Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 12
Review of the U.S. Department of Energy Office of Fossil Energy's Research Plan for Fine Particulates 2 Background and Summary INTRODUCTION The major concern about elevated concentrations of ambient PM2.5 is based on persistent evidence (e.g., statistical associations) of human health effects associated with airborne particles, although most PM measurements to date are for PM10 or the more inclusive measure of PM. Another concern related to PM2.5 is regional haze in the atmosphere that impairs visibility. The EPA is committed to a strategy of systematically improving visibility in both rural and urban areas. DOE-FE's PM2.5 research program is focused on determining (1) the environmental impact of energy production from the combustion of fossil fuels and (2) the options for emission-control technologies of gases and particles that would reduce ambient PM2.5 loadings. Sulfur dioxide and NOx emissions from coal-fired power plants are areas of focus because these precursor gases lead to the formation of secondary particles containing sulfates and nitrates, which are the main contributors to concentrations of fine particles in the ambient air. A smaller contribution is the direct emission to the atmosphere of fine particles consisting of minerals and trace metals found in coal. In keeping with DOE's mission, the DOE-FE program does not include studies on the health effects of air pollution, which are the responsibility of EPA, the National Institute of Environmental Health Sciences, and others. Thus, the current DOE-FE program does not have a health science component, although DOE could collect valuable data for epiderniologic and exposure studies conducted by others. Impaired visibility associated with emissions from fossil-fueled power plants is primarily a regional problem, and the DOE-FE measurement program will collect data on the origins of regional haze (e.g., the eastern United States). An understanding of the air quality issues associated with PM2.5 must include: the relationship between ambient PM2.5 and human health effects; an identification of the sources of particulates; emission and ambient monitoring techniques; source-receptor relationships; the relationship of ambient PM2.5 concentrations to personal exposure; and possible control strategies. Although this report is not a comprehensive study of all activities related to PM2.5, this chapter provides a brief overview for readers unfamiliar with PM2.5 research. HEALTH EFFECTS The primary reason for the EPA's new PM2.5 standards is the epiderniologic association of ambient particles with adverse health effects. This section reviews the current understanding of these effects as they relate to the DOE-FE program. In 1996, the EPA issued the "Air Quality Criteria Document for Particulate Matter," in which numerous epiderniologic studies were cited (EPA, 1996). The document included the following conclusion: By far the strongest evidence for ambient PM exposure health risks is derived from epiderniologic studies. Many epiderniologic studies have shown statistically significant associations of ambient PM levels with a variety of human health end points, including mortality, hospital admissions and emergency room visits, respiratory illness and symptoms measured in community surveys, and physiologic changes in mechanical pulmonary function. Associations of both short-term and long-term PM exposure with most of these end points have been consistently observed. The general internal consistency of the epiderniologic database and available findings have led to increasing public health concern, due to the severity of several studied end
OCR for page 13
Review of the U.S. Department of Energy Office of Fossil Energy's Research Plan for Fine Particulates points and the frequent demonstration of associations of health and physiologic effects with ambient PM levels at or below the current U.S. NAAQS for PM10. The weight of epidemiologic evidence suggests that ambient PM exposure has affected the public health of U.S. populations (EPA, 1996, vol. 3, p. 13-30). The majority of the studies reviewed in the 1996 criteria document were related to the association of acute health effects and exposure to particulates, mainly PM10 and TSP (total suspended particulates). Prospective studies of chronic exposure have associated ambient air pollution with mortality (Dockery et al., 1993; Pope et al., 1995; Abbey et al., 1999). Respirable particles (PM2.5) have stronger correlations with health effects than other air pollutants (Dockery et al., 1993), and deaths were associated with cardiopulmonary diseases (Dockery et al., 1993; Pope et al., 1995). Indeed, one reason for the increased interest in the effects on cardiac function is that seven times as many individuals suffer from cardiac disease than from