United States
Environmental Protection
Agency
National Risk Management
Research Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/SR-96/128 January 1997
SEPA Project Summary
Evaluation of Emissions from the
Open Burning of Land-Clearing
Debris
Christopher C. Lutes and Peter H. Kariher
The exposure of persons to combus-
tion emissions during land clearing has
become an issue of increasing con-
cern. This study identifies and quanti-
fies a broad range of pollutants that
are discharged during small-scale,
simulated, open combustion of land-
clearing debris and reports these emis-
sions relative to the mass of material
combusted. Two types of land-clearing
debris (representing the typical land-
clearing debris found in Florida and
Tennessee; primarily wood and other
organic debris) were combusted in a
facility designed to simulate open burn-
ing. One debris sample was also com-
busted in the same facility using a simu-
lated air curtain incinerator. Volatile,
semivolatile, and particulate-bound or-
ganics were collected and analyzed by
gas chromatography/mass spectrom-
etry. The emphasis of the analyses was
on the quantification of hazardous air
pollutants listed in Title III of the Clean
Air Act Amendments of 1990 (CAAA),
although further efforts were made to
identify and quantify other major or-
ganic components. Fixed combustion
gases (carbon dioxide, carbon monox-
ide, nitric oxide, oxygen, and total hy-
drocarbons) were monitored continu-
ously throughout the test period.
This project produced estimated
emissions data for a broad range of
atmospheric pollutants from a simu-
lated open debris combustion process.
Both air pollutant concentrations within
the facility where combustion was tak-
ing place, and estimated emissions ex-
pressed as mass of pollutant per mass
of debris material consumed by combus-
tion were reported for volatile, semivolatile,
and particulate bound organics, typical
combustion gases, and particulate. Sub-
stantial emissions of a large number of
pollutants including carbon monoxide,
PM10, PM25, benzene, acetone, toluene,
ethyl benzene, pinene, naphthalene, phe-
nol, and 14 polycyclic aromatic hydrocar-
bons were observed. QA/QC requirements
apply to this project. Data are supported
by QA/QC documentation as required by
the US EPA's QA Policy.
These tests did not provide conclu-
sive evidence of the effectiveness of
air curtain combustors in reducing
emissions. While the emissions of some
pollutants seemed to be decreased, oth-
ers were unchanged or, in a few cases,
appeared to increase.
This Project Summary was developed
by EPA's National Risk Management
Research Laboratory's Air Pollution
Prevention and Control Division, Re-
search Triangle Park, NC, to announce
key findings of the research project
that is fully documented in a separate
report of the same title (see Project
Report ordering information at back).
Introduction
Disposal of debris generated by land
clearing or landscaping has long been
problematic. Land clearing is required for
a wide variety of purposes such as con-
struction, development, and clearing after
natural disasters. The resultant debris is
primarily vegetative, but may include inor-
ganic material. Landscaping such as prun-
ing often generates similar vegetative de-
-------
bris. This debris is often collected and
disposed of by municipalities. Open burn-
ing or burning in simple air curtain incin-
erators is a common means of disposal
for these materials, and has long been a
source of concern. Air curtain incinerators
use a blower to generate a curtain of air,
(for brevity, in this document "with blower"
is used to describe tests in which an air
curtain incinerator was used) to enhance
combustion taking place in a trench or a
rectangular shaped, open-topped refrac-
tory box. For instance, in Detroit, the prob-
lem of municipal burning of brush, logs,
and stumps became so severe that in
September 1958, the mayor appointed a
committee to study this problem among
others. This eventually led to the design
and construction of a specially designed
incinerator in 1961-62 for brush and log
burning, which was more complex than
an air curtain incinerator, at a cost of
$250,000. In many locations open burning
or the use of simple air curtain incinera-
tors is still the method of choice for the
disposal of these materials.
An evaluation of literature on emissions
from open air burning of debris shows a
limited amount of information on emission
factors for specific pollutants measured in
such a way that emissions could be esti-
mated and therefore modeled. Several
similarities can be drawn from the litera-
ture reviewed. Most of the available data
focus on only a few classes of pollutants.
