United States Industrial Environmental Research EPA-600/2-78-004o
Environmental Protection Laboratory June 1978
Agency Research Triangle Park NC 27711
Research and Development
Source
Assessment
Coal-fired
Combustion
Equipment
Tests, June 1977
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-78-004o
June 1978
Source Assessment: Coal-fired
Residential Combustion Equipment
Field Tests, June 1977
by
D.G. DeAngelis and R.B. Reznik
Monsanto Research Corporation
1515 Nicholas Road
Dayton, Ohio 45407
Contract No. 68-02-1874
ROAPNo. 21AXM-071
Program Element No. 1AB015
EPA Project Officer: Ronald A. Venezia
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
The Industrial Environmental Research Laboratory (IERL) of the
U.S. Environmental Protection Agency (EPA) has the responsibility
for insuring that pollution control technology is available for
stationary sources to meet the requirements of the Clean Air Act,
the Federal Water Pollution Control Act, and solid waste legisla-
tion. If control technology is unavailable, inadequate, or
uneconomical, then financial support is provided for the develop-
ment of the needed control techniques for industrial and extrac-
tive process industries. Approaches considered include: process
modifications, feedstock modifications, add-on control devices,
and complete process substitution. The scale of the control
technology programs ranges from bench- to full-scale demonstra-
tion plants.
Monsanto Research Corporation (MRC) has contracted with EPA to
investigate the environmental impact of various industries which
represent so jes of pollution in accordance with EPA's responsi-
bility as outlined above. Dr. Robert C. Binning serves as MRC
Program Mana tr in this overall program entitled "Source Assess-
ment," which includes the investigation of sources in each of
four categories: combustion, organic materials, inorganic mate-
rials, and open sources. Dr. Dale A. Denny of the Industrial
Processes Division at Research Triangle Park serves as EPA Pro-
ject Officer. Reports prepared in this program are of three
types: Source Assessment Documents, State-of-the-Art Reports,
and Special Project Reports.
Source Assessment Documents contain data on emissions from
specific industries. Such data are gathered from literature,
government agencies, and cooperating companies. Sampling and
analysis are also performed by the contractor when the available
information does not adequately characterize the source emissions.
These documents contain all of the information necessary for IERL
to decide whether emissions reduction is necessary.
State-of-the-Art Reports include data on emissions from specific
industries which are also gathered from the literature, govern-
ment agencies and cooperating companies. However, no extensive
sampling is conducted by the contractor for such industries.
Results from such studies are published as State-of-the-Art
Reports for potential utility by the government, industry, and
others having specific needs and interests.
ii
-------
Special projects provide specific information or services which
are applicable to a number of source types or have special
utility to EPA but are not part of a particular source assessment
study. This special project report, "Source Assessment: Coal-
Fired Residential Combustion Equipment Sampling Tests," was
prepared to provide a general characterization of air emissions
from the residential combustion of coal. In this study, Dr.
Ronald A. Venezia of the Chemical Processes Branch served as EPA
Task Officer.
iii
-------
ABSTRACT
This report presents the results of a study conducted to char-
acterize atmospheric emissions from coal-fired residential heat-
ing equipment. Flue gases from a hot-water boiler and a warm-air
furnace, each rated at about 200,000 Btu/hr, were sampled during
the combustion of three western coals. Resulting samples were
analyzed for particulates, sulfur oxides, nitrogen oxides, car-
bon monoxide, individual elements, and organic species.
Particulate emissions factors from the warm-air furnace were
found to be about an order of magnitude higher than those from
the boiler while burning a high volatile western coal. High
volatile coals with high free swelling index produced the highest
particulate emission factors. No correlation was observed
between particulate emission factors and the ash content of the
coals fired. The composition of particulate emissions was
primarily carbon, rather than elements present in the ash. In
most cases, missions of individual elements amounted to less
than 5% of the elemental content of the coal burned.
The low-fire, or "off", portion of a typical heating cycle made
a significant contribution to the total emissions from the com-
bustion equipment. In the case of polynuclear organic materials
(POM's) the greatest contribution came during the off period.
Over 50 organic species, including many POM's, were identified
in the organic material collected from the flue gas during com-
bustion of high volatile coals. Maximum POM emission factors
occurred when high volatile coals were burned at low excess air
levels.
Results of this study are compared with those of a similar study
made by Battelle Memorial Institute. Both studies are found to
be in overall agreement.
This report was submitted in partial fulfillment of Contract No.
68-02-1874 by Monsanto Research Corporation under the sponsor-
ship of the U.S. Environmental Protection Agency. The study
described in this report covers the period September 1976 to
April 1978.
-------
CONTENTS
Preface ii
Abstract iv
Figures vi
Tables viii
Abbreviations x
Acknowledgments xi
1. Introduction 1
2. Summary and Conclusions 3
3. Test Program and Procedures 8
Design of test program 8
Test conditions 10
Test site and facilities 12
Test coal 16
Sampling methods and equipment 21
Analytical procedures 34
4. Results 43
Emissions . 43
Ash residue and leachate 57
Discussion of results 62
References 71
Appendices
A. Determination of Average Heating Cycle 74
B. Determination of Uncertainty Caused by Low Stack
Gas Velocity 77
Glossary 80
Conversion Factors and Metric Prefixes 82
-------
FIGURES
Number
1 Will-Burt stoker assembly * 13
2 Long Arrow boiler/Will-Burt stoker hot-water
heating system and sampling port locations. . . 15
3 Hardin furnace/stoker forced-air heating system
and sampling port locations 17
4 Barometric damper employed on coal-fired
combustion equipment to equalize exhaust stack
pressure drops 18
5 Schematic of EPA Method 5 sampling train for
particulate and condensable organic material
collection 22
6 Expand 1 view of a typical Drager gas
detector tube 24
7 Schematic of Source Assessment Sampling System. . 26
8 Sample handling and transfer-nozzle, probe,
cyclones, and filter 28
9 Sampling handling and transfer - XAD-2 module . . 30
10 Sample handling and transfer - impingers 31
11 POM train components and sample recovery
procedure employing modified Method 5
equipment 33
12 SASS train component separation and analysis
scheme 36
13 Organic component analysis flow diagram 40
14 Effect of coal ash content on boiler particulate
emissions 46
15 Effect of coal volatile content on boiler
particulate emissions 47
16 Effect of free swelling index on boiler
particulate emissions 47
17 Effect of coal ash content on condensable organic
emissions 50
vx
-------
FIGURES (continued)
Number Pag(
18 Effect of coal sulfur content on SOX emissions. . 51
19 Effect of coal volatile content on POM
emission 53
20 Effect of excess air on POM emissions 53
21 Ash clinker containing partially combusted coal
from boiler 58
22 Ash clinkers and rocks found in boiler
bottom ash 58
23 The effect of heating cycle variations on
emission factors 68
vii
-------
TABLES
Number
Average Emission Factors for Coal-fired
Residential Combustion Equipment Operating on
a 20-min ON/40-min OFF Heating Cycle ,-....
Average Emission Rates for the ON and OFF
Heating Cycle Segments of the Warm-air
Furnace Burning Coal B
Emission Sampling Program for Coal-fired
Residential Combustion Equipment .......
4 Proximate and Ultimate Analyses, Free Swelling
Index, and Ash Fusion Temperatures for Test
Coals Compared with 86 Rocky Mountain Province
Coal Samples (Proximate and Ultimate Only) . . 19
5 Elemental. Analysis of Two Test Coals Compared
with 124 Rocky Mountain Province Coal
Samples . 20
6 SASS Train Impinger System 27
7 Emission Factors for POM and Criteria Pollutants
from Coal-fired Residential Heating Equipment
Operated on a 20-min ON/40-min OFF Heating
Cycle 44
8 Experimental Data for the Coal-fired Heating
Equipment Operated on a 20-min ON/40-min OFF
Cycle 45
9 Size Distribution of Particulate Emissions
During ON Segment of a Coal-fired Residential
Furnace Heating Cycle 48
10 Carbon, Hydrogen, and Nitrogen Content of
Particulate Emissions from Coal-fired
Residential Heating Systems 49
11 Elemental Emission Factors from Coal-fired
Residential Heating Equipment Operated on a
20-min ON/40-min OFF Heating Cycle 55
12 Emission Factors for C-j to Ci6 Hydrocarbons
During the ON Segment of a Coal-fired
Residential Furnace Heating Cycle. ...... 55
viii
-------
TABLES (continued)
Number
13 Emission Factors of Major Organic Species from
Residential Combustion of Coal 56
14 Ash Residue from Coal B Burned in the Warm-air
Furnace 59
15 Elements Emitted from the Coal-fired Warm-air
Furnace as Bottom Ash 60
16 Fraction of Coal Elemental Content Emitted to
the Atmosphere and Total Material
Balance 61
17 Elements Leached from Ash Removed from the
Coal-fired Warm-air Furnace 63
18 Average Emission Factors from the Combustion of
Coal B in the Residential Boiler as Compared
to the Residential Furnace 64
19 Comparison of Emissions from the ON and OFF
Segments of the Warm-air Furnace Heating Cycle
While Burning Coal B . 65
20 Comparison of Emission Data from the SASS
Train to Conventional Sampling Methods .... 69
3X
-------
ABBREVIATIONS
AA atomic absorption
acf actual cubic feet
acfh actual cubic feet per hour
amu atomic mass units
cfm cubic feet per minute
dscf dry standard cubic feet
El election impact
FSI free swelling index
GC-MS gas chromatograph-mass spectrometer
ICAP inducti/ely coupled argon plasma
KD Kuderr i-Danish
AP pressure drop in inches of water
PCB polychlorinated biphenyl
POM polynuclear organic material
ppm parts per million
SASS Source Assessment Sampling System
SIM selected ion mode
-------
ACKNOWLEDGMENTS
The authors wish to acknowledge the assistance of the EPA Task
Officer, Dr. Ronald A. Venezia, and Terry Thoem of EPA Region
VIII who participated in planning this program.
Special acknowledgment is due J. Scott Kinsey, Air Pollution
Specialist with the Colorado Department of Health, for locating
a test site and arranging for combustion equipment to be provided.
Mr. Kinsey and his staff also performed preliminary tests on the
equipment and provided laboratory analysis of a large portion of
the samples.
The authors are indebted to Mr. and Mrs. John O'Brien of Solid
Fuel Systems, Inc. and Don Young and Thomas Neill of Engineered
Products Company for their cooperation in providing a test site
and combustion equipment, and for donating their services in
operating and maintaining this equipment.
XI
-------
SECTION 1
INTRODUCTION
The majority of homes within the United States are now heated by
either natural gas (56%) or fuel oil (24%). These fuels have
traditionally been inexpensive, easy to obtain, clean, and simple
to burn in home furnaces. However, as natural gas and oil
reserves diminish, it has become apparent that alternate energy
sources must be developed for the residential sector. One
candidate for a replacement fuel is coal.
Coal is the nation's most plentiful fuel resource, and in the
past it was the predominant home heating fuel. At the present
time about 1% of the nation's homes are heated with coal. Recent
advances in furnace design have made coal-fired residential fur-
naces easier to operate and nearly automatic. Consequently, the
use of coal for home heating is attractive both economically and
technically, and there are indications in some regions of the
country of a trend toward increased home heating with coal.
One major drawback to residential coal combustion is an increased
level of air pollution. Both natural gas and fuel oil are
cleaner fuels because they produce lower levels of pollutants
when burned in home heating devices. As a result, both regional
and national EPA officials have become concerned over the poten-
tial environmental impact from increased residential coal com-
bustion. A major changeover from natural gas or fuel oil to coal
could produce a dramatic adverse effect on local air quality, and
the EPA is responsible for averting or minimizing such effects.
The major problem confronting the EPA has been the absence of an
adequate data base upon which to make policy decisions. Although
air emissions from combustion of coal have been extensively char-
acterized for larger units, such as electric utility boilers,
relatively little work has been done to measure air emissions
from coal-fired residential furnaces. Until these emissions are
adequately characterized, the EPA will lack sufficient informa-
tion to formulate policies regarding an increased usage of coal
in the residential sector.
The objective of this special project was to assemble an emis-
sions data base for coal-fired residential combustion. A
stoker-fed boiler and a stoker-fed warm-air furnace were tested
while burning western coals. Two types of coals were prepared
at three ash levels and tested in the boiler. One of these
-------
coal types was also tested In the warm-air furnace. A third,
cleaner burning western coal was tested to a limited extent in
the furnace. Exhaust gases were measured for particulates, sul-
fur dioxide (S02)/ nitrogen oxides (NOX), carbon monoxide (CO),
organic species including polynuclear organic materials (POM's),
and individual elements. Testing was performed in Denver,
Colorado, during June 1977, and samples were analyzed later in
the year.
-------
SECTION 2
SUMMARY AND CONCLUSIONS
Renewed interest in coal as a residential heating fuel has raised
concern over the subsequent effect on the environment. This
report presents results of a study conducted by the U.S. Environ-
mental Protection Agency, Monsanto Research Corporation, and the
Colorado State Health Department to quantify criteria pollutants
and characterize other emissions from coal-fired residential
heating equipment.
Emission testing was conducted on a warm-air furnace and a hot-
water boiler each rated at about 200,000 Btu/hr. Both units were
coupled with automatic stokers to feed and burn the test coal.
Emission testing was conducted during the combustion of three
western coals (designated Coals A, B, and C), of which two (Coals
A and B) were burned in the boiler and two (Coals B and C) in the
furnace. All tests were conducted with the stokers operating on
a 20-minute ON/40-minute OFF cycle. The ON and OFF portions of
the cycle correspond respectively to "high-fire" and "low-fire"
conditions in the fuel bed. During the OFF segment the stoker
does not operate, but residual coal in the fuel bed continues to
burn.
Samples collected from the boiler using particulate trains were
integrated over the total heating cycle, while corresponding
samples from the furnace were collected from cycle segments
independently.
Quantitative testing employed EPA methods from the Federal
Register to measure particulates, sulfur oxides (SOX), nitrogen
oxides (NOX)/ and condensable organics. Semiquantitative test-
ing was performed using modified EPA methods and a Source Assess-
ment Sampling System (SASS train) to measure POM's, polychlorin-
ated biphenyls (PCB's), organic species, and 21 elemental species.
Combustion residues were analyzed for their elemental content,
relative leachability of elements, and ash content.
Average emission factors for the combustion equipment tested in
this program are presented in Table 1 for a 20-minute ON/40-
minute OFF heating cycle. Table 2 presents average emission
rates for each cycle segment of the warm-air furnace. Results
indicate that emission rates are greater during the ON segment
except in the case of POM's where emission rates are higher for
the OFF segments.
3
-------
TABLE 1. AVERAGE EMISSION FACTORS FOR COAL-FIRED RESIDENTIAL
COMBUSTION EQUIPMENT OPERATING ON A 20-MIN ON/
40-MIN OFF HEATING CYCLE
(Ib/ton)
Boiler
Emission species
Particulate
SOX
NOX
CO
Condensable organics
POM
Elements:
Aluminum
Antimony
Arsenic
Barium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Tin
Titanium
Vanadium
Zinc
Coal A
6.6,(0.28).
8.8?<0.37)°
9.2D(0.39)D
1.6 (0.07)
2.2 (0.09)
0.4 (0.02)
0.014 ,
<0.001Q
0.0008
0.0006
0.004 .
<0.004
0.024 ,
-------
TABLE 2. AVERAGE EMISSION RATES FOR THE ON AND OFF HEATING CYCLE
SEGMENTS OF THE WARM-AIR FURNACE BURNING COAL B
(10-3 Ib/hr)
Heating cycle segment
Emission species
Particulates
SOX
NOX
CO
POM
Condensable organics
Elements:
Aluminum
Antimony
Arsenic
Barium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Tin
Titanium
Vanadium
Zinc
ON
183
167
61
127
0.057
28
0.85 .
<0.048D
0.019
0.055
0.15 .
<0.0007D
1.8
0.017
0.007
0.031
1.3
<0.016D
0.64
0.016
<0. 00007°
<0.083b
<0.048b
<0.069b
0.007
0.42
<0.26D
0.12
0.024
<0.21b
0.097
0.013
0.16
OFF
36
10
12
31
0.17
13
0.26.
<0.12b
0.006
0.020
0.051.
<0.005D
0.13
0.021
0.011
<0.003b
0.15
<0.081D
0.036
<0.001D
0.004
<0.030D
0.13 .
<0.071b
<0.012b
0.50
<0.067b
<0.052b
<0.003b
<0.33b
<0.010b
0.013
<0.018b
Emissions are presented as pounds per hour
rather than pounds per ton because quantify-
ing the amount of coal combusted during
each cycle segment was not possible.
Value is the detection limit.
organics in the form of insoluble tars and resins or carbon
black-like material.
Sulfur oxide emissions increased with rising coal sulfur content.
However, there was evidence that a high calcium coal produced
lower SOX emissions than expected based on coal sulfur content.
Emissions of NOx were independent of major variables. Carbon
monoxide emission rates were highly variable and appeared to be
sensitive to changing fuel bed conditions.
-------
Emissions of POM's were found to increase as coal volatile con-
tent increased. High excess air levels in the combustion equip-
ment had a tendency to reduce POM emissions. Polychlorinated
biphenyls were below the detection limit of 1.0 x 10~5 Ib/ton in
the stack gas.
Characterization of the organic material present in the stack gas
identified 61 organic compounds, 20 of which were POM's. A large
portion of the organic material was found to be aliphatic hydro-
carbons between C^ and C33, napthalene compounds, and phenols.
Emissions of each elemental constituent of coal in the flue gas
were, in most cases, less than 5% of the quantity of that element
in the coal feed. Emissions of individual elements to the air
were highest for aluminum, calcium, iron and silicon (0.16 Ib/ton
to 0.58 Ib/ton). Analysis of the combustion residue (bottom ash)
showed the highest elemental concentrations to be aluminum, cal-
cium, iron, and magnesium (greater than 2.0 Ib/ton of residue).
Material balance calculations for individual elements revealed
that on the average over 50% of the coal elemental impact was
accounted for in the stack gas emissions plus the bottom ash
residue. Additionally it was determined that unburned coal
amounted to up to 50% of the combustion residues of some coals.
Leaching tests on the ash residue with distilled water showed
that aluminum, ba: ium, calcium, silicon, sodium, and strontium
were leached in the,greatest quantities (0.02 Ib/ton of ash to
12 Ib/ton of ash). Elements most susceptible to leaching were
calcium, mercury, sodium, selenium, silicon, and strontium, with
20% to 95% of each leached from the residue.
Certain results and conclusions from this study deserve special
emphasis. The combustion efficiency for coal-fired residential
units is lower than that for coal-fired utility boilers or
industrial boilers. This is evidenced by the higher level of
organic emissions (including POM's), the carbonaceous nature of
the particulate emissions, and the carbon content of the ash.
As a result, an increase in the number of homes using coal for
heating may have deleterious effects on local air quality.
Major variables that affect emission rates are combustion equip-
ment design and coal composition. For example, particulate and
CO emissions were higher from the furnace than from the boiler,
while sulfur oxide emissions increased with increasing coal sul-
fur content. The most unexpected result was that particulate
emissions were not a function of the coal ash content, but instead
correlated with coal v >.latile content and free swelling index.
This behavior can be understood by considering the composition of
the particulate emissions (approximately 80% carbon). Unlike
those from large coal-fired units, the emissions do not arise
from inorganic matter contained in the coal; instead particles
are formed from coal dust and condensed organic materials.
-------
Further studies on residential combustion of coal would be use-
ful. In particular it would be helpful to characterize emissions
over a broader range of coal types and for a wider variety of
combustion devices.
