Literature review of Particulate Matter (PM) from transport with a special focus on organic aerosols

Report no. 2/22: Several recent scientific studies on urban air quality have suggested that air quality modelling based on current emission inventories for mobile sources systematically underestimates the contribution of these sources to ambient particulate matter (PM) levels, and organic aerosol levels in particular. This document discusses a number of factors that may explain the reasons for this apparent systematic underestimation. An obvious cause is that current road transport emission factors do not fully account for the contribution of primary organic particulate.

Secondly the emissions, and the chemical activity of precursor gases (in the atmospheric formation of secondary organic aerosol (SOA)) are not yet fully accounted for in the inventories nor in all air quality models. Organic aerosol (OA) can be classified based on the volatility of the compounds it consists of. Organic particulate and condensable particulate matter (CPM) may consist of non-volatile compounds (NVOCs), organic compounds with low volatility (LVOCs), semi-volatile organic compounds (SVOCs) and to some degree organic compounds with intermediate volatility (IVOCs). In addition, gaseous SVOCs, IVOCs and certain NMVOCs are very important classes of secondary organic aerosol (SOA) precursors. S/IVOCs in road transport emissions are therefore central to this document. SOA precursors, such as S/IVOCs, are not distinguished separately in official emission inventories. They may or may not be included in the emissions of PM or hydrocarbons and more specific in those of non-methane volatile organic compounds (NMVOCs).

Due to the lack of specified emission data on SOA precursors, air quality models are adapted to estimate SOA precursor emissions and include modules to simulate the transformation of these precursors to SOA. This document starts with a general description on how these models estimate the emission/concentration of SOA precursor gases and how the resulting SOA formation is estimated. A descriptive analysis is given on the following subjects in order to provide a better understanding on how road transport emission rates for PM are currently assessed and which factors are of influence in this respect:

• Relevant emission legislation for road vehicles;
• Currently used emission control technologies;
• Testing procedures for characterizing PM emissions;
• How PM emissions under testing conditions compare to real-driving cycles;
• The contribution by non-exhaust emission of PM;
• Fleet composition and renewal rate;
• Current PM and hydrocarbons (HC) emissions from road transport.

The current assessment of PM emission factors for road vehicles does not however specifically address emission of S/IVOCs. A method is described on how S/IVOCs have been measured in the exhaust of US gasoline and diesel-fuelled road vehicles. The ratio between S/IVOCs and other organic compounds was found to be fairly consistent among different vehicle technologies, suggesting that with the decrease of primary PM and HC emissions, SOA precursors are reduced at about the same pace.

Due to this reduction in primary PM exhaust emission, the emission from wear process will become the main remaining primary PM emission source from road transport. A second study focused specifically on the SOA yield of non-methane organic gases (NMOGs) emitted by road vehicles, based on smog chamber measurements. Total SOA production was determined to be around 0.06 g/kg fuel for pre-LEV, 0.04 for LEV (California legislation CARB Low Emission Vehicles), 0.015 for ULEV (Ultra Low Emission Vehicles), and around 0.002 g/kg fuel for SULEV (Super Ultra Low Emission Vehicles) vehicles. Based on the above referenced studies, we have furthermore concluded that current measurement methods for vehicular PM emissions may underestimate direct primary CPM emission to some degree, as additional CPM may form when exhaust gases are further cooled from the filter temperature (52°C) in the outside air temperature. At this stage, it is not yet clear how much CPM this involves.

However, it is certain that towards 2030, emissions from direct CPM will decrease fairly rapidly, with a few percent per year, due to the natural phase out of older vehicles and the planned introduction of new stringent vehicle emission limit values for PM, particulate number (PN) and HC. The emission of SOA precursors such as S/IVOCs by road transport has also decreased significantly and just as primary CPM, will continue to decrease towards the year 2030, due to the abovementioned reasons. First order estimates made in this study suggest that in the Netherlands in 2019, the PM burden on ambient air by SOA formation from gaseous S/IVOCs and NMVOCs emitted by road transport, may have been of the same order as the burden caused by primary PM emissions by road transport in 2019. Since the Dutch vehicle fleet is not that different from the European average, albeit the share of diesel vehicles is lower than the European average, as can be deduced from the EEA vehicle registration database but their mileage is higher (Geilenkirchen et al., 2020), this is likely true for other European countries as well.

SOA formation is found to be highly dependent on NOx concentrations. When NOx concentrations decrease, this may increase the SOA yields of SOA precursors emitted by road transport and by other precursor sources such as stationary fuel combustion and food preparation. Both NOx and SOA precursor emissions (such as aromatics and S/IVOCs) from road vehicles is projected to decrease as a result of emission control measures and the phase-out of older vehicles. What the overall effect on SOA production from vehicle exhaust will be, needs to be further investigated. It should be remarked that precursor emission by other sources besides road transport (e.g. wood combustion, product use) is expected to decrease to a much lesser extent, up to a point that these other sources dominate SOA precursor emission.