S9 activation, and whether the toxic
effect was dose responsive.
4.6.2 Mutagenicity Testing/Data Analysis an Interpretation
The mutagenicity of a test compound can be quantified as the mutagenic
ratio (MR). The MR is calculated as follows:
n . - T> ^ - no. of revertant colonies in sample
Mutagenic Ratio = r : : : r*-—-
6 no. of revertant colonies in solvent
The sample was considered mutagenic with strains TA98 and TA100 if the MR
was 2.0 or greater and the response increased at three increasing doses.
The sample can be further categorized as having high or low mutagenicity
based upon the smallest amount of sample required to produce a MR of 2.0 or
greater. This lowest dose is called the Minimum Effective Concentration
(MEC) and was identified in the tables by an asterisk under the column with
the "MEC" heading. Table 7 list the criteria used for categorizing a
sample's mutagenic activity based upon the MEC. This study used the criteria
in the column labeled "Solids."
The slope of the linear portion of the dose-response curve is also
utilized for categorizing a sample's mutagenic activity. The slope is
expressed in revertants per microgram of sample. Because the Salmonella
tester strains respond differently to mutagens, the slopes of the dose-
response curves should only be compared within and not between strains.
105
-------
TABLE 7. IERL/LEVEL I AMES ASSAY EVALUATION CRITERIA
Muta genie
Activity
High
Moderate
Low
Not detectable
Criteria Used
Solids
(MEC in fjg/plate)
<50
.50-500
500-5000
5000
Liquids
(MEC in pi/plate)
<2
2-20
20-200
>200
Organic Extracts
(MEC in pi/plate)
<2
2-20
20-200
>200
NOTE: These cagegories at these doses are based upon a mutagenic ratio
of 2.0 for strains TA98 and TA100 and 3.0 for strains TA1535,
TA1537, and TA1538.
106
-------
5.0 Results and Discussion
Table 8 summarizes the toxicity and mutagenicity results. The raw
data and statistical analysis can be found in Appendix A. In the front of
the Appendix is a legend key for the computer printout headings.
5.1 Toxicity Pre-Screen
The toxicity pre-screen was performed on strain TA98 with and without
S9 activiation. The results are listed in Tables Al-2 and Graphs Al-4 in
Appendix A. The asterisk in the "TOX" column denotes a viability ratio of
less than 50% survival. Sample 110201 was more toxic to TA98 than sample
11007 and was toxic at the 500 |Jg/plate dose with or without the S9 homogen-
ate. Sample 11007 was toxic at the 1000 pg/plate dose in the absence of S9
and demonstrated a reduced toxic activity in the presence of S9. Because
of a limited amount of sample only one plate was poured with sample 110201
and duplicate plates were poured for sample 11007. A toxicity prescreen
was not performed on strain TA100. The analysis of the toxicity results
suggested the following doses for mutagenicity testing: 1,000, 500, 100,
50, 10 (Jg/plate. Adequate amounts of sample were available for triplicate
plates in the mutagenicity test for these samples. Triplicate plates are
recommended for quality control purposes.
5.2 Mutagenicity Testing
The results from the mutagenicity testing are listed in Tables AA-5
and Graphs A5-8 for TA98 and Tables A5-6 and Graphs A9-12 for TA100. Under
the heading, "Activation," a batch code (i.e., RLI002) will appear if the
S9 activation system was added to the test mixtures. Under the column
headed, "MEC," an asterisk appears by the minimum dose which produced a
mutagenic ratio (MR) of 2.0 or greater. The MEC value can then be catego-
rized to rank the sample for mutagenic activity.
Sample 110201 demonstrated mutagenic activity with strains TA98 and
TA100 both with and without the addition of a S9 rat liver activation
system. The MEC necessary to produce a MR of 2.0 or greater was 50
107
-------
TABLE 8. MUTAGENICITY OF 110201 AND 11007 TA98 (DETECTS
FRAMESHIFT MUTAGENS), DETERMINED AT NONTOXIC
DOSES
Sample Code
110201
11007
S9
Activation
+
•*•
MEC
(pg/plate)
50
50
50
50
Mutagenicity
Ranking
H
H
H
H
Slope
(rev/|jg)
0.47
0.74
0.67
0.97
TA100 (DETECTS BASE PAIR SUBSTITUTION MUTAGENS)
S9
Sample Code ' Activation
110201
11007
MEC
(pg/plate)
500
500
500
500
Mutagenicity
Ranking
M
M
M
M
Slope
(rev/pg)
0.79
1.28
0.96
1.74
3H = high mutagenic activity; M = moderate mutagenic activity.
108
-------
with TA98 and 500 pg/plate with TA100. The MRs at these doses were greater
than 2.0. Based upon the criteria listed in Table 7, sample 110201 was
categorized as highly (H) mutagenie with strain TA98 (frameshift mutagens)
and moderately (M) mutagenic with strain TA100 (base pair substitution
mutagens).
The mutagenic activity of the samples with or without the addition of
a S9 rat liver activation system was analyzed by comparing the slopes from
the linear portion of the sample's dose-response curves. The mutagenic
activity demonstrated in the absence of S9 was interpreted as activity from
direct-acting mutagens. Direct-acting mutagens do not require the addition
of a S9 activation system to produce a mutation in the histidine operon.
An increase in the mutagenic activity with the addition of the S9 activation
system was interpreted as increased activity from promutagens.