-------
SECTION 3
TEST PROGRAM AND PROCEDURES
DESIGN OF TEST PROGRAM
This project was designed to measure emission rates of selected
criteria pollutants (SOx/ NOX/ CO and particulates) from typical
coal-fired residential combustion units and to characterize other
possible emissions. The criteria pollutant emission rate deter-
minations were obtained in part to provide the Colorado State
Health Department with a data base for local coal-fired residen-
tial combustion. In addition other species were measured as part
of a larger EPA effort on the environmental assessment of combus-
tion sources. Table 3 presents a summary of the test program
including run number, test parameters, and emission species
sampled. Run numbers 1-12 were conducted on a coal-fired boiler;
runs 13-29 were conducted on a coal-fired furnace.
Two coal-fired cc^nustion units were selected for testing in this
program: 1) a 20 ")00-Btu/hra Long Arrow Model 101 hot-water
boiler combined w_tn a Model S30 Will-Burt stoker, and 2) a
200,000-Btu/hr Harbin warm-air furnace/stoker. The design of
these units is typical of those currently on the market for
residential heating. The Long Arrow boiler is one of several
types of coal-fired residential boilers in current use. However,
the stoker rather than the boiler is the source of emissions in a
residential furnace. In this case the Will-Burt stoker that
feeds the coal and contains the fuel bed has been on the market
for many years and is representative of most underfeed stokers
employed for residential boilers (1). The Will-Burt stoker, in
addition, is capable of firing most coal except anthracite (2),
while the Hardin stoker is usually limited to firing subbitumi-
nous or lignite coals. The Hardin forced-air rates during each
Nonmetric units are used, when they are the most common form,
for greater utility.
(1) Personal communication with John S. Kinsey, Colorado Depart-
ment of Health, Penver, Colorado, 1 September 1976.
(2) Giammar, R. D., R. B. Engdahl, and R. E. Barrett. Emissions
from Residential and Small Commercial Stoker-Coal-fired
Boilers Under Smokeless Operation. EPA 600/7-76-029, U.S.
Environmental Protection Agency, Research Triangle Park,
North Carolina, October 1976. 77 pp.
8
-------
TABLE 3. EMISSION SAMPLING PROGRAM FOR COAL-FIRED
RESIDENTIAL COMBUSTION EQUIPMENT
Test
run
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Combustion
equipment
Boiler
X
X
X
X
X
X
X
X
X
X
X
X
Furnace
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Test coal
Designation8
ABC
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Ash level
Low Medium High
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Heating cycle
segment sampled
ON OFF Total
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Sample types collected
Particulates
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
SOx
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
NC*
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
CO POM
X X
X
X
X
X X
X
X
X
X
X
X
X
X
X X
X X
X X
X
X
X
X
X X
X X
X X
X X
X
X
X
X
Condensable
organics Elements SASS
X
X
X
X
X
X
X
X
X
X
X
X
x )
x } x
X
X
X \
x | x
X
X
X
X
X
X
X
X
X
X
X
Coal A rank: high volatile B bituminous; Coal B rank: high volatile C bituminous; Coal C rank: subbituminous C.
-------
furnace plus stoker is representative of most coal-fired forced-
air units on the market today. These units are all of the same
basic design with only minor cosmetic differences (3).
In terms of heat output, coal-fired units for residential heating
are rated near either 100,000 Btu/hr or 200,000 Btu/hr. Larger
units heat houses with floor space greater than 2,000 ft2. Both
units sampled in this program are in the larger size range. A
complete discussion of the test equipment appears later in this
section.
Two coals were chosen for testing at the beginning of this pro-
gram and a third was added during testing. The two original
coals were in the bituminous ranking and were obtained washed and
unwashed so that each coal was available at two ash levels.
Portions of the washed and unwashed coals were blended together
to produce a third ash level. The third coal was a cleaner
burning subbitiminous coal introduced late in the program for
limited testing. A more detailed discussion of the test coals
appears later in this section.
TEST CONDITIONS
Combustion Cycle
Characteristic o esidential heating units is the thermostat-
ically controller heating cycle sometimes referred to as the use
burning cycle. \ jn the thermostat senses a drop in temperature,
i.e., a demand for heat, the unit turns on and fuel is fed to the
burner. When demand for heat is satisfied the unit turns off and
the flow of fuel ceases.
Coal-fired residential heating units differ somewhat from gas-
and oil-fired units in that air must be provided by a fan to
maintain combustion in the fuel bed. Another difference is that
residual coal continues to burn during the OFF segment even
though no fuel is fed and the combustion fan is off. Combustion
during the OFF segment of a heating cycle is at a much slower
rate than during the ON segment because combustion is maintained
only by natural draft from air leaks in the unit or draft vents
in the combustion chamber door. The ON and OFF segments are
sometimes referred to as high-fire and low-fire conditions,
respectively.
The combustion cycle is a signficant factor influencing emission
rates from coal-fired residential combustion units. Emission
data from similar units show significant differences in emission
rates during the ON u^d OFF segments for the same emission
species (2). It was decided therefore to quantify the emission
(3) Personal communication with John S. Kinsey, Colorado Depart-
ment of Health, Denver, Colorado, 16 May 1977.
10
-------
rates during each cycle segment as one part of the test program.
Because of program constraints, separate sampling of ON and OFF
segments was limited to tests on the warm-air furnace. Emissions
from the boiler were measured for the total heating cycle.
The ratio of heating unit ON/OFF time varies with the heating
season. Thus the coldest months require a longer ON period
while the warmer months require a longer OFF period. For coal-
fired residential heating units this ratio is estimated to be
1:2 during the average portion of the heating season in moderate
climates (see Appendix A). This corresponds to a total of 20 min
out of every hour of forced combustion. It was desired to employ
a similar ratio of ON time to OFF time during this program.
Because testing was conducted in June the cycle had to be
artificially induced with a timer.
The timer was set for a 20-min ON/40-min OFF heating cycle; that
is, high-fire conditions were maintained continuously for 20 min
and then low-fire conditions were maintained continuously for
another 40 min to make up each 1-hr cycle. Such a cycle does not
duplicate the fluctuations in thermostatically controlled resi-
dential heating. However, it was necessary to maintain a con-
stant combustion cycle during the entire sampling program in
order to obtain reproducible results that could be meaningfully
compared when other test parameters were varied. Battelle's
experience in the sampling of similar units confirmed this pro-
cedure (4) .
Fuel Feed Rate
Coal feed rate during testing was set according to the maximum
amount of coal that could be effectively burned in each unit
tested. The coal-fired boiler burned bituminous coals (Coals A
and B) effectively at a feed rate of about 20 Ib/hr during
stoker operation. However, because the stoker only operated
20 min out of each hour, actual average feed during each elapsed
hour was approximately 6.6 Ib/hr. Subbituminous coal (Coal C)
was not tested in the boiler because of program limitations.
The coal-fired furnace burned subbituminous coal (Coal C) and
high volatile C bituminous coal (Coal B) effectively at a feed
rate of about 15 Ib/hr, or about 5 Ib/hr of elapsed time.
Attempts to effectively burn high volatile B bituminous coal
(Coal A) in the furnace were unsuccessful. Combustion was slow
and agglomeration prevented much of the coal from being burned.
Coal feed rates for each coal type were measured by adding a
known weight of coal to the empty stoker hopper at the beginning
of each run and weighing the residual coal after each run.
(4) Personal communication with Robert D, Giammar, Battelle-
Columbus Laboratories, Columbus, Ohio, 9 June 1977.
11
-------
Whenever the test plan called for a change in coal type, the new
coal was burned for a minimum of 1 hr before testing resumed in
order to purge the equipment of the previous coal.
Fuel Bed
Fuel bed combustion conditions were maintained by regulating the
combustion air at the blower until optimum burning was confirmed
by visual observation. Manipulation was performed by representa-
tives of Engineered Products Co., and Solid Fuel Systems, Inc.
Conditions were therefore optimum for coal combustion. It was
beyond the scope of this program to determine typical combustion
conditions of units installed in residential structures.
Ambient Conditions
Testing was performed during June 1977, at which time tempera-
tures induced less stack draft during OFF segments of the heating
cycle than would occur during colder months. This should not
have affected combustion conditions because of the barometric
damper; however, it probably reduced the volume of dilution air
in the stack. Wind gusts during several days of testing also
caused fluctuations in dilution air volume and, in turn, pollu-
tant concentration, Variations in the amount of dilution air had
the least influence on Method 5 (particulates) sampling, which is
flow proportiona and the greatest influence on Method 7 (NOX)
grab sampling. T>~ improve the representativeness of grab
samples, three we * > taken during each test (near the beginning,
middle and end of each run). Sampling was performed inside a
warehouse to minimize the effects of changing atmospheric
conditions on the sampling equipment itself.
TEST SITE AND FACILITIES
Investigations were conducted ori coal-fired residential combus-
tion equipment located at Solid Fuel Systems, Inc., Englewood,
Colorado, a sales outlet for coal- and wood-fired heating equip-
ment in the residential size range. Combustion equipment con-
sisted of a 200,000-Btu/hr Long Arrow Model 101 hot-water boiler
(Longero, Inc., Denver, Colorado) combined with a Model S30
Will-Burt stoker (Orville, Ohio), and a 200,000-Btu/hr Hardin
warm-air furnace/stoker (S. S. Manufacturing Company, Lyons,
Colorado). Both units were set up adjacent to eacft- other to
facilitate emission sampling from one scaffolding unit. Exhaust
gases were discharged to the atmosphere through 8-in. galvanized
ducts at an elevation of 23.3 ft above the base of the combustion
equipment. The equipment is described in detail in the following
sections.
12
-------
Stokers
A stoker is a mechanical device that feeds solid fuel to a com-
bustion operation. Stokers employed for residential combustion
are of the underfeed type and in addition to feeding the coal,
also provide supporting mechanisms for combustion such as the
retort, windbox, and combustion air supply. Figure 1 (5) illus-
trates a typical stoker assembly.
TOP VIEW
SIDE VIEW
-t^-
1. HOPPER
2. ELECTRIC MOTOR
3. TRANSMISSION
4. COAL FEED TUBE
5. FEED WORM
6. RETORT
7. RETORT AIR CHAMBER
8. COMBUSTION CHAMBER
9. WIND BOX AND TUYERES
Figure 1. Will-Burt stoker assembly (5).
(5) Domestic Stokers, Hopper and Bin Feed by Will-Burt (manu-
facturer's brochure). Form W346-75-2M, The Will-Burt Co.,
Orrville, Ohio. 4 pp.
13
-------
During stoker operation a worm-feed mechanism conveys coal from
a hopper to the fuel bed inside the furnace or boiler. A multi-
ple grooved pulley on the motor controls the coal feed rate.
Underfeed stokers deliver fresh coal to the fuel bed by feeding
it underneath the hot coals. Below the fuel bed coal is devola-
tized and ignited in a cast iron chamber, or retort. The retort
is surrounded by a wind box that delivers combustion air to the
fuel bed through slotted holes called tuyeres. The tuyeres are
part of a cast iron ring which in the Hardin unit is rotated by
the feed screw to break up clinkers and push ash to the outside
of the fuel bed. Air for combustion is supplied by a blower
located under the coal hopper and driven by the coal feed motor.
Air flow is controlled by a cover on the fan housing and, on the
Will-Burt stoker, by a counterweight damper located in the air
delivery tube. The motor driving the feed screw and the blower.
is controlled by a room thermostat, a limit switch, and a hold-
fire timing relay.
In order to obtain consistent, reproducible stack gas samples it
was necessary to maintain a predictable combustion cycle. A
timer was installed to override the above control devices by
activating the stoker according to a predetermined 20-min ON/40-
min OFF cycle.
Boiler
The Long Arrow Moael 101 boiler was operated in conjunction with
a 303 (30 Ib/hr ma ) Will-Burt stoker. This arrangement was
such that, as f'esn fuel was fed to the combustion chamber, com-
bustion residue, or ash, accumulated at the outer edge of the
fuel bed. About once a day ash was removed from the combustion
chamber with a shovel, but never during sampling.
Heat generated by the burning coal was used to raise the tempera-
ture of water circulating through the boiler. Hot water was
pumped through two unit heaters, which dissipated heat to the
atmosphere and then returned water to the boiler. Figure 2 shows
the overall boiler system layout used for this test program and
includes the location of stack sampling ports.
The Will-Burt stoker design is such that, in conjunction with the
boiler, it is capable of burning most coals except anthracite.
Under this arrangement the fuel bed is deeper than that of the
furnace and, therefore, hotter.
Furnace
The Hardin furnace was equipped with a stoker that has a rotating
ring above the retort to break up clinkers as they form and push
ash out of the combustion area into a removable ash pan. This
ring was activated by the worm-feed mechanism and slowly rotated
in increments. The Hardin furnace was operated to exchange heat
14
-------
(Jt
STACK CAP
ELEV. 23'-4"
AIR VENT n
ELEV. 17'-7"(S02)
ELEV. 16'-2" (CO,CO?,OJ
ELEV. 14'-7" (NO )
A
<; ELEV. 12'-10" (PARTICULATES, ELEMENTS,
ORGAN 1C MATER I AD
SUPPLY
THERMOMETER
BAROMETRIC DAMPER
ELEV. 5'-8"
ELEV. 5'-2" (THERMOMETER)
RELIEF VALVE
ELEV. 4'-11-1/2"
-------
with room air. Figure 3 illustrates the overall furnace system
layout and location of exhaust stack sampling ports.
The Hardin unit, as originally designed, is capable of burning
only lignite and subbituminous coals, possibly because of its
shallow fuel bed. To make this unit more marketable in the
Denver area, the ring supplied with the stoker, having conven-
tional tuyeres (air slots), was replaced with one fabricated by
Solid Fuel Systems, Inc. that differed only in the geometry of
the air slots. This ring appeared to provide higher velocity
combustion air directed more toward the center of the fuel bed.
A version of this ring is now being employed by the manufacturer.
This modification enabled the Hardin unit to burn some bituminous
coals. Test Coal B in this program, a high volatile C bituminous
coal, burned reasonably well in this unit, but test Coal A, a
high volatile B bituminous coal, could not sustain adequate
combustion.
Barometric Damper
Most solid fuel combustion systems rely on a barometric damper
to control draft through the fuel bed. Barometric dampers allow
room air to enter the exhaust stack when the stack is placed
under negative pressure due to atmospheric conditions. This is
beneficial during the portion of a heating cycle when the stoker
and fan are off ? j. the fuel bed remains hot. An externally
induced draft wouj'"' cause more rapid combustion of residual coal
during the OFF se i tent and could result in the fire burning out.
During testing the barometric damper was allowed to operate
normally to maintain steady-state combustion conditions. As a
consequence, stack gas volumetric flow rate and emission concen-
trations varied with time during the OFF period. Figure 4 is a
photograph of the barometric damper.
TEST COAL
Combustion units in this study were fired with western coals
obtained from Swisher Coal Co., Price, Utah; Rosebud Coal Co.,
Hanna, Wyoming; and Imperial Coal Co., Erie, Colorado. These
coals are referred to as test coals A, B, and C, respectively.
Commercial Testing and Engineering Company (CT&E), Denver,
Colorado, was contracted to procure, prepare, analyze and deliver
two of the test coals (A and B). Coal C was obtained late in the
program by Solid Fuel Systems to evaluate a cleaner burning coal.
Coals A and B were purchased in two fractions, high ash (unwashed)
and low ash (washed). Equal portions of each fraction were com-
bined to generate a medium ash coal thereby giving three ash
levels of each coal. Concentrations of other materials in the
coal, such as sulfur content, also changed as seen later in this
section. Test coals were delivered to the test site in sealed
16
-------
CONTROL PANEL
ELEV. 23'-4" STACK CAP
ELEV. 17'-8" (S02)
ELEV. 16'-9" < PARTI CULATES, ELEMENTS, ORGANIC MATERIAL)
ELEV. 16'-5-1/2" (CO,C02,02)
ELEV. 12'-7" (NO)
A , .
ELEV. 11'- 7-1/2" (PARTICULATES, ELEMENTS, ORGAN 1C MATERIAL,
SASS TRAIN)
ELEV. 5'-8" BAROMETRIC DAMPER
ELEV. 5'-2" (C02,02»
NOTE: FLOOR ELEVATION 0'- 0"
Figure 3. Hardin furnace/stoker forced-air heating system and sampling port
locations. (Drawing by D. Young of Engineered Products Co.,
Denver, Colorado.)
-------
Figure 4. Barometric damper employed on coal-fired
combustion equipment to equalize exhaust
stack pressure drops.
polyethylene-lined metal drums. Proximate and ultimate analyses
for Coals A and B were obtained by CT&E, who also analyzed them
for free swelling index and ash fusion temperature. MRC analyzed
Coals A and B for elemental composition and obtained a proximate
analysis for Coal C. Results of the test coal analyses are
presented in Tables 4 and 5 along with analyses of 86 Rocky
Mountain Province coals obtained from the literature for compar-
ison (6). Test coals compare reasonably well with most coals in
(6) Swanson, V. E., J. H. Medlin, J. R. Hatch, S. L. Coleman,
G. H. Wood, S. D. Woodruff, and R. T. Hildebrand. Collection,
Chemical Analysis, and Evaluation of Coal Samples in 1975.
Open-File Report 76-468, U.S. Department of the Interior,
Denver, Colorado, 1976. 503 pp.
18
-------
TABLE 4. PROXIMATE AND ULTIMATE ANALYSES, FREE SWELLING INDEX, AND ASH FUSION
TEMPERATURES FOR TEST COALS COMPARED WITH 86 ROCKY MOUNTAIN PROVINCE
COAL SAMPLES (PROXIMATE AND ULTIMATE ONLY) (6)
Test coals (as
Moisture, %
Ash, %
Sulfur , %
Carbon, %
Hydrogen , %
Nitrogen, %
Oxygen , %
Chlorine , %
Volatile matter, %
Fixed carbon, %
Heat value, Btu/lb
Free swelling index
Ash fusion temperatures, °F
Oxidizing atmosphere:
Initial deformation
First softening
Second softening
Fluid
Reducing atmosphere:
Initial deformation
First softening
Second softening
Fluid
Low
ash
8.77
4.26
0.42
69.23
5.22
1.46
10.64
0.00
42.27
44.70
12,368
1
2,285
2,305
2,325
2,345
2,160
2,180
2,200
2,220
A
Low/high
ash blend
8.36
7.50
0.38
66.55
4.98
1.36
10.87
0.00
40.52
43.62
11,894
1
2,295
2,315
2,330
2,360
2,230
2,250
2,270
2,290
High
ash
7.82
10.92
0.41
64.26
4.84
1.48
10.27
0.00
39.08
42.12
11,510
1
2,310
2,330
2,350
2,370
2,255
2,275
2,295
2,315
Low
ash
10.00
5.04
0.58
65.80
4.72
1.55
12.31
0.00
38.68
46.28
11,593
1/2
2,180
2,190
2,200
2,220
2,100
2,110
2,120
2,140
received)
B
Low/high
ash blend
11.18
7.09
0.97
62.56
4.75
1.31
12.14
0.00
38.79
42.94
11,079
1/2
2,220
2,240
2,250
2,270
2,170
2,190
2,215
2,240
Rocky Mountain Province coal
High
ash
12.35
9.06
1.45
59.79
4.44
1.36
11.52
0.00
37.51
41.05
10,592
0
2,230
2,260
2,280
2,310
2,175
2,210
2,240
2,245
C
as mined
21.15
3.26
0.47
_a
a
ct
34.72
40.87
9,638
0
a""
Average
of 86
samples
12.9
9.1
0.6
59.7
5.6
1.2
23.8
_D
36.0
42.0
10,480
_b
h
U
~h
L?
