Characterization of Utility Boiler Contributions to PM2.5

ABSTRACT
Emissions from coal-fired utility boilers may contribute significantly to ambient particulate concentrations, especially in the very fine (below 2.5 micron) particle size range. Ambient PM2.5 speciation studies often show that geologic materials, which comprise the majority of fly ash produced by coal combustion, make up only a small fraction of the particulate. Sulphate, nitrates, and organic compounds typically comprise more than 80 percent of the ambient PM2.5. Therefore emission of secondary aerosol precursors (e.g., nitrogen oxides, sulfur dioxide, sulfur trioxide, ammonia, certain volatile organic compounds) contributes significantly to ambient PM2.5. Coal-fired boilers emit both primary PM2.5 and secondary particle precursors. Because of the large quantity of gaseous fuels burned, even gas-fired devices may have potential to contribute significantly to ambient PM2.5.

The chemical speciation of primary particles emitted from coal-fired boilers provides important markers that can be used to identify the relative contribution of a specific category of sources to regional ambient PM2.5. Existing emission factors and speciation profiles for PM2.5 and PM2.5 precursors from combustion sources are often dated and incomplete. Also, the wide variety of coal mineral matter compositions, boiler designs and pollution control equipment makes site-specific emissions characterization the most reliable manner of attributing source contributions. A test protocol for developing PM2.5 emission factors and chemical speciation profiles has been developed. The paper discusses approaches for characterization of PM2.5 and PM2.5 precursors from stationary combustion sources and preliminary results of field measurements made using traditional and dilution sampling techniques. Results of tests on gas-fired units using both traditional source testing methods and a dilution sampling approachare presented and compared.

INTRODUCTION
In July 1997, the U. S. Environmental Protection Agency (USEPA) promulgated new National Ambient Air Quality Standards (NAAQS) for fine particulate matter and ozone, including a new standard for particles 2.5 mm or less in diameter, referred to collectively as PM2.5. Although this standard was recently remanded by the U. S. courts, the EPA intends to appeal the decision and it is expected that this represents only a short delay in the implementation of the new NAAQS. USEPA is continuing to implement a national network to monitor and speciate ambient PM2.5 while litigation continues. In Canada, the development of Canada-wide ambient air quality standards for PM2.5 actively continues with passage in the foreseeable future likely¹. In Europe, the European Commission (EC) has proposed tighter standards for PM10 and is considering new legislation on PM2.5².

Chemical speciation of PM2.5 emissions provides important markers for determining the contribution of source categories to ambient PM2.5. Also, it is widely believed that only certain components of PM2.5 cause the various adverse human health effects that have been observed. Chemical analysis of ambient PM2.5 samples collected in various parts of the United States and Canada show that sulfates, nitrates, carbon (elemental and organic) dominate in most urban and many non-urban areas; ammonium and mineral (soil) elements also are present^3,4. Organic compounds are important components of particulate matter and most of the particulate organic carbon is believed to reside in the fine particle fraction⁵. For example, in a study of the Los Angeles area organic compounds constituted approximately 30 percent of the ambient fine particle mass⁶.

Particles may be either directly emitted into the atmosphere (primary particulate) or formed there by chemical reactions and physical transformations (secondary particulate). The majority of primary emissions from combustion are found in the PM2.5 or smaller size range, especially with clean burning fuels, such as gas. Sulfates and nitrates are the most common secondary particles, although organic carbon can also result from volatile organic compounds(VOCs)⁷. The gaseous (SO₂) and sulfur trioxide (SO₃); oxides of nitrogen (NO and NO2, the sum of which is designated NO_X), respectively; and ammonia (NH₃). Secondary organic aerosol formation mechanisms are not well understood due to the multitude of precursors involved and the rates of formation which are heavily dependent on meteorological variables and the concentrations of other pollutants. It is believed, however that atmospheric transformations leading to the formation of secondary aerosol from gas-phase primary organic emissions may be very significant in some areas, particularly during the summertime.

The reliability of studies to apportion the contribution of regional sources to local ambient PM10 and PM2.5 relies to a large extent on having accurate emission inventory and speciation data for model input. The Chemical Mass Balance Model is one common approach to particulate source apportionment, favored for its simplicity. This model relates chemical analysis of ambient air samples to sources based on emission profiles for those sources and other factors⁸. Generic source emission profiles, e.g. those developed by EPA for sourcereceptor modeling, are available for many source categories; however, these must be used with great caution since they may not accurately represent specific sources because of site-specific process differences, data based on measurements using older, less sensitive and selective techniques, and/or incomplete data. The National Research Council Committee on Research Priorities for Airborne Particulate Matter concluded that one of the 10 most important research priorities for fine particulate studies is to "develop advanced mathematical, modeling and monitoring tools to represent the relationships between specific sources of particulate matterand human exposures⁹". Thus, there is a need for new source emission data for specific sources and locations using the latest measurement technologies to provide more reliable source apportionment results¹⁰.