William P. Linak

Toxic metal emissions from incineration: Mechanisms and control

Abstract Toxic metals appear in the effluents of many combustion processes, and their release into the environment has come under regulatory scrutiny. This paper reviews the nature of the problems associated with toxic metals in combustion processes, and describes where these problems occur and how they are addressed through current and proposed regulations. Although emphasis in this paper is on problems associated with metals from incineration processes, conventional fossil fuel combustion is also considered, insofar as it pertains to mechanisms governing the fate of metals during combustion in general. This paper examines the release of metals into the vapor phase, with the particle dynamics of a nucleating, condensing, and coagulating aerosol that may be subsequently formed, and with the reactive scavenging of metals by sorbents. Metals can be introduced into combustion chambers in many physical and chemical forms. The subsequent transformations and vaporization of any volatile metal depend on the combustion environment, the presence of chlorine and other species (reducing or oxidizing), on the nature of the reactive metallic species formed within the furnace, and on the presence of other inorganic species such as alumino-silicates. Some insight into how these factors influence metal release can be gained by considering the release of organic sodium during coal char combustion. Once vaporized, a metal vapor cloud will normally pass through its dewpoint to form tiny nuclei, or condense around existing particles. These aerosols are then affected by other dynamic processes (including coagulation) as they evolve with time. This paper shows how current mathematical descriptions of aerosol dynamics are very useful in predicting metal aerosol size distributions in combustion systems. These models are applied to two prototype problems, namely: the prediction of the temporal evolution of a particle size distribution of a self-coagulating aerosol initially composed of nuclei; and the scavenging of nuclei by coagulation with larger sorbent particles. A metal vapor can also react with certain aluminosilicate sorbents. This process, which will occur at temperatures above the dewpoint, is described, and is important, since it allows the high temperatures in incineration processes to be exploited to allow the formation of water-unleachable metal-containing compounds that can be isolated from the environment. Future research problems are also identified.

Trace metal transformation mechanisms during coal combustion

Abstract Mechanisms governing the fate of trace metals during coal combustion are reviewed, and new theoretical results interpreting existing data are presented. Emphasis is on predicting the size-segregated speciation of trace metals in pulverized coal-fired power plant effluents. This facet, which determines how trace metals originally in coal impact the environment, is controlled by fuel composition and combustion conditions. Multicomponent equilibrium calculations are used to predict vaporization/condensation temperatures for antimony, arsenic, beryllium, cadmium, chromium, lead, mercury, nickel, and selenium compounds in coal combustion flue gases, for a representative Illinois No. 6 coal. Experimental data show that equilibrium provides a good guide on the effect of chlorine on the partitioning of pure nickel, cadmium, and lead salts, introduced separately into a gaseous turbulent diffusion flame within an 82 kW combustor. Metal nuclei coagulation mechanisms are examined using existing computer codes, and these predict that coagulation does not allow condensed metal nuclei to be scavenged by existing coal ash particles. Rather, literature data on trace metal enrichment on small particles are consistent with processes of reactive scavenging of metals by larger particles, and it is suggested that these processes might be exploited further to convert these metals into environmentally benign forms.

Nitrous oxide behavior in the atmosphere, and in combustion and industrial systems