The list of pollutants for which emission
factors are available does not include most
of the air toxic compounds listed in the
Clean Air Act Amendments of 1990
(CAAA). However, the rough order of mag-
nitude agreement in the total particulate
and total hydrocarbon (THC) emission fac-
tors reviewed over a wide variety of source
types is notable.
Local air regulatory agencies, including
those in Tennessee and Broward County
(Florida), requested that more detailed in-
formation on the emissions from these
processes be made available. Therefore,
EPA's Control Technology Center (CTC)
steering committee proposed a research
project examining emissions from the open
burning of debris.
In response to these concerns, through
the guidance of EPA's Air Pollution Pre-
vention and Control Division (APPCD), a
study was undertaken to measure emis-
sions from the simulated open combus-
tion of land-clearing debris. This study in-
cluded replicated simulated open burning
tests of debris from Florida and Tennes-
see and replicate tests with a simulated
air curtain incinerator for the Tennessee
debris. The study was designed to collect,
identify, and quantify a wide range of air
emissions and to report these emissions
per mass of debris material combusted.
The emphasis of these analyses was on
the quantification of air toxics compounds
listed in the CAAAs, although further ef-
forts were made to identify and
semiquantify other major organic compo-
nents.
Experimental
The project consisted of a replicate study
to collect and qualitatively and quantita-
tively characterize organic and particulate
emissions resulting from the simulated
open combustion of land-clearing debris.
Small quantities (11.3 to 17.8 kg [25 to 39
Ib]) of wood, sticks, twigs, leaves, and
other organic matter were combusted in a
refractory lined pit within a test facility
designed to simulate open-combustion
conditions. Sampling was conducted within
the facility through a modified dichotomous
sampler using 142 mm filter heads for
particulate with an aerodynamic diameter
< 2.5 urn (PM2.5) and particulate with an
aerodynamic diameter < 10 |j,m (PM10).
Volatile organics were sampled using
SUMMAŽ canisters, and semivolatiles were
sampled using a PUF/XAD TO-13 sam-
pling train. A portion of the combustion
effluent was diverted to an adjacent sam-
pling facility via an induced draft duct. A
portion of the sample from the induced
draft duct was also analyzed by a series
of continuous emission monitors for car-
bon dioxide (CO2), carbon monoxide (CO),
nitric oxide (NO), oxygen (O2), and THC.
The organic constituents were analyzed
both qualitatively and quantitatively using
a gas chromatograph/mass spectrometer
(GC/MS). Measured concentrations were
related to dilution air volumes and mea-
sured net mass of debris combusted to
derive emission rates.
The EPA's Open Burning Simulation
Facility was used in this study. A simu-
lated air curtain combustor was con-
structed for the tests of this system based
on an analysis of specifications of pilot-
and full-scale units of this type.
Before each test, a sample of debris
was removed from the crate of either
Florida or Tennessee samples provided
by cooperating state and local govern-
ment personnel and placed in the refrac-
tory burn box (RBB). The wood and other
materials were arranged in the RBB to
allow for easy lighting and total consump-
tion of burn material. For these tests, 11.3
to 17.8 kg (25 to 39 Ib) of material was
placed in the RBB. Before and after each
test, or before and after each change of
sample media (if this occurred more fre-
quently), all sampling trains were leak
checked. Before the beginning of each
test day, at least 15 min of background
data were acquired on the continuous
emission monitors (CEMs), thermocouples,
and the scale platform. The burn was then
ignited by a brief application of a handheld
propane torch, which was removed before
sampling began. During a typical test, suf-
ficient combustion began after less than 5
min of torch operation. The air curtain
was started immediately after the removal
of the lighting torch in tests involving this
system. All sampling started 2-min after
removal of the torch from the burn hut.
This 2-min period was designed to ensure
exhaust of any propane combustion va-
pors.
To allow adequate time for all neces-
sary emissions samples to be obtained,
some tests had another charge of debris
added. Combustion of charge was allowed
to go to apparent completion (as signified
by unchanging weight and near back-
ground concentrations of combustion
gases) before completion of the run. Com-
bustion of one charge was allowed to go
to apparent completion before another
charge was introduced.
