Virus Fate and Transport in Groundwater : Organic matter, uncertainty, and cold climate

Sammanfattning: Water managers must balance the need for clean and safe drinking water with ever-increasing amounts of waste-water. A technique for treating and storing surface water called “managed aquifer recharge” (MAR) is frequently used to help maintain this balance. When MAR is used to produce drinking water, water managers must ensure that disease-causing microbial contaminants are removed from the water prior to its distribution. This thesis examined the processes responsible for removing a specific class of microbial contaminants called “enteric viruses” during MAR. Viruses are naturally removed in groundwater through adsorption and inactivation mechanisms. This thesis investigated how these virus removal mechanisms were affected by ionic strength (IS), dissolved organic carbon (DOC), and the age of the sand used in a MAR infiltration basin. This was done using batch and flow-through column experiments designed to mimic conditions characteristic of a basin infiltration MAR scheme in Uppsala, Sweden. Bacteriophage MS2 was used as a proxy for enteric viruses. All of the experiments were conducted at 4°C. Experimental data were modeled to describe the fate and transport of viruses in the infiltrated groundwater. Conventional least-squares optimization and generalized likelihood uncertainty estimation (GLUE) were compared as model fitting-approaches in order to determine how data uncertainty affects parameter estimates and model predictions. Results showed that the sand used in the infiltration basins accumulates adsorbed organic matter as it is exposed to infiltrating surface waters. This reduced the amount of MS2 that was removed due to adsorption and inactivation. Results from GLUE indicated that MS2 is more likely to inactivate in a time-dependent manner when in the presence of sand with high concentrations of organic matter. Both model fitting techniques indicated that virus attachment rates were significantly lower for sand with high organic carbon content. Neither methodology was capable of adequately capturing the kinetics of virus adsorption. Uncertainties in the experimental data had a large effect on the conclusions that could be drawn from fitted models. This study showed that the presence of natural organic matter reduces the value of the infiltration basin as a microbial barrier.