Stress-Induced Anomalous Transport in Natural Fracture Networks

Peter K. Kang, Qinghua Lei, Marco Dentz, Ruben Juanes

Research output: Contribution to journalArticlepeer-review

13 Scopus citations

Abstract

We investigate the effects of geological stress on fluid flow and tracer transport in natural fracture networks. We show the emergence of non-Fickian (anomalous) transport from the interplay among fracture network geometry, aperture heterogeneity, and geological stress. In this study, we extract the fracture network geometry from the geological map of an actual rock outcrop, and we simulate the geomechanical behavior of fractured rock using a hybrid finite-discrete element method. We analyze the impact of stress on the aperture distribution, fluid flow field, and tracer transport properties. Both stress magnitude and orientation have strong effects on the fracture aperture field, which in turn affects fluid flow and tracer transport through the system. We observe that stress anisotropy may cause significant shear dilation along long, curved fractures that are preferentially oriented to the stress loading. This, in turn, induces preferential flow paths and anomalous early arrival of tracers. An increase in stress magnitude enhances aperture heterogeneity by introducing more small apertures, which exacerbates late-time tailing. This effect is stronger when there is higher heterogeneity in the initial aperture field. To honor the flow field with strong preferential flow paths, we extend the Bernoulli Continuous Time Random Walk model to incorporate dual velocity correlation length scales. The proposed upscaled transport model captures anomalous transport through stressed fracture networks and agrees quantitatively with the high-fidelity numerical simulations.

Original languageEnglish (US)
Pages (from-to)4163-4185
Number of pages23
JournalWater Resources Research
Volume55
Issue number5
DOIs
StatePublished - May 2019

Fingerprint Dive into the research topics of 'Stress-Induced Anomalous Transport in Natural Fracture Networks'. Together they form a unique fingerprint.

Cite this