EPA-600/R-96-146
December 1996
FINAL REPORT
ROANOKE WOODSTOVE
EMISSION TESTS
Prepared by:
M. Buckland
Acurex Environmental Corporation
4915 Prospectus Drive
P.O. Box 13109
Research Triangle Park, NC 27709
EPA Contract No. 68-DO-0141
TD 94-237
EPA Project Officer: Robert C. McCrillis
U.S. Environmental Protection Agency
National Risk Management Research Laboratory
Research Triangle Park, NC 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, D.C. 20460
-------
TECHNICAL REPORT DATA .• .... „ ,.V™, ,^7.-.7mi777.~.. "•
(Please read Instructions on the reverse before complei 111 Illl 11III1II 111 1111IJ] II111
1. REPORT NO, 2.
EPA-600/R-96-146
3. Ill llll.l lllll! Illlllllllll 111
V ... PB97-13.1387
4. TITLE AND SUBTITLE
Roanoke Woodstove Emission Tests
5. REPORT DATE
December 1996
6. PERFORMING ORGANIZATION CODE
7. AUTHOB(S)
M. Buckland
8. PERFORMING ORGANIZATION REPORT NO.
9, PERFORMING ORGANIZATION NAME AND ADDRESS
Acurex Environmental Corporation
P. 0. Box 13109
Research Triangle Park, North Carolina 27709
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-D0-0141, Task 94-237
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COWERED
Final report; 2/89-10/89
14. SPONSORING AGENCY CODE
EPA/600/13
16. supplementary NOTES APPCD project officer is Robert C. McCrillis, Mail Drop 61,
919/541-2733.
is. abstractrj-^g rep0rtf discusses' a project, part of the Integrated Air Cancer Project
(IACP) Roanoke study, that characterizes and quantifies emissions generated by-
burning authentic Roanoke cordwood. -The burning oce.ur.r_ed in a controlled labora-
tory setting using two woodstoves, each operated at two different burn ratesj-JThe
project goal was to collect organic and inorganic emissions produced by burning
Roanoke wood during high- and low-burn rate conditions similar to those in a home.
The two stoves, a LCPI 380/440 conventional and a LCPI 1988 EPA-certified Answer
low-emission model, were run at high- and low-burn rates simulating burn condi- ;
tions found in a typical home. Eight sampling runs were conducted consisting of du-
plicate runs at both burn rates of the two stoves. After sampling, the sampling me-
dia, filters, canisters, and raw data were distributed to various analytical groups
for analysis. The 380/440 stove generated higher levels of emissions than the Ans- .
wer stove because the latter incorporated secondary combustion technology. The
narrow burn rate range of the Answer stove and the scatter of all the data in general
make drawing definitive conclusions on trends difficult. It'appears that the conven-
tional stove showed a direct relationship with burn rate for volatile organic emis-
sions and an inverse relationship for extractable organic emissions.
17. KEY WORDS AND DOCUMENT ANALYSIS
a, DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Wood
Combustion
Stoves
Burning Rate
Emission
Pollution Control
Stationary Sources
Woodstoves
13 B
11L
2 IB
13 A
14G
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport;
Unclassified
21. NO. OF PAGES
41
20. SECURITY CLASS {This page}
Unclassified
22. PRICE
EPA Form 2220*1 (9-73)
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policyand
approved for publication. Mention of trade names
or commercial products does not constitute endorse
ment or recommendation for use.
t
ii
-------
FOREWORD
The U.S. Environmental Protection Agency-is charged by Congress with pro-
tecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and implement actions lead-
ing to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA1 s research
program is providing data and technical support for solving environmental pro-
blems today and building a science knowledge base necessary to manage our eco-
logical resources wisely, understand how pollutants affect our health, and pre-
vent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for
investigation of technological and management approaches for reducing risks
from threats to human health and the environment. The focus of the Laboratory's
research program is on methods for the prevention and control of pollution to air,
land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites and groundwater; and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze
development and implementation of innovative, cost-effective environmental
technologies; develop scientific and engineering information needed by EPA to
support regulatory and policy decisions; and provide technical support and infor-
mation transfer to ensure effective implementation of environmental regulations
and strategies.
This publication has been produced as part of the Laboratory's strategic long-
term research plan. It is published and made available by EPA's Office of Re-
search and Development to assist the user community and to link researchers
with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
PROTECTED UNDER INTERNATIONAL COPYRIGHT
{^tSwJtcotKwfohmation service
U.S. DEPARTMENT OF COMMERCE
•s 111
i * ¦ ,
Li
-------
ABSTRACT
As part of the Integrated Air Cancer Project (IACP) Roanoke study, this project
characterizes and quantifies emissions generated by burning authentic Roanoke
cordwood. The burning occurred in a controlled laboratory setting using two
woodstoves, each operated at two different burn rates. The project goal was to collect
organic and inorganic emissions produced by burning Roanoke wood during high and
low burn rate conditions similar to those in a home. The two stoves, a LOPI 380/440
conventional and a LOPI 1988 EPA-certified Answer low-emission model, were run at
high and low burn rates simulating burn conditions found in a typical home. Eight
sampling runs were conducted consisting of duplicate runs at both burn rates of the two
stoves. After sampling, the sampling media, filters, cartridges, canisters, and raw data
were distributed to various analytical laboratory groups for analysis. The 380/440 stove
generated higher levels of emissions than the Answer stove because the latter
incorporated secondary combustion technology. The narrow burn rate of the Answer
stove and the scatter of all the data in general make drawing definitive conclusions on
trends difficult. It appears that the conventional stove showed a direct relationship with
burn rate for volatile organic emissions and an inverse relationship for extractable
organic emissions. Historically, burn rate has been shown to be the major variable
affecting emission rates. Data are presented that may be used to calculate emission
factors for woodstove use during the IACP Roanoke oil furnace field study.
