Development and application of a particle-tracking model for nitrogen transport and transformation in aquifers

This work focuses on quantifying reactive groundwater nitrogen transport through development and application of a new numerical model. I first developed the model "SF Monod" to incorporate multispecies biodegradation and geochemical reactions into an existing particle-tracking code to simulate reactive transport in three-dimensional variably saturated media, with a focus on nitrification and denitrification processes. Biogeochemical reactions are modeled using Monod kinetics; reactive air-phase transport is included so that gases such as nitrogen gas can be tracked. I verified the new code by showing good agreement with analytical solutions, other numerical codes, and laboratory observations. I demonstrated model capability to track sharp fronts by simulating a leaking sewer plume in heterogeneous variably saturated media. In the second part of the work I demonstrated application of the SF Monod model to strategically design placement of woodchip bioreactors for step-pool and channel re-meandering stream restoration scenarios. With only 50% and 53% of full-reach design lengths, the strategically designed bioreactors achieve about 65% and 85% the performance of the original design for the step-pool and channel re-meandering scenarios, respectively, as well as maximum denitrification efficiencies. These results suggest that shorter bioreactors can be installed while still achieving the goal of the restoration project. This work represents a first application of a multispecies reactive transport code to stream restoration design. In the third part of this work, I used the SF Monod model to quantify transport of subsurface nitrogen from septic contamination to a stream network at a Piedmont catchment. I evaluated spatial and temporal distribution of nitrate concentration and denitrification rates, as well as the nitrate concentrations and flux at the groundwater-stream interface. Although the modeled catchment is small and represented by only one kind of contamination (multiple 50- to 60-year-old septic tanks in a 0.23 km2 watershed), the model can be applied to any scale. Because it is particle-tracking based, relatively low spatial resolution can be used for watershed-scale simulations. To my knowledge, no previous efforts have been made to quantify subsurface transport of nitrogen to receiving water bodies at watershed scale using a three-dimensional multispecies reactive transport model that considers biomass metabolism.