EPA/600/A-97/114
Research on Emissions and Mitigation of POPs from Combustion Sources
C. W. Lee, P. M. Lemieux, B. K. Gullett, J. V. Ryan, and J. D, Kilgroe
Air Pollution Technology Branch, Air Pollution Prevention and Control Division, National Risk
Management Research Laboratory, U.S. Environmental Protection Agency, Research Triangle
Park, NC 27711, US A
ABSTRACT
The environmental consequences of persistent organic pollutants (POPs) are of increasing
concern due to the serious health effects on animals and humans including reproduction,
development, and immunological function. Several major classes of POPs, including polycyclic
aromatic hydrocarbons (PAHs), chlorobenzenes, chlorinated dioxins, and chlorinated furans,
have been identified as products of incomplete combustion (PICs) produced in trace levels in
combustion systems, A wide variety of combustion processes, ranging from power plants,
industrial boilers, industrial furnaces, and incinerators, to home heating devices, are believed to
be potential sources of POPs. Full-scale combustion facilities can be significant sources of POPs
due to the large mass flow of flue gas released from a plant. Total emissions of POPs from small
combustion devices, such as wood stoves and residential oil furnaces, can also be significant due
to the large numbers of existing units near high population areas. It becomes increasingly
important to understand the formation of POPs from different combustion processes to identify
sources of POPs and to develop strategies for their prevention and mitigation.
Research on POP emissions from combustion sources conducted by EPA is largely
driven by the need for regulating the emissions of hazardous air pollutants as required by Title HI
of the 1990 Clean Air Act Amendments and by the Resource Conservation and Recovery Act,
This paper provides a summary of EPA's research on emissions and control of POPs from
combustion sources with emphasis on source characterization and measurement, formation and
destruction mechanisms, formation prevention, and flue gas cleaning. Laboratory experiments
conducted to examine the PAH emissions from a wide variety of combustion processes, ranging
from pulverized coal utility boilers to wood stoves, have shown that they exhibit widely different
emission characteristics. Waste incineration research conducted by the National Risk
Management Research Laboratory, Air Pollution Prevention and Control Division
(NRMRL/APPCD) has also shown that complex mechanisms, including physical mixing and
chemical kinetics, are involved in the formation of chlorinated PICs.* Research has also
indicated that the formation of ultra-trace levels of chlorinated-dioxins and -furans in
combustion/incineration processes includes the complex interaction of several factors including
temperature, chlorine content, and catalyst. The beneficial effect of sulfur and sorbents for
dioxin formation prevention is demonstrated. This Laboratory's effort to develop and evaluate
state-of-the-art technologies for on-line measurements of PAHs, volatile PICs, dioxins, and
furans is also discussed. The promising potential of applying artificial-intelligence-based control
systems for improving combustion processes operating conditions as a POP prevention approach
is demonstrated.
1.0. INTRODUCTION
The wide distribution of persistent organic pollutants (FUFSJ m toe environment nas
become a global issue of growing concern. International intentions to control such substances
are expressed in the Washington Declaration made at the conclusion of a United Nations
NRMRL/APPCD was previously named the Air and Energy Engineering Research Laboratory (AEERL),
-------
Environmental Programme conference attended by environment ministers of 108 countries held
in Washington, DC, during November 1995 [1]. POPs are long-lived organic compounds which
survive long distance migration in the global environment and become concentrated as they
move through the food chain with serious health effects on animals and humans including
reproduction, development, and immunological function. Much of the concern on POPs involves
12 chemicals or chemical classes, including polychlorinated biphenyls (PCBs), polychlorinated
dibenzo-p-dioxins (PCDDs) and -furans (PCDFs), polycyclic aromatic hydrocarbons (PAHs),
and pesticides such as DDT and chlordane.
Although the production and use of many of the anthropogenic chemicals which are
considered as POPs, such as PCBs and dichlorodiphenyltrichloroethane (DDT), are banned in
most developed countries, they are still widely distributed and used in developing countries.
However, there are also POPs which are produced unintentionally as byproducts from
combustion processes. Several major classes of POPs, including PAHs, chlorobenzenes,
PCDDs, and PCDFs, have been identified as products of incomplete combustion (PICs) emitted
at trace levels from various combustion systems. The development of control and mitigation
strategies for POPs requires a better understanding of the major sources of these pollutants. It is
very important to identify the industrial and residential combustion processes which may also be
the potential sources of POPs. This paper summarizes EPA's research on POP emissions and
control from different combustion sources.