~h
M
~h
V
k
M
-.
~h
v
Range
Minimum Maximum
1.6
2.1
0.2
27.1
4.4
0.5
8.2
U
22.7
17.1
4,660
_b
h
V
~h
U
-
k
U
~k
U
~K
\J
35.0
32.2
5.1
75.2
6.7
1.6
47 9
Q
46.7
52.5
13,390
_b
K
U
~h
""K
u
h
U
~*h
D
aAnalysis not performed.
bNo data reported in reference.
-------
TABLE 5. ELEMENTAL ANALYSIS OF TWO TEST COALS COMPARED WITH
124 ROCKY MOUNTAIN PROVINCE COAL SAMPLES (6)
(Ib/ton of coal)
Test Coal B
(as received)
Element
Aluminum
Antimony
Arsenic
Barium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Tin
Titanium
Vanadium
Zinc
Low ash
content
9
0
0
0
5
<0
8
0
0
0
6
<0
3
0
<0
<0
0
0
<0
0
0
4
0
<0
0
0
<0
.8
.028
.006
.26
.0 _
.014a
.4
.086
.004
.066
.0 a
.14a
.0
.042
.00002
. °c 5a
. ->4
. ' r
.J72
.28
.054
.0
.13
. 11
.60
-11 a
.0008d
Rocky Mountain
High ash Average of
content 124 samples
12
0.
0.
0.
1.
13
0.
0.
0.
12
3.
0.
<0.
-------
Comparison of test coal properties to the ASTM classification of
coal by rank (D 388-66) gives the following classification (7):
Coal A: High volatile B bituminous coal
Coal B: High volatile C bituminous coal
Coal C: Subbituminous B coal
In addition, Coals A and B received at the test site contained an
occasional rock, ranging from about 0.75 in. to 1.25 in. in dia-
meter, and excessive fines. Rocks could not be eliminted since
they were not always seen when coal was charged to the hopper;
however, because fines were always near the bottom of a drum,
only the top three-fourths of the coal in a drum was used for
testing. Poor coal quality is common for coal used in the
residential sector (8).
SAMPLING METHODS AND EQUIPMENT
Flue gas constituents were collected for analysis using a variety
of sampling systems including standard EPA procedures (9) and the
Source Assessment Sampling System (SASS).
Particulate and Condensable Organics
/
The procedure and equipment used for quantitative particulate
collection met specifications outlined in Method 5 of the Federal
Register (9). A preliminary velocity traverse indicated a flat
velocity profile across the stack. Because of the small stack
diameter (8 in.), single-point sampling was used to determine
the mass emission rates with the probe tip placed in the center
of the exhaust stack.
Mass emissions from the boiler were collected over the entire
heating cycle by a single Method 5 train. The sampling rate was
adjusted to be isokinetic to the flue gas flow rate as the
boiler cycled. Minimum volume collected was 59 dry standard
cubic feed (dscf), and minimum sampling time was 180 min. Twelve
sampling runs were made on the boiler ranging from 94% to 102%
isokinetic.
(7) 1975 Annual Book of ASTM Standards, Part 26: Gaseous Fuels;
Coal and Coke: Atomspheric Analysis. American Society for
Testing and Materials, Philadelphia, Pennsylvania, 1975.
792 pp.
(8) Personal communication with Stratton C. Schaeffer, Consult-
ing Engineer, Camp Hill, Pennsylvania, 10 November 1977.
(9) Standards of Performance for New Stationary Sources.
Federal Register, 36 (247):24876-24895, 1971.
21
-------
Mass emissions from the furnace were determined by using two
Method 5 trains, one during the ON segment and the other during
the OFF segment of the heating cycle. A light signaled the
transition from one segment to the other to facilitate train
shutdown and startup because each run covered several cycles.
Samples collected during the ON segment ranged from 22 dscf to
70 dscf in volume; and were collected over 60 min to 160 min at
98% to 105% isokinetic except for one run which was 61% isokin-
etic . Samples collected during the OFF segment had volumes of
22 dscf to 117 dscf and were collected over 100 min to 240 min at
97% to 103% isokinetic except for one run which was 148%
isokinetic. Time constraints limited the volume of samples
obtained.
Determination of condensable organic material was made from the
back half portion of the sampling train (Figure 5). The back
half of the train consists of water-filled impingers that collect
most materials passing through the front half filter. This
material is usually considered a part of the total sample (9).
Examination of the back half catch during these tests showed that
it was condensable organic material, and it is reported separately
in this way in the results section.
THERMOMETER
=R IMPINGER TRAIN OR BACKHALF THERMOMETER
TEMPERATURE
SENSOR
PROBE
REVERSE-TYPE
PITOTTUBE
CHECK VALVE
VACUUM LINE
/ AIR-TIGHT \
\ PUMP I
^C
Figure 5. Schematic of EPA Method 5 sampling train for particu-
late and condensable organic material collection (9).
22
-------
Particulate and condensable organic samples obtained at the sam-
pling site were sent to the Colorado State Health Department for
analytical workup. Method 5 equipment was also used to measure
stack gas temperature and velocity for later use in converting
analytical results into stack emission rates.
Sulfur Oxides
Sulfur oxides were collected using equipment and procedures out-
lined in Method 6 of the Federal Register (9). Sulfur oxides
were collected from the boiler during the ON segment only and
from the furnace during each segment separately. Samples
collected during the ON segment were typically 2 to 4 actual
cubic feet (acf) in volume and were collected at a rate of about
3 actual cubic feet per hour (acfh). During the OFF segment
sample volumes were typically 7 acf to 9 acf collected at about
3 acfh.
Nitrogen Oxides
EPA Method 7 was employed, as specified in the Federal Register
(9), for the determination of the oxides of nitrogen. Because
this method is a grab sample method, three samples were taken
over the period of each particulate determination run: one each
near the beginning, middle, and end of a run. Each sample con-
sisted of approximately 2 £ of flue gas collected over 30 s.
Samples were obtained from the boiler during the ON segment only
and from the furnace during the ON and OFF segments separately.
Carbon Monoxide
Concentrations of carbon monoxide (CO) in the flue gas were
determined at the test site by Drager tube analysis of a gas
sample collected in a Tedlar bag (10). Orsat analysis was not
employed because anticipated CO concentrations were below the
detection limit of the equipment. Carbon monoxide analyzers
could not be justified because they would have required an
additional man in the field to collect a relatively small amount
of data.
Drager tubes are normally used in workplace environments where it
is necessary to determine very small concentrations of CO in
ambient air with maximum reliability in a short time. Tubes are
packed with a reagent that reacts when contacted with CO to
produce a color change. The volume of reagent changing color
indicates the volume of CO present in a fixed volume of sample.
(10) Detector Tube Handbook, Air Investigations and Technical
Gas Analysis with Drager Tubes, 2nd Edition, compiled by
Kurt Leichnitz. Dragerwerk Ag., Lubeck, Federal Republic
of Germany, October 1973. 164 pp.
23
-------
In this case the reaction is:
SCO
H2S2Oy
I2 + 5C02
(1)
iodine
(color: brownish
green)
Figure 6 shows an expanded view of one type of Drager CO detector
tube. The detector tube is used in conjunction with a hand
operated pump. The gas detector pump exhausts 100 cm3 with each
stroke, simultaneously drawing in a measured volume of gas sample
through the detector tube. Flue gas samples were withdrawn from
the Tedlar bag which was connected to the Drager tube by rubber
tubing.
n = 5
VOL. - % CO,
4 6
DRAGER TUBE CARBON DIOXIDE 0.1 %
1 AND 2 FUSED TIPS
3 RECORD ING SURFACE
4 INDICATING LAYER (WHITE)
WITH MEASURING SCALE
(NUMERICAL VALUES IN VOL.
5 ARROW (SHOULD POINT
TOWARDS PUMP DURING
TESTING)
6 COVER FILM
%C00
Figure 6.
n = NUMBER OF PUMP STROKES
Expanded view of a typical Drager
gas detector tube (10).
The Drager tube employed (Carbon Monoxide 5/c) contains a pre-
cleanse layer that retains interfering gases (e.g., petroleum
distillates, benzene, hydrogen sulfide). Acetylene and hydrogen
in concentrations greater than 50% are indicated as CO but were
not a problem in this program. The detection range of the tube
was from 5 parts per million (ppm) to 150 ppm with a relative
standard deviation of 10% to 15% (10).
Carbon Dioxide and Oxygen
Field analysis of carbon dioxide (CO2) and oxygen (O2) was per-
formed to determine stack gas molecular weight and combustion
24
-------
excess air. Combustion excess air was determined employing a
Fyrith test kit to measure 02 and CO2 below the barometric damper
of each unit (11). Three measurements were made and averaged over
the ON segment of the heating cycles during Method 5 sampling.
Flue gas composition above the barometric damper was determined
by collecting a composite flue gas sample in a Tedlar bag over
the duration of Method 5 sampling. The sample was analyzed in
the field using the Fyrite kit for O2 and C02 and Drager tubes
for CO concentration.
Fyrite test kits employ the Orsat method of volumetric measure-
ment in which the gas sample is chemically absorbed by a liquid
chemical absorbent which is also used as the indicating fluid.
Fresh fluid was charged to the fluid reservoir at the start of
the test program and observed for color changes as a sign of
exhaustion throughout the program. Fyrite is an acceptable
procedure for molecular weight determination although it is
normally not acceptable for excess air measurements. However,
MRC has conducted tests comparing Fyrite with Orsat and the
results show that at high excess air levels (greater than 100%),
such as those encountered in this program, both methods give
comparable results.
SASS Train Samples
The SASS train employed in the environmental assessment portion
of this study was used to sample the Hardin stoker furnace unit
for particulate loading, particle size data, organic components
including POM1s and PCB's, and individual elements. This system,
depicted in Figure 7 (12), employs a set of three cyclones and a
filter for particle size fractionation, a solid sorbent (XAD-2)
trap, a trace inorganic impinger collection trap, and a system
for flow measurement and gas pumping.
A major feature of the SASS train is that it samples inorganic
and organic emissions simultaneously. Inorganic species are
primarily collected by the cyclones and filters. Volatile
inorganics are also collected in the solid sorbent trap and the
impinger solutions. Organic species are primarily collected in
the solid sorbent trap, although other portions of the train
(particulate samples, impinger solutions, rinse solutions) are
solvent extracted to recover any other organic material.
(11) Bacharach Fyrite Gas Analyzers for Measuring Carbon Dioxide
(C02) or Oxygen (02). Bulletin 4042/2-75, Bacharach
Instrument Company, Pittsburgh, Pennsylvania, 1975. 2 pp.
(12) Hamersma, J. W., S. L. Reynolds, and R. F. Maddelone.
IERL-RTP Procedure Manual: Level I Environmental Assessment,
EPA-600/2-76-160a, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, June 1976. 147 pp.
25
-------
KK
COUPLE
C
HEATER
CON-
TROLLER
1
lOum
CONVECTION
OVEN
FILTER
GAS COOLER
STAINLESS STEEL PROBE
to
3 urn
\ /
L __
imn
\~ ' 7
r
%
£1
.J
GAS
TEAAPERATURE
THERAAOCOUPLE
I Oyo
I -cEri
OVEN
THERMOCOUPLE
XAD-2
CARTRIDGE
CONDENSATE
COLLECTOR
^PINGER/COOLER
TRACE ELEMENT
COLLECTOR
DRY GAS METER/OR I Fl CE METER
CENTRAL! ZED TEMPERATURE
AND PRESSURE READOUT
CONTROL MODULE
IMPINGER
THERMOCOUPLE
10-CFM VACUUM PUMP (2)
Figure 7. Schematic of Source Assessment Sampling System (12).
-------
The impinger portion of the train consists of four impingers.
The impinger order, impinger contents, and purpose of each
impinger are shown in Table 6.
TABLE 6. SASS TRAIN IMPINGER SYSTEM (12)
Impinger
Reagent
Quantity
Purpose
6 M H2O2
0.2 M (NHif)2S208
+ 0.02 M AgN03
0.2 M (NHit)2S2O8
+ 0.02 M AgNO3
3-8 Mesh silica gel
(color indicating)
750 ml Trap reducing gases such as
SO2 to prevent depletion of
oxidative capability of trace
element collecting impingers
2 and 3.
750 ml Collection of volatile trace
elements by oxidative
dissolution.
750 ml Collection of volatile trace
elements by oxidative
dissolution.
750 g Prevent moisture from
reaching pumps.
M represents molar.
Before sampling, the SASS train components were passivated with
1:1 (on a volume basis) aqueous nitric acid. All surfaces
associated with organic collection were cleaned with distilled
water, isopropyl alcohol, and methylene chloride in succession.
These components were then dried with a stream of clean air or
nitrogen. Impingers were cleaned with distilled water and
subsequently with isopropyl alcohol.
Once on site, the train was assembled and the oven heated to
400°F before each run. The resin trap was maintained at 68°F
during a run. A leak check was made before the start of a run,
with a leak rate of less than 0.05 cubic feet per minute (cfm)
at 20 in. Hg being acceptable.
SASS train sampling was performed during the ON segment of the
furnace heating cycle. To obtain the required sample volume of
30 m3, it was necessary to sample for 8 hr. Because sampling was
performed only during the ON segment of a 20-min ON/40-min OFF
cycle, the total elapsed time for a run was about 24 hr. High
particulate loadings were encountered and necessitated filter
changes after every 40 min or two heating cycles, of sampling.
Cleanup procedures used after completion of a sampling run are
specified in the Level I Procedure Manual and are shown in
Figures 8 to 10 (12).
27
-------
PROBE AND
NOZZLE
METHYLENE CHLORIDE:METHANOL
RINSE INTO AMBER GLASS CONTAINER
ADD TO 10 ^m
CYCLONE RINSE
\
to
00
STEP 1: TAP AND BRUSH
CONTENTS FROM WALLS
AND VANE INTO LOWER
CUP RECEPTACLE
STEP 2: RECONNECT LOWER CUP
RECEPTACLE AND RINSE ADHERED
MATERIAL ON WALLS AND VANE
INTO CUP (METHYLENE CHLORIDE:METHANOL)
STEP 1: TAP AND BRUSH CONTENTS
FROM WALLS INTO LOWER CUP RECEPTACLE
STEP 2: RECONNECT LOWER CUP
RECEPTACLE AND RINSE ADHERED
MATERIAL WITH METHYLENE CHLORIDE:
METHANOL INTO CUP
STEP 3: RINSE WITH METHYLENE CHLORIDE:
METHANOL INTERCONNECT TUB ING JOINING
10/urn TO 3ftm CYCLONES INTO AMBER
GLASS CONTAINER
REMOVE LOWER CUP RECEPTACLE
AND TRANSFER CONTENTS INTO
A TARED NALGENE CONTAINER
REMOVE LOWER CUP RECEPTACLE
AND TRANSFER (METHYLENE CHLORIDE:
METHANOL) INTO PROBE RINSE CONTAINER
REMOVE LOWER CUP RECEPTACLE
AND TRANSFER CONTENTS INTO
A TARED NALGENE CONTAINER
REMOVE LOWER CUP RECEPTACLE
AND TRANSFER CONTENTS INTO
AN AMBER GLASS CONTAINER
COMBINE
ALL RINSES
FOR SHIPPING
AND ANALYSIS
Figure 8. Sample handling and transfer-nozzle, probe, cyclones, and filter (12).
(continued)
-------
1 Mm CYCLONE
to
VQ
FILTER HOUSING
STEP 1: TAP AND BRUSH
CONTENTS FROM WALLS
INTO LOWER CUP RECEPTACLE
STEP 2: RECONNECT LOWER CUP
RECEPTACLE AND RINSE ADHERED
MATERIAL WITH METHYLENE CHLORIDE:
METHANOL INTO CUP
STEP 3: RINSE WITH METHYLENE CHLORIDE:
METHANOL INTERCONNECT TUBING JOINING
3 Mm TO 1 Mm CYCLONES INTO AMBER GLASS
CONTAINER
STEP 1: REMOVE FILTER AND
SEAL IN TARED PETRI DISH
STEP 2: BRUSH PARTICULATE FROM
BOTH HOUSING HALVES INTO A
TARED NALGENE CONTAINER
STEP 3: WITH METHYLENE CHLORIDE:
METHANOL RINSE ADHERED PARTICULATE
INTO AMBER GLASS CONTAINER
STEP 4: WITH METHYLENE CHLORIDE:
METHANOL RINSE INTERCONNECT TUBE
JOINING 1 Mm CYCLONE TO HOUSING
INTO AMBER GLASS CONTAINER
REMOVE LOWER CUP RECEPTACLE
AND TRANSFER CONTENTS INTO
A TARED NALGENE CONTAINER
REMOVE LOWER CUP RECEPTACLE
AND TRANSFER CONTENTS INTO
AN AMBER GLASS CONTAINER
NOTE - ALL METHYLENE CHLORI DE:METHANOL
MIXTURES ARE 1:1.
ALL BRUSHES MUST HAVE NYLON BRISTLES.
ALL NALGENE CONTAINERS MUST BE HIGH
DENSITY POLYETHYLENE
Figure 8 (continued)
-------
STEP NO.1
COMPLETE XAD-2 MODULE
AFTER SAMPLING RUN
STEP NO. 2
RELEASE CLAMP JOINING XAD-2
CARTRIDGE SECTION TO THE UPPER
GAS CONDITIONING SECTION
REMOVE XAD-2 CARTRIDGE FROM
CARTRIDGE HOLDER. REMOVE FINE
MESH SCREEN FROM TOP OF CART-
RIDGE. EMPTY RESIN INTO WIDE
MOUTH GLASS AMBER JAR.
REPLACE SCREEN ON CARTRIDGE
REINSERT CARTRIDGE INTO MODULE.
JOIN MODULE BACK TOG ETHER.
REPLACE CLAMP.
OPEN CONDENSATE RESERVOIR
VALVE AND DRAIN AQI^OUS
CONDENSATE INTC i>£PARATORY
FUNNEL. EXTRACT tolTH METHYLENE CHLORIDE.
B AS IFY ONE-HALF
w pH12
ACIDIFY ONE-HALF
pH LESS THAN 2
CLOSE CONDENSATE RESERVOIR VALVE
REEASE UPPER CLAMP AND
LI FT OUT INNER WELL
WITH GOTH UNITIZED WASH BOTTLE
(METHYLENE CHLORIDE:METHANOL)
RINSE INNER WELL SURFACE INTO AND
ALONG CONDENSER WALL SO THAT RINSE
RUNS DOWN THROUGH THE MODULE AND
INTO CONDENSATE COLLECTOR
WHEN INNER WELL IS CLEAN,
PLACE TO ONE SIDE
RINSE ENTRANCE TUBE INTO MODULE
INTERIOR. RINSE DOWN THE CONDENSER
WALL AND ALLOW SOLVENT TO
FLOW DOWN THROUGH THE SYSTEM
AND COLLECT IN CONDENSATE CUP
RELEASE CENTRAL CLAMP AND
SEPARATE THE LOWER SECTION
(XAD-2 AND CONDENSATE CUP)
FROM THE UPPER SECTION (CONDENSER)
THE ENTIRE UPPER SECTION IS NOW
CLEAN._
RINSE THE NOW EMPTY XAD-2 SEC-
TION INTO THE CONDENSATE CUP
RELEASE LOWER CLAMP AND
REMOVE CARTRIDGE SECTION
FROM CONDENSATE CUP
THE CONDENSATE RESERVOIR NOW
CONTAINS ALL RINSES FROM THE
ENTIRE SYSTEM.DRAIN INTO AN
AMBER BOTTLE VIA DRAIN VALVE.
