Performance audits and laboratory comparisons for SCOS97-NARSTO measurements of speciated volatile organic compounds

Abstract

Abstract Performance audits and laboratory comparisons were conducted as part of the quality assurance program for the 1997 Southern California Ozone Study (SCOS97-NARSTO) to document potential measurement biases among laboratories measuring speciated nonmethane hydrocarbons (NMHC), carbonyl compounds, halogenated compounds, and biogenic hydrocarbons. The results show that measurements of volatile organic compounds (VOC) made during SCOS97-NARSTO are generally consistent with specified data quality objectives. The hydrocarbon comparison involved nine laboratories and consisted of two sets of collocated ambient samples. The coefficients of variation among laboratories for the sum of the 55 PAM target compounds and total NMHC ranged from ±5 to 15 percent for ambient samples from Los Angeles and Azusa. Abundant hydrocarbons are consistently identified by all laboratories, but discrepancies occur for olefins greater than C 4 and for hydrocarbons greater than C 8 . Laboratory comparisons for halogenated compounds and biogenic hydrocarbons consisted of both concurrent ambient sampling by different laboratories and round-robin analysis of ambient samples. The coefficients of variation among participating laboratories were about 10–20 percent. Performance audits were conducted for measurement of carbonyl compounds involving sampling from a standard mixture of carbonyl compounds. The values reported by most of the laboratories were within 10–20 percent of those of the reference laboratory. Results of field measurement comparisons showed larger variations among the laboratories ranging from 20 to 40 percent for C 1 –C 3 carbonyl compounds. The greater variations observed in the field measurement comparison may reflect potential sampling artifacts, which the performance audits did not address.