underlying pulmonary disease (Pope et al., 1995). In recent epiderniologic studies, the contributions of copollutants (other pollutants present in addition to particulates) have received greater attention. Ozone (Loomis et al., 1996; Stieb et al., 1996; Hoek et al., 1997; Delfino et al., 1997, 1998; Burnett et al., 1997; Moolgavkar et al., 1997), nitrogen dioxide (Pantazopoulou et al., 1995; Dab et al., 1996; Moolgavkar et al., 1997), sulfur dioxide (Dab et al., 1996; Moolgavkar et al., 1997), and carbon monoxide (Moolgavkar et al., 1997; Burnett et al., 1998) have also been positively and significantly associated with adverse health effects. Uncertainties in epidemiologic studies include the magnitude of risk for PM, the attribution of observed health effects to specific PM constituents, the time intervals over which PM health effects are manifested, the extent to which findings in one location can be generalized to other locations, the nature and magnitude of the overall public health risk from ambient PM exposure, and the biologic mechanisms of the observed responses. Differentiating adverse responses in individuals, specific toxicity of ambient particle exposures, and the effects of the "pollution mix" may require comprehensive monitoring specifically related to individual human exposures, as well as sensitive measures of health effects, to reduce uncertainties in the definitions of health effects from ambient air pollution. No epiderniologic study has identified a particular chemical constituent of PM as a causative agent. Indications are that physical characteristics of the particles may be important. PM2.5 relationships to health can differ from those of PM10 (EPA, 1996), and one study showed that ultrafine particles (< 0.1 μm) are more sensitive indicators of effect than fine particles (Peters et al., 1997a). Ultrafine particles as a potential cause of the observed health effects are the focus of several European epidemiologic studies and a component of the Atlanta-Based Aerosol Research Inhalation Epidemiology Study (ARIES), which is partly supported by the DOE-FE program (see Chapter 3). Many ongoing health studies are attempting to define the pollutants or components of PM that are the most important to health effects and the biological mechanisms that would establish causality. In fact, new epidemiologic research has focused on the biologic mechanisms of response. A primary biological hypothesis is that inflammation is the underlying cause of adverse cardiopulmonary responses (Godleski and Clarke, 1999). Peters et al. (1997b) supported a systemic inflammatory response mechanism, based on increases in blood viscosity associated with elevated TSP and sulfur dioxide levels in Augsburg, Germany. Another biological hypothesis focused on autonomic nervous system mechanisms, assessing cardiac responses by electrocardiogram and variations in heart rate (Godleski, 1999), which have been associated with adverse health effects in the medical literature (Dougherty and Burr, 1992; Ewing et al., 1981; Lombardi et al., 1987; van Ravenswaaig-Arts et al., 1993; and Bigger and Schwartz, 1994). Recently, Pope et al. (1999) showed significant increases in sympathetic and parasympathetic influences on the heart associated with increases in ambient PM10 based on variable heart rates in seven elderly subjects. These studies, coupled with comprehensive assessments of exposure, could make a significant contribution to the understanding of morbidity and mortality associated with increases in PM. In a recent study, the EPA reviewed toxicologic studies of primary and secondary particle emissions from power plants (EPA, 1996). Although a number of historical data sets exist, studies of coal fly ash using samples collected more than 20 years ago are unlikely to be applicable to emissions from current facilities. The health effects of sulfate have been debated for many years, but it appears that associated metal or hydrogen ions may be the bases of "sulfate toxicity." Pathologic responses to particles containing silicon vary with the surface composition of the particle. Silica
OCR for page 14
