Diesel Particulate Matter Number Emissions: Evaluation of Existing Modal Emission Modeling Approaches

Recent health and exposure studies have indicated that emissions and air quality standards of particulate matters (PM) numbers could be beneficial since they account for the more harmful ultrafine particles that have negligible mass. This study examines existing modal emission modeling approaches [e.g., polynomial regressions, fuel rate-based approach, and vehicle specific power (VSP)-based approach] in order to evaluate their ability of predicting diesel particle number concentration emitted from a transit diesel bus. This is a necessary and useful first step to quantifying vehicular emissions in terms of PM number. The existing modeling approaches were applied to and then the technique of bootstrapping was used on the onboard measurement data of PM number concentrations emitted from a diesel transit bus. The predictive ability of the models with respect to particle number concentration is assessed by examining the statistical significance of the key regressors used in those approaches and the adjusted $R$ -squared values of the models. Results indicate less than satisfying predictive capability of the examined approaches for PM number emission that behaves in a different manner from gaseous emissions. The study, however, demonstrates that vehicle operating factors such as VSP, speed, and acceleration are significant in explaining the variability of transient PM number concentration measurements from the bus.