Old Manufactured Gas Plant site in New York.

PAH PARTITIONING AND ORGANIC MATTER CHARACTERIZATION IN MANUFACTURED GAS PLANT SITE SEDIMENTS
Investigator: Upal Ghosh
Duration: January 2003 - December 2004.
Funding agencies: Gas Technology Institute and Northeast Gas Association

Background. US EPA’s Risk Assessment Guidance for Superfund calls for the determination of site-specific characteristics that may influence fate and transport of contaminants. The partitioning of PAHs between soil and water is typically determined using standard empirical correlations between sorbate octanol-water partition coefficient (Kow) and sorbent organic carbon normalized partition coefficient (Koc). Such empirical correlations are based on PAH partitioning to natural organic matter present in soils/sediments, and do not account for strong binding with carbonaceous matter like coal, coke, or lampblack that may exist in industrial sites. Our earlier work has shown that it is important to understand the nature of contaminant association with sorbents because hydrophobic organic contaminants such as PAHs may bind very strongly to certain organic carbon types found in soils and sediments impacted by industrial processes.

Research Objectives. This research provides fundamental understanding of particle-scale binding and assessment of site-specific partition coefficients for PAHs in sediments from several sites impacted by Manufactured Gas Plant (MGP) operations. The sites range from freshwater in the upper reaches of the Hudson River to the estuarine region downstream. Equilibrium partitioning measurements for PAHs are hindered by strong partitioning of the higher molecular weight compounds to colloidal particulates that are difficult to separate from the aqueous phase. Two recently developed methods are being used to measure equilibrium aqueous concentrations of PAHs: an air-bridge method to keep colloids out of the water phase, and an alum-flocculation method to precipitate colloidal particles from the aqueous phase.

A major objective of this research is to explain scientifically differences between measured and estimated partition coefficients for PAHs. This explanation is aided by density separation of particulate organic matter in sediments, particle-scale PAH measurements, and petrographic characterizations to identify the nature of the organic matter responsible for PAH binding. In our recent work we have demonstrated that PAHs sorbed on carbonaceous particles such as soot, coal, coke, and charcoal are strongly bound and less available than PAHs sorbed on natural organic matter. Thus, identification of the geochemical nature of the soil carbon, specifically the presence of coal, coke, lampblack, and pitch helps explain differences in site-specific values of PAH partitioning from what would have been estimated using standard empirical correlations. Improved understanding of the nature of the sorption process will improve site-specific risk assessment of the impacted sites.