Research on POP emissions from combustion sources conducted by the Air Pollution
Prevention and Control Division (APPCD) of EPA's National Risk Management Research
Laboratory (NRMRL) is motivated by the need for regulating the emissions of air toxic
pollutants as required by Title III of the 1990 Clean Air Act Amendments (CAAAs) [2] and by
the Resource Conservation and Recovery Act (RCRA) [3], Title HI of the CAAAs lists 189
compounds and compound classes, and application of maximum achievable control technology
(MACT) is required by any non-utility source that emits over 9,091 kg/year (10 t/year) of any
one air toxic pollutant, or 22,727 kg/year (25 t/year) of total air toxic pollutants. PAHs,
chlorobenzenes, and PCDDs/PCDFs are among the classes of the 189 regulated air toxic
pollutants which are also considered as important POPs. In addition, revised RCRA standards
were proposed for hazardous waste combustors (hazardous waste burning incinerators, cement
kilns, and lightweight aggregate kilns) in April 1996 [4]. These proposed standards would limit
the PCDD/PCDF emissions to 0.2 ng International -Toxic Equivalency (I-TEQ)/dscm.
Characterization of ah* toxic emissions from a wide range of sources and development of their
control strategies have become an important part of the Agency's CAAAs implementation
efforts. Research on dioxin emission prevention and control from hazardous waste combustion
is very important to the Agency's RCRA regulation development.
2.0. POP EMISSIONS
Characterization of organic air toxic emissions from combustion sources in general is
much more difficult compared to that from chemical production facilities which involve
predominately fugitive emissions of individual product compounds. Air toxic emissions
generated by combustion, including the three major classes, of POPs mentioned above, are
typically contained in a complex mixture of organic compounds at trace levels. Because of large
volumes of flue gas produced during the combustion process, even trace levels (ppm) of air toxic
compounds in the flue gas can exceed the limits specified under Title HI of the CAAAs, Due to
a wide variety of combustion boilers that exist for utility, industrial, commercial, and
institutional applications across the country, and because of the relatively small amount of
information available characterizing organic air toxic emissions from these sources, research has
been undertaken recently by EPA to provide such information.
-------
2.1 PAH Emissions
PAHs are among the most common PICs found in combustion flue gases at trace levels,
PAHs will likely be produced as a result of localized improper mixing of fuel and combustion air
in a large size industrial combustor. A study was conducted by EPA to characterize organic air
toxics emissions in the flue gases from the combustion of pulverized coal [5]. A small-scale
combustor was operated under different conditions to simulate high excess air firing and nitrogen
oxide (NOX) controls by combustion modifications of a utility boiler. Only a few organic air
toxics were found above the detection limits, with naphthalene as the only PAH identified from
the tests. Results of the tests also indicate that total air toxics emissions from a large coal-fired
utility plant are not likely to increase as a result of installation of combustion modifications for
NOX control. In other research performed recently by EPA to determine the emissions levels of
organic air toxics from combustion of various grades of oils, tests were conducted on a
commercial fire tube package boiler running on fuel oil [6]. It was found that carbonyls
dominated the trace organic emissions, and PAHs constituted only minor components of the
emissions.
The emissions of PAHs have been identified as the major pollutants emitted from
residential combustion units in studies conducted by EPA. Large numbers of such units,
including wood stoves, oil combustors, and coal-fired furnaces, are operating in urban/suburban
regions, and their emissions may have serious impacts on indoor and ambient air quality. With
collaboration of EPA's National Health and Environmental Effects Research Laboratory as part
of the Integrated Air Cancer Project, we have conducted studies to characterize the organic
emissions from different residential combustion devices. High levels of PAHs (0.7 g/kg wood
burned) have been observed in EPA's wood stove research [7]. Some operating variables,, such
as wood species, stove type, and altitude of the stove, were found to have a strong effect on PAH
emissions, while burn rate has very little effect on PAH emissions. Pine was found to produce
more PAHs than oak; conventional stoves showed higher PAH emission rates than those from a
catalytic stove; and lowering the operating altitude of the stove from 825 to 90 m caused an
increase in PAH emission rates. It was also found from this study that emissions from burning
oak are less mutagenic than those from pine, and emissions from the catalytic stove were more
mutagenic than those from conventional stoves.
The organics emitted from residential heating furnaces have been found to be closely
related to the chemical structure of the fuel burned in EPA's home heating fuels studies [8]. This
is particularly true for wood and coal which contain mainly polymeric chemical structures where
thermal cleaving during combustion accounts for a significant portion of the emitted organics. It
was found that combustion of wood, which contains largely lignin and cellulose, produces
mainly oxyaromatics, and naphthalene accounts for almost all the PAHs in its emissions. The
burning of coal, which contains fused ring structures, produces emissions with three-, four-, and
five-ring PAHs, a class that includes benzo(a)pyrene and other known carcinogens. In the case
of oil, the organic emissions contain mainly the unburned droplets of the oil itself, with
substituted naphthalenes dominating the PAH emissions. Natural gas was found to be a clean-
burning home heating fuel. Only a few PAHs with levels at least 100 times less than those
emitted from wood stoves were identified in natural gas furnace emissions [9].
PAHs were identified as the most common pollutants produced from open burning of a
wide variety of waste materials. Open burning is still widely practiced as a waste management
method for several types of wastes, ranging from agricultural plastics to land-clearing debris.