Review of the U.S. Department of Energy Office of Fossil Energy's Research Plan for Fine Particulates (α quartz) is highly toxic, although most silicates have limited toxicity (Godleski, 1994). Other recent studies have focused on residual oil fly ash (Killingsworth et al., 1997; Dreher et al., 1997; Kodavanti et al., 1998; Watkinson et al., 1998), because this material is a source of transition metals that may be linked to responses in both respiratory and cardiovascular systems (Chapman et al., 1997) and could provide a plausible mechanism for some of the health effects observed with PM. Because residual oil fly ash exhibits reproducible responses in biological systems, it has become a favorite with some laboratory investigators of PM toxicology. Some of the metals identified in these studies as important are iron, copper, zinc, and nickel (Gavett et al., 1997; Dreher et al., 1997). The responses observed in these studies tend to support the hypothesis of inflammatory processes. One study specifically evaluated pathologic and immunologic effects of coal fly ash in rats (Dormans et al., 1999). However, the lowest concentration used was 10 mg/m3, which is much higher than ambient exposures. In vivo and in vitro toxicologic studies using ambient urban particles have the greatest potential to address the question of specific toxicity. However, the variability of the samples makes assessing their effects exceedingly complex. The development of an ambient particle concentrator (Sioutas et al., 1997) has facilitated the collection of daily samples for in vitro studies (Goldsmith et al., 1999) and made it possible to carry out laboratory experiments with animal models (Clarke et al., 1999; Godleski et al., 1999). These studies will require careful characterization of the particles used in the exposure. Initial experience with this approach indicates that dose-response relationships are complicated by the day-to-day variability in the toxicity of ambient particles, which has necessitated large numbers of replications. Overall, the toxicologic studies to date have provided a basis for further studies testing specific hypotheses of mechanisms of toxicity and have begun to reveal mechanistic links that support the morbidity and mortality observed in epidemiologic studies. However, the contributions of specific sources to the toxicity of ambient particulates have not been determined. Furthermore, it is not clear whether specific components of the ambient mixture are responsible for biologic responses or all components acting in concert produce the response. Similarly, although the toxicity of components of both primary and secondary particles that result from coal-fired power plants have been studied individually, the contributions of these constituents to the toxicity of ambient air particulates are unknown. In summary, there are considerable gaps in the knowledge of how current sources of ambient particles contribute to the ambient mixture of particulate constituents that result in adverse health effects. REGIONAL HAZE AND IMPAIRED VISIBILITY The degradation of the visual environment is the most direct and noticeable effect of fine particles in the atmosphere. The effects include decreased visual range, the discoloration of distant objects, and the whitening of the blue sky caused by the scattering and absorption of sunlight by atmospheric particles. The most effective scatterers are particles in the size range of 0.3 to 1 μm. The most significant contributors to fine mass in the optically active size range are sulfates and organics, as well as nitrates and dust in some regions of the country. Based on simultaneous measurements of chemical composition and light scattering over the eastern United States, sulfates have been shown to contribute more than their mass share in nonurban areas, primarily because of water absorption at higher humidities. Regional haze is created when the contributions of many individual sources are combined during the long-range transport of particles, which results in homogeneous hazy air masses. Stagnating air masses over the eastern United States can extend more than a thousand kilometers (600 miles). Regional-scale hazy episodes are also frequently associated with elevated concentrations of ozone and other secondary pollutants. Climatologically, the lowest visibility has been recorded east of the Mississippi River. Visibility is worse during the summer than during the winter. Historically, the visibility over the eastern United States declined during the 1960s and early 1970s (NAPAP, 1991), but it appears to have improved in the 1990s. Poor visibility has also been recorded over the western air basins, such as Los Angeles and the San Joaquin Valley in California, and in many other areas, such as Big Bend National Park in Texas. Recently, EPA introduced rules for regulating regional haze over the United States for the next 60 years. Sources of sulfur, such as coal-burning power plants, are explicitly targeted for emission controls for reducing regional haze over the eastern part of the country, as well as in western regions. The best visibility is over the pristine and arid regions of the southwestern United States where many national parks
OCR for page 15