iv
-------
TABLE OF CONTENTS
Section Page
1.0 INTRODUCTION 1
2.0 EXPERIMENTAL APPROACH 2
2.1 Sampling 2
2.2 Test Apparatus 6
3.0 ANALYTICAL METHODS AND PROCEDURES 8
3.1 Aldehydes 8
3.2 Volatile Organic Compounds 8
3.3 Continuously Recorded Data 8
3.4 Total Capture 8
3.5 Elemental Analysis and Organic Carbon/Elemental Carbon
Determinations 8
3.6 Sample Extraction and Analysis 9
3.7 Total Chromatographable Organic Compounds Analysis 9
3.8 Gravimetric Analysis 9
3.9 Polynuclear Aromatic Hydrocarbons Analysis 9
3.10 Bioassay Analysis 10
4.0 QUALITY ASSURANCE 11
5.0 PRESENTATION OF RESULTS 13
6.0 DISCUSSION OF RESULTS 21
6.1 Sampling 21
6.2 Total Organic Mass Emitted 21
6.3 Polynuclear Aromatic Hydrocarbons 25
6.4 Aldehydes 25
6.5 Fine Particulate Matter 25
6.6 Total WSDSS Filter Capture 27
7.0 CONCLUSIONS 34
8.0 REFERENCES 35
IB';
V
-------
LIST OF TABLES
Table Page
1. Samples Collected and Data Generated 4
2. Sampling and Analysis Responsibilities 5
3. Sampling Schedule 6
4. Sampling Conditions 7
5. Data Completeness 12
6. Overall Emissions (Mg/kJ) 14
7. VOC Results
-------
SECTION 1.0
INTRODUCTION
In support of the Integrated Air Cancer Project (IACP) Roanoke Study, Acurex
Environmental Corporation was contracted to characterize and quantify emissions
generated by burning authentic Roanoke cordwood. The burning occurred in a
controlled laboratory setting using two woodstoves, which were each operated at two
different burn rates. The project goal was to collect organic and inorganic emissions
produced by burning Roanoke wood during high- and low-burn rate conditions similar to
1 I • I 1 I ¦ /% r-t 1 AAA li 4 A A" I I a A m A A A A Vm fh A
those in a home. The two stoves, a LOPI 380/440 conventional and a LOPI 1988 EPA-
certified Answer low-emission model, were run at high and low target bum rates
simulating burn conditions found in a typical home. Eight sampling runs were
conducted, four for each stove over a range of burn rates. The sampling media, filters,
cartridges, canisters, and raw data were distributed to various analytical laboratory
groups for analysis after sampling. This document represents a compilation of the work
of those researchers.
Acurex Environmental modified and operated the facilities for this study; conducted
all sampling activities; performed non-volatile organic compound, semivolatile organic
compound, and polyaromatic hydrocarbon (PAH) analyses; and coordinated the
results.
l
-------
SECTION 2.0
EXPERIMENTAL APPROACH
2.1 SAMPLING
Sampling data collected by the operator consisted of the load cell readings,
barometric pressure, and data pertinent to the various sampling trains. Barometric
pressure was recorded once per day. Load cell and sampling train data were recorded
every 30 minutes during sampling runs. Sampling interruptions for sampling media
changes and fuel additions were carefully recorded.
Sampling was conducted while burning a mixture of pine and apple wood
purchased in Roanoke and burned in the laboratory in Research Triangle Park, NC.
One run lasting approximately 4 h was performed per day. The bums and sampling
were performed in a portable building adjacent to the G-Wing High Bay at the
Environmental Research Center (ERC) in Research Triangle Park, NC. Figure 1
presents the overall layout of the test facility. Table 1 shows sample collection and
data generation. Table 2 details sampling and analysis responsibilities.
The test stove, mounted on a Toledo 8142, temperature-compensated electronic
scale with a 0,045-kg readability and 220 kg capacity, was vented through the roof via a
15-cm diameter stove pipe. The scale measured fuel additions and monitored short-
term fuel consumption. The exhaust pipe incorporated a sliding joint to isolate the
mass borne by the scale from the sampling apparatus. The stove pipe terminated 4.57
m (15 ft) above the scale platform. A thermocouple was mounted in this exhaust pipe
0.46 m (1.5 ft) below the exit to monitor stack temperature. Molar concentrations of
oxygen, carbon dioxide, and total hydrocarbons (THC) in the exhaust gases were
measured using continuous emission monitors (CEMs). The CEMs and thermocouple
were connected to a data logger that recorded data at five minute intervals and
transmitted the collected data once per day via modem to the Project Officer's
computer.
A secure platform was constructed above the portable building to support the
sampling apparatus, A tube axially inserted into the stack exit sent a portion of the
exhaust gas to a Woodstove Dilution Source Sampler (WSDSS). The WSDSS,
developed by the Air and Energy Engineering Research Laboratory (AEERL) of the
U.S. Environmental Protection Agency (EPA), dilutes the collected combustion gases
with clean air before sample collection. This dilution process cools the sample to
ambient temperature so that condensable gases (those analytes with low vapor
pressures at ambient temperature) will be filter collectible." The WSDSS was operated
by Acurex Environmental as per instruction received from AEERL personnel. The
WSDSS has been used extensively to collect emission samples from residential wood
and oil combustion devices.2 3i 4,5
The cooled gases were pulled through a Pallflex® Teflon®-coated quartz fiber filter
and an XAD-2 sorbent resin cartridge to collect non-volatile and semivolatile organic
compounds, respectively. Depending on test conditions, the filter could become fully
2
-------
-------
TABLE 1. SAMPLES COLLECTED AND DATA GENERATED
(See text for explanation of acronyms)
CEMs DNPH XAD-2 Filter Summa® Dichot
02
co2
Total Hydrocarbon (THC)
Temperature
Total Capture
Total Chromatographable
Organics (TCO)
Gravimetric (GRAV)
PAH
Volatile Organic Compounds
(VOC)
Aldehydes
Elemental Analysis
Organic/Inorganic
Carbon
x
x
x
x
x
x
x
x
x
x
x
x
loaded; a fresh filter would be installed, and testing resumed. This filter loading was
monitored by the pressure drop across the system.
Before sample collection, Pallflex® filters were desiccated and tared. XAD-2 resin
was cleaned and QC-checked using a modified version of AEERL Recommended
Operating Procedure (ROP) 40.6 The dry resin was packed into custom stainless steel
cartridges. These cartridges were capped and sealed in Teflon® bags before delivery to
the sampling site.
VOCs were collected in an evacuated stainless steel canister (SUMMA®) as a
side stream between the Pallflex® filter and the XAD-2 cartridge. VOC collection post-
filter ensures a particulate matter-free sample. A critical orifice at the inlet of the
canister controlled the flow so that a time averaged sample was collected. A dry gas
meter determined the total sample volume collected. These canisters were delivered to
Acurex Environmental by the Atmospheric Research Exposure Assessment Laboratory
(AREAL) ready for sample collection and were returned to AREAL for analysis.