A "hut blank" test, in which the propane
torch was briefly introduced into the facil-
ity but no debris was combusted, was
conducted for comparison purposes. In
addition, various field and laboratory blank
samples were collected for each sampling
train.
After completion of the chemical analy-
ses, analyte concentration data were
coupled with sample volume, facility air
flow, and combustible material mass loss
data to derive estimated emissions (ex-
pressed as mass of analyte produced per
mass of debris material consumed in the
combustion process).
Results and Discussion
Estimated emissions on a mass emitted
per mass consumed by combustion basis
of CO and THC appear broadly similar for
the Tennessee and Florida materials in
the no-blower case (Table 1). These val-
ues appear to agree within a factor of two
with those measured by Gerstle and
Kemnitz for "Landscape Refuse." Esti-
mated emissions of CO and THC for the
Tennessee material appear to be little im-
pacted or at best slightly decreased by
the use of the air curtain incinerator.
Substantial emissions of PM10 and PM2.5
particulate matter were observed with both
types of debris materials combusted (Table
2). Particulate catches on a mass/volume
basis during hut blank tests were at least
tenfold lower than during any actual com-
-------
Table 1. Targeted Volatile Compounds Estimated Emissions (mg/kg)
Test No.
Sample ID
Compound
dichlorodifluoromethane
dichlorotetrafluoroethane
chloromethane
vinyl chloride
1 ,3-butadiene
bromomethane
chloroethane
trichlorofluoromethane
dichlorotrifluoroethane
trichlorotrifluoroethane
1,1-dichloroethene
acetone
carbon disulfide
methylene chloride
3-methylpentane
1, 1-dichloroethane
butyl methylether
cis-1 ,2-dichloroethene
2-butanone
ethyl acetate
chloroform
1, 1,1-trichloroethane
carbon tetrachloride
benzene
1,2-dichloroethane
trichloroethene
1,2-dichloropropane
cis- 1 ,3-dichloropropene
dimethyl disulfide
4-methyl-2-pentanone
octane
toluene
trans-1 ,3-dichloropropene
1, 1,2-trichloroethane
tetrachloroethene
butyl acetate
1,2-dibromoethane
chlorobenzene
nonane
ethyl benzene
m,p-xylene
o-xylene
styrene
pinene
1, 1 ,2,2-tetrachloroethane
decane
4-ethyltoluene
1 , 3,5-trimethylbenzene
1,2,4-trimethylbenzene
limonene
1 , 3-dichlorobenzene
1 , 4-dichlorobenzene
benzyl chloride
undecane
1 ,2-dichlorobenzene
dodecane
1,2,4-trichlorobenzene
hexachlorobutadiene
naphthalene
1
TN
nb
<2
<7
6
<1
141
<2
<2
<3
<3
<4
<2
224
<1
4
<2
<2
<2
<2
42
42
<2
<3
<3
346
<2
<3
<2
<2
<2
<2
8
207
<2
<3
<3
<2
<4
<2
<2
37
89
21
76
54
<3
<3
29
5
18
99
<3
<3
2
4
<3
4
<3
<5
69
2
TN
nb
<3
<8
6
<1
116
<2
<2
<3
<3
<4
<2
198
<2
4
<2
<2
<2
16
36
36
<3
<3
<4
325
<2
<3
<2
<2
<2
<2
6
179
5
<3
<4
<3
<4
<2
<3
29
70
17
70
137
<4
<3
23
4
14
84
<3
<3
<3
4
<3
<4
<4
<6
73
3
FL
nb
<2
<6
133
<1
108
2
<1
<2
<3
<3
<2
209
<1
<2
<2
<2
<2
31
40
40
<2
<2
<3
258
<2
<2
<2
<2
<2
<2
5
147
<2
<2
<3
<2
<3
<2
<2
21
46
15
40
<2
<3
<3
12
3
11
<2
<3
<3
<2
<3
<3
<3
<3
<5
48
4
FL
nb
<2
<6
55
<1
41
<2
<1
<2
<2
<3
<2
84
<1
2
<1
<2
2
33
16
16
<2
<2
<3
132
<2
<2
<2
<2
<2
<2
3
65
<2
<2
<3
<2
<3
<2
<2
9
18
7
17
<2
<3
<2
5
<2
4
<2
<2
<2
<2
<3
<2
<3
<3
<4
24
5
Hut
Blank
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
