Bias in air sampling techniques used to measure inhalation exposure

Factors have been evaluated which contribute to the lack of agreement between inhalation exposure estimates obtained by time-weighted averaging of samples taken with mini hi-volume samplers, and those measured by time integrating, low-volume, lapel mounted, personal monitors. Measurements made with real-time aerosol monitors on workers at a Be-Cu production furnace show that part of the discrepancy results from variability of the aerosol concentration within the breathing zone. Field studies of sampler inlet bias, the influences of the electrostatic fields around polystyrene filter holders, and resuspension of dust from work clothing, were done in three areas of a Be plant. No significant differences were found in Be air concentrations measured simultaneously by open and closed face cassettes, and "mini hi-volume" samplers mounted on a test stand. No significant influence on Be collection was detected between either positively or negatively charged monitors and charge neutralized control monitors. The effect of contaminated work clothing on dust collection by lapel mounted monitors is most important. Beryllium release from the fabrics affected air concentrations measured by fabric mounted monitors more than it affected concentrations measured by monitors positioned above the fabrics. The latter were placed 16 cm from the vertically mounted fabrics, to simulate the position of the nose or mouth. We conclude that dust resuspended from work clothing is the major source of the observed discrepancy between exposures estimated from lapel mounted samplers and time-weighted averages.