Source identification of sub-10 nm particles through air dispersion modeling

Previously reported rooftop ambient aerosol measurements in Raleigh, NC, USA, detected episodic events where sub-10 nm particle number concentrations (PNC) exceeded 3.73 × 10⁵ cm^-3. Their small size and temporally stable modal diameter (sometimes persisting for days) indicated origins from nearby primary emission sources rather than mesoscale new particle formation (NPF) events. To investigate potential sources, simulations were conducted using the U.S. Environmental Protection Agency's Gaussian plume-based model, AERMOD. Campus surveys and Google Earth analyses identified three candidate sources near the measurement site, including two combined heat and power (CHP) facilities with high-efficiency natural gas turbines and heat recovery steam generators that provide energy to NC State's campus. Distinct point sources were modeled for each facility using an emission factor of 5 × 10^-4 g s^-1. The study explored source contributions under varying micrometeorological conditions (e.g., wind speed, wind direction, solar radiation, and planetary boundary layer height). Wind pattern analysis revealed distinct plumes from individual power plants reaching the receptor site. Statistical analyses confirmed wind direction and speed as the strongest predictors of modeled mass concentrations, and that observed PNC profiles during NPF and particle burst events are fundamentally distinct. Exceptionally high sub-10 nm particle growth rates were observed during plume transport, averaging 104-120 nm hr^-1. These findings reveal that expanding deployment of CHPs for distributed power generation may pose unrecognized health risks through sub-10 nm particle emissions with demonstrated respiratory and neurological impacts. New emission standards may be needed to address ultrafine particle production from natural gas combustion technologies.