In Situ Vitrification (ISV) is a promising technology, under development by DOE, for long-term stabilization of hazardous waste buried in shallow sites. It is based on melting contaminated soil at the site and letting it cool. The melt solidifies into a glassy or polycrystalline rock incorporating the waste, thus significantly reducing the leaching of contaminants to groundwater. Melting is induced by applying electrical power to the ground through electrodes inserted vertically into the soil to be melted. Melting proceeds downward and produces a roughly hemispherical body upon cooling. A hood is placed over the site to confine gases and particulates released from the melt and direct them to an off-gas treatment system. The stages of the ISV process are depicted in Figure 1.

Buried wastes, containing radioactive, organic, metallic, and combustible materials, are a large component of DOE waste sites that will require remediation. The advantages of ISV for such wastes include: (1) hazardous organic contaminants and combustible materials are pyrolyzed and destroyed, leading to volume reduction and avoidance of future site subsidence, (2) radionuclides are incorporated into glassy or crystalline phases upon cooling of the melt, resulting in reduced mobility, and (3) metallic components are melted, thus minimizing volume and surface area.

A well-instrumented pilot-scale field test was conducted at ORNL in May 1991. We have analyzed the massive data collected and developed various models in order to: check consistency of data; understand the processes involved and which ones dominate; determine effective values of parameters (such as thermal and electrical conductivities); explain what is observed, geochemically and thermally; and develop effective estimation and simulation tools. For example, among other things, we developed a simple non-invasive method for determining the melt temperature from amperage and voltage data (Figure 2). Such indirect procedures are indispensable in applications to existing highly contaminated sites. Among the crucial issues that modeling can elucidate is the fate of water vapor beneath the melt and the conditions under which a water-saturated zone can form.

Cooling and solidification of the melt is of great geological interest. We developed a detailed model and simulation code for cooling and solidification of a binary magma, featuring: coupled heat conduction and solute diffusion, binary crystal-melt thermodynamics, constitutional supercooling, temperature and composition dependent thermophysical properties, conductive cooling of surrounding soil. The description is macroscopic in terms of local variables (concentration, enthalpy, temperature, solid fraction); conservation laws valid everywhere in weak (integral) sense, phases distinguished only by values of solid fraction. In this "volume-of-fluid" approach, no explicit tracking of fronts is needed, which is particularly convenient for computations. We have applied it to Diopside-Anorthite binary and also to Feldspar-Pyroxene pseudobinary with very good results: matching of simulated and experimental cooling curves (Figure 3), can determine effective values of parameters (e.g. conductivities) and sensitivities, and can simulate various cooling scenarios.