Symposium on Energy Engineering in the 21st Century
January 9-13, 2000
Hong Kong, P.R. China
ENVIRONMENTAL CONTROL OF TOXIC METAL AIR EMISSIONS FROM THE
COMBUSTION OF COAL AND WASTES
Jost O.L. Wendt, Sheldon B. Davis, Thomas K, Gale, and Wayne S. Seames
Department of Chemical and Environmental Engineering
University of Arizona, Tucson, AZ 85721 USA
William P. Linak
Air Pollution Prevention and Control Division, MD-65
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711 USA
Keywords: toxic metals, combustion, incineration, sorbents
Abstract: The emission of toxic metals from combustion of fossil fuels and wastes is an important global
environmental issue. Toxic metals, such as arsenic, selenium, mercury, chromium, lead, and cadmium, are
present in coals and in many municipal and industrial wastes. This paper is concerned with the partitioning of
these metals during combustion, and with the mitigation of their effect on the environment using high
temperature sorbents. The paper is divided into three parts. First, the partitioning of arsenic and selenium
during coal combustion in a 17 kW laboratory down-fired furnace is discussed, and appropriate mechanisms
identified. Second, the speciation of mercury and chromium during combustion is addressed, through special
experiments on an 82 kW refractory-lined combustor. Third, experimental results on the sorption of individual
and multiple metals on sorbents are presented. These sorbents were kaolinite and lime, and were injected
directly into flue gas containing lead and cadmium, which had vaporized in the main flame. Results suggest
that toxic metals from coal and waste combustion can interact with lime or kaolinite sorbents and that, for some
multiple metal mixtures, designer sorbents containing calcium, aluminum, and silicon might be useful to
capture them and render them environmentally benign.
INTRODUCTION
Toxic metal emissions pose a problem in both fossil fuel and waste combustion [1,2,3], Toxic metals can be
classified as non-volatile (barium, beryllium, chromium, nickel), as semi-volatile (antimony, arsenic, cadmium,
lead ), or as volatile (mercury, selenium), depending on their volatility at stack temperatures [1]. This paper
focuses on the volatile metals, mercury and selenium, and the semi-volatile metals, arsenic, lead, and cadmium.
The fates of selenium and arsenic are investigated as they relate to the combustion of pulverized coal. Both
metals occur in many coals around the world, and arsenic can reach major concentrations in certain Chinese
coals. Mercury speciation, a major issue in both fossil fuel and waste combustion, is investigated through
special tests in which a clean flame is doped with a mercury salt. Lead and cadmium are discussed insofar as
their emissions can be diminished through reaction of the metal vapor on sorbents such as kaolinite or lime.
Cadmium and lead can vaporize in the combustion zone, subsequently to nucleate and form a difficult-to-collect
fume. The fact that both lead and cadmium vaporize may allow solid substrates, such as kaolinite and lime, to
react with metal vapors and render them insoluble in water. This process may be used to scavenge these metals
preventing subsequent gas-to-solid transformations that would result in submicron particle formation.
Fig 1 shows the possible mechanisms that control the partitioning of toxic metals during combustion. These
metals may be introduced as aqueous solutions, as hydrocarbon fluids or solids, or as solids, municipal wastes,
and sludges. For pulverized coal, metal partitioning follows the route through porous char, where the metal
may be contained extraneously in excluded mineral matter fragments, contained in inherent occlusions of
-------
mineral matter within the char matrix, or organically bound directly to the organic matrix. The initial
occurrence of the metal may well have a significant bearing on the subsequent partitioning routes it will follow
as it passes through the combustion process. Of special interest in this paper are the routes that prevent toxic
metals being emitted from the stack in submicron form.
EXPERIMENTAL APPROACH
The approach followed was to introduce the appropriate metal into a combustion chamber and, except in the
case of mercury speciation, to infer its partitioning mechanisms through size segregated analyses of the ensuing
particulate matter in the exhaust. For mercury speciation, the wet chemical method developed by Ontario
Hydro [4], was used.
