Simulation analysis of inhalable dust sampling errors using a multi-component error model

Measurements to characterize inhalable aerosol exposure are subject to rand om error even after sources of systematic error have been eliminated. For a fixed aerosol sampler geometry the random errors are due to the variabilit y of measured and unmeasured parameters including ambient variables, quanti fication technique, and operation parameters. In this discussion we apply a multi-component error estimation model to size selective aerosol sampling with the well-known Institute of Occupational Medicine (IOM) inhalable aero sol sampler. Random errors due to typical variations in sampler flow contro l, timing, and mass determination were small, being approximately 3%. Simil arly, random errors due to variations in wind velocity were reasonably smal l at approximately 10%. However, the bias introduced by wind velocity was n otable, ranging from peak values of 17 to 27% depending on aerosol mass med ian aerodynamic diameter and geometric standard deviation. This modeling in dicated that the combined influence of variations in sampler flow control, timing, mass determination, and ambient wind velocity on IOM performance ap peared to be less than approximately 10%; however, bias at moderate wind ve locities was shown to be important for the IOM sampler as suggested by othe r studies. The effects of sampler placement, angle of incidence of ambient wind velocity on the sampler, and head orientation of the exposed person ar e unknown at this time and need additional research.