United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-84-094 Nov. 1984
Project Summary
Characterization of
Emissions from the
Combustion of Wood and
Alternative Fuels in a
Residential Woodstove
R. S. Truesdale, K. L. Mack, J. B. White, K. E. Leese, and J. G. Cleland
This study was undertaken to com-
pare the emissions from the combustion
of alternative fuels to those from wood
in a residential woodstove, and to check
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 emissions of SO2, particu-
late, and NO,. 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. NO, emissions
are controlled by fuel nitrogen content
and combustion air flow rate. Organic
emissions are affected by fuel consump-
tion rate, fuel structure, and amount of
air through the stove. PAH formation is
affected by combustion airflow, firebox
temperature, and fuel structure. Bio-
assay results indicate the presence of
both mutagens and promutagens in the
organic extracts of flue gas samples
from both wood and coal combustion
tests.
This Project Summary was developed
by EPA's Industrial Environmental Re-
search Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering infor-
mation at back).
Introduction
The purpose of this study is to measure
the emissions from the residential com-
bustion of alternative fuels to wood,
including coal, in a conventional wood-
stove. Fuels tested include compressed
wood, treated wood, newspapers, com-
mercially available paper logs, and peat,
in addition to untreated oak wood and
bituminous coal. Pollutants including
particulates, SO,, NOX, CO, PAH, organics,
and benzo(a)pyrene were measured
during the course of this study for the
alternative fuels tested, and their emis-
sion levels are compared to those from
wood combustion. The effects of the
stove operation parameters on emission
levels of these pollutants are also con-
sidered. This information should be useful
in estimating the overall effect of these
emissionsfrom residential solid fuel units
on ambient air quality.
Procedure
During the planning phase of this
project, eight fuels were chosen as likely
alternatives to wood for use in residential
combustion units. Dry oak wood wasalso
tested so that emissions from alternative
fuels could be compared to it. The alterna-
tive fuels chosen were coal (both bitumi-
nous and anthracite), peat, newspaper
logs, cardboard logs, compressed wood-
chip logs (both with and without binders).
-------
and pine lumber pressure-treated with
copper compounds to retard rot.
The woodstove used in this study was
free-standing and air-jacketed, with a
simple open firebox. Originally designed
as a fireplace insert, this type of stove is
being installed in increasing numbers of
new homes. The stove utilizes forced air
circulation through the air jacket to con-
vect heat into the room.
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. These tests were used as a
baseline for comparison to tests with
other fuels.
Temperature was measured during
each run by several thermocouples: in the
firebox and the stack, at the air jacket
blower inputs and outputs, andinthetest
room (ambient). Temperature data and
certain gas data were automatically
recorded by the online DEC PDP-1100
computer. An RTI-designed turbine meter
(vane anemometer) was used for continu-
ous flow measurements of the stack flow
during each test.
Stack gas composition was continu-
ously monitored during the tests. Carbon
monoxide, carbon dioxide, and methane
were analyzed using infrared detectors.
Nitrogen oxides (NO,) were measured
using a photolummescent detector. Sul-
fur dioxide (SO2) was measured using a
photometric detector. In addition to con-
tinuous gas analysis, gas bulb samples
were also taken and analyzed by gas
chromatography for total organic carbon.
A whole-test-integrated sample of poly-
cyclic organic matter and other organic
emissions was collected by a modified
Method 5 sampling train similar to the
one described in the trial protocol. The
train was assembled and checked out
according to a test protocol developed to
meet RTI's situation as well as to incor-
porate the trial protocol. Mass emissions
were collected over a 45 to 1 20 minute
interval depending on the volume of
sample required for analysis.
Modified Method 5 samples which
were analyzed for organics include: probe
wash (CH2CI2 + CH2OH), filter, condenser
and XAD module wash (CH2CI2), XAD-2
adsorbent, condensate catch, and im-
pinger water. These samples were ex-
tracted using an EPA procedure. Each
test produced two samples: one was the
concentrated extract of the XAD and
particulate; and the other was the extract
of the aqueous impinger solutions and
the aqueous condensate.