Figure 9. Sample handling and transfer - XAD-2 module (12).
30
-------
ADD RINSE FROM
CONNECTING LINE
LEADING FROM
XAD-2 MODULE
TO FIRST IMPINGER
IMPINGER 1
t
TRANSFER TO
NALGENE CONTAINER
RINSE WITH 1:1
ISOPROPYL ALCOHOL (IPA)/-
DI STILLED WATER AND ADD
IMPINGER 2
TRANSFER TO
NALGENE CONTAINER
1/1
CXL
RINSE WITH 1:1
1 ISOPROPYL ALCOHOL (IPA)/
DISTILLED WATER AND ADD
J
IMPINGER 3
r
TRANSFER TO
NALGENE CONTAINER
COMBINE AND
MEASURE TOTAL
VOLUME FOR
SINGLE ANALYSIS
RINSE WITH 1:1
ISOPROPYL ALCOHOL (IPA)/-
DI STILLED WATER AND ADD
IMPINGER 4
SILICAGEL
DISCARD
Figure 10. Sample handling and transfer - impingers (12)
31
-------
Modified Method 5 Train Samples
A modified Method 5 procedure was used to obtain samples of
particulate and POM emissions. Because of problems in SASS train
operation the modified Method 5 samples were also used for
elemental analysis and organic characterization. The modifica-
tion added an XAD-2 resin trap between the filter and impinger
system of the standard Method 5 train to collect organic species.
The train oven was operated above stack temperature to allow
organic vapors to pass through the filter. A cooler trap was
inserted between the filter and the resin trap to lower the tem-
perature of the gas to 70°F before its entry into the resin trap.
The impinger portion of the train used for POM collection was
modified so that the first impinger contained 100 mJl of 10%
potassium hydroxide (KOH) to remove sulfur oxides and prevent an
increase in acidity in the subsequent impingers. The risk of oxi-
dizing what little organic may pass through the XAD resin during
the short contact time with KOH is less than the risk of reacting
organics in acidic impinger liquids. The second and third
impingers contained 200 m£ each of toluene, and the fourth con-
tained 200 g of 6-mesh to 16-mesh silica gel. The purpose of this
impinger system was to provide additional assurance that all POM-
type materials were collected. Silica gel was added to the last
impinger to remove moisture and protect the pump and dry gas meter.
At the completion of a sampling run, the resin trap was removed
and capped. :he entire train from probe tip to filter holder was
cleaned with 1:1 CH2C12-CH3OH and the sample bottled. KOH
impinger and cooler trap contents were measured and the liquids
poured into amber bottles. Contents of the toluene impingers
were measured in order to determine the quantity of condensed
water, and the contents were poured into a separate amber bottle.
The silica gel impinger contents were weighed and the material
discarded. All impingers and connecting glassware were rinsed
with 1:1 CH2C12-CH3OH, and the washings were poured into amber
bottles. This system was also used to determine particulate
emissions in accordance with the standard Method 5 procedure.
Particulates were then analyzed to determine their elemental and
organic composition.
The basic Method 5 schematic is shown earlier in Figure 5;
Figure 11 shows the components and sample recovery procedure for
the modification.
Coal
A 2-lb coal sampxe was obtained from the coal feed of each sam-
pling run, and selected samples were later sent for elemental
analysis. As the coal was shoveled into the stoker hopper, some
of it was added to a polyethylene jar until about a 2-lb
composite was collected.
32
-------
,GLASS PETRI DISH
PROBE
fl
RINSE
(POTASSIUM
HYDROXIDE)
RINSE.
METHYLENE
CHLORIDE:
METHANOL
FI
CONTENTS
(MEASURE
VOLUME)
RINSE,
METHYLENE
CHLORIDE:
METHANOL
n
K
WEIGH,
DISCARD
PLACE IN THE COOLER
SO AS NOT TO BECOME
IMMERSED IN WATER
RINSE
TOLUENE
1. XAD TRAPS
a, CLOSE WITH CLASS CAPS AND PINCH CLAMPS.
b. PUDGE WITHIN POLYETHYLENE BAG USING NITROGEN.
c. SEAL SO AS 10 EXCLUDE WATER.
a. PLACE ON ICE. EXCLUDING LIGHT.
2. FILTERS
a. PLACE IN PETRI DISH, TOP SIDE UP.
b. TAPE DISH CLOSED.
t PURGE VIITH NITROGEN WITHIN POLYETHYLENE BAG.
fl. SEAL TO EXCLUOE WATER. CHILL. EXCLUDE LIGHT.
!. LIQUIDS
a. PURGE WITH NITROGEN: TAPE CAPS CLOSED.
b. CHILL ON ICE: EXCLUDE LIGHT.
«. USE SPECTRO-VAt-2 GREASE THROUGHOUT.
5. IF IMPINGER CONTFtlTS ARE MULTIPHASE, MEASURE
SEPARATELY AND IDENTIFY AS TO ORDER Of LAYERS:
R1COMBINE IN RESPECTIVE SAMPLL 60TTLE.
6. RETAIN BLANKS Of MITFR.XAD, AND All REAGENTS.
Figure 11.
POM train components and sample recovery procedure employing modified
Method 5 equipment.
-------
Ash
Ash samples were collected from the coal-fired furnace only
because there was no provision for separating ash from the boiler
fuel bed. Each sampling run was begun with the slide-out ash pan
empty. At the end of the run the ash pan contents were allowed
to cool, and then they were weighed and bagged in zip-lock bags
for transport to the analytical lab.
ANALYTICAL PROCEDURES
The Dayton Laboratory of MRC and the Denver Laboratory of the
Colorado Department of Health (CDH) shared analytical responsib-
ilities in this program. Samples collected by Methods 5, 6, and
7 were analyzed by CDH using analytical methods as specified in
the Federal Register (9). Results were used by MRC to calculate
emission rates and emission factors. Samples collected with the
SASS train and modified Method 5 train were analyzed by MRC.
Particulates and Condensable Organics
Samples for particulate analysis from the Method 5 train (front
half washings and filter) were sealed in glass and turned over
to CDH for analysis by standard Method 5 procedures. The back
half portion cf the train (impinger solutions) was checked in
the field fc jioisture gain, sealed in glass, and turned over to
CDH for gravimetric determination of condensable organic
material. T" f nonstandard Method 5 train (modified for POM col-
lection) was returned to MRC. Mass emission rates were deter-
mined as specified in the Federal Register (9), and the samples
were saved for extraction of organic material.
Sulfur Oxides and Nitrogen Oxides
Method 6 samples for SOX analysis were sealed in polyethylene
containers and refrigerated until picked up by CDH. They were
analyzed using the Federal Register procedures (9).
Samples collected by Method 7 for NOX analysis were partially
work up at the test site by recording flask temperature, baro-
metric pressure, and internal flask pressure (Federal Register
section 4.2.1). The flasks and their contents were then turned
over to CDH for the remainder of the analysis as specified in
the Federal Register (9).
SASS Train Workup and Analysis
The Source Assessment Sampling System employed in this program
allowed the collection of many components of an emission source
in a single test. However, the separation and analysis schemes
were complex because of the many components. Once separation
34
-------
was accomplished, the analytic method for any particular compon-
ent class of compound (e.g., trace metals or POM) was identical
regardless of the source of the sample. As a result, a major
portion of the analytical effort concerned the separation of
components prior to analysis. Figure 12 illustrates the complete
separation and analysis protocols. Specified Level I procedures
were modified to be consistent with the objectives of this
program.
At the completion of a sample run, the samples collected by this
system included the contents of the three cyclones, the filter,
the combined probe and cyclone washes, the XAD-2 resin trap, the
XAD-2 trap aqueous layer, the XAD-2 organic layer and washes, and
the combined impinger collection and washes.
In order to determine a mass loading, the materials, collected
from the cyclones and filter were individually weighed. The
probe and cyclone wash was evaporated to dryness and weighed,
and this plus the cyclone collected material and the filter
catch provided a "front half" mass for the calculation of the
particulate emission rate. All of the solid materials were then
combined and extracted with methylene chloride for 24 hr in a
Soxhlet extractor in order to extract organic materials for later
organic analysis. The solid residue was then retained for
elemental analysis.
The XAD-2 resin containing trapped organics was homogenized to
assure an even distribution of organic and inorganic components,
and a 2-g portion was removed for trace element analysis. The
remaining sample was extracted with pentane for 24 hr in a Soxh-
let extractor to extract organic species. The aqueous condensate
of the XAD-2 module was extracted with methylene chloride, and
the organic phase was combined with the XAD-2 organic layer and
washes. The aqueous phase of the condensate was used for
elemental analysis. Impinger contents were diluted to a known
volume, and an aliquot was withdrawn and analyzed for its
elemental content.
At this point in the separation scheme the organic components
from the SASS collection were contained in 1) the CH2C12 extract
of the cyclones, filter catch, and solid residue from the probe
and cyclone washes; 2) the pentane extract of the XAD-2 resin
trap; and 3) the combined XAD-2 organic layer, XAD-2 module
organic wash, and CH2Cl2 extract of the XAD-2 aqueous condensate.
Portions of each extract were analyzed for low molecular weight
(C7 to Ci2) organic compounds with a flame ionization gas chroma-
tograph using a 1.5% OV-101 on Gas Chrom Q 100/120 mesh (1/8 in.
x 6 ft) stainless steel column. Following chromatography, the
liquids were all combined and then subjected to rotary evapora-
tion. Pentane and CH2C12 were replaced with hexane by solvent
exchange to be more compatible with the silica gel column used
for separation.
35
-------
GAS
CHROMATOGRAPHY
ANALYSIS
C7 to C16
HYDROCARBONS
GAS
CHROMATOGRAPHY
ANALYSIS
C7toC16
HYDROCARBONS
Figure 12.
SASS train component separation
and analysis scheme.
(continued)
36
-------
Figure 12 (continued)
(continued)
37
-------
XAD-2
ORGANIC WASH
XAD-2
AQUEOUS
CONDENSATE
FIRST 1MPINGER
SECOND AND THIRD
IMPING ERS
ACIDIFY
NITRIC ACID
AA ANALYSIS
SELENIUM,
ARSENIC,
MERCURY
Figure 12 (continued).
38
-------
After volume reduction, the sample was separated into eight
fractions on a silica gel column using the solvent systems shown
in Figure 13. Each fraction was then reduced in volume with a
Kuderna-Danish evaporator, transferred to a tare-weighted micro-
weighing pan, and remaining solvent evaporated in air. Each
dried fraction was weighed and then redissolved in a minimum
quantity of methylene chloride.
Following separation, the second, third, and fourth fractions
(containing the POM and PCB components) were combined, diluted if
necessary with methylene chloride, and transferred to a Teflon
septum-sealed vial which was covered with aluminum foil and
refrigerated until required for analysis. Prior to analysis, the
sample underwent one more volume reduction via the Kuderna-Danish
method depending on sample matrix.
The method used for POM and PCB analysis employed a peak-area
quantitation technique with computer reconstructed chromatograms
from the (HP 5982-A) GC/MS. All data were collected in the
electron impact (El) mode because of the abundance of available
El-mass spectra.
Gas chromatographic separation was achieved using a 6-ft Dexsil
400 glass column with temperature programming from 160°C for
2 min, rising to 280°C at 8°C/min (4°C/min for PCB's), and
becoming isothermal at 280°C. The carrier gas was helium at a
flow rate of 30 m£/min.
The mass spectrometer, operating in the El mode, was programmed
to scan the 35-350 atomic mass unit (amu) range as the POM and
PCB components eluted from the gas chromatograph. The data
system reconstructed the chromatogram using the total ion mode.
POM's and PCB's were located by their molecular mass ions which
were displayed using the selected ion mode (SIM). Their identity
was confirmed by examination of their mass spectra and retention
times. Samples and standards were run in SIM for quantitation.
Calibration curves were prepared for each POM and PCB of interest
using varying concentrations of standards in methylene chloride,
plotting mass ion peak area vs. concentration, and determining
response factors. Sample peaks were compared with standard
curves that were obtained under the same conditions of attenua-
tion, injection volume (2 y£), and tuning condition.
Organics and POM
A Method 5 particulate sampling train modified with an XAD-2
resin trap was used to collect organic material for POM analysis.
Later in the program it was also decided to characterize and
quantify all the major chromatographable organic species present
in the organic extract. The separation and analysis of the
probe, filter, and resin trap contents followed the procedure
for the SASS train just described.
39
-------
CYCLONE AND
FILTER PARTICU-
LATE SAMPLE
XAD~ 2
PROBE WASHES
Jl
EXTRACTION
LIQUID
CHROMA TOGRAPHIC
SEPARATION
8 FRACTIONS
CONCENTRATED
HYDROCHLORIC ACID,
METHYL ALCOHOL,
METHYLENE CHLORIDE
(5/70/30)
20% METHYLENE
50% METHYLENE
KUDERNA -DANISH
EVAPORATION;
WEIGHT 8 SAMPLES
DISSOLVED IN METHYLENE CHLORIDE
4 5
FRACTIONS FOR ORGANIC ANALYSIS
Figure 13. Organic component analysis flow diagram.
-------
The 10% KOH impinger solution was extracted with methylene
chloride and the volume reduced by rotary evaporation to approx-
imately 5 ma. Contents of the toluene impinger were reduced in
volume by the same technique. These extracts and solutions were
combined with the probe washing and the material evaporated in
air. The residue was dissolved in 25 m£ of pentane and then
separated into fractions using the silica gel column. The second,
third, and fourth fractions collected using the mixed solvent
procedure (Figure 13) were analyzed by GC/MS for POM's.
In addition other major organic species were identified and
quantified for all fractions from three sampling runs. Results
were obtained by using the standard-area-quantitation method for
GC/MS analysis. When standards of a particular compound could
not be obtained, quantitation was achieved using the response
of a closely related compound. Gas chromatographic separation
was achieved on a 6-ft 3% Dexsil 400 on Chromasorb W glass
column.
Elemental Analysis
Analysis of elemental composition was performed on SASS train
components, particulates collected in the modified Method 5
trains, coal, bottom ash, and ash leachate. Atomic absorption
was used to analyze arsenic, mercury and selenium, while the
AtomComp technique employing an inductively coupled argon plasma
(ICAP) excitation source was used to quantify aluminum, antimony,
barium, boron, cadmium, calcium, chromium, cobalt, copper, iron,
lead, magnesium, manganese, molybdenum, nickel, phosphorus, sili-
con, silver, sodium, strontium, tin, titanium, vanadium, and zinc.
Mercury was determined by flameless atomic absorption; arsenic
and selenium by conversion to their hydrides and aspiration into
an argon-hydrogen flame. (A more detailed description of these
methods can be found in Reference 13.)
The Jarrell-Ash AtomComp with ICAP forms an analytical system for
simultaneous multielement determinations of trace metals at the
sub-ppm level in solutions. The basis of the method is atomic
emission promoted by coupling the sample, nebulized to form an
aerosol, with high temperature argon gas produced by passage of
argon through a powerful radio-frequency field (14).
(13) Metals by Atomic Absorption Spectrophotometry. In: Stand-
ard Methods for the Examination of Water and Wastewater,
14th Edition. American Public Health Association, Washing-
ton, D.C., 1976. 1193 pp.
(14) Jarrell-Ash Plasma AtomComp for the Simultaneous Determina-
tion of Trace Metals in Solutions (manufacturer's brochure)
Catalog 90-975, Jarrell-Ash Company, Waltham, Massachusetts
5 pp.
41
-------
All of the solid samples were digested before analysis using the
acid digestion Parr bomb technique originally developed by Bernas
and modified by Hartstein for trace metal analyses of coal dust by
atomic absorption (15, 16). This method employs the Parr 4145
Teflon-lined bomb and involves digestion of powdered samples in
ULTRAR brand (69% to 71%) redistilled nitric acid at 150°C. The
accuracy of this method for coal dust analysis, reported for 10
metals, ranged from 94% for beryllium to 106% for nickel using
10-mg samples. Sample solutions produced by acid digestion were
diluted with distilled water to reduce acid concentrations to
approximately 2% and submitted for analysis.
Aqueous impinger solutions from the SASS train were analyzed for
the volatile elements which could not be 100% collected by the
filter; i.e., mercury, arsenic, selenium and antimony. In addi-
tion to the above analyses, particulate samples were submitted to
the Physical Science Center of Monsanto Company for analysis of
carbon, hydrogen, and nitrogen. A Perkin Elmer Model 240
Elemental Analyzer was used to perform the analysis.
Ash Leaching
Bottom ash collected during sampling runs was mixed with dis-
tilled water at a l-to-10 ash-to-water ratio. The mixture was
shaken for 24 hr, after which it was filtered through a Whatman
#1 filte , The filtrate was then diluted to a known volume and
submitted "ror elemental analysis.
Other Coal Properties
Analysis of the test coal was conducted by Commercial Testing
and Engineering Company, Western Division, located in Denver,
Colorado. Analyses were performed according to standard proced-
ures to measure proximate and ultimate analyses, ash fusion
temperature, and free swelling index.
(15) Bernas, B. A New Method for Decomposition and Comprehensive
Analysis of Silicates by Atomic Absorption Spectrometry.
Analytical Chemistry, 40 (11) -.1682-1686, 1968.
(16) Hartstein, A. M., R. W. Freedman, and D. W. Platter. Novel
Wet-Digestion Procedure for Trace-Metal Analysis of Coal
by Atomic Absorption. Analytical Chemistry, 45 (3) :611-614
1973.
42
-------
SECTION 4
RESULTS
Test results for the sampling and analysis of emissions from coal-
fired residential heating systems are discussed in this section.
Each emission species measured is addressed separately, and emis-
sion factors are correlated where possible to other test param-
eters. Overall conclusions and potential implications of the
test results are then considered. Caution should be exercised in
extrapolating results from this test program to other combustion
equipment and other coal types.
EMISSIONS
Emission factors for the test conditions used in this program are
summarized in Table 7 for criteria pollutants and POM emissions.
Emission factors are, in most cases, for duplicate test runs. As
discussed in Section 3, emissions from the furnace were measured
independently for the ON and OFF heating cycle segments. These
emission rates were combined and are presented as an emission
factor for the total heating cycle (20 min ON/40 min OFF). Data
for the separate segments are presented later in this section.
Other parameters and stack gas conditions for each run are given
in Table 8.
Particulates
Table 7 shows that particulate emissions from the warm-air
furnace are an order of magnitude higher than those from the
boiler. This difference is related to differences in firing geom-
etry of each unit. The effect of combustion equipment on particu-
late emissions is discussed in more detail later in this section.
Particulate emission factors from the boiler compare reasonably
well with recent studies (2, 17) on similar equipment and similar
coals and are in most cases less than that suggested in the EPA
(17) Hangebrauck, R. P., D. J. Von Lehmden, and J. E. Meeker.
Emissions of Polynuclear Hydrocarbons and Other Pollutants
from Heat-Generation and Incineration Processes. Journal of
the Air Pollution Control Association, 14(7):267-278, 1964.
