Discrete Fracture Network Multiapp

Example models of discrete fracture networks (DFN) loosely coupled with the surrounding matrix using the Multiapps system are given in this section. These examples are based on the multiapp fracture flow example given in porous_flow module. The first example provides a comparison to an analytic solution by Gringarten et al. (1975) for a single infinite crack in an infinite domain. The second example demonstrates the use of automatic mesh refinement to refine the matrix mesh around the region containing the DFN from the porous_flow multiapp fracture flow example.

Gringarten solution

This fracture-flow validation problem was created by Koenraad Beckers at NREL to ensure that FALCON for fracture-based reservoirs is set up correctly by comparing FALCON simulation results with the analytic solution by Gringarten et al. (1975). Table 1 lists the FALCON parameter values for this validation problem. The FALCON matrix domain and embedded fracture, with temperature after 3 years of operation, are shown in Figure 1 and Figure 2. Given that the Gringarten solution neglects heat transfer effects in the vertical direction, only one mesh layer was assumed in the vertical direction with a no-flux boundary condition on top and bottom. Figure 3 indicates that the FALCON simulation result for the production temperature is in good agreement with the analytical solution for finite spaced and single fracture solutions. Input files for these simulations are found here:

Matrix Mainapp Input File

Fracture Subapp Input File

Table 1: Parameter values for fracture flow validation problem with FALCON

ParameterValue
Rock initial temperature90C
Rock density2875 kg/m
Rock heat capacity825 J/kg-K
Rock thermal conductivity2.83 W/m-K
Rock permeability1e-16 m
Rock porosity0.1
Water flow rate0.1 kg/s
Water injection temperature30C
Domain length100 m
Domain width40 m or 200 m
Domain height10 m
Well spacing100 m
FALCON matrix domain with temperature (in K) after 3 years of operation for fracture-flow validation problem.  The domain width is 40m where the fracture half width spacing is 20m . A fracture is embedded in the center of this matrix domain.

Figure 1: FALCON matrix domain with temperature (in K) after 3 years of operation for fracture-flow validation problem. The domain width is 40m where the fracture half width spacing is 20m . A fracture is embedded in the center of this matrix domain.

FALCON embedded fracture mesh with temperature profile (in K) after 3 years of operation. Fluid injection in the fracture occurs at the left side and fluid production is from the right side.

Figure 2: FALCON embedded fracture mesh with temperature profile (in K) after 3 years of operation. Fluid injection in the fracture occurs at the left side and fluid production is from the right side.

Production temperature of fracture-flow validation problem as simulated with FALCON is in good agreement with analytical solution by Gringarten for finite and infinitely spaced fractures.  The legend gives the fracture half width spacing.

Figure 3: Production temperature of fracture-flow validation problem as simulated with FALCON is in good agreement with analytical solution by Gringarten for finite and infinitely spaced fractures. The legend gives the fracture half width spacing.

References

  1. AC Gringarten, PA Witherspoon, and Yuzo Ohnishi. Theory of heat extraction from fractured hot dry rock. Journal of Geophysical Research, 80(8):1120–1124, 1975.[Export]