Review of the U.S. Department of Energy Office of Fossil Energy's Research Plan for Fine Particulates are located, and preserving the visibility in those areas is one goal of the new regulations. Several major power plants in the area of the Grand Canyon National Park, for example, are hypothesized to threaten the visibility in that region during certain times of the year. CHARACTERISTICS OF AMBIENT PARTICULATE MATTER Particulate matter is not a single compound but a complex mixture whose composition and morphology vary significantly in time and space. Airborne particles may contain numerous inorganic and organic components. The distribution of atmospheric particles tends to be trimodal with peaks in the volume of particles at about 0.01 μm (ultrafine mode), between 0.1 and 1 μm (accumulation mode), and between 2.5 and 10 μm (coarse mode). Measurements indicate that the number concentration of airborne particles is dominated by very small, ultrafine particles (less than 0.1 μn in diameter); the surface area of nonporous smooth particles is dominated by fine particles (between 0.1 and 2.5 μm in diameter); and the mass concentration is dominated by the sum of the mass concentration of fine and coarse particles (0.1 to 10 μm in diameter). Particle size and mass attributes are of particular interest because they may be related to health effects, but a number of other characteristics (e.g., surface area, porosity, rugosity) also affect health. For example, metals are ubiquitous constituents of ambient PM, as are acids and organic compounds. A small portion of ambient PM is believed to contain biological components, such as bacteria, viruses, pollens, and detritus. Nitrate and sulfate salts, peroxides, and elemental carbon might also have health-related aspects. Currently, mass data for PM2.5 are only available for about 100 monitoring sites in the United States. These data are collected mostly by regional networks, such as the Interagency Monitoring of Protected Environments (IMPROVE) and the Clean Air Status and Trends Network (CASTNet), and by some urban sites.1 This network is too sparse for detailed spatial mapping. Nevertheless, based on available PM10 and PM2.5 data, various investigators have estimated concentration patterns of PM2.5 throughout the United States (Falke, 1999). These estimates, which contain many uncertainties, suggest that many areas will be out of compliance with the new PM2.5 NAAQS. EPA is currently engaged in an extensive sampling program to determine which areas are out of compliance; the completion of this sampling program will help to resolve many of the current uncertainties. Current analyses and assessments indicate that the proposed daily ambient PM2.5 EPA standard of 65 μg/m3 is exceeded only in California. However, large areas of the eastern United States appear to be near or to exceed the proposed annual average of 15 μg/m3. Concentrations above 10 μg/m3 are common throughout the eastern United States, and an eight-state area surrounding the Appalachian Mountains shows PM 2.5 levels above 12 μg/m3. In some urban areas, especially the Washington to New York corridor and the Pittsburgh area, the three-year average (1994-1996) annual PM2.5 levels were above 15 μg/m3 (the annual PM 2.5 standard). The data show a relatively consistent composition of PM2.5 across the eastern United States with PM2.5 dominated on a regional scale by ammonium sulfate and carbonaceous particles. The higher concentrations of PM2.5 in the eastern United States are attributed to higher emissions of sulfur dioxide in the region and the resulting concentration of ammonium sulfate. Several areas in California, such as the San Joaquin Valley and the Rubidoux area of the South Coast Basin, also have high concentrations of ammonium nitrate. Although PM2.5 concentrations are known to vary substantially, both daily and seasonally, long-term trends have been difficult to study because of the paucity of data. Lipfert (1998) combined PM2.5, PM3.5 , and cascade impactor data and concluded that "fine PM" declined about 5 percent per year from 1960 to 1990. His analysis suggested that primary PM2.5 may have decreased more than secondary (sulfate) PM2.5 precursors. The contribution to ambient PM from emissions of large power plants has long been a concern in the United States because of acid precipitation and regional haze. Studies in the 1970s and 1980s revealed a great deal about the characteristics of the contributions of the power-generation sector to PM2.5 haze and acid precipitation. The findings of the National Acid Precipitation Assessment Program (NAPAP) and similar studies 1 The Interagency Monitoring of Protected Environments (IMPROVE) is a collaborative monitoring program to establish present visibility levels and trends and to identify sources of manmade impairment. The IMPROVE program has been collecting data since 1987 in 20 Class I areas nationwide. There are 69 sites in the network. The current Clean Air Status and Trends Network (CASTNet) was developed by EPA in response to the Clean Air Act Amendments of 1990 requiring implementation of a national network to measure national status and trends in air quality.