43
-------
.fa-
it*
TABLE 7. EMISSION FACTORS FOR POM AND CRITERIA POLLUTANTS FROM COAL-FIRED
RESIDENTIAL HEATING EQUIPMENT OPERATED ON A 20-MIN "ON"/40-MIN
"OFF" HEATING CYCLE9
Test run
number
1,2
3,4
5,6
7,8
9,10
11,12
13,14,15,16
25,26,27,28
17,18,19,20
21,22,23,24
29
Heating u
equipment
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Furnace
Furnace
Furnace
Furnace
Furnace
Most emission factors
Both units
when stoker
As received
Data for ON
Designation
A
A
A
B
B
B
B
B
B
B
C
represent the
Coal
Ash c
content ,
%
10.9
7.5
4.3
9.1
7.1
5.0
5.0
7.1
9.1
9.1
3.3
average of
Sulfur
content ,
%
0.41
0.38
0.42
1.5
1.0
0.58
0.58
1.0
1.5
1.5
0.47
duplicate
are rated at about 200,000 Btu/hr; boiler fuel
was ON.
. Front half Method 5.
segment of
heating cycle
Stethod 6
. Method
Excess
air.
%
238
123
171
128
151
160
129
k
117
182
207
Emission factor.
.
Particulates
5.0 (0.22)
7.8 (0.33)
6.8 (0.28)
2.8 (0.13)
4.8 (0.22)
5.8 (0.25)
44 (1.9)
20 (0.90)
26 (1.2)
34 (1.6)
4.0 (0.21)
a
SOx
6.0^ (0.26)
11. « (0.46)
9.6J. (0.39)
17. 8J (0.84)
30^ (1.4)
14J (0.60)
12.6 (0.54)
30 (1.4)
28 (1.3)
30 (1.4)
_k
f
NOX
12.8^ (0.56)
7.0^ (0.29)
7.6^ (0.31)
2.8^ (0.13)
4.6J(0.21)
6.4 (0.28)
13.6 (0.59)
13.2 (0.60)
7.8 (0.37)
6.0 (0.28)
k
Ib/ton (lb/106 Btu)
0.26 (0.011)
4.0 (0.17)
0.30 (0.012)
O.OB (0.0038)
0.08 (0.0036)
0.08 (0.0035)
24 (1.0)
8.8 (0.40)
26 (1.2)
26 (1-2)
k
Condensable
organics
2.0 (0.087)
2.4 (0.10)
2.2 (0.089)
2.4 (0.11)
3.6 (0.16)
5.0 (0.22)
k
9.4 (0.42)
5.2 (0.25)
k
3.6 (0.19)
\
POM
0.26 (0.011)
k
0.58 (0.023)
k
"k
"k
0.070 (0.0030)
k
"k
0.036 (0.0017)
_k
sampling runs.
feed rate
g
7 . 'Drager
averaged 19.8
tube. Back
Ib/hr when stoker was ON, furnace fuel feed rate averaged
half Method 5.
Modified Method 5 with XAD-2
only. No data obtained due to program limitations.
^
resin trap.
15.5 Ib/hr
-------
TABLE 8. EXPERIMENTAL DATA FOR THE COAL-FIRED HEATING
EQUIPMENT OPERATED ON A 20-MIN ON/40-MIN OFF CYCLE
Test
run
number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Heating
equipment
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Furnace
Furnace
Furnace
Furnace
Furnace
Furnace
Furnace
Furnace
Furance
Furance
Furnace
Furnace
Furnace
Furnace
Furnace
Furnace
Furnace
Coal
type
A
A
A
A
A
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
C
Stoker
Average stack gas conditions3
feed
rate , Temperature ,
Ib/hr °F
21.3
21.3
22.0
22.0
21.6
21.6
21.1
21.1
17.2
17.2
15.4
15.4
15.6
15.6
b
~b
15.8
15.8
b
b
15.8
15.8
b
b
15.0
15.0
_b
b
22.0
226
188
200
179
240
233
214
204
222
189
239
217
277
252
199
186
256
318
180
162
278
338
202
204
282
260
185
167
187
Flow
Velocity , rate ,
fpm acfm
164
265
159
325
177
288
170
164
192
240
216
204
496
527
318
330
530
589
343
270
498
504
129
186
398
488
192
288
374
57
92
56
114
62
100
59
57
67
84
75
71
173
184
111
115
185
206
120
94
174
176
45
65
139
170
67
100
141
H20,
5.5
3.0
4.2
1.9
4.3
3.1
4.0
4.2
2.7
3.0
3.1
3.2
3.0
2.9
2.0
1.9
5.2
2.5
2.5
0.9
2.7
3.7
1.8
1.4
2.9
4.8
1.6
1.8
2.3
C02,
0.2
0.8
1.2
0.1
0.8
0.6
0
0
2.4
0.5
0.3
0.5
3.2
1.8
0
0.2
2.6
2.6
0.2
0.2
1.8
1.6
0.1
0
0
0
0.8
0.3
1.0
O2,
21.0
20.8
18.8
18.8
20.2
20.3
20.7
20.5
17.7
20.5
21.0
20.2
17.3
18.0
21.0
21.1
17.3
17.0
20.0
20.8
18.9
16.5
20.8
19.5
20.5
20.5
20.3
21.0
21.2
Total
particulate
run time,
min (cycles)
300 (5)
180 (3)
300 (5)
240 (4)
300 (5)
180 (3)
300 (5)
300 (5)
420 (7)
360 (6)
240 (4)
240 (4)
80 (4)
80 (4)
120 (4)
120 (4)
160 (8)
160 (8)
240 (8)
240 (8)
80 (4)
80 (4)
120 (4)
120 (4)
80 (4)
60 (3)
120 (4)
100 (3)
180 (3)
Heating
cycle
segment
tested
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
Total
ON
ON
OFF
OFF
ON
ON
OFF
OFF
ON
ON
OFF
OFF
ON
ON
OFF
OFF
Total
Conditions are at sampling location.
Stoker off; no fuel fed.
-------
document "Compilation of Air Pollutant Emission Factors" (18) .
The EPA report predicts particulate emissions as a function of
coal ash content. However, this study and one performed by
Battelle (2) indicate that ash content has only a secondary effect
on particulate emission rates. Other parameters of equal or
greater importance include coal volatile content, free swelling
index (FSI), ash fusion temperature, and fuel bed temperature.
Particulate emissions from the boiler are correlated with coal
ash level, volatile content, and FSI in Figures 14, 15, and 16.
These figures also include comparison emission data from a study
by Battelle-Columbus Laboratories on a small stoker-fed boiler
(2).
Particulate emissions were found to vary inversely with coal ash
content (Figure 14). This is contrary to estimating procedures
found in Reference 18 and indicates influence by other coal param-
eters. In this case, the FSI is obviously an important parameter.
Figure 15 shows that particulate emissions increased as coal vola-
tile content increased and again demonstrates the significance of
free swelling index. At an FSI less than 1, an increase in coal
volatile content from 38% to 43% increased particulate emissions
16.0
14.0
3 12.0
10.0
6
p
6.0
4.0
2.0
0
COAL A
COAL B
» BATTaLE STUDY. EASTERN BITUMINOUS COAL (2)
BATTELLE STUDY, WESTERN SUBBITUMINOUS COAL (2)
FSI - 5, 7-1/2
6 8 10 12 14
ASH CONTENT OF COAL, %
Figure 14.
Effect of coal ash content on
boiler particulate emissions.
(18) Compilation of Air Pollutant Emission Factors. Publication
AP-42, Part A, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, February 1976.
216 pp.
46
-------
o
16.0
14.0
12.0
10.0
§ 8.0
6.0
o
CD
4.0
2.0
FSI - 5. 7-1/2
COAL A
COALB
BATTELLE STUDY, EASTERN BITUMINOUS COAL (2)
BATTELLE STUDY, WESTERN SUBBITUMINOUS COAL (2)
0 20
Figure 15
25 30 35
VOLATILE CONTENT OF COAL. %
40
45
Effect of coal volatile content on
boiler particulate emissions.
16.0
35 % to 40 % COAL VOLATILE CONTENT
22% COAL VOLATILE CONTENT
COAL A
" COALB
* BATTELLE STUDY, EASTERN BITUMINOUS COAL (2)
» BATTELLE STUDY, WESTERN SUBBITUMINOUS COAL (2)
3 4 5
FSI OF COAL
Figure 16.
Effect of free swelling index on
boiler particulate emissions.
47
-------
from about 3.6 Ib/ton to 8.0 Ib/ton. Likewise, at a fixed coal
volatile content of 40%, a rise in the coal FSI from 1 to 5
increased particulate emissions from 6.0 Ib/ton to 15 Ib/ton. To
examine the effect of FSI on particulate emissions, all emission
factors for coal of the same FSI were averaged and plotted versus
FSI (Figure 16). Results indicate increased particulate emis-
sions with increasing FSI when coal volatile content was between
35% and 40%.
The size range of particulates collected by the SASS train is
given in Table 9. From 80% to 85% of the particulate emissions
during the ON segment of the warm-air furnace heating cycle pass
through the cyclones and are trapped on the filter. The filter
catch normally represents particulates of less than 1-ym diameter;
however, proper operation of the cyclones requires a volumetric
flow rate at the dry gas meter of 3 cfm to 5 cfm. In this pro-
gram, the volumetric flow rate for the SASS runs could not be
maintained above an average of 2.8 cfm. Lower velocities result
in more particles greater than 1 ym passing through the cyclones
and being caught by the filter.
TABLE 9. SIZE DISTRIBUTION OF PARTICULATE EMISSIONS
DURING "ON" SEGMENT OF A COAL-FIRED
RESIDENTIAL FURNACE HEATING CYCLE
Emission factors
SASS run 2, SASS run 3,
Coal B, Coal B,
Particle size low ash content high ash content
range by SASS Percent Percent
train component Ib/ton of total Ib/ton of total
Large cyclone
Intermediate cyclone
Small cyclone
Filter
Total
2.1
0.78
1.1
16
20
10.7
4.0
5.4
79.9
100
1.6
0.42
0.34
14
16
9.9
2.5
2.1
85.5
100
Test run of 100-min duration. Test run of 480-min duration.
Carbon, hydrogen, and nitrogen analysis was also performed on
particulate samples, and results are listed in Table 10 for vari-
ous test conditions. A comparison of these results to the coal
analysis in Table 4 reveals that the particulates emitted
actually have a higher carbon content (75% to 90%) than the feed
coal (68% to 76% dry basis), indicating that much of the particu-
lates are either unburned coal or condensed hydrocarbons. On
run 1, the analysis was carried out on the particulate sample
after it had been extracted with methylene chloride. There was
no significant decrease in carbon content for this run. Conse-
quently, any condensed organic material must be in the form of
nonextractable tars and resins or carbon black-like material.
One SASS filter contained particulates having only 50% carbon.
48
-------
TABLE 10. CARBON, HYDROGEN, AND NITROGEN CONTENT OF PARTICULATE
EMISSIONS FROM COAL-FIRED RESIDENTIAL HEATING SYSTEMS
Composition,
percent of particulate
Run
2
6
1
SASS 2
SASS 3
SASS 3
Sample identification
a
Particulate train filter catch
a
Particulate train filter catch
POM train front half particulate catch
after extraction of organics
b
SASS train filter catch of one filter
b
SASS train filter catch of filter 1
b
SASS train filter catch of filter 2
Carbon
77.84
82.33
79.81
88.66
50.38
83.61
Hydrogen
1.53
1.24
1.97
0.97
0.93
0.93
Nitrogen
0.87
.0.53
0.94
0.50
1.20
1.20
Sample collected from total heating cycle.
Sample collected during ON segment of heating cycle. Each SASS filter
represents approximately 40 min of sampling time.
There is no explanation for this low value. Another filter from
the same run analyzed 81% carbon.
The carbonaceous nature of the particulate emissions explains the
lack of correlation between particulate emission factors and coal
ash content. In large coal-fired combustion units (e.g., utility
boilers), particulate emissions arise from inorganic matter (ash)
contained in the coal. Consequently, emissions increase as the
ash content of the coal rises. However, in residential units the
combustion process is less efficient and particles are apparently
formed by volatilization and subsequent condensation of carbon-
aceous matter in the coal. This hypothesis is supported by the
positive correlation observed between particulate emission
factors and coal volatile content.
Condensable Organics
Condensable organic emissions were determined by measuring the
mass of material collected in the back half of the EPA Method 5
particulate train. The back half of this train consisted of
impingers containing distilled water which, when evaporated, left
an amber-colored residue of condensed coal tars. This material
would condense in the atmosphere and could be considered as part
of the particulate emissions; however, for this study it is
reported separately.
Emission factors for condensable organic material are presented
in Table 7 and ranged from 2.0 Ib/ton to 9.4 Ib/ton. Emission
factors obtained for the furnace were about three times higher
than boiler emission factors for the same coal type. The
difference in combustion geometry that affected the particulate
49
-------
emissions also affected the condensable organic emissions, but to
a lesser degree. There was an inverse correlation of emission
factor to coal ash content when burning Coal B. For Coal A, the
emission factors remained unchanged and were in fact lower than
those for Coal B of the same ash level (Figure 17).
s 10.0
8.0
6.0
o 4.0
g 2.0
o
COAL A BURNED IN BOILER
COALB BURNED IN BOILER
* COALB BURNED IN FURNACE
6 8 10
ASH CONTENT OF COAL, %
12
14
Figure 17,
Effect on coal ash content on
condensable organic emissions.
It was anticipated that condensable organic emissions would be a
function of coal volatile content, but this relationship was not
as strong as that with coal ash content.
Sulfur Oxides
Sulfur oxide emission factors presented in Table 7 ranged from
6.0 Ib/ton to 30 Ib/ton. In terms of coal sulfur content, emis-
sion factors ranged from 15 S Ib/ton to 30 S Ib/ton with an aver-
age of 22 S Ib/ton. (S is the sulfur content of the coal in
weight percent.) The relationship of SOX emissions to coal
sulfur content is shown in Figure 18 together with results for
similar studies. Up to a coal sulfur content of 1.0%, the rela-
tionship is linear, but above 1% coal sulfur content the results
are scattered. The point falling above the maximum SOX line
(calculated SO2 emissions based on coal sulfur content) must be
in error and should probably fall slightly below this line. The
low SOX value may be influenced by the coal lime content. It has
been predicted that the lime (i.e., Calcium) in coal can influence
SOX emissions by combining with sulfur and reducing the formation
of SOX- The 1.5% sulfur coal in Figure 18, which produced slightly
less SOX than the 1.0% sulfur coal, had 50% more calcium than the
50
-------
B
Q£
O
* 30
20
10
MRC STUDY
BATTH1E STUDY <2)
HANGEBRAUCK,ErAL(17)
CURVE BASED ON STOICHIOMETRIC
CONVERSION OF TOTAL SULFUR IN
COAL TO SULFUR DIOXIDE
0.5 1.0
SULFUR CONTENT OF COAL, %
1.5
2.0
Figure 18. Effect of coal sulfur content
on SOx emissions.
1% sulfur coal. These data are insufficient for definite conclu-
sions, but they do demonstrate that a high calcium coal can pro-
duce lower SOX emissions than expected based on sulfur content
alone.
Other test variables did not appear to affect SOX emissions.
Several of the SOx results were ignored because they were
unreasonably low. These values were one to two orders of magni-
tude lower than the remaining values and ranged from 0.1 Ib/ton
to 0.5 Ib/ton of S02, or only 0.2% to 3.0% of the coal sulfur.
It is suspected that a three-way valve on the sampling equipment
was improperly positioned during these tests to that the sample
collected was mostly ambient air.
Nitrogen Oxides
Emission factors for NOX, shown in Table 1, ranged from 2.8 lb/
ton to 14 Ib/ton. NOX emissions were determined from three 30-s
grab samples collected during each test run. Wind gusts caused
the barometric damper to open and close during testing, and fre-
quently changed the dilution of the stack gases, resulting in a
wide variation of NOX levels. Averaging the results of three
grab samples for each test run reduced this variability, but not
to the extent that would be expected with continuous sampling.
Correlation of NOX emissions to test variables was not observed,
and it can be reasonable assumed that these emissions will remain
unchanged with changes in major variables. Previous studies have
reported NOX emissions from 3.2 Ib/ton to 9.8 Ib/ton or within
51
-------
reported NOX emissions from 3.2 Ib/ton to 9.8 Ib/ton or within
the range observed in this study (2, 17).
Carbon Monoxide
Carbon monoxide emission factors presented in Table 7 ranged from
less than 0.08 Ib/ton to 26 Ib/ton. This large variability could
not be explained completely by known variables. However, such
variability has also been observed in tests conducted by other
researchers (2, 17). Apparently CO is the emission most sensitive
to changing fuel bed conditions. CO emissions from the residen-
tial boiler were higher when burning Coal A (0.26 Ib/ton to 4.0
Ib/ton) than when burning Coal B (less than 0.08 Ib/ton). The
higher free swelling index of Coal A may have led to higher CO
emissions due to less complete combustion associated with these
coals. Combustion of Coal B in the warm-air furnace yielded CO
emission factors from 8 Ib/ton to 28 Ib/ton, compared with less
than 0.08 Ib/ton when the same coal was fired in the boiler. As
mentioned in the discussion of particulate emissions, firing geom-
etry of the furnace was significantly different from that of the
boiler and may have influenced combustion efficiency. Excess air
did not appear to influence CO emissions. CO emission factors in
this study are generally lower than those of previous studies, in
which CO emissions varied from 20 Ib/ton to 100 Ib/ton for simi-
lar equipment (2, 17).
POM
Emission factors for polynuclear organic materials ranged from
0.036 Ib/ton to 0.58 Ib/ton (Table 7). Individual POM compound
emission factors are presented later in this section along with
POM emission factors from the separate heating cycle segments.
Combustion of Coal A in the boiler resulted in an average POM
emission factor of 0.42 Ib/ton, while combustion of Boal B in
the warm-air furnace resulted in an average POM emission factor
of 0.052 Ib/ton. Whether this difference is due to the coal
types or combustion equipment cannot be determined from the data.
It would be expected that higher volatile content coals would
have the potential for emitting greater quantities of POM's.
POM emission factors versus coal volatile content are presented
for this study and the Batelle study (2) in Figure 19. Points
representing MRC data are for one test each, while the Battelle
points are averages of four tests each. The Battelle study
showed a high variability of POM emission factors for each test
condition, indicating the influence of other variables. Figure
20, showing POM emission factors versus excess air, indicates
that high excess air reduced POM emissions. The large Difference
between POM emissions from bituminous coal in the Battelle and
MRC studies may result from the differences in combustion equip-
ment design.
52
-------
MRC COAL A BURNED IN BOILER
i MRC COAL B BURNED IN FURNACE
h » BATTaLE BITUMINOUS COAL BURNED IN BOILER 0
' BATTELLE SUBBITUMINOUS COAL BURNED IN FURNACE (21
3
3
0.1
0.01
38 31 40 41 42 43
COAL VOLATILE CONTENT. %
Figure 19. Effect of coal volatile content on POM emissions
i.o
B
a
o
I-H
IS)
(/>
0.1
0.01
MRC COAL A BURNED IN BOILER
MRC COAL B BURNED IN FURNACE
* BATTELLE BITUMINOUS COAL BURNED IN BOILER (2)
» BATTELLE SUBBITUMINOUS COAL BURNED IN BOILER (2)
40 60
100 120 140 160 180 200 220 240
EXCESS AIR, %
Figure 20. Effect of excess air on POM emissions.
53
-------
Elemental Emissions
Combustion releases a portion of the elemental constituents found
in coal to the atmosphere. Measurement of elemental emissions in
this study is the first attempt to characterize the magnitude of
these emissions. Table 11 presents emission factors of 27 ele-
ments for residential combustion of four coal types. Emission
factors are given as pounds of element emitted per ton of coal
burned. Combustion of Coal A in the boiler resulted in emissions
of aluminum, calcium, and iron ranging from 0.014 Ib/ton to
0.036 Ib/ton. Remaining elements had emission factors that were
lower by at least an order of magnitude. Elemental emissions
from the warm-air furnace burning Coal B exceeded those from the
boiler burning Coal A by at least a factor of 2 and in most cases
by a factor of 10. This ratio corresponds to the ratio of partic-
ulate emission factors for the same condition. The highest ele-
mental emissions were aluminum, calcium, iron, and silicon,
ranging from 0.17 Ib/ton to 0.34 Ib/ton. In both cases, elemen-
tal emission factors from the low ash coals were similar to those
from the high ash coals because the particulate composition was
primarily carbon, as discussed earlier in this section.