Sample 110201 clearly demonstrated mutagenic activity with both strains
TA98 and TA100 in the absence of S9 and therefore contains direct-acting
mutagens. This sample demonstrated a greater mutagenic activity with the
addition of S9. The greater activity of sample 110201 with the addition of
S9 suggests that this sample also contained promutagens. Promutagens have
been previously detected in stove combustion products by Austin et al.
(Appendix B).
Sample 11007 was also categorized as highly mutagenic with strain TA98
(MEC of 50 pg/plate) and moderately mutagenic with strain TA100 (MEC of 500
|jg/plate) with or without the addition of S9. As in sample 110201, greater
activity occurred with the addition of S9 suggesting the presence of both
direct-acting mutagens and promutagens. Comparing the two samples for
mutagenic activity by comparing the slopes of the dose-response curve
demonstrates that sample 11007 was more mutagenic than sample 110201 under
all conditions tested.
109
-------
6.0 References
Ames, B. N. , McCann, J. and E. Yamasaki. Methods for Detecting Carcinogens
and Mutagens with Salmonella Mammalian Microsome Test, Mutation Res.
31, 374-364 (1975).
Brusick, D. J. "Bacterial Mutagenesis and Its Role in the Identification
of Potential Animal Carcinogens", Carcinogens: Identification and
Mechanics of Action, (A. C. Griffin and C. R. Shaw, eds.), Raven
Press, New York, pp. 95-107 (1979).
Brusick, D. J. , Young, R. R. , Hutchinson, C., Dilkas, A. G. and T. A. Geyo.
IERL-RTP Procedures Manual: Level I Environmental Assessment Biological
Test. EPA Contract No. 68-02-2681, Environmental Protection Agency
(ORD), Washington, D. C. (1980).
deSerres, F. and M. Shelby. The Salmonella Mutagenicity Assay: Recom-
mendations. Science 203, 563-565 (1979).
Lowry. 0. H. , Rosebrough, N. J. , Farr, A. L. and R. J. Randall. Protein
Measurement with the Folin Phenol Reagent. J. Biol. Chem. 193,
265-275 (1951).
McCann, J. , Choi, E. , Yamasaki, E. and B. N. Ames. Detection of Carcino-
gens or Mutagens in the Salmonella/Microsome Test: Assay of 300
Chemicals, Proc. Nat. Acad. Sex. 72, 5135-5139 (1975).
Rinkus, R. and N. Legator. Chemical Characterization of 465 Known or
Suspected Carcinogens and Their Correlation with Mutagenic Activity in
the Salmonella typhimurium System, Cancer. Res. 39, 3289-3318 (1979).
110
-------
APPENDIX A
111
-------
COMPUTER PRINTOUT LEGEND
EXP Date = Date of the experiment.
Test Type = Pour or preincubation test (left blank or 01).
Dilution = Milligrams of S9 to milligrams of S9 mix.
Mixture = Microliters of S9 mix per plate.
Chemical = Code for chemical tested.
Strain = Salmonella strain used.
Plated = Microliters of S9 mix plated for sterility plate.
Technician = Code for the technician performing the work.
Contract No. = Contract number of project.
Activation = Code number for the batch of S9 mix. Blank for the testing
without S9.
Batch = Code for the strain batch.
Colonies - Number of colonies on S9 sterility plate. Generally blank if
S9 was sterile.
S054 = Solvent (dimethylsulfoxide).
P = Positive controls.
P-01 = Sodium azide.
P-02 = 9-aminoacridine.
P-03 = 2-nitrofluorene.
P-04 = 2-aminoanthracene.
STD DEV = Standard deviation of duplicate plate counts.
Q/C = Quality control column.
A = acceptable i.e., within the ranges specified for each project.
NA = not acceptable.
VR = Viability ratio, the ratio of surviving colonies with sample to the
surviving colonies of the solvent control.
112
-------
COMPUTER PRINTOUT LEGEND (continued)
TOX = Toxicity - an asterisk appears in this column for the lowest dose
with
a viability ratio of 0.5 or less.
MR = Mutagenic ratio - the ratio of the number of revertants with sample
over
the number of revertants with the solvent control.
MEC = Minimum effective concentration, the lowest dose with a MR of 2.0 or
greater.
The units for the dose are in microliters/plate for liquid samples and
in micrograms/plate for solids. Please not that for the mutagenicity data
the positive controls and the samples are in microgram units while the
solvent control is in microliter units.
