The Golden System for particle number measurement (Matter Engineering rotating diluter, evaporation tube and ejector diluter plus a TSI Condensation Particle Counter) performed well. Daily validation checks at each lab did not highlight any problems in terms of system leakage, particle counter high and low responses and linearity. Comprehensive calibrations at the beginning and middle of the test programme confirmed the stable operation of the system. Minor damage was sustained to the first diluter unit in the Golden System but this was attributable to laboratories unfamiliarity with the equipment, its repair did not affect the performance of the measurement system or cause any shift in observed particle number measurements.
Mean particle number emissions were less than 2x1011 particles/km for DPF equipped diesels, including light goods vehicle derivatives, with repeatabilities of 27-78% (expressed as coefficients of variance). Repeatability was typically around 30% and the one major deviation from this (78%) was due to the DPF being in an unstabilised fill state resulting in emissions from the vehicle decreasing test after test as the DPF filled up and the DPF's filtration efficiency progressively increased. Subsequent to testing of this vehicle a DPF stabilisation protocol was adopted. One DPF equipped diesel did give higher mean results of around 6x1011 particles/km. This vehicle differed from the other DPF equipped vehicles in being fitted with a more porous cordierite DPF substrate than the more commonly used silicon carbide DPF substrates. The particle emissions trace from this vehicle showed solid particle emission levels following the drive cycle (as they do for a conventional diesel vehicle) unlike the trace for a more efficient DPF where solid particle emissions are practically eliminated except for during the cold start and final acceleration of the Type 1 test cycle.
Reproducibility of the measurement was assessed by testing a single DPF equipped ‘Golden Vehicle’ in all laboratories. This gave an all-labs mean of ~8x1010 particles/km with a reproducibility of 31% i.e. similar to the repeatability of the measurement on DPF equipped vehicles.
Conventional diesel vehicles gave particle number emissions of around 5x1013 particles/km i.e. more than two orders of magnitude higher than the DPF equipped vehicles. Direct injection petrol-engined vehicles mean particle number emissions were in the range 3x1012 to 1x1013 particles/km. The conventional, multi-point injection petrol-engined vehicle tested gave particle number emissions similar to the DPF equipped diesels.
A number of alternative measurement systems using the same operating principles as the Golden System were tested alongside it at the various laboratories. Full performance data for these systems demonstrating the extent to which they meet the PMP specification in terms of volatile particle removal and solid particle penetration efficiency was available for only one system. This system, which meets the PMP specification, gave good correlation with the Golden System results (R2 of 0.93) with absolute numbers being around 15% lower than the Golden System results. Direct clones of the Golden System predictably gave even better correlation (R2 over 0.98), again with absolute measurements being around 15% lower than those from the Golden System. All other measurement systems gave good correlation with Golden System measurements (R2 between 0.8 and 0.9), although absolute particle number levels were around 40% lower than from the Golden System. These systems comprised components adapted for PMP use rather than specifically designed to meet PMP requirements and may well give substantially improved results if redesigned/optimised to meet the PMP system specification.