BMT3 is a model of the near-field of a hypothetical repository set up in a 2D space of size 50 m in height and 50 m in width, located 500 m below the ground surface, containing 6580 fractures that was created from a 2D simplification of one of the realizations in 3D, generated from on site mapping data from the Stripa Mine, Sweden, see Figure 4a and 4b.
The aim of the BMT3 problem was to evaluate approaches relating to numerical methods, material characterization and computational capacities for coupled THM processes of fractured rocks with a large number of fractures for a near-field study. Eight research teams participated in BMT3 using different computer codes and numerical methods (UDEC of DEM used by CNWRA, INERIS, NGI and VTT; NAPSAC of DFN used by AEA; FLAC of FDM used by ITASCA; THAMES of FEM used by KPH/PNC; and CASTEM of FEM used by CEA), and developed different models (Fig. 4c), and generated different but convergent results (see Fig. 4d and 4e).
The results demonstrated a well-developed BMT model of coupled THM processes of fractured rock masses for performance assessments of near-field problems of GDRW repositories, for different models using different codes and different characterization approaches of equivalent continuum of porous media and discrete blocks systems of fractured rocks. There was remarkable agreements among the teams for the thermal results, but less agreement for the stress results, and very divergent results for the flowrate—indicating the need for more comprehensive and in-depth research on characterization and modeling fractured rock masses, especially crystalline rock masses.