United States
Environmental Protection
Agency
Industrial Environmental Researc
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-81-127 Sept. 1981
Project Summary
High Altitude Testing of
Residential Wood-Fired
Combustion Equipment
J. A. Peters and D. G. DeAngelis
The increase in residential wood
burning across the nation has sparked
interest in assessing the potential
effects of such burning on air quality.
In EPA Region VIII, residential wood
burning for aesthetic reasons is in-
creasing in the ever-expanding ski
areas. Since most ski areas are located
in narrow valleys which have a small
carrying capacity, the impact of resi-
dential wood burning could be signifi-
cant.
To determine whether emissions
from operating a wood stove at high
altitude differ from those at low alti-
tude, a high altitude sampling program
was conducted which was compared
to previously collected low altitude
data. The low altitude data used for
comparison were that from the study
entitled "Preliminary Characterization
of Emissions from Wood-Fired Resi-
dential Combustion Equipment" pre-
pared by Monsanto Research Corpo-
ration (EPA-600/7-80-040, March
1980).
Emission tests were conducted in
the identical model stove using the
same type of wood with the same
moisture content, amount of wood
charged, burning rate, air flow rate,
and identical sampling intervals and
port locations. Particulate emissions,
carbon monoxide, and polycyclic or-
ganic matter were analyzed and no
statistically significant difference in
emissions was found. Thus, for a
particular type of wood burned, emis-
sion tests conducted at any altitude
are applicable to regions of different
altitude.
This Project Summary was devel-
oped by EPA's Industrial Environmen-
tal Research Laboratory, Cincinnati,
OH, 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
information at back).
Introduction
As North Americans turn increasingly
to wood-fired heat equipment to soften
the impact of rising fossil fuel costs, a
new problem is arising—the possibility
of increased levels of air pollution from
wood burning in stoves and fireplaces.
The potential effects of such burning on
air quality are of particular concern in
EPA Region VIII, where residential wood
burning for home heating and for aes-
thetic reasons is increasing in the valley
communities and ski resort areas, often
causing a winter haze. It is felt that the
reason for the problem is the near
simultaneous lighting of hundreds of
apres-ski blazes in condominium fire-
places and woodburning stoves.
A study was conducted in 1979 by
Monsanto Research Corporation (MRC)
for EPA to characterize the emissions
and determine the environmental impact
of residential wood combustion equip-
ment (EPA-600/7-80-040 and EPA-
600/2-80-042b); however, the emission
testing was .done at low altitude. Since
-------
combustion processes and associated
emissions often differ at elevated alti-
tudes, e.g., automobile emissions, it
could not be assumed that existing data
from near sea level were applicable to
the high altitude Region VIII communi-
ties. An earlier study conducted in Vail,
Colorado by PEDCo-Environmental, Inc.
indicated that emission factors from
residential wood combustion at higher
altitudes differed from those at lower
altitudes by a factor of two or more.
The objective of this sampling pro-
gram was to determine whether emis-
sions from operating a wood stove at
high altitude differ from those at low
altitude. The low altitude data from the
previous MRC study were used for
comparison. Accordingly, test conditions
duplicated the previous study, with
altitude the only variable which was
changed.
Experimental
The sampling program was designed
to duplicate the earlier, low altitude
study—specifically, the emission test
runs on a nonbaffled airtight stove
burning seasoned red oak. Under the
low-altitude study a matrix of test
conditions was set up involving three
types of combustion devices burning
four types of wood. One of the twelve
test conditions was strictly duplicated:
identical stove, wood type, wood mois-
ture, size of logs, fuel bed configuration,
wood burning rate, stack gas flow rate,
height of sampling ports above stove,
and sampling methods and equipment.
Also available for comparison were
results of an emission test run on a
stove conducted by PEDCo in their Vail
sampling program. Table 1 illustrates
the experimental design scheme for
analysis comparison.
Samples were also collected for
measuring polycyclic organic materials
(POMs), which also yield front-half
particulates, and carbon monoxide,
which can vary over an order of magni-
tude throughout the burn of one charge
of wood.
