Background: Accurate quantification of polychlorinated biphenyl (PCBs) at contaminated sediment sites is critical, as PCBs are often risk and remedy drivers. Two quantification methods are commonly employed: Aroclor analysis under Method 8082 and PCB congener analysis under Method 1668.
The former method generates concentrations with significant uncertainty at a reasonable cost, while the latter method produces comprehensive and accurate data at great expense. Historically, quantification was limited to Aroclor analyses but over the last 20 years congener analyses have become commercially available. Nonetheless, reliance on Aroclor analyses has persisted until recently. Increased awareness among regulators of the large uncertainties associated with Aroclor Method 8082 are leading to a new PCB sediment sampling regime: site-wide quantification of Aroclors, with a subset of those samples also undergoing congener analysis. To date, these co-located datasets have been used to correct underestimated Aroclor-based total PCB concentrations. In the future, it is likely that this new PCB sampling regime will be applied to other environmental media, such as surface water and passive samplers. Statistical methods for generating correction factors have yet to be evaluated in these media.
Approach: Sediment collected from two PCB-contaminated Superfund Sites was transported to a laboratory where it was homogenized and divided into triplicates. A passive sampler was placed into each sediment sample and allowed to equilibrate. A commercial laboratory provided Aroclor and congener concentrations for each of the sediment, porewater and passive samplers. While small, this novel dataset has significant power, as this data is typically collected at different locations and times at contaminated sites.
Results: Analysis of Aroclor and congener results is twofold and ongoing. First, single (total PCB) and multiple (homologue) linear regressions will be used to derive correction factors for Aroclor-based total PCB concentrations. Differences in the models across all three media in both sediment sources are expected and will be evaluated in light of large differences in their congener compositions. Second, tracker pair analysis, a valuable but rarely applied method, will be used to quantify the extent of weathering in both sediments. These results will be compared to the magnitude of the total PCB correction factors, which is expected to scale with the extent of weathering. Additionally, tracker pair analysis will be explored in the context of environmental forensics. Congener tracker pairs whose concentration ratios are not Aroclor-like and whose shifts cannot be explained by typical weathering patterns will be considered for their potential as inadvertently produced congeners. Such congeners cannot be detected with Aroclor Method 8082. Results generated from this uniquely complete dataset will provide a roadmap to optimizing the quality and extent of information obtainable from an emerging congener and Aroclor sediment sampling regime, as well as evaluate its extension to porewater and passive samplers.