113
-------
TABLE Al. TOXICITY OF 110201 - TA98
T 0
EXP
TEST
X I
DATE
TYP
C I
: o
E:
T
lie
01
Y
82
CHEMIC
STRAIN
DI LUTION= :
M IX
S OL
POS
S;i
T 0
EXP
TES
TURE =
VENT/
ITIVE
51
X I
DATE
T TYP
c
L
1
5
10
50
100
C I
: o
E :
D I LUTION=
M I X
S OL
POs
S i
TURE =
VFNT/
ITIVE
51
50
CSE
EVE
0.0
0.0
0.0
0.0
0.0
0.0
T
IIP
01
i :
0
L
0
0
0
0
0
0
Y
82
PL
TE
PLATE
COUNT
138
572
552
199
82
0
ATED
CHNI
A:
BG
CHEMIC
STRAIN1
10
DOSE
L
1
5
10
50
100
EVE
0.0
0.0
0.0
0.0
0.0
0.0
L
0
0
0
0
0
0
PL
TE
PLATE
COUKT
57°
572
552
1 c 9
62
0
A TEC
CHM
A:
BG
AL:
: T
110201
A98
CONTR
ACT NO
.: 2075
ACTIVATION:
BATCH
: A172
= COLONIES=
CIA
PL
N= RDE
ATE B:
COUNT BG
AL:
: T
= 1
387
110201
i98
CO
PLATE C:
COUNT BG
CONTR
ACTIV
BATCH
COLON
AVER
AGE STD DEV
COUNT COUNT
112
572
552
199
82
0
ACT NO
ATION:
: A172
1ES =
.5 36.1
.0
.0
.0
.0
.0
.: 2075
RLI002
C I A N = R D E
PL
ATE P:
COUNT PG
592
PLATE C:
COUNT PG
589
AVER
AGE STD DEV
COUNT COUNT
586
572
552
199
82
0
.7 6.8
.0
.0
.0
.0
.0
VB
1.0
1.1
1.3
1.2
0.2
0.0
VR
1.0
1.0
0.9
0.9
0.1
0.0
TOX
TOX
-------
TABLE A2. TOXICITY OF 11007 - TA98
TOXICITY
EXP DATE: oiib82
TEST TYPE: 01
DILUTION :
MIXTURE
SOLVENT/ DOSE
POSITIVE LEVEL
SI 54 0.00
10.00
50.00
100.00
500.00
1000.00
T^O X I C I T Y
EXP DATE : 011882
TEST TYPC: 01
DILUTION': i: 10
M IXTURE= 500
SOLVENT/ DOSF
POSITIVE LEVEL
Si 54 O.DO
10.00
bO.OO
100.00
500.00
1 GOO. 00
CONTRAC
CHEMICAL:
STRAIN
PL
TL
PLATE
COUNT
43fl
4°1
473
555
307
131
ATED
CHNI
A:
BG
11007
: TA98
ACT
IVAT
BATCH:
= COLONIE
T
I
A
S
NO
ON:
172
—
•
• •
2075
CIAN= RDE
PL
ATE B:
COUNT PG
387
518
458
461
320
141
PLATE
COUNT
c:
BG
CONTRAC
CHEMIC
ST
PL
TE
PLATL
COUNT
57"
6?5
657
72F
441
387
PAIN
ATED
CHM
A:
L-G
AL:
: T
= i
110C7
A98
00
ACT
IVAT
BATCH:
COL
OME
AVERAGE
COUNT
T
I
A
s
412
504
465
507
313
136
NO
ON:
172
-
.5
.5
.5
.0
.5
.0
•
• •
RL
STD
DEV
COUNT
36
19
10
65
9
7
2075
1002
.1
.1
.6
.1
.2
.1
CIAM= R[
-------
GRAPH Al. TOXICITY OF 110201 - TA98
-S9
S=SOLVENT P=POSITIVE N=ANTHR APINJE y/C S9 * = TE:ST COMPOUND
CONTRACT=2075 CHEK1CAL=11020I S7RAIM=98 DATE=011882 ACTIVATN=
550.0
PLOT OF AVGCOUNT*DOSE
SYMBOL IS VALUE OF SPINO
250.0 *
i
i
200.0 *
i
i
i
150.0 •»•
i
t
t
100.0 *
'
I
I
50.0 *
i
t
i
0.0 •>
0.0
0.3
-4-
3.0
30.0
300.0
.--.*-
3000.0
DOSE I-N MICRCGPAMS
116
-------
GRAPH A2. TOXIC1TY OF 110201 - TA98 +S9
S=SOLVENT P=POSITIVE N=ANTHRAHINE U/0 S9 *=TEST COMPOUND
CHEM 1C AL = 1 1 020 I STRAIN = 98 CATE = 011882 ACT I V ATN=RL 1002
PLOT OF AVGCOUNT'DOSE SYMBOL IS VALUE OF SPIND
i
900.0 «•
I
1
I
I
800.0 *
i
t
t
i
700.0 *
i
i
! S
«
t
500.0 *
i
«
t
i
^tOO.D *
i
t
300.0 *
i
;
i
t
200.0 *
i
«
i
t
100.0 *
I
•
t
I
D.O *
0.0 0.3 3.0 ' 30.0 300.0 3000.0
POSE *N MICROGRAMS
117
-------
GRAPH A3« TOXICITY OF 11007 - 7A98 -S9
S=SOLVENT P=POSITIVE N=ANTHRAMINE U/0 S9 *=TEST COMPOUND
CONTRACT=2075 CHEMICAL=11007 STRAIN=98 DATE=011882 ACTIVATN=
PLOT OF AVGCOUNT*DOSE SYMBOL IS VALUE OF SPIND
550*0 *
i
500.0
^50.0 *
i
t
»
400*0 •»
i
30C.O
350.0 *
i
i
i
*
i
t
i
250. P •»
i
t
i
200.0 *
i
i
i
150.0 *
i
i
i
100.0 *
I
I
I
50.0 *
i
i
0.0
._4_.
0.0
>-—4 —.
30.0
.._ ^ __,
300.0
0.3
3.0
3000.0
DOSE IN MJCROGRAMS
118
-------
GRAPH A4. TOXICITY OF 11007 - TAgs -fS?