Because of concerns of their potential impacts on ambient air quality, EPA has undertaken a
series of experimental studies to characterize the emissions from simulated open burning of
waste materials. It was found that emissions produced from simulated open burning of
agricultural plastic contain a complex mixture predominated by high molecular weight PAHs,
only a minor fraction of which can be identified [10]. No mutagenic effects were found for the
-------
emissions as a whole; however, organic extracts of the particulate emissions contained
rautagenicity comparable to that measured from wood stoves. In response to public concern over
health hazards caused by the growing incidence of tire fires, EPA has conducted research to
characterize the emissions generated from simulated scrap tire fires. It was found that the
emissions contained mainly PAHs, which ranged from 10 to 50 g/kg of tire material burned, and
alkyl-substituted PAHs were the predominant PAHs identified [11], EPA has also conducted a
study to characterize the emissions produced from open burning of land-clearing debris, in order
to assess the environmental risk associated with such land-clearing practice. Substantial
emissions of a large number of pollutants including carbon monoxide (CO), fine particulate
matter, volatile organic compounds (VOCs), and high molecular weight organics dominated by
PAHs were observed from simulated open burning experiments [12]. Only 14 of the PAHs were
identified, with a majority of PAHs only tentatively identified through searches of mass spectral
libraries.
2.2. Formation of Chlorinated PICs
We are continuing our research to understand the formation and control mechanisms
associated with toxic organic pollutants from waste incineration, with emphasis placed on the
emissions of chlorinated PICs. Chlorocompounds are commonly contained in waste streams,
which are difficult to destroy thermally during incineration with the resulting formation of
chlorinated PICs. Many of the chlorinated PICs, such as chlorobenzenes, are toxic and also
listed as a major class of POPs. Our research has focused mainly on PIC formation and control
from rotary kiln incinerators. This particular incinerator design is very versatile and is widely
used for treating industrial wastes in the form of liquids, sludges, or solids.
The secondary combustion chamber (SCC) is an important piece of control equipment for
rotary kiln incinerators [13-14], The SCC should be capable of destroying any unburned organic
material that exits the primary combustion chamber due to rogue droplets, transients, quenching,
or incomplete mixing. SCCs are also commonly used to combust liquid wastes that have high
heating values. Design criteria in the past have been mostly limited to a time-temperature
requirement, such as 2 s at 1000 °C (1800 °F). Although a time-temperature requirement is not
written into the hazardous waste incinerator regulations as defined in the Resource Conservation
and Recovery Act (RCRA), it appears to have been adopted as a criterion by regulators and the
regulated community alike. A disadvantage of this "apparent" policy is that mixing, known to be
of critical importance in incineration systems [15], is largely ignored; and no economic
incentives exist with which to improve afterburner designs, given that any new design would
likely require a certain time-temperature profile before being allowed to be installed, even if such
a design could meet required emissions limits with a much more compact configuration. .
The emissions that the SCC must deal with generally result from some sort of system
failure in the primary chamber, since steady-state operation of the primary chamber generally
eliminates the need for an SCC. Liquid injection incinerators, for example, typically don't
require an SCC. The failure modes that can cause elevated levels of organic compounds to enter
the SCC include mixing failures, such as those caused by: poor microscale mixing intensities or
poor macroscale mixing; poor atomization; flow stratification; batch charging and depletion of
oxygen in the primary chamber; and reaction quenching, such as that caused by unburned
material entering cold regions of the combustion device, or by cold walls. Rotary kilns in
particular exhibit high levels of flow stratification [16-18], and typically have some of their
waste feed fed in batches and, as such, generally employ an SCC.
Part of the reason that a time-temperature requirement is used as a common SCC design
criterion is that the effects of turbulence and complex chemical kinetics are not understood well
enough to incorporate their use into the permitting process. It is very important, however, to
work toward gaining an understanding of kinetics and mixing in incinerators, since it is possible
-------
to have excessive levels of PICs even after having successfully met the necessary time-
temperature requirement. EPA, in cooperation with the New Jersey Institute of Technology
(NJIT) and Massachusetts Institute of Technology (MIT), has been performing research on a
pilot-scale rotary kiln incinerator simulator (RKIS) to complement laboratory-scale research
being performed at both of the previously mentioned academic institutions, with the ultimate
goal of furthering the state-of-the-art of SCC design by incorporating gas-phase mixing and
kinetic considerations into the design criteria, particularly in regards to chlorocarbon combustion.
Initial pilot-scale experiments have consisted of system characterization tests.
