Accuracy of particle number measurements from partial flow dilution systems

The measurement of the particle number (PN) concentration of non-volatile particles >23 nm was introduced in the light duty vehicles regulation; the heavy duty regulation followed. Based on the findings of the Particle Measurement Programme (PMP) heavy duty inter-laboratory exercise, the PN concentration measurement can be conducted either from the full dilution tunnel with constant volume sampling (CVS) or from the partial flow dilution system (PFDS). However, there are no other studies that investigate whether the PN results from the two systems are equivalent. In addition, even the PMP study never investigated the uncertainty that is introduced at the final result from the extraction of a flow by a PN system from the PFDS. In this work we investigate the uncertainty for the three possible cases, i.e. considering a constant extracted flow from the PFDS, sending a signal with 1 Hz frequency to the PFDS, or feeding back the extracted flow to the PFDS. We also discuss what kind of tests could be conducted in order to ensure that the reported extracted flow is correct. The experimental results confirmed that if the extracted flow rate is accurately measured or fed back, then CVS and the PFDS give equivalent PN results with differences in the order of ±15% for PN emissions >1×10-11 p/kWh. At lower levels, the PN backgrounds of the CVS and the PFDS become important.