< back to main site


Particle measurement programme - analysis of errors.

Gilham, R; Quincey, P (2007) Particle measurement programme - analysis of errors. NPL Report. AS 17

[img] Text

Download (3MB)


This document has been prepared for the Department for Transport (DfT) as part of the Particle Measurement Programme (PMP). The aim of the document is to evaluate uncertainties (errors) in specific vehicle emission particle measurements, linking them to the calibrations and checks that are required or that may be needed in the measurement procedures. These requirements and procedures [1-3] were themselves being modified during the time that this document was being written, to some extent in response to its early drafts. This final document takes into account changes agreed at the PMP meetings at DfT on 8th and 9th October 2007.
A provisional rough estimate of the uncertainty for particle concentration measurements, based on the figures and assumptions used here, is 15%. This uncertainty figure corresponds to a level of around 95% confidence. It should be emphasised that it is not possible to be fully rigorous or definitive, because there will be variations in how the methods are carried out in practice, and some of the factors are not well characterised. Nevertheless, this should be a realistic approximate figure.
The major factors are the calibration of the particle number counter (PNC), an area where it is acknowledged that international standardisation is required, and the reproducibility of the Particle Conditioning and Measuring System (PCMS).
The error calculation applies when total dilution factors of 150 are used. In all cases, high dilution factors make the measurements more prone to errors, because the actual particle dilution factor is more difficult to determine. When a dilution factor around 600 is used, for example with a Gasoline Direct Injection (GDI) engine, additional care would be required to ensure that statistical variations, and the effects of leakage and “noise”, were addressed during the dilution factor measurement. Similar uncertainties could still be obtained.
It is important that significant factors such as reproducibility and test source stability are properly quantified and controlled within the procedures. Consideration should be given to retrospective correction of data using subsequent calibration results.
Although more than 99% of volatile particles are removed by the Volatile Particle Remover (VPR), the presence of a 1% fraction of the volatile particles emitted by the vehicle could have a significant influence on results, but this effect is not investigated in detail here.
For volatile particle removal efficiency measurements, there are potentially large errors in the absolute determination due to the small numbers of particles at the outlet of the VPR. These can be minimised by using the hot/cold method together with the highest available concentration of volatile particles at the inlet. However, in general VPR performances appear to meet the 99% removal requirement comfortably, and the errors have the effect of reducing the apparent efficiency, so that even quite large errors of this kind do not affect the validity of results.

Item Type: Report/Guide (NPL Report)
NPL Report No.: AS 17
Subjects: Environmental Measurement
Environmental Measurement > Air Quality and Airborne Particulates
Last Modified: 02 Feb 2018 13:15
URI: http://eprintspublications.npl.co.uk/id/eprint/4019

Actions (login required)

View Item View Item