Experimental data presented here were obtained from two different laboratory combustion rigs. One rig located
at the U.S. Environmental Protection Agency (EPA) was a 59 kW (actual), 82 kW (maximum rated) horizontal
tunnel combustor. This refractory-lined research combustor was designed to simulate the time/temperature and
mixing characteristics of practical industrial liquid and gas combustion systems. Natural gas fuel, aqueous
metal solutions, gas dopants, and combustion air were introduced into the burner section through an
International Flame Research Foundation (IFRF) moveable-block, variable-air, swirl burner. Further details
regarding the experimental combustor can be found elsewhere [5,6]. The other rig, located at the University of
Arizona (UA), consisted of a 7 m, 17 kW downflow combustor. Details can be found elsewhere [7,8]. Like the
EPA furnace, this facility was also designed to have a time and temperature history similar to that of a full-scale
system, yet it was well defined and could be considered one-dimensional. In both systems, metals could be
introduced into a natural gas flame, through an atomizer in the form of the aqueous stream. Metal salts were
dissolved in distilled water in a quantity such that the concentration of each metal on an elemental basis is 100
ppm in the flue gas. In this paper, all the pulverized coal results reported were obtained from the 17 kW UA
down-fired furnace, while mercury and chromium speciation data were gleaned from the 59 kW EPA
refractory-lined furnace. Results describing the metal/sorbent interactions reported here are taken from data
from the UA furnace. In that unit, kaolinite or lime sorbents were introduced as powders suspended in nitrogen
through a port downstream of the combustion zone.
Particulate samples were taken by the isokinetic sampling and size segregation system shown in Fig 2. The
samples were drawn into a sampling probe, and immediately mixed with a controlled amount of nitrogen. The
nitrogen immediately quenched the stream and stopped all subsequent reactions. The diluted sample was
directed immediately to an impactor, which size-segregated the particles. A variety of impactors were used.
The atmospheric pressure Andersen impactor allowed segregation onto nine stages, while the Micro Orifice
Uniform Deposition Impactor (MOUDI) allowed better resolution for the smaller particle size range. The
Bemer low-pressure impactor (BLPI) was the experimental centerpiece for much of the data presented here and
is also shown in Fig 2. Whereas conventional cascade impactors operate at atmospheric pressure under which
separations are performed purely by varying the gas velocity through successively smaller orifices, the BLPI, by
operating at low pressures, is able to take advantage of compressible flow properties and can attain much
smaller particle cut-off diameters than atmospheric impactors.
The BLPI is composed of 11 stages having aerodynamic cut-off diameters ranging from 15,7 to 0.0324 ujn.
Each stage is composed of an orifice plate above a substrate and a substrate holder. The substrate is held in
place by a ring, which also acts as a spacer between the orifice plate and the substrate. The flow rate through
the BLPI is regulated by the last orifice plate, which acts as a sonic orifice. The absolute pressure after this
final orifice plate is approximately 8 kPa. A full description of this apparatus and its performance can be found
in reference [9]. Each substrate was analyzed by atomic absorption spectroscopy after digestion in a 3 HF: 1
HNO3:1 HC1: 5 H2O solution.
The mode within the particle size distribution in which the metal of interest escapes is used to infer pertinent
mechanisms controlling its fate. For example, the efficacy of reactive capture of toxic metals by sorbents can
be determined by determining how sorbent addition shifts the particle size distribution (PSD) containing the
metal. Previous work [7,8] has shown that, when metal vapors are sampled, they appear as a very fine nuclei
fume at about 0.1 |im diameter.
-------
RESULTS
Arsenic and Selenium During Pulverized Coal Combustion
The fate of arsenic and selenium during coal combustion is discussed here in the context of the mechanism
pathways shown in Fig 1. Experiments were conducted on the UA 17 kW laboratory combustor, in which
pulverized coal was burned in a self-sustaining mode. Experimental results, not shown here for lack of space,
have established [10] that most of the arsenic in fly ash from pulverized coal combustion is contained in
particles smaller than 1-2 |im, which is in the fine particle range. However, although arsenic is a semi-volatile
metal, there is little evidence that arsenic vapors form arsenic-rich nuclei in the exhaust. Rather, arsenic is
associated with calcium, when that metal is present in coals. Fig 3 shows the correlation between arsenic and
calcium for three U.S. coals. The upper two panels show good correlation between these two elements for
samples taken at the exhaust (2.2 s residence time), but not necessarily closer to the flame at 0.5 s residence
time. This may indicate that the reaction time scales are different for the two metals. Selenium may interact
with the calcium faster than the arsenic. The lower panel, for the Ohio coal, does not show a correlation
between arsenic and calcium, suggesting that any association between these elements is weak for that coal. The
two coals in the upper panels contained sufficient calcium to react with the available arsenic. The Ohio coal in
the bottom panel did not contain enough calcium for sorption of the arsenic available. These data suggest that
arsenic will follow calcium, and that the ensuing particle-size-segregated speciation particle is determined by
the aerosol mechanisms followed by the calcium.
