Upscaling particle transport in discrete fracture networks: 2. Reactive tracers

Abstract

We study sorbing tracer transport through discrete fracture networks using a stochastic Lagrangian framework, combined with the methodology for upscaling particle breakthrough curves developed in the first part of this article series. Results indicate that this procedure can accurately predict expected normalized tracer discharge for an upscaled distance of 1 order of magnitude in terms of transport scale, which for our simulations is about 2 orders of magnitude greater than the mean fracture segment scale. Specifically, we show the importance of retaining the correlation between the water residence time τ and the hydrodynamic control of retention β in order to make accurate tracer discharge predictions. Also, we show that the extreme tails of τ and β distributions have essentially no impact on tracer discharge. These results are illustrated using the unlimited diffusion model, and for two hypothetical tracers with properties designed to capture the behavior of many commonly occurring natural radionuclides.