Organics analyses were performed
separately on the two types of concen-
trated samples. Total organics with a
boiling point of 100 to 300°C were
determined by total chromatographable
organics (TCO). Organics with a boiling
point above 300°C were determined by a
gravimetric technique.
Glass capillary gas chromatography
(GC2) was used to determine PAH con-
centrations in the modified Method 5
sample extracts. PAH analyses were
performed separately on the two samples
described previously. PAH-spiked sam-
ples were used to identify PAHs in the
unspiked samples. PAH concentrations
were quantified using an internal stand-
ard. GC/MS was used to confirm GC2
identifications for selected samples.
In addition to GC2 analysis, a PAH
sensitized fluorescence spot test was
used to screen the XAD extract, con-
denser and probe wash, and m the
methylene chloride extract from the
aqueous impingers and condensate sam-
ples for the presence of PAHs. Both
original and concentrated extracts were
tested.
To properly compare the emissions of
the alternative fuels tested, it was neces-
sary to sample only at steady-state stove
operating conditions. Start-up and shut-
down conditions were too variable for
reproducible testing. Steady-state condi-
tions were chosen to approximate condi-
tions the typical stove owner would
achieve for most of the stove's operation.
Results and Discussion
Comparisons of the emission factors of
the fuels tested are given 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.
Particulate—Particulate emission re-
sults for the eight fuels successfully
tested are given in Figure 1. This figure
shows that the fuels may be ranked by
particulate emissions as follows (highest
to lowest):
1. Compressed wood-chip logs with
binders (CWB)
2. Bituminous coal (BC)
3. Newspaper logs (N)
4. Treated Lumber (TW)
5. Peat(P)
6. Compressed wood-chip logs (no
binders) (CW)
7. Cardboard logs (C)
8. Wood (W)
Sulfur Dioxide (SOg)—Figure 2 gives
SO2 emission factors for the fuels tested.
The eight fuels ranked as follows with
regard to sulfur emissions (highest to
lowest):
1. Bituminous coal
2. Compressed wood-chip logs with
binders
3. Peat
75-
5-
•n
Jl
g/kg fuel consumed
Jl
40
30
-20
W CW CWB C
Figure 1. Emission factors: particulate.
BC
-------
8-
6
4-
2-
rfl ^
B p/*0 foe/ consumed • 1 7.30
ft g/hr
1
1
Jl
1 <^i r^ '
39.32
24
•22
20
18
16
14
•12
10
•8
6
•4
•2
X
Ol
W CW* CWB
'Factors based on single test.
Figure 2. Emission factors' S
4. Cardboard logs
5. Wood
6. Compressed wood-chip logs (no
binders)
7. Treated lumber
8. Newspaper logs
SO2 emissions varied directly with fuel
sulfur content.
Nitrogen Oxides (NOJ—NOX emission
factors are shown in Figure 3. Two rank-
ings of fuel by NO, emissions are possible.
First, considering NOX emission rates
(g/hr), the fuels may be ranked as follows
(highest to lowest):
1. Peat
2. Compressed wood-chip logs with
binders
3. Bituminous coal
4. Wood
5. Compressed wood-chip logs (no
binders)
6. Cardboard logs
7. Newspaper logs
8. Treated logs
Considering NO, emission factors(g/kg
fuel consumed), the fuels may be ranked
as follows (highest to lowest):
/v*
TW*
BC
1. Compressed wood-chip 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-chip logs (no
binders)
The difference in ranking between g/hr
and g/kg emission factors is due to differ-
ence in fuel consumption rates. Higher
heating value fuels (BC and CWB) have
low fuel consumption rates because less
fuel has to be burned to produce a unit
heat output. Two factors were found to
influence NO, emission magnitude: fuel
nitrogen content and stack gas flow rate.
Fuel nitrogen content was judged to be
the most important factor affecting NOX
emissions from the combustion of these
fuels.