In order to incorporate gas-phase mixing and kinetic phenomena into afterburner design,
it is necessary to achieve several goals, including:
1) Development of reaction pathways and kinetic data for combustion of the principal
organic hazardous constituents (POHCs) present in the waste, along with possible mechanisms of
formation of PICs from POHC decomposition products. Although mechanistic information is
not available for complex compounds, mechanisms do exist for Cl and C2 chlorocarbon
combustion [19]. Initial tests focused on combustion of carbon tetraehloride (CCLj) and
methylene chloride (Cl^Cla), compounds for which kinetic mechanisms exist. Combustion of
CH2C12 results in levels of 1,2 dichlorobenzene and monochlorobenzene much higher than those
found from CC14 combustion. It may be possible that CH2C12 can readily form chlorinated
intermediate structures that are ring-growth precursors, resulting in direct formation of
monochlorobenzene rather than from chlorination of benzene. Combustion of CH2C12 produced
higher quantities of identified PICs than combustion of CC14, particularly during fuel-lean
combustion. Although CCU may be useful as a POHC due to its high thermal stability, and
provide a useful measure of a system's ability to meet the required 99.99 % destruction and
removal efficiency (DRE), it may not challenge an incinerator's ability to produce or destroy
PICs.
2) Development of models that take into account macromixing and micromixing phenomena
to aid in the scale-up of results from very small-scale experiments to pilot and full-scale systems.
Kinetics and thermodynamics alone cannot account for emissions of PICs from incinerators.
Mixing must eventually be considered [20]. Initial tests have attempted to characterize the
macromixing in the EPA's RKIS SCC using sulfur dioxide (862) as a tracer. By measuring
residence time distributions (RTDs) in various portions of the SCC, it is possible to determine
the unknown reactor volumes in a series of ideal reactors. These reactor volumes can then be
modeled numerically using detailed reaction mechanisms [21].
3) Development of techniques to measure trace organic species or surrogates for trace
organic species in the field, given that many of the advanced diagnostics available in a laboratory
setting cannot easily be transferred to a field application. Semi-continuous measurement of key
organic compounds can potentially be used to characterize the overall destruction of all
hazardous trace organics of concern. Since POPs are frequently in the semi-volatile range of
boiling points (>150 °C), and thus more difficult to measure using existing monitoring
methodologies, it is useful to find a surrogate compound for the POPs that is more easily
measured than the POP of interest.
2.3. Dioxins
On-going research has been investigating the fundamental mechanisms behind formation
of PCDDs/PCDFs, collectively termed "dioxins." Dioxins are formed in the post-combustion
region from components of chlorine (Cl), unburned organics, and catalytic surfaces. A complex
interaction of surface-induced catalysis, organic ring structure formation, and chlorination
countered by destruction/dechlorination reactions can result in the formation and emission of
dioxins and other chloro-organics The synthesis reactions that lead to the formation of dioxins
-------
occur at temperatures ranging from approximately 200 to 600°C [22,23]. They may be
catalyzed by combustor deposits, and by entrained and collected fly ash. These deposits can be
composed of fly ash or sooty materials containing catalytic metals such as copper [24,25]. With
appropriate fly ash properties and operating temperatures, air pollution control devices such as
electrostatic precipitates and fabric filters can act as chemical reactors that form dioxins and
other chloro-organics[25].
While commonly associated with waste combustion, dioxin formation can also occur
from burning coal, oil, and wood. The mechanisms of formation are being studied at EPA to
enable source-specific predictions and to develop pollution prevention strategies that discourage
or prevent dioxin formation from occurring. Work at EPA has coupled mechanistic aspects
related to the effect of different types of metal catalysts [26], the ability of Cl to chlorinate
aromatic ring structures [27], and the role of catalysts in formation of the biaryl structure [28].
The effect of these interactive mechanistic parameters, along with combustor operating
parameters, has been examined for their relationship with dioxin formation [29-30]. This work
has shown on a combustor-specific basis that reduction of hydrogen chloride/chlorine (HC1/C12)
concentration, completeness of combustion, and/or increases in quench rate can reduce formation
of dioxin.
Recent work has focused on the effect of sorbent injection technologies to remove Cl and
thereby prevent formation of dioxins [31]. Research programs on two pilot-scale combustors
have shown that injection of calcium (Ca)-based sorbents at moderately high temperatures has
the effect of preventing formation of dioxins [29,32]. This is likely to hold for all systems in
which formation is Cl limited. This technology is the subject of two U.S. patents [33, 34] and is
currently undergoing field demonstration in the U.S.
Another preventive strategy derived from mechanistic studies is the suppression of dioxin
formation by the presence of sulfur (S) as SO2, The effect of S(>2 may relate to its effect on an
important, catalytic chlorination reaction and biaryl synthesis [35] and/or its effect on gas-phase
reactions with chlorinating compounds [36]. This SO2 effect has been shown on a large pilot
scale and is currently undergoing field demonstration. Results to date [30, 32] suggest
significant (90%) suppression of dioxin formation across all congener classes at S/C1 ratios >1/1.
Upcoming work at EPA will undertake a mobile sampling program for dioxin emissions
from heavy duty diesel vehicles. This work, part of EPA's Dioxin Reassessment effort, will.
provide important information in the effort to quantify dioxin sources in the environment This
effort is significant in that U.S. mobile sources, with the exception of a recently completed tunnel
study (still in draft form), are entirely uncharacterized as potential dioxin sources. Further, it is
possibly the first mobile sampling effort for diesel engine dioxins.
