A Risk Based, Multi-Component Model to Identify Contaminant Loadings and Transport through Groundwater Systems Under Uncertainty
Todd A. Wang
William F. McTernan
AbstractWe developed and applied a suite of risk-based methods for characterizing the contaminant potentials from a former munitions plant in East Texas. The site was originally "clean closed" when a subsequent groundwater monitoring program disclosed areas of contamination by the chlorinated solvent, trichloroethylene (TCE) and others. As part of an overall decision model developed for the site, a series of probability-based mathematical and statistical models were developed to address off-site contamination and plume configuration. As with most historic hazardous waste sites, there was virtually no information relative to contaminant loading rates to the water table aquifer. These loads were reconstructed by comparing the results generated from a Monte Carlo-based technique which linked the vadose and saturated zone models to minimal groundwater data previously collected. The contaminant flux in the aquifer was assumed to coincide with activities at the munitions plants peaking as the plant was decommissioned and tailing off through subsequent years. This curve followed the classic boundary condition where the contaminant source is terminated after a period of flux into the aquifer. Comparisons between simulated data and the site activity curve indicated that the peak of the contamination had occurred before the monitoring program was initiated, generally matching concentrations along the recession limb. Probabilistic transport modeling through the water table aquifer produced a series of statistical distributions of off-site contamination. These curves further corroborated the observation that peak contamination at this site had occurred before the monitoring data were collected. A Bayesian updating technique was applied to compare the revised probabilities associated with various management alternatives and a conditional simulation was completed to define the plume configuration with some statistical confidence.
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.