Date of Award


Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Engineering and Public Policy


Allen Robinson


Combustion emissions from on-road sources such as light duty gasoline vehicles (LDGV), medium duty diesel vehicles (MDDV) and heavy duty diesel vehicles (HDDV) as well as small off-road engines (SORE) such those used in lawn and garden equipment are a major source of fine particulate matter (PM) pollution in the ambient atmosphere. Existing regulations have restricted direct PM emissions, especially for on-road sources; however, recent studies suggest that organic PM formed from the photo-oxidation of gaseous precursor emissions—so-called secondary organic aerosol (SOA)—contributes at least as much to the overall PM burden as PM “emitted from the tailpipe.” A major limitation of many of these studies is that they attempt to induce from the behavior of simple emission surrogates (e.g., vaporized whole fuel) the behavior of actual combustion emissions from real world sources.

This research investigates combustion emissions directly. The primary gas- and particle-phase emissions, SOA production and SOA yields from a range of different on-road and off-road combustion sources were characterized. LDGV, MDDV and HDDV were driven on chassis dynamometers over realistic, urban driving cycles. Off-road sources, including 2- and 4-stroke lawn and garden equipment and a diesel transportation refrigeration unit were tested using engine dynamometers operated over certification cycles.

For nearly all gasoline engines (LDGV and SOREs), photo-oxidizing dilute combustion emissions for 3 hours produced at least as much SOA as the directly emitted primary PM. SOA increased net PM production for LDGV by a factor of 1-10, depending on the vehicle emission standard. SOA yields were found to increase with newer vehicles, which have lower primary emissions. SOA for diesel vehicles, while still large on an absolute basis, was a smaller fractionof the primary PM emissions (between 10-30%), due to the very high elemental carbon (EC) emissions from vehicles without diesel particulate filters (DPF). Aftertreatment systems utilizing a DPF and a diesel oxidation catalyst essentially eliminated primary PM and SOA. Among the off-road sources, SOA from 2-stroke emissions increased the net PM by roughly a factor of 2.

Primary emission and SOA production factors from the various combustion sources tested in this work were combined with fuel consumption data for California’s South Coast Air Basin (SoCAB) to determine the impact on the aggregate PM from on- and off-road sources in the region. These estimated impacts were compared to the PM values derived from the regulatory models EMFAC and OFFROAD. Our data indicate that PM from on-road gasoline sources is significantly under-represented by existing emissions models due to the dominant role of SOA. When SOA production is included, newer LDGV are one of the largest sources of PM in SoCAB. Furthermore, LDGV will become an even more important PM source once existing regulations requiring DPF retrofits on both on- and off-road diesel sources are implemented over the next few years. While on-road diesel vehicles are currently an important source of PM based on total fuel consumption, LDGV are responsible for a larger fraction of total PM. The primary and secondary PM contribution of off-road sources also appear to be high, but the magnitude remains highly uncertain, pending further experimental data. Evidence is presented that suggests that existing models may dramatically overpredict primary PM emissions from off-road sources