Hydraulic-conductivity reduction, reaction-front propagation, and preferential flow within a model reactive barrier

Recent work has demonstrated the utility of reactive barriers for transformation or immobilization of contaminants in ground water. However, reaction-induced changes in hydraulic conductivity may compromise the reactive barrier by diverting flow or focusing breakthrough of contaminants. For an acidic, ferric solution flowing through calcareous sand in column experiments, we investigated how hydraulic-conductivity reduction, reaction-front propagation, and fingering depend upon the reactive solid volume, reactive surface area, and pore-water velocity. Hydraulic-conductivity reductions are greater with larger initial pore-water velocity, calcite surface area, or calcite volume. Reductions in hydraulic conductivity (up to four orders of magnitude) result primarily from CO2 (g) exsolution rather than from ferric oxyhydroxide precipitation and may be at least partly reversed as bubbles migrate. Although fingering is both driven and repressed by pore-scale hydraulic changes, the velocity of the reaction front, width of the primary reaction zone, and maximum length of the largest finger appear to be insensitive to macroscopic changes in hydraulic conductivity at a constant flow rate. Reaction-front velocity increases as the ratio of initial calcite volume to pore-water velocity decreases, whereas zonal width and maximum finger length appear to increase as the Damköhler number decreases for a given calcite grain-size distribution. These results offer guidelines for improving the efficiency of reactive barriers when the reaction rate equals or exceeds the rate of mass transfer.