Discussion of the relationship of elemental emissions to coal
elemental composition is postponed until the discussion of resid-
ual ash elemental content, later in this section.
Organib Spocies
SASS train samples were used to identify hydrocarbon emissions in
the Cj to Ci5 range during the ON segment of the heating cycle.
Emission factors for these are presented in Table 12. Approxi-
mately 85% of the hydrocarbons emitted were between CJ/CQ and C\2
hydrocarbons, with the total Cy to Ci6 hydrocarbon emission
factor averaging 1.1 Ib/ton.
Organic samples collected in the POM trains were analyzed by
GC/MS to identify and quantify the major organic components in
the flue gas from the coal-fired combustion equipment. Table 13
presents the organic compounds identified and their emission
factors for combustion of 1) Coal A in the boiler and 2) Coal B
in the warm-air furnace for each cycle segment of a 20-min ON/
40-min OFF cycle. Over 50 organic compounds were identified, 20
of which are POM's.
A comparison of the total emission factors for all organic
species (6.0 Ib/ton and 4.4 Ib/ton) with the emission factors
for condensatle organic material reported earlier (2.0 Ib/ton to
9.4 Ib/ton) shows excellent agreement. It is concluded therefore
that Table 13 does list all the major chromatographable organic
species emitted from the residential combustion units during
this test program.
54
-------
TABLE 11.
ELEMENTAL EMISSION FACTORS FROM COAL-FIRED
RESIDENTIAL HEATING EQUIPMENT OPERATED ON
A 20-MIN "ON"/40-MIN "OFF" HEATING CYCLE
(Ib/ton)
Boiler burning
Coal A
Element
Aluminum
Antimony
Arsenic
Barium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Tin
Titanium
Vanadium
Zinc
High ash
content
0.014b
<0.002D
0.002
<0.0008
0.002h
<0.006
0.028h
<0.002u
<0.002
<0.001
0.030h
<0.008
0.006
<0.0008 .
<0. 00002°
<0.00045
<0.0006
<0.032P
<0.003
0.006h
<0.002°
<0.015ch
<0.0008
0.004
0.002h
<0.008D
0.008
Low ash
content
0.014
0.00008
0.0002
0.001
0.004
0.0004
0.02
0.0004
0.0002
0.0014
0.036
0.002
0.002
0.0001
0.0004
0.00002
0.0002
0.002 b
<0.0004°
<0.012C
<0.0008C
0.034
0.0006
0.006
0.004
0.0002
0.004
Furnace burning
Coal B
High ash
content3
0.17
0.054
0.004
0.014
0.016
0.0002
0.28
0.001
0.001
0.008
0.22
0.002
0.072
0.001
0.0006
<0.066b
0.012
0.014.
<0.006D
0.17
0.12
0.001
0.004
0.014
0.010
0.004
0.04
Low ash
content
0.18 h
<0.010
0.002
0.014
O.OSOu
<0.001
0.34
0.006
0.001
0.004
0.19
0.024
0.11
0.002 .
<0. 00002
<0.014b
0.068
0.042.
<0.007D
0.19
0.010
0.024
0.002
0.14
0.016
0.006
0.016
Average of duplicate runs.
Value is the detection limit.
°Value is that of the reagent blank used as upper limit.
TABLE 12. EMISSION FACTORS FOR C7 TO C16 HYDROCARBONS
DURING THE "ON" SEGMENT OF A COAL-FIRED
RESIDENTIAL FURNACE HEATING CYCLE
Emission factors
SASS run 2,a
Hydrocarbon
C7
C7 to Ca
Cg
C9
Cii
c12
Cl3
Cm
Cl6
Total
Coal
low ash
Ib/ton
0.004
0.090
0.012
_c
_c
_c
0.016
-C
_c
0.008
-c
0.13
B,
content
Percent
of total
3.1
69.2
9.2
0
0
0
12.3
0
0
6.2
0
100
SASS run 3,b
Coal
high ash
Ib/ton
_c
0.038
_c
0.004
0.008
0.006
0.022
-C
0.002
0.008
0.002
0.090
B,
content
Percent
of total
0
42.2
0
4.4
8.9
6.7
24.5
0
2.2
8.9
2.2
100
Test run of 100-min duration.
Test run of 480-min duration. °None found.
55
-------
TABLE 13. EMISSION FACTORS OF MAJOR ORGANIC SPECIES
FROM RESIDENTIAL COMBUSTION OF COAL
(Ib/ton)
- ;r _ - -
Identified compound
Cs-alkylbenzenes
Ci,-alkylbenzenes
Indane
Indene
Methyl indenes
Phenol
Methyl phenol
Dimethyl phenol
C3-alkyl phenol
Naphthalene
Methylnapthalenes
Dimethylnapthalenes
Cs-alkyl-napthalenes
Ci,-alkyl-naphthalenes
Cs-alkyl-naphthalenes
Biphenyl
Acenapthalene
Fluorene
Methylfluorene
Phenyl phenol
Benzofuran
Methylbenzofuran
Fluorenone
Di-t-butyl cresol
Methyl resorcinols
Anthr 8 tuinone
Methy laurate
Methyl myristate
C i i -alkyl phenol
Methyl palmitate
Methyl stearate
Di-C2-alkyl-phthalate
Di-C 3 -alkyl-phthalate
Di -Ci, -alkyl -phthalate
Di-ethylhexyl-phthalate
Di-octyl-phthalate
Aliphatics (Cii» to C33)
POM:
Dibenzothiophene
Anthracene/phenanthrene
Methylanthracenes/phenanthrenes
9-Methylanthracene
Dimethylanthracenes/phenanthrenes
Fluoranthene
Pyrene
Methy If luoranthenes/pyrenes
Benzo (c) phenanthrene
Chrysene/benz (a) anthracene
7,12-Dimethylbenz (a) anthracene
Benzof luoranthene (s)
Benzopyrene (s) (and perylene)
3-Methylcholanthrene
Indeno (1,2,3 -cd ) pyrene
Dibenz (a,h) anthracene
Dibenzo (c , g) carbazole
Dibenzopyrenes
Methy lchrysenesc >d
Cij-alkyl p^cnanthrerie
Total
i
Coal A
Boiler .-
total cycl
b
~b
07004
0.050
0.058
0.108
0.088
0.032
b
0.56
0.44
0.40
&.176
0.056
0.016
0.018
0.144
0.096
0.064
0.008
b
0704
0.042
b
~b
oToio
0.004
0.004
b
07012
0.006
b
oTooos
0.0012
0.024
0.30
2.6
0.0006
0.058
0.030
<0.0004
0.02
0.02
0.018
0.008
0.0008
0.014
0.30
0.014
0.012
0.006
0.008
0.012
<0.0004
0.034
0.02
b
6.0
Low ash content
Coal
Warm-air
e ON segment
0.086
0.038
b
~b
~b
0.008
0.010
0.004
b
0.030
0.026
0.02
0.014
0.004
0.004
0.0004
0.002
0.0014
0.0016
b
~b
~b
0.0012
0.032
0.20
b
~b
~b
0.010
0.00012
0.00006
0.06
0.002
0.002
0.016
0.002
3.2
0.00004
0.0012
0.0012
<0. 00012
0.002
0.0006
0.0004
0.0002
0.00004
0.0016
0.006
0.00002
0.00004
<0. 00016
0.00002
<0. 00002
<0.0001
<0. 00004
0.0001
_b
3.8
, :
~B
furnace
OFF segment
_b
~b
~b
0.0004
b
0.052
0.122
0.128
0.034
0.004
0.016
0.026
0.038
0.006
0.008
0.002
0.008
0.008
b
~b
^
0.006
0.0014
b
~b
~b
~b
~b
~b
~b
~b
"b
~b
~b
~b
"b
"b
0.036
0.0002
0.006
0.008
<0. 00012
0.010
0.0010
0.0010
0.002
0.00014
0.0012
0.024
0.0012
0.0006
0.0008
0.0006
0.0016
<0. 00008
0.002
0.0012
0.006
0.56
Determined from POM train samples by GC/MS.
Slay include isomers.
Not identified in sample.
Quantitation based on response of benzofluoranthene.
56
-------
species emitted from the residential combustion units during this
test program.
Combustion of Coal A in the boiler produced emissions of at least
50 organic compounds. Aliphatic hydrocarbons between GI and C33
accounted for 44% of the total emission factor of 6.0 Ib/ton;
naphthalene, methylnaphthalenes, and dimethylnaphthalenes
accounted for 24%; and POM's accounted for 10%. Combustion of
Coal B in the furnace also resulted in emission to the atmosphere
of at least 50 organic compounds. The ON segment of the heating
cycle yielded a greater number of organic compounds and in
greater quantities than the OFF segment; however, POM emissions
were much higher in the OFF segment.
Aliphatic hydrocarbons in the range of Cm to C33 were the major
organic constituents of the flue gas during the ON segment,
accounting for 84% of the total organic emission factor of 3.8
Ib/ton. Most of the organic emissions identified in the OFF seg-
ment were phenols, napthalenes, POM's and aliphatics ranging from
Ci*,. to C33. Methyl phenol and dimethyl phenol each accounted for
22% of the OFF segment organic emission factor of 0.56 Ib/ton,
while POM's were about 12% of the total.
A special effort was made to detect polychlorinated biphenyl com-
pounds (PCB's) by GC/MS, but none were found in any of the runs.
Based on the sensitivity of the instrument, it can be concluded
that if such compounds are present, they must have emission
factors less than 10~5 Ib/ton.
ASH RESIDUE AND LEACHATE
Bottom Ash
Attempts to recover ash residue were hindered by the physical
arrangement of the combustion equipment. The coal-fired boiler
had no provisions for physical separation of ash from fuel, and
any attempt to quantitatively recover ash resulted in removal of
partially burned coal from the fuel bed or incomplete recovery of
ash. A fused mass of ash, shown in Figure 21, removed from the
boiler includes pieces of partially combusted coal. For this
reason, further attempts to recover ash from the boiler were
abandoned. The coal-fired warm-air furnace was equipped with a
slide-out ash pan located below the fuel bed. This reduced the
chance of unburned coal contaminating the ash sample. However,
there was the potential for some ash to remain on the fuel bed at
the end of each run. Another problem in accurate ash recovery
resulted from rocks in the coal feed which, although a small
fraction of the coal, became a much larger fraction of the ash.
Figure 22 shows ash clinkers and several rocks that entered the
boiler with the coal and were recovered in the ash. Rocks are
shown as the white or gray objects near the edges of the figure.
57
-------
Figure 21. Ash clinker containing partially
combusted coal from boiler.
Figure 22.
Ash clinkers and rocks found
in boiler bottom ash.
58
-------
Quantities of ash residue recovered from the combustion of Coal B
in the warm-air furnace are listed in Table 14 as pounds of resi-
due per ton of coal burned- Because the amount of residue is
greater than the ash content of the coal, the combustion process
must be incomplete, with unburned coal remaining in the ash pan.
Three samples of residue were analyzed and found to contain from
15% to 56% volatile and combustible material, or 44% to 85%
actual ash. This accounts for the high variability of residue
recovery.
TABLE 14. ASH RESIDUE FROM COAL B BURNED
IN THE WARM-AIR FURNACE
(Ib/ton)
Coal Average
Test ash ash
number content jresidue
13
14
17
18
22
25
100
100
182
182
182
142
600
242
310
298
236
220
Elemental Composition
Elements not emitted to the air through the flue gas leave the
combustion equipment with the combustion residue (ash). Table 15
lists the elemental emission factors for bottom ash from Coal B
as pounds of element per ton of coal and also presents the concen-
tration of the elements in ash. Emission factors for elements in
the ash range from 0.001 Ib/ton to 12 Ib/ton. The highest values
are for aluminum, calcium, iron, and magnesium; they all have
emission factors greater than 2.0 Ib/ton and ash concentrations
greater than 0.7%.
Elemental Material Balance
Elemental emissions are limited by the amount of elements in the
coal. The fractions of elements in the coal emitted to the air
are a function of the chemical and physical properties of the
elements and the conditions in the combustion equipment. One
report estimated that in residential combustion of bituminous
coal, 1% of most elements in coal are released to the atmosphere
in the flue gas (19). For volatile elements, the percentage is
(19) Surprenant, N., R. Hall, C. Young, D. Durocher, S. Slater,
T. Susa, and M. Sussman. Final Report, Volume II: Pre-
liminary Emissions Assessment of Conventional Stationary
Combustion Systems. Contract 68-02-1316, Task Order 11,
U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, March 1976. 523 pp.
59
-------
TABLE 15. ELEMENTS EMITTED FROM THE COAL-FIRED
WARM-AIR FURNACE AS BOTTOM ASHa
Element
Aluminum
Antimony
Arsenic
Barium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Tin
Titanium
Vanadium
Zinc
low
Emission
factor,
Ib/ton
11
0.046
0.002
0.34
0.06
0.001
11
0.07
0.028
0.094
9.4
0.052
4.2
0.11
<0. 000002
0.010
0.084
0.52
0.0004
0.034
0.042
0.90
0.22
0.02
0.84
0.12
0.12
Coal B,
ash content
Concentration
in ash, %
2.6
0.011
0.0006
0.082
0.014
0.0003
2.7
0.017
0.007
0.022
2.2
0.012
0.99
h °-026 b
b <0. 0000005
0.002
0.02
0.12
0.00007
0.008
0.01
0.22
0.053
0.006
0.20
0.028
0.028
high
Coal B,
ash content
Emission
factor , Concentration
Ib/ton in ash, %
13
0.062
0.002
0.26
0.030
0.004
7.4
0.058
0.030
0.14
9.8
0.032
2.0
0.096
<0. 000002
0.016
0.086
0.66
<0.002C
0.034
0.07
0.12
0.30
0.022
1.2
0.12
0.012
4.2
0.021
0.001
0.085
0.01
0.001
2.4
0.019
0.01
0.046
3.2
0.012
0.67
. 0.031 h
D <0. 0000005
0.006
0.029
0.22
<0.001C
0.013
0.023
0.045
0.097
0.007
0.40
0.04
0.004
20-min ON/40-min OFF heating cycle.
Value is the detection limit.
cValue is that of the reagent blank used as upper limit.
greater: Releases of 90% for mercury and 70% for selenium were
predicted. The value of 1% is believed to be a good estimate
based on test results in this program for particulates. Particu-
late emission factors were found to vary from 2.8 Ib/ton to 44
Ib/ton; i.e., approximately 1% of the mass of the coal is emitted
to the air. Analysis for carbon, hydrogen, and nitrogen revealed
the carbon content of the particulate samples (approximately 80%)
was slightly higher than that in the coal (approximately 70%).
It follows that elemental concentrations in the particulate emis-
sions should be similar to those in the coal for the less vola-
tile elements. Consequently, emission factors for these elements
would be in the same ratio as particulates; i.e., 1% of the mass
of an element in coal would be emitted to the air.
Emissions of individual elements from the combustion equipment
are compared to the concentrations of those elements in the feed
coal in Table 16. The amounts of elements emitted to the air
represent less than 10% of each element in the coal feed in most
cases. Elements emitted to the air at 10% to 50% of the coal
elemental content were antimony, molybdenum, lead, nickel, silver,
cobalt, and arsenic, while those emitted at greater than 50% of
the coal elemental content were silicon, molybdenum, silver, tin,
zinc, selenium, and mercury. Other studies have reported all of
these elements except silicon as being enriched in fly ash or
60
-------
TABLE 16. FRACTION OF COAL ELEMENTAL CONTENT EMITTED TO
THE ATMOSPHERE AND TOTAL MATERIAL BALANCE3
(percent)
Coal B, low ash content
Element
Air
emission
Solid
residue"
Total
Coal B, high ash content
Air i Solid .
emission residue Total
Aluminum
Antimony
Arsenic
Barium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Tin
Titanium
Vanadium
Zinc
1.8
<36
33
5.4
1.0
8.6C
4.1
7.0
35
6.1
3.2
<18
3.8
4.8
100
25°
20
6-8
9.7°
70
19
0.6
1.5
>128
2.7
5.3
>2,000
110
164
33
131
1.2
>10
136
81
700
142
157
>38
140
267C
10
>18
25
84
>0.6
12
78
23
169
>19
140
104
>14,750
112
<200
66
136
2
19C
140
88
735
148
160
d
144
272
110r
43C
45
91c
ioc
82
97
24
171
>147
143
109
>16,750
1.4
19
40
4.4
1.0
2.2
1.8
4.0
7.6
1.8
>1.5
2.1
1.5
3,000
118°
6.0
2.0
<250
70
400
0.2
1.2
>13
1.1
3.0
>667
105
22
20
81
2
>29
58
74
100
132
80
>23
59
100
ioc
>29
43
94
<83
14
233
28
88
>20
130
91
>200
106
41
60
85
3
>30
60
76
104
140
82
>25
61
102
3,010*:
147C
49
96
<333
84
633
28
89
>33
131
94
>867
Warm-air furnace employing a 20-min ON/40-min OFF heating cycle.
emission factor, Ib/ton
coal content, Ib/ton
(100)
Value determined by using the detection limit in numerator and
denominator; actual value may be lower or higher.
Cannot be determined from data.
potentially volatilized during combustion (19, 20). Large emis-
sions of silicon are unlikely, and it is suspected that contam-
ination from glass containers accounts for these high values.
Zinc emissions are much higher than possible based on coal zinc
content and probably represent emissions from combustion equip-
ment components such as the galvanized exhaust stack.
Quantities of elements found in bottom ash exceeded the amount
in the feed coal for 13 elements in low ash Coal B and for 6
elements in high ash Coal B. However, only cobalt, manganese,
(20) Magee, R. A., F. B. Meserole, and R. G. Oldham. Coal-fired
Power Plant Trace Element Study, Volume I: A Three Station
Comparison. U.S. Environmental Protection Agency, Denver,
Colorado, September 1975. 75 pp.
61
-------
silver, and zinc were unreasonably high. It was difficult to
accurately recover the bottom ash, and an error of up to 50%
would be expected in any quantification of bottom ash constitu-
ents. The unexpectedly high zinc levels probably arose from the
galvanized steel ash pan. A material balance indicates that
about one-half of the elements emitted could be accounted for
within 50% of the quantity fed to the furnace. The extremes were
boron, cadmium, and sodium, where less than 30% of the material
in the feed coal was recovered in air emissions and ash residue,
and cobalt, manganese, zinc, and silver, where the amount of ele-
ments leaving the furnace was 3 times to 170 times higher than
the input.
Leachate
The elemental content of ash residue may be an environmental prob-
lem if the ash is exposed to rainfall and elements in the ash
leach out and enter water supplies. A test was conducted on
bottom ash from the furnace to quantify the leachate material.
Ash was shaken for 24 hr in distilled water at an ash-to-water
ratio of 1 to 10, and the elemental content of the water was
determined after shaking. Final pH of the water rose to 11.6
from an initial pH of 7.1, reflecting the alkalinity of the ash
and the calcium content of the coal.
Emission ff \_ors for the leached material are presented in
Table 17 as pounds of element leached per ton of bottom ash.