Analogous selenium data are shown in Fig 4 and show that selenium is rapidly associated with calcium early in
the flame. This is true even for the low calcium Ohio coal, and indicates that calcium reacts more rapidly with
selenium that it does with arsenic. For all three coals, selenium is associated with calcium very early in the
combustion process, at 0.5 s residence time. Calcium is clearly a very important metal in determining the
ultimate partitioning of both arsenic and selenium.
Mercury Speciation
While the sources and fate of elemental mercury constitute a global environmental issue, the risk assessments
often associated with the emissions and deposition of water soluble oxidized mercury species are often more
closely related to their local and regional environmental impacts. Mercury speciation is also important, since
the solubility of oxidized forms promotes their potential removable from stack gases by scrubbing. Fig 5 shows
results from 73 kW natural gas flames doped with mercury nitrate [Hg(NO3)2] such that 100 ppb mercury in the
stack gas was produced at a fuel/air stoichiometric ratio (SR) of 1.2 (fuel lean), or 140 ppb mercury at SR=0.8
(fuel rich). Diatomic chlorine (C12) and sulfur dioxide (SO2) levels were 500 and 1000 ppm, respectively, at
SR=1.2. The Ontario Hydro method was used for mercury analysis, and the sample was withdrawn at a
temperature of 640 K. The data show that C12, added to the air stream at the burner, resulted in almost complete
conversion of mercury to the ionic form. Note that less than 1 ppm C12 can be expected to leave the flame, with
the remainder probably converted to hydrochloric acid (HC1). It is not surprising that similar results were
obtained when C12 was added to the flue gas downstream. Slight oxidation was also observed with addition of
SO2, Surprisingly, only elemental mercury was observed in the absence of C12 or SO2, and under all rich
conditions, with or without C12 or SO2.
Chromium Speciation and PSD
Speciation of chromium is an important issue, since the hexavalent form of chromium is much more harmful to
human health than the trivalent form, which is relatively environmentally benign. Fig 6 shows results from the
EPA 59 kW refractory-lined furnace, with hexavalent chromium oxide [CrO3(VT)] and trivalent chromium
nitrate [Cr(NO3)3(III)] added to distilled water and sprayed into a gas flame. Due to chromium (III) digestion
problems, the analysis of total chromium is quite difficult, and a caustic fusion method was employed as
described in Reference [11], Clearly, at these conditions, the initial form of chromium has a large influence on
the particle size distribution (PSD) emitted, while it appears to have no influence on the speciation of
chromium. The fraction of chromium that is generated as hexavalent chromium is very small, and less than 1%
in the absence of chlorine. In further experiments, hexavalent chromium formation is enhanced by chlorine (to
5% fraction emitted) but greatly diminished by sulfur. In fact, the addition of sulfur (even in the presence of
chlorine) can completely suppress the formation of hexavalent chromium in combustion systems.
-------
Further tests including both sulfur and chlorine indicate that even a stoichiometric amount of sulfur suppresses
chromium (VI) formation [6].
Reactive Scavenging of Hot Toxic Metal Vapors by Sorbents
The efficacy of reactive capture of toxic metals by sorbents can be determined by determining how sorbent
addition shifts the particle size distribution (PSD) containing the metal. Previous work [8] has shown that metal
vapors appear as a very fine particles (~0.1 u.m aerodynamic diameter) when extracted by the sampling system
described earlier. Further, there is insufficient time within the sampling system for significant coagulation to
occur. Discrimination between physical condensation and reactive scavenging is more difficult. The former
process cannot occur above the metal dewpoints, and, at the concentrations encountered here, generally requires
a greater external surface area than that afforded by the sorbent injection rates here. In this paper, we restrict
ourselves to three systems: 1) lead capture by kaolinite and lime, 2) cadmium capture by kaolinite and lime, and
3) mixtures of lead and cadmium interacting with kaolinite sorbent.
Lead/Kaolinite and Lead/Lime Systems. Figs 7 and 8 show the changes in lead aerosol size distribution
when kaolinite (Fig 7) and hydrated lime (Fig 8) are individually injected at -1450 K. Samples were taken at
the bottom of the furnace (-900 K), below the dewpoint of the metals.
