Sediment PAH Allocation Using Parent PAH Proportions and a Least Root Mean Squares Mixing Model

Abstract

This article presents the case study of a forensic approach used to apportion polycyclic aromatic hydrocarbon (PAH) sediment impacts among two upland legacy contamination source areas-a former manufactured gas plant (MGP) Site and a mixed-use landfill/lumber mill/wood-treating facility-at a Great Lakes National Priorities List (NPL) site. The method uses "parent" (non-alkylated) PAH proportions to empirically define the in-situ mass signature of each source and the sediment impact. Six different PAH proportions, based on more than 450 individual soil and sediment samples, were used to characterize the PAH mass within each area. A mixing model was then used to predict the mass percent contribution from each source that most closely matched the observed sediment impact, based on least root mean square error (RMSE) between the modeled and measured sediment PAH proportions. The model was able to match the observed sediment impact with a high degree of correlation (R 2 of 96%).