S=SOLVENT P=POSITIVE N=ANTHRAKINE U/0 S9 *=TEST COMPOUND
CONTRACT=2075 CHE«ICAL=11 007 STRAIN=98 DATE = 011882 ACT I V ATN = RL 1002
PLOT OF AVGCOUNT*DOSE SYKPCL IS VALUE OF SPIMD
i
900.0 *
I
800.0 «•
t
t
i
t
700.0 *
i
t
! * *
t
600.0 *
i <•
i
i
500.0* * *
i
i
i •
400.0 *
i
i
i
300.0 *
i
t
i
t
200.0 *
i
i
i
100.0 «
I
I
I
I
0.0 *
0.0 0.3 3.0. 30.0 300.0 3000.0
DOSE 7>N M1CRCGRAMS
119
-------
TABLE A3 MUTAGEMCITY OF 110201 - TA98
1 U T A G E N I C
:XP DATE: 012282
'EST TYPE: 01
)ILUTICN= :
1IXTLRE =
JOLVENT/ DOSE
'OSITIVE LEVEL
SI 54 O.OC
PI 04 5.00
P2 03 5.00
P3 03 10. OC
1 0 . C C
50.00
100.00
5CO.CC
10CO.CC
2000.00
1UTAGENIC
:XP CATE: C12282
TEST TYPE: ci
!ILUTXCN= l: 10
IIXTL'RE = 500
I T Y
CHEMIC
STRAIN
PLATED
TECHM
PLATE A:
CCUNT BG
JC
30
402
795
18
56
94
184
247
0
I T Y
CKE^IC
STRAIN
PLATED
TECHM
AL:
: TA
s
CIAN
PLA
110201
98
= PDE
TE B :
COUNT 9G
3
8
1
2
AL:
: TA
= 10
CIAN
20
21
40
fe
27
60
77
80
46
0
11C2CI
58
0
= ROE
cc
AC
BA
CC
PLATE
CCUNT
19
3?
368
771
19
58
108
200
238
0
CO
AC
BA
NTRA
TIVA
TCH:
LCM
c:
BG
NTkA
TIV*
TCH:
COLONI
CT NO
TION:
A212
ES =
•
• •
AVERAGE
COUNT
19
33
370
808
21
5P
93
1P8
242
0
CT NO
TION:
A212
ES =
.3
.3
.0
.0
.3
.0
.0
.0
.7
.0
•
• •
RL
2075
STC
DEV
G/C M
R
CCUN'T
0
4
31
44
4
2
15
10
4
0
2075
IC02
.6
.9
.0
.9
.9
.0
.5
.6
.9
.0
A 1
1
A 19
A 41
1
3
4
c
12
0
.0
.7
.2
.9
.1
.0
.8
.7
.6
.0
MEC
SOLVENT/
'OSI
SI
PI
P2
TI VE
54
04
04
CCSE-
LEVEL
o.co
l.OC
5.00
10.00
50. OC
1CO.CC
500.00
1000.00
2CCO.OC
PLATE A:
C CLNT BG
22
178
1361
46
85
147
2bl
242
281
PLATE 5 :
COUNT P G
24
229
1326
43
92
146
261
306
267
PLATE C:
CCUNT EG
20
221
1400
33
95
143
278
321
277
AVERAGE
CCLNT
25.3
209.3
1363.0
40.7
90.7
14E.3
263.3
323.0
275.0
STC CEV
CCUNT
7.6
27.4
26. P
6.8
5.1
2.1
13.7
16.1
7.2
G/C PR
t 1.0
A 8.2
A 53.9
1.6
3.6
5.7
10.4
12.8
10.=
MEC
120
-------
TABLE A4 PUTAGEMCITY OF 11007 - TA9E
MUTAGENIC
EXF CATE: C12262
TEST TYPE: 01
DILUTICN= :
MIXTL'RE =
T Y
CHEKICAL: 31007
STRAIN: TA58
PLATED=
TECHNICIAN: RDE
CONTRACT NC.:
ACTIVATION:
BATCH: A212
COLCMIFS=
075
SOLVEN
POSITI
SI 54
PI 04
P2 03
P3 03
K U T
EXF CA
TEST T
DILUTI
MIXTUP
T/ CCSE PLATE A:
VE LE
0
5
5
10
10
50
100
500
10CO
20CO
AGE
TE : Cl
YPE: o
CN! = 1
E= 500
VEL CCLNT BG
.Ct
.CC
.00
.00
.OC
.CC
.00
.00
.OC
.OC
fi I C
2262
1
: 10
15
30
402
795
27
60
£7
206
11C
176
I T Y
CHEMIC
STRAIN
PLATED
TECHM
PLA
TE B:
COUNT BG
3
8
1
3
1
AL:
: TA
= 10
CIAN
20
21
40
58
26
£5
C9
C8
57
65
11007
56
0
= F.DE
PLATE C:
COLNT BG
19
35
366
771
31
68
89
293
93
165
CONTRA
AC TIVA
BATCH:
COLOM
AVERAGE
COUNT
15.3
32.3
370.0
808.0
26.0
65.7
95.0
302.3
100.0
166.7
CT NO.:
TION: RL
A212
ES =
STC
DEV
C/C P
R
COUNT
0
4
31
44
2
4
12
8
6
6
2075
IOC2
.6
.5
.0
.5
.6
.5
.2
.1
.9
.4
A 1
1
A 1«
A 41
1
3
4
15
5
8
.0
.7
.2
.9
.5
.4
.c
.7
.2
.7
SCLVEN
POSI
SI
PI
P2
TI
54
04
04
T/ CCSE.