Fine particulate samples (particles < 2.5 txm) were collected in parallel with the
Pallflex® filter and XAD-2 cartridge. Fine particulate sampling media consisted of
4
-------
TABLE 2. SAMPLING AND ANALYSIS RESPONSIBILITIES
Analyte
Acurex Environmental AREAL3 AEERLb
02
X
C02
X
THC
X
Temperature
X
Total Capture
X
TCO
X
GRAV
X
PAH
X
VOCs
X
Aldehydes
X
Elemental Analysis
X
Organic/Elemental Carbon
X
a Renamed National Exposure Research Laboratory in 1996.
b Renamed Air Pollution Prevention and Control Division of National Risk Management
Research Laboratory in 1996,
Teflon® and quartz filters in parallel. The quartz filters were ordered from Sunset
Laboratories and were ready for sample collection when delivered. The Teflon® filters
were provided by AREAL. The filters were tared by Acurex Environmental before
sample collection and reweighed after sample collection to determine particulate
capture. The quartz and Teflon® filters were delivered to AREAL. The quartz filters
were then sent to Sunset Laboratories in Oregon for organic and elemental carbon
analysis. Elemental analysis was conducted on the Teflon® filters at AREAL using
energy-dispersive X-ray fluorescence.
Aldehyde samples were collected on dinitrophenylhydrazine (DNPH) tubes.
Aldehydes react with the DNPH to provide non-volatile derivatives that are ready for
analysis by high-performance liquid chromatography (HPLC). Aldehyde sample
collection was also performed in parallel to the Pallflex® filter and XAD-2 cartridge, front
and back pairs. The two front/back pairs are typically operated with a factor of 2 flow
rate between them. Each front/back pair was connected serially. The DNPH tubes
were delivered to Acurex Environmental by AREAL ready for sample collection and
were returned to AREAL for analysis.
5
-------
2.2 TEST APPARATUS
Two stoves, typical of Roanoke, VA usage, were purchased for testing. These
were an LOP1 380/440 conventional stove and an LOPI Answer low-emission, non-
catalytic stove. The Answer model is capable of secondary combustion operation and
is certified to the EPA Phase ! standard. Each stove was tested at high and low burn
rates. The target low burn rate was 0.45 to 0.91 kg/h, and the target high burn rate was
2.27 to 2.72 kg/h. Duplicate bums were performed for each stove at each targeted
rate. Each test burn required at least one refueling step.
Table 3 details the test schedule for each stove. Before changes in heating
appliance or operating conditions, the stove and chimney were cleaned with a stove
pipe brush to reduce cross contamination from the previous test. The stove and
chimney were conditioned by operating the facility for a minimum of 2 h at the target
TABLE 3. SAMPLING SCHEDULE
Operation Day 1 Day 2 Day 3 Day 4 Day 5 Day 6
Install Stove x
Clean Chimney x x
Condition for Low Burn Rate x
Condition for High Bum Rate x
Sample for Low Burn Rate x x
Sample for High Burn Rate x x
CEM Data via Data Logger x x x x
Wood Moisture Reading x x x x
Load Cell Data Each Half-Hour x x x x
WSDSS Data Each Half-Hour x x x x
Collect XAD-2 Sample x x x x
Collect Filter Sample x x x x
Collect Aldehyde Sample x x x x
Collect Summa® Canister x x x x
Collect fine Particle Sample x x x x
Collect XAD-2 Filter Blank x
Collect Filter Field Blank x
conditions without sample collection. The fire was allowed to die and the stove allowed
to cool to ambient conditions before a sample run was started. Sampling was from cool
6
-------
to cool conditions. Table 4 lists the actual sampling conditions. The impact on data
due to the fact that target burn rates were not met for some tests on the Answer stove
is discussed in Section 7.0.
TABLE 4. SAMPLING CONDITIONS
Run
No.
Sampling
Date
Stove Model
Bum Rate
(kg/h)
Stack Flow
(Nm3/ min)
Sampling
Time (min)
1
5/11/89
LOPI 380/440
3.40 (High)
0.845
240
2
5/15/89
LOPI 380/440
2.42 (High)
0.693
240
3
5/17/89
LOPI 380/440
1.05 (Low)
0.461
240
4
5/18/89
LOPI 380/440
0.72 (Low)
0.301
240
5
5/23/89
LOPI Answer
1.28 (High)
0.639
240
6
5/24/89
LOPI Answer
1.42 (High)
0.602
240
7
5/31/89
LOPI Answer
0.85 (Low)
0.331
235
8
6/1/89
LOPI Answer
0.98 (Low)
0.372
240
7
-------
SECTION 3.0
ANALYTICAL METHODS AND PROCEDURES
3.1 ALDEHYDES
Aldehydes were analyzed by HPLC in the laboratories of Roy Zweidinger of
AREAL/EPA by procedures established in that laboratory. Each tube was analyzed
individually. Collecting four tubes from each burn, two parallel pairs in series, provided
QA checks on the analysis and sample collection. Analysis of the back tubes detects
the presence of break-through during sample collection. Comparison of results
obtained from the parallel sample collections detects questionable results because of
such factors as tube overload and clogging.
3.2 VOLATILE ORGANIC COMPOUNDS
VOCs were analyzed by GC/FID in the laboratories of Robert Seila of AREAL/EPA
by procedures established in that laboratory. An aliquot of gas from the evacuated
stainless steel Summa® canister was injected. Compound identification was based on
comparing the retention time of the compound to a library of well-characterized
standard compounds. Identified compounds were quantified from stored calibrations.
Where identification was not possible, an averaged response factor was used.
3.3 CONTINUOUSLY RECORDED DATA
Continuously recorded data during this project included stack temperature, total
hydrocarbons, carbon dioxide, and oxygen concentration in the stack. These data were
collected by a computer-run "data logger" that averaged the data over 5-min intervals
and periodically transmitted the data via modem to the Project Officer's computer. CEM
data for CO were not collected because concentration exceeded the 1,000 ppm range
of the instrument the majority of the time.
