RELATIONSHIPS BETWEEN SIMULTANEOUS AIRBORNE DUST SAMPLES TAKEN WITH FIVE TYPES OF INSTRUMENTS AT SOUTH AFRICAN ASBESTOS MINES AND MILLS

摘要

Simultaneous airborne dust samples were taken with five types of instrument, namely, konimeter, thermal precipitator, membrane filter, tyndallometer and VC 25 gravimetric sampler, on the surface and underground at two types of asbestos mine, producing amosite and Cape crocidolite respectively. The results were analysed and compared in terms of five different concentration parameters, namely, particles plus short fibres (p+sf), long fibres (If), particles plus short fibres plus long fibres (p+sf+If), the volume (tyndallometer readings) and the mass of dust (VC 25). The Kolmogorov-Smirnov statistic was used to test for normality and the Bartlett statistic for equality of variance of the transformed concentration data recorded by each type of instrument.Fourteen linear relationships were investigated. In those cases where replicates were given these were used to obtain estimates of the error variances for the predictor and the response variables. With these estimates an additional regression equation to the least squares equation was determined. This new equation is recommended in preference to the corresponding least squares equation, as it is based on information which is discarded when only the position means are used.All but one of the fourteen relationships reached statistical significance.The konimeter and the thermal precipitator can be linked with the membrane filter, the tyndallometer can reveal short time variations in the fibre level and the VC 25 can disclose the asbestos content of the fine dust.