OCR for page 16
Review of the U.S. Department of Energy Office of Fossil Energy's Research Plan for Fine Particulates could serve as a starting point for DOE-FE's PM2.5 research program. MONITORING OF AMBIENT PARTICULATE MATTER To monitor the concentrations of PM2.5 for regulatory purposes, filter samples are collected over a 24-hour period every third day. Consistent measurements of fine particulate mass are made by using the standardized FRM (federal reference method), which is appropriate for establishing compliance with the ambient air quality standard but is not suitable for determining sources of fine particles or characterizing ambient PM. Because the FRM does not account for the substantial loss of semivolatile compounds, its PM samples are not truly representative of ambient conditions. Research-quality monitoring of ambient PM includes measurements of detailed chemical composition of fine particles, as well as the size distribution. To account for the entire fine mass concentration, the measurement of chemical composition must first reconstitute the mass balance. Second, trace constituents must be measured to identify specific source types, such as coal-fired power plants, which can be inferred from concentrations of selenium or arsenic as a tracer of fly ash. Third, the chemical composition and/or size distribution must include measurements of the materials suspected of contributing to adverse health effects, such as toxic metals and ultrafine aerosols. All of these data are important for source modeling and receptor modeling. From the perspective of the DOE and FETC, key features of a research-quality PM monitoring site should include chemical speciation measurements that, in principle, could identify the contribution of coal-fired power plants to the ambient PM2.5 concentrations. The monitoring site, or network of sites, should be located within the zone of influence of major power-plant emissions, such as in the Ohio River Valley (see Chapter 3). Ideally, research-quality monitoring would focus on regional-scale air pollution episodes during which acute health effects and exceedances of the PM standard are expected to occur. EPA Supersites EPA has established a "supersites" fine-particle monitoring program to augment the 1,000 to 1,200 regulatory FRM sites and the 200 to 300 sites collecting samples for speciation (Scheffe, 1999). The three main goals of the supersites monitoring program are listed below: Support the development of SIPs (state implementation plans) through a better understanding of source-receptor relationships. Gather monitoring data for health and exposure studies. Evaluate and compare emerging sampling and analysis methods. The supersites will be set up in four to seven airsheds representing a spectrum of conditions across the country. According to the EPA, implementation of the supersites will be flexible to meet "the needs posed by questions and hypotheses related to the coordinated research objectives for that location" (EPA, 1998c). For example, at some locations, the site could have a central high-grade monitoring site surrounded by satellite monitoring sites. At other locations, it could be a flexible, mobile sampling system that yields cross-sectional and pollutant transport data. Also, some supersites may focus on testing new instrumentation while others focus on determining source-receptor relationships. EPA has established the following guiding principles for supersites: A comprehensive supersite monitoring network should be an integrated part of the larger PM monitoring network. A supersite should be designed as a "learning" rather than a "measurement" site. Consistent, but not necessarily identical, measurements should be provided by the supersites. The supersite network should leverage other governmental and private investments. Data analysis and evaluation capabilities should be built in from the start. Measurements should be organized based on the following questions: what are the major questions to be answered; what are the hypotheses; where, when, and what should be measured. These general principles would also provide useful guidelines for designing the DOE monitoring programs.
OCR for page 17
Review of the U.S. Department of Energy Office of Fossil Energy's Research Plan for Fine Particulates CHARACTERIZATION OF SOURCE EMISSIONS Two types of fine particles are formed from emissions from coal-fired power plants: primary particles, which represent a relatively minor contribution to ambient PM2.5 mass; and secondary particles, which represent a much more significant contribution to PM2.5 mass. The primary particles are emitted directly to the atmosphere and consist mainly of minerals and trace metal compounds found in coal. The amount and composition of these emissions depends on the composition of the coal and the nature of the emission controls on individual plants. The secondary particles are formed in the atmosphere from gaseous precursors. The secondary fine particles of most concern from coal combustion are sulfates and nitrates. Sulfates are formed from the oxidation of sulfur dioxide and consist primarily of sulfuric acid, ammonium bisulfate, letovicite, and ammonium sulfate. The relative distribution of the sulfate compounds depends on the availability of ammonia in the ambient air. Nitrates are formed from the oxidation of gaseous nitrogen oxides (nitric oxide and nitrogen dioxide), mainly as particulate ammonium nitrate. Because ammonium nitrate readily reacts with sulfuric acid to form ammonium bisulfate and gaseous nitric acid, particulate nitrate does not readily accumulate unless there is sufficient ammonia to neutralize the sulfate. The chemical composition of the primary fine particle