Date of Award

12-2009

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Environmental Engineering and Science

Advisor

Falta, Ronald W

Committee Member

Molz , Fred J

Committee Member

Murdoch , Lawrence C

Committee Member

Fjeld , Robert A

Abstract

The complex processes and expensive costs of source and plume remediation of dense, non-aqueous phase liquid (DNAPL) complicate the decision-making process for site remediation. Selection of remediation alternatives has been a big challenge due to the lack of tools that simultaneously evaluate the effectiveness of source and plume remediation and access the uncertainties in all major parameters. In this research, a new probabilistic remediation model, Probabilistic Remediation Evaluation Model for Chlorinated solvents sites (PREMChlor), has been developed. This is achieved through linking the analytical model REMChlor to a Monte Carlo modeling simulation package GoldSim via a FORTRAN Dynamic Link Library (DLL) application. PREMChlor can simultaneously evaluate the effectiveness of source and plume remediation considering the inherent uncertainties in all major parameters. In PREMChlor, all of the uncertain input parameters are treated as stochastic parameters represented by probability density functions (PDFs). The outputs from the PREMChlor model are probability distributions and summary statistics of those distributions. This new model considers common technologies for DNAPL source removal and dissolved plume treatment. A license-free file containing the graphical user interfaces has been generated to make the PREMChlor model available for use by others.
In model demonstration, probabilistic simulations show the different probabilities of meeting a remediation goal for different combinations of source and plume remediation scenarios considering uncertainties in input parameters. The PREMChlor model has been applied to a trichloroethene (TCE) plume in a shallow aquifer at a manufacturing plant. The calibrated model using a deterministic approach is able to closely match the pre-remediation site condition. Probabilistic simulations predicting the effects of remediation show the overall uncertainty in TCE concentration propagates over time given uncertainties in key input parameters. Probabilistic simulations capture most uncertainties in key parameters based on estimated PDFs. The PREMChlor model has also been used to conduct sensitivity analyses by assessing the influence or relative importance of each input parameter on plume behavior, in terms of contaminant mass concentration, for three different plume types. It is found that the degree of influence of different input parameters on the contaminant mass concentration varies widely for different plume types. The overall uncertainty of the contaminant mass concentration is reduced greatly by the remediation effort in all three plume types.

Share

COinS