Carbon Monoxide (CO)—CO emission
factors are given in Figure 4. CO emis-
sions for the various fuels tested did not
vary as much as with the previously
discussed pollutants. The ranking of fuels
according to CO emission factors (g/kg
fuel consumed) is as follows (highest to
lowest):
1. Newspaper logs
2. Compressed wood-chip logs with
binders
3. Peat
4. Bituminous coal
5. Cardboard logs
6. Compressed wood-chip logs (no
binders)
5-
4-
3-
01
X
O)
2-
1-
ff/^ff fuel consumed
, I
12
W
IV CW CWB C N
Figure 3. Emission factors: fl/O,.
TW
BC
-------
7. Treated lumber
8. Wood
Ranking of fuels according to CO emis-
sion rates (g/hr) is as follows (highest to
lowest):
1. Newspaper logs
2. Compressed wood-chip logs (no
binders)
3. Peat
4. Cardboard logs
5. Compressed wood-chip 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 earlier for N0« emis-
sions.
CO emission levels could not be suc-
cessfully correlated with stove operating
parameters, including stack flow rates,
firebox temperatures, stack temperatures,
fuel consumption rate, and heat output.
The physical structure of the fuels tested
probably is the major factor affecting CO
emissions: compressed man-made fuels
have higher CO emissions than the
naturally formed fuels (wood and coal).
Results from duplicate tests for each fuel
suggest that combustion air flow also
affects CO emissions: reduced air flow
leads to increased emissions.
Organics—Results from total chromat-
ographical organics and gravimetric
analyses indicated that total organic
emissions in the flue gas were similar for
all fuels except N andP, which had higher
organic emission factors. This is some-
what surprising since N had the lowest
PAH emission factors. BC, with organic
emissions comparable to most other fuels,
had the highest proportional contribution
of heavy organics.
Polynuclear Aromatic Hydrocarbons
(PAH)—?M\ formation was affected by
combustion air flow, firebox temperature,
and fuel structure. Composite structured
fuels had higher PAH formation except
for N which, in contrast to high total
organic emissions, had very low emis-
sions of heavier PAHs. It was concluded
that, during the tests with N, firebox
temperatures were too low for extensive
cyclization reactions leading to PAH
formation. Other composite fuels had
relatively high PAH production rates:
700-
80
60
f
40
20
g/kg fuel consumed
300
200
h700
W CW CWB
Figure 4. Emission factors: CO.
N TW P BC
5-
4-
n
n
w
CW
CWB
N
TW
BC
Figure 5. Emission factors, benzfajpyrene
attributed to their structure, which limits
the availability of air during combustion
and creates starved air conditions favor-
able to PAH production. Tests with TW
also had relatively high levels of PAHs in
the flue gas effluent stream: attributed to
low air flow through the stove during
these tests. W and BC had similar PAH
emissions: BC emitted less PAHs than
wood. PAH emissions from BC could
possibly be pyrolysis products from the
coal itself. Figure 5, a comparison of
benz(a)pyrene (BaP) emission factors for
eight fuels tested, shows that BaP emis-
-------
sion factors are lowest for N. W and BC
have similar levels. Other fuels had
higher BaP emission factors probably due
to burning characteristics which reduce
air to the fuel (particulate fuels) or to very
low air flow through the stove (TW).
Bioassay—Method 5 sample extracts
from one W combustion (W1) and one BC
combustion test (BC1) were subjected to
an Ames Salmonella mutagenicity assay
to measure their mutagenic potential.
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 demonstrated
an increase in mutagenic activity with the
addition of S9, a metabolic activator of
promutagen compounds. Therefore, both
samples contain direct-acting mutagens
and promutagens.
The BC combustion sample was more
mutagenic than the W combustion sam-
ple, 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 BC extract is more mutagenic than
the W extract by a factor of two. Since
emission factors (g/kg) for the PAHs
analyzed in this report are only slightly
higher for BC1 than f or W1, this suggests
that compounds other than the 24 PAHs
analyzed in this report may be contribu-
ting to the mutagenicity of these samples.
