Abstract
Power-frequency magnetic fields were measured with two types of instruments at a large automobile transmission plant and the results were compared. A group of machining and assembly workers (n = 57) wore a datalogging three-axis personal monitor, and also had each of their work locations measured with a hand-held meter. Time estimates at each of these locations were combined with averages of the meter readings to give surrogate daily-weighted average (DWA) exposures, which were compared with measures of central tendency derived from the personal monitor. Nonparametric correlation coefficients were used to compare the two estimates. Closer agreement was found between the DWA and the datalogger's geometric mean (Spearman coefficient, rs = 0.85) than between the DWA and the arithmetic mean (rs = 0.77), apparently because both the geometric mean and DWA were relatively insensitive to brief peaks. Principal sources were identified and their distances from the usual work position were estimated. Demagnerizers were unexpectedly found to be important sources in some jobs. Median distance from source to worker was about 1 m. For one cluster of demagnetizers, which influenced exposures up to 15 m away, the power law exponent (b) for falloff with distance was estimated with measurements in four directions. We found 1.9 < b < 2.6, consistent with physical principles for sources which are small relative to the distances measured.