VE LEVEL
0.00
l.CC
5.00
10.00
50. OC
1 C 0 . 0 G
500.00
1000.00
2CCO.CC
PLATE A :
C CLNT BG
22
176
1361
27
62
7C
245
2£.4
247
PLATE B:
COUNT f- G
24
225
1326
28
c c
£7
271
351
24C
PLATE C:
CCLNT EC
20
221
1400
35
46
5?
260
330
AVEFAGE
COLNT
25.3
205.3
1363.0
34.7
54.3
72. 0
258.7
348.3
242.5
STC CEV
CCUNT
7.6
27.4
26. r.
5.9
e.c
14.1
13.1
17.2
4.9
0/C MR
A 1.0
A 8.2
A 53.5
1.4
2.1
2.P
10.2
13. P
q .6
121
-------
TABLE A5. KUTAGENICITY OF 110201 - TA100
r A G E N I
;ATE: 012282
TYPE: 01
riON= :
^RE =
:NT/ DOSE
'IVE LEVEL
>4 0.00
14 5.00
11 1.00
11 5.00
10.00
50.00
100.00
500.00
moo. oo
r A G r N i
LATE: 012282
TYPE: 01
FION= i: 10
[|RE= 5CO
[NT/ DOSE
FIVE LEVEL
i
h o.oo
[4 l.CO
U 5.00
10.00
50.00
100. CO
500.00-
1000.00
CITY
CHEMIC
STRAIN
AL: IIOPOI
: TAIOO
PLATED=
TECHNI
PLATE A:
COUNT bG
198
302
547
974
216
266
26*
436
514
CITY
CHEKIC
S T R # I \
PLATE:
TECHM
PLATE A:
COUNT PG
"l5F
345
1300
226
352
400
51?
454
CIAN= RDE
PLATE B:
COUNT PG
193
289
516
963
187
2R2
277
474
506
AL: 110201
: TAIOO
= 100
CIA\= FDE
PLATE E:
COUNT PG
168
374
1 ?Q6
217
319
365
497
461
CONTRAC
ACTIVAT
BATCH:
T NO.:
ION:
A212
2075
COLONIES=
PLATE C:
COUNT BG
225
272
578
931
217
268
266
435
414
CONTPAC
ACTIVAT
EATCH:
COLONJE
PLATE C:
COUNT PG
167
349
1290
216
307
410
480
428
AVERAGE
COUNT
205.3
287.7
547.0
956.0
206.7
272.0
277.0
448.3
478.0
T NO.:
ICK: PL
A212
s =
AVERAGE
COUNT
164.3
356.0
1296.0
220.3
326.0
391.7
498.7
447.7
STD DEV
COUNT
17.2
15.0
31.0
22.3
17.0
8.7
11.0
22.2
55.6
2075
1002
STD DfV
COUNT
5.5
15.7
5.3
6.7
23.3
23.6
19.6
17.4
G/C MR
A
A
A
A
1.0
1.4
2.7
4.7
1.0
1.3
1.3
2.2
2.3
G/C MR
1.0
2.2
7.9
1.3
2.0
2.4
3.0
2.7
MEC
KEC
122
-------
TABLE A6. MUTAGENICITY OF 11007 - TA100
MU T A G
EXP DATE:
TEST TYPE:
DILUTIONS
MIXTURE=
SOLVENT/
POSITIVE
S 1 54
PI 04
P2 ni
P3 Dl
1
c,
10
2C
MU TAG
EXF DATE:
TEST TYPE. :
DILUTIONS
MIXTURES 5
E N I
012282
01
:
CITY
CHEMIC
STRAIN
PLATED
CONTRACT NO
AL: 11007
: TAIOO
—
ACTIVAT
BATCH:
COLONIE
ION:
A212
s=
.
•
2075
TECHNICIANS RDE
DOSE
LEVEL
0.00
5.00
1.00
5.00
10.00
50.00
00.00
00.00
00.00
oc.oo
r fy T
0122P2
01
i : 10
OC
PLATE A:
COUNT BG
199
302
547
974
201
237
335
667
227
270
CITY
CHEMC
S T F. A I N
PLATEP
TECHM
PLATE e:
COUNT BG
193
289
516
963
220
272
323
645
?89
2B1
AL: 11007
: TA1CO
= 100
CIATJ= RTE
PLATE C:
COUNT BG
225
272
576
931
203
227
330
691
322
2fi2
CONTPAC
ACTIVAT
BATCH:
COLGME
AVER
AGE
COUNT
205
287
547
956
208
245
330
667
279
277
T NO
ION:
A212
S =
.
•
.
.
.
•
•
.
.
•
•
3
7
0
0
0
3
3
7
3
7
•
•
RL
STD
DEV
Q/C MR
COUNT
17
15
31
22
10
23
7
23
48
6
2075
1002
.2
.0
.0
.3
.4
.6
.5
.0
.2
.7
A 1.
1.
A 2.
A 4.
1.
1.
1.
3.
1.
1.
0
4
7
7
0
2
6
3
4
4
SOLvrr.
