EPA-600/R-9 5-106
July 1995
EFFECTS OF CHANGING COALS ON THE
EMISSIONS OF METAL HAZARDOUS AIR
POLLUTANTS FROM THE COMBUSTION OF
PULVERIZED COAL
Final Report
By
C. Andrew Miller
Air Pollution Prevention and Control Division
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
U.S. EPA Interagency Agreement RWIL935725-01-0
Prepared for:
The Illinois Clean Coal Institute
P.O. Box 8
Carterville, Illinois 62918
and
U.S. Environmental Protection Agency
Office of Research and Development
Washington, D.C. 20460
-------
TECHNICAL REPORT DATA /-
(Please read ImOuctions on the reverse before comple
1. REPORT NO.
EPA-600/R-95-106
illinium
PB.95-246385.....
4. TITLE AND SUBTITLE
Effects of Changing Coals on the Emissions of Metal
Hazardous Air Pollutants from the Combustion of
Pulverized Coal
5. REPORT DATE
July 1995
6, PERFORMING ORGANIZATION CODE
7, AUTHORIS)
C. A. Miller
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
See Block 12
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
EPA IAG RWIL93572 5-01-1
12, SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 6/92 - 12794
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES
541-2920.
APPCD project officer is C. Andrew- Miller^MaiUDrop-65y--&L-9-,/..
p/, r
1B*ABSTRA^The report dicusses tests,
conducted at.'EPA;'s'*Air Pollution Prevention and
Control Division, to evaluate the effects of changing coals on emissions of metal haz
ardous air pollutants from coal-fired boilers. Six coals were"burned in a 29-kW
(100,000 Btu/hr) down-fired;.combustor,1 under similar conditions. Flue gases were
sampled for 10 metals: antimony, arsenic, beryllium, cadmium,-- chromium, lead,
manganese, mercury, nickel, and selenium. No general correlation was found be-
tween coal sulfur content and metal emissions. The tests showed correlations be-
tween as-fed metal content in the coal, and uncontrolled measured emissions of the
metal changed as the coals changed. The factor determining the degree of correlation
appears to be metal vapor pressure. The study illustrates that predictions of metal... -
emissions based only on the trace metal content of the coal may not be accurate,'"The
small scale results were also compared with field data. Field measurements up-
stream of any pollution .control equipment "showed the small scale results to be 30
lower than the full scale emissions for manganese, nickel, and selenium, with re-
maining emissions similar between the two scales. However, trends between the two
scales are expected to be similar, to the degree that similar changes in coals are
made.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c, cosati Field/Group
Pollution
Coal
Pulverized Fuels
Combustion
Emission
Toxicity
Metals
Pollution Control
Stationary Sources
Hazardous Air Pollutants
13 B
2 ID
2 IB
14 G
06T
07 B
18, DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO, OF PAGES
87
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 <9-73)
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
ii
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with protecting the Nation's
land, air, and water resources. Under a mandate of national environmental laws, the Agency
strives to formulate and implement actions leading to a compatible balance between human
activities and the ability of natural systems to support and nurture life. To meet these mandates,
EPA's research program is providing data and technical support for solving environmental
problems today and building a science knowledge base necessary to manage our ecological
resources wisely, understand how pollutants affect our health, and prevent or reduce
environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for investigation of
technological and management approaches for reducing risks from threats to human health and the
environment. The focus of the Laboratory's research program is on methods for the prevention and
control of pollution to air, land, water, and subsurface resources; protection of water quality in
public water systems; remediation of contaminated sites and groundwater, and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze development and
implementation of innovative, cost-effective environmental technologies; develop scientific and
engineering information needed by EPA to support regulatory and policy decisions; and provide
technical support and information transfer to ensure effective implementation of environmental
regulations and strategies.
This publication has been produced as part of the Laboratory's strategic long-term research plan. It
is published and made available by EPA's Office of Research and Development to assist the user
community and to link researchers with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
k
s
IL
I^
111
-------
Abstract
A series of tests were conducted at the U.S. Environmental Protection Agency's National
Risk Management Research Laboratory, Air Pollution Prevention and Control Division
(APPCD), formerly the Air and Energy Engineering Research Laboratory, to evaluate the effects
of changing coals on the emissions of metal hazardous air pollutants (HAPs) from coal-fired
boilers. The tests were conducted on a small scale combustor, and samples were taken prior to
any pollution control equipment to allow application of different control efficiencies to the
uncontrolled emissions. Six different coals were burned in APPCD's Innovative Furnace
Reactor (IFR) under the same combustion conditions, and each coal was sampled for 10 metals:
antimony, arsenic, beryllium, cadmium, chromium, lead, manganese, nickel, selenium, and
mercury. Each of these metals is on the list of 189 compounds and compound classes listed as
HAPs under Title III of the 1990 Clean Air Act Amendments. The results of the tests showed
that changes in the uncontrolled emissions tended to correlate well with the corresponding
changes in the as-fed metal content of the coals in the cases of mercury, selenium, and arsenic.
For beryllium, chromium, manganese, and nickel, changes in the uncontrolled emissions with
different coals did not correlate well with the changes in the as-fed trace metal contents. The
remaining three metals, antimony, cadmium, and lead, did not show conclusive results when
comparing emissions to as-fed trace metal contents. The factor that determines the degree of
correlation between the as-fed trace metal concentration and the uncontrolled stack emissions
appears to be the vapor pressure of the metal. Metals that have high vapor pressures tend to
exhibit strong correlations between the as-fed metal concentration in the coal and the
uncontrolled emissions, while metals with low vapor pressures tend to show a much weaker
correlation. In summary, the study illustrates that predictions of metal emissions based only on
the trace metal content of the coal do not yield accurate results in all cases. Such predictions
cannot be used with any confidence for refractory metals, but do have some degree of validity for
the more volatile metals of interest.
Based on average emission factors (in lb/106 Btu), the Illinois coals had higher emissions of
arsenic, beryllium, cadmium, lead, and selenium than did the western coals, while the western
coals had higher emissions of chromium, manganese, and nickel. Antimony and mercury
emissions were similar for both coals. These results must be carefully interpreted, however,
given the limited scope of testing and the fact that the tests were conducted on a small scale unit.
Comparisons of the small scale results with U.S. Department of Energy (DOE) field data
show that the highest measured full scale emissions tend to be lower than the small scale results, •
with the exception of mercury, which was higher in the DOE field data. This is in contrast to the
data taken-from a full scale test using one of the coals used in this study. In this case,
measurements upstream of any pollution control equipment showed the small scale results to be
30-50% lower than the full scale emissions for manganese, nickel, and selenium, with the
remaining emissions being similar between the two tests. The correlation between full scale and
small scale emissions remains unclear in general. However, trends seen in the small scale tests
are expected to be similar to trends from the full scale testing, to the degree that similar changes
in coals are made.
-------
Table of Contents
Abstract Lf-V
List of Tables
List of Figures "Y"•
Executive Summary -j i
I. Introduction 1
II. Metal Emissions from the Combustion of Pulverized Coal 2
A. Partitioning and Enrichment 3
B. Trace Metals in Coal 6
IE. Project Description..... 7
A. Facility and Equipment 7
B. Experimental Approach 8
C. Sampling and Analysis 12
D. Quality Assurance and Quality Control 12
IV. Results 13
A. Calculated Emissions 14
B. Measured Emissions 16
C. Comparison of Measured and Calculated Emissions 17
V. Discussion 18
A. Scaled Results 18
B. Comparison of Small Scale Results to Full Scale Tests 27
C. Effects of Air Pollution Control Systems 30
D. Effects of Coal Cleaning 33
E. Other Approaches to Predicting Emissions 33
VI. Conclusions 34
VII. References .,35
Appendix A. Data Sheets 37
I, Coal Analysis Data Sheets 37
II. Sample Analysis Data Sheets 51
-------
List of Tables
Table Page
1. Submicron flyash elemental enrichment in coal combustion
investigations 5
2. As-received proximate analysis of the six coals used in the combustion
tests, 10
3. Dry proximate and ultimate analyses of the six coals used in the
combustion tests. 10
4. Trace metal concentrations for the six coals used in the combustion tests 11
5. Maximum calculated trace metal emissions 15
6. Measured trace metal emissions 16
7. Ratio of measured trace metal emissions to calculated emissions 18
8. Full scale emission factors for hazardous metal air pollutants, from DOE
. test program 27
9. Removal efficiencies of air pollution control devices for seven metals 31
V
-------
List of Figures
Figure Page
1. Potential pathways of metal compounds in a utility boiler 4
2. Schematic of Innovative Furnace Reactor 8
3. Schematic of coal burner used on IFR .,9
4. IFR exhaust duct showing sampling locations 9
5. Concentrations of trace metals in as-fed coal for the six coals tested. 11
6. Maximum calculated emission factors for the six coals tested. 15
7. Measured emission factors for the six coals tested 17
8. Comparison of scaled calculated and measured emissions of antimony
for the six coals tested 19
9. Comparison of scaled calculated and measured emissions of arsenic
for the six coals tested. 20
10. Comparison of scaled calculated and measured emissions of beryllium
for the six coals tested 21
11. Comparison of scaled calculated and measured emissions of cadmium
for the six coals tested. 22
12. Comparison of scaled calculated and measured emissions of chromium
for the six coals tested. 22
13. Comparison of scaled calculated and measured emissions of lead for
the six coals tested 23
14. Comparison of scaled calculated and measured emissions of
manganese for the six coals tested 23
15. Comparison of scaled calculated and measured emissions of nickel for 1
the six coals tested. 24
16. Comparison of scaled calculated and measured emissions of selenium
for the six coals tested. 25
17. Comparison of scaled calculated and measured emissions of mercury
for the six coals tested. ,25
18. Comparison of measured emissions from small scale test program to
DOE field test data. 28
19. Comparison of uncontrolled emissions from a full scale pulverized
coal utility boiler to those from the small scale IFR. 29
ft
x
, 1 '
\!
-------
Executive Summary
I, Introduction
Title III of the 1990 Clean Air Act Amendments includes significant provisions for the
reduction of hazardous air pollutants (HAPs) from a wide range of sources, including combustion
sources. The Act defines 189 compounds and compound classes as HAPs, and requires
maximum available control technology (MACT) to be applied to sources that emit over 10 tons
(9.07 tonnes) per year of any single HAP, or 25 tons (22.7 tonnes) per year of any combination
of HAPs. These provisions do not immediately apply to utility units, pending the results of a
Congressionally mandated Study and Report to Congress on the risk to human health from the
emissions of HAPs from utility boilers. That study is currently being conducted by EPA's Office
of Air Quality Planning and Standards.
The provisions of Title III may result in a significant impact on coal-fired utility boilers
because of the presence of metal compounds in coal, many of which are listed as hazardous
under Title III. The immense volumes of coal burned in a large boiler may result in annual
emissions of such compounds that exceed the 10 ton per year level, making utility boilers subject
to the application of MACT if the Title III provisions are applied to utility boilers without
modification.
Other provisions of the Act require utilities to reduce their total annual emissions of sulfur
dioxide (SO2). Because the sulfur content of some coals is lower than others, one control
strategy that has been applied is to replace the use of higher sulfur coals with those having a
lower sulfur content, thereby removing the need for installation of flue gas desulfurization
systems. However, this approach (known as fuel switching) will also alter the emissions of metal
compounds due to the differences in metal content of the different coals. This adds an additional
factor to the process of planning for compliance with the provisions of the Act.
II. Behavior of metals in coal combustion systems
Coal naturally contains a variety of different metal compounds, and the combustion of
pulverized coal results in the release of those compounds from the coal into the combustion
environment. Unlike organic compounds, which can be chemically broken down into more basic
and nonhazardous constituents, metal compounds are not chemically transformed by the
combustion process to the degree that they lose their toxic nature. Therefore, when a fuel-bound
metal compound is introduced into a combustion environment, it must exit the furnace or
combustor. However, the routes each metal may take through the combustor can differ based on
combustor design, operating conditions, or the properties of the coal.
vi 11
J
-------
The percentage of a given metal that exits via a particular effluent stream is referred to as the
"partitioning" behavior of that metal toward that stream. Different metals will exhibit different
partitioning behavior for similar conditions, and changes in those conditions can change the
partitioning behavior of a single metal. Of considerable concern are the metals which exit the
combustor with the flue gases, since these may result in emissions into the atmosphere if
adequate pollution control strategies have not been implemented. The metals which pass through
the combustion process may exit as a metal vapor, an aerosol composed of small particles, in
condensed form on larger solid-phase particles entrained in the flue gases, or as a combination of
the three.