Another recently initiated program will be using isotopically labeled fuels to study the
mechanisms and rates of formation. This work is geared toward discerning the critical pathways
of dioxin formation and determining rate information for development of a global reaction
model. Expected results will discern the importance of incomplete combustion (either gas phase
or carbonaceous, solid phase) on supplying organic dioxin precursors and, thus, will be relevant
to a wide variety of fuel types and combustors. This program is being designed with the
assistance and advice of a team of international researchers and reviewers.
3.0. ON-LINE POP MEASUREMENT
Part of the importance of determining what POPs are released from combustion systems
is the measurement of those POPs. This creates a major technical challenge since many of the
compounds of interest are present in very small quantities, plus are normally in the semivolatile
boiling point range (>150 °C), which partitions the organic compounds of interest between the
-------
gas phase and condensed phase (usually bound on particulate matter). This partition makes
continuous measurement difficult. A useful approach to circumvent this limitation is to measure
another, more easy-to-measure compound, that is a precursor to the POP of interest, or at least is
a well-correlated indicator of not only the presence/absence of the POP, but also of the relative
concentration of the POP in the stack. This concept of using a surrogate indicator has been used
in current regulations of municipal waste combustors (MWCs). CO, for example, has been used
as a useful surrogate for PCDDs/PCDFs in MWC systems; however, its usefulness as a surrogate
indicator breaks down in a system that is very well operated.
A potentially more useful approach is to measure a volatile organic precursor to the
semivolatile POPs of concern. In a recently completed test program [37] at the U.S. EPA
Incineration Research Facility (IRF) in Jefferson, Arkansas, several continuous emission
monitors (CEMs) for measuring trace quantities of various organic and inorganic pollutants were
tested. One of the CEMs tested was a dual-detector on-line gas chromatograph (GC) system
developed by EPA/APPCD in-house. This system consisted of a sample delivery system, a
sample concentrator, and a GC equipped with a flame ionization detector and an electron capture
detector [38], The IRF's rotary kiln incineration system was operated at conditions indicative of
normal incinerator operation, while injecting varying concentrations of 10 VOCs that are usually
found in incinerator stack gases as PICs, Target VOC concentrations included low level (i-2
|Hg/m3), medium level (10-20 Jig/m3), and high level (160-240 jig/m3). Tests were performed to
compare CEM results to the EPA's standard reference methods using a relative accuracy test
audit protocol [39]. The EPA/APPCD on-line GC performed successfully for most compounds
at all concentration levels. Instrument sensitivities were sufficient to measure all 10 compounds
at levels typically found in well-operated systems.
The high molecular weight PAHs are predominantly adsorbed on soot (carbon aerosols)
due to their low vapor pressure at stack temperatures. Currently, PAHs are measured using a
sampling train (EPA Modified Method 5, MM5) followed by solvent extraction and analysis
using GC and mass spectrometry (MS) [40 - 42]. The measurement is time consuming, and no
real-time data can be obtained from such captive sampling techniques. EPA recently conducted
tests to evaluate the application of a photoelectric aerosol sensor for on-line measurement of
particle-bound PAHs in combustion flue gas. The PAH monitor works on the principle of
photoionization of carbon aerosols. After being exposed to the ultraviolet (UV) light of the
monitor, the carbon aerosols which have PAH molecules adsorbed on the surface pass through
the monitor, then emit electrons, and the resulting electric current is proportional to the
partieulate-bound PAHs. Initial tests of the PAH monitor for measuring emissions from burning
tire-derived fuel in EPA's pilot-scale RKIS indicated that the monitor appeared to track transient
operation of the combustor well [43], The PAH monitor was also evaluated during a recently
completed test program, which gave excellent relative accuracy for measuring the three selected
PAHs (naphthalene, phenanmrene, and pyrene) at an intermediate concentration level [37],
Operating problems caused by the high moisture content of the combustion flue gas were
experienced and the monitor's manufacturer is modifying the design of the instrument's moisture
removal system.
A recent international research program has assisted in the development of Jet-REMPI™
(resonance-enhanced multiphoton ionization) as an analytical technique for monitoring
extremely low concentrations of chlorinated dioxins. The DLR Jet-REMPI apparatus was tested
in the EPA laboratories for its ability to detect chlorinated dioxins. This work [44] has detected
spectra for a dichlorodibenzodioxin isomer and determined a detection limit of less than 30
ng/dscm. Identification work for more highly chlorinated isomers is currently underway. It is
likely that suitable correlations between the lower chlorinated dioxin isomers and the more toxic,
higher chlorinated isomers can be established, enabling Jet-REMPI to be used as a correlative
monitor for dioxin toxic equivalency (TEQ) emissions.