Conclusions
1. Overall oak wood (W) was the best
fuel, considering both emissions
and stove operation. Cardboard logs
(C) were almost as good as W.
Although they did emit more CO
and PAHs than W, 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, partic-
ulate, and NO,. In addition, CWB
emissions were high in CO and
PAHs.
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 and had the highest
PAH emission rates of all fuels.
4. Treated wood (TW) should not be
burned under any circumstances
because of the presence of arsenic
compounds which probably volatil-
ize during combustion. Other stud-
ies have shown that, in the com-
bustion of chlorophenol-treated
wood products, polychlorinated
dibenzo-p-dioxins (PCDD) and poly-
chlorinated dibenzofurans (PCDF)
are emitted.
5. Peat (P) emissions had relatively
high levels of NO*, SO2, CO, and
PAHs.
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, news-
paper (N)) produced particulate
emissions higher than those of W.
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 man-made,
compressed particulate structure
(CW, CWB, C, P) and N had high CO
emissions because their structure
inhibited air flow to the combustion
zone. W, TW, and BC had the lowest
CO emissions: these fuels would
shrink and crack when burned,
permitting sufficient air to reach
the burning fuel. Results from dupli-
cate tests for each fuel suggest that
air flow through the stove is also a
factor affecting CO emissions: re-
duced air flow leads to increased
emissions.
8. S02emission levels could be related
to fuel sulfur content: higher fuel
sulfur content causes higher S02
emissions. SO2 emissions were at
levels of environmental concern
only for P, BC, and CWB.
9. NOX emissions were controlled by
fuel nitrogen content and combus-
tion air flow rate. High nitrogen
content f uels (P and BC) had highest
NO, emissions. Increased air flow
through the stove also led to in-
creased NO, emissions. NO, levels,
generally low, were not as much of
a concern as other pollutants.
10. Organic emission levels were com-
parable for all fuels except P and N,
which had high levels of organics in
the flue gas effluent stream.
Organic emissions were affected
by fuel consu mption rate, fuel struc-
ture, andamountof air through the
stove. Higher fuel consumption
sometimes led to increased organ-
ics. Lowering air flow through the
stove increased organic emissions.
N had high organic emissions be-
cause of their physical structure,
which inhibited air from reaching
the combustion zone leading to
increased pyrolysis products.
11. PAH formation was affected by
combustion airflow, firebox temper-
ature, and fuel structure. Composite
structured fuels had higher PAH
formation except for N which, in
contrast to high total organic emis-
sions, had very low emissions of
heavier PAHs. It was concluded
that, during the tests with N, firebox
temperatures were too low for
extensive reactions leading to PAH
formation. Other composite fuels
had relatively high PAH production
rates: attributed to their structure,
which limits the availability of air
during combustion and creates
starved air conditions favorable to
PAH production. Tests with TW also
had relatively high levels of PAHs in
the flue gas effluent stream: attrib-
uted to low air flow through the
stove during these tests. W and BC
had similar PAH emissions, with
BC emitting less PAHs than wood.
PAH emissions from BC could
possibly be pyrolysis products from
the coal itself.
12. Bioassays on organic extracts from
one W test and one BC test demon-
strated the presence of both muta-
gens and promutagens in the
sample extracts. Organics from BC
combustion were about twice as
mutagenic as those from W com-
bustiorf on a mutagenicity per unit
mass of fuel consumed basis.
-------
/?. S. Truesdale, K. L Mack, J. B. White, K. E. Leese, andJ. G. C lei and are with
Research Triangle Institute, Research Triangle Park, NC 27709.
Michael C. Osborne is the EPA Project Officer (see below).
The complete report, entitled "Characterization of Emissions from the Combustion
of Wood and Alternative Fuels in a Residential Woodstove," (Order No. PB
85-105 336; Cost: $14.50, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
irUSGPO: 1984—559-111/10722
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
PS 0000129
4GENCV
------- |