In many instances, one or more of the effluent streams will exhibit a concentration of a given
metal that is higher than would be expected if the metal were evenly distributed among all the
various streams. This partitioning behavior is known as "enrichment" of that particular stream
by that metal. Enrichment is of interest because air pollution control systems are typically much
less efficient in removing particles smaller than 1 [im in diameter than they are in removing
larger particles. Larger particles are often removed from the flue gases even without air pollution
control systems, due to their inability to remain entrained in the flue gases. Thus, if metals
partition preferentially to the smaller size ranges, they are more likely to be emitted from the
stack, even if air pollution control systems are present. The combustion conditions, presence of
other compounds, and the way the metal is contained in the coal matrix can all affect the
partitioning behavior of the metal. Thus, the partitioning and enrichment behavior make the
prediction of metal compound emissions very difficult, even when the metal content of the coal
is well known.
III. Experiment
To study the behavior of metal emissions from different coals in a single combustion
environment, the Illinois Clean Coal Institute and the National Risk Management Research
Laboratory, Air Pollution Prevention and Control Division (formerly the Air and Energy
Engineering Reseach Laboratory) of the Environmental Protection Agency's Office of Research
and Development entered into a cooperative research project. This project was designed to
combust several different coals in the same small scale research combustor, and sample and
analyze the resulting flue gases for metal compounds.
Six different coals were tested during the test program. An Illinois #5 and Illinois #6, two
western coals, and two other "test comparison" coals were used. The last two coals were used in
other tests evaluating HAP emissions from coal to provide comparison with other tests, including
a full scale field test. The Illinois coals were higher in sulfur than the two western coals, and
represented the "high sulfur" cases, with the western coals representing the "low sulfur" cases.
ix
L ' \1^ •
-------
The coals were burned in APPCD's Innovative Furnace Reactor (IFR), which is a down-fired
refractory-lined furnace rated at 29.3 kW (100,000 Btu/hr), capable of firing natural gas, fuel oil,
or pulverized coal. Continuous emission monitors (CEMs) were used to provide continuous
readings and records of the levels of carbon monoxide (CO), nitrogen oxides (NOx), sulfur
dioxide (SO2), oxygen (02), and carbon dioxide (CO2).
Combustion conditions were chosen to most closely simulate the combustion of pulverized
coal in a typical utility boiler. A stoichiometric ratio of 1.25 (25% excess air) was chosen to
provide a stable flame and relatively low CO and NOx values. Gas and particulate samples were
taken upstream of any flue gas cleaning equipment in order to evaluate the uncontrolled emission
levels. The metals emissions were collected using the Multiple Metals sampling train (MMT),
and mercury emissions using the Method 101A sampling train.
The project was conducted following an EPA-approved Quality Assurance (QA) Project
Plan. CEMs were calibrated before and after each test ran to check zero and span. Samples were
collected in complete accordance with EPA method guidelines. Duplicate samples were
collected for each test condition, and field and method blanks were used to evaluate potential
contamination during sampling and analysis. The sampling and analysis procedures followed
during the testing were identical to those that are required under more stringent QA levels.
The major problem affecting data quality was the contamination of one sample with
permanganate solution during the sampling and analysis process. Extremely high levels of
manganese were found in this sample (WC2 test number 1) resulting from permanganate
contamination, and this value was not included in the data used to calculate emissions. Analyses
of the method blanks showed that the analysis of all metals except manganese correctly indicated
the levels of of metal compounds sampled from the flue gases. For manganese, the reported
levels are (at worst) higher than actually present in the gases, but are of the same order of
magnitude.
IV. Results
The study compares the measured results to the maximum possible emissions, both on a per
unit energy basis. Using an energy-specific basis for comparison accounts for differences in
energy, ash, and moisture contents in the different coals. An estimate of the maximum possible
emissions of metals exiting the stack is based on an assumption that 100% of the metals' mass
passes through the combustor to the stack. This assumption is extremely simplified, but such
calculations can provide a benchmark against which measured metal emissions can be compared.
On a per unit energy basis, the maximum theoretical metal emissions (or the calculated
emissions) can be determined by dividing the trace element concentration of the raw coal by the
energy content of the coal, resulting in a maximum emission factor in pounds per trillion (lO1^)
Btu or grams per megajoule. The measured emissions factors per unit-energy were determined
-------
by measuring the total mass of metal in each collected sample and dividing by the total amount
of energy input from that test run.
For the calculated emissions, the highest theoretical emissions were found for chromium,
lead, manganese, and nickel, with a maximum value of 1.34x10* lb/1012 Btu for manganese
from one of the "comparison coals" (coal B2). Cadmium and mercury yielded the lowest of
these calculated maximum emissions, with the lowest calculated maximum being 6.41 lb/10^
Btu for mercury from coal B2. These results are similar to the distribution of metals
concenu'ations in the coals; however, some differences are noted due to the difference in heating
values between the coals, which in general increase the emission factors of the western coals WT .
and W2 relative to the Illinois coals.
While the metals that were found in the highest concentrations in the coals tended also to
result in the highest measured emission factors, there were considerable differences in the
measured emisson factor results and the as-found concentrations. As was the ease for the
calculated emissions, the highest measured emissions were for nickel, lead, and chromium, with
the maximum measured emission being 2740 lb/1012 Btu for nickel from coal Wl, The lowest
measured emissions were from mercury and cadmium, with the minimum being 0.14 lb/1012 Btu
for cadmium from coal B2.
The ratio of measured to calculated emissions of metal HAPs can provide a considerable
amount of information. In some cases, the measured emissions may exceed the calculated
emissions (which are based on 100% of the metal in the coal exiting the combustor). This can be
due to one or more reasons: the measured value of the trace metal content of the coal is in error,
and indicates a lower level than is actually present; the measured value of the concentration of
the metal in the flue gas is in error, and indicates a higher level than actually present; or both
measurements may be correct, but the particular sample of coal used to determine the trace metal
content of the coal had a lower than average amount of the particular metal. Because of the low
concentrations of trace metals, the accuracy of the measurements can also add to the uncertainty
of the final results. However, these sources of error and uncertainty do not remove the
usefulness of this ratio. By focusing on consistent results and trends, much can be determined
from these values. First, if the ratio of theoretical to measured emissions of a given metal is
similar for a wide range of coals, then it is likely that the partitioning behavior of the metal is
similar for the different type coals tested, and may therefore be more dependent on the
combustion and furnace characteristics than on the coal type or composition.
As noted above, there is some concern that the measurements of cadmium, chromium, lead,
manganese, and nickel in coal B2 were higher than the levels actually in the coal. This
possibility is strengthened by the very low percentages of those metals that are measured in the
stack relative to the rest of the coals. While there is no quantitative evidence from the laboratory
XI
-------
data that the as-received trace metal analyses of coal B2 were faulty, the comparisons of these
values with the other cases indicate that these values may not be as reliable as the remaining data.
V. Discussion
To enable a clearer view of how the emissions change between the different coals, both the
calculated and measured emissions results are linearly scaled to the maximum value in each
respective category, and the two plots are presented side by side. This provides two distributions
of emissions, one presenting a relative view of how the emissions from the different coals would
be if 100% of the metal were emitted from the stack, and the other presenting the relative view of
how the emissions were measured. This enables an immediate comparison of how the emissions
of the tested metals are affected by their presence in the different coals. For example, the relative
emission levels of nickel change significantly when comparing calculated emissions based on the
as-fed trace metal concentration to the measured emissions from the six different coals. On the
other hand, the comparison of the scaled calculated and measured emissions for arsenic or the six
different coals tested showed relatively little difference in the distributions. This information-
indicates that the relative levels of emissions from burning the six different coals are roughly the
same as the relative levels of arsenic found in the as-received coals.
The data gathered from these plots indicate that the more volatile metals such as mercury and
arsenic tend to be emitted in similar ratios for all coal types. However, the more refractory
metals such as nickel differ significantly between the different coals, indicating that there arc
mechanisms that affect the pardoning of these metals that change as the coal characteristics
change. This is not a new result, but one that must be re-emphasized as increasing focus is
placed on predicting toxic metal emissions using the trace metal content of the coal. The ability
to predict how metals behave in combustion systems is currently beyond the state of the art.
Therefore, broad statements regarding the behavior of these metals during the combustion of coal
cannot be considered accurate in general.
A further comparison can be made to full scale results that used the same coal to evaluate the
ability of small scale tests to predict behavior in full scale systems. This was done using one of
the test coals, which was taken from a coal pile being used during tests of a prototype pollution
control system. The prototype system was tested on a slip stream of flue gases from a full scale
utility boiler, and emissions upstream of the pollution control system should accurately reflect
the uncontrolled emissions in a manner similar to those measured in the small scale system used
in these tests. However, it is reasonable that the uncontrolled pilot scale emissions will be lower
than uncontrolled emissions from full scale units, for several reasons: (1) lower gas velocities in
the pilot scale compared to full scale units, making it more difficult for particles to remain
entrained in the flue gases; (2) a smaller ratio of gas volume to wall surface area in the pilot scale
unit, leading to increased condensation of the metals in the flue gas at the small scale; and (3)
\ \
\ : \ I S
XI i yN/\ A
-------
tighter gas turns in the pilot scale that are likely to lead to increased amounts of particles being
deposited on the walls and removed from the flue gases. These changes should affect the
refractory metals to a greater degree than the volatile metals, and lower values were indeed seen
in the small scale for chromium and manganese. However, the small scale results also were less
for mercury and selenium, indicating other scale-related factors may also play a role. It is
important to keep in mind, however, that although the small scale emissions of mercury were six
times higher than those in the full scale, there was only a net difference of approximately 200 fig
in the total sample collected. This small change in total sample collected highlights the difficulty
in drawing significant conclusions based on such small levels of a compound.
VI. Conclusions
The primary conclusion to be drawn from this study is that there are mechanisms and
processes that influence the partioning of metal compounds during the combustion of pulverized
coal that are dependent upon the eharacterisics of the coal being burned. Because these
mechanisms are not well understood, particularly in quantifiable ways, the prediction of metal
emissions based on the trace metal content of the coal is likely to be inaccurate. While trends
can likely be predicted based on changes in trace metal content, quantifying the levels of stack
emissions is likely to involve a significant level of uncertainty, even when removal efficiencies
of pollution control equipment are well known. These conclusions are especially true for
refractory metals. While the results indicate that such predictions are more likely to be adequate
for volatile metals, much is yet unknown regarding the behavior of trace metal compounds in
pulverized coal combustion systems.
In general, the Illinois coals tended to show higher emissions of arsenic, beryllium, cadmium,
lead, and selenium than did the western coals, based on average emission factors (in pounds per
trillion Btu). The western coals had higher average emissions of chromium, manganese, and
nickel than did the Illinois coals. Antimony and mercury emissions were similar for both Illinois
and western coals. Because of the limited number of tests, these results cannot be used to make a
general conclusion that Illinois coals are higher emitters of air toxics than western coals. The
differences in trace element content and combustion behavior of those elements are not well
enough quantified to generate such conclusions based on the results of these tests. The use of a
small scale combustor also introduces a measure of uncertainty into how these results can be
compared at full scale.
Based on comparisons of the highest measured emissions from DOE field tests, the full scale
emissions tend to be lower than the small scale results, with the exception of mercury. In the
case of mercury, the highest measured emission from a full scale unit was higher than any of the
small scale emissions. No information was given regarding the coal origin or sulfur content in
the DOE field data, making it impossible to compare emissions based on these categories.
-------
Additional comparisons between full scale and small scale emissions were made based on a
series of full scale tests using a coal which was also used in this study. In this case, full scale
measurements upstream of any pollution control equipment showed emissions of manganese,
nickel, and selenium to be 30-50% lower for the small scale than for the full scale, with the
remaining emissions being similar between the two tests.
In general, there is no clear correlation between small scale and full scale results. However,
with the exception of manganese, nickel, and selenium, there was relatively good correlation
between the two tests, and trends seen in the small scale tests are expected to be similar to trends
from the full scale testing, to the degree that similar changes in coals are made. It must be kept
in mind that the most common generalization that has been noted in testing of air toxics
emissions is that there is significant scattering of results even for repeat tests at a single site using
a single coal. The degree of variability in the trace element contents and the operating
characteristics of any particular unit make it extremely difficult to obtain repeatable data over
any period of time, and discrete data points must be compared with caution.