-------
4.0. APPLICATIONS OF AI FOR COMBUSTION CONTROL
The objective of this research [45] was to apply fuzzy logic artificial intelligence (AI) to
control combustion systems, in particular the pilot-scale RKIS and its secondary combustion
chamber (SCC) mentioned above. The mechanistic details of system response in a system such
as the RKIS are complex and are not usually known a priori. However, a relatively small number
of generic control rules based on operating experience can be linguistically stated and then
translated into a fuzzy logic system for automatic control. The purpose of the control system in
this case was to reduce emissions of PICs from the RKIS, when significant transients were
artificially imposed on the system by spraying liquid surrogate wastes into the RKIS based on a
time-based algorithm.
Rotary kilns have the advantage of being flexible enough to handle a wide variety of
waste streams in a wide variety of forms, including flammable liquids, aqueous streams, sludges,
and whole drums. When a drum containing volatile material is fed into a rotary kiln, it ruptures
and releases its contents into the hot kiln environment in a transient event that occurs over a short
period of time. If the instantaneous stoichiometric requirements to combust the volatile waste are
greater than the amount of oxygen being added to the kiln through the main burners or auxiliary
air sources, then a plug of unburned material, called a transient "puff leaves the kiln and must be
dealt with by the SCC. The transient puffs include large quantities of soot, CO, and organic
compounds of both a volatile and semivolatile nature. Past experiments have shown high
quantities of PAHs contained in transient puffs (46).
One possible control option is to equip the SCC with a system that can inject additional
oxidizer in response to a puff leaving the kiln. This option has been shown to dramatically
decrease the emissions of PAHs and other PICs (35). If air is used as the oxidizer, then the
downstream equipment, such as baghouses or scrubbers, must be sized to handle the increased
volume of flue gases, based on maximum flow rates. If oxygen (C»2) is used as the oxidizer, then
the downstream equipment can be sized smaller; however, significant costs are associated with
the use of pure Oa- A tradeoff is realized between increased capital costs and increased operating
costs. If a means is available to maximize the efficiency of 62 usage, then O2 injection will
become more economically attractive.
A series of experiments were performed to evaluate the effectiveness of fuzzy logic
control schemes by injection of pure 62 into the SCC's afterburner. A statistically designed test
matrix was repeated for each of four control schemes: no control, feedback control (based on
stack 02), and two fuzzy logic control schemes based on CO and total hydrocarbon (THC)
emissions at the kiln exit. Results were evaluated using a combined performance indicator that
utilizes stoicbiometrically weighted integrated emissions of CO, THC, and soot at the inlet and
outlet of the SCC. The following conclusions were observed:
• The speed of analyzer response is a significant hindrance to the effectiveness of any sort
of process control for reduction of transients from incinerators.
« The feedback control tended to open the Oj valve all the way in order to attempt to
restore the stack 02 to a level near the set point.
• The fuzzy logic control scheme was more effective than feedback in controlling transients
of short duration.
* The fuzzy logic control scheme was significantly more efficient in its use of the injected
oxygen, particularly where the transients were small.
-------
5.0. CONCLUSIONS
Major classes of POPs including PAHs, chlorobenzenes, and chlorinated dioxins and
furans can be emitted from various combustion processes as unintentional and unwanted
byproducts. Research has been conducted by EPA to study POP emissions from combustion
sources, with emphasis on understanding the fundamental mechanisms which lead to POP
emissions and identifying mitigation strategies for such emissions. Our research found that good
combustion conditions achieved in well-operated industrial units in general prevent significant
POP emissions. POP emissions result from complex interactions of a wide variety of physical
and chemical mechanisms occurring in combustion systems. POP emission potentials strongly
depend on the design and operating characteristics of the combustion units. The presence of
fuel/waste components which contain difficult-to-destroy chemical structures in fuels/wastes also
plays an important role for promoting POP emissions. Significant progress has also been made
in our research on developing technologies for continuous monitoring and reducing POP
emissions from combustion sources.
REFERENCES
1. F. Wania and D. Mackay, Env. Sci. Techno!., 30 (9), 190A-196A, 1996.
2. Public Law 101-549, Clean Air Act Amendments of 1990, November 15, 1990.
3. 40 CFR Part 60, "Hazardous Waste Combustors; Revised Standards; Proposed Rule," Part II,
U.S. Environmental Protection Agency, April 19,1996.
4. 40 CFR Part 60, et al, "Hazardous Waste Combustors, Revised Standards, Proposed Rule",
Part n, U. S. Environmental Protection Agency, April 10,1996.
5. C.A. Miller, R.K. Srivastava, and J.V. Ryan, Environ. Sci. Techno!., 28 (1994) 1150.
6. C.A. Miller, J.V. Ryan, and T. Lombardo, J. Air & Waste Manage. Assoc., 46 (1996) 742.
7. R.C. MeCrillis, R.R. Watts, and S.H. Warren, J. Air & Waste Manage. Assoc. 42 (1992) 691.
8. R.S. Steiber, J. Air & Waste Manage. Assoc., 43 (1993) 859.
9. J.V. Ryan and R.C. McCrillis, "Analysis of Emissions from Residential Natural Gas
Furnaces," in Proceedings: 87th Annual Air and Waste Management Association Meeting, Paper
No. 94-WA75A.04, Cincinnati, OH, June 1994.
