The Department of Defense (DoD) is responsible for the cleanup of groundwater that is contaminated with chlorinated volatile organic compounds (CVOCs) at thousands of sites. Many of these sites are underlain by fractured rocks or soils with significant matrix porosity. As dissolved CVOCs and dense nonaqueous phase liquids (DNAPLs) move through fracture networks, the CVOCs diffuse into the lower permeability matrix materials, where they can remain for hundreds of years. Remediation options for treating fractured geologic media are extremely limited because the low matrix permeability and unknown fracture locations make any type of fluid or chemical delivery difficult or impossible.

Thermal methods hold promise for remediation of fractured media, because heat can be efficiently transferred without any fluid flow by the mechanisms of thermal conduction and electrical resistance heating. Once a fractured rock or soil is heated above the water boiling point, subsequent depressurization of the fracture network by vacuum extraction may induce boiling in the matrix, leading to large gas phase pressure gradients and a steam stripping effect that can remove the contaminants from the matrix.

The objectives of this project were to (1) experimentally and theoretically evaluate the contaminant mass removal process from heated low permeability matrix materials that are bounded by a depressurized fracture network; (2) assess the feasibility of field-scale implementation of thermal remediation techniques at fractured sites, and (3) develop design and operational strategies for maximizing the effectiveness of field-scale applications of thermal remediation methods at fracture sites.

The research used an integrated bench-scale experimental and numerical modeling approach. The primary fluid, heat, and contaminant transport processes during matrix boiling and fracture depressurization occur at the scale of a single matrix block. Laboratory tests consisted of heating contaminated low permeability rocks and soil cores to temperatures above the normal boiling point and depressurizing a fracture at one end of a sealed column. Temperature, steam discharge, and contaminant recovery data from the cores during boiling were collected and used to assess the phenomena. A multiphase heat and mass transfer numerical model was used to analyze the experiments and to evaluate the boiling and contaminant mass transfer over a wider range of fractured porous media characteristics.

Key findings of this work were that CVOCs were readily removed from low to moderate permeability fractured materials through the boiling mechanism, and it was not necessary to boil all of the pore water to get complete contaminant removal by this mechanism. Simulation results showed that remediation efficiency was sensitive to the amount of heat added to the system. In some cases, the addition of a relatively small additional amount of heat can greatly improve the remediation efficiency. The contaminant recovery from fractured rock masses in simulations was relatively insensitive to the fracture properties, assuming that the fractures were continuous and well-connected. The simulations were sensitive to the unfractured matrix permeability, with lower permeability corresponding to somewhat lower removal efficiencies for a given thermal input. However, at higher thermal input, even the low permeability matrix simulations showed effective remediation.

This project confirms the ability of thermal remediation to remove volatile contaminants from the matrix in fractured geologic media. The experimental and numerical methods developed will aid the future design and analysis of thermal remediation projects at fractured sites.