Development of source functions for modeling dissolution of residual DNAPL fingers in the saturated zone

Abstract

After more than a decade of effort to remediate many waste sites, it has become evident that a major obstacle to meeting F-PA,s 'clean-closed, standard is the presence Of dense non-aqueous phase liquids at many con contaminated sites. These chemicals are characterized by low aqueous solubility, a greater density water, toxicity to humans at low concentrations, and Persistence of the source for tens to hundreds of years. The particular source of interest in this research is a residual DANIEL fmger in the saturated zone. Modeling dissolution from a residual fmger can be approached in one of two ways: as an equilibrium process or as a rate dependent kinetic process. Development of a source term for modeling the dissolution from remediation DNAPL is necessary in Order to Predict the fate of the dissolved con evaluate the persistence of the source, and plan the remediation of a DNAPLcontaminanted site. Two source terms for modeling the dissolution from a residual fmger m the saturated zone are developed: a local equllibrium model and a Idnetic mass transfer rate coefficient model. Fach of these terms were then -mcorporated into the United States Geological Survey's method of Characteristic (MOC) code and their accuracy and reliability were evaluated by comparing model predictions to published results of DNAPL dissolution experiments.. The local equllibrium approach assumes that local equilibrium conditions are achieved rapidly at the interface between the residual DNAPL globule and the flowing ground water. Assumming that local equilibrium conditions were acheived within the first hour hour of contact produced a poor fit of predicted to experimental values. However, an equilibrium model which assumes that equilibrium conditions are not achieved until the ground water has been in contact with approximately 10 centimeters of contaminated media produced an excellent fit between predicted and experimental values. In the mass transfer rate coefficient model, a mass transfer rate coefficient is determined from dimensionless numbers which are dependent on porous media and fluid characteristics. Several experimental correlations were used to calculate the mass transfer rate coefficient with the correlations determined by Pfannnkuch (1984) and Wilson and Geankopolis (1966) producing the best fit of numerical to experimental data.