Also shown is the fraction of elements in the ash that leached to
the water. About half of the elements were found to be below the
detection limits, corresponding to emission factors of less than
0.002 Ib/ton with some as low as less than 0.0002 Ib/ton. Highest
quantities leached were from aluminum, barium, calcium, silicon,
sodium, and strontium, ranging from 0.02 Ib/ton to 12 Ib/ton.
Elements most susceptible to leaching were calcium, sodium,
selenium, silicon, mercury, and strontium with 20% to 95% of the
element leached from the ash. These were followed by barium,
boron, and molybdenum at about 2%.
DISCUSSION OF RESULTS
Major Process Variables Affecting Emissions
Major process variables investigated during this study include
1) type of combustion equipment, 2) type of coal, 3) ash level in
the coal, and 4) ON versus OFF segments of the heating cycle.
Obviously the variation in coal type encompasses a number of
other parameters; e.g., sulfur level, free swelling index, vola-
tile content, and trace element levels. The variation in emis-
sions with coal type and coal ash level have been covered
previously in this section. Therefore, this discussion centers
on the other two major parameters.
62
-------
TABLE 17. ELEMENTS LEACHED FROM ASH REMOVED FROM
THE COAL-FIRED WARM-AIR FURNACE
Element
Amount leached
per quantity
of ash, Ib/ton
Fraction of element
in ash leached
to water, %
Aluminum
Antimony
Barium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Tin
Titanium
Vanadium
Zinc
0.024
<0. 0004
0.038
0.004
<0.0001
11.6
<0. 00004
<0.0002
<0.0002
<0. 00002
<0.001
0.004
<0. 00002
<0. 000004
0.002
<0.001
<0.002
0.0006
0.094
<0.0008
2.4
0.30
<0.0008
<0. 00004
<0.0001
<0. 00004
0.035
<0.13
2.3
1.9
<0.77
23
<0.011
<0.12
<0.029
<0. 00004
<0.42
0.024
<0.004
<92
2.0
<0.20
<0.059
74
45
<0.23
91
20
<0.62
<0.0007
<0.015
<0.013
Combustion Equipment
An unambiguous evaluation of the effect of combustion equipment
on emissions would require that all other variables be held
constant. It was impossible to accomplish this in the present
study because of the large number of variables and limited number
of tests. However, the effect of many variables appears less
significant than the effect of the combustion equipment, and
certain conclusions can be drawn.
Emission factors for the warm-air furnace and boiler burning
Coal B are presented in Table 18. In general, the furnace gener-
ated more emissions than the boiler, except in the case of POM's
where measurements were not made on identical fuels. Particulate
emissions were eight times higher from the furnace than from the
boiler. Emissions of SOX were essentially the same, in support
of the hypothesis that variables other than coal sulfur and calci-
um content exert little if any influence on SOX emissions.
NOX emissions during the ON segment of the heating cycle were 2
times higher from the warm-air furnace than those from the boiler,
Measurements were not made on the total heating cycle of the
boiler because of program limitations. The higher NOX emissions
63
-------
TABLE 18. AVERAGE EMISSION FACTORS FROM THE COMBUSTION
OF COAL B IN THE RESIDENTIAL BOILER AS
COMPARED TO THE RESIDENTIAL FURNACE
(Ib/ton)
Emission species
Particulates
S0xa
NOX3
CO .
POM
Condensable organics
Emission
Residential
boiler
4.4
20
4.6
<0.08
0.42
3.6
factors
Residential
furnace
32
22
8.2
22
0.05
7.4
Measurements for the ON segment of the heating
cycle only.
POM emissions from the boiler were measured
while burning Coal A and therefore may also
indicate the influence of the coal type.
from the furnace indicate that the furnace operated at higher
temperatures and/or longer residence times. Other factors that
influence NOX formation, namely fuel nitrogen content and excess
air, were not a factor in these tests. Fuel nitrogen remained
unchanged because the same coal was used in both units, and
excess air overall was the same for the furnace and the boiler.
CO emissions from the furnace were almost two orders of magnitude
higher than those from the boiler. Increased CO emissions indi-
cate a loss in combustion efficiency. Condensable organic mater-
ials are also products of incomplete combustion, and they too had
higher emission factors from the furnace.
POM emission factors were higher by an order of magnitude from
the boiler, in contrast to the results for CO and condensable
organic material. However, in the case of POM's, two different
test coals were used, and the results demonstrate the effect of
coal volatile content discussed earlier in this section. No con-
clusions can be drawn with regard to combustion efficiency
because both the coal type and the combustion equipment varied.
Heating Cycle
Automatic coal-fired heating equipment differs from other forms
of automatic heating in that a bed of fuel continues to burn when
the demand for heat is satisfied and the combustion-air fan shuts
off. Consequently the equipment continues to release pollutants
to the atmosphere during the OFF period. However, the nature of
the combustion process and the flue gas composition are not the
same as during the ON period.
In this program, emissions from each cycle segment were quanti-
fied for the warm-air furnace burning Coal B. Results are pre-
sented in Table 19 as pounds emitted per hour of operation. A
64
-------
TABLE 19. COMPARISON OF EMISSIONS FROM THE ON AND
OFF SEGMENTS OF THE WARM-AIR FURNACE
HEATING CYCLE WHILE BURNING COAL B
(10-3 lb/hr)
Emission species
Particulates
so*
NO*
CO
POM
Condensable organics
Elements :
Aluminum
Antimony
Arsenic
Barium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Tin
Titanium
Vanadium
Zinc
Low ash
content
ON
240
97
42
191
0.055
26
0.86
<0.009
0.013
0.062
0.26
<0.0004
2.4
0.02
0.009
0.02
1.3
<0.020
0.84
0.018
<0. 00004
<0.079
0.055
0.062
<0.013
0.48
0.013
0.19
0.018
0.35
0.13
0.015
0.12
OFF
53
4.2
31
0
0.21
_b
0.26
<0.033
0.004
0.020
0.062
<0.007
0.059
C.013
0.009
<0.004
0.092
<0.086
0.020
<0.001
0.007
<0.026
0.24
<0.14
<0.012
0.51
0.035
<0.058
<0.0004
<0.37
<0.011
0.018
<0.029
Medium ash
content
ON
134
176
92
1.33
b
44
b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
>
OFF
29
20
2.6
90fl
b
13
b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"b
"fa
"b
"b
"b
"b
"b
"b
"b
"b
~b
High ash
content
ON
174
229
48
190
0.059
15
0.84
<0.086
0.024
0.046
0.033
<0.001
1.2
0.013
0.004
0.042
1.2
<0.011
0.44
0.013
0.0001
0.086
<0.040
<0.075
0.0002
0.35
<0.51
0.057
0.029
<0.077
0.064
0.011
0.20
OFF
26
6.4
2.6
2.2
0.12
12
0.26
<0.21
0.007
0.020
0.040
<0.002
0.20
0.029
0.013
0.002
0.2
<0.075
0.051
<0.0009
0.0002
<0.033
0.018
<0.001
0.011
0.48
<0.099
<0.046
<0.006
0.29
<0.009
0.007
0.007
Average
ON
183
167
61
127
0.057
28
0.85
<0.048
0.019
0.055
0.15
<0.0007
1.8
0.017
0.007
0.031
1.3
<0.016
0.64
0.016
<0. 00007
<0.083
<0.048
<0.069
0.007
0.42
<0.26
0.12
0.024
<0.21
0.097
0.013
0.16
OFF
36
10
12
31
0.17
13
0.26
<0.12
0.006
0.020
0.051
<0.005
0.13
0.021
0.011
<0.003
0.15
<0.081
0.036
<0.001
0.004
<0.030
0.13
<0.071
<0.012
0.50
<0.067
<0.052
<0.003
<0.33
<0.010
0.013
<0.018
Data are not consistent with the majority of the results but are consistent with Battelle data.
b
No data collected.
similar comparison was made for organic species in Table 13.
Emission rates are presented rather than emission factors because
it was impossible to measure separately the actual quantity of
coal burned during the ON and OFF segments of the heating cycle.
An estimate of the relative amounts of coal burned during each
segment can be made based on the emission rates of SOX. Sulfur
oxides are formed solely by the oxidation of sulfur in the coal.
Therefore their emission rate is proportional to the amount of
coal burned. The ratio of SOX emission rates for ON versus OFF
ranges from 10:1 to 50:1. Consequently, the relative amount of
coal burned during each segment of the heating cycle will be in
approximately the same ratio. Emission rates in Table 19 can
65
-------
therefore be converted into relative emission factors by multi-
plying the OFF values by a factor of 10 to 50.
Emission rates for all emission species in Table 19 were higher
during the ON segment except for POM emissions which were about
five times higher in the OFF segment. Because POM emissions are
products of incomplete combustion of volatiles, this finding was
not unexpected. If emission rates are converted into relative
emission factors, then values for particulates, condensable
organic materials, and NOX are also higher during the OFF period.
Again, this is not surprising except in the case of NOX.
The behavior of CO is puzzling. CO is a product of incomplete
combustion and should be generated in the largest amounts during
the OFF period when insufficient oxygen exists for stoichiometric
combustion. Battelle found this relationship to hold true in
their study where CO emission rates were higher during the OFF
segment (2). The difference in CO emission rates between this
study and that conducted by Battelle cannot be fully explained,
although the operation of the barometric damper may be the key to
any logical explanation. The position of the barometric damper
can play an important role in determining the nature and extent
of combustion during the OFF segment. Battelle chose to position
the damper "closed" during all of their sampling. By doing this,
any draft induced by the hot flue gas or meteorological condi-
tions pul'ed air through the combustion chamber and through the
hot fuel ^ed. In contrast, sampling for the present study was
conducted with the barometric damper free to open and close as
the stack pressure drop varied. As a result, a minimum amount of
draft was induced through the combustion chamber during the OFF
segment. The actual effect of this variation in damper operation
during the OFF segment is unknown.
Emission rates for particulates were about five times higher
during the ON segment due to more rapid combustion and higher air
velocities through the fuel bed that entrained more particles in
the flue gas. Elemental species and NOX were emitted from the
furnace at rates 17 and 5 times higher, respectively, during the
ON segment.
Emissions data for this study were determined for heating equip-
ment operating at a 20-min ON/40-min OFF heating cycle. This
cycle was determined to be representative of the average total ON
and total OFF time for a heating season (see Appendix A). Fur-
naces controlled by thermostat may have ON/OFF cycles of shorter
duration. Extrapolation of the data obtained in this study to
predict emissions for other heating cycles is possible if the
emission rateb are not strongly dependent on the length of the ON
and OFF periods. Although variations in the heating cycle were
not studied during this investigation, it is believed that condi-
tions during the ON and OFF periods approximate steady-state
conditions. It is true that if the furnace is left ON too long
66
-------
it will overheat, and if it is left OFF too long the fire will
die out. However, automatic timing mechanisms prevent this from
happening in actual equipment, so that the steady-state assump-
tion is reasonable.
Battelle observed that particulate and POM emissions were not sig-
nificantly different during transient conditions (initial ON and
initial OFF) (2). This evidence supports the belief that emis-
sion factors measured over a 20-min ON/40-min OFF heating cycle
can be used to predict emissions for cycles of different lengths.
The effect of heating cycle length on hourly emission rates is
predicted in Figure 23. The linear relationship observed depends
on the validity of the above assumption. The lines are not con-
tinued through complete OFF or complete ON because no coal-fired
heating equipment operates in this manner. Emission factors for
the lower percentages of ON time are representative of the early
and late parts of the heating season while the emission factors
for high percentages of ON time represent the coldest part of the
heating season. Meteorological data in conjunction with this
information can be used to determine annual emission rates.
Comparison of SASS Train Results
Three SASS train sampling runs were originally scheduled for this
program. Mechanical problems prevented one run from being con-
ducted and forced another to be terminated early. The two major
problems affecting these runs were 1) a pinhole in a weld that
caused a leak in the organic module and 2) a galled fitting on
the filter holder that prevented filter changes.
Because the SASS train uses a series of three cyclones to separ-
ate particles by size range, it is important to maintain the
proper velocity flow through the train. During the two actual
test runs, this could not be done. A layer of fine particles
quickly built up on the filter element, causing a large pressure
drop and slowing the flow of flue gas through the train. Filters
had to be changed periodically to continue sampling. This low
velocity through the cyclones resulted in more large particles
passing through them and depositing on the filter. Consequently
particulate size fraction data presented earlier are reported
only in a qualitative way.
Even though the flow velocity through the cyclones was lower than
design specifications, sampling rates were about 280% isokinetic
versus the 90% to 110% required for emission compliance sampling.
Inasmuch as the SASS train is designed to obtain qualitative or
semiquantitative data, such nonisokinetic conditions are some-
times unavoidable. In this case, the resulting particulate data
from the SASS train are biased on the low side because fewer of
the larger particles were collected compared to the number that
would have been collected under isokinetic conditions.
67
-------
s
o
o
en
en~
o
cc.
o
CD
<
on
O
o
on"
LU
o
DC
Q.
UJ
o:
z
o
on
0.16 -
0.14 -
0.12 -
CONDENSABLEORGANICS
0.02 -
30 40 50 60 70 80
DURATION OF " ON " SEGMENT, percent of 1 hr
Figure 23.
The effect of heating cycle
variations on emission factors.
Sampling was performed under the same test conditions using both
the SASS train and a modified Method 5 train. Species measured
were particulates, organic material, POM's, and trace elements.
A comparison of test results for the two methods is given in
Table 20. Particulate emission rates were approximately 20%
lower using the SASS train, and condensable organics and POM's
were 60% and 77% lower, respectively. As expected, the particu-
lates collected by the SASS train were slightly less than those
from Method 5. However, with the larger sample volume collected
by the SASS train, it was expected that POM's and organics would
be collected more efficiently and therefore yield larger emis-
sion factors. It may be that the physical nature of the organic
material in the flue gas affected the collection efficiency
during nonisokinetic sampling. For example, some of the organic
material could have been present as droplets or condensed on
particulates. Elemental emission rates varied, but there were
68
-------
TABLE 20. COMPARISON OF EMISSION DATA FROM THE SASS TRAIN
TO CONVENTIONAL SAMPLING METHODS3
(10-3 Ib/yr)
Emission rate
Emission species
Particulate
Organic material
POM
Trace elements:
Aluminum
Antimony
Arsenic
Barium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Tin
Titanium
Vanadium
Zinc
SASS
Low ash
coalb
154 f
8.1 f
0.013
0.88
<0.004
0.022
0.048
0.46
0.002
2.2
0.040
0.013
0.024
1.1
<0.009
0.68
0.02
0.00009
0.018
0.077
0.099
0.024
0.79
0.015
0.35
0.018
0.70
0.20
0.015
0.097
train
High ash
coalc
128 f
8.1T f
0.013
0.84
<0.0004
0.022
0.024
0.013
<0.002
0.16
0.031
0.004
0.011
1.2
<0.004
0.40
0.02
0.00009
<0.007
<0.04
<0.095
0.0009
0.53
<0.004
0.17
0.026
0.024
0.064
0.009
0.24
Method
Low ash
coal
187d
26 a
0.0559
0.84
<0.011
0.007
0.070
0.064
0.002
2.9
0.004
0.004
0.020
1.5
<0.031
0.99
0.015
<0. 0000007
<0.14
0.033
0.026
<0.002
0.17
<0.009
<0.08
0.17
0.024
0.053
0.013
0.14
5
High ash
coal
163f
14 h
0.059
0.86
<0.13
0.024
0.057
0.042
0.001
1.7
0.004
0.004
0.055
1.3
0.013
0.46
0.011
0.0002
<0.13
0.040
0.07
<0.002
0.29
<0.75
<0.07
0.029
0.103
0.064
0.011
0.17
Emissions from the warm-air furnace ON segment of the heating
cycle while burning Coal B.
SASS run 2. CSASS run 3.
Data are from run 13 sampled simultaneously to SASS run 2.
O
Data are from runs 17 and 18 sampled simultaneously to SASS run 3.
Average value from SASS runs 2 and 3; samples were accidentally
combined.
Data are from run 14 sampled at conditions similar to those for
SASS run 2 but not sampled simultaneously.
Data are from runs 21 and 22 sampled at conditions similar to
those for SASS run 3 but not sampled simultaneously.
69
-------
about as many lower values as there were higher values when using
the SASS train, probably due to the close similarity in particu-
late emission rates.
Comments on Results and Conclusions
This investigation represents one of the most comprehensive pro-
grams undertaken to characterize emissions from coal-fired resi-
dential combustion equipment. Nevertheless, test results apply
only to the types of combustion equipment and varieties of coal
sampled. Extrapolating these results to other test conditions
should be done carefully. Because data obtained during this pro-
gram are in general agreement with other studies on residential
coal combustion, a certain degree of extrapolation seems justifi-
able. Further study on different combustion systems and a
broader range of coal types would provide valuable information on
variations in emissions, and is to be recommended.
Emissions data collected during this study are believed to be
typical of what would actually be observed in homes heated with
coal. Combustion systems, coal quality/ and operating conditions
were all chosen to fall within, although not encompass, the
normal range expected in the residential sector. Test results
show that all three parameters influence emission rates, and that
there is a wide variation in emission rates even under normal
conditions.
In contrast to previous predictions, particulate emissions were
not a function of coal ash content but instead correlated with
coal free swelling index and volatile content. Whereas in larger
combustion systems particles are formed from inorganic ash in the
coal, in these residential units particles were formed from the
organic material in the coal, and the particulate composition
averaged about 80% carbon. This behavior is a consequence of
incomplete combustion in the residential units studied.
Because combustion efficiencies for coal-fired residential combus-
tion systems were lower than those for larger units such as
utility boilers, emissions of organic species were relatively
high. Ash residues contained a high percentage of unburned
carbon and, as just mentioned, particulate emissions were primar-
ily carbonaceous materials. Because the organic emissions
included a number of carcinogenic polynuclear organic compounds
(e.g., benzopyrene), the trend toward greater residential coal
usage may have a significant impact on local air quality. It is
suggested that this impact be evaluated in terms of its potential
health and environmental effects before a large number of home-
owners install coal-fired heating equipment.
70
-------
REFERENCES
1. Personal communication with John S. Kinsey, Colorado Depart-
ment of Health, Denver, Colorado, 1 September 1976.
2. Giammar, R. D., R. B. Engdahl, and R. E. Barrett. Emissions
from Residential and Small Commercial Stoker-Coal-fired
Boilers Under Smokeless Operation. EPA 600/7-76-029, U.S.
Environmental Protection Agency, Research Triangle Park,
North Carolina, October 1976. 77 pp.
3. Personal communication with John S. Kinsey, Colorado Depart-
ment of Health, Denver, Colorado, 16 May 1977.
4. Personal communication with Robert D. Giammar, Battelle-
Columbus Laboratories, Columbus, Ohio, 9 June 1977.
5. Domestic Stokers, Hopper and Bin Feed by Will-Burt (manu-
facturer's brochure). Form W346-75-2M, The Will-Burt Co.,
Orrville, Ohio. 4 pp.
6. Swanson, V. E., J. H. Medlin, J. R. Hatch, S. L. Coleman,
G. H. Wood, S. D. Woodruff, and R. T. Hildebrand. Collec-
tion, Chemical Analysis, and Evaluation of Coal Samples in
1975. Open-File Report 76-468, U.S. Department of the
Interior, Denver, Colorado, 1976. 503 pp.
7. 1975 Annual Book of ASTM Standards, Part 26: Gaseous Fuels;
Coal and Coke; Atomspheric Analysis. American Society for
Testing and Materials, Philadelphia, Pennsylvania, 1975.
792 pp.
8. Personal communication with Stratton C. Schaeffer, Consult-
ing Engineer, Camp Hill, Pennsylvania, 10 November 1977.
