Elucidating long-term trends, seasonal variability, and local impacts from thirteen years of near-road particle size data (2006-2019).

Significant attention, especially in the last decade, has been focussed on elevated concentrations of ultrafine particulate matter (UFP) in urban areas and the adverse health effects associated with exposure to UFP. Despite this, there is a relative scarcity of long-term ambient UFP measurements. This study examined trends in UFP measurements made continuously near a busy roadway in downtown Toronto, Canada, between the years 2006 and 2019 using a fast mobility particle sizer (FMPS). These long-term trends were associated with other air pollutant concentrations-namely: nitric oxide (NO), nitrogen dioxide (NO2), sulphur dioxide (SO2), and fine particulate matter mass concentrations (PM2.5)-and persistent declining trends were observed for each during the study period. From 2006 to 2019, reductions of 45%, 68%, 39%, 83%, and 41%, for UFP, NO, NO2, SO2, and PM2.5, respectively, were observed. These reductions are in part associated with a total phase-out of coal-fired electricity generation in Ontario, Canada, between 2004 and 2015, and continuous improvements in vehicle emissions control technologies. Additionally, deconvolution of the time-series yielded seasonal fluctuations which were analysed as a function of particle diameter and ambient temperature, the results from which may aid in the comparison of UFP measurements made in climates with different ambient temperature ranges in a meaningful way. Finally, the UFP data were background-subtracted and it was found that local sources (such as vehicle traffic) contributed ~45% to total concentrations and this fraction remained relatively constant throughout the study. A multilinear function regressed on these local and background concentrations better elucidated the sources contributing to UFP variability-background concentrations were largely covariate with SO2 emissions whereas local concentrations were more affected by NO emissions. The data in this study shows clear co-benefits to reducing UFP concentrations by targeting NOx and SOx emissions.

Quantifying metal emissions from vehicular traffic using real world emission factors.

Road traffic emissions are an increasingly important source of particulate matter in urban and non-road environments, where non-tailpipe emissions can contribute substantially to elevated levels of metals associated with adverse health effects. Thus, better characterization and quantification of traffic-emitted metals is warranted. In this study, real-world emission factors for fine particulate metals were determined from hourly x-ray fluorescence measurements over a three-year period (2015-2018) at an urban roadway and busy highway. Inter-site differences and temporal trends in real-world emission factors for metals were explored. The emission factors at both sites were within the range of past studies, and it was found that Ti, Fe, Cu, and Ba emissions were 2.2-3.0 times higher at the highway site, consistent with the higher proportion of heavy-duty vehicles. Weekday emission factors for some metals were also higher by 2.0-3.5 times relative to Sundays for Mn, Zn, Ca, and Fe, illustrating a dependence on fleet composition and roadway activity. Metal emission factors were also inversely related to relative humidity and precipitation, due to reduced road dust resuspension under wetter conditions. Correlation analysis revealed groups of metals that were co-emitted by different traffic activities and sources. Determining emission factors enabled the isolation of traffic-related metal emissions and also revealed that human exposure to metals in ambient air can vary substantially both temporally and spatially depending on fleet composition and traffic volume.

Evaluation of a Passive Sampling Method for Long-Term Continuous Monitoring of Volatile Organic Compounds in Urban Environments.

Environmental Protection Agency Method 325 was developed for continuous passive monitoring of volatile organic compounds (VOCs), particularly benzene, at petroleum refinery fencelines. In this work, a modified version of the method was evaluated at an Ontario near-road research station in winter to assess its suitability for urban air quality monitoring. Samples were collected at 24 hour and 14 day resolution to investigate accuracy for different exposure times. Tubes were analyzed by thermal desorption-gas chromatography-mass spectrometry, and 11 VOCs were quantified, including aromatic air toxics. The same VOCs were simultaneously monitored using traditional canister sampling for comparison, and a subset of four were also monitored using a novel miniature gas chromatograph. Good agreement (within 10%) was observed between the 14 day passive tube samples and the canister samples for benzene. However, field-calibrated uptake rates were required to correct passive tube concentrations for less volatile aromatics. Passive tube deployment and analysis is inexpensive; sampling does not require power, and accurate measurements of benzene are demonstrated here for an urban environment. The method is expected to be advantageous for the generation of long-term continuous benzene datasets suitable for epidemiological research with greater spatial coverage than is currently available using traditional monitoring techniques.

Near-Road Air Pollutant Measurements: Accounting for Inter-Site Variability Using Emission Factors.