-------
I. Introduction
In 1990, the Clean Air Act Amendments of 19901 (CAAAs) were passed into law, resulting
in lowering the emission limits for the acid rain precursor gases of sulfur dioxide (SO2) and
nitrogen oxides (NOx) from existing utility sources, with the intent of significantly lowering
emissions of these pollutants on a national scale. In addition to addressing emissions of the
traditional criteria pollutants such as SO2, NOx, and carbon monoxide (CO), Title HI of the
CAAAs also placed emission limits on a list of 189 distinct compounds and compound classes
defined in the Act as hazardous air pollutants (HAPs), and which are released by a wide variety
of sources.
While electric utility steam generating units were not made immediately subject to the
requirements of Title III, the CAAAs mandated that the Environmental Protection Agency (EPA)
conduct a study to quantify the risks to human health posed by utility emissions of HAPs, and
placed the responsibility on the EPA Administrator of determining the regulatory status of
utilities based on the findings of that study.1 This study is now being conducted by EPA's Office
of Air Quality Planning and Standards (OAQPS) with inputs from other government agencies
and industry groups. The CAAAs also required EPA to conduct a study on the health and
environmental effects of mercury from all sources, plus a study of the effects of atmospheric
deposition of all pollutants to the Great Lakes, Chesapeake Bay, Lake Champlain, and coastal
waters on human health and the environment. A further study by the Electric Power Research
Institute (EPRI) has been underway to quantify the emissions of a limited number of HAPs under
their Power Plant Integrated Systems: Chemical Emissions Study (PISCES)'through a review of
the existing literature and other data sources as well as a program of field testing at a number of
operating power generating sites ,2>3 The result of these legislative requirements and industry
efforts has been to significantly increase attention regarding HAP emissions from utility sources,
and to emphasize the shortage of quality data available to address these concerns.
Because HAPs from utility sources have not been subject to regulation in the past, the
quantity of information regarding HAP emissions from coal is limited, resulting in large
uncertainties for utilities considering the potential advantages and disadvantages of fuel
switching as an SO2 control strategy. Fuel switching is one SO2 reduction option which is being
considered by a number of utilities as a means of reducing emissions. In this strategy, coals with
a lower sulfur content (found in the U.S. primarily in the western part of the country) would be
used rather than higher sulfur coals common to the eastern U.S. The use of the low sulfur coals
would eliminate or reduce the requirement of Hue gas cleaning equipment such as wet scrubbers
in order to meet the lower emission limits. It has been suggested that an increased use of western
1
-------
coals because of fuel switching may lead to higher uncontrolled emissions of HAPs.4 With the
regulatory status of HAP emissions from utility sources yet to be determined, a thorough
evaluation of potential SO2 control strategies should also address the impacts of those
approaches on HAP emissions. For example, as part of the OAQPS study on HAP emissions
from'utilities, trace metal contents of high, medium, and low sulfur coals are being evaluated to
provide information concerning the effects of fuel switching on the metal HAP emissions.
Although evaluating the trace metal contents of different coals can provide valuable
information about the potential for metal HAP emissions, it is difficult to accurately predict
actual stack emissions of metal HAPs due the complex behavior of metals in combustion
systems. The assumption that a fixed percentage of a metal will be emitted from the stack based
only on data from a limited set of emissions tests may not be valid for the full operating range of
the unit, for other plants of similar design, or from different coals. Such an assumption in some
cases may provide order of magnitude accuracy; however, changes in furnace design, fuel, or
operating conditions can have a significant impact on the percentage of the as-fed coal which is
ultimately emitted from the stack of a particular unit. Therefore an examination of how toxic
metal emissions are related to the concentrations of those metals in coal under controlled
conditions is very important. In order to conduct such an evaluation, the Illinois Clean Coal
Institute (ICCI) entered into a cooperative research agreement with EPA's National Risk
Management Research Laboratory, Air Pollution Prevention and Control Division (formerly the
Air and Energy Engineering Research Laboratory) to evaluate the differences in HAP emissions
from the combustion of pulverized coals of different geographic origin in a single small-scale test
combustor. These tests were designed to focus on the emissions of metal HAPs, as coals from
different regions tend to show relatively large variations in their contents of such metals; while
variations in energy content and volatility may affect organic HAP emissions, it was felt that the
effect of changing coals would have a larger impact on the metals emissions. Therefore, a series
of combustion tests designed to provide information on the emissions of metal HAPs was
performed at AEERL, located at EPA's Environmental Research Center in Research Triangle
Park, NC.
II. Metal Emissions from the Combustion of Pulverized Coal
Coal naturally contains a variety of different metal compounds, and the combustion of
pulverized coal results in the release of those compounds from the coal into the combustion
environment. Unlike organic compounds, which can be chemically broken down into more basic
and nonhazardous constituents, metal compounds are not chemically transformed by the
combustion process to the degree that they lose their toxic nature. In this respect, organic
2
-------
compounds are analagous to NOx, which is primarily controlled via combustion process
modifications, while metal compounds find their analog in SO2, where the control focus shifts to
the fuel properties and post-combustion flue gas treatment. As with sulfur, there is no
breakdown of the metals during the combustion process. Therefore the total mass of metal
entering the furnace in the coal is the same as the mass of metal exiting the boiler, whether via
the stack or other routes.
A. Partitioning and Enrichment
When a metal compound is introduced into a combustion environment due'to its presence in
the fuel, it may exit the furnace or combustor via several pathways. The metal may exit the
combustor via the slag, bottom ash, or other furnace ash, or it may pass completely through the
combustor with the combustion gases. The major entrance and exit paths that metal compounds
may take through a pulverized coal boiler are shown in Figure 1. The metals that are retained in
the slag or bottom ash or are removed by the pollution control equipment may be of concern
from the perspective of solid or liquid waste disposal. However, investigation of solid or liquid
effluents containing toxic metal compounds is beyond the scope of this study; the emphasis here
is on the emissions of metal hazardous air pollutants.
The percentages of the input metal exiting through these various waste streams are referred to
as the partitioning behavior of the metal; in general, different metals exhibit different partitioning
behavior. Of specific concern here are metals which exit the combustor with the flue gases, since
these may result in emissions into the atmosphere if adequate pollution control strategies have
not been implemented. The metals which pass through the combustion process may take one or
more physical forms: a metal vapor; an aerosol* composed of small particles; in condensed form
on larger, solid-phase particles entrained in the flue gases; or as a combination of the three. In
many instances, one or more of the effluent streams will exhibit a concentration of a given metal
that is higher than would be expected if the metal were evenly distributed among all the various
streams. This behavior is known as enrichment of that particular stream by that metal.
The different exit paths that a metal may take are determined by complex physical and
chemical processes which occur during the combustion of the fuel and the flow of the flue gases
through the unit.5 Thermodynamic properties of the metal, combustion temperature, presence of
other compounds, flue gas velocities, and the physical means by which the metal is bound in the
coal are among the factors which can affect the ways a metal will be partitioned to the different
* An aerosol is defined as a distribution of particles suspended in a gas. Aerosols may be composed of a
combination of large (supermicron) and small (submicron) particles, or either of the two individually.
Aerosol particles can be either solid or liquid, or a combination of both.
3
-------
I
Scrubber
Liquids
Precipitator
Ash
Stack
Figure 1. Potential pathways of metal compounds through a utility boiler and
air pollution control systems.
combustor exit streams. A number of studies have been conducted to determine the partitioning
behavior of different metals,+ with the finding that some metals tend to partition toward the
submicron aerosol form.6"11
This "submicron enrichment" results in higher levels of the metal being present in the stack
as a submicron aerosol than would be expected if that metal were to exit the unit via all possible
waste streams in a percentage identical to its percentage of the input fuel. Because of this
enrichment process, and the less efficient removal efficiencies of particulate control equipment
for these smaller particles, for some metals the total mass of that metal in the form of a
+A comprehensive list of references on the subject of the partitioning behavior of different metals in
combustion systems can be found in Reference 5.
4
-------
submicron aerosol may be on the same order as the total mass of the' metal condensed on
particulate material of supermicron size.12 Table 1 shows the enrichment behavior of several
metals; arsenic, antimony, cadmium, chromium, lead, mercury nickel, and selenium arc seen to
be metals which tend to be enriched in the submicron particulate fraction, making them more
difficult to control.
Table 1. Submicron fly ash elemental enrichment in coal combustion investigations (adapted
¦ from Reference 5).
Source
Submicron enriched
No enrichment trend
Submicron depleted
Ref. 6
Sb, As, Cd, Cr, Pb, Ni,
Se, S, Tl, Zn
Al, Be, C, Fe, Mg, Mn,
Si, V
Bi, Ca, Co, Cu, K, Sn,
Ti
Ref, 7
Sb, As, Cu, Pb, Mo, Po,
Se, Zn
Al, Fe, Nb, Rb, Sr, Y
Ref. 8la)
As, Cu, Cr, Ga, Ge, Pb,
Mo, Ni, Se, Sn, V, Zn
Al, Ba, Ca, Ce, Fe, La,
Mn, Nb, K, Rb, Si, Sr,
Ti, Y, 7.r
Ref. 9(b)
Sb, As, Cd, Pb, Mo, Se,
W, Zn
Bate), Cr, Co, Ni, Mn,
Na, Sr, V
AlW), Ca, Ce, Hf, Fe,
Mg, K, Si, Ti
Ref. 10
As, Pb, K, Na, Zn
Al, Ca, Fe, Mg, Mn, Si,
Ti
Ref. 11
Ca, Cd, Cu, Pb, Sr, S, V
Al, Fe, Mg, Mn, Na, Si,
Ti, Zn
Ref. 12
Sb, As, Cd, Cr, Ni, Rb,
Se, V, Zn
Fe, Ti
Al, Hf, Mg, Mn, Ta
aSpecies As, Br, CI, I, Hg, Se in vapor phase, high filter penetration,
^Literature review.
cSliglit enrichment or no change.
llNo change or slight depletion.
The enrichment process is influenced by a number of factors. One of these is the physico-
chemical environment into which the metals are released during the combustion process. As the
temperature increases, a higher fraction of bound metals susceptible to submicron enrichment
will be released as vapor into the combustion gases. The actual fraction is highly dependent
upon the equilibrium vapor pressure of each metal, which in turn may vary due to the presence of
other compounds. Further, the composition of the gases, including the presence of particulates,
will strongly determine the final form the metal vapors take; for example, the lack of available
oxygen will limit the formation of metal oxides, resulting in higher levels of pure metal aerosol.
In addition, the presence of chlorine has been shown to have a strong influence on the
equilibrium behavior of metal vapors, which can result in significant differences in the size
distribution of the metal aerosols. ^ Because of these complex interacting influences, the
approach of estimating metal emissions only from the original trace metal content of the fuel is
highly simplified. This is because a number of variables may strongly affect the form each metal
takes as it exits the combustion zone in addition to affecting the size distribution of the metal
5
-------
aerosols and condensed metals on particulate matter. Thus, changes in the combustion
environment or changes in fuel properties (aside from trace metal content) can lead to significant
changes in the ability of pollution control equipment to remove metal compounds from the gas
stream.
B. Trace Metals in Coal
Metals exist in pulverized coal in three ways: as included mineral matter, which is present in
inorganic forms trapped as crystalline or glassy structures throughout the fuel particles; as
inherent mineral matter chemically bound as individual atoms within the coal organic matrix;
and as excluded mineral matter which is composed of particles distinct and separate from the
majority of fuel particles containing the combustible fraction, and which, originally having been
included mineral matter, are either released during the grinding process or have their origins in
the overburden added through the mining/transportation process. It has been found that the
inherent ash content is important in determining the likelihood of a coal to produce subrnicron
ash particles (particles which are less than 1 nm in diameter) during the combustion process.5
Of these, it is much more difficult to control metal aerosol and vapor emissions than metals
condensed on the particulate. This is because the metal aerosols are typically much smaller in
size than the particulate matter which act as nuclei for condensation of the metal vapors. The
metals in aerosol form are often less than 1 um in diameter (often referred to as subrnicron
particles), and most commercially installed pollution control equipment is less efficient in
removal of these small particles, with removal efficiencies in the range of 80-97%, versus greater
than 99% for larger particles (those over 1 Jim in diameter, or supermicron particles). A study of
five full-scale electrostatic precipitators (ESPs) found moderate to high particulate removal rates
for particles larger than a few micrometers or smaller than a few hundredths of a micrometer in
diameter, while minimum removal efficiencies were found to occur for particles of a few tenths
of a micrometers in diameter. ^ a more recent study presented removal efficiency curves for
veniuri scrubbers, high efficiency particulate air filters, and ESPs.14-15 For all three types of
equipment, the minimum collection efficiency occurred for subrnicron particles, with minima at
different sizes for the three device types. Therefore, although the subrnicron particle fraction
may not comprise the majority of the particulate entering the pollution control system, it may
account for the largest fraction of toxic metal emissions from the stack.