Without sorbent, the sampled lead aerosol has a predominant size of 0.1 \im (Fig 7), suggesting a freshly
coagulated aerosol of lead nuclei. Upon introduction of kaolinite powder, the lead PSD shifts by over a factor
of 10 in particle size, and (except for a fine aluminum tail) appears to correlate with the aluminum PSD, which
represents the injected sorbent powder. However, with lime injection (Fig 8), only a minimal change in the lead
aerosol PSD is discerned. Clearly lead is almost completely reactively captured by kaolinite (Fig 7); whereas,
only a small portion is scavenged (most probably by physical condensation) by lime (Fig 8). Therefore,
comparison of Figs 7 and 8 allows reactive scavenging and simple physical condensation processes to be
distinguished in samples taken after the metal dewpoint. At the metal and sorbent concentrations used here
(100 ppmv Pb and 5/1 AlrPb molar ratio), reactive scavenging is demonstrated by the data in Fig 7, while other
forms of interaction are demonstrated in Fig 8 (although partial reaction cannot be completely ruled out, unless
samples are also taken above the dewpoint).
Cadmium/Kaolinite and Cadmium/Lime Systems. Figs 9 and 10 show the analogous data for cadmium
capture on kaolinite (Fig 9) and cadmium capture on lime (Fig 10) where each experiment consisted of injection
of one metal and one sorbent at a lime. Fig 9 shows that cadmium is partially captured by kaolinite, but
substantially more so than lead was captured by lime. The coarse cadmium peak is believed to represent partial
scavenging, rather than physical condensation because of the large fraction of cadmium that was transferred to
the coarse particle size. Note that the rate of cadmium reaction with lime appears to be significantly faster than
the reaction between lead and lime. Results below support this conclusion.
Multiple Metals (Lead and Cadmium) Interacting with Kaolinite. Experiments were performed in which
the flue gas contained 100 ppm lead vapor together with 100 ppm cadmium vapor. Dewpoints for lead and
cadmium were very close, within 25 K of each other. Hence times for reaction between metal vapor and
sorbent are the same for both metals. Kaolinite was injected such that the molar ratio of aluminum to total
metal in the system was approximately unity. Results are shown on Figs 11 and 12, a comparison of which
clearly demonstrates that lead is preferentially reactively scavenged by kaolinite. Essentially all of the lead
follows the coarse aluminum mode; whereas, essentially none of the cadmium is now affected by the kaolinite.
Additional experiments in which the multi-metal mixture was exposed to lime sorbent showed that, as expected,
the lime captured the cadmium, but not the lead. These data are not presented here.
CONCLUSIONS
Arsenic and selenium from pulverized coal appear to be associated with calcium when there is sufficient
calcium present in the coal. When calcium levels are low, as in the Ohio coal, there is no correlation between
arsenic and calcium, although there is a correlation between selenium and calcium. Selenium and calcium
appear to react rapidly in the combustion zone, while arsenic and calcium react more slowly downstream in the
post-flame region.
-------
Mercury speciation, oxidized versus elemental, seems to be profoundly influenced by chlorine, especially
molecular chlorine. Results suggest that oxidized mercury can be found only under lean-combustion
conditions. Chromium in the exhaust appears primarily as trivalent chromium, with less than 1% present as
hexavalent chromium. The outlet valency form appears to be independent of the form that was fed to the
furnace. The outlet PSD depended strongly on the initial form fed to the furnace.
Emissions of toxic metals, which constitute an environmental problem in the combustion of waste and fossil
fuels, can possibly be managed through the injection of commercial sorbents, such as kaolinite and lime.
Kaolinite is very effective at reactively scavenging lead vapor, and moderately effective at capturing cadmium,
provided there is little competition for reactive sites. When both lead and cadmium are present, the kaolinite
scavenges lead preferentially to the cadmium. The situation is reversed for lime as a sorbent. Lime captures
cadmium very effectively but is ineffective for lead. These data suggest that effectively designed sorbents for
multiple metals might consist of mixtures of kaolinite and lime, for example. Additional work is required to
determine the pertinent kinetics, to develop models necessary for optimization, and to evaluate the total
environmental impact of the products.
ACKNOWLEDGEMENTS
The authors acknowledge financial support of this research by NEDO, a Japanese government agency, under
their cooperative international program on energy and environmental research. They also acknowledge the
sponsorship by the U.S. Environmental Protection Agency under Cooperative Agreement 826317 and Grant R
825389-01, and by the U.S. Department of Energy under Contract DE-AC22-95PC95101. Author Linak
acknowledges support under EPA Internal Grant OMIS-2529. The authors acknowledge the assistance
rendered by Ichiro Naruse of Toyohashi University, Japan, by I. Olmez of MIT, Cambridge, USA, for
performing the neutron activation analysis, and by Jeffrey Ryan, of EPA, for the mercury and chromium
analyses The authors also acknowledge the assistance of the many undergraduate students who worked in the
University of Arizona combustion group during the performance of this research. The research described in this
paper has been reviewed by the Air Pollution Prevention and Control Division, U.S. Environmental Protection
Agency, and approved for publication. The contents of this paper should not be construed to represent Agency
policy, nor does mention of trade names or commercial products constitute endorsement or recommendation for
use.