3.4 TOTAL CAPTURE
Teflon® filters from the fine particle dichotomous sampler and 142-mm Pallflex®
filters were transferred to the laboratories of AEERL after sampling. After desiccation
for 24 h, they were weighed on a Mettler A-250 balance. Total capture was determined
as the difference between this weight and the tare weight. After determining total
capture, the Pallflex® filters were extracted and analyzed for organic constituents while
the dichotomous sampler Teflon® filters were transferred to AREAL for elemental
analysis.
3.5 ELEMENTAL ANALYSIS AND ORGANIC CARBON/ELEMENTAL
CARBON DETERMINATIONS
Charles Lewis of AREAL/EPA was responsible for analyzing the quartz and Teflon®
filters collected from the dichotomous sampler. The quartz filters were sent to Sunset
Laboratories in Forest Grove, OR for organic carbon and elemental carbon analysis
while the Teflon® filters were retained and used in Lewis' elemental analysis
determinations by X-ray fluorescence.
8
-------
3.6 SAMPLE EXTRACTION AND ANALYSIS
Pal If lex® filters for each run were extracted by ultrasonic solvent extraction
techniques. Filters from a run were placed in a cleaned beaker. To completely
submerge the filters, 100 mL of reagent-grade dichloromethane were added to the
beaker. An aluminum foil cover was placed over the mouth of the beaker, and the
beaker was placed in an ultrasonic water bath. The liquid in the bath was shallow
enough so that the beaker sat firmly on the bottom. The ultrasonic bath was then
turned on for 15 min. After sonicating, the dichloromethane was poured off into a
collection flask. These steps were repeated three more times for a total of 400 mL
dichloromethane extract.
After sampling, the XAD-2 cartridges were resealed in Teflon® bags and stored in a
freezer until extraction. Each XAD-2 cartridge was extracted by pump-through elution
as described in AEERL/ROP 41.6
All dichloromethane extracts were concentrated using a Kudema-Danish apparatus
as described in AEERL/ROP 41. According to the ROP, concentration is stopped at the
first evidence of saturation, and the extract is transferred to a volumetric flask and made
up to a known volume with dichloromethane. The extracts are stored in a freezer until
analysis and returned to the freezer after analysis is performed.
3.7 TOTAL CHROMATOGRAPHABLE ORGANIC COMPOUNDS ANALYSIS
TCO mass is defined as those compounds with boiling points in the range of 98 to
300 °C, corresponding to straight chain alkane carbon numbers C7 to C,7.
Dichloromethane extracts of Pal If lex® filters and XAD-2 samples were subjected to TCO
compounds analysis according to the established procedures of AEERL/ROP 13.6
Each sample was analyzed in duplicate by direct injection GC/FID, and the reported
result is the average of these determinations,
3.8 GRAVIMETRIC ANALYSIS
GRAV mass is defined as those compounds with boiling points of higher than 300
°C. Dichloromethane extracts of Pal If lex® filters and XAD-2 samples were subjected to
GRAV analysis according to the established procedures of AEERL/ROP 12.6 According
to the procedure, each sample is analyzed in duplicate, and the reported result is the
average of these determinations. Acurex Environmental deviated from the ROP in that
the sample aliquot added to each pan was 0.25 mL, rather than 1.0 mL, to preserve
sufficient sample for subsequent analyses. Each GRAV run includes the analysis of
blank samples to detect contamination by laboratory particulate.
Balance data were transferred directly to a computer spreadsheet via an RS-232
interface and Lotus Measure®. This exchange eliminates data transfer and arithmetic
errors. QC checks are built into the spreadsheet to ensure valid data reporting. Any
sample that fails these QC checks is repeated with additional weighings or fresh extract
in new pans until all samples pass.
3.9 POLYNUCLEAR AROMATIC HYDROCARBONS ANALYSIS
The term PAH describes a series of fused-ring aromatic hydrocarbons that are
9
-------
frequent products of incomplete combustion associated with woodsmoke. Certain
members of this class are potent carcinogens, PAHs were analyzed by GC/FID on a
Hewlett-Packard model 5880 instrument equipped with a J&W Scientific 30 m by 0.32
mm DB-5 column. The injector and detector were held at a temperature of 300 °C
throughout the period of the analytical run. During analysis, the column oven was
initially held at 40 °C for 3 min. This was followed by an 8° per min ramp to 280 °C.
The oven was then operated isothermally until the total run time of 45 min was
completed, Calibration was performed for the 18 compounds naphthalene,
acenaphthalene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene,
pyrene, retene, benz[a]anthracene, chrysene, benzo[e]pyrene, benzo[b]fluoranthene,
benzo[k]fluoranthene, benzo[a]pyrene, dibenz[a,h]anthracene, benzo[ghi]perylene, and
indeno[1,2,3-cd]pyrene.
Analysis was performed on composited extracts of the XAD-2 and filter samples
collected during a sample run. For sampling run Number 8, for example, (LOPI Answer
stove at low burn rate conditions) this composite was prepared from extracts of XAD-2
cartridge Number 9 and filters numbered 21 and 22. It must be noted that, for sample
run Number 1, the XAD-2 extract was lost prior to analysis. PAH results for run Number
1 are based solely upon the composited filter extract. Dilution was typically required
prior to analysis and this was performed with dichloromethane.
A Hewlett-Packard model 7672A autosampler injected 1 jjL aliquot of an
appropriately diluted extract into a split injector operating at a 14:1 split ratio. The
results provided in this report are averages of duplicate injections. Identification of a
peak as a calibrated analyte was based upon retention time of the standard ± 0.04 min.
in most cases, retention time agreement was within 0,02 min of the calibration run.
Additionally, as a Quality Assurance step, injection of a spiked extract was performed to
confirm identifications based upon retention times. Spiking was performed with the 18
compounds previously mentioned.
3.10 BIOASSAY ANALYSIS
Bioassay analysis was conducted by EPA's Health Effects Research Laboratory
(HERL). Microsuspension and Ames assays 70 were performed and results converted
to units of Revertants/kJ of fuel consumed. Analyses were performed both with and
without activation factor S9 (indicated by + and - S9, respectively).
10
-------
SECTION 4.0
QUALITY ASSURANCE
This project was performed using procedures established by EPA's Integrated Air
Cancer Project (IACP). These procedures ensure a quality control approach greater
than that required by the Category IV QA rating assigned by AEERL.
Field blanks were not collected for the Summa® canisters used for VOC collection.