6
TN
wb
<1
<4
4
<1
135
<1
<1
<1
<2
<2
<1
180
<1
<2
<1
<1
<1
28
30
30
<1
<1
<2
273
<1
<1
<1
<1
<1
<1
<1
165
<1
<1
<2
<1
<2
<1
<1
27
86
17
59
80
<2
<2
27
4
15
51
<2
<2
2
2
<2
2
<2
<3
42
7
TN
wb
<2
<4
5
<1
140
<1
<1
<2
<2
<2
<1
123
<1
<2
<1
<1
<1
13
19
19
<2
<2
<2
270
<1
<2
<1
<1
<1
<1
5
212
<1
<2
<2
<1
<2
<1
<1
35
151
19
86
124
<2
2
51
5
25
92
<2
<2
3
6
<2
4
<2
<3
53
8
Hut
Blank
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
nb = no blower, wb = with blower, NA = not applicable, nd = not detected.
-------
Table 2. Particulate Data
Test
No.
1
2
3
4
5
6
7
8
Test Conditions
TN No Blower
TN No Blower
FL No Blower
FL No Blower
Hut Blank
TN With Blower
TN With Blower
Hut Blank
Concentration
PM25
(mg/m3)
30.51
18.75
3.95
11.63
0.11
45.15
35.73
0.07
Concentration
PM,0
(mg/m3)
36.30
19.13
17.54
11.90
0.29
45.77
37.82
0.26
Estimated
Emission PM25
(g/kg)
14.13
10.04
1.75
4.56
NA
12.07
8.33
NA
Estimated
Emission PM10
(g/kg)
16.81
10.25
7.75
4.66
NA
12.23
8.82
NA
bustion test (Table 2). This indicates that
most of the participate collected was ac-
tual combustion emissions and not par-
ticulate being resuspended from the burn
hut walls or present in the ambient air fed
into the facility. Estimated emissions on a
mass particulate per mass material com-
busted basis from the Tennessee material
appeared to be substantially higher than
those from the Florida material. The Ten-
nessee material without the blower gave
fairly consistent values in replicate tests.
The Tennessee material with the blower,
in one case, gave a value that appeared
similar to the value without the blower. In
the next (duplicate) test, it gave values
somewhat lower than those typical with-
out the blower. However, in this test the
sample was obtained for only a short pe-
riod due to an equipment malfunction, and
the flow rate did not meet data quality
indicator goals. In other tests, data quality
was acceptable for this measurement. In
almost all cases, regardless of source of
material or use of blower, most of the
PM10 appears to be composed of very fine
material (<2.5 |im diameter). This is an
important observation because many be-
lieve that fine particulate is more strongly
associated with health effects than coarse
particulate. Our average estimated PM10
emissions agree within +25% with those
measured by Gerstle and Kimnitz for total
particulate, perhaps due to this predomi-
nance of fine particulate.
The volatile organic data set produced
from these tests included concentration
measurements for more than 55 targeted
and several dozen tentatively identified
species. Targeted species are defined as
those for which the analytical instrument
was calibrated. Tentatively identified spe-
cies are other compounds found in the
sample that can be tentatively identified
through searches of mass spectral librar-
ies checked by investigator examination
of the mass spectral match. Approximately
19 of the targeted species were consis-
tently detectable. Results of the volatiles
analyses of the targeted analytes are pre-
sented in Table 1 as estimated emissions
on a mass of pollutant per mass of mate-
rial consumed by combustion basis.
