This work presents a detailed three-dimensional finite element based model for wave propagation, combined with a postprocessing procedure to determine the fracture intensity caused by blasting. The data generated during this project includes output files of all simulations with detailed fields, geometries and meshes. The model incorporates the Johnson-Holmquist-2 constitutive model, which is designed for brittle materials undergoing high strain rates and high pressures and fracturing, and a tensile failure model. Material heterogeneity is introduced into the model through variation of the material properties at the element level, ensuring jumps in strain. The algorithm for the combined Johnson-Holmquist-2 and tensile failure model is presented and is demonstrated to be energy-conserving, with an open-source MATLABTM implementation of the model. A range of sub-scale numerical experiments are performed to validate the modelling and postprocessing procedures, and a range of materials, explosive waves and geometries are considered to demonstrate the model's predictive capability quantitatively and qualitatively for fracture intensity. Fracture intensities on 2D planes and 3D volumes are presented. The mesh dependence of the method is explored, demonstrating that mesh density changes maintain similar results and improve with increasing mesh quality. Damage patterns in simulations are self-organising, forming thin, planar, fracture-like structures that closely match the observed fractures in the experiments. The presented model is an advancement in realism for continuum modelling of blasts as it enables fully three-dimensional wave interaction, handles damage due to both compression and tension, and relies only on measurable material properties. The uploaded data are the specific simulation outputs for four explosion models occurring on two different rock types, and the specific fracture patterns generated.

Herein lies the supporting data for the paper 'Small-scale capillary heterogeneity linked to rapid plume migration during CO2 storage'. We supply experimental, analytical and numerical simulation data used in the paper. The supplied zipped folders follow the same order as the main paper, with codes to reproduce each figure (and those in the supporting information PDF). There are also video files (in the 5_Field_scale_simulation zipped folder) showing the final CO2 plume evolution from the static images in the main paper Figure 4. Descriptions of each of the folders are given below: 0 - README. This contains detailed instructions on the data and using the supplied files. 1 - Scaling analysis. This contains the scaling analysis analytical methods, with figure generation for Figure 1 in the main paper. 2 - Petrophysics. This contains all the petrophysical experimental data, analysis files and core flood simulation files. This is used to produce Figure 2 in the main paper. 3 - Fine_resolution_simulations. This contains the simulation files, Matlab post processing files and figure generation for the fine resolution simulations, presented in Figure 3 in the main paper. 4 - MIP_upscaling. This contains simulations files, Matlab post processing files and figure generation for the macroscopic invasion percolation scheme. The results of this are presented in the supporting information document. 5 - Field_scale_simulation. This contains the simulations files, Matlab post processing files and figure generation for the final field scale simulations in the main manuscript Figure 4 and in the supporting information. In each folder are separate READMEs containing specific information relevant for the included files.