QC procedures were performed on the canisters after cleaning and were considered
sufficient by AREAL. Field blanks for the filters and XAD-2 canisters were taken for
Burns 1 and 8. Aldehyde field blanks were taken for Burns 1, 2, and 8. AREAL was
responsible for QC of the Teflon® and quartz fine particle filters. CEMs were calibrated
before each test using three different concentrations of span gas appropriate to each
instrument. Zero and span checks were performed after each test. Completeness data
are presented in Table 5.
The VOC, aldehyde, elemental, and total capture analyses were conducted by EPA
in Reasearch Triangle Park, NC, and the organic carbon and elemental carbon
analyses were conducted by Sunset Laboratories in Oregon, according to their methods
and QA procedures.
The XAD-2 canisters were prepared according to AEERL/ROP 40.6 The XAD-2
cartridges were extracted, and the dichloromethane extracts from Pallflex® filters were
concentrated according to AEERL/ROP 41.6 GRAV analysis was performed on the
dichloromethane extracts of the Pallflex® filters and XAD-2 samples using AEERL/ROP
12.5 Samples were run in duplicate, and each GRAV run included analyzing blank
samples to detect contamination by laboratory particulate matter.
Balance data were transferred directly to a computer spreadsheet, which has a built
in QC checks. If a sample failed a QC check, it was repeated until all samples passed.
TCO analysis was performed on the dichloromethane extracts of the Pallflex® filters and
XAD-2 samples using AEERL/ROP 13.6 Each sample was analyzed in duplicate.
PAHs were analyzed by GC/FID. Blanks were run and produced no integrable
peaks. Samples were run in duplicate. In addition, each sample was identically diluted
and spiked with a PAH standard mix containing each of the 18 PAH compounds of
interest. The samples and spikes were not permitted to deviate more than 0.04 min
retention time.
Loss of sample was experienced for the XAD-2 portion of Burn 1. For reporting
purposes, an estimated value has been included based on 60 percent of the mass
collected on the filter. This value was consistent for the other three tests using the
LOP I 380/440 stove. Analysis of the complete data set shows that the relative impact
of this estimation procedure on overall conclusions reached is insignificant.
Data quality are sufficient to meet the overall project objectives of collecting
information on organic emissions from two types of woodstove over a range of burn
rates.
This report considers temperature in degrees Fahrenheit. Readers more familiar
with the metric system may convert to degrees Celsius by the equation 0.56 * (°F - 32).
ii
-------
TABLE 5. DATA COMPLETENESS
Analysis
Data Points
Completeness {%)
CO,
8
100
CO
01
0
0,
8
100
Temperature
8
100
THC
42
50
TCO Compounds
143
87
GRAV
16
100
Filter Capture
8
100
Teflon® Catch
8
100
Organic Carbon
8
100
Elemental Carbon
8
100
Elemental Analysis
8
100
VOCs
8
100
Aldehydes
8
100
Load Cell Data
8
100
Overall
122
90
3 Instrument calibration was exceeded during each sample run.
b THC data was collected but not transmitted by the data logger for the first four runs.
c Two extracts were lost in the laboratory.
12
-------
SECTION 5.0
PRESENTATION OF RESULTS
Table 6 presents results from the eight tests as of analyte emitted/kj of fuel
consumed. A value of 20,000 kJ/kg (8,598 Btu/lb) of wood on a dry basis was used in
calculation of these values after correcting scale values for a measured average of 23.9
percent moisture for the wood used. Extracted organic material (EOM) represents the
sum of measured TCO and GRAV values. Filter and XAD data are combined in the
presentation of TCO and GRAV values because the distribution of SVOCs between the
sampling substrates is a function of filter temperature, an uncontrolled variable in these
tests. VOC data represents the sum of compounds analyzed individually and
presented in Table 7. Table 8 presents the results of elemental analysis of material
collected on the Teflon® filters. Results of GC/FID analyses for PAHs are summarized
in Table 9. Aldehyde measurements are shown in Table 10. Finally, Table 11 presents
data for the bioassay analyses from all eight tests.
13
-------
TABLE 6. OVERALL EMISSIONS, LOPI STOVES (ug/kJ)
Run #
1
2
3
4
5
6
7
8
Stove Model
380/440
380/440
380/440
380/440
Answer
Answer
Answer
Answer
Burn Rate
(kq/h)
3.40
2.42
1.05
0.72
1.28
1.42
0.85
0.98
TCO
124.68
82.24
577.94
239.18
159.50
73.90
67.00
39.22
GRAV
62.73
177.03
770.28
361.43
406.09
455.25
335.22
246.07
EOM
187.41
259.28
1348.22
600.60
565.59
519.15
402.22
285.29
filter GRAV
62.75
135.41
511.30
187.45
103.29
227.20
58.94
110.97
filter catch
363.14
205.52
1470.24
525.16
205.78
171.68
69.12
204.53
VOCs
964.5
590.2
38.9
26.9
84.7
*
299.2
8.9
Org. Carbon
363.99
186.62
187.81
368.06
96.24
71.08
32.19
118.44
El. Carbon
19.12
14.06
3.63
8.41
4.05
2.66
1.24
2.45
Ratio
19.04
13.28 ,
51.71
43.77
23.76
26.72
26.05
48.32
* Blank Entries had activities indistinguishable from blank runs.
-------
TABLE 7. VOC RESULTS, LOPI STOVES (ug/kJ)
Run No.
1
2
3
4
5
6
7
8
Stove Model
380/440
380/440
380/440
380/440
Answer
Answer
Answer
Answer
Burn Rate (kg/h)
3.40
2.42
1,05
0.72
1.28
1.42
0.85
0.98
Ethylene/acetylene/ethane
157.52
155.42
13.07
*
Propene
19.12
19.97
Propane
9.13
9.92
2-Methyl propylene,
butene-1
12.74
1,3-Butadiene
8.61
10.31
t-2-Butene
6.90
5.28
c-4 Olefin
8.28
c-5 Olefin
10.90
8.15
2-Methyl-2-butene
11.43
18.66
84.67
299.15
8.87
c-2 Pentene
412.47
c-6 Paraffin
9.20
14.91
t-4-Methyl-2-pentene
9.53
14.52
c-6 Olefin
11.96
12.81
Benzene
121.72
114.37
Toluene
32.19
33.83
14.39
(Continued)
-------
TABLE 7. VOC RESULTS, LOPI STOVES (ug/kj) (Concluded)
Run No.