POSITI
S 1 54
PI 04
P2 04
T/ DOSf
VL LEVEL
0.00
1.00
5.00
10.00
5 C . 0 0
100.00
toe. oo
1COO.OO
2f 00.00
PLAT£ A:
COUNT tG
15P
345
1300
159
214
277
568
fi6I
467
PLATE F:
COUNT FG
16P
374
1298
140
239
?74
569
P38
497
PLATE C:
COUNT EG
167
349
1290
166
233
279
656
763
491
AVERAGE
COUNT
164.3
356.0
1296.0
155.0
226.7
276.7
596.3
821.3
4R5.0
STD DEV
COUNT
5.5
15.7
5.3
13.5
13.1
2.5
51.7
52.0
15.9
C/C MR
A 1.0
2.2
A 7.9
O.1'
1.4
1.7
3.6
5.0
3.0
KEC
123
-------
GRAPH A5 MUTAGENICITY OF 110201 - TA98 S9
S=SOLVENT P=FCSITIVE N=ANThR*MINE W/0 S9 *=TFST COMPOUND
CCNTRACT = 2075 ChE K I CAL= 11 02 01 STRAIK=S6 DATE=C122f2 fiC7JVATN
PLOT OF AVGCOUNT*DOSE SYPBOL IS VALUE OF SPIND
500.0 •»
i
i
600.0 *
i
i
;
i
700.0 *
t
i
i
i
£CO.O •»
i
i
5 CO.C *
i
i
;
i
4 00.0 *
!
i
»
i
300. 0 *
i
i /
2CO.O * /
! X *
i
100.0 •» •/
i x
! X
! N
! S *
0.0 *
~~*~~~"*~~~~~™'™~i~ ~~ ~~ •••••••••«•••* — — — ••••••••„
0. C 0.3 2.0 30.0 300.0 3000
DCSE IK KICRCGFAKS
124
-------
GRAPH A6 MUTAGENICITY OF 110201 - TA98 S9
S=SCLVENT F = FCSITIVE N=ANThRAMIKE U/0 S9 *=TEST CCPPOUNC
CONTRACT=207£ CHEKICAL = 11020 I STRAIN=?P DATE=012?F2 ACT I VATN = RLIOC
PLOT OF AVGCCUNT-DOSE SYMBOL IS VALL'E OF SPIND
R
E
V
E
3
T
A
N
T
C
0
u
N
T
15
14
13
12
11
10
c
£
7
£
c
«
3
2
1
00.
CO.
CO.
00.
00.
CO.
CO.
00.
00.
CO.
CO.
80.
00.
00.
CO.
0.
0
0
0
0
0
c
c
0
0
0
0
0
0
0
0
0
1
4
;
! P
1
1
4
1
1
4
J
1
4
J
1
4
1
!
4
1
1
4
1
1
4
1
1
4
1
1
4
1
4
! » •
! X"
4 p X
1 ^ X^
; s
! S «
4
O.C 0.3 3.0 3C.O 300.0 3C
DOSE IN KICROGRAMS
125
-------
GRAPH A7 MUTAGENICITY OF 11007 - TA98 S9
S=SCLVENT P=POSITIVE N=AM HP AK UE WO SS «=TEST CCPFOUNC
CONTRACT=2075 CHEM 1C AL =1 10 07 STRAIN=56 DATE=C122f2 ACTIVATN=
PLCT CF AVGCCUNT*CCSE SYMBOL IS VALUE OF SPINO
f
soo.o
t
6CO.O *
;
i
i
i
700.0 •»
;
i
;
i
600.0 *
i
i
500.0 •»
i
i
•
i
ACO.O •»
!
;
i
3CO.O *
I
I
200.0 *
i
;
i
100.0 *
! S
0.0 *
0.0 0.3 3.0 30.0 300.0 3000
DOSE IN KICROGPAMS
126
-------
GRAPH A8 MUTAGENICITY OF 11007 - TA98 SS
S=SOLVENT P=POSITIVE N=ANTHRAHINE fc/0 S9 *=TEST COMPOUND
CCMTRACT = 2075 CHEKICAL=llC07 STRAIN=98 DATE = 012282 ACT I VATN = RLIC
FLCT OF AVGCCLNT*DCSE SYPECL IS VALLE CF SPINC
15
CO
1400
R
E
V
E
R
T
A
T
C
0
u
M
T
13
12
11
10
9
e
7
6
5
4
3
2
1
00
CO
CO
00
00
CO
CO
00
00
CO
CO
00
00
0
.0
.0
.0
.0
.0
.0
.0
.0
.C
.0
.0
.0
.0
.0
.0
.0
1
<*
1
1
*
! F
I
4-
1
'
1
4
1
I
!
1
4
1
I
*
1
*
1
1
4
!
1
4
1
1
4
1
1 J
4. '
\ /.
• s *
•» p
4
; */*
\ s *
4
0.0 0.3 3.0 30.0 300.0 3
HCSE IK K1CROGRAMS
127
-------
GRAPH A9. MU7AGENICITY OF 110201 - TAlOO -S9
S=SOLVENT P=POSITIVE N=ANTHRAMINE U/0 S9 *=TEST COMPOUND
CDNT*ACT=2075 CHEM 1C AL = 11 020 I STRAIN=100 DATE=012?fi2 ACTIVATN
PLOT OF AVGCOUMT*DOSE SYMBOL IS VALUE OF SPIND
i
i
1100. D *
1000. 0 •»
!