9. Standards of Performance for New Stationary Sources.
Federal Register, 36 (247):24876-24895, 1971.
10. Detector Tube Handbook, Air Investigations and Technical
Gas Analysis with Drager Tubes, 2nd Edition, compiled by
Kurt Leichnitz. Dragerwerk Ag., Lubeck, Federal Republic
of Germany, October 1973. 164 pp.
71
-------
11. Bacharach Fyrite Gas Analyzers for Mearuring Carbon Dioxide
(C02) or Oxygen (02). Bulletin 4042/2-75, Bacharach
Instrument Company, Pittsburgh, Pennsylvania, 1975. 2 pp.
12. Hamersma, J. W., S. L. Reynolds, and R. F. Maddalone.
IERL-RTP Procedure Manual: Level I Environmental Assessment,
EPA-600/2-76-160a, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, June 1976. 147 pp.
13. Metals by Atomic Absorption Spectrophotometry. In: Stand-
ard Methods for the Examination of Water and Wastewater,
14th Edition. American Public Health Association, Washing-
ton, D.C., 1976. 1193 pp.
14. Jarrell-Ash Plasma AtomComp for the Simultaneous Determina-
tion of Trace Metals in Solutions (manufacturer's brochure).
Catalog 90-975, Jarrell-Ash Company, Waltham, Massachusetts.
5 pp.
15. Bernas, B. A New Method for Decomposition and Comprehensive
Analysis of Silicates by Atomic Absorption Spectrometry.
Analytical Chemistry, 40 (11):1682-1686, 1968.
16. Hartstein, A. M., R. W. Freedman, and D. W. Platter. Novel
Wet-Digestion Procedure for Trace-Metal Analysis of Coal
by Atomic Absorption. Analytical Chemistry, 45(3):611-614,
197^.
17. Hancjebrauck, R. P., D- J. Von Lehmden, and J. E. Meeker.
Emissions of Polynuclear Hydrocarbons and Other Pollutants
from Heat-Generation and Incineration Processes. Journal
of the Air Pollution Control Association, 14(7):267-278,
1964.
18. Compilation of Air Pollutant Emission Factors. Publication
AP-42, Part A, U.S. Environmental Protection Agency,
Research Triangle"Park, North Carolina, February 1976.
216 pp.
19. Surprenant, N., R. Hall, C. Young, D. Durocher, S. Slater,
T. Susa, and M. Sussman. Final Report, Volume II: Prelim-
inary Emissions Assessment of Conventional Stationary Com-
bustion Systems. Contract 68-02-1316, Task Order 11, U.S.
Environmental Protection Agency, Research Triangle Park,
North Carolina, March 1976. 523 pp.
20. Magee, R. A., F. B. Meserole, and R. G- Oldham. Coal-fired
Power P^ant Trace Element Study, Volume I: A Three Station
Comparison. U.S. Environmental Protection Agency, Denver,
Colorado, September 1975. 75 pp.
72
-------
21. Statistical Abstracts of the United States 1975. U.S.
Department of Commerce, Washington, D.C., July 1975.
1050 pp.
22. Current Housing Reports; Bureau of the Census Final Report
H-150-74; Annual Housing Survey: 1974, Part A; General
Housing Characteristics for the United States and Regions.
U.S. Department of Commerce, Washington, D.C., August 1976.
179 pp-
23. Mineral Industry Surveys, Bituminous Coal and Lignite Dis-
tribution, Calendar Year 1974. U.S. Department of the
Interior, Washington, D.C., 18 April 1975. 74 pp.
24. 1973 NEDS Fuel Use Report. EPA 450/2-76-004 (PB 253 908) ,
U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, April 1976. 124 pp.
25. Guide for Compiling a Comprehensive Emission Inventory
(Revised). Publication No. APTD-1135, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina,
March 1973. 204 pp.
26. Standard for Metric Practice. ANSI/ASTM Designation E 380-
76e, IEEE Std 268-1976, American Society for Testing and
Materials, Philadelphia, Pennsylvania, February 1976. 37 pp.
73
-------
APPENDIX A
DETERMINATION OF AVERAGE HEATING CYCLE
To find the average heating cycle for residential combustion
systems, it is first necessary to calculate the average annual
heat load. This can be done by weighting the average number of
degree-days for each state by the number of housing units burning
bituminous coal in each state. Therefore,
50
Weighted average of heating degree-days =
["(Degree-days), (housing units burning bituminous coal in state i) . ~|
L U.S. total housing units burning bituminous coal J
where i = each state
Table A-l lists the average heating degree-days for each state
for a 30-yr period from 1941 to 1970 (21) and the number of
housing units that burned bituminous coal in 1974 (22, 23, 24).
Using these data in Equation A-l yields an average heat load of
4,470 degree-days.
(21) Statistical Abstracts of the United States 1975. U.S.
Department of Commerce, Washington, D.C., July 1975.
1050 pp.
(22) Current Housing Reports; Bureau of the Census Final Report
H-150-74; Annual Housing Survey: 1974, Part A; General
Housing Characteristics for the United States and Regions.
U.S. Department of Commerce, Washington, D.C., August 1976.
179 pp.
(23) Mineral Industry Surveys, Bituminous Coal and Lignite Distri-
bution, Calendar Year 1974, U.S. Department of the Interior,
Washington, D.C., 18 April 1975. 74 pp.
(24) 1973 NEDS Fuel Use Report. EPA 450/2-76-004 (PB 253 908),
U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, April 1976. 124 pp.
74
-------
TABLE A-l. ANNUAL HEATING DAYS FOR ALL STATES (21-24)
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Degree-days
1,684
9,007
1,552
3,354
2,331
6,016
6,350
4,940
4,211
767
3,095
0
5,833
6,113
5,577
6,710
4,687
4,640
1,465
7,498
4,729
5,621
7,806
9,034
2,300
4,956
7,652
6,049
6,022
7,360
4,946
4,292
6,221
3,366
9,044
5,642
3,695
4,792
5,398
5,972
2,598
7,838
3,462
2,134
5,983
7,876
3,714
6,010
4,590
7,444
7,255
Number of housing units
burning bituminous coal
in 1974b
20,384
779
92
971
286
3,427
0
0
3,333
183
9,412
34
5,114
28,889
4,771
1,421
174
51,321
22
0
0
575
4,773
1,782
2,002
1,406
1,125
272
175
0
0
248
357
18,207
0
9,923
165
613
40,625
0
8,842
309
47,885
134
4,464
0
35,426
2,991
29,926
3,227
875
Data in Reference 21 are given for major cities in each
state. For this study, it was assumed that these numbers
approximated state averages.
Values are derived from data in References 22, 23, and 24.
75
-------
Annual fuel usage is then letermined by the amount of coal needed
per degree-day per dwelling unit. This is reported in Reference
25 to be 2.4 Ib/degree-day per dwelling unit. Consequently, the
average annual coal usage is 5.4 tons/yr. For a typical heating
season of 212 days (October 1 to April 30), the corresponding
average hourly feed rate is:
10,730 Ib/yr /
(212 days/yr) (24 hr/day) z-x iD/nr
In order to determine the heating cycle, it is necessary to know
the actual furnace feed rate. In Reference 25, the dwelling unit
was defined as a five-room residence. A home of this size is
generally heated with a 100,000-Btu furnace which can be assumed
to operate at 75% load and burn coal with a heating value of
11,500 Btu/lb. Therefore the furnace feed rate is:
(100,000 Btu/hr) (75%) _ fi , .. /.
11,500 Btu/lb 6 ' 5 lb/hr
Thus the average hourly heating cycle must be (2.1/6.5)(60 min) =
19 min ON and 41 min OFF. During this program, two 200,000-Btu
combustion units were tested, but it is believed the average
heating cycle is similar because these units are found in larger
home s.
(25) Guide for Compiling a Comprehensive Emission Inventory
(Revised). Publication No- APTD-1135, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina,
March 1973. 204 pp.
76
-------
APPENDIX B
DETERMINATION OF UNCERTAINTY CAUSED BY LOW STACK GAS VELOCITY
Low stack gas velocities during the OFF portion of the heating
cycle were difficult to measure because manometer readings were
often near zero. Precautions were taken to maintain a reason-
able degree of accuracy by zeroing and leveling the manometer
every 10 min to 15 min.
A micromanometer having a 6-in. scale was used to measure a veloc-
ity head (AP) of 0 to 0.25 in. of water. Major divisions on this
scale are in increments of 0.01 in. of water with minor divisions
of 0.005 in. of water. A velocity head reading of 0.005 in. of
water is represented by 0.06 linear inch on the scale and can be
read reasonable well. However, during the OFF segment of a heat-
ing cycle, stack gas velocities were so low that observed AP's
were less than 0.005 in. of water, and actual AP's were often
estimated to be about 0.001 in. of water. Obviously, there is a
great degree of uncertainty in attempting to read 0.001 in. of
water AP, even with a micromanometer. It is known, however, that
the readings were not actually zero because the manometer fluid
showed movement above zero, and visible emissions were observed
coining from the stack during the OFF segments. Considering this,
AP readings of 0.001 in. of water were the best estimates for
field measurement.
Manometer readings of 0.001 in. of water occurred during most of
the OFF segments of heating cycles in 8 of the 13 tests conducted
on the boiler. For the remaining five tests, these low veloci-
ties were experienced in less than 50% of the readings during the
OFF segment. Particulate samples that are affected by these
measurements were collected over the total boiler heating cycle.
The tests conducted during the ON segment of the furnace had AP
readings of greater than 0.01 in. of water, and only two of the
eight tests conducted during the OFF segment resulted in AP read-
ings of 0.001 in. of water.
Because AP readings of 0.001 in. of water are uncertain, it is
important to estimate the relative uncertainty and determine its
effect on emission factors. This can be done by estimating the
upper and lower limits on the AP measurements and using them to
recalculate emission factors. A reading of 0.0005 in. of water
was chosen as the lower limit (50% of actual reading), and a AP
77
-------
of 0.0025 in. of water was chosen as the upper limit (150% higher
than actual reading). The lower limit was arbitrarily chosen as
midway instead of zero because stack gas flow was visually
observed. The upper limit was chosen as the minimum reading that
can reasonable be estimated on a micromanometer.
These values were substituted for all AP readings of 0.001 in.
of water in two Method 5 particulate sampling runs. Run 1 was a
particulate sample collected from the boiler and integrated over
the ON and OFF segments of six heating cycles. Run 23 was a par-
ticulate sample collected from the furnace during the OFF seg-
ments of four heating cycles. All of the AP readings during the
OFF segments were 0.001 in. of water. Table B-l shows the
effects of this substitution on stack gas velocity, sampling per-
cent isokinetics, and particulate and POM emissions. For run 1
the substitutions result in a decrease of 16% and an increase of
24% in the particulate emission rate of the total heating cycle.
Substitution in run 23 results in a decrease of 24% and an
increase of 47% in the particulate emission rate during the OFF
segment. However, the combined emission rates for the total heat-
ing cycle were only lower by 5% and higher by 7% than that calcu-
lated from actual field data. This is because the ON segment had
higher emission rates than the OFF segment, and an error in veloc-
ity measurement during the OFF segment had a much smaller effect
on the overall emission rate.
TABLE E 1.
EFFECT OF STACK GAS VELOCITY MEASUREMENT ERROR ON
PARTICULATE AND POM EMISSION RATE DETERMINATION
Test run 1
Calculated parameter
Case A
Average stack gas velocity, fpm 141
Average stack gas flow rate, acfm 49
Percent isokinetic 109
d
Case B
164
57
94
a
e
Case C
209
73
74
Test run 23
Case A
99
34
131
d
Case B
129
45
101
b
e
Case C
188
66
69
Particulate emission rate, Ib/hr:
OFF segment
ON segment
Overall^
POM emission rate, Ib/hr:
OFF segment
ON segment
OverallS
f
f
0.021
f
f
0.00078
f
f
0.025
f
~f
0.00093
f
f
0.031
f
"f
0.0012
0.016
0.218
0.083
0.000083
0.000067
0.000078
0.021
0.218
0.087
0.00011
0.000067
0.000096
0.031
0.218
0.093
0.00016
0.000067
0.00013
Integrated sampling of the boiler. Sampling of OFF heating cycle segment of warm air furnace.
Case A: All manometer readings of 0.001 in. of water changed to 0.0005 in. of water
d
Original field d\ca used.
g
Case C: All manometer readings of 0.001 in. of water changed to 0.0025 in. of water.
Heating cycle segments were not sampled separately. ^20-min ON/40-min OFF heating cycle.
78
-------
In the case of POM's, where the emission rate is greater during
the OFF segment, the combined emission factors deviated from the
field measurement to a greater degree. The combined results were
lower by 19% and higher by 35% than when using actual field data.
This degree of inaccuracy is tolerable, being within the varia-
tion expected from field data.
Another factor to consider is that errors in velocity measurement
can adversely affect sampling isokinetics and result in biased
particulate measurement. However, nonisokinetic conditions are
important only when larger-sized particles (larger than 10 ym)
are entrained in the gas stream. In the present study, the flow
velocities during the OFF phase were so low that these particles
could not have been carried from the fuel bed and suspended in
the exhaust gas.
79
-------
GLOSSARY
anthracite: Dense hard coal with a very high precentage of
fixed carbon, usually above 90%. Most anthracite is mined
in Pennsylvania.
barometric damper: Balanced butterfly gate located outside the
flue gas flow on an exhaust stack in a tee fitting. It is
used to control draft.
bituminous: Coal covering a wide range of properties, but in
general with a fixed carbon content under 80% and volatile
matter over 20%.
boiler: Closed vessel in which fuel is burned to generate
steam or heat water.
clinker: Fused mass of the residue (ash) from coal combustion.
criteria pollutants: Those for which air quality standards have
been established.
cycle: Pattern of operation of automatic heating equipment
characterized by four segments; initial ON, steady-state
ON, initial OFF, and steady-state OFF.
damper: Valve or plate used to regulate the flow of air to a
combustion process.
draft: Pressure difference causing flow of a fluid, usually
applied to convection flow as in chimneys.
emission factor: Quantity of emissions per quantity of mass
burned.
free swelling index: Measure of the caking properties of coal
determined by the percent swelling of a heated coal.
flue: Enclosed passage for conveying combustion gases to the
atmosphere.
80
-------
lignite: Brown, noncaking woody coal with high moisture and
low heating value. It is usually mined in Texas and
North Dakota.
proximate analysis: Fuel analysis on the basis of percent
fixed carbon, volatile matter, moisture, and ash.
rank of coal: Method of classifying coal by chemical and
physical properties.
retort: Cast iron chamber in the shape of a cup or trough
used to devolatilize and ignite coal in a stoker-feed
furnace or boiler.
stoker: Mechanical device used to feed solid fuel to a com-
bustion unit.
tuyeres: Slotted holes in the retort for directing combustion
air to the fuel bed of a coal-fired heating device.
ultimate analysis: Fuel analysis on the basis of elemental
content; namely, carbon, hydrogen, oxygen, nitrogen,
sulfur, and ash.
underfeed: Method of feeding solid fuel to a fuel bed where the
fresh fuel is introduced from beneath the fuel bed.
underfire: Method of providing combustion air to a fuel bed by
forcing the air through the fuel bed from underneath.
warm-air furnace: Closed vessel in which fuel is burned to heat
room air.
wind box: Enclosure around a stoker which directs combustion air
through the retort.
worm-feed mechanism: Motor-driven screw which conveys solid fuel
from storage and discharges the fuel into the retort of a
combustion operation.
81
-------
CONVERSION FACTORS AND METRIC PREFIXES (26)
To convert from
Degree Fahrenheit
Degree Celsius
Pound-mass
Pounds/hour
British thermal unit
(Btu)
Pound mass
(avoirdupois)
Ton (short, 2,000
lb mass)
Pound mass/foot3
Mile2
Foot
Inch
Foot3
Pound-mass
Pound-force/in2
(psi)
CONVERSION FACTORS
To
Gram (g)
Gram/second
Joule (J)
(g/s)
Multiply by
Degree Celsius (°C)
Kelvin (°K)
t« = (1
4 =
*C +
32)/1.8
273.15
Kilogram (kg)
Kilogram (kg)
Kilogram/meter3 (kg/m3)
Kilometer2 (km2)
Meter (m)
Meter (m)
Meter3 (m3)
Metric ton
Pascal (Pa)
4.535 x 102
1.260 x 10"1
1.055 x 103
4.535 x 10-1
9.074 x 102
1.602 x 101
2.591
3.048 x ID"1
2.540 x 10~2
2.832 x 10~2
4.535 x IQ-1*
6.897 x 103
Prefix Symbol
Giga
Mega
Kilo
Milli
Micro
G
M
k
m
y
METRIC PREFIXES
Multiplication
factor
109
106
103
io-3
10~6
Example
!Gg=lxl09 grams
1 MJ = 1 x IO6 joules
1 kPa = 1 x 103 pascals
1 mg = 1 x 10"3 gram
1 ym = 1 x 10~6 meter
(26) Standard for Metric Practice. ANSI/ASTM Designation E 380-
76£, IEEE Std 268-1976, American Society for Testing and
Materials, Philadelphia, Pennsylvania, February 1976.
37 pp.
82
-------
4. TITLE AND SUBTITLE Source Assessment: Coal-tired Resi-
dential Combustion Equipment Field Tests, June 1977
TECHNICAL REPORT DATA
(riease read Instructions on the reverse before completing)
REPORT NO.
EPA-600/2-78-0040
2.
3. RECIPIENT'S ACCESSION-NC
5. REPORT DATE
June 1978
6. PERFORMING ORGANIZATION CODE
.AUTHOR(S)
D. G. DeAngelis and R. B. Reznik
8. PERFORMING ORGANIZATION REPORT NO
MRC-DA-786
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Monsanto Research Corporation
1515 Nicholas Road
Dayton, Ohio 45407
10. PROGRAM ELEMENT NO.
1AB015: ROAP 21AXM-071
11. CONTRACT/GRANT NO.
68-02-1874
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task; 9/76-5/78
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES JERL-RTP task officer is Ronald A.
541-2547.
Venezia, Mail Drop 62, 919/
16. ABSTRACT
The report gives results of a study to quantify criteria pollutants and char-
acterize other atmospheric emissions from coal-fired residential heating equipment.
Flue gas was sampled from a warm air furnace and a hot water boiler which burned
three western coals. Tests were conducted with the stokers operating on a 20-minute
ON/40-minute OFF cycle, corresponding to high- and low-fire conditions in the
fuel bed. Variations in coal composition and type of heating equipment both influenced
emission rates, and the OFF portion of the heating cycle contributed significantly to
total emissions. The report gives a number of correlations between emission rates
and test parameters. Combustion efficiencies for coal-fired residential heating
equipment were lower than for larger coal-fired systems (e.g., utility boilers), as
evidenced by the higher emission levels for CO and organic species, including
POMs. In contrast to previous estimates, particulate emissions were not a function
of the coal ash content, but did correlate with the coal free swelling index and
volatile content. The particulate composition was primarily carbon, indicating that
the particles were not formed from coal ash but from carbonaceous material vola-
tilized during combustion.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI I icId/Group
Pollution Flue Gases
Industrial Processes Sampling
Assessments Warm Air Heating
Coal Hot Water Heating
Combustion Carbon Monoxide
Heating Equipment Volatility
Residential Buildings Dust
Pollution Control
Stationary Sources
Free Swelling Index
Particulate
Polynuclear Organic
Material
13B
13H
14B
21D
21B
13A
13M
07B
20M
11G
8. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
94
2O. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
83
------- |