The performance and improvement of dust sampling instruments

Abstract

The design of accurate dust sampling instruments for use in mines should logically depend on the process by which dust causes silicosis, the extent to which dust particles of different sizes are retained in the lungs, and on various physical properties of the dust. These subjects are reviewed in the Introduction, which concludes with a history of dust sampling on Witwatersrand gold mines and a description of the t wo instruments in common use - the konimeter and the thermal precipitator. The period of time over which dust samples should be taken is discussed in Section 2. Experiments are quoted which lead to the conclusion that 10 minutes is a suitable sampling time. The very short sampling time of the konimeter - approximately ¼ second - is criticised. Investigations of the performance of the konimeter are described in Section 3. The konimeter is shown to have a low sampling efficiency for high concentrations of fine dust, but to overestimate low concentrations of coarse dust. The need for a more accurate instrument is indicated. Although the thermal precipitator has been proved to collect practically all the dust particles in the air sampled, down to the limits of visibility of the electron microscope, errors arise in estimating the dust concentration with it. The sources and magnitude of these errors are discussed in Section 4, and a single thermal precipitator measurement is shown to have a standard error of approximately 13%. The standard form of thermal precipitator is not suitable for routine dust sampling in mines. A modified thermal precipitator which overcomes most of the disadvantages of the standard form and which has other advantages, is described in Section 5. Section 6 describes experiments made to improve the microscope technique at present used for counting dust samples. These experiments have resulted in finer dust particles being made visible. A high contrast photomicrographic method which reveals still smaller particles is described, and recent preliminary studies of Witwatersrand gold mine dusts with an electron microscope are mentioned. A photo-electric method of assessing dust samples is described in Section 7. This method is shown to possess advantages of speed in measurement, to eliminate the human factors and statistical errors inherent in the microscope method, and to measure a property of the dust, related to the surface area, which is probably a better measure of the danger to health than the number concentration.