Impact of Matrix Diffusion on Heat Transport Through Heterogeneous Fractured Aquifers

International audience ; Transport in subsurface fractured rocks is strongly affected by the variability of the velocity field, which in complex fracture networks may vary broadly. Because the rock matrix is almost impervious, the flow occurs in the network of connected fractures, which exhibits multiscale features spanning several orders of magnitude, that is, fracture size ranging from millimeters to hundreds of meters, with a small number of fractures acting as preferential flow paths where most of the fluid flow channelizes (Bonnet et al., 2001; Goc et al., 2010; C. F. Tsang & Neretnieks, 1998). Heterogeneity at different scales in fracture aperture and hydraulic conductivity implies strong heterogeneity in flow rates and diffusive processes, leading to behaviors that cannot be captured by traditional theories, and field-scale observations that are difficult to predict (Becker & Shapiro, 2003). Understanding the impact of multiscale heterogeneity and network connectivity on advective and advective-diffusive transport is critical for many scientific and engineering applications including geothermal energy, nuclear waste disposal, waste-water injection, and water resource protection. It is key to improve our capacity of characterizing the geometry of the fracture network and identifying connected structures by means of field tests, such as conservative solute or heat tracer experiments. They typically comprise the interpretation of breakthrough curves (BTCs) to yield information on the structure of the fractured systems which drive flow and transport processes. In natural fractured media, transport of both solutes and heat behaves different from traditional Fickian models, where a constant dispersion coefficient can describe the impact of flow fluctuations occurring at the small scales on solute and thermal spreading. Solute transport through fractured media often exhibits anomalous (non-Fickian) behavior, with non-symmetrical BTCs characterized by power-law post-peak tails (Berkowitz & Scher, ...