Impacts of PCB analytical interpretation uncertainties on dechlorination assessment and remedial decisions

Abstract

Laboratory analyses of polychlorinated biphenyls (PCBs) often do not quantitate the 209 individual PCB congeners, thereby requiring analyst interpretation to determine individual congener concentrations. Error introduced during this interpretation is subsequently propagated to calculated surrogate variables, such as the number of chlorines per biphenyl (CPB), and the molar dechlorination product ratio (MDPR), which are used to assess the extent of dechlorination and inform remedial decisions. The present work applies a Monte Carlo (MC) analysis to assess current methods for quantitating co-eluting congeners and the errors that could occur in individual congeners and derived CPB and MDPR estimates. Synthetic chromatograms, which were created using two alternative methods (random assignment and assignment based on relative proportions in Aroclors) for assigning mass to co-eluting congeners, were compared to their fully-quantitated counterparts. The percent error introduced in total PCB (∑PCB) concentration ranges from approximately −60% to +50%. Similarly, the errors associated with CPB and MDPR estimates range from approximately −20% to +20% and −120% to +30%, respectively. Uncertainties introduced during congener analysis and propagated to surrogate variables can thus be substantial, and should be considered in assessments of the extent of dechlorination and associated remedial decisions.