10. W.P. Linak, J.V. Ryan, E. Perry, R.W. Williams, and D.M. Demarini, J. Air & Waste
Manage. Assoc. 39 (1989) 836.
11. P.M. Lemieux and J.V. Ryan, J. Air & Waste Manage. Assoc, 43 (1993) 1106.
12. C.C. Lutes, and P.H. Kariher, "Evaluation of Emissions from the Open Burning of Land-
Clearing Debris," Final Report, EPA-60Q/R-96-128 (NTIS PB97-115356), October 1996.
13. P.M. Lemieux, W.P. Linak, J.A. McSorley, J.O.L. Wendt, and J.E. Dunn, Combust. Sci. &
Technol., 74 (1990) 311.
14. R.W. Rolke, R.D. Hawthorne, C.R. Garbett, E.R. Slater, T.T. Phillips, and G.D. Towell,
"Afterburner Systems Study," EPA-R2-72-062 (NTIS PB-212-560), August 1972.
-------
15. J.O.L. Wendt, W.P. Linak, and P,M. Lemieux, Hazardous Waste & Hazardous Materials, 7
(1) (1990) 41.
16. V.A. Cundy, T.W. Lester, A.M. Sterling, A.N, Montestrue, J.S. Morse, C.B. Leger, and S.
Acharya, J. Air Poll. Cont. Assoc,, 39 (1989) 944.
17. V.A. Cundy, T.W. Lester, A.M. Sterling, A.N. Montestrue, J.S. Morse, C.B. Leger, and S.
Acharya, J. Air Poll Cont. Assoc., 39 (1989) 1073.
18. V.C. Cundy, A.M. Sterling, T.W. Lester, A.L. jakway, C.B. Leger, C. Lu, A.N. Montestruc,
and R.B. Conway, Environ. Sci. Technol., 25(2) (1991) 223.
19. W.P. Ho and J.W. Bozelli, "Validation of a Mechanism for Use in Modeling CH2C12 and/or
CHsCl Combustion and Pyrolysis," Proceedings of the Twenty Fourth Symposium
(International) on Combustion, Sidney, Australia, (1992) 743.
20. J. Brouwer, G. Sacchi, J.P. Longwell, and A.F. Sarofim, Combust. Sci. Technol., 101(1-6)
(1994)361.
21. C. Bass, R. Barat, G. Sacchi, and P.M. Lemieux, "Fundamental Studies on the
Characterization and Failure Modes of Incinerator Afterburners," Proceedings of the 1995
International Incineration Conference, University of California, Irvine, CA, (1995) 223.
22. R. Addink and K. Olie, Env. Sci. & Technol., 29 (6) (1995) 1425.
23. S.B. Gorishi, and E.R. Altwicker, Env. Sci. and Technol., 29 (5) (1995) 1156.
24. C.W. Lee, J.F. Ryan, R.E. Hall, G.D. Kryder, and B.R. Springsteen, Combust. Sci and
Tech., 116-117 (1996) 455.
25. J.D. Kilgroe, Journal of Hazardous Materials, 47 (1996),163.
26. B.K. Gullett, K.R. Brace, andL.O. Beach, Chemosphere, 20 (1990) 1945.
27. K.R. Bruce, L.O. Beach, and B.K. Gullett, Waste Management, 11 (1991) 97.
28. B.K. Gullett, K.R. Bruce, and L.O. Beach, Chemosphere, 25 (1992) 1387.
29. B.K. Gullett, P.M. Lemieux, and J.E. Dunn, Environ. Sci. Technol., 28 (1994) 107.
30. B.K. Gullett and K. Raghunathan, Chemosphere, 34:5-7 (1997) 1027. •
31. B.K. Gullett, W. Jozewicz, andL.A. Stefanski, Ind. Eng. Chem., 31(11) (1992) 2437.
32. K. Raghunathan, B.K. Gullett, et ai, "Prevention of PCDD/PCDF Formation by Coal Co-
firing," Proceedings of the Fifth Annual North American Waste to Energy Conference, RTP, NC,
April 22-25, (1997) 779.
33. B.K. Gullett, Reduction of Chlorinated Organics in the Incineration of Wastes, U.S. Patent
No. 5 021 229, (1991).
34. B.K. Gullett, Reduction of Chlorinated Organics in the Incineration of Wastes, U.S. Patent
No. 5 185 134, (1993).