The nonbaffled stove is an airtight,
boiler plate radiant heater with a com-
• 0.44m-
Front View
Adjust able ^
Air Inlets'
O
O
EPA-5 and
POM Trains
(Elev. 2.2 m)
I*—0.33m—H
0.42m i
O. 62m
3.36m
JO.
3.36m
^0.09 m
Side View
-0.71 m-
*—0.33 m—»f* 0.38 m
=3
Figure 1.
^•Furnace Air
Nonbaffled airtight stove showing generalized combustion-air flow
pattern.
bustion zone approximately 0.61 m
high, 0.4 m wide, and 0.71 m long. Two
air inlets are located on the door of the
stove. This unit, shown in Figure 1, is
lined with firebrick and weighs approxi-
mately 160 kg (350 Ib). The stove and its
associated flue pipe were mounted on a
mechanical scale to provide readout of
mass. Weights were taken prior to and
immediately after charging wood to de-
termine the mass of wood charged. A
weight reading was also taken at the
startup and shutdown of sampling for
emission factor calculations. Burning
rate was monitored by readings at 10-
minute intervals.
As a precaution against sampling
probes affecting weight readings, the
sampling trains were self-supporting,
never touching the stove pipe. A dif-
ference of less than 0.05 kg influence
was shown with the sampling train in
place.
Time of sampling spanned at least
one wood charge to .the stove. Steady-
state burning conditions were sampled
Table 1. High Altitude Wood Stove Test Comparison Program (Number of Samples Collected and Analyzed)
MRC-High altitude MRC-Low altitude PEDCo-High altitude
Front-half
particulates
4
Condensable
organics
2
Front-half
particulates
2
Condensable
organics
1
Front-half
particulates
1
Condensable
organics
1
i
-------
rather than stove startup or burndown
conditions, duplicating MRC-low altitude
and PEDCo-Vail studies.
Total paniculate emissions (front-half
filterable and condensable organics)
were collected by EPA Method 5. A
backup filter was inserted between the
third and fourth impinger. Particulate
emissions (front-half filterable only) and
ROMs were collected with a modified
Method 5 sampling train, which adds an
XAD-2 resin trap between the filter box
and impinger system. A cooler trap
(empty impinger) was inserted prior to
the resin' cartridge to reduce the gas
temperature to 21 °C before entry into
the resin for increased collection ef-
ficiency. Single point sampling with the
probe tip placed in the stack center was
used because flat velocity profiles have
been demonstrated on solid fuel resi-
dential heating devices. Isokinetic
sampling rates were maintained.
Results
The results of MRC's high altitude
paniculate and carbon monoxide emis-
sion testing are given in Tables 2 and 3.
Excellent control over the wood burning
rate was exercised among the four test
runs, as demonstrated by a standard
deviation for the burning rate of ±14%
of the mean value. The high altitude
tests' burning rates nearly duplicated
those at low altitude, differing by only
2.5% for the averages. Particulate
emission rates (front-half filterables)
also showed good repeatability between
runs, with a standard deviation which
was only ±28% of the mean vajue. Thus
replicate runs at a fixed burning rate will
give repeatable particulate emission
rates by EPA Method 5 for a given stove
and wood type.
The small differences in emission
factors between MRC's high and low
altitude emission test runs are attributed
to the difference in ambient air temper-
ature and, hence, stove inlet air temper-
ature. During the low altitude testing
the room temperature was 70-75°F
versus 49°F for the high altitude test.
This is shown in the large difference
between stack gas temperature at the
sampling point—493°F vs. 718°F.
Cooler stove inlet air may tend to
"quench" combustion in and above the
flame zone in the stove and increase
filterable particulate emissions to a
small degree.
A comparison of POM results is given
in Table 4. The total POM emission
factor from the high altitude test was
0.05 g/kg, which was about one-fourth
that of the low altitude test. Another
noticeable difference was that the high
altitude test results had POM compounds
identified which tended to be lower in
molecular weight. Although unconfirmed,
it is felt that the lower combustion zone
temperatures of the high altitude test,
as indicated by stack gas temperature,
suppressed the POM chemical formation
mechanisms. This would result in the
formation of lower molecular weight
POM compounds as well as reduced
total POMs on a mass basis. An even
cooler fire should, at some point, result
in zero POM emissions since very high
temperatures (at least 420-500°C) and
a chemically reducing environment are
necessary for POM formation from the
long-chain molecules present in wood
cellulose, lignin, and hemi-cellulose.
