PCB source assessment in the lower Clinton River, Clinton River Area of Concern, Mount Clemens, Michigan

Abstract

Polychlorinated biphenyls (PCBs), some of the earliest “forever chemicals,” were used for decades in the United States before 1979 when PCB manufacturing was banned. High PCB concentrations were found recently in the lower Clinton River in the Great Lakes drainage. To determine the possible existence, location, and significance of a current source of PCBs, the U.S. Geological Survey (USGS) deployed passive water samplers (SPMDs, semipermeable membrane devices) in the river at 15 sites for 1 month in 2019 near outfalls of interest and other locations. USGS also deployed passive stream sediment samplers at a subset of four sites during the same period and collected bank sediment samples at a subset of four sites. Sediment from nearby catch basins was also collected. Samples were assayed for 209 individual PCB congeners, and patterns in total and individual congeners were evaluated; ancillary sediment data included grain size, total organic carbon, and moisture. U.S. Army Corps of Engineers (USACE) data for total PCBs and 209 PCB congeners in surficial sediment samples collected in 2019 were also evaluated. In general, total PCBs were highest in streambed sediment, followed by catch basin sediment, bank sediment, and then water as estimated from SPMDs. Total PCBs in sediment were low in all catch basins but one (sample CB19–02) that drains from an historical landfill area to one of two adjacent outfalls of interest: the outfall for a nearby wastewater treatment plant and adjacent outfall MTC–R–060, where the highest total PCBs in USGS stream sediment samples were found (site 14, sample 14STRM; 1,260,000 picograms per gram). Also, the SPMD at site 14 was the only water sample with more “light” (three or fewer chlorine atoms) than “heavy” (four or more chlorine atoms) PCB congeners, and the passive sediment sample had the highest proportion of light PCBs in USGS sediment samples. Light PCB congeners degrade more quickly than heavy PCB congeners and results may indicate that one or more current sources of PCBs are contributing to total PCBs in sediment at four river sites. Of 209 possible PCB congeners assayed, 117 congeners were detected in water samples; 155 and 154 congeners were detected in USGS and USACE sediment samples, respectively. PCBs 28, 73, 31, and 18 (highest to lowest) contributed most toward total PCBs in water samples overall; PCBs 20/28, 31, 52, and 44/47/65 contributed most toward total PCBs for sediment in USGS stream samples overall and USACE samples overall; these rankings were also true for catch basins overall except for PCB–31. After omitting coeluting congeners to allow further comparison, 5 key PCB congeners are in the top 20 congeners across all assay groups: 17, 31, 52, 95, and 118. The importance of these congeners in multiple assays aligns with their importance as components of certain Aroclors. Sediment from the high PCB catch basin (sample CB19–02) had a different pattern of top congeners than the other catch basins, and multivariate analyses indicated a high degree of similarity in its overall congener pattern with that of the highest PCB sediment sample (sample 14STRM) collected by the outfalls for the catch basin and the wastewater treatment plant. Similarities in overall congener patterns across sample media as determined by multivariate analyses confirmed some site linkages and the possibility of more than one source of PCBs to the reach. Furthermore, equilibrium partitioning calculations indicated that water concentrations as estimated by SPMDs were high enough to result in the PCB concentrations measured in USGS passive sediment samples but not USACE surficial sediment samples when normalized by organic carbon. However, the SPMDs and passive sediment samples reflect only one month of contribution to the river and higher concentrations would be expected to result with years of PCB accumulation. PCBs contributed to the river water by outfalls could eventually partition to sediment in the reach. Thus, the river could have a current source or sources of PCBs, perhaps one or more outfalls near four sites. Additional investigation is needed to better define the relative significance of each outfall and areas in nearby drainage systems that may be contributing PCBs to outfalls and the river.