REFERENCES
1. Linak, W.P., and Wendt, J.O.L., Prog. Energy Combust. Sci., 19:145-185 (1993)
2. Abbas, T., Costen, P.G., and Lockwood, F.C., Twenty-Sixth (International) Symposium on Combustion,
pp. 3041-3058, The Combustion Institute, Pittsburgh, PA (1996)
3. Wendt, J.O.L., Twenty-Fifth (International) Symposium on Combustion, pp. 277-289, The Combustion
Institute, Pittsburgh, PA (1994)
4. Brown, T.D., Smith, D. N., Hargis, R. A., and O'Dowd W.J., J. AWMA. 49(6), 628-640(1999)
5, Linak, W.P., Srivastava, R.K., and Wendt, J.O.L., Combust. Sci. and Technol.. 101, 7-27(1994)
6. Linak, W.P., and Wendt, J.O.L., Combust. Sci. and Technol.. 134(1-6), 291-314 (1998)
7, Davis, S.B., and Wendt, J.O.L., "Quantitative Analysis of High Temperature Toxic Metal Sorption Rates
Using Aerosol Fractionation" in press to Journal of Aerosol Science and Technology, November 1998
8. Davis, S.B., Gale, T.K., Wendt, J.O.L., and Linak, W.P., Twenty-Seventh (International) Symposium on
Combustion, pp. 1785-1791, The Combustion Institute, Pittsburgh, PA (1998)
9. Hillamo. R.. and Kauppinen. E.. Aerosol Sci. Technol.. 14: 33-47 (1991)
-------
10, Wendt, J.O.L. and Seames, W.S., "The Partitioning of Arsenic, Selenium, Cadmium, and Cesium During
tit
Pulverized Coal Combustion in a 17kW Downflow Combustor," Invited Keynote Paper, 4 ' International
Symposium on Coal Combustion, August 18-21, 1999, Beijing, P.R. China
11, Linak, W.P., Ryan, J.V., and Wendt, J.O.L., Combust. Sci. and Technol.. 116-117, pp. 479-498 (1996)
-------
Nucleation / Condensation/
Coagulation
Submicron Aerosol
Aqueous Solutions
Distillate Oil and
Organometallics
Dissolved,
Inherent,
Reactive Metal
Submicron Inclusions,
Residual Particles,
Secondary Atomization
I Heavy Oil and Coal j
Solid and Municipal Waste
Mineral Excluded
Inclusions
Supermicron (Collectable)
Particles
Inorganic Mixture
Fig 1. Toxic metal partitioning paths
-------
Process qas
Water cooled
isokinetic
sampling
probe
25.4 L/min
Nitrogen quench gas
Berner
low-pressure
impactor
Wettest
meter
Fig 2. Schematic of sampling and LPI size segregation system
-------
150
100 -
50 -
5
E
CL
X
to
o>
ra
en
c
o
c
.0
'ra
-------
5
E
Q.
a.
0)
D>
ra
en
c
o
c
o
c
o
o
c
o
o
.2
01
150
100 -
50 -
150
A 0.5 Res. Sec.
0.5 Res. SBC. Trendllne
2.2 Res. Sec.
2.2 Res. Sec. Trondlme
Pittsb urgh
100 -
50 -
200
150 :
100 ^
50 :
Ohio
1 2
1 6
Illinois #6
12
16
0.5 Res. Sec.
0.0 0,3 0.6 0.9 1.2 1.5
Calcium Concentration on Stage x (wt%)
Fig 4. Selenium/calcium correlation for three U.S. coals
-------
SO2 @ burner
SR = 0,8
SO2 @ burner
SR= 1.2
Chadded @ -1000K
SR=0.8
Chadded @ -1000K
SR=1.2
Chadded @ burner
SR=0.8
CI2 added @ burner
SR=1.2
Baseline
SR=0.8
Baseline
SR=1.2
K
y/////
-,H
100 80 60 40 20 0 20 40 60 80 100
Oxidized Elemental
Percentage of Hg reported as
Fig 5. Mercury speciation - doped gas flames
-------
1 10
Aerodynamic Diameter [pm]
CO
CO CO
° te
10%
5%
1
0% O
10% ]g
o
-
5% ~
O
0%
Fig 6. Chromium wastes: effects of inlet speciation [chromium(VI) vs.
