Role of upper mixed layer depth in forecasting contaminated sediment recovery rates

Abstract

A critical step in evaluating and selecting contaminated sediment remediation options is the simulation of a long-term future trend of exposure concentrations that will occur under natural conditions without active in situ remediation. This ‘natural attenuation’ forecast provides a baseline reference against which to compare alternative sediment management options. Model development and application at a number of contaminated sediment sites has indicated that one of the most important parameters for accurately simulating natural attenuation is the depth of the upper mixed (i.e., ‘active’) sediment layer. The upper mixed sediment layer essentially acts as a biologically available contaminant reservoir, determining risks to ecological and human health. Contaminant concentrations in the upper mixed sediment layer are dynamic and determined by: the sediment inventory created by historical loadings to the sediment; a balance of fate and transport processes both into and out of the mixed layer at the sediment water interface; and burial of contaminants from the mixed layer into deeply buried ‘inactive’ sediments. Model sensitivity analyses at one of our sediment management sites, the Lower Fox River, has demonstrated that the rate at which polychlorinated biphenyl exposure decreases with time under natural processes is inversely proportional and very sensitive to the depth of mixing in the surficial sediments of the system. These studies have shown that increasing the assumed mixed layer depth in the model from 10 cm to 30 cm increases the simulated recovery half-time—as measured by the rate of change of the upper layer sediment polychlorinated biphenyl concentration—from about 20 years to more than 100 years. In the case of the Lower Fox River, radioisotope and polychlorinated biphenyl concentration depth profiling supports a 10 cm mixed depth layer. Implementing a 30 cm mixed depth in the model smears the existing polychlorinated biphenyl profile, artificially mixing higher concentrations in the 10 to 30 cm layer with lower concentrations in the 0 to 10 cm layer, resulting in the increased simulated recovery times.

Keywords:

• PCBs
• contaminant fate and transport modeling
• remediation
• natural attenuation

Acknowledgments

We gratefully acknowledge the support of the Fox River Group, who funded studies on which this paper is based. We also thank colleagues at Blasland, Bouck and Lee, who conducted the 2000–2001 Fox River sampling program. Nevertheless, the views expressed here are our own, as is the responsibility for any errors or omissions.