1
2
3
4
5
6
7
8
Stove Model
380/440
380/440
380/440
380/440
Answer
Answer
Answer
Answer
Burn Rate (kg/h)
3.40
2.42
1.05
0,72
1.28
1.42
0.85
0.98
c-9 Paraffin
41.45
49.53
m&p-Xylene
10.77
13.07
c-9 Olefin
16.88
28.18
c-10 Paraffin
14.85
12.61
o,m,p-Methyl styrene
10.90
14.78
c-10 Aromatic
18.52
c-11 Aromatic
47.63
54.06
* Blank entries had activities indistinguishable from blank runs.
-------
TABLE 8. ELEMENTAL ANALYSES OF TEFLON FILTER CAPTURE, LQPI STOVES (pg/kJ)
Run #
T
2
3
4
5
6
7
8
Stove Model
380/440
380/440
380/440
380/440
Answer
Answer
Answer
Answer
Burn Rate (kg/h)
3.40
2.42
L 105
0.72
1.28
L_ 1.42
0.85
0.98
Argon
0.031
*
0.014
Barium
0.016
0.095
0.026
0.026
Bromine
0.016
0.095
Calcium
0.001
Cesium
0.017
Chlorine
0.235
0.357
0.698
0.235
0.093
0.106
0.032
0.068
Chromium
0.002
Copper
0.005
Lanthanum
0.037
0.070
Lead
0.004
0.000
Nickel
0.002
0.004
Potassium
0.525
0.932
1.229
0.757
0.185
0.261
0.030
0.206
Rubidium
0.002
Silicon
0.043
Strontium
0.002
Sulfur
0.152
0.191
0.239
0.189
0.131
0.076
0.024
0.066
Tellurium
0.008
Zinc
0.035
0.079
0.023
0.040
0.010
0.009
0.002
0.002
* Blank entries had activities indistinguishable from blank runs.
-------
TABLE 9. GC/FID ANALYSES FOR PAHs, LOPI STOVES (ug/kJ)*
Run No.
1
2
3
4
5
6
7
8
Stove Model
380/440
380/440
380/440
380/440
Answer
Answer
Answer
Answer
Burn Rate (kg/h)
3.40
2.42
1.05
0.72
1.28
1.42
0.85
0.98
Naphthalene
3.02
16.03
36.20
9.59
49.00
12.35
2.19
31.01
Acenaphthalene
0.91
15.96
96.17
10.38
49.86
12.61
4.69
20.69
Acenaphthene
1.55
3.49
43.62
2.12
7.82
2.99
1.78
4.16
Fluorene
1.7
4.19
4.40
1.91
4.35
2.95
0.98
2.13
Phenanthrene
2.58
7.03
6.77
1.71
9.39
3.27
2.00
3.30
Anthracene
0.91
1.68
1.95
BQL1
2.91
0.85
0.37
0.91
Fluoranthene
3.22
1.93
2.82
BQL
1.97
1.12
0.81
1.16
Pyrene
3.25
1.93
2.49
BQL
1.35
1.16
0.65
0.75
Retene
BQL
2.52
4.75
1.13
BQL
BQL
0.81
1.59
Benzfalanthracene
0.14
BQL
BQL
BQL
BQL
BQL
BQL
BQL
Chrysene
0.62
BQL
1.65
BQL
BQL
BQL
BQL
BQL
Benzo[e]pyrene
BQL
BQL
4.47
BQL
BQL
0.79
BQL
BQL
Analytes benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene, dibenz[a,h]anthracene,
benzo[ghi]perylene, and ideno[1,2,3-cd]pyrene were Below Quantitation Limits for all samples and are not
reported separately in this table.
BQL = Below Quantifiable Limit (QL < 0.1 ;ug/kJ)
-------
TABLE 10. ALDEHYDE MEASUREMENTS, LOPI STOVES (ug/kJ)
Run No.
1
2
3
4
5
6
1
8
Stove Model
380/440
380/440
380/440
380/440
Answer
Answer
Answer
Answer
Burn Rate (kg/h)
3.40
2.42
1.05
0.72
1.28
1.42
0.85
0.98
Formaldehyde
182.03
109.57
290.41
155.42
107.80
78.11
39.48
107.47
Acetaldehyde
90.32
56.08
206.73
99.19
48.74
38.30
17.87
66.02
Acrolein
33.37
15.96
48.61
18.52
10.44
3.91
1.16
6.77
Acetone
32.19
20.36
58.33
44.67
25.68
15.63
7.36
32.06
Propionaldehyde
16.29
8.67
41.84
20.76
12.94
10.25
7.36
9.92
Crotonaldehyde
18.46
11.69
22.20
14.78
11.36
6.77
3.41
8.87
Butyraldehyde
2.03
0.88
7.09
2.21
2.34
4.10
0.87
2.73
Benzaldehyde
51.37
11.23
41.19
12.94
27.20
18.26
6.42
23.12
Isovaleraldehyde
2.96
4.14
11.10
7.09
3.75
2.17
0.94
5.87
Valeraldehyde
4.99
4.79
11.63
9.85
13.27
5.39
4.61
11.69
o-Tolualdehyde
3.64
~
1.60
m-Tolualdehyde
0.67
22.47
45.52
25.62
24.44
37.18
18.72
37.44
p-Tolualdehyde
29.82
0.96
17.80
Hexanaldehyde
4.99
0.15
* Blank entries had activities indistinguishable from blank runs.
-------
TABLE 11. BIOASSAY RESULTS, LOPI STOVES (Revertants/kJ)
Run No.
1
2
3
4
5
6
7
8
Stove Model
380/440
380/440
380/440
380/440
Answer
Answer
Answer
Answer
Burn Rate (kg/h)
3.40
2.42
1.05
0.72
1.28
1.42
0.85
0.98
MSA+S9 XAD
*
109.30
6.41
19.47
MSA+S9 Filter
14.19
51.11
26.59
20.01
13.21
4.69
Ames +S9 XAD
Ames -S9 XAD
6.07
4.04
2.67
Ames +S9 Filter
3.28
4.88
7.08
11.87
9.60
7.49
4.04
Ames -S9 Filter
2.18
2.21
26.31
4.27
3.47
3.96
1.78
* Blank entries had activities indistinguishable from blank runs.