chromium(ni)] on outlet PSD and speciation
-------
0.60
0.55
050
0.45
0.40
0.35
^ 0.30
^ 02S
S °-2°
ffl 0.15
IU
0.10
0.05
0.00
-O- FtCMy
-O-Rjwflhkaoilrite
A Fbwttikaolirtte [replicate]
-- Aluirina
10
Aerodynanric Oamster farri)
Fig 7. Lead capture by kaolinite in a single metal, sorbent
system.
-------
Q1 1
Aerodynamic Damper (nm)
Fig 8.
Lack of lead capture by lime in a single metal,
sorbent system.
CdOnly
Cd with kadinile
Cd with Kaoiinite [repiicale]
Alumina
Alumina f replicate!
1
Aerodynamic Diameter (urn)
Fig 9. Cadmium capture by kaolinite in a single
metal, single sorbent system.
-O- CO Only
-D- CdwiltiLime
-A- cd rtlh Ume [replteatel
-- Calcium
-A- Catelum freplicalel
Aerodynamic Diameter (n
Fig 10. Cadmium capture by lime in a single metal,
single sorbent system.
-------
Ft with &rterl [radicate]
Alurina
Alumina [replicate]
0.70
0-1 1
/terodynarric Diameter (jam)
-OQl-NDSofbent
-Q- CtiwilhScrtjent
-A- Ctl wiih Sorbent [replicate)
- Alurrina
-A- AJurrina [replicate]
10
Aerodynarric D'ameter (urn)
Fig 11. Lead capture by kaolinite in a multi-metal
lead-cadmium, single sorbent system.
Fig 12. Cadmium capture by kaolinite in the multi-
metal lead-cadmium, single sorbent system.
PROTECTED UNDER INTERNATIONAL COPYRIGHT
ALL RIGHTS RESERVED
NATIONAL TECHNICAL INFORMATION SERVICE
U.S. DEPARTMENT OF COMMERCE
Reproduced from
best available copy.
-------
NRMRL-RTP-P-461
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completin'
1. REPORT NO.
EPA 600/A-00/001
2.
3. B
4. TITLE AND SUBTITLE
5. REPORT DATE
Environmental Control of Toxic Metal Air Emissions
from the Combustion of Coal and Wastes
6. PERFORMING ORGANIZATION CODE
PB2000-102971
7'AUTHOR(si J. C. L. Wendt, S.B.Davis, T.K.Gale, and
W. S. Seames (Univ. of AZ); and W. P. Linak (EPA)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Department of Chemical and Environmental Engrg.
University of Arizona
Tucson, Arizona 85721
11. CONTRACT/GRANT NO.
CR826317, R825389-01
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
Published paper;
14. SPONSORING AGENCY CODE
EPA/600/13
".SUPPLEMENTARY NOTES
project officer is William P. Linak, Mail Drop 65, 919 /
541-5792. For presentation at Symposium on Energy Engineering in the 21st Century,
Hong Kong, P. R. China, 1/9-13700. *
16'ABSTRACTThe paper is concerned with the partitioning of toxic metals (e.g. , arsenic,
selenium, mercury, chromium, lead, and cadmium) during combustion, and with the
mitigation of their effect on the environment using high-temperature sorbents. The
paper is divided into three parts: (1) the partitioning of arsenic and selenium during
coal combustion in a 17-kW laboratory down-fired furnace is discussed, and appro-
priate mechanisms identified; (2) the speciation of mercury and chromium during
combustion is addressed, through special experiments in an 82-kW refractory-lined
combustor; and (3) experimental results are presented on the sorption of individual
and multiple metals on sorbents. The sorbents, kaolinite and lime, were injected
directly into flue gas containing lead and cadmium, which had vaporized in the main
flame. Results suggest that toxic metals from coal and waste combustion can inter-
act with lime «or kaolinite sorbents and that, for some multiple metal mix-
tures, designer sorbents containing calcium, aluminum, and silicon might be useful
to capture them and render them environmentally benign.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IOENTIFIEHS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Coal
Wastes
Combustion
Metals
Toxicity
Emission
Sorbents
Incinerators
Pollution Control
Stationary Sources
13B
2 ID
14G
2 IB
11F.07B
06T
11G
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
15
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
E!PA Form 2220-1 (9-73)
------- |