In summary, the emissions of metals from the combustion depend not only upon the content
of the metals in the coal, but also upon a variety of other factors. These include, but are not
limited to, how the metals are found in the coal, the presence of other elements (both functions of
the particular coal of interest), the preparation of the coal prior to its use as a fuel, the thermal
. 6
-------
conditions to which the metals are exposed, and the particulate control system efficiency over the
entire range of particle sizes (functions of the design and operation of the combustion and
pollution control systems). For these reasons it is important to measure actual emissions from
different coals under the same combustion environment to isolate the effects of changes in coal
properties on metal emissions.
III. Project Description
A. Facility and Equipment
EPA's Air Pollution Prevention and Control Division of the National Risk Management
Research Laboratory, located in Research Triangle Park, NC, has conducted a wide variety of
research on air pollution control technologies. As part of this research, a number of research
combustors have been constructed, some of which are capable of filing pulverized coal at
relatively low rates, but which are also capable of reasonably simulating combustion conditions
typical of full scale coal combustion systems. One of these units, the Innovative Furnace Reactor
(IFR), was utilized for the experiments reported here.
The IFR is a down-fired refractory-lined furnace rated at 29.3 kW (100.000 Btu/hr). and is
capable of firing natural gas, fuel oil, or pulverized coal. A schematic of the IFR is shown in
Figure 2. The furnace has a number of access ports along the vertical length of the unit which
can be used for sampling probes, injection of additional fuel or air for staged combustion tests, or
injection of sorbents for pollution control. A horizontal section at the bottom of the vertical
section (labelled as the Flue Gas Sampling Section in Figure 2) is used to sample flue gases
downstream of the combustion zone, and provides adequate length for isokinetic sampling. For
the current series of tests, a smaller diameter horizontal duct was installed in order to increase the
flow velocity and the total amount of flue gas collected. The inside diameter of the vertical
furnace section is 15.2 cm (6 in), with a 10.2 cm (4 in) diameter horizontal sampling duct. The
length of the vertical combustion section is 4.57 m (15 ft). The flue gas cleaning system for the
IFR consists of a wet packed-bed scrubber for removal of acid gases, and a pulse-jet baghouse
for particulate emissions control.
The burner utilized for combustion of pulverized coal is specially designed for use in
conjunction with the IFR. A schematic view of the burner is presented in Figure 3. The burner
is of a variable-swirl design, providing adjustment of the axial air to tangential air ratio as a
means of changing the flame shape. For this series of tests, a flame with high swirl was used.
The coal is fed into a primary hopper by a screw feeder, from which it is pneumatically fed into
7
-------
the burner through the central axial burner port. The axial tip was used in the coal/primary air
port during this series of tests.
Continuous emission monitors (CEMs) were used to provide continuous readings and records
of the levels of CO, NOx, SO2, oxygen (O2), and carbon dioxide (C02). These values were
used for evaluation of the combustion process as well as determination of pollutant levels. The
location of the monitor probes is shown in Figure 4.
B. Experimental Approach
Six different bituminous coals were tested during the test program. The coals were chosen
for specific reasons to evaluate the potential differences in emissions of toxic metal compounds.
Two Illinois coals (Illinois #5 and Illinois #6, referred to here as II and 12, respectively), two
western coals (referred to as W1 and W2), and two other coals (designated B1 and B2). Coals
B1 and B2 were used in other air toxics test programs; B1 was used in a small scale test at
Battel le that was designed to mimic load changes in a full scale unit, and B2 was used in actual
8
-------
IFR Exhaust Duct
Figure 4. Schematic of the IFR exhaust duet showing sampling locations. Shown are the
locations of the multi-metals train (MMT) sample probe and the Method 5 (M5) probe.
The duct diameter is not drawn to the same scale as the length.
9
-------
full scale testing of air toxics emissions by Babcoek & Wilcox. The results from these two tests
have not yet been published.
Coals II and 12 were higher in sulfur than W1 and W2, and represent the "high sulfur" cases,
with coals W1 and W2 representing the "low sulfur" eases. B1 and B2 were used to provide a
comparison to other tests that used the same coals as a measure of compatibility of test results.
Tables 2 and 3 present the proximate and ultimate analyses of these six coals, respectively, and
the concentrations of trace metals are shown in Table 4 and plotted in Figure 5. The
concentration of manganese in coal B2 was the highest of any of the trace elements.
Concentrations of antimony and cadmium were at the detection limit of the analysis methods for
all coals, except for the cadmium concentration of coal B2. In general, the metal concentrations
of coals II and 12 were greater than for coals W1 and W2, with the exception of manganese,
which was higher in coal W1 than for either II or 12. Trace metal concentrations for coal B2
were generally the highest of all the coals tested, except for arsenic, beryllium, and mercury, in
which cases coal B1 had the maximum concentrations. In the case of mercury, B2 showed the
lowest concentration of the six coals.
Table 2. As-received proximate analysis of the six coals used in the combustion tests.
11
12
Wl
W2
B1
B2
Moisture
4.01%
6.48%
20.22%
18.50%
1.90%
2.08%
Volatile Matter
33.64%
33.65%
35.34%
35.22%
32.97%
37.02%
Fixed Carbon
53.44%
51.46%
36.87%
39.72%
52.62%
49.24%
Ash
8.91%
8.41%
7.57%
6.56%
12.51%
11.66%
Heat Content,
Btu/lb
12831
,12128
9033
9623
12501
12483
Table 3. Dry proximate and ultimate analyses of the six coals used in the combustion tests.
11
12
Wl
W2
B1
B2
Volatile Matter
35.05%
35.98%
44.30%
43.22%
33.61%
37.81%
Fixed Carbon
55.67%
55.03%
46.21%
48.73%
53.64%
50.28%
Ash
9.28%
8.99%
9.49%
8.05%
12.75%
11.91%
Heat Content,
Btu/lb
13367
12968
11322
11807
12743
12748
Carbon
74.53%
71.79%
66.04%
69.53%
71.42%
70.74%
Hydrogen
5.01%
4.97%
4.77%
4.70%
4.84%
4.95%
Nitrogen
1.67%
1.51%
1.18%
1.03%
1.44%
1.29%
Sulfur
1.51%
2.95%
0.47%
0.67%
2.43%
3.79%
Oxygen (a)
8.00%
9.79%
18.05%
16.02%
7.30%
7.32%
(a) Oxygen content is calculated by difference.
10
-------
The concentrations of cadmium, chromium, lead, manganese, and nickel in coal B2 were
much higher than expected, and there is some concern that these values may not accurately
reflect the levels of metal that were actually in the coal. Unfortunately, the remainder of the coal
sample was combusted prior to the receipt of the final results, making additional analysis of the
coal sample impossible. The potential for the as-received concentrations to be higher than the
actual case should be considered when evaluating the results in the following section.
Each of the coals was burned in the IFR under conditions chosen to most closely simulate the
combustion of pulverized coal in a typical utility boiler. A stoichiometric ratio of 1.25 (i.e. 25%
excess air) was chosen to provide a stable flame and relatively low CO and NOx values at the
Table 4. Trace metal concentrations for the four coals used in the combustion tests. All
concentrations are given in (ig/g (10"6 lb/lb), on a dry basis.
11
12
W1
W2
B1
B2
Antimony
1
< 1
< 1
< 1
< 1
< 1
Arsenic
6
4
1
2
15
4
Beryllium
0.9
1.3
0.2
0.2
2
0.8
Cadmium
<0.2
<0.2
<0.2
<0.2
0.2
4
Chromium
21
20
17
5
18
47
Lead
21
36
3
3
14
50
Manganese
12
19
9
21
30
167
Mercury
0.10
0.15
0.09
0.10
0.33
0.08
Nickel
53
18
42
3
15
70
Selenium
2
1
< 1
1
3
3
i i i I I r i
Antimony Arsenic Beryllium Cadmium Chromium Lead Manganese Mercury Nickel Selenium
Metals
Figure 5. Concentrations of trace metals in as-fed coal for the six coals tested.
11
-------
test firing rate of approximately 13.2 kW (45,000 Btu/hr). While the actual time-temperature
history of the IFR is not the same as a full-scale unit, the combustion zone temperatures are quite
similar, providing conditions for metal release and formation of metal aerosols similar to full-
scale.
C. Sampling and Analysis
The metals emissions were collected using the Multiple Metals sampling train (MMT),1C| and
mercury emissions using the Method 101A sampling train.1*7 Sampling ports were installed in the
horizontal section at locations specifically determined for the conditions expected for this series
of experiments. Emissions samples were collected simultaneously at two different axial
locations. These locations were located according to isokinetic sampling requirements, given the
duct size and flue gas velocity for these tests. This ability allowed the particulate to be collected
without excessive fluid influence on the particle stream from the duct walls, entrance, or exit.
Sampling locations are shown in Figure 4. Note that all samples were collected upstream of any
pollution control equipment. This was done to provide data that would be applicable to all
systems firing pulverized coal; collection efficiencies of various pollution control equipment can
then be used to estimate emissions from a wide range of full scale systems.
The collected samples were then analyzed in accordance with the respective EPA analytical
methods.16-17 The MMT does provide for simultaneous sampling of all the metals listed in
Table 3; however, this procedure must be done with care when both mercury and manganese
measurements are required. Individually, mercury can be collected using Method 101A or by the
MMT. In both methods, mercury is captured using a permanganate solution. Even though the
mercury capture and analysis procedures are identical in both trains, the permanganate solution
can alter measured manganese concentrations through glassware contamination. However, it
was desirable to use only a single sampling train, and the first series of tests included mercury
sampling and analysis using the MMT. This initial series included all test runs for II, 12, and
W1 coals, and one am using W2. The remaining tests were conducted as the additional coals
were received, and in these tests, mercury was collected using a separate train (i.e., a Method
101A train).
D. Quality Assurance and Quality Control .
Several steps were taken to maintain data quality. The project was conducted following an
EPA-approved Quality Assurance (QA) Project Plan, using AEERL QA Level IV. QA/QC
requirements apply to this project. Data are supported by QA/QC documentation as required by
the U.S. EPA's QA policy. Although Level IV is the least stringent QA level for AEERL
experimental projects, the measurement, sampling, and analytical work incorporated several
12
-------
procedures designed to ensure data quality at a higher QA level. CEMs were calibrated before
and after each test run to check zero and span. Samples were collected in complete accordance
with EPA method guidelines. These methods include measures for leak checks, isokinetic
sample collection, and sample custody records. Duplicate samples were collected for each
sampling and analytical method, and field and method blanks were also used to evaluate
potential contamination during sampling and analysis. Since the data were not intended to be
used for regulatory purposes or enforcement, more stringent QA procedures were not deemed to
be necessary. However, the sampling and analysis procedures followed during the testing were
identical to those that are required under more stringent QA plans.
With the exception of W2 test number 2, all samples were collected within 85% of isokinetic
conditions. The major problem affecting data quality was the contamination of one sample with
permanganate solution during the sampling process. Extremely high levels of manganese were
found in this sample (W2 test number 1) resulting from permanganate contamination, and this
value was not included in the data used to calculate emissions. Analysis of the field/method
blank yielded detectable concentrations for antimony, cadmium, and manganese, with levels of
all other metals (including mercury) below the method detection limits. In the cases of antimony
and cadmium, the quantities detected in the blank were at least one order of magnitude less than
the lowest value measured in the samples. The amount of manganese in the blank was, however,
of the same order of magnitude as. found in the test samples, although all test samples were
higher than the blank. The conclusion drawn from these results was that the analyses of all
metals except manganese correctly indicated the levels of of metal compounds sampled from the
flue gases. For manganese, the reported levels are (at worst) higher than actually present in the
gases, but are of the same order of magnitude.
IV. Results
Rather than focusing only on the measured emissions data, it is more useful to compare the
measured results to the maximum possible emissions. It is a relatively straightforward
calculation to estimate the maximum possible emissions of metals exiting the stack by assuming
that none of the metals' mass remains in the com bus tor as slag or bottom ash, or is collected in
the flue, gas cleaning systems. This assumption is extremely simplified, and does not accurately
reflect the true behavior of metals in either small scale or full scale coal combustion systems.
However, such calculations do provide a benchmark against which measured metal emissions
can be compared. Because the primary goal of the study was to evaluate the differences in toxic
metal emissions between different coals, it is necessary to consider the basis on which such
comparisons are made. While a direct comparison of flue gas concentrations is of some interest,
13
-------
it is in this case much more informative to evaluate emissions based ori equivalent energy inputs
to the combustor. Using an energy-specific basis for comparison accounts for differences in
energy, ash, and moisture contents in the different coals.