-------
SECTION 6.0
DISCUSSION OF RESULTS
6.1 SAMPLING
Four runs were successfully completed with each stove. For the conventional
stove, a factor of nearly 5 (4.72) was achieved between the lowest and highest burn
rates. This range was smaller {1.67) for the LOPI Answer stove. This was caused by
the manufacturer's design restrictions that allowed operation over only a narrow wood
consumption range. The primary restriction consisted of high and low limits on the air
inlet draft control.
The secondary combustion feature on the Answer stove consisted of a pierced
tube feeding auxiliary air to the combustion gases. When the stove was operated at a
sufficiently high burn rate, auxiliary air would bleed out because of expansion.
Secondary combustion would occur when the combustion gases plus auxiliary air were
hot enough. Secondary combustion could be observed as individual flames at the
auxiliary air tube ports.
An objective measurement of secondary combustion operation was not part of
the experimental design for this task. Subjectively, secondary combustion was quite
consistent at higher bum rates. At the low burn rates of Burns 7 and 8, secondary
combustion did not always occur.
As noted previously, instrument difficulties made collection of carbon monoxide
data impossible, Total hydrocarbon data were not available for the conventional stove.
Electronic problems in the interface prevented transmission of these data. When the
problem was recognized, it was corrected for the Answer stove. Stack temperature
data were based on manual records at half-hour intervals. As shown in Figure 2, stack
temperature showed a general increasing trend with increasing burn rate, although
Test 3, at 1.05 kg/h had an average stack temperature below that for Run 4 at 0.72
kg/h. As seen in the analysis of data given in the following discussion, this may have
had a biasing effect on measured emissions. Data on 02, and C02 concentrations are
shown in Figures 3 and 4, respectively.
As expected, 02 levels decrease with increasing burn rate. As per the
anticipated relationship, C02 concentrations show a corresponding increase with
increasing bum rate (and decreasing 02).
Stack gas velocity measurement was unsuccessful. A Sierra Instruments hot
wire anemometer would not function because condensable materials quickly coated the
wire. Total stack gas volume was calculated from stoichiometric relationships.9
6.2 TOTAL ORGANIC MASS EMITTED
The sum of the total EOM and Summa® canister results represent the total
organic mass emitted. EOM from the filter represents the non-volatile organic
compounds (NVOCs) and some of the semivolatile organic compounds (SVOCs); the
XAD-2 EOM includes the rest of the NVOCs and SVOCs; and the Summa® canister
21
-------
Average of 4 hour burn.
Burn Rate, kg wood/h
Figure 2. Stack temperature versus burn rate.
-------
19
is. 5
IB
17 . 5
17
;16.5
16
15 . 5
15
14 . 5
14
(
ire 3.
380/440
Answer
I | |
1.5 2 2.5
Burn Rate, kg wood/hr
3 . 5
Oxygen concentration versus burn rate.
-------
Average of 4 hour burn.
Figure 4. Carbon dioxide concentration versus burn rate.
-------
represents the VOCs. These data are plotted versus bum rate in Figure 5, showing no
clear trend, although a trend indicating that emitted organics increase with burn rate
might be inferred especially if Test 3 is temporarily ignored. This effect has been
observed in earlier tests.10 The Burn 1 XAD-2 EOM is estimated as described earlier.
Figure 5 shows an error band indicating the effect of eliminating Bum 1 XAD-2 EOM
from the calculation of total organic mass emitted, demonstrating that the effect on the
noted trend is minimal.
VOCs as a function of burn rate are presented graphically in Figure 6. The
graph shows that VOCs generally increase for higher bum rates, resulting in a trend
toward increasing total organics with increasing bum rate. Here again, this trend has
been noted inn earlier tests10. A general trend of decreasing EOM as a function of
increased burn rate can be seen in Figure 7, although it can be noted that EOM for
Burn 3 is more than twice that of the other tests.
EOM is of primary importance in woodstove emissions. The LOPI 380/440 stove
clearly shows decreased EOM with increasing burn rate. The burn rate range for the
LOPI Answer stove is too narrow to confidently allow conjecture on the effect of burn
rate on EOM emissions for this stove.
As burn rate increases, stack temperature and stack flow increase while the
stack gas oxygen content decreases. As a consequence of the increased stack flow,
stove residence times decrease. In looking at general trends in the data it may be
noted that the total organic emissions increase with burn rate. This number, the sum of
NVOC, SVOCs and VOCs, increases as a direct result of increasing VOCs. This must
be so because EOM, the sum of NVOCs and SVOCs, decreases with burn rate. VOCs
may be regarded as the product of a cracking process on the original lignin molecule
and on the NVOCs and SVOCs (intermediate products in this cracking process). This
cracking process increases with temperature resulting in the observed decreasing
NVOCs and increasing VOCs with burn rate.
The primary "consumption" process for VOCs is oxidation resulting, primarily, in
the production of carbon dioxide and water. As noted above stack gas oxygen
decreases with increasing burn rate. It would, therefore, be expected that the "survival"
of VOCs will increase with increasing burn rate and this is the observed trend.
6.3 POLYNUCLEAR AROMATIC HYDROCARBONS
No significant trend can be noted between total PAHs and burn rate or stove
type, as shown in Figure 8. Bum 3 is again significantly higher than the other tests.
6.4 ALDEHYDES
Figure 9 shows that no significant trends can be seen between aldehydes and
either burn rate or stove type. Results for Burn 3 are again much greater than those of
the other tests.
6.5 FINE PARTICULATE MATTER
While significant scatter in the data exists, organic and elemental carbon emitted
tend to increase with increased burn rate. The ratio between the two remains relatively
constant. This is likely a result purely of the tendency of increased stack flow rate to
25
-------
1400 1
1200
•i
M
1000
3
TJ
0)
4-»
JJ
e
u\
c
fd
o»
u
o
m
o
Eh
600
ro
O)
400
200
380/440
Answer
0 . 5
1.5 2 2.5
Burn Rate, kg wood/h
Figure 5. Total organics emitted versus burn rate.
-------
entrain more particulate matter. The same general trend {again with some data scatter)
is seen for potassium emissions, representative of inorganic ash emitted, thus
reinforcing this suggestion.
