BS EN ISO 28439:2011 pdf download – Workplace atmospheres — Characterization of ultrafine aerosols/nanoaerosols — Determination of the size distribution and number concentration using differential electrical mobility analysing systems
8.2 Sampling
If worker exposure is to be estimated, the aerosol shall be sampled in the breathing zone of the worker. All existing DMASs at the time of publication are large static instruments. The sample should be taken at a representative position in or very near the worker’s breathing zone. A sampling line as specified in 6.2 can be applied. If the worker is not stationary, samples from different locations can be taken by moving either the entry of the sampling line or the complete DMAS. Diffusion losses in the line shall be taken into account (see Clause 11).
If the instrument performs a scanning or stepping voltage mode, the time for one measurement is a few minutes, typically 3 min to 6 min. During this time, the aerosol should not differ in size and number of particles. If highly fluctuating concentrations, such as in welding workplaces, are to be measured, it is recommended that a buffer vessel of a few litres’ volume be used in the sampling line to avoid artificial peaks during a scan. Another option is to average a series of samples if short-term emissions appear regularly. Some DMAS devices allow a selection of instrument parameters like sampling air flow rate, sheath air flow rate, scan times or even a change of the DEMC type. These settings affect the measurement range. It is recommended that the range be selected in such a way as to cover the whole size distribution of the ultrafine particles at the workplace. Practical experience shows that the maxima of the particle number size distributions lie between approximately 10 nm and 500 nm, depending on the occurrence of primary, aggregated and agglomerated ultrafine particles. The total number concentration integrated over the measurement range found at workplaces varies between 10 9 particles/m 3 and 10 14 particles/m 3 in clean air sections or urban background and welding plumes, respectively (see Reference [10]).
9. Presentation and evaluation of data
The resulting distribution of the number concentration, C N , shall be given as differential concentration in terms of the equivalent particle electrical mobility particle diameter, d p , e.g. dC N /dd p or dlog C N /dlog d p, in order to make different instrument settings (number of channels per decade, width of channels) comparable. Depending on the object of the measurements, the number size distribution data can be presented in a number of different ways, for example:
a) the calculation of statistical data like maximum or median values is of value and is possible if the entire distribution is sufficiently covered;
b) calculation of the total number concentration in the measurement range, the total number concentration for the range below 100 nm, or any other sub-fraction could be reported as a single value for comparison of different exposure scenarios;
c) the time-weighted average size distribution and the total number concentration can be determined for the whole reference period (e.g. 8 h or 15 min);
d) in order to present the variation in concentration over the whole size distribution for the whole reference period or measurement period, plot the size distributions as a two-dimensional contour diagram (with time on the x-axis, particle size on the y-axis and with concentration as contours) or as a three-dimensional surface or wire-frame diagrams (with time on the x-axis, particle size on the y-axis, and concentration on the z-axis);
e) peaks related to process or workplace activities.
NOTE The particle surface area and particle volume distribution can be estimated from the number size distribution if mainly compact or spherical particles or agglomerates and aggregates are sampled. For other non-spherical particle shapes, e.g. tubes and ropes, both these estimates as well as the number size distribution are less accurate. These distributions can be presented in similar ways to the number size distribution data listed above.