On a per unit energy basis, the maximum theoretical metal emissions can be determined by
merely dividing the trace element concentration by the energy content of the coal. This provides
a maximum emission factor in grams per megajoule (or pounds per trillion [1012] Btu). The
measured emissions factors per unit energy were determined by measuring the total mass of
metal in each collected sample and dividing by the total amount of energy input from that
sample's coal. It is expected that the measured values will be lower than the theoretical
maximum values for several reasons. The primary reason is that a portion of each metal will exit
the combustor via pathways other than out the stack as discussed earlier. A second reason is that
the measurements of metals in the flue gas stream are taken over a considerable period of time,
effectively integrating out short-term fluctuations in the concentration of a particular compound.
The measurements of trace metals in the coal, however, are based on discrete samples of the
coal. Because.of the heterogeneous nature of coal, the possibility of significant fluctuations in
the trace element concentrations is likely, leading to corresponding fluctuations, in the calculated
emissions (this could also lead to values of the measured emissions being higher, rather than
lower, than the calculated emissions). Finally, the concentrations of these metals in both the coal
and in the flue gases are typically very low. In some cases, the concentration of a particular
metal in either the coal or the flue gas may be below the method detection level. In these
instances, the only assumptions that can reasonably be made are that the actual concentration is
either at the detection level (the maximum possible concentration) or at zero concentration (the
minimum possible concentration). Assuming either of these values introduces further, and often
significant, uncertainty into the calculations.
A. Calculated Emissions
The theoretical maximum calculated emission factors (hereafter referred to as the calculated
emission factors) foreach of the metals analyzed are presented in Table 5 for the six coals tested,
in pounds per trillion (1012) Btu, and are shown graphically in Figure 6. In general, the
calculated emission factors follow the trends shown in Figure 5, as would be expected. As noted
above, this calculation assumes 100% of the metal entering via the coal will exit the combustor,
either in condensed form on the fly ash or in the vapor phase. The highest calculated emissions
were for nickel, lead, and chromium, with a maximum value of 1.34xl04 lb/1012 Btu for
manganese from coal B2. The lowest calculated emissions were for mercury and cadmium, with
the minimum emission being 6.41 lb/1012 Btu for mercury, also from coal B2. These results are
similar to the distribution of metals concentrations in the coals; however, some differences are
14
-------
noted due to the difference in heating values between the coals, which in general increase the
emission factors of the western coals W1 and W2 relative to the Illinois coals.
The calculated emissions' are based upon the as-received concentrations in the coals, and as
noted above, the values reported for the trace concentrations in coal B2 seemed high relative to
the rest of the coals. This results in high values for the calculated emissions, and alters the
Table 5. Maximum calculated trace metal emissions in lb/1012 Btu. Values are
calculated based on as-received metal and Btu contents of the coal. The
calculated values assume 100% of the metal in the coal will pass through the
combustor to the stack.
11
12
W1
W2 •
B1
B2
Antimony
75
77
88
85
78
78
Arsenic
449
308
88
169
1177
314
Beryllium
67
100
18
17
157
63
Cadmium
15
15
18
17
16
314
Chromium
1571
1542
1501
423
1413
3687
Lead
1571
2776
265
254
1099
3922
Manganese
898
1465
795
1779
2354
13100
Mercury
7
12
8
8
26
6
Nickel
3965
1388
3709
254
1177
5491
Selenium
150
77
88
85
235
235
ioooqo
Antimony Arsenic Beryllium Cadmium Chromium Lead Manganese Mercury Nickel Selenium
Metals
Figure 6. Maximum calculated emission factors for the six coals tested.
15
-------
relative comparisons between calculated and measured emissions. This must be kept in mind
when evaluating the comparative results presented below.
B. Measured Emissions
The measured uncontrolled emission factors (hereafter referred to as the measured emission
factors) of each of the metals of interest are tabulated in Table 6 for the six coals tested, in
pounds per trillion (1012) Btu, and are presented graphically in Figure 7. In contrast to the
similarity between the as-fed trace element concentrations and the calculated emisson factors,
Figure 7 shows significant differences in the emission factors compared to the as-fed
concentrations. While the metals that were found in the highest concentrations in the coals tend
also to result in the highest measured emission factors, there are considerable differences in the
measured emisson factor results and the as-found concentrations.
As was the case for the calculated emissions, the highest measured emissions were for nickel,
lead, and chromium. The maximum measured emission was 2740 lb/1012 Btu for nickel from
Wl. In some cases, the measured value exceeds the maximum calculated emission factor.
However, as mentioned above, this occurrence is not uncommon; this problem will be discussed
further in the following section. The lowest measured emissions were from mercury and
cadmium, with the minimum being 0.14 lb/1012 Btu for cadmium from coal B2. It is interesting
to note that, based on the concentration of metal found in the coal, it would be expected that coal
B2 would have the highest cadmium emission factor, rather than the lowest. In several other
instances as well, the calculated emission factors based on the trace metal concentrations in the
coals seem to have a very weak relationship to the measured emission factors, highlighting the
differences in the coals' trace metal contents and the metals emissions resulting from the
combustion of those coals.
Table 6. Measured trace metal emissions in lb/1012 Btu.
11
12
Wl
W2
B1
B2
Antimony
19.7
35.7
29.2
23.5
20.1
6.72
Arsenic
419
329
• 126
229
666
275
Beryllium
35.2
68.9
15.7
' 25.8
31.4
11.8
Cadmium
3.96
13.9
3.77
5.28
2.58
1.36
Chromium
545
957
1150
743
254
199
Lead
707
1310
152
132
192
107
Manganese
408
685
655
712
394
216
Mercury
3.20
7.02
3.62
5.78
6.15
1.55
Nickel
1210
1650
2740
1378
286
167
Selenium
51.9
65.2
22.5
25.2"
92.9
52.6
16
-------
Antimony Arsenic Beryllium. Cadmium Chromium Lead Manganese Mercury Nickel Selenium
Metals
Figure 7. Measured emission factors for -the six coals tested.
C. Comparison of Measured arid Calculated Emissions
The ratio of measured to calculated emissions of metal HAPs can provide a considerable
amount of information. These values are presented in Table 7 below, expressed as the percent of
as-fed maximum calculated emissions. In some cases, the measured emissions may exceed the
calculated emissions (which are based on 100% of the metal in the coal exiting the combustor).
This can be due to one or more reasons: the measured value of the trace metal content of the coal
is lower than is actually the case; the measured value of the concentration of the metal in the flue
gas is higher than actual; or both measurements may be correct, but the discrete sample of coal
used to determine the trace metal content of the coal had a lower than typical amount of the
particular metal. Because of the low concentrations of trace metals, the accuracy of the
measurements is also an area that can add to the uncertainty of the final results. However, these
sources of error and uncertainty do not remove the usefulness of this ratio. By focusing on
consistent results and trends, much can be determined from these values. First, if the ratio of
theoretical to measured emissions of a given metal is similar for a wide range of coals, then it is
likely that the partitioning behavior of the metal is similar for the different coals tested, and may ¦
therefore be more dependent on the combustion and furnace characteristics than on the coal type
or composition.
17
-------
Table 7.
Ratio of measured trace metal emissions to calculated emissions, in percent.
11
12
W1
W2
B1
B2
Antimony
26
46
35
27
26
9
Arsenic
93
107
74
259
57
88
Beryllium
52
69
93
146
¦ 20
19
Cadmium
26
90
22
30
16
0
Chromium
35
62
272
49
18
5
Lead
45
47
60
50
17
3
Manganese
45
47
37
884
17
2
Mercury
43
61
43
73
24
• 25
Nickel
30
119
1079
37
24
3
Selenium
35
85
27
29
39
22
The cases where the measured emissions are below the calculated emissions are more
frequent because of the fact that some percentage of each metal typically exits the combustor via
the bottom ash and is therefore not measured in the stack sampling. This is not to say that there
are no measurement errors in the instances where the measured emissions are lower than the
calculated emissions; however, such errors are more difficult to detect and require repeated
measurements to determine average values and identify outlying points.
As noted above, there is some concern that the as-received concentrations for coal B2
showed levels of cadmium, chromium, lead, manganese, and nickel higher than those actually in
the coal. This possibility is strengthened by the very low percentages of those metals that are
measured in the stack relative to the rest of the coals. While there is no quantitative evidence
from the laboratory data that the as-received trace metal analyses of coal B2 were faulty, the
comparisons of these values with the other cases indicate that these values may not be as reliable
as the remaining data.
V. Discussion
A. Scaled Results
To more clearly see the differences between the calculated emissions and the measured
emissions and how they change for the different coals and the different metals as the coals are
combusted, it is useful to plot both in the same figure. In order to clearly view the information, it
is also helpful to look at the results for each metal separately. Because the absolute values of the
emissions (both calculated and measured) can sometime differ by one or more orders of
magnitude, it is helpful to normalize the results so that they both have the same maximum value.
The scaling is done by dividing the value for each coal by the maximum for that particular series
of calculations or measurements. For example. Table 5 shows that the maximum calculated
18
-------
emission of arsenic is for coal Bl, at 1177 lb/1012 Btu. Each of the values for calculated arsenic
emissions is then divided by 1177 lb/1012 Btu to obtain scaled values between 0 and 1.
Likewise, each measured arsenic emission value in Table 6 is scaled by the value for coal Bl,
666 lb/1012 Btu. For each metal, then, plots are produced that show the distribution of relative
emissions (both measured and calculated) for each of the six coals tested. Figures 8 through 17
plot the scaled as-received distributions and the scaled measured emission factors for each metal
tested.
Figure 8 provides relatively little information concerning changes in the calculated and
measured antimony emissions, since the calculated emissions are different due only to the
differences in heat content of the coals. In each case, the concentration of antimony in the coal
was below the method detection level, making it impossible to accurately compare the calculated
emissions which are based on the as-received trace compound levels; however, the figure is
included for completeness.
Figure 9 compares the scaled calculated and measured emissions for arsenic for the six
different coals tested. While there is some difference in the heights of the columns in the two
plots, the same general shape is obvious for both. The only substantial difference is that coals
Sb, Calculated emissions Sb, Measured emissions
Figure 8. Comparison of scaled calculated and measured emissions of antimony for the six coals
tested. Results are scaled to the maximum value in each category.
19
-------
W2 and W1 have traded rankings, with W2 showing the lowest calculated emissions and Wl the
second lowest, but Wl having the lowest measured emissions and W2 the second lowest. Coal
B1 has an as-received concentration approximately twice as great as the next highest coal (II),
and similarly, emissions of roughly twice those of II. The information from Figure 9 indicates
that the relative levels of emissions from burning the six different coals are roughly the same as
the relative levels of arsenic found in the as-received coals.
As, Calculated emissions
As, Measured emissions
II
12
Wl
W2
B1
B2
Figure 9. Comparison of scaled calculated and measured emissions of arsenic for the six coals
tested. Results are scaled to the maximum value in each category.
The situation changes for beryllium, as shown in Figure 10. The left side of Figure 10 shows
that the as-received concentration of beryllium is greatest in coal Bl. with the second greatest
concentration in coal 12 at roughly 65% that of B1. However, the measured emissions show that
12 has the greatest level of beryllium emissions, with II at about 50% of 12 and Bl at less than
50%. The relative levels of emissions from the other coals also changed, making it impossible to
accurately predict even the relative levels of beryllium emissions based on the amounts of
beryllium in the coals.
20
-------
Be, Calculated emissions Be, Measured emissions
Figure 10. Comparison of scaled calculated and measured emissions of beryllium for the six
coals tested. Results are scaled to the maximum value in each category.
Because the calculated emission levels of cadmium were so low (at the. detection limit of the
test method), the comparison of calculated versus measured emissions shown in Figure 11 does
not provide a significant amount of information. Again the figure is included for completeness.
As was the case in Figure 10, the relative levels of calculated emissions of chromium versus
measured emissions show a marked change, as can be seen in Figure 12. Here coal B2 shows the
highest level of calculated emissions, while W1 actually measured the highest, with B2's
measured emissions being less than 20% of Wl. Again, the conclusion to be drawn from this
comparison is that the relative emissions of chromium cannot be accurately predicted from the
relative amounts of trace chromium in the coal, based on the behavior exhibited by these results.
Similar results are seen in Figures 13 through 15, which compare lead, manganese, and
nickel, respectively. In each ease, the relative emission levels change significantly when
comparing calculated emissions based on the as-fed trace metal concentration to the measured
emissions of the same metal. The change in relative lead emissions is not as great as for
manganese and nickel, but in each of these cases the use of the as-found trace metal
concentration as a predictor of the relative emissions is seen to give significantly different results
that what might be expected. In the cases of lead, manganese, and nickel, the high calculated
21
-------
II
12
W1
W2
B1
B2
Cd, Calculated emissions
Cd, Measured emissions
Figure 11. Comparison of scaled calculated and measured emissions of cadmium for the six
coals tested. Results are scaled to the maximum value in each category.