•
i
900.0 *
i
t
i
eoD.o «
700.0 *
i
£00.0 *
*
! F
t
500. D * ^
i *
/
t *
«OD.D * X
;
\
20D.O •» N /
j *'*
X
i
200.0 * 5 *
«
«
i
100.0 *
1
t
»
0.0 +
._ +_._«.._._._.^.. _--•---.-.*-------- ----4 ------------ 4- -----------+-
0.0 C.3 3*0 30.0 300.0 3000.0
COSE I.N MICRCGRArS
128
-------
GRAPH A10. MUTAGENICITY OF 110201 - TAlOO + S9
S=SOLVENT P=POSITIVE N=ANTHPAMINE U/0 S9 *=TEST COMPOUND
CDNTRACT = 2D75 CHE?1CAL = 11020 I STRAIN = 100 OATE = 012282 ACTIVATN = RLI 002
PLOT OF AVGCOUNT-DOSE SYMPOL IS VALUE OF SPIKC
1500.0 *
i
1400.0 «
i
i
1300.0 *-
t
1200.0 *
R
E t
V 1100.0 *
E !
R
1000.0 *
A I
N "500.0 *
T i
O
u
N
800.0 *
i
i
700.0
600.0 *
500.0 •»
i
400.0 *
i
i
300.0 *
x
X
/
X
200.0 *
! S
i
100.0 «
I
I
0.0 «•
.. ^ -.
0.0
..4..
0.3
3.0 30.0
HOSE TN MI CROCRAMS
129
»__ ««._,
300.0
.„.--+-
3000.0
-------
GRAPH All. MUTAGENICITY OF 11007 - TA 100 -S9
S=SOLVENT P=POSIT1VE N=ANTHRAMINE y/0 S9 *=TEST COMPOUND
CHEM I CAL=11 007 STRAIN=100 DATE=012282 ACTIVATN
PLOT OF AVGCOUNT*DOSE SYMBOL IS VALUE OF SPIND
1100*0 +
I
I
I
1000.0 *
500.0 *
i
t
600.0 *
t
i
i
700.0 +
i
i
fcOO.C «• '
I /
! P /
i
500.0 «•
I
I
4CC.O *
30C.C * M /
*
t
200.0 •» S
i
i
100.0 +
I
I
I
0.0 >
-.*.--.-.------*.------. -.--*------------*------------*-«--.-.- ----4-
0.0 0.3 3*0 30.0 300.0 3000.0
DOSE JN HIC
130
-------
GRAPH A12. KUTAGEMCITY OF 11007 - TA100 +S9
S=SOLVENT P=POSITIVE N=ANTHPAMINE W/0 S9 *=TEST COMPOUND
CO\TRACT=2075 CHEMC AL = 11 007 STRAIN=100 DATE=012282 ACT IVATN=RL1002
PLOT OF AVGCOUNT*DOSE SYMBOL IS VALUE OF SPIND
1500.0
1400.0
1300.0
1200.0
R
E 1100.0
V
E
R 1000.0
T
A 900.0
N
T
eoo.o
700.C
I
*
I
I
I
I
*
I
I
*
I
I
t
I
*
I
/
C 600.0 *
o '
u
N
T
500.0
4DO.O *
t
i
300.0 •»
200.0
100.0
0.0 «•
.-••«•»-
3000*0
0.0
0.3
3.0 30.0
DOSE IN MICROGRAPS
._..*...
300.0
131
-------
APPENDIX B
132
-------
UnifecJ Slates
Environmental Protection
Agency
Research and Development
SEPA 1982 Symposium on the
Application of Short-Term
Bioassays in the
Analysis of Complex
Environmental Mixtures
Registration and
Abstract Book
January 25-27, 1982
The Carolina Inn
Chapel Hill, North Carolina
Sponsored by the
U.S. Environmental
Protection Agency
Office of Research and
Development
Research Triangle Park,
North Carolina
133
-------
RESULTS OF HEALTH AND ECOLOGICAL EFFECTS BIOASSAYS FOR THIRTY-TWO
RESIDENTIAL WOOD COMBUSTION RESIDUE SAMPLES
Robert Young,1 Curt Hutchinson,2 D.R. Jagannath,1 David J. Brusick, * and
Ray Merrill3
Litton Bionetics, Inc./ Kensington, MD 20895, %nion Carbide Corp.,
Tarrytown, NY 10591, and Industrial Environmental Research Laboratory,
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
This report presents the results of U.S. Environmental Protection Agency
Level 1 environmental assessment bioassays on 32 Source Assessment Sampling
System train samples collected from 12 separate residential wood-fired
combustion source configurations (Auburn project). Three types of samples
were supplied for testing from each combustion configuration, namely:
(1) particulate catch extract, (2) combined organic module rinse and XAD-2
extract, and (3) combustion residue (bottom ash). The particulate catch
extract was a composite sample consisting of (1 ) the front half wash residue,
(2) the combined cyclone catch, and (3) the filter catch.
