An interim ASTM method for analysis of metalworking fluids was evaluated for estimation of personal exposure concentrations in a 79-plant survey of machine shops and metalworking facilities across the United States. Both thoracic and total particulate samples were collected on tared polytetrafluoroethylene filters. Metalworking fluid concentrations were estimated by determination of the total- and extractable-sample weights using ASTM Method PS-42-97. This procedure employs a ternary solvent blend of dichloromethane:methanol:toluene to separate the fluids from comingled insoluble particulate. Following the initial extraction, re-extraction of 322 samples with the ternary solvent removed, on average, <2.5% of the sample weight, indicating that the majority of extractable material had been removed during the first extraction. Evaluation of the field study blank data permitted estimation of the mean, median, and upper 95th percentile of the limits of quantitation to be 0.1 mg, 0.1 mg, and 0.3 mg, respectively, for both the total weight and the extractable weight samples. For the 79-plant survey, the fractions extracted (weight extracted/weight of sample) were studied as a function of four metalworking fluid types and three main work operations-grinding, milling, and turning-according to a restricted maximum likelihood statistical procedure. This evaluation of the data indicated that the fractions extracted generally decreased in the order: straight > semisynthetic or soluble > synthetic; the differences in the fractions extracted for the straight and the synthetic fluids were statistically significant only for the grinding operation at two sample levels tested. Studies of the stability of quality assurance (QA) samples spiked separately with a straight, a soluble, a semisynthetic, and a synthetic fluid indicated that the QA samples all lost weight according to simple linear decay equations. These decay equations were used to estimate the amounts expected to be reported for QA filters by the performing laboratory. For storage periods ranging from 18-26 days, the total weight of sample recovered for each QA sample was ³ 80% of that expected from the decay equations. For these QA samples, the fractions extracted (extracted weight/total sample weight) from all four fluid types were ³0.90.
National Insitute for Occupational Safety and Health, 4676 Columbia Parkway, Cincinnati, OH 45226