Various hydrocarbon, aromatic, and oxy-
genated species, such as benzene, ac-
etone, toluene, ethyl benzene, m, p-xy-
lene, pinene, limonene, naphthalene, and
styrene, were among the highest concen-
tration targeted volatiles observed. In gen-
eral, emissions of these species were
higher with the Tennessee material than
with the Florida material. This trend was
most dramatic for pinene and limonene,
two compounds that belong to the ter-
pene group that is often isolated from
plants.
The data set is inconclusive on the ef-
fect of the air curtain incinerator on volatiles
emissions. Emissions of many compounds
appear unchanged and, while some spe-
cies appear to be emitted at a lower rate
with the air curtain in operation, emissions
of others may be increased.
Alkenes, ketones, heteroaromatics, and
alkyl-substituted aromatics are prominent
among the tentatively identified volatile
compounds.
More then 100 semivolatile species were
targeted in these analyses. Approximately
23 of these species were consistently de-
tected in the combustion samples at lev-
els significantly above blank levels. Of
these 23 species, 14 are polycyclic aro-
matic hydrocarbons (PAHs). These have
been detected in numerous studies of
wood combustion, so their appearance in
a study of the combustion of land-clearing
debris is expected. The range of estimated
emissions reported in this document agree
broadly with those reported by Cooper for
various PAH species from wood combus-
tion in fireplaces. Of the 23 species de-
tected, 4 were phenol and its methyl sub-
stituted derivatives. Phenols have also
been previously established as wood com-
bustion byproducts. The values measured
here for estimated emissions of phenol
are slightly higher then those measured
by Cooper for wood combustion in fire-
places. The remaining five consistently
detected species were biphenyl, styrene,
cumene, 2-methylnaphthalene, and
dibenzofuran.
The results of the tests without the air
curtain incinerator showed that concentra-
tions of individual semivolatile species were
usually similar for the Florida and Tennes-
see materials, but a few species were
emitted at a moderately higher rate from
combustion of the Tennessee material. A
brief analysis of this data set suggests
that, for most semivolatile species, no dis-
cernible difference in emission factor be-
tween the with and without air curtain in-
cinerator tests can be observed. However
for a few species, such as pyrene,
benzo(a)pyrene, and biphenyl, use of the
air curtain does appear to reduce emis-
sions.
Numerous tentatively identified species
were also identified in the semivolatile
analyses. These species consist primarily
of alkylated and oxygenated aromatics,
heteroaromatics, and polyaromatics.
Summary and Conclusions
This project produced estimated emis-
sions data for a broad range of atmo-
spheric pollutants from a simulated open
-------
debris combustion process. Both air pol-
lutant concentrations within the facility
where combustion was taking place and
estimated emissions expressed as mass
of pollutant per mass of debris material
consumed by combustion were reported
for volatile, semivolatile, and particulate-
bound organics, typical combustion gases,
and particulate. Substantial emissions of
a large number of pollutants including CO,
PM,n, PM,,, benzene, acetone, toluene,
ethyl benzene, pinene, naphthalene, phe-
nol, and 14 PAHs were observed. These
tests did not provide conclusive evidence
regarding the effectiveness of air curtain
blowers in reducing emissions. While the
emissions of some pollutants seemed to
be decreased slightly, others were un-
changed or, even in a few cases, ap-
peared to increase. A definitive assess-
ment of the value of the air curtain device
cannot be made without a detailed statisti-
cal and relative risk analysis. Measure-
ments of a variety of pollutants in the
emissions of full-scale models of this de-
vice operating under realistic work site
conditions would also be helpful.
This project has yielded estimated emis-
sions values for open debris combustion
processes that can be used to assess the
risks of these processes.
-------
Christopher C. Lutes and Peter H. Kariher are with Acurex Environmental Corp.,
Research Triangle Park, NC 27709.
Paul M. Lemieux is the EPA Project Officer (see below).
The complete report, entitled "Evaluation of Emissions from the Open Burning of
Land-Clearing Debris," (Order No. PB97-115356; Cost: $28.00, 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:
Air Pollution Prevention and Control Division
National Risk Management Research Laboratory
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use $300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/600/SR-96/128
------- |