Abstract

The colloid-facilitated solute transport model, based on HYDRUS-1D, was evaluated using the column experimental data of Pang et al. (2005) for cadmium (Cd) transport facilitated by B. subtilis spores or E. coli in saturated coarse alluvial gravels. We simulated Cd transport involving convection, dispersion, kinetic adsorption/desorption to/from the aquifer media and to/from mobile/immobile bacteria, and kinetic attachment/detachment of the bacteria to/from the aquifer media. To reduce the number of parameters to be optimized, we independently estimated Cd sorption/desorption rates to mobile bacteria from a batch study. The model described the collected experimental data reasonably well. Extensive sensitivity analysis to various reaction parameters was carried out to obtain an understanding of the relative importance of individual model parameters on model predictions. Our modeling results suggest that the rates of Cd sorption or desorption differ not only between different bacterial species but also between unattached and deposited bacteria. The results of the sensitivity analysis indicated that the Cd sorption rate to unattached bacteria had a significantly greater impact on the model results than its sorption rate to deposited bacteria. For the experimental system investigated here, model results were most sensitive to parameters describing interactions between Cd-aquifer media, bacteria-aquifer media, and Cd-mobile bacteria, and they were less sensitive to interactions between Cd-immobile bacteria and desorption rate from mobile bacteria.

view journal