Tropospheric measurements show that nitrous oxide (N2O) concentrations are increasing over time. This demonstrates the existence of one or more significant anthropogenic sources, a fact that has generated considerable research interest over the last several years. The debate has principally focused on (1) the identity of the sources, and (2) the consequences of increased N2O concentrations. Both questions remain open, to at least some degree. The environmental concerns stem from the suggestion that diffusion of additional N2O into the stratosphere can result in increased ozone (O3) depletion. Within the stratosphere, N2O undergoes photolysis and reacts with oxygen atoms to yield some nitric oxide (NO). This enters into the well known O3 destruction cycle. N2O is also a potent absorber of infrared radiation and can contribute to global warming through the greenhouse effect. A major difficulty in research on N2O is measurement. Both electron capture gas chromatography and continuous infrared methods have seen considerable development, and both can be used reliably if their limitations are understood and appropriate precautions are taken. In particular, the ease with which N2O is formed from NO in stored combustion products must be recognized; this can occur even in the lines of continuous sampling systems. In combustion, the homogeneous reactions leading to N2O are principally NCO + NO → N2O + CO and NH + NO → N2O + H, with the first reaction being the most important in practical combustion systems. Recent measurements have resulted in a revised rate for this reaction, and the suggestion that only a portion of the products may branch into N2O + CO. Alternatively, recent measurements also suggest a reduced rate for the N2O + OH destruction reaction. Most modeling has been based on the earlier kinetic information, and the conclusions derived from these studies need to be revisited. In high-temperature combustion, N2O forms early in the flame if fuel-nitrogen is available. The high temperatures, however, ensure that little of this escapes, and emissions from most conventional combustion systems are quite low. The exception is combustion under moderate temperature conditions, where the N2O is formed from fuel-nitrogen, but fails to be destroyed. The two principal examples are combustion fluidized beds, and the downstream injection of nitrogen-containing agents for nitrogen oxide (NOx) control (e.g., selective noncatalytic reduction with urea). There remains considerable debate on the degree to which homogeneous vs heterogeneous reactions contribute to N2O formation in fluidized bed combustion. What is clear is that the N2O yield is inversely proportional to bed temperature, and conversion of fuel-nitrogen to N2O is favored for higher-rank fuels. Fixed-bed studies on highly devolatilized coal char do not indicate a significant role for heterogeneous reactions involving N2O destruction. The reduction of NO at a coal char surface appears to yield significant N2O only if oxygen (O2) is also present. Some studies show that the degree of char devolatilization has a profound influence on both the yield of N2O during char oxidation, and on the apparent mechanism. Since the char present in combustion fluidized beds will likely span a range of degrees of devolatilization, it becomes difficult to conclusively sort purely homogeneous behavior from potential heterogeneous contributions in practical systems. Formation of N2O during NOx control processes has primarily been confined to selective noncatalytic reduction. Specifically, when the nitrogen-containing agents urea and cyanuric acid are injected, a significant portion (typically > 10%) of the NO that is reduced is converted into N2O. The use of promoters to reduce the optimum injection temperature appears to increase the fraction of NO converted into N2O. Other operations, such as air staging and reburning, do not appear to be significant N2O producers. In selective catalytic reduction the yield of N2O depends on both catalyst type and operating condition, although most systems are not large emitters. Other systems considered include mobile sources, waste incineration, and industrial sources. In waste incineration, the combustion of sewage sludge yields very high N2O emissions. This appears to be due to the very high nitrogen content of the fuel and the low combustion temperatures. Many industrial systems are largely uncharacterized with respect to N2O emissions. Adipic acid manufacture is known to produce large amounts of N2O as a by-product, and abatement procedures are under development within the industry.

Comparison of Particle Size Distributions and Elemental Partitioning from the Combustion of Pulverized Coal and Residual Fuel Oil

ABSTRACT U.S. Environmental Protection Agency (EPA) research examining the characteristics of primary PM generated by the combustion of fossil fuels is being conducted in efforts to help determine mechanisms controlling associated adverse health effects. Transition metals are of particular interest, due to the results of studies that have shown cardiopulmonary damage associated with exposure to these elements and their presence in coal and residual fuel oils. Further, elemental speciation may influence this toxicity, as some species are significantly more water-soluble, and potentially more bio-available, than others. This paper presents results of experimental efforts in which three coals and a residual fuel oil were combusted in three different systems simulating process and utility boilers. Particle size distributions (PSDs) were determined using atmospheric and low-pressure impac-tion as well as electrical mobility, time-of-flight, and light-scattering techniques. Size-classified PM samples from this study are also being utilized by colleagues for animal instillation experiments. Experimental results on the mass and compositions of particles between 0.03 and >20 μm in aerodynamic diameter show that PM from the combustion of these fuels produces distinctive bimodal and trimodal PSDs, with a fine mode dominated by vaporization, nucleation, and growth processes. Depending on the fuel and combustion equipment, the coarse mode is composed primarily of unburned carbon char and associated inherent trace elements (fuel oil) and fragments of inorganic (largely calcium-alumino-silicate) fly ash including trace elements (coal). The three coals also produced a central mode between 0.8- and 2.0-μm aerodynamic diameter. However, the origins of these particles are less clear because vapor-to-particle growth processes are unlikely to produce particles this large. Possible mechanisms include the liberation of micron-scale mineral inclusions during char fragmentation and burnout and indicates that refractory transition metals can contribute to PM <2.5 μm without passing through a vapor phase. When burned most efficiently, the residual fuel oil produces a PSD composed almost exclusively of an ultrafine mode (~0.1 μm). The transition metals associated with these emissions are composed of water-soluble metal sulfates. In contrast, the transition metals associated with coal combustion are not significantly enriched in PM <2.5 μm and are significantly less soluble, likely because of their association with the mineral constituents. These results may have implications regarding health effects associated with exposure to these particles.