The high altitude POM emission
factor was in the same range as several
POM emission factors reported for other
stove tests in which the burning rate
was slightly lower than the MRC-low
altitude tests. In addition to POM identi-
fication and quantification, twenty-one
other high molecular weight organic
compounds were identified.
Conclusions
A comparison between the high alti-
tude and low altitude emission test
results leads to the conclusion that,
when the operating variables are held
constant, there is no statistically sig-
nificant difference in particulate or CO
emissions. The question remains, how-
ever, as to why the results of the PEDCo
emission tests are three times as high?
Table 2. Comparison Summary of Wood Stove Emission Test Conditions
Description (68°F) Units MRC-High altitude
MRC-Low altitude
PEDCo-Vail
Barometric pressure
Ambient temperature
Wood type
Burning rate
Excess air
Stack temperature
Stack flow rate (dry std)
Stack flow rate (actual)
in. Hg
°F
kg/hr
%
°F
dscfm
acfm
24.64
40
Seasoned oak
6.9
44.2
493
30.0
75.3
29.35
75-80
Seasoned oak
7.1
25.0
718
33.5
90.5
21.58
50
Seasoned pine
6.2
47.1
415
63.1
152
Table 3. Comparison Summary of Wood Stove Emission Results
Description (68° F) Units MRC-High altitude
MRC-Low altitude
PEDCo-Vail
Particulate loading (front)
Particulate loading (back)
Particulate loading (total)
Emission rate (front)
Emission rate (back)
Emission rate (total)
Emission factor (front)
Emission factor (back)
Emission factor (total)
CO emission factor
gr/dscf
gr/dscf
gr/dscf
Ib/hr
Ib/hr
Ib/hr
g/kg
9/kg
9/kg
9/kg
0.2143
0.3645
0.5788
0.056
0.085
0.141
3.73
5.33
9.06
181
0.1405
0.2426
0.3249
0.041
0.066
0.107
2.50
6.00
8.50
230.5
0.1408
0.5756
0.7164
0.076
0.311
0.387
5.56
22.74
28.30
219
3
-------
Table 4. POM Emissions from Nonbaffled Stove Burning Seasoned Oak fg/kg)
POM compounds
Biphenyl
Methyl-biphenyl
Ci-alkyl-biphenyl
Naphthalene
Methyl naphthalene
Cz-alkyl naphthalene
Anthracene/ phenanthrene
Methyl-anthracenes/ -phenanthrenes
Cz-alkyl-anthracenes/ -phenanthrenes
Cyclopenta-anthracenes/-phenanthrenes
Fluoranthene
Pyrene
Methyl- fluoranthenes/-pyrenes
Cz-alkyl-fluoranthenes/-pyrenes
Benzo(ghi)fluoranthene
Cyclopentafedfpyrene
Benzofcjphenanthrene
Benzo(a)anthracene/chrysene
Methyl-benzanthracenes-
benzophenanthrenes/-chrysenes
Ct-alkyl-benzanthracenes-
benzophenanthrenes/-chrysenes
Benzofluoranthenes
Benzopyrenes/perylene
Indenopyrene
Methyl-benzofluoranthenes/-perylene
Benzo(ghi)perylene
Dibemop yrenes
Dibenzanthracenes/ -phenanthrenes
Low
altitude tests
—
—
—
—
—
—
0.06/8
0.0/67
0.0045
0.0030
0.0208
0.0169
0.0103
—
0.0047
0.0051
0.0016
0.0076
0.0062
0.0037
0.0112
0.0084
—
—
0.0043
0.0007
0.00/0
High
altitude tests'
0.0016
0.0020
0.0003
0.0127
0.0024
0.0011
0.0100
0.0012
0.0006
—
0.0044
0.0026
0.00/4
0.0006
—
—
0.0006
0.0023
0.0004
—
0.002 /
0.0020
0.00/0
0.0003
0.0010
—
0.0003
Total
0. 1885
0.0503
Note: Blanks indicate POM compound not detected.