II
U 12
wi
W2
B1
B2
Cr, Calculated emissions
Cr, Measured emissions
Figure 12. Comparison of scaled calculated and measured emissions of chromium for the six
coals tested. Results are scaled to the maximum value in each category.
22
-------
Pb, Calculated emissions Pb, Measured emissions
Figure 13. Comparison of scaled calculated and measured emissions of lead for the six coals
tested. Results are scaled to the maximum value in each category.
Mn, Calculated emissions Mn, Measured emissions
Figure 14. Comparison of scaled calculated and measured emissions of manganese for the six
coals tested. Results are scaled to the maximum value in each category.
23
-------
Ni, Calculated emissions Ni, Measured emissions
Figure 15, "Comparison of scaled calculated and measured emissions of nickel for the six coals
tested. Results are scaled to the maximum value in each category.
emissions from coal B2 shown in Figures 13 through 15 shift the remaining values strongly
lower. It is very possible that the as-received coal analyses do not accurately reflect the true
levels of these compounds, and that the reported values are high. If this is the case, the relative
metal emission levels of the remaining five coals should be compared.
In Figure 16, relative calculated and measured emissions are plotted for selenium. Here the
differences between calculated and measured are much less dramatic, and while there are some
changes between the two plots, the same general shape is maintained, with B1 exhibiting the
greatest relative emissions for both the calculated and measured values, coals W1 and W2 being
the lowest, and II and 12 lying between the western coals and the "B" coals.
Figure 17, which plots calculated and measured relative emissions for mercury, also exhibits
less dramatic differences than the earlier plots. There is a significant change in that the measured
emissions of 12 were the highest, while B1 had the highest calculated emissions. However, there
is little change in the qualitative shape of the two plots, in contrast to the marked change shown
by nickel (Figure 15).
24
-------
Se, Calculated emissions
Se, Measured emissions
Figure 16. Comparison of scaled calculated' and measured emissions of selenium for the six
coals tested. Results are scaled to the maximum value in each category.
Hg, Calculated emissions
Hg, Measured emissions
II
12
W1
W2
B1
B2
Figure 17. Comparison of scaled calculated and measured emissions of mercury for the six coals
tested. Results are scaled to the maximum value in each category.
25
-------
From the above comparisons, a significant conclusion can be drawn. In this series of tests,
the distributions of emissions change most markedly for beryllium, chromium, manganese, and
nickel; least markedly for arsenic, selenium, and mercury; and are inconclusive for antimony,
cadmium, and lead. These results show that emissions of the least refractory metals (arsenic,
selenium, and mercury) tend to be more strongly affected by the levels of those metals in the
coals, while the more refractory metals (beryllium, chromium, manganese, and nickel) do not
show such a strong correlation. This behavior is to be expected, since the refractory metals tend
to be enriched in the larger particles, which are typically collected as bottom ash or at other
points prior to the combustor exit. The most volatile metals tend to be enriched in the submicron
particle fraction, and are therefore much more likely to pass through the combustor with the flue
gases.
The dominating mechanisms governing this behavior are the vapor pressures of the
individual metals. For the refractory metals, such as nickel, the vapor pressure is so low that no
additional nickel can vaporize from the coal or ash particles. No matter how much additional
nickel may be in the coal, it will not be released into the flue gases to form particles small
enough to pass through the combustor and be collected, but rather remains in the larger particles
and falls out of the flue gas flow stream. On the other hand, the metals with high vapor
pressures, such as mercury, will easily vaporize and form either vapors or particles small enough
to pass through to the combustor exit. For mercury, selenium, and arsenic, the flue gases and the
coal- or ash-bound metal particles have not yet reached a state of equilibrium; therefore, higher
levels of these metals fed into the combustor will easily vaporize and allow collection at the
combustor exit For this reason, a doubling of a volatile metal will result in roughly a doubling
of the emissions. Conversely, for the refractory metals, a doubling of a coal's metal content will
not result in a doubling of the uncontrolled emissions, since the dominant mechanism governing
the release of these metals is no longer the vapor pressure. Because the solid and vapor phases of
the refractory metals are essentially in equilibrium, other mechanisms will govern the release of
the metals into the flue gases and the subsequent aerosol formation. Therefore, factors such as
the way the metal is bound in the coal, the presence of other compounds such as chlorine, or
local combustion conditions must be taken into account when attempting to predict the
uncontrolled metals emissions.
In summary, the tests show that, for volatile metals, changes in the amounts of those metals
from one coal to another can be a good predictor as to the changes in the emissions of those
metals from the combustor. For the other metals, however, predictions of emission changes due
to changes in the metal contents of the coal are much more likely to be in error. One may be able
-------
to predict the direction of change for these metals, but attempting to predict the magnitude of that
change is not likely to be accurate unless factors other than the trace metal contents of the
comparative coals are also considered.
As mentioned above, some of the other factors that can play a significant role are the way the
metals are bound in the coal, the presence of other elements in the coal, the preparation of the
coal prior to feeding into the furnace, the combustion conditions, and the geometry of the
furnace. As these tests have shown, however, even maintaining the same combustion conditions
in the same furnace does not ensure that the emissions of refractory metals from different coals
will change in the same manner as the as-fed' trace concentrations of those metals in the coals.
Since the coals were not specially prepared prior to burning them, the differences noted here are
likely to be due to differences in the ways the metals were bound in the coal, or to the presence
of other elements in the coals that affected the metal aerosol formation process during
combustion. Unfortunately, studies of these effects were beyond the scope of this project;
therefore, reasons for the differences in emissions cannot be explained definitively.
B. Comparison of Small Scale Results to Full Scale Tests
With the recent emphasis on HAP emissions from utility boilers, several significant field
tests of HAP emissions have been undertaken by EPA, DOE, and EPRI. DOE recently reported
results from coal-fired boilers18 (shown in Table 8), and it is useful to compare the results
obtained in the small scale tests reported here to the field data reported by DOE. Figure 18
shows the average emission factors for the six coals tested in this study along with the maximum
values reported by DOE. In this figure, the small scale studies yielded the highest emission
factors for all metals tested except for cadmium, selenium, and mercury. For these three metals,
the DOE field tests had higher emissions on a per energy input
Table 8. Full scale emission factors for hazardous metal air pollutants, from DOE test
¦ program.18 Emission factors are in lb/1012 Btu.
Metal
Low value
High value
Antimony
<0.1
2.4
Arsenic
0.1
42
Beryllium
<0.1
1.4
Cadmium
<0.1
3.0
Chromium
< 0.1
51
Lead
0.6
29
Manganese
1.1
22
Nickel
0.3
40
Selenium
<0.1
130
Mercury
0.5
14
27
-------
Antimony Arsenic Beryllium Cadmium Chromium Lead Manganese Mercury Nickel Selenium
Figure 18, Comparison of measured emissions from small scale test program to upper range of
DOE field test data,
basis. Again, it is worth noting that these three metals are very volatile, and the higher full scale
emissions are likely to be due to higher levels of these metals in the coals.
While the remaining small scale results are veiy similar to the maximum reported values
from the DOE study, it must also be emphasized that the small scale results are for uncontrolled
emissions. The minimum and maximum reported emission factors from the full scale tests differ
by as much as four orders of magnitude, with the differences likely due not only to differences in
the metal contents of the coals, but also to the efficiency of pollution control equipment present
at the different sites. Unfortunately, the DOE data are not in sufficient detail to allow a more
definitive evaluation of the effect of controls; however, they do show that pollution control
equipment can remove as much as 99% of some metals that exit the boiler. If the effects of
pollution control equipment are accounted for, the emissions reported for the small scale tests
would likely drop from 30 to 99%, depending upon the metal.
As a further measure of the correlation between emissions from the small scale combustor
used in this study and emissions from a full scale unit, a set of small scale test runs was
conducted using a coal that was also being used in full scale air toxics tests. The B2 coal was
taken from a coal pile used during a full scale demonstration test of an innovative pollution
control system for coal fired utility boilers. Therefore, the sampled emissions from the small
28
-------
scale testing can be compared to the full scale results using this coal to quantify the similarities
and differences between the full scale and small scale tests. The full scale tests were done at a
160 MWe wall fired boiler of a pre-NSPS (New Source Performance Standards) design. Coal B2
is a blend of bituminous Ohio coals with sulfur content of approximately 3.8% (see Table 3).
Figure 19 shows the emission factors for the 10 metals tested in this study compared to the
emission factors for the same 10 metals measured upstream of any pollution control equipment
in the full scale tests. In general, the results correlate well, with the major differences being for
chromium and manganese. In each case, the full scale emissions were higher by approximately
50%. The uncontrolled mercury emissions from the full scale system was also significantly
higher at about 9 versus 1.55 lb/1012 Btu measured in the small scale tests. Uncontrolled
selenium emissions were also higher in the full scale tests than in the small scale tests, by over
80%. For both mercury and selenium, the levels of input of both metals were found to be very
similar between the two tests. However, the trace element concentrations of chromium and
manganese seemed to be significantly higher for the small scale tests samples than for the full
scale tests. These results were also much higher than the remaining coals, indicating that the
analyses of the trace element concentrations were higher than actually in coal B2.
350.00-,
300.00-
250.00-
3 -
PS
2 200.00-
' o
7: 150.00-
2 :
a :
^ 100.00-
c
o
' 'I
• 1 50.00-
W ;
0,00-
Anumor.y Arsenic Beryllium Cadmium Chromium Lead Manganese Mercury Nickel Selenium
Figure 19. Comparison of uncontrolled emissions from a full scale pulverized coal utility
boiler to those from the small scale IFR. The same coal was used in both cases.
29
-------
It is reasonable that the uncontrolled small scale emissions will be lower than uncontrolled
emissions from full scale units, for several main reasons. First, the gas velocities in the pilot
scale sampling duct were not as high as those found in full scale units, making it more difficult
for particles to remain entrained in the flue gases. Second, the ratio of gas volume to wall
surface area is much smaller in the pilot scale unit, leading to increased condensation of the
metals in the flue gas at the small scale. This increased cool wall area in the pilot scale is likely
to have the effect of minimizing the volatilization of the metals, followed by more rapid
condensation. This may lead to increased particle sizes and an accompanying increase in the
amount of metal-containing particulate that falls out of the flue gas stream. The third main
reason that the emissions from the pilot scale are likely to be less is that the smaller volume gas
flow are the tighter turns that would be seen in the pilot scale, leading to increased amounts of
particulate being deposited on the walls and removed from the flue gases. It was expected that
the more volatile metals (mercury and selenium) would probably not exhibit this effect nearly as
strongly as the refractory metals, and would be closer to the actual uncontrolled emissions seen
in full scale units. However, the volatile metals seemed to show the largest differences between
small and full scale tests.
The above differences in flows emphasize the fact that pilot scale results cannot be directly
translated to full scale. The value of the pilot scale studies lies in their ability to more closely
maintain consistency between test runs, allowing for a more reliable comparison between the
emissions from different coals. The comparisons should be. viewed as trends that are likely to
occur at full scale. For instance, the conclusion that the emissions of refractory metals are likely
to correlate less closely with their content in the as-fed coal than are the emissions of more
volatile metals should hold regardless of scale.
C. Effects of Air Pollution Control Systems
The emissions reported here are uncontrolled. In comparing these values to those from full
scale units, or in estimating stack emissions from the metal content in the coals, the effects of
pollution control systems must be taken into account. This is not a simple task, however, since
the different forms of control technologies will have different effects on the metals in question.
The more volatile metals such as mercury and selenium will not be nearly as strongly affected by
the presence of flue gas cleaning equipment as will be the more refractory metals such as
chromium and nickel. Likewise, a wet flue gas desulfurization (FGD) scrubber will have less
effect on emissions of the refractory metals than will an efficient particulate removal system such
as an ESP. In addition, the treatment of flue gases by more than one type of pollution control
system will have synergistic effects that can alter the removal efficiencies of a single piece of
equipment. For instance, the removal of sulfur trioxide (SO3) upstream of an ESP may have a
-------
significant effect on the ESP's particulate removal performance, and therefore on the removal
efficiency of metal-laden particles.
Currently, a significant amount of information is being generated on the capabilities of air
pollution control equipment for removing metal HAPs from coal combuslion flue gas streams.
