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

-------