Critique of 1985 ACGIH report on particle size-selective sampling in the workplace.

The criteria recently proposed by ACGIH for judging the acceptability of respirable and other dust fraction samplers are analyzed. Implications on the sampling of workplace aerosol are determined. With the consideration of both bias and imprecision, the overall accuracy limited by the criteria is estimated for the sampling of coal mine dust as characterized by various researchers. The accuracy limits thus found appear to be excessively broad. As an example with actual workplace dust distributions in the sampling of a single aerosol (mass median diameter = 18.6 micrometers and geometric standard deviation = 2.3) with respirable dust concentration near 2 mg/m3, two samplers acceptable according to the proposed criteria could be found giving respirable dust measurements equal to 0.71 mg/m3 and 4.3 mg/m3 (even after excluding 5% of the low and high measurements from each sampler, respectively). Large variation in samplers acceptable according to the criteria is found for many other distributions as well; this indicates that tighter requirements are necessary. Seldom attained are both the single-sample +/- 25% accuracy at the 95% confidence level required of sampling/analytical methods endorsed by the National Institute for Occupational Safety and Health (NIOSH) and the tighter ISO dust fraction measurement requirement that 67% of the measurements of a sampled dust fall within 10% of a true value. Suggestions are given for sharpening the criteria without eliminating all samplers from acceptability.