EPA's Electric Utility Study Report to Congress19 and EPRI's PISCES field test program2'? are
two primary efforts to quantify the emissions of HAPs from coal combustion, and the PISCES
program in particular is evaluating the effects of pollution control systems on HAP emissions.
Other data are also available to indicate the capabilities of pollution control systems to reduce the
emissions of metal HAPs. EPA released a report in 1989 that estimated air toxic emissions from
coal-fired utilities, and provided some data on the effects of pollution control systems on those
emissions,20 The report's estimates of pollution control device metal HAP removal efficiencies
are shown in Table 9 for seven metal compounds (although copper is included in this table, it is
not on the list of HAPs under Title III). Note that the testing data on which these estimates are
based are very limited, and therefore do not adequately characterize the full range of coals,
combustion equipment, or pollution control devices that are in place throughout the utility
industry. Nevertheless, these values do show that significant levels of control can be attained for
some metals with existing control technologies.
As noted earlier in Section II, particulate control is strongly dependent upon particle size, and
also varies significantly according to the type of control technology applied. Venturi-type wet
scrubbers usually remove at or below 50% of these smaller particles, as well as the metals that
tend to be enriched in these smaller particle sizes. Packed-bed scrubbers may show metal
removal levels of over 90% (see Table 9).20>21 However, utility scrubbers are usually designed
for acid gas removal rather than for particulates, and will show a much greater ability to remove
Table 9. Removal efficiencies of air pollution control devices for seven metals. Values are
given as percent removal measured between device inlet and device outlet. Data
are from Reference 20.
Control
Technology
Arsenic
(a)
Beryllium
Cadmium
Chromium
Copper
(b)
Manganese
Nickel
Wet Scrubber
6-97
94,3
94.4
91.8
91.4
89.1
96.4
ESP
87.5
91.9
74.6
71.5
85.0
78.1
79.1
ESP/Scrubber
98.9
NA
>67
92.9
97.4
97.7
97.2
Fabric Filter
99.6
NA
NA
99.1
NA
NA
NA
(a) Only two values were reported for arsenic removal efficiency lor wet scrubbers. Because the values
varied so significantly, the arsenic removal efficiency of a scrubber is given here as a range rather than as an
average value.
(b) Copper is not listed as a hazardous air pollutant under Title III of the Clean Air Act Amendments.
NA - Data not available.
31
-------
large particles (1 |im and larger) than small particles. Therefore, where significant amounts of
metals are condensed on these larger particles, scrubbers may show relatively high removal rates
of metal HAPs. Where the metals are concentrated in the smaller particles, scrubbers will not be
as efficient.
ESPs comprise the primary particulate removal system for utility boilers firing pulverized
coal, and are relatively efficient in removing a significant fraction of particles, both by mass and
by number. As with most particulate removal equipment, ESPs are more efficient when the
particle sizes are larger, and are less efficient when removing submicron particles. Particles of
0.1-0.3 [im in size are less likely to be captured by an ESP than are panicles of 1 Jim or larger.22
ESP performance can be significantly improved by "conditioning" the flue gas, Conditioning
involves the injection of a chemical, typically ammonia (NH3) or SO3, or in some cases sodium
carbonate (NaCOj), into the flue gas. Conditioning can improve the average capture rate by 2-
5%. It is not clear what interactions may occur between conditioners and metals; however, it is
likely that these interactions will primarily affect the volatile metals rather than the refractory
metals, since the relatively low temperatures found in ESPs are not adequate to cause significant
chemical interaction with the refractory metals. Emissions of metals such as mercury and
selenium may, however, be affected by the conditioning agent, although currently no data are
available to evaluate these potential effects.
Baghouses or fabric filters are the most efficient means of particulate removal, and in some
instances are able to reduce the emissions of metal HAPs by over 99%. This is primarily true for
those compounds that have condensed onto fly ash or into a solidified metal aerosol; the more
volatile meials such as mercury are captured by fabric filters only when they are first absorbed
onto particles that are then captured on the filter.
Spray dryer absorbers (SDAs) in conjunction with particulate removal equipment are also
capable of high removal efficiencies, even in the case of mercury. In a study of eight power
plants that used SDA systems, mercury captures of over 99% were measured across the SDA and
particulate removal system.23 However, the removal efficiencies ranged from a low of 6% to
one case of over 99% removal. The most important factor that appeared to affect the removal
efficiencies was the particulate loading of the flue gases, with the higher removal efficiencies
correlating with higher particulate loadings. An additional factor appeared to be the chlorine
content of the coal, which showed an apparent increase in mercury reduction efficiency with
increasing chlorine content. Although these correlations were not examined in detail, it is clear
from these results that it is difficult, if not impossible, to predict the emissions of mercury a
priori based on the mercury content in the coal without considering other factors.
32
-------
D. Effects of Coal Cleaning
The above results are based on the trace element analyses and combustion of the different
coals with little or no preparation beyond grinding. In many cases, however, coals are washed to
some degree prior to being fed into the furnace. This washing can vary from a simple washing
with water to remove at least a portion of the excluded mineral fraction to much more
sophisticated coal cleaning processes which can remove significant amounts of both sulfur and
trace metals.
Coal cleaning is used to remove or reduce the inorganic fraction of the raw coal. The degree
to which trace metals are removed from the coal during washing processes therefore depends
strongly upon the degree to which each metal is physically or chemically associated with the
organic fraction of the raw coal. Studies by Finkelman et al. examined the organic and mineral
associations of trace elements in coal.24*25 When trace elements are embedded in the organic
matrix of the coal, the ability of washing to remove that element decreases. The Finkelman
results indictated that antimony, beryllium, and selenium were associated with the organic
matrix, while arsenic, lead, and mercury were more closely associated with pyrite in the coals. In
the coals tested, therefore, coal cleaning will reduce arsenic, lead, and mercury by a greater
degree than antimony, beryllium, and selenium. As with the emissions, however, the degree to
which coals can be washed of trace elements will be a-function of the characteristics of the
individual coals.
In general, the levels of metals in washed or cleaned as-fed coal will most likely be less than
the trace concentrations of raw, as-mined coal, and the as-fed concentrations can be significantly
less if the coal is cleaned extensively.26"28 The degree of reduction will vary from coal to coal
depending on the cleaning process and on the composition of the coal. As with the as-found
trace metal concentrations, the as-fed trace metal levels will vary among different coals, with no
generally applicable correlation between trace metal concentration and sulfur content.
E. Other Approaches to Predicting Emissions
EPRi's PISCES field test program2-3 has gathered a considerable amount of full scale
emissions data from a variety of utility power plants. These data have been taken from a range
of plant sizes and types, with different pollution control equipment, and using different coal
types and sources. In their approach to estimating emissions for the purpose of assessing the risk
to human health from toxic emissions from utility boilers, EPRI has developed a series of metal-
specific correlations to empirically estimate the emissions of trace metals from coal-fired utility
boilers based on the trace metal and ash contents of the coals being used. These correlations do
33
-------
not account for plant type or installed pollution control devices. This approach has been chosen
by EPRI researchers as the best alternative available to them for providing a nationwide estimate
of toxics emissions.
The approach taken by EPA's OAQPS, on the other hand, does distinguish emissions
between plant and pollution control device type. While this approach requires more complex
calculations to be performed to estimate emissions, it removes a large number of the
uncertainties associated with a single correlation for all plants and control device types. Both
EPRI and EPA estimates of HAP emissions are to be used as the basis for assessments of the risk
to human health posed by those emissions from utility boilers. The risk assessment procedures
contain considerable levels of uncertainty in themselves, and it is most desirable to minimize the
uncertainties wherever possible. The OAQPS approach then seems to provide data that are more
accurate than the EPRI approach, by accounting for differences associated with individual plant
characteristics, rather than assuming that all changes in HAP emissions are due to changes in
coal characteristics and feed rates.
In both cases, however, changes in trace element content of the feed coal was linearly
translated into changes in metal emissions. While it is desireable to use a more accurate model
that would account for the different behavior seen in these small scale tests, the state of the art is
not yet adequate to produce an accurate predictive model.
VI. Conclusions
From the scaled distributions, it has been demonstrated that the coals tested behave
differently with regard to the emissions of refractory metals. As is often the case in studies of
metal emissions, the heterogeneity of the fuel and the relatively low metals concentrations typical
of combustion systems lead to variations in the data from one run to another. It is therefore
impossible to state that these differences will hold in all cases. However, the data are-consistent
enough to provide a strong indication that the common practice of estimating metals emissions
based entirely upon the metals content of the coal is not completely accurate. Further, since the
metals content does not in general correlate with sulfur content, it is therefore impossible to
accurately make sweeping statements concerning the effects of using low sulfur coal on HAP
emissions. While western coals typically contain higher ash and have a lower Btu content, the
amounts of trace metals in these coals vary significantly when compared to high sulfur eastern
coal, making it impossible to make generalizations concerning the effects of fuel switching on
emissions of metal HAPs. A significant amount of additional study would help to provide a
34
c
-------
more comprehensive means of estimating metals emissions based on the coal and combustion
parameters.
It is therefore not possible to make a single broad statement about how changing coal will
affect the emissions of toxic metals for the industry as a whole. Predicting metals emissions
based entirely upon the trace metal concentrations of the coal is an overly simplified approach,
although this approach may be necessary when evaluating broad national or regional impacts.
While there will likely be some increase as concentrations of metals in the coal increase, it is
impossible to accurately predict the actual changes, due to the effects noted above, .
Predicting emissions of metals from the combustion of coal based only upon the as-received
concentration of those metals in the coal is not accurate in all cases,
VII References
1. Public Law 101-549, Clean Air Act Amendments of 1990, Nov. 15, 1990.
2. W.D. Balfour, Chow, W„ and Rubin, E.S., "PISCES: A Utility Database for Assessing the
Pathways of Power Plant Chemical Substances," Paper No. 89-71.6, presented at the 1989
AWMA Annual Meeting, Air & Waste Management Association, Anaheim, CA, 1989.
3. Behrens, G.P. and Chow, W., "Use of a Multi-Media Database for Chemical Emission
Studies of Conventional Power Systems," Paper No. 90-131.1 presented at the 1990 Air &
Waste Management Association Annual Meeting, Pittsburgh, PA, 1990.
4. The Mcllvaine Company, "Fuel Combustion," in Air Toxics and VOCs: Markets and
Technology for Compliance Analysis. Control Technology, and Measurement. The
Mcllvaine Company, Northbrook, IL, October 1991.
5. Linak, W.P. and Wendt, J.O.L., "Toxic Metal Emissions from Incineration: Mechanisms
and Control," Progress in Energy and Combustion Science, 19, pp. 145-185, 1993.
6. Davison, R.L., Natusch, D.F.S., Wallace, J.R,, and Evans, C.A. Jr., "Trace Elements in Fly
Ash: Dependence of Concentration on Particle Size, Environ. Sci. Teehnol., 8, pp. 1107-
1113, 1974.
7. Kaakinen, J.W., Jorden, R.M., Lawasani, M.H., and West, R.E., "Trace Element Behavior
in a Coal-Fired Power Plant," Environ. Sci. Teehnol., 9, pp. 862-869,1975.
8. Smith, R.D., Campbell, J. A., and Nielson, K.K., "Volatility of Fly Ash and Coal," Fuel, 59,
pp. 661-665, 1980.
9. Damle A.S., Ensor, D.S., and Ranade, M.B., "Coal Combustion Aerosol Formation
Mechanisms: A Review," Aerosol Sci. Teehnol., 1, pp. 119-133, 1982.
10. Linak, W.P. and Peterson, T.W., "Mechanisms Governing the Composition and Size
Distribution of Ash Aerosol in a Laboratory Pulverized Coal Combustor," 21st Symposium
(International) on Combustion, pp. 399-410, Combustion Institute, Pittsburgh, PA, 1986.
11. Kauppinen, E.I. and Pakkanen, T.A., Coal combustion aerosols: a field study, Environ. Sci.
Teehnol, 24, pp. 1811-1818,1990.
12. . Markowski, G.R., Ensor, D.S., Hooper, R.G., and Can-, R.C., "A Submicron Aerosol Mode
in Flue Gas from a Pulverized Coal Utility Boiler," Environ, Sci. Teehnol. 14, pp. 1400-
1402, 1980.
13. McCain, J.D., Gooch, J.P., and Smith, W.B., "Results of Field Measurements of Industrial
Particulate Sources and Electrostatic Precipitator Performance," JAPCA 25, pp. 117-121,
1975.
35
-------
14. Seeker, W.R., "Overview of Metals Behavior in Combustion Systems," AS ME/EPA
Workshop on Metals in Incineration, Cincinnati, OH, November 1991.
