Measurement of fine particle emissions from combustion sources: Design and evaluation of dilution samplers and the effects of sampling conditions

In 1997 the U.S. Environmental Protection Agency (EPA) promulgated new standards to address ambient air concentrations of particulate matter with an aerodynamic diameter of 2.5 micrometers (PM2.5) because of concerns over human health effects associated with fine atmospheric particles. It is important to understand these transformations in order to relate measured emissions to ambient air quality. Dilution sampling is used to simulate the effects of cooling and dilution on fine particle emissions that occur as the exhaust mixes with ambient air. This dissertation examines the effects of dilution sampling on fine particle emissions from combustion sources, a major source of PM in the atmosphere.; Dilution samples of a pilot-scale coal combustor, a diesel engine, and a wood stove were analyzed for fine particle emissions. Emissions from coal combustion are dominated by inorganic species. PM2.5 mass emission rates were independent of dilution ratio or residence time, and Selenium was enriched in diluted filter measurements when compared with hot filter measurements. Results from wood combustion and a diesel engine show emissions dominated by organic species. Fine particle emission rates vary strongly with dilution ratio when sampling from a combustion source emitting high levels of organic material, due to shifts in the phase partitioning of the semi-volatile species. As the emissions are diluted, semi-volatile species are forced into the gas phase to maintain equilibrium. An absorptive partitioning model was used to explain the changes in fine particle mass emission rates with dilution ratio. This effect can lead to over-estimations of particle emissions as large as 50% at low dilution ratios (∼20:1) when measuring from combustion sources such as gasoline/diesel engines and biomass burning using standard emissions measurement protocols. Organic carbon emission rate measurements can be severely affected by the collection of sampling artifact on filter measurements. Back-up and a denuder filter measurements are used to demonstrate that significant positive artifact is collected on bare-quartz filters leading to over-estimates of carbon emissions. Measurements for all three sources demonstrated that filter measurements of fine particle mass, OC, and EC are independent of residence time, eliminating the need for a residence time chamber.