Salt rock deformation and evolution: damage development and healing in rock salt

Abstract

Rock salt formations have already proven to be excellent caprocks for containing gas and oil over geological timescales