Real-World Vehicle Emissions Characterization for the Shing Mun Tunnel in Hong Kong and Fort McHenry Tunnel in the United States.
Xiaoliang Wang,Andrey Khlystov,Kin-Fai Ho,Dave Campbell,Judith C Chow,Steven D Kohl,John G Watson,Shun-Cheng Frank Lee,Lung-Wen Antony Chen,Minggen Lu,Steven Sai Hang Ho
2019-01-01
Abstract:INTRODUCTION:Motor vehicle exhaust is an important source of air pollutants and greenhouse gases. Concerns over the health and climate effects of mobile-source emissions have prompted worldwide efforts to reduce vehicle emissions. Implementation of more stringent emission standards have driven advances in vehicle, engine, and exhaust after-treatment technologies as well as fuel formulations. On the other hand, vehicle numbers and travel distances have been increasing because of population and economic growth and changes in land use. These factors have resulted in changes to the amount and chemical composition of vehicle emissions.Roadway tunnel studies are a practical way to characterize real-world emissions from the on-road vehicle fleet in an environment isolated from other combustion pollution sources. Measurements in the same tunnel over time allow evaluation of vehicle emission changes and the effectiveness of emission reduction measures. Tunnel studies estimate the impacts of vehicle emissions on air quality and traffic-related exposures, generate source profile inputs for receptor-oriented source apportionment models, provide data to evaluate emission models, and serve as a baseline for future comparisons.The present study characterized motor vehicle emission factors and compositions in two roadway tunnels that were first studied over a decade ago. The specific aims were to (1) quantify current fleet air pollutant emission factors, (2) evaluate emission change over time, (3) establish source profiles for volatile organic compounds (VOCs) and particulate matter ≤2.5 μm in aerodynamic diameter (PM2.5), (4) estimate contributions of fleet components and non-tailpipe emissions to VOCs and PM2.5, and (5) evaluate the performance of the latest versions of mobile-source emission models (i.e., the EMission FACtors vehicle emission model used in Hong Kong [EMFAC-HK] and the MOtor Vehicle Emission Simulator used in the United States [MOVES]).METHODS:Measurements were conducted in the Shing Mun Tunnel (SMT) in Hong Kong and the Fort McHenry Tunnel (FMT) in Baltimore, Maryland, in the United States, representing the different fleet compositions, emission controls, fuels, and near-road exposure levels found in Hong Kong and the United States. These tunnels have extensive databases acquired in 2003-2004 for the SMT and 1992 for the FMT. The SMT sampling was conducted during the period from 1/19/2015 to 3/31/2015, and the FMT sampling occurred during the periods from 2/8/2015 to 2/15/2015 (winter) and 7/31/2015 to 8/7/2015 (summer). Concentrations of criteria pollutants (e.g., carbon monoxide [CO], nitrogen oxides [NOx], and particulate matter [PM]) were measured in real time, and integrated samples of VOCs, carbonyls, polycyclic aromatic hydrocarbons (PAHs), and PM2.5 were collected in canisters and sampling media for off-line analyses. Emission factors were calculated from the tunnel measurements and compared with previous studies to evaluate emission changes over time. Emission contributions by different vehicle types were assessed by source apportionment modeling or linear regression. Vehicle emissions were modeled by EMFAC-HK version 3.3 and MOVES version 2014a for the SMT and the FMT, respectively, and compared with measured values. The influences of vehicle fleet composition and environmental parameters (i.e., temperature and relative humidity) on emissions were evaluated.RESULTS:In the SMT, emissions of PM2.5, sulfur dioxide (SO2), and total non-methane hydrocarbons (NMHCs) markedly decreased from 2003-2004 to 2015: SO2 and PM2.5 were reduced by ~80%, and total NMHCs was reduced by ~44%. Emission factors of ethene and propene, key tracers for diesel vehicle (DV) emissions, decreased by ~65%. These reductions demonstrate the effectiveness of control measures, such as the implementation of low-sulfur fuel regulations and the phasing out of older DVs. However, the emission factors of isobutane and n-butane, markers for liquefied petroleum gas (LPG), increased by 32% and 17% between 2003-2004 and 2015, respectively, because the number of LPG vehicles increased. Nitrogen dioxide (NO2) to NOx volume ratios increased between 