ABSTRACT
Hosted in predominantly volcanic material, the Snake River Plain aquifer in south east Idaho (United States) is an extremely heterogeneous medium where hydraulic conductivity (K) can vary by up to five orders of magnitude over distances of less than a meter vertically and a few hundreds of meters horizontally. While the bulk of the aquifer is only moderately conductive, the zones of contacts between lava flows, often containing rubble, are highly conductive and form preferential pathways for contaminant migration. In this study, a numerical experiment using a Monte Carlo approach is carried out to study the impact of K variability on the effectiveness of detection monitoring networks. Several scenarios with different types of source and different distances from the source to the monitoring network were considered and the probability for monitoring failure or inaccuracy was quantified for each of these scenarios. The results of this experiment confirm the challenges posed by ground water monitoring in the Snake River plain aquifer.