1. Ames Mutagenicity Assay
a. The samples of combustion residue (bottom ash) that were evaluated
did not show any mutagenic activity.
b. Making very general comparisons between burning- areas (fireplace and
baffled and nonbaffled stoves), there is an apparent difference between the
fireplace and the stoves. The mutagenicity of the filter particulate catch
extract and of the XAD-2 extract of the fireplace samples was consistently
lower than that of the same samples from the stoves. The significance of
this is difficult to interpret without more information on the chemistry of
these samples.
c. No consistent difference between oak or pine and green or seasoned
wood was observed in this study. Both produced roughly equivalent mutagenic
effects.
2. Chinese Hamster Ovary Cell Clonal Toxicity Assay
The particulate catch and XAD-2 extracts consistently showed high levels
of toxicity to Chinese hamster ovary cells in culture, regardless of burning
area, wood type, or moisture content. These effects appeared to support the
Salmonella results.
3. Rabbit Alveolar Macrophage Cytotoxicity Assay
This test was conducted on the bottom ash samples and produced results
showing low toxicity to rabbit alveolar macrophages.
4. Freshwater Invertebrate Test
The acute toxicities of two combustion residues were determined in the
ecological effects bioassay using the freshwater invertebrate Daphnia maana.
Based on the defined LC£Q ranges, both samples tested against Daphnia magna
were considered nontoxic.
This research was supported by U.S. Environmental Protection Agency Contract
No. 68-02-2681 to Litton Bionetics, Inc.
-------
MUTAGENICITY OF EMISSIONS FROM AN AIRTIGHT WOODSTOVE
Ann C. Austin, Robert E. Hall, Ray Merrill, and Joellen Lewtas
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Woodstoves, now widely used for residential heating purposes, emit relatively
high concentrations of particulate matter and organics. The objective of this
study was to determine the mutagenicity of the organics extracted from the
diluted particles emitted during wood combustion. The wood, seasoned pine
and oak, was burned in a Johnson Energy Converter Model J-1000 airtight
woodstove. The stove was operated with the following periods: a cold start
up, a constant burn, a fuel addition, and a burndown. The emissions from each
burning period were collected separately using a standard dilution tunnel
sampling system, and the organics were removed by Soxhlet extraction with
dichloromethane and prepared for bioassay in dimethylsulfoxide. The extracts
were bioassayed in the Ames Salmonella typhimurium plate incorporation assay
and were found to be mutagenic in Salmonella typhimurium TA98. Although there
was some direct-acting activity, the addition of an S9 metabolic activation
system increased the activity of each sample four- to twenty-one-fold. During
the cold start up period for pine and the cold start up and fuel addition
periods for oak, the organic emission rate (mg/m^) was significantly greater
than during the constant burn period. Similarly, the mutagenicity of the
organics (rev/pg) during these respective periods was also greater. However,
since the cold start up and fuel addition periods are relatively short
compared to the total burn, the time averaged mutagenicity over the burn of
all four periods closely resembles the mutagenicity of the constant burn
period. The pine exhibited a higher particulate and organic emission rate
than did the oak, and the organics emitted from the pine were somewhat more
mutagenic. The mutagenicity of the organic emissions from woodstove and
residential oil combustion were compared. Although the woodstove organics
were two to five times less mutagenic than the oil, the woodstoves exhibited
a much greater organic emission rate per joule of heat and the resultant
mutagenicity of the woodstove emissions per joule of heat was greater than
oil combustion.
135
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-84-094
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Characterization of Emissions from the Combustion
of Wood and Alternative Fuels in a Residential
Woods to ve
5. REPORT DATE
September 1984
6. PERFORMING ORGANIZATION CODE
'.AUTHORis)R>Sf Truesdale, K. L. Mack, J. B. White,
K. E. Leese, and J. G. Cleland
8. PERFORMING ORGANIZATION REPORT NO.
RTI/1914-39-01F
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Research Triangle Institute
P. O. Box 12194
Research Triangle Park, North Carolina 27709
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3170, Task 39
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 2/81 - 3/84
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES IERL-RTP project officer is Michael C. Osborne, Mail Drop 65,
919/541-4113.
16. ABSTRACT
repOrt gives results of a comparison of emissions from the combustion
of alternative fuels to those from wood in a residential woodstove, and of a study of
the effects of woodstove operating parameters on combustion emissions. Overall,
oak wood is the best fuel tested, considering both emissions and stove operation.
Compressed wood logs with binders and bituminous coal produce the highest emis-
sions of SO2, particulate, and NOx. Compressed wood logs without binders and
treated lumber produce the highest PAH emissions. Important parameters affecting
CO emission levels are fuel structure and, to a lesser degree, combustion air flow.
SO2 emission levels are related directly to fuel sulfur content. NOx emissions are
controlled by fuel nitrogen content and combustion air flow rate. Organic emissions
are affected by fuel consumption rate, fuel structure, and the amount of air through
the stove. Total discharge severities for PAHs measured during this study indicate
that PAHs are the pollutants of highest concern in the flue gas effluent stream. PAH
formation is affected by combustion air flow, firebox temperature, and fuel struc-
ture. Bioassay results indicate the presence of both mutagens and promutagens in
the organic extracts of flue gas samples from both wood and coal combustion tests.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Flue Gases
Combustion
Coal
Wood
Wood Products
Stoves
Bioassay
Mutagens
Aromatic Polycyclic
Hydrocarbons
Pollution Control
Stationary Sources
Woodstoves
13 B
21B
21D
11L
ISA
06A
06E
07C
13. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
141
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
136
------- |