'Average of two runs.
A major aspect of the PEDCo-Vail
study was the absence of weight deter-
minations for the burning rate calcula-
tions. It should be pointed out that "the
amount of wood burned during a test
run could only be estimated because the
amount remaining at the end of the test
could not be weighed." A visual estimate
of the wood charge left in the stove (or
fireplace) at a run's end based on the
apparent volume will tend to bias the
error in weight determination toward
the low side, and, hence, inflate the
emission factor. The filterable panicu-
late stack gas concentration for PEDCo's
test was within the same range as
either MRC study. PEDCo's stove had a
higher stack gas flow rate, probably due
to more inlet air admitted. Tables 2 and 3
summarize the test conditions and test
results, respectively.
The PEDCo test had much greater
condensable organic emissions—0.311
Ib/hr vs. 0.085 Ib/hr for MRC's study.
This was probably due to two main
factors: PEDCo burned dry pine and
burned it at a lower temperature. The
use of pine, a more resinous wood, can
have an increased effect on emissions
over the use of a hardwood such as oak.
MRC's low altitude study determined
that condensable organic emission
doubled (to 10-12 g/kg) when burning
green pine, although this was not true of
dry pine. However, the MRC low altitude
study had greater stack gas tempera-
tures (700°F vs. 415°F), which indicates
a higher stove combustion zone tem-
perature. In the PEDCo-Vail stove test,
the stack temperature was measured at
the stove exit. The stove inlet tempera-
ture was unusually low (similar to the
MRC high altitude tests) at 50°F. Thus, it
was likely that the lower combustion
zone temperature created more unburned
volatiles which were "quenched" by the
cool inlet air rising above the combustion
zone. Also, because the weights of
unburned wood charges were estimated,
it is likely that PEDCo's reported burning
rate was lower and the emission factor
became inflated. Other researchers
have shown, with hi-vol particulate
sampling systems, that particulate
emission rates are inversely propor-
tional to burning rate.
Carbon monoxide emissions measured
by PEDCo appear to be within the range
of both MRC emission test results.
PEDCo's carbon monoxide sampling
technique was by EPA Method 3 inte-
grated gas bag collection, same as the
MRC high altitude tests, followed by
Orsat analysis. Comparison of PEDCo's
CO results with other tests conducted
on fireplace emissions showed that
stove CO emissions (in g/kg) were 2-5
times higher than fireplace CO emis-
sions, a result which MRC observed in
the low altitude testing program.
Data from other wood stove sampling
programs indicate that updraft and
sidedraft type stoves typically have CO
emission factors in the 150-400 g/kg
range. The degree of incomplete com-
bustion at high burning rates in these
types of stoves appears to remain fairly
constant over the burning period of at
least one charge of wood, although
instantaneous CO readings fluctuate
wildly. Wood stove tests at Battelle have
shown that at a high burning rate CO
and unburned hydrocarbon emissions
(measured as THC with a GC/FID) cor-
relate well; the THC/CO emission ratio
is roughly 1:5.
-------
J. A. Peters and D. G. DeAngelis are with Monsanto Research Corporation,
Dayton. OH 45407.
Paul DePercin is the EPA Project Officer (see below).
The complete report, entitled "High Altitude Testing of Residential Wood-Fired
Combustion Equipment," (Order No. PB 81-226 151; Cost: $6.50. subject to
change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U. S. Environmental Protection Agency
Cincinnati, OH 45268
* U& GOVERNMENT FRWTMaOfncemf 757-01Z/7Z97
-------
United States Center for Environmental Research Postage and
Environmental Protection Information £ees rala
Agency Cincinnati OH 45268 Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
RETURN POSTAGE GUARANTEED
Third-Class
Bulk Rate
TtRL01690h4
US EPA REGION V
LIBRARY
230 S DE.ARBURN ST
CHICAGO IL 60604
------- |