6.6 TOTAL WSDSS FILTER CAPTURE
A comparison of burn rate and total filter capture is of interest since the total filter
capture is that measured parameter from this study which most closely relates to
measurements by EPA Method 5G. This is found in Figure 10. As observed in several
previous sections, there is no clear relationships between stove burn rate and the filter
catch results. As has been observed elsewhere, Burn 3 produces the highest filter
capture value which is attributed to the low stack temperature of that burn.
Comparisons of the filter capture results and PAHs, EOM, and total VOCs (all in
units of /ig/kJ) have been examined and are not separately presented here. No clear
relationships were observed. In the cases of PAHs and EOM, this result is expected
since a portion of these analytes are captured post-filter, on the XAD-2 cartridge.
Figure 11 presents a comparison of filter capture and total aldehydes. Here, a
monotonic relationship can be observed. A power curve (y = a*xb) has been drawn to
illuminate the trend only and is not intended to suggest the specific form of any
relationship between these variables. It should be noted that filter catch and aldehyde
results are based upon unrelated sampling and analytical events. It is unlikely,
therefore, that the observed trend is due to a sampling or analytical artifact. It is also
unlikely that there is a real dependent.independent relationship between aldehydes and
filterable material.
27
-------
Burn Rate, leg wood/h
Figure 6. VOCs versus burn rate.
-------
o
u
(D
Z
o
CO
oo
£
on
<
¦
¦4
u>
fvi
TJ
O
O
5
m
O
HI
h-
-------
210
1 80
M
3
S 150
o
A
U
d
O
o
k
120
>i
4J
«l
£
o
k
o
&
N
f-H
o
0|
o
fc*
90
fiO
30
3 8 0/440
Answer
0 . 5
1.5 2 2.5
Burn Rate, kg wood/h
Figure 8. PAH versus burn rate.
-------
Lrt
CO
o
o
00
CO
0
£
0)
c
<
ui
o
o
5
Q
OS
a
m
o
o
00
o
o
r*
un
o
o
o
<0
o
o
m
o
Q
O
O
m
o
o
a
o
o
's^pZq^pxv T^oj.
cn
0
k.
3
Q)
31
-------
1600 1
CO
IV)
1400
1200
*10 0 0
o»
a
.c
a
+J BOO
«
u
t)
+J
¦h 600
b
400
200
0 . 5
380/440
Answer
T
T
I
1.5 2 2.5
Burn Rate, leg wood/h
Figure 10. Filter catch versus burn rate.
-------
Figure 11.
Filter Catch, ug/fcj
Total aldehydes versus filter catch.
-------
SECTION 7.0
CONCLUSIONS
The target sampling conditions were 0.45 to 0,9 kg/h for the low bum rate and
2,25 to 2.7 kg/h for the high burn rate. Although these specific values were not met
(see Table 4) because of operational characteristics of the Answer stove, the data show
some trends. Differences in emissions from the stoves were primarily a function of burn
rate and consisted mostly of va^ing levels of VOC and EOM emissions. VOCs
generally increased with increasing bum rate, while EOM decreased. No significant
trends in filter capture, total PAHs, or aldehydes were noted as a function of either burn
rate or stove type. Fine particulate matter, as typified by organic and elemental carbon
and potassium tended to increase with increased burn rate. Historically, burn rate has
been shown to be the major variable affecting emission rates.11
There was a wide disparity between emissions produced during duplicate Burns
3 and 4. The EOM, PAH, and aldehyde emissions from Burn 3 test are significantly
greater than from Bum 4.
Tables 6 through 11 present data that may be used to calculate emission factors
for woodstove use during the IACP Roanoke Oil Furnace study.
34
-------
SECTION 8.0
REFERENCES
1. Williamson, A. D. and D. B. Harris, "Measurement of Condensable Vapor
Contribution to PM10 Emissions," in Proceedings of the 78th Annual Meeting, Air
& Waste Management Association, Detroit, 1985, Paper No. 85-14.4.
2. Merrill, R. G.and D.B. Harris, "Field and Laboratory Evaluation of a Woodstove
Dilution Sampling System," in Proceedings of the 80th Annual Meeting. Air &
Waste Management Association, New York, 1987, Paper No. 87-64.7.
3. McCrillis, R. C.and P.G. Burnet, "Effects of Operating Variables on Emissions
from Woodstoves," in Proceedings: 1988 EPA/APCA International Symposium:
Measurement of Toxic and Related Air Pollutants. Research Triangle Park, NC,
May 1988, EPA-600-/9-88-015 (NTIS PB90-225863).
4. Steiber, R. S., R. C. McCrillis, J.A. Dorsey, and R.G. Merrill, Jr.,
"Characterization of Condensible and Semivolatile Organic Materials from Boise
Woodstove Samples," in Proceedings of the 85th Annual Meeting of the AWMA,
Air & Waste Management Association, Kansas City, MO, 1992, Paper No. 92-
118.3.
5. McCrillis, R. C. and R. R. Watts, "Analysis of Emissions from Residential Oil
Furnaces," in Proceedings of the 85th Annual Meeting of the AWMA. Air &
Waste Management Association, Kansas City, MO, 1992, Paper No. 92-110.6.
6. Tufts, M. and D. Natschke, Emissions from Burning Cabinet Making Scraps.
EPA-600/R-93-213 (NTIS PB94-130408), November 1993.
7. R. Stevens, C. Lewis, T. Dzubay, L. Cupitt, and J. Lewtas, "Source of Mutagenic
Activity in Urban Fine Particles," Toxicology and Industrial Health, 6, 81-94
(1990).
8. DeMarini, D., M. Dallas, and J. Lewtas, "Cytotoxicity and Effect of Mutagenicity
of Buffers in a Microsuspension Assay," Teratogen, Carcinogen, and Mutagen.
9:287 (1989).
9. Burnet, P., "Northeast Cooperative Woodstove Study Volume II—Technical
Appendix, November 1987. EPA-600/7-87-026b (NTIS PB88-140777).
10. McCrillis, R.C. and R.G. Merrill, "Emission Control Effectiveness of a Woodstove
Catalyst and Emission Measurement Methods Comparison," in Proceedings:
78th Annual Meeting. APCA. Paper No. 85-43.5, Detroit, June 1985.
35
-------
11. McCrillis, R.C., R.R, Watts, and S.H. Warren, "Effects of Operating Variables on
PAH Emissions and Mutagenicity of Emissions from Woodstoves,' Journal of
The Air and Waste Management Association, Vol. 42, No. 5: 691-694, May
1992.
36
------- |