Following an evolutionary process, the DTT consortium has developed both laboratory and portable versions of a sampling system going through a gradual procedure, basing on fundamental investigations around the lab system (Figure 2 and 3) and then solving practical problems in converting the system for portable use (Figure 4 and 5). This was then deployed in several laboratories to assess the effects of different vehicle technologies, fuels, regulatory cycles, temperatures, operating conditions and aftertreatment systems. From the overall results (Figures 6, 7 and 8), a number of important messages emerged:
General
• A novel sampling and measurement system was developed, for which the DTT consortium has demonstrated capability to measure both solid and total particle numbers. It offers a reasonable penetration efficiency for primary particles and a reasonably stable dilution ratio. When a catalytic stripper (CS) is in-line, the system was found to be artefact free.
• Most vehicle technologies were found to be compliant with the 6x1011 #/km limit for both >23 nm and >10 nm particles. For moderate power drive cycles, emissions of vehicles with particle filters are always below the limit value. Improvement in current filter applications will be required to bring PN10 emissions below the limit with a suitable engineering safety margin in severe. This is also expected to bring PN<10 below the limit value. The highest emissions were observed from PFI passenger cars, motorcycles and mopeds, currently not subject to PN legislation. Technology improvements (including particle filters) are required to meet the current limit for PN10. Average PN emissions from gasoline hybrids in ICE mode under severing conditions may also require a GPF even for PN23.
Specific cases with significant number of particles below 23nm
• PN emissions from the cold start are higher from the average cycle in general. PN<23 from this phase can dominate the overall emissions from spark ignition technologies.
• Diesel PN emissions during active DPF regenerations can be dominated by PN<10 for short periods.
• GDI vehicles may emit <23 nm soot particles under very high load transients; these particles seem to mostly fall above the PN10 threshold rather than below 10 nm.
• CNG vehicle emissions were found to be significantly elevated below 10 nm. A simple retrofit of a GPF has been seen to reduce both PN10 and PN<10 to well below the limit value. A prototype CNG operating on in-cylinder injection exhibited PN<23/PN23 ratios that were below 2.5 times.
• In RDE measurements, the measured PN10 and PN23 emissions are influenced by the soot loading both in diesel and gasoline filter applications. Higher PN emissions are detected after regeneration. PN10 and PN23 emission levels over RDE are usually above the WLTP levels.
• Total Particle Number (TPN) emissions also seem lower than levels measured for corresponding combustion technologies one decade ago. This suggests that emission control technologies introduced for solid particle control also led to reductions in total particle number. Specific GDI operation events, such as high accelerations and steady state high speeds, can increase emissions of such particles, to levels more than an order of magnitude higher than SPN emissions.