ON TRIMODAL PARTICLE SIZE DISTRIBUTIONS IN FLY ASH FROM PULVERIZED-COAL COMBUSTION

Combustion-generated fine particles, defined as those with aerodynamic diameters less than 2.5 μm, have come under increased regulatory scrutiny because of suspected links to adverse human health effects. Whereas classical theories regarding coal combustion suggest that mechanisms of ash vaporization and fragmentation lead to bimodal ash particle size distributions (PSDs), this paper presents experimental results supporting other existing hypotheses that three distinct ash modes may be more appropriate. This paper focuses on the existence and generality of a central mode, between approximately 0.7 and 3.0 μm diameter. This central mode is presumably caused by fragmentation mechanisms, but is still important from a health perspective, because a large portion is contained within the 2.5 μm particle size fraction. Presented here are experimental results from two different laboratory combustors and one industrial boiler, all burning pulverized coals. Use of a variety of particle-sampling and size classification methods, including electrical mobility, time-of-flight, and inertial (low-pressure impaction) methods, confirms that the central mode is not an artifact of the particle-sampling and -sizing methods used. Results from the combustion of 10 different coals consistently show that this central mode is significant for both high-and low-rank coals. Size-segregated elemental distributions of calcium, iron, and aluminum provide additional insight into mechanisms of formation of each mode. Field tests show that the central mode can be the major contributor to fine particle emissions leaving an electrostatic precipitator (ESP). The new experimental results presented here are interpreted in the light of complementary existing data and available theories from the literature.

Electrophilic and redox properties of diesel exhaust particles

The adverse health effects of air pollutants have been associated with their redox and electrophilic properties. Although the specific chemical species involved in these effects are not known, the characterization of their general physical and chemical properties is important to our understanding of the mechanisms by which they cause health problems. This manuscript describes results of a study examining the partition properties of these activities in aqueous and organic media. The water and dichloromethane (DCM) solubility of redox active and electrophilic constituents of seven diesel exhaust particle (DEP) samples were determined with assays developed earlier in this laboratory. The constituents exhibiting redox activity, which included both metals and nonmetal species, were associated with the particles in the aqueous suspensions. Portions of the redox active compounds were also DCM-soluble. In contrast, the electrophilic constituents included both water-soluble and DCM-soluble species. The role of quinones or quinone-like compounds in redox and electrophilic activities of the DCM-soluble constituents was assessed by reductive acetylation, a procedure that inactivates quinones. The results from this experiment indicated that most of the activities in the organic extract were associated with quinone-like substances. The partition properties of the reactive species are important in exposure assessment since the toxicokinetics of particles and solutes are quite distinct.

Differential Pulmonary Inflammation and In Vitro Cytotoxicity of Size-Fractionated Fly Ash Particles from Pulverized Coal Combustion

Abstract Exposure to airborne particulate matter (PM) has been associated with adverse health effects in humans. Pulmonary inflammatory responses were examined in CD1 mice after intratracheal instillation of 25 or 100 μg of ultrafine (<0.2 μm), fine (<2.5 μm), and coarse (>2.5 μm) coal fly ash from a combusted Montana subbituminous coal, and of fine and coarse fractions from a combusted western Kentucky bituminous coal. After 18 hr, the lungs were lavaged and the bronchoalveolar fluid was assessed for cellular influx, biochemical markers, and pro-inflammatory cytokines. The responses were compared with saline and endotoxin as negative and positive controls, respectively. On an equal mass basis, the ultrafine particles from combusted Montana coal induced a higher degree of neutrophil inflammation and cytokine levels than did the fine or coarse PM. The western Kentucky fine PM caused a moderate degree of inflammation and protein levels in bronchoalveolar fluid that were higher than the Montana fine PM. Coarse PM did not produce any significant effects. In vitro experiments with rat alveolar macrophages showed that of the particles tested, only the Montana ultrafine displayed significant cytotoxicity. It is concluded that fly ash toxicity is inversely related with particle size and is associated with increased sulfur and trace element content.