2003-2004 and 2015, indicating an increased NO2 fraction in primary exhaust emissions. Although geological mineral concentrations were similar between the 2003-2004 and 2015 studies, the contribution of geological materials to PM2.5 increased from 2% in 2003-2004 to 5% in 2015, signifying the continuing importance of non-tailpipe PM emissions as tailpipe emissions decrease. Emissions of CO, ammonia (NH3), nitric oxide (NO), NO2, and NOx, as well as carbonyls and PAHs in the SMT did not show statistically significant (at P < 0.05 based on Student's t-test) decreases from 2003-2004 to 2015. The reason for this is not clear and requires further investigation.A steady decrease in emissions of all measured pollutants during the past 23 years has been observed from tunnel studies in the United States, reflecting the effect of emission standards and new technologies that were introduced during this period. Emission reductions were more pronounced for the light-duty (LD) fleet than for the heavy-duty (HD) fleet. In comparison with the 1992 FMT study, the 2015 FMT study demonstrated marked reductions in LD emissions for all pollutants: emission factors for naphthalene were reduced the most, by 98%; benzene, toluene, ethylbenzene, and xylene (BTEX), by 94%; CO, NMHCs, and NOx, by 87%; and aldehydes by about 71%. Smaller reductions were observed for HD emission factors: naphthalene emissions were reduced by 95%, carbonyl emissions decreased by about 75%, BTEX by 60%, and NOx 58%.The 2015 fleet-average emission factors were higher in the SMT for CO, NOx, and summer PM2.5 than those in the FMT. The higher CO emissions in the SMT were possibly attributable to a larger fraction of motorcycles and LPG vehicles in the Hong Kong fleet. DVs in Hong Kong and the United States had similar emission factors for NOx. However, the non-diesel vehicles (NDVs), particularly LPG vehicles, had higher emission factors than those of gasoline cars, contributing to higher NOx emissions in the SMT. The higher PM2.5 emission factors in the SMT were probably attributable to there being more double-deck buses in Hong Kong.In both tunnels, PAHs were predominantly in the gas phase, with larger (four and more aromatic rings) PAHs mostly in the particulate phase. Formaldehyde, acetaldehyde, crotonaldehyde, and acetone were the most abundant carbonyl compounds in the SMT. In the FMT, the most abundant carbonyls were formaldehyde, acetone, acetaldehyde, and propionaldehyde. HD vehicles emitted about threefold more carbonyl compounds than LD vehicles did. In the SMT, the NMHC species were enriched with marker species for LPG (e.g., n-butane, isobutane, and propane) and gasoline fuel vapor (e.g., toluene, isopentane, and m/p-xylene), indicating evaporative losses. Source contributions to SMT PM2.5 mass were diesel exhaust (51.5 ± 1.8%), gasoline exhaust (10.0 ± 0.8%), LPG exhaust (5.0 ± 0.5%), secondary sulfate (19.9 ± 1.0%), secondary nitrate (6.3 ± 0.9%), and road dust (7.3 ± 1.3%). In the FMT, total NMHC emissions were 14% and 8% higher in winter than in summer for LD and HD vehicles, respectively. Elemental carbon (EC) and organic carbon (OC) were the major constituents of tunnel PM2.5. De-icing salt contributions to PM2.5 were observed in the FMT in winter.Emission estimates by the EMFAC-HK agreed with SMT measurements for CO2; the modeled emission factors for CO, NOx, and NMHCs were 1.5, 1.6, and 2.2 times the measurements, respectively; and the modeled emission factor for PM2.5 was 61% of the measured value in 2003. The EMFAC-HK estimates and SMT measurements for 2015 differed by less than 35%. The MOVES2014a model generally overestimated emissions of most of the pollutants measured in the FMT. No pollutants were significantly underestimated. The largest overestimation was observed for emissions measured during HD-rich driving conditions in winter.CONCLUSIONS:Significant reductions in SO2 and PM2.5 emissions between 2003 and 2015 were observed in the SMT, indicating the effectiveness of control measures on these two pollutants. The total NMHC emissions in the SMT were reduced by 44%, although isobutane and n-butane emissions increased because of the increase in the size of the LPG fleet. No significant reductions were observed for CO and NOx, results that differed from those for roadside ambient concentrations, emission inventory estimates, and EMFAC-HK estimates. In contrast, there was a steady decrease in emissions of most pollutants in the tunnels in the United States.