Formatted Title:

Development and Application of a Predictive Framework for Evaluating Remedial Actions for PCB-Impacted Water Bodies

Background/Objectives:

Multiple past studies have characterized the concentrations of polychlorinated biphenyls (PCBs) in sediments, water, and fish in the Anacostia River. However, several gaps remain in the interpretation of the data that make it difficult to link concentrations in water and sediment simultaneously to the concentrations in fish and anticipate changes in the future from natural processes or management actions. The main objectives of this work were: a) development of a site-specific food web model for PCB bioaccumulation in key aquatic species in the Anacostia River; b) development of a congener-level PCB mass balance model for different segments of the Anacostia River and coupled with the food web PCB bioaccumulation model, and c) use of the coupled models to predict the impact of sediment remediation and source control scenarios on the congener level PCB concentrations in sediment, porewater, surface water, and key aquatic species.

Approach/Activities:

An aquatic food-web model for the Anacostia River was developed based on previous work by Arnot and Gobas (2004) using site-specific information from the Anacostia River Sediment Project (ARSP) Remedial Investigation (RI) and published literature. Freely dissolved PCB concentrations in the surface water measured using passive samplers and sediment PCB concentrations measured as part of the RI were used as model inputs to predict PCB concentrations in food-web species. Model predictions of the sum of 85 PCB congeners (∑PCB) in benthic and pelagic invertebrates, and in fish were validated against measured data from multiple studies. For the PCB mass balance model, the Anacostia River was divided into six segments and incorporated most recent available field data and results from the ARSP hydrodynamic modeling. Freely dissolved PCB concentrations from the mass balance model were validated against measurements from Ghosh et al. (2020) and the ARSP hydrodynamic model predictions. The mass balance and food web models were coupled and used to assess the impact on PCB bioaccumulation in fish from a range of management options including no action, source control, dredging and capping, in situ treatment using activated carbon, and combinations thereof.

Results/Lessons Learned:

Results of PCB uptake in the aquatic food web using the Arnot and Gobas (2004) model were indicative of good model performance, with 90% of the predicted ∑PCB concentration in fish lipids within a factor of 2 of measured data and all predictions within a factor of 5. Predictions of freely dissolved total PCB concentrations from the mass balance model also showed good agreement with available direct measurements. Predicted results from coupling of the food-web and mass balance models showed that dissolved PCBs in surface water are an important driver of bioaccumulation in aquatic organisms at this site and significant reductions are required in surface water PCB concentrations to enable achievement of target concentrations in fish. A remedy that combines upstream source control with remediation of contaminated sediment hot-spots was generally the most effective at achieving substantial reductions or PCBs in fish in most segments of the river. Modeling results also indicate that additional remedial actions may need to be explored for downstream segments to expedite recovery faster than the natural attenuation processes. The approach presented here demonstrates the utility of passive sampling and mechanistic bioaccumulation modeling for informing and improving the remedial decision-making process at contaminated sites.

Alfred Pinkney

Affiliation

Refine Results

Development and Application of a Predictive Framework for Evaluating Remedial Actions for PCB-Impacted Water Bodies