emissions is the basis for receptor models to determine the contributions of various sources to ambient fine particle mass. Although chemical fingerprints of emissions from generic coal-fired power plants have been identified, the fingerprints have probably been altered as a result of recent and planned reductions in emissions of sulfur dioxide and NOx required by the 1990 Clean Air Act Amendments and the recent SIPs for achieving NOx reductions. Consequently, updating source profiles should be an important component of any research on identifying sources of PM2.5 based on receptor conditions. The relative contribution of coal-fired power plants to the formation of secondary fine particles will be determined using dispersion models incorporating the distribution of emission sources and nitrogen and sulfur chemistry. Speciated ambient measurements will be used to evaluate and improve these models. This will require that gridded emission inventories be developed for all gaseous precursors in relevant spatial and temporal scales. Although some inventories have been developed, they will have to be improved before they can be used as a basis for planning strategies. DOE could contribute by providing a high-quality inventory for coal-fired power plant emissions. This should be a straightforward task because all U.S. coal-fired power plants are required to monitor emissions of sulfur dioxide and NOx continuously. RELATIONSHIP BETWEEN EMISSIONS AND CONCENTRATIONS OF AMBIENT PARTICULATE MATTER The development of control strategies for fine PM will require that relationships between emissions and ambient PM concentrations, so-called source-receptor relationships, be established. If the receptor region were the city of Pittsburgh, for example, a general question would be which sources contribute to the PM levels in the area. A more specific question would be what fraction of the sulfate in Pittsburgh is emitted as sulfur dioxide by power plants in the state of Ohio. The development of these source-receptor relationships will be essential to an understanding of ambient PM. The development of reliable source-receptor relationships is a challenging task. Carefully planned field experiments represent one possible method for determining these relationships. Potential experiments include the release of passive tracers for sources of concern, the use of tracers of opportunity (e.g., transition metals), large-scale changes in emissions, and releases and ambient measurements of stable isotopes of sulfur or nitrogen. Unfortunately, each of these approaches has major drawbacks, ranging from the magnitude and cost of the undertaking to the inability of passive tracers to mimic gas-to-particle conversion processes (Hidy, 1984). Because of these difficulties, mathematical modeling will have to be the main tool for deriving source-receptor relationships. Potential modeling approaches could be statistical (including statistical receptor models) or deterministic (including atmospheric chemistry models). Statistical receptor models attempt to relate measured concentrations at a given site to their sources without reconstructing the dispersion patterns of the material. The CMB (chemical mass balance) model combines the chemical and physical characteristics of particles measured at the sources and the receptors to quantify the contributions of individual sources (Miller et al., 1972; Watson, 1984). Application of the CMB model requires aerosol composition measurements in
OCR for page 18
Review of the U.S. Department of Energy Office of Fossil Energy's Research Plan for Fine Particulates the receptor, the identification of all sources contributing significantly to the receptor, and reliable source profiles of all sources (Hopke, 1985). If the source profile measurements correspond to the period of the ambient measurements, uncertainties or errors in the CMB will be reduced (Glover et al., 1991). A successful CMB model can quantify the source contributions for the primary aerosol components (e.g., geological, motor vehicle, or others). Currently available receptor models are unable to quantify the sources of secondary aerosol components (sulfates, nitrates, ammonium, and secondary organics). Deterministic atmospheric chemistry models attempt to simulate all of the transformations of gaseous and particulate pollutants during their atmospheric lifetime. These tools require descriptions of emission patterns, meteorological conditions, chemical transformations, and removal processes (Seinfeld and Pandis, 1998). Some deterministic models simulate changes in the chemical composition of a given nair parcel as it is advected in the atmosphere (Lagrangian models); others describe the concentrations in a three-dimensional array of computational cells (Eulerian models). Although several deterministic models are currently available, their applications depend on suitable input, including meteorological fields (three-dimensional fields of wind speed and direction, temperature, relative humidity, cloud cover, rain intensity), emission fields, and boundary conditions. Ideally, this information would be available for the whole modeling domain for the entire modeling period. The evaluation of model simulations requires comprehensive ground and upper-air gaseous and aerosol measurements in multiple locations. The advantage of simulation tools is that they can provide information about the contribution of any source to any receptor in the modeling domain. The input data and computational requirements of deterministic models often limit their application to specific air pollution episodes. Although quantifying the contribhtion of a given source to the ambient PM concentrations of a given area is valuable, it is