No abstract

No abstract

This paper presents a comprehensive set of ultrafine particles (UFPs) emission factors (EFs) for heavy duty vehicles (HDVs) as a function of vehicle flow rate, speed, and mode of operation (free flow and congestion) using 664 measurements of UFPs, carbon dioxide (CO 2), meteorology and traffic conditions near a major roadway (average daily traffic 300,000 day-1). 5-min samples were collected for 2 to 3 hour time period on 60 days between 2015 and 2018. The average traffic-induced concentration of UFPs was 11,300 pt cm-3 for free flow and 12,400 pt cm-3 for congestion. Results demonstrate that HDVs produce significantly more dispersion (30x) than light duty vehicles (LDVs). The additional dispersion from HDVs results in the minimum pollutant concentrations occurring at the highest vehicle flow rate. EFs for UFPs are determined using inverse modeling based on the calculated CO 2 dispersion. This eliminates the need to rely on air-quality models to estimate dispersion. The EFs for HDVs range from 4 × 10 14 to 20 × 10 14 (pt km-1 veh-1). The variations in EFs are correlated with variations in vehicle flow rate and speed. The average UFP EFs for HDVs are significantly higher (3x) for congestion compared to free flow. UFP EFs for HDVs are more sensitive to speed in congestion compare to in free flow conditions. Thus, even a moderate increase in HDVs speed or mitigation of congestion will have a significant impact on lowering UFP concentrations.

A parameterization of initial vertical dispersion coefficient (σ z,init) was developed for incorporation into California line source dispersion model, version 4 (CALINE4) and AMS/EPA regulatory model (AERMOD) to better predict pollutant concentrations near roadways. The momentum wake theory of moving vehicles indicates that both vehicle-induced turbulence (VIT) and dispersion occur in the vehicle wake. Based on a literature review, it is postulated that σ z,init near roadways can be estimated using a “wake area model” concept of effective wake area defined as the vehicle height times the wake length, vehicle density, and vehicle type. A total of 523 5-min near-roadway simultaneous measurements (2016–2018) of pollutant concentrations and meteorological and traffic information were used to evaluate the model. Two roadways with distinct fleet composition and simple road configurations were selected for monitoring. The near-roadway σ z,init ranged from 1 to 4 m for light-duty vehicles (LDVs) and from 3 to 7 m for fleet-mix (LDVs and heavy-duty vehicles (HDVs)). The results demonstrate that the dispersion contribution from one HDV was 31 times larger than that from one LDV. Calculated pollutant dispersion using the wake area model compared favorably with measurements (R 2 = 0.91, slope = 1.07). These results indicate that σ z,init varies with vehicle density and HDVs. Pollutant dispersion related to the vehicle wakes can be used to correctly parameterize dispersion models and improve prediction of pollutant concentrations near roadways.