Characterization of Fine Particulate Matter Produced by Combustion of Residual Fuel Oil

ABSTRACT Combustion experiments were carried out on four different residual fuel oils in a 732-kW boiler. PM emission samples were separated aerodynamically by a cyclone into fractions that were nominally less than and greater than 2.5 |j.m in diameter. However, examination of several of the samples by computer-controlled scanning electron microscopy (CCSEM) revealed that part of the PM2.5 fraction consists of carbonaceous cenospheres and vesicular particles that range up to 10 |j.m in diameter. X-ray absorption fine structure (XAFS) spectroscopy data were obtained at the S, V, Ni, Fe, Cu, Zn, and As K-edges and at the Pb L-edge. Deconvolution of the X-ray absorption near edge structure (XANES) region of the S spectra established that the dominant molecular forms of S present were sulfate (26-84% of total S) and thiophene (13-39% of total S). Sulfate was greater in the PM2.5 samples than in the PM25+ samples. Inorganic sulfides and elemental sulfur were present in lower percentages. The Ni XANES spectra from all of the samples agreed fairly well with that of NiSO4, while most of the V spectra closely resembled that of vanadyl sulfate (VO•SO4•xH2O). The other metals investigated (i.e., Fe, Cu, Zn, and Pb) also were present predominantly as sulfates. Arsenic was present as an arsen-ate (As+5). X-ray diffraction patterns of the PM2.5 fraction exhibit sharp lines due to sulfate compounds (Zn, V, Ni, Ca, etc.) superimposed on broad peaks due to amorphous carbons. All of the samples contain a significant organic component, with the loss on ignition (LOI) ranging from 64 to 87% for the PM2.5 fraction and from 88 to 97% for the PM2.5+ fraction. Based on 13C nuclear magnetic resonance (NMR) analysis, the carbon is predominantly condensed in graphitic structures. Aliphatic structure was detected in only one of seven samples examined.

Sorbent capture of nickel, lead, and cadmium in a laboratory swirl flame incinerator

Abstract The in situ capture of toxic metals by sorbents was investigated in a small semi-industrial scale 82 kW research combustor. Metals considered were nickel, lead, and cadmium. These metals were introduced into the system as aqueous nitrate solutions, sprayed down the center of a natural gas flame, supported on a variable swirl burner. Kaolinite, bauxite, and hydrated lime were injected along the centerline in the postflame, near the peak system temperature. Measurements of both the submicron acrosol size distribution and the size segregated particulate composition in the exhaust allowed the effects of sorbent injection to be ascertained, both with and without the presence of chlorine. Lead and cadmium could be almost completely scavenged by kaolinite, which formed melted particles. Bauxite, which did not melt, was exceedingly effective in capturing cadmium. However, chlorine inhibited metal capture in these instances. Hydrated lime also captured cadmium to form a eutectic melt, and this process was slightly enhanced by chlorine. Nickel alone did not significantly vaporize and was not captured by kaolinite. However, in the presence of chlorine, nickel did vaporize and was effectively captured. These results are interpreted and compared to bench scale results in the literature. Two mechanisms, or scenarios, for toxic metal capture are presented.

Differential Potentiation of Allergic Lung Disease in Mice Exposed to Chemically Distinct Diesel Samples

Numerous studies have demonstrated that diesel exhaust particles (DEP) potentiate allergic immune responses, however the chemical components associated with this effect, and the underlying mechanisms are not well understood. This study characterized the composition of three chemically distinct DEP samples (N, C, and A-DEP), and compared post-sensitization and post-challenge inflammatory allergic phenotypes in BALB/c mice. Mice were instilled intranasally with saline or 150 microg of N-DEP, A-DEP, or C-DEP with or without 20 microg of ovalbumin (OVA) on days 0 and 13, and were subsequently challenged with 20 microg of OVA on days 23, 26, and 29. Mice were necropsied 18 h post-sensitization and 18 and 48 h post-challenge. N-DEP, A-DEP, and C-DEP contained 1.5, 68.6, and 18.9% extractable organic material (EOM) and 47, 431, and 522 microg of polycyclic aromatic hydrocarbons (PAHs), respectively. The post-challenge results showed that DEP given with OVA induced a gradation of adjuvancy as follows: C-DEP approximately A-DEP > N-DEP. The C- and A-DEP/OVA exposure groups had significant increases in eosinophils, OVA-specific IgG1, and airway hyperresponsiveness. In addition, the C-DEP/OVA exposure increased the T helper 2 (T(H)2) chemoattractant chemokine, thymus and activation-regulated chemokine and exhibited the most severe perivascular inflammation in the lung, whereas A-DEP/OVA increased interleukin (IL)-5 and IL-10. In contrast, N-DEP/OVA exposure only increased OVA-specific IgG1 post-challenge. Analysis of early signaling showed that C-DEP induced a greater number of T(H)2 cytokines compared with A-DEP and N-DEP. The results suggest that potentiation of allergic immune responses by DEP is associated with PAH content rather than the total amount of EOM.