Modeling the impact of natural roughness of tension joints on heat transport

The understanding of heat transport in fractures is crucial for mining geothermal systems. Studies of heat transport in natural fractures at scales comprised between those of laboratory experiments and those of field tracer tests are seldom. To bridge the gap, a joint surface with characteristic plumose was scanned in the field using LiDAR technology. The scanned surface was used to build a numerical model of mode 1 fracture. Fluid flow and heat transport were modeled solving the steady-state Stokes equation and assuming Fourier transport, respectively. We considered three different fracture apertures and varied systematically roughness in order to investigate the impact of plumose on fluid and heat transport. The 3D velocity flow fields were characterized by mean hydraulic aperture and by statistics on the directional components of the velocity vector. The method of temporal moments was used to extract first and second moments from temperature breakthrough curves. Heat transport parameters (local and macroscopic) were calculated from first and second moments.

We show that hydraulic aperture and the longitudinal component of the velocity vector decrease with increasing roughness. The local variation of heat transport parameters is controlled by fracture roughness. For the macroscopic transport parameters, several transport regimes were identified. At low fracture aperture (i.e. 1 mm), conductive regime dominates heat transport in agreement with low Péclet numbers. In this case, fracture roughness affects the transport parameters via the loss of hydraulic aperture. With higher aperture (i.e. 3 mm) geometrical dispersion regime is dominant, roughness controlling the amplitude of transport parameters. At 5 mm aperture, transition from geometrical to Taylor dispersion occurs and the roughness tends to decrease dispersion and dispersivity according to the mean flow velocity.