PM2.5 From Gas Combustion: Emissions Characterization Issues and Speciation Results

摘要

Reliable PM2.5 source apportionment will require accurate inventories of speciated emissionsrnfrom stationary sources. Although gas combustion is typically thought of as a "clean" process, thernenormous quantities of gas burned in boilers, process heaters, and engines potentially makes evenrnsmall pollutant concentrations significant. Tests were performed on a gas-fired industrial boilerrnand a gas-fired process heater at different oil refineries to characterize air emissions contributingrnto fine particulate matter in the ambient air.rnTotal particulate, PM10 and PM2.5 were measured using in-stack cyclones followed by an instackrnquartz filter at stack temperature (EPA Method 201A). Downstream of the filter, thernsample was passed through a series of impingers in an ice bath (EPA Method 202) to collect bothrnfine particles which pass through the filter and aerosols which condense at the temperature of therngas leaving the impingers (typically 60-70°F). PM2.5 mass and chemical speciation also wererndetermined using a dilution tunnel and ambient air sampling/analytical methods. The dilutionrntunnel included an aging chamber which provided a total system bulk residence time ofrnapproximately 80 seconds to ensure representative sampling of organic aerosols.rnThe amount of filterable particulate collected using the in-stack methods was essentially near orrnbelow the overall method detection limit. However, condensable particulate matter (CPM) asrndefined by EPA Method 202 was measured at much higher levels. Analysis of the impingerrncontents from the boiler test suggests a large fraction of the measured CPM can be accounted forrnin sulfates and chlorides; however, the mass of sulfates in the impingers was much greater than inrnthe dilution tunnel samples. Laboratory results indicate most of this can be accounted for byrnoxidation of gaseous SO2 in the impingers leading to a positive bias in the results. These resultsrnindicate the potential for a large positive bias in the existing emission inventory for PM2.5rn(which is based on EPA methods) from gas fired stationary sources.