H088-0007
Effects of Recirculating Flows on Solute Transport with Solid-Fluid Reactions in Rough Fractures

Thursday, 10 December 2020
Poster
Woonghee Lee, University of Minnesota Twin Cities, Minneapolis, MN, United States, Seonkyoo Yoon, University of Minnesota Twin Cities, Department of Earth and Environmental Sciences, Minneapolis, MN, United States and Peter K. Kang, University of Minnesota, Department of Earth and Environmental Sciences, Minneapolis, MN, United States
Abstract:
Anomalous transport is widely observed in fractured geologic media, and recirculating flows in rough fractures have been shown to induce anomalous transport. Recirculating flows are developed by the interplay between fluid inertia and wall roughness and induce preferential flows by reducing effective cross-sectional area and increase residence times by trapping effects. However, the effects of recirculating flows on reactive transport with heterogeneous (solid-fluid) reactions have not been comprehensively studied.

In this study, we investigate solute transport with heterogeneous reactions and propose an upscaled reactive transport model that effectively captures anomalous reactive transport. We use a particle tracking method to simulate irreversible heterogeneous bimolecular reactive transport at different levels of fracture roughness, Reynolds (Re), and Peclet (Pe) numbers. We inject A particles as reactants at the inlet and generate C particles as products when A particles hit fracture surfaces (Fig. 1(a)). We found that recirculating flows cause anomalous transport by increasing the residence time of C particles through trapping effects and also facilitate the reaction by increasing residence time of A particles near fracture walls. The fracture roughness and Re determine the development of recirculating flows, and Pe also plays an important role in the particle trapping and reaction process. Based on the improved understanding, we propose an upscaled reactive transport with velocity-dependent effective reaction rates. We compare upscaled models with velocity-dependent and velocity-independent reaction rates and show that the velocity-dependent reaction rate improves the predictability of the upscaled model (Fig. 1(b)).