Comparing enhanced natural recovery and enhanced natural recovery with activated carbon: a case study in the Lower Duwamish Waterway

The use of activated carbon (AC) to augment enhanced natural recovery (ENR) is an increasingly recognized remedy to reduce the bioavailability of hydrophobic, bioaccumulative compounds. The U.S. Environmental Protection Agency (USEPA) and Washington Department of Ecology (Ecology) were interested in whether the performance of ENR with AC would enhance the effectiveness of ENR in the Lower Duwamish Waterway (LDW), a tidally influenced, salt-wedge estuary. In 2014, USEPA and Ecology directed the Lower Duwamish Waterway Group (LDWG) to evaluate the potential effectiveness of using AC (Coconut Fine Mesh Activated Carbon graded 200-1,000 µm) with ENR (referred to herein as ENR+AC) to remediate polychlorinated biphenyls (PCBs) in aquatic sediment in the LDW. This three-year pilot study established three one-acre areas within the LDW representing different site conditions (an intertidal area, an area prone to scour, and a subtidal area) where ENR+AC and ENR would be compared. The target ENR and ENR+AC thickness was 15-30 cm with 4% AC in the ENR+AC plots; actual thicknesses across all plots were 15-46 cm, with a mean depth of material across plots that ranged from 24 to 35 cm. Over the three-year study period, the ENR and ENR+AC placements were relatively stable, and AC remained stable within the ENR+AC plots. While final ENR applications were somewhat thicker than expected, benthic community results demonstrated substantial biological activity during the study, including the presence of organisms that burrow deeper than the ENR layer depth. Both treatments performed similarly at plots where the performance could be most accurately assessed (i.e., at the Intertidal and Subtidal Plots). For the Intertidal Plot, the average (±SD) Year 3 freely dissolved (Cfree) total PCB concentration in the ENR subplot was 1.6 ± 0.26 ng/L, compared with 0.78 ± 0.19 ng/L in the ENR+AC subplot; the difference in Year 3 Cfree concentrations, while small, was statistically significant (p = .011) and reflected 95% and 97% decreases from average baseline Cfree concentrations, respectively. The Subtidal Plot had a 96% decrease from baseline Cfree PCBs in Year 3 in the ENR+AC subplot compared to an 89% decrease in the ENR-only subplot. Average Year-3 Subtidal Plot Cfree concentrations were 4.3 ± 1.1 ng/L and 3.8 ± 0.42 ng/L, respectively; the difference between the subplot concentrations in Year 3 was not statistically significant (p = .588), suggesting that the larger decrease seen in the ENR+AC subplot was influenced in part by a higher baseline Cfree PCB concentration in the ENR+AC subplot (108 ng/L) compared to the ENR subplot (36 ng/L). In the Scour Plot, low baseline Cfree PCB concentrations in the ENR (1.5 ng/L) and ENR+AC (11 ng/L) subplots made it difficult to statistically compare the Scour Plot performances. In the Intertidal and Subtidal Plots, ENR reduced PCB bioavailability so well that the additional improvements by AC were difficult to detect or very minor, and the Year 3 results at ENR and ENR+AC subplots were not meaningfully different. In all three plots, the Year 3 AC measurements confirmed the continued presence of AC. Overall, results indicate that both ENR and ENR+AC were successful in reducing PCB bioavailability under a wide variety of conditions in the LDW. The ENR reduced PCB bioavailability so well that no substantive improvements as a result of adding AC were detected.