Enhanced quantitative characterisation of the geometrical features and variability of the fracture network properties is an essential prerequisite to model fracture networks in crystalline reservoirs. By combining regional digital elevation model analysis and local outcrop investigation, eleven outcrops from various lithologies were investigated through a structural framework (fracture dimension, orientation, clustering and spacing). The structural dataset presented here contains the regional DEM interpretation at two resolutions (25 m , 5 m) of the Northern Upper Rhine Graben and the Odenwald. Additional statistics on LiDAR and GIS structural interpretation of eleven outcrops are provided. Fracture length, orientation, dip, density and intensity are calculated for each profile. Power-law parametrisation is extracted from the length distribution of 2D interpreted maps. This parametrisation of the fracture network is then used to build 24 standalone DFN models for two sites, from the Lichtwiese and Heppenheim datasets, in Fracman ®. These models consider various configurations of the fracture termination, percentage of active fracture, and aperture estimation methods. The outputs of the models in terms of permeability and fracture aperture are provided here. This dataset supplements another publication (Bossennec et al., submitted) that presents the structural organisation of crystalline rocks from several crystalline lithologies in the Odenwald and first-order DFN modelling of associated properties at different specific depths (800 m and 4000 m).