15. Seeker, W.R., "Metals Behavior in Waste Combustion Systems," presented at the 1992
International Joint Power Generation Conference, Atlanta, GA, October 16-22, 1992.
16. Gang, S., "Methodology for the Determination of Metals Emissions in Exhaust Gases from
Hazardous Waste Incineration and Similar Combustion Processes," in Methods Manual for
Compliance with the BIF Regulations, EPA/530-SW-91-010 (NTIS PB91-120006), U.S.
EPA Office of Solid Waste, Washington, DC, December 1990.
17. U.S. Environmental Protection Agency, "Method 101A - Determination of Particulate and
Gaseous Mercury Emissions from Sewage Sludge Incinerators," in Code of Federal
Regulations, Title 40, Part 61, Appendix B, U.S. Government Printing Office, Washington,
DC, 1991.
18. Schmidt, C.E. and Brown, T.J., "Results from the Department of Energy's Assessment of
Air Toxic Emissions from Coal-Fired Power Plants," presented at the 1994 International
Joint Power Generation Conference, Phoenix, AZ, October 2-6, 1994.
19. U.S. Environmental Protection Agency, "Study of Hazardous Air Pollutant and Mercury
Emissions from Electric Utility Steam Generating Units Pursuant to Section 112(n) of the
Clean Air Act Amendments of 1990 — Interim Report to Congress," EPA-453/R-93-051 a
(NTIS PB94-187820), U.S. EPA Office of Air Quality Planning and Standards, Research
Triangle Park, NC, November 1993.
20. Brooks, G., "Estimating Air Toxics Emissions from Coal and Oil Combustion Sources,"
EPA-450/2-89-001 (NTIS PB89-194229), U.S. EPA Office of Air Quality Planning and
Standards, Research Triangle Park, NC, April 1989.
21. Bounicore, A.J., "Wet Scrubbers" in Air Pollution Control Equipment: Selection, Design,
Operation and Maintenance, Theodore, L. and Bounicore, A. J., eds., Prentice-Hall,
Englewood Cliffs, NJ, 1982.
22. Bibbo, P.P., "Electrostatic Precipitators," in Air Pollution Control Equipment: Selection,
Design, Operation and Maintenance, Theodore, L. and Bounicore, A J., eds., Prentice-Hall,
Englewood Cliffs, NJ, 1982.
23. Felsvang, K., Gleiser, R., Juip, G., and Nielsen, K.K.. "Control of Air Toxics by Dry FGD
Systems," presented at Power-Gen '92 Conference, November 17-19, 1992, Orlando, FL.
24. Finkelman, R.B., "Modes of Occurence of Trace Elements and Minerals in Coal: an
Analytical Approach," in Atomic and Nuclear Methods in Fossil Energy Research. R.H.
Filby, B.S. Carpenter, and R.C. Ragaini, eds., Plenum Press, New York, 1982, pp. 141-149.
25. Finkelman, R.B., Stanton, R.W., Cecil, C.B., and Minkin, J.A., "Modes of Occurrence of
Selected Trace Elements in Several Appalachain Coals," American Chemical Society, Fuel
Chemistry Division, Vol. 24, pp. 236-241, 1979.
26. Finkelman, R.B., Palmer, C.A., Krasnow, M.R., Aruseavage, P.J., Sellers, G.A., and
Dulong, F.T., "Combustion and Leaching Behavior of Elements in the Argonne Premium
Coal Samples," Energy and Fuels, Vol. 4, pp. 755-766,1990.
27. DeVito, M., Rosendale, L., and Conrad, V., "Comparison of Trace Element Contents of
Raw and Clean Commercial Coals," presented at the DOE Workshop on Trace Elements in
Coal-Fired Power Systems, Scottsdale, AZ, April 1993.
28. Ford, C. and Price, A., "Evaluation of the Effects of Coal Cleaning on Fugitive Emissions:
Final Report, Phase in," DOE/EV/04427-62, July 1982.
36
-------
Appendix A. Data Sheets
I. Coa! Analysis Data Sheets
37
-------
02^19/93 14:52 CT&E, SOUTH HOLLAND. IL -» 1 919 544 5630
NO.082 P003
COMMERCIAL TESTING St ENGINEERING CO.
GEWERW. OFFICBS: 1919SOUTH HIGHLAND AVE., SUITE ZlfrB, LOMBAKO. tLLIM6)S OOI«a« 17031963-8300
Member ol Iha 5G$ Owp (SowW at Sunwtuica)
February 19, 1993
ACOREX CORP., DURHAM NC
ENVIRONMENTAL SYSTEMS DIVISIOM
4915 PROSPECTUS DRIVB
DURHAM, NC 27713
ATTN; Kathy Hinton
PLEASE AOORESS ALL OOW»eS«5NDEMCS TO:
161S0 VAN 0RUN*£M RO, P.O. 0OX (27
SOUTH HOLLAND. H ©>473
TELEPHONE; (TO) SS1-2900
TELEX. Z8S950 COMTeGO SH UR
FAX; (?08| 33MOSO
Sampl« identification by
ACTTREX CORP., DURHAM NC
Kind of
reported to ua Western Coal
Sample taken at C-Wing C-225-C
Sample taken by ••••«
Date sampled
Date receivad February a, 1393
Saropie ID;
ti_*S
BC # s
P.O. Ho. CH 38900 B
Analysis Report Ho. 11-49767
Page i of 2
PROXIMATE AHM.YSIS
Pl/ftKATE XHALVSia
As Raaaived
Dry Baaia
.
Ai* Received
Dry Basie
% Moisture
4.01
JOaOCX
% Moisture
4 .01
xwcxx
% Ash
a.si
9 .26
% Carbon
71.54
74,53
% Volatile
33.64
35,05
% Hydrogen
4 . SI
5.01
Fixed Carbon
S3. 44
55.67
% Nitrogen
1.60
1.67
100.00
100.00
% Sulfur
1 ,45"
1.51
% Rfih
8.81
3.26
Btu/lb
15831
1336?
% Oxygen(dlff)
7.68
8 .00
100.00
100,00
% Sulfur
1.45
1 .51
MAP Bfcu
14734
Method: Moisture ASTM Designation D 3173
Ash per ASTM Designation d 3174
volatile per ASTM Designation D 3175
Etu per ASTM Designation D 201S or 3286
Sulfur per ASTO Designation D 4239 (Method C>
Fixed Carbon [calculated Value) is the
Resultant of the syranax ion of percentage Moisture,
Ash, and Volatile matter subtracted, for 100.
Cartoon arxJ Hydrogen by infrar©<3 detection, Nitrogea
by Thermal - conductivity.
Rejgpecffutiy
COUU^RCIAL TEST1«
Msnagef. Sewdi Housed UibOr&^ry
OVER 40 6BANCH LABORATOFUeS STRATEGICALLY u^tB) W P«NQP« COAL WUMG *R£AS, TIDEWATER AWO CRCAT UKKS PORTS, AND f»VEB LOAO«C FACJUTJ^C
Jflgjna." WAfc*'rt«Vk«/'J ?or Youf <'ioie
-------
02/19^33 14:53 CT££, SOUTH HOLLAND, IL ¦» 1 919 541 5630
NO.002 P004
COMMERCIAL TESTING & ENGINEERING CO.
GeneralarriCCS: m9 South highland ave., sutxe 210-6. tOMBARO. Illinois G01 «3 ~ C7<») 953-9300
MUCE 10M
NG fACILTCS
• nej
Original V(»yi Prc>Us3-ji;\
fims KH0 CQf
-------
02/19/33 1^:53 CT&Er SOUTH HOLLPHDi IL -» 1 gi9 5/14 5530
NO.002 P005
COMMERCIAL TESTING & ENGINEERING CO.
QENCHALOFFICCS; 1919 30IJTH HIGHLAND AVt„ SUITE 210-e. LOMBMO, ILLINOIS6Q148~ 0081 9KJ-B3M
M«rr«er c* ens SGS Group noetic Q«h#t»<> =* SufrHi***!
February 19, 1993
ACURSX. CORP ¦ , DURHAM MC
ENVIRQNMEHTAIi SYSTEMS DIVISION
4915 PROSPECTUS DRIVE
DURHAM, KC 27713
ATTF; Kathy Hinton
fiXasg adoress au. cotoesponosmcg to-
18130 VAN ORWffiN RD,, P.O. BOX 127
SOUTH HOLLAND, R, ®X?3
. by
ACURKX CORP. , DURHAM EC
- Kind of sample
reported to ua Illinois Coal
Sample taken, at C-King C-225-C
Sample taken by * ¦ - - -
Date sampled —
Date received February ft, 1993
Sample ID: xh # 6
P.O. PTO, CH 35900 E
Aaalycio Report ho. 71-49763
. Page
PROXIMATE MJALYfllS
s
jl
£
K
V)
H
m
AO
Received
Drv Basie
AR
ROCQivad
Dry Bai
% Moisture
6 .48
KXJMC
% Kqioture
6 .48"
3CXJOOC
% Ash.
8 .41
a.99
% Carbon
67 .14
71.79
% Volatile
33 .€5
35.9a
% Sydrogeji
4 .65
i .97
% Fixed Carbon
51 .46
55.03
% Hitrogoa
1 .41
1. 51
100 ~00
100.0O
% Sulfur
2 .76
2.95
% Ash
8 .41
8,99
Btu/lb
12128
12968
% Oxygen(di££)
. 3.15
9 .79
100 .00
100.00
% Sulfux
2.76
a .95
MAT Btu
142«$
1 of 2
Method: Moisture ASTU. Obsignation D 3x73
Ash per aStM Designation d 3174
Volatile par ASTM E>»sIgaation, D 3175
Btu per ASHK Designation D 2015 or 3286
Sulfur per JiSTM. Designation D 4239 (Method C)
Fixed, Carbon (calculated. Value) ic the
Raeultant of the sunmatioh of percentage Moisture,
Ash, and Volatile matter subtracted for 100.
Carbon and Hydrogen by infrared detection. Nitrogen
by Thermal - conductivity•
RaSfMcUuUy »uom(!ta
-------
33/19/03 14:54 CT2E, SOUTH HOLLAND, 1L -) 1 919 544 5630
no,00a peag
COMMERCIAL TESTING fie ENGINEERING CO.
GENERAL OFFICES: 1818 SOUTH HIGHLAND AVE , SUfTE ZIO-B, UOMBAHO, 11-LtNClS60148 • 08J 353-9300
SiuCC 1*»
MtfntWf e( th4 6GS Gtovp (Soc«i<< oa Surw«*»c*)
February 19, 1993
ACUREX CORP., DORHAK HC
ENVIRONMENTAL SYSTEMS DIVISION
491S PJROSPECTCS DRIVE
DURHAM, NC 27713
ATTN: K&Ztoy Hinton
PLEASE *D0R£38 «JL COn
ttftMs m> 03MKH0NS m wmsz
41
-------
02/19/93 16:4G CTg.G> SOUTH HOLLAND, IL •» 1 913 54,, 5Sga
HQ.024 P001
COMMERCIAL TESTING & ENGINEERING CO.
GSNCflAt. OFFICES: 1919 SOUTH HIGHLAND AVE.. SUITS 210 B. lOMHABn, ILLINOIS 6014S • (708! SS3-S3O0
M«TO«r Of the SGS GraJp (asomc* 0*>4i6t
Btu/lb
9033
11322
% Oxygen(dif£)
14 .40
18. OS
100.00
100.00
% Sulfur
0.37
0.47
MAP Bt\j
12509
Method; Moisture AfTB-i Designation D 3173
Ash per ASTK Designation D 3174
Volatile per ASTtfl Designation a 3175
Btu per ASTH Designation, D 2015 or 3286
Sulfur per &STM designation D 4239 (Method Cl
fixed Carbon (calculated Value} is the
Resultant of the surtsnation of percentage Moisture,
Ash, end Volatile Wctet subtracted for 100.
Carbon and Hydrogen by infrared detection, nitrogen
by Thermal - conductivity.
F»«cp«vT«. COAL aidlMG AREAS. TlflDWATtn AN0 GftfAT UK£S t-owa AM0 WVCO U3AO»4a IWUT1E5
•4C£# *
Iilo.'nel Wj:i««w;ir»i
-------
02/19/33 IS:47 CTS.E. SOUTH HOLLAND, IL ¦* 1 919 544 5690
HO,024 P002
COMMERCIAL TESTING & ENGINEERING CO.
general 0«=ICKS: tats SOUTH HIGHLAND Ave., SUITE 2t0 B, LOMBARD, ILLINOIS 60I4S* rraftl 933-S300
MMOtr cS |