A daily integrated emission factor (EF) method was applied to data from three near-road monitoring sites to identify variables that impact traffic related pollutant concentrations in the near-road environment. The sites were operated for 20 months in 2015-2017, with each site differing in terms of design, local meteorology, and fleet compositions. Measurement distance from the roadway and local meteorology were found to affect pollutant concentrations irrespective of background subtraction. However, using emission factors mostly accounted for the effects of dilution and dispersion, allowing intersite differences in emissions to be resolved. A multiple linear regression model that included predictor variables such as fraction of larger vehicles (>7.6 m in length; i.e., heavy-duty vehicles), vehicle speed, and ambient temperature accounted for intersite variability of the fleet average NO, NO x, and particle number EFs (R2:0.50-0.75), with lower model performance for CO and black carbon (BC) EFs (R2:0.28-0.46). NO x and BC EFs were affected more than CO and particle number EFs by the fraction of larger vehicles, which also resulted in measurable weekday/weekend differences. Pollutant EFs also varied with ambient temperature and because there were little seasonal changes in fleet composition, this was attributed to changes in fuel composition and/or post-tailpipe transformation of pollutants.

Real-World Emission of Particles from Vehicles: Volatility and the Effects of Ambient Temperature.

A majority of the ultrafine particles observed in real-world conditions are systematically excluded from many measurements that help to guide regulation of vehicle emissions. To investigate the impact of this exclusion, coincident near-road particle number (PN) emission factors were quantified up- and downstream of a thermodenuder during two seasonal month-long campaigns with wide-ranging ambient temperatures (-19 to +30 °C) to determine the volatile fraction of particles. During colder temperatures (<0 °C), the volatile fraction of particles was 94%, but decreased to 85% during warmer periods (>20 °C). Additionally, mean PN emission factors were a factor of 3.8 higher during cold compared to warm periods. On the basis of 130 000 vehicle plumes including three additional campaigns, fleet mean emission factors were calculated for PN (8.5 × 1014 kg-fuel-1), black carbon (37 mg kg-fuel-1), organic aerosol (51 mg kg-fuel-1), and particle-bound polycyclic aromatic hydrocarbons (0.7 mg kg-fuel-1). These findings demonstrate that significant differences exist between particles in thermally treated vehicle exhaust as compared to in real-world vehicle plumes to which populations in near-road environments are actually exposed. Furthermore, the magnitude of these differences are dependent upon season and may be more extreme in colder climates.

Assessing the Climate Trade-Offs of Gasoline Direct Injection Engines.

Compared to port fuel injection (PFI) engine exhaust, gasoline direct injection (GDI) engine exhaust has higher emissions of black carbon (BC), a climate-warming pollutant. However, the relative increase in BC emissions and climate trade-offs of replacing PFI vehicles with more fuel efficient GDI vehicles remain uncertain. In this study, BC emissions from GDI and PFI vehicles were compiled and BC emissions scenarios were developed to evaluate the climate impact of GDI vehicles using global warming potential (GWP) and global temperature potential (GTP) metrics. From a 20 year time horizon GWP analysis, average fuel economy improvements ranging from 0.14 to 14% with GDI vehicles are required to offset BC-induced warming. For all but the lowest BC scenario, installing a gasoline particulate filter with an 80% BC removal efficiency and <1% fuel penalty is climate beneficial. From the GTP-based analysis, it was also determined that GDI vehicles are climate beneficial within <1-20 years; longer time horizons were associated with higher BC scenarios. The GDI BC emissions spanned 2 orders of magnitude and varied by ambient temperature, engine operation, and fuel composition. More work is needed to understand BC formation mechanisms in GDI engines to ensure that the climate impacts of this engine technology are minimal.

Field Measurements of Gasoline Direct Injection Emission Factors: Spatial and Seasonal Variability.

Four field campaigns were conducted between February 2014 and January 2015 to measure emissions from light-duty gasoline direct injection (GDI) vehicles (2013 Ford Focus) in an urban near-road environment in Toronto, Canada. Measurements of CO2, CO, NOx, black carbon (BC), benzene, toluene, ethylbenzene-xylenes (BTEX), and size-resolved particle number (PN) were recorded 15 m from the roadway and converted to fuel-based emission factors (EFs). Other than for NOx and CO, the GDI engine had elevated emissions compared to the Toronto fleet, with BC EFs in the 73rd percentile, BTEX EFs in the 80-90th percentile, and PN EFs in the 75th percentile during wintertime measurements. Additionally, for three campaigns, a second platform for measuring PN and CO2 was placed 1.5-3 m from the roadway to quantify changes in PN with distance from point of emission. GDI vehicle PN EFs were found to increase by up to 240% with increasing distance from the roadway, predominantly due to an increasing fraction of sub-40 nm particles. PN and BC EFs from the same engine technology were also measured in the laboratory. BC EFs agreed within 20% between the laboratory and real-world measurements; however, laboratory PN EFs were an order of magnitude lower due to exhaust conditioning.