by no means sufficient for the design of PM control strategies. Effective strategies will also require an understanding of changes in the PM concentrations in response to changes in the source emissions. For example, how will a reduction in sulfur emissions of 50 percent by the source affect its contribution to sulfate in a given area? If the source contribution to sulfate is also reduced by 50 percent, then the response is considered to be linear. Several atmospheric chemical and physical processes are strongly nonlinear, however, which complicates the design of control strategies (NAPAP, 1991). The response of the sulfate concentrations in the eastern United States to changes in sulfur dioxide emissions was investigated during the NAPAP, which concluded that the most significant deviations from linearity occur during the winter close to the source regions (NAPAP, 1991). For specific episodes, a 50 percent reduction in emissions by all sulfur dioxide sources was predicted to result in reductions of the sulfate concentrations of 35 percent. On an annual average basis, the system response was predicted to be closer to linear with reductions more than 40 percent for all areas. A recent study (Husain et al., 1998) at Mayville, New York, and Whiteface Mountain, New York, shows an approximately linear proportional decline in sulfate following the reported decline in midwestern sulfur dioxide emissions. In other studies of the eastern United States, the realized reductions in sulfate concentrations appear to be less than expected (i.e., a nonlinear reduction of ambient sulfate levels in response to sulfur dioxide emissions) (NAPAP, 1998; Shannon, 1999). These results may reflect regional, temporal, and spatial differences of the studies (Shannon, 1999). Interactions between aerosol components can introduce further nonlinearities. For example, reductions in sulfate can free the associated ammonia in aerosol ammonium sulfate, and the additional ammonia can then react with available nitric acid vapor to form aerosol ammonium nitrate (Seinfeld and Pandis, 1998). These chemical interactions result in the replacement of the controlled sulfate by nitrate, thereby reducing the effectiveness of the sulfur dioxide controls on ambient PM levels. In extreme cases in specific areas in the northeastern United States, controls for sulfur dioxide emissions could result in a small increase in PM concentrations during the winter (West et al., 1999). Similar nonlinearities are expected in the response of PM concentrations to changes in NOx, volatile organic compounds, and ammonia emissions. The partitioning of semivolatile organic aerosol components between the gas and aerosol phases suggests that the organic aerosol system is often nonlinear (Odum et al., 1996). These nonlinearities are still not well understood. Although deterministic models have been the major source of insight, the lack of complete data has limited their application. The lack of suitable measurements has also limited the quantification of responses of aerosols to emission changes.
OCR for page 19
Review of the U.S. Department of Energy Office of Fossil Energy's Research Plan for Fine Particulates AMBIENT PARTICULATE MATTER CONCENTRATIONS AND PERSONAL EXPOSURE A recent NRC report on health-related priorities for EPA research on PM noted that, based on current information, ambient PM concentrations and their associated health effects could not be linked with actual human exposures that include indoor environments (NRC, 1998). Individual indoor exposures to PM have consistently differed from estimates based on corresponding outdoor concentrations because of the variable contributions of outdoor and indoor environments, indoor sources of PM, and the ''personal cloud" (PM concentrations recorded by individual personal monitors that exceed readings by fixed monitors located either indoors or outdoors) (EPA, 1996; Ozkaynak et al., 1996). Personal exposure information is also lacking for subpopulations that may be highly susceptible to air pollution. Several studies, in conjunction with high-quality ambient modeling, are currently either under way or in development (EPA, 1998b). Because larger particles tend to be captured more effectively by existing particulate emission control devices, primary PM emissions from power plants are generally in the fine-particle range, often consisting of glassy spheres with high aluminum and silicon content. These fly-ash particles may also be coated with trace metals, such as selenium, arsenic, nickel, or vanadium; mercury and the majority of selenium are emitted primarily as vapors. Selenium has been used as an atmospheric tracer for coal combustion, as has arsenic, because there are few other sources of airborne selenium (Eldred, 1997). Vanadium and nickel are tracers for residual oil combustion. Electric utilities are responsible for a large portion of the secondary PM in the atmosphere, notably sulfates in the eastern part of the United States. The role of power plants in the formation of other secondary particles, such as nitrates or organics, has not been quantified. Of the primary and secondary particles from coal combustion, sulfates have been studied the most, partly because of the relative ease of measuring them. Personal exposure studies of sulfate have found no excess in the "personal cloud" (Ozkaynak et al., 1996), and indoor sulfate levels are equal to or lower than outdoor levels (Suh et al., 1992, 1993). Furthermore, the reduction in indoor levels has been found to be greater in homes with air conditioning, which extends the residence times of indoor air allowing time for increased sulfate deposition on indoor surfaces (Suh et al., 1992). However, this type of information is generally not available for other types of secondary particles (e.g., nitrates) or for primary particles (e.g., fly ash) from coal combustion. The limited data on trace metals suggest that personal exposure to primary PM depends on particle size (Ozkaynak et al., 1996) and is higher for people with occupations that take them outside and near traffic sources (Riveros-Rosas et al., 1997). In general, outdoor exposures are higher than indoor exposures, as would be expected in the absence of indoor sources. No data are available for indoor or personal exposure to selenium (partly because of the difficulty of measuring low concentrations on lightly loaded filters), and only one study has been done on exposure to arsenic (Pellizzari et al., 1999). As a result, the present database is inadequate for estimating the contributions of primary particles from coal combustion to personal exposures to primary PM. Because most epidemiologic studies must rely on ambient air quality data rather than individual personal measures that constitute actual exposure, these studies are affected by "measurement errors" (the difference between ambient and average personal measures). These errors can have various statistical implications. EFFECTIVENESS OF CONTROL TECHNOLOGIES An emission-control program for reducing emissions of NOx from coal-fired power plants would help the utility industry make the transition from existing units, which are 50 years old on average, to power plants that can meet new and proposed stringent local and federal emissions standards. Recently imposed regulations (e.g., Title IV requirements of the Clean Air Act Amendments and the 22-state SIPs) and regulations that may emerge in the next decade (e.g., concerns related to ambient fine particulates, ozone, visibility, acidification, and eutrophication) will require that a large number of the coal-fired power plants operating today meet more stringent NOx controls. Regulatory requirements for controlling emissions of NOx from steam-electric power plants have increased rapidly. The technological approaches for meeting the requirements of Phase I, Title IV, emission limits have been low-NOx burners, reburning, and other combustion controls. However, in the present regulatory environment, higher performance low-NOx burners will be required to meet Phase II, Title IV, emission limits. Moreover, with the imposition of the NOx SIP requirements, combustion controls for reducing NOx
OCR for page 20
Review of the U.S. Department of Energy Office of Fossil Energy's Research Plan for Fine Particulates emissions will have to be supplemented by post-combustion controls. Advanced cost-effective NOx control technologies that can be retrofitted to existing coal-fired electric utility boilers and are capable of meeting a target NOx emission limit of 0.15 lb NOx/million Btu will have to be developed. Although, in general, currently available selective catalytic reduction control technology can achieve this emission limit, high capital costs and the lack of long-term availability of catalysts are major concerns. In most cases, existing control technologies for the removal of sulfur dioxide will be adequate to meet future mandated reduction goals. However, retrofitting existing equipment could pose problems for many utilities because of space limitations, and innovative retrofitting concepts will be required for these sites. Currently, coal-fired electric utility boilers built or modified after August 17, 1971, must comply with a New Source Performance Standard (NSPS) limit on primary particulate emissions of 0.10 lb/million Btu. Units built or modified after September 18, 1978, must comply with a more stringent NSPS of 0.03 lb/million Btu. At present, average primary particulate emissions from all coal-fired utility boilers are about 0.043 lb/million Btu (FETC, 1999c). To comply with current PM emission limits, the majority of coal-fired electric utility boilers control primary particulates with electrostatic: precipitators, which use electric fields to remove particulates from boiler flue gas. A smaller but growing number of boilers employ fabric filter collectors (bag houses), which control PM by passing the flue gas through a tightly woven fabric that collects the particles in the form of dust cake. When operating properly, state-of-the-art electrostatic precipitators and bag houses can achieve overall collection efficiencies of 99.9 percent of total primary particulate mass to below the 1978 NSPS of 0.03 lb/million Btu. However, even with high-performance particulate control systems, collection is less efficient for particle sizes of less than 1 μm. Particles in the 0.1 to 0.5 μm size range are an especially challenging problem for stationary sources that employ wet scrubbers and electrostatic precipitators. Moreover, more than half of the existing electrostatic precipitators installed on electric-utility boilers in the United States have been in operation for more than 30 years, and almost 17 percent for 40 years. These older models of electrostatic precipitators are less efficient than more recent models. Potential national strategies for reducing PM may include limiting coal-fired units to a more stringent emissions rate than the current 1978 NSPS. In the near term, it may be possible to improve the operation and maintenance of existing particulate control technologies and achieve higher overall collection efficiencies. However, strategic research, development, and testing of efficient, cost-effective, innovative PM control technologies, processes, and concepts may be necessary. If so, new fine particulate control technologies are likely to require long development times.
Representative terms from entire chapter: