Conference report: 26^th ETH Nanoparticles Conference

4 August 2023

The 26^th ETH Nanoparticles Conference (NPC) was held June 20-22, 2023, as an in-person conference and supported by the Swiss Chemical Society. The conference program included 44 presentations and 60 posters covering particle fundamentals, particles from aircraft, emission measurement, emission control, occupational and indoor exposure, and health effects of air pollution. The conference closed with a Focus Event—“Indoor air filtration of biogenic and combustion nanoparticles”.

Engine Emissions

The US EPA provided a summary of the recently proposed emission standards for MY 2027 Light-Duty and Medium-Duty Vehicles with a focus on the 0.5 mg/mile PM requirements [5922]. This PM limit must be met across three test cycles: 25°C FTP, US06, and -7°C FTP. The EPA chose mass-based emissions standards in the proposed rule because health benefits can be directly quantified using epidemiological studies that use ambient air mass PM_2.5, and unlike solid PN, PM mass includes semi-volatile PM which can have high toxicity and is emitted even from GPF-equipped vehicles.

The -7°C FTP is considered an important real-world temperature that can differentiate between vehicles with GPF-level PM from vehicles with Tier 3-level of PM. The US06 represents high load real-world driving and ensures good PM control during and immediately after GPF regeneration by inducing passive GPF regeneration.

In support of the proposed standard, PM mass measurements were conducted according to the existing EPA test procedures (40 CFR Part 86, 1065, and 1066) using five vehicles (2011 F150, 2019 F150, 2021 Corolla, 2021 F150 HEV, 2022 F250) tested across three organizations (EPA, ECCC, and FEV) without a GPF, with a MY2019 GPF (catalyzed), and with a MY2022 GPF (bare). The results show that the existing EPA test procedures can resolve PM mass emissions below 0.5 mg/mi. To improve PM signal to noise ratio, it is recommended to use the lower half of the allowable dilution factor range, increase filter face velocity from 90 to 140 cm/s and load 1 filter/test rather than 1 filter/phase. Both GPF technologies reduced PM below the proposed limit over the -7°C and 25°C FTP cycles but over the US06, only the 2022 GPF technology was compliant.

The regulatory proposal also eliminates the allowance of commanded enrichment for power or component protection and requires that MDVs with GCWR > 22,000 lbs comply with the heavy-duty engine dynamometer-based criteria pollutant standards.

In further support of EPA’s proposal, SwRI also showed that GPF technology can comply with the proposed PM standard [5923]. A CO₂ penalty of less than 1% at all temperatures is expected. The required GPF volume to engine displacement ratio is expected to be about 0.55 and GPF cost was estimated to range from about $50 for a 1-liter engine to about $110 for a 5-liter engine. E-fuels with an ultra-low value of PM Index (PMI~0.27) show potential to significantly reduce all particle metrics that could benefit efforts to reduce PM and GHG emissions from the legacy fleet.

With regards to measurement variability, it was shown that the current CFR Part 1065/66 procedures are sufficient to demonstrate that PM emissions from a 2019 F-150 with a close-coupled catalyzed GPF over the 25°C FTP were well below the 0.5 mg/mile proposed limit even after adding three standard deviations to the measurement (99.7% confidence interval). Over the US06 cycle, PM emissions and their variability were higher with the result that the 99.7% confidence interval was much closer to 0.5 mg/mile limit. No variability data over the -7°C FTP was presented.

The VERT Association attempted to quantify the climate change benefit of the particulate filters currently in use [5924]. They estimate that more than 250 million DPFs and GPFs have been sold worldwide over the last 10 years with the result that 3 Mt of black carbon emissions have been avoided. Using a global warming equivalence of BC/CO₂ of 2083, this is equivalent to about 6,200 Mt CO₂ over the 10-year period. Annual reduction in black carbon would be about 600 Mt or about 15% of the EU’s annual CO₂ emissions.

Researchers at Tokyo “Denki” University Graduate School and the National Institute for Environmental Studies presented some work on off-cycle emissions from Japanese certified gasoline and diesel passenger cars [5925]. They tested several Japanese light-duty vehicles at low ambient temperature (<0°C) over the WLTP. Some vehicles were also tested with the extra high phase included in the WLTP. The extra high phase portion of the WLTC is not included in Japanese certification testing. For one gasoline fueled car with a small engine (0.66 L displacement), enrichment over the extra high phase led to increased CO and PN (mainly below 23 nm). For the other gasoline cars, both PFI and GDI, CO and PN were high during warm up from low ambient temperature over the Japanese certification WLTP. With the diesel car, the extra high phase triggered passive regeneration of the DPF that increased PN (mainly below 23 nm) and NOx.

Researchers at the University of Minnesota and Cummins Filtration examined the reasons for the unexpected temperature dependent performance of crankcase particle control devices [5926]. They found that there is a significant measurement artefact. During sampling and measurement, particles are in dynamic equilibrium with their vapors and since there is no dilution, no vapors are lost. With the apparatus used, particles shrink in the heated inlet system but grow again when they pass from upstream measurement location to the instrument. Particles remain small in the isothermal removal device and are filtered or processed at this reduced size. The particles then grow again when they pass from the downstream measurement port to the instrument. To better determine the performance at a given temperature, the measurement system should be operated at the same temperature as the filtration device and dilution probes could be used. However, in actual application, particles will regrow as they cool after passing through the filtration device.

Emission Measurements

Cummins and Horiba evaluated a couple of options to make tailpipe solid PN measurements and compared them to measurements made from a constant volume sampler (CVS) tunnel and a partial flow dilution system (PFDS) [5927]. Measurements were made from natural gas and diesel vehicles. The direct tailpipe measurement systems evaluated were (1) a system that used a 0.5 m heated line upstream of a PN measurement system and (2) a system that used a heated diluter close to the sampling point followed by a heated transfer tube. While there were some differences between the tailpipe measurement systems and the reference systems, the greatest challenge seemed to occur with natural gas engines operating at high exhaust temperature conditions where the thermophoretic losses were on the order of 30-40%.

Development of a miniaturized exhaust emission measuring system at Aristotle University shows some promise as a low-cost emission measuring system that could be used for large scale testing beyond type-approval [5928]. The current iteration of the system uses low-cost electrochemical (NO and CO) and NDIR (CO₂) sensors for gaseous species and an optoacoustic black carbon sensor based on a low-cost commercial laser diode (LD) and sensitive Quartz Tuning Fork (QTF) for sound detection. Compared to the PEMS system used as a reference, CO₂ emissions results had a comparable time response and level, NO followed the trend but with a slight lag and overall deviation of 10-15%, CO peaks were detected but underestimated with a mean deviation of ~40-50%, and BC showed a fast time response and excellent correlation with the PEMS PN. Sensor durability was unclear. Further miniaturization and optimization of the sampling system is required as well as the integration of a hydrocarbon sensor.

Brake and Tire Wear

Ricardo summarized work it is doing to support the UK’s Department for Transport to develop a measuring system and methodology for characterizing particles emitted from brake and tire wear under real driving conditions with the primary aim of developing emission factors [5929]. Tire wear particles were sampled using an open duct located immediately behind the tire’s contact patch while brake wear particles were sampled from an enclosed fixed volume created by enclosing the pad and disc. Tire emissions sampling proved more challenging than brake sampling, with background contributions often dominating overall collected masses. For brakes emissions, particle number emissions corresponded to real-time braking events, with emissions around 2 - 5 × 10⁹ #/km/brake while mass emissions (~PM_2.5) were around 1 mg/brake/km. For tire emissions, non-volatile PN and PM emissions could be related to individual braking events on the chassis dynamometer, but not easily on the track or road. The exact quantification of tire PM_2.5 and PN is dependent on the sampling duct efficiency, which is currently unknown and is to be investigated further. Without correction, measured tire emission per km were lower than from brakes. Repeated aggressive braking events appear to lead to outgassing of particles from tires that were volatile below about 200°C.

Work at the University of California Riverside’s CE-CERT assessed the level of exposure to non-tailpipe emissions in near-road environments by using the metal contents and size distributions of PM_2.5 and PM₁₀ measured next to two major highways in California [5930]. The results showed non-tailpipe emissions were at similar levels to tailpipe emissions. Non-tailpipe emissions were present mostly above 1 μm size with particles containing brake markers (Ba, Sb) peaking between 1-10 μm and those with tire markers (Zn) peaking between 10-18 μm.

CARB provided updated brake emission factors (EMFAC) in 2021 based on dynamometer tests for light-duty vehicles [5931]. Previous emission factors provided in 2017 showed brake emissions would dominate (80%+) PM_2.5 and PM₁₀ emissions from light-duty vehicles out to 2050. While the updated emission factors from 2021 still show that non-exhaust emissions will dominate PM_2.5 and PM₁₀ emissions from light-duty vehicles out to 2050, the contribution from tires is expected to be more significant than the previous estimate. The updated estimate suggests that the contributions from brakes and tires are about equivalent with brakes contributing slightly more than tires to PM_2.5 and tires contributing slightly more than brakes to PM₁₀.

* * *

The 27^th ETH Nanoparticles Conference, NPC-24, is scheduled for June 11-13, 2024 as an in-person event.

Conference website: npc23.scg.ch