Precision and accuracy of foundry dust exposure estimates from air sampling data.

Errors in the sampling methods and in the determination of respirable size quartz in foundry dust have been evaluated and discussed. In general, the total precision of dust sampling and analysis was better than 20%. The term "exposure dose" is introduced and defined. The temporal variation associated with measuring exposure dose is assessed with reference to the use of a log-normal distribution of air sampling results. The logarithmic standard deviations of dust concentrations at fixed sites and jobs are calculated.