10
-------
35. B.K. Gullett, K.R. Brace, and L.O. Beach, Environ. Sci. Technol, 26 (1992) 1938.
36. K. Raghunathan and B.K. Gullett, Environ. Sci. Technol. 30(6) (1996) 1827.
37. L.R. Waterland, E. Whitworth and M.K. Richards, "Innovative Continuous Emission
Monitors: Results of the EPA/DOE Demonstration Test Program," Proceedings of the 1996
International Conference on Incineration and Thermal Treatment Technologies, University of
California, Irvine, CA, (1996) 373.
38. P.M. Lemieux, J.V. Ryan, C. Bass, and R. Barat, J. Air & Waste Manage. Assoc.,
Nashville, TN, 46 (1996) 309.
39. 40 CFR Part 266, Appendix IX, Section 2 "Performance Specifications for Continuous
Emission Monitoring Systems."
40. U. S. EPA, "Modified Method 5 Sampling Train," Method 0010, in Test Methods for
Evaluating Solid Wastes, SW-846 (NTIS PB88-239223), Revision 0, September 1986.
41. U. S. EPA, "Extraction of Semivolatile Analytes Collected Using Modified Method 5,"
Method 3542, in Test Methods for Evaluating Solid Wastes, SW-846 (NTIS PB88-239223),
Revision 0, January 1995.
42. U. S. EPA, "Semivolatile Organic Compounds By Gas Chromatography/Mass Spectrometry
(GC/MS): Capillary Column Technique," Method 8270C, in Test Methods for Evaluating Solid
Wastes, SW-846 (NTIS PB88-239223), Revision 3, January 1995.
43. P.M. Lemieux, "Pilot-Scale Evaluation of the Potential for Emissions of Hazardous Air
Pollutants from Combustion of Tire-derived Fuel," Final Report, EPA-600/R-94-070 (NTIS
PB94-169463), May 1994.
44. H. Oser, R. Thanner, H.-H. Grotheer, B.K. Gullett, N. French, and D. Natschke, "On-line
Detection of Dichlorodibenzodioxins by DLR Jet-REMPI," Presented at: Japanese Flame Days
"97, Japanese Flame Research Committee, May 16-17, 1997, Osaka, Japan.
45. P.M. Lemieux, C.A. Miller, KJ. Fritsky, and PJ. Chappell, "Development of an Artifieial-
Intelligence-Based System to Control Transient Emissions from Secondary Combustion
Chambers of Hazardous Waste Incinerators," Proceedings of the 1995 International Conference
on Incineration and Thermal Treatment Technologies, University of California, Irvine, CA,
(1995) 527,
46. W.P. Linak, J.A. McSorley, J.O.L. Wendt, and J.E. Dunn, J. Air Poll. Cont. Assoc., 37 (8)
(1987) 934.
11
-------
-------
1. REPORT
4. TITLE AND SUBTITLE
Research on Emissions and Mitigation of POPs from
Combustion Sources
NRMRL-RTP-P-263
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before compter"
5. REPORT DATE
6. PEflFORMING ORGANIZATION CODE
. AUTHORS a w> Lee> p^ M> LemieuX) B>
Ryan, and J. D. Kilgroe
J. V.
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO,
See Block 12
11. CONTRACT/GRANT NO.
MA (Inhouse)
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; 1989-1997
14. SPONSORING AGENCY CODE
EPA/60.0 A3
is.SUPPLEMENTARY NOTES AppCD prOject officer is Chun Wai Lee, Mail Drop 65. 919/541-
7663. Presented at Conference, Air Pollution in the 21st Century: Priority Issues
and Policy Trends, Noordwijk, The Netherlands, 4/13-17/97.
i6. ABSTRACT The paper summarizes EPA's researchonemissions and controlof per sis-
tent organic pollutants (POPs) from combustion sources, with emphasis on source
characterization and measurement, formation and destruction mechanisms, forma-
tion prevention, and flue gas cleaning. Laboratory experiments conducted to exa-
mine polycyclic aromatic hydrocarbon (PAH) emissions from a wide variety of com-
bustion processes, ranging from pulverized coal utility boilers to wood stoves, have
shown that they exhibit widely different emission characteristics. Waste incineration
research has also shown that complex mechanisms, including physical mixing and
chemical kinetics, are involved in the formation of chlorinated products of incom-
plete combustion (PICs). Research has also indicated that the formation of ultra-
trace levels of chlorinated dioxins and chlorinated furans in combustion/ incineration
processes includes the complex interaction of several factors including temperature,
chlorine content, and catalyst. The beneficial effect of sulfur and sorbents for dioxin
formation prevention is demonstrated. EPA's effort to develop and evaluate state-of-
the-art technologies for on-line measurements of PAHs, volatile PICs, dioxins, and
furans is also discussed. The promising potential of applying artificial-intelligence-
based control for improving combustion as a POP prevention approach is shown. .
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.tDENTIFlERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Organic Compounds
Combustion
Emission
Research
Artificial Intelligence
Pollution Control
Stationary Sources
Persistent Organic Pol-
lution (POP)
13 B
07 C
21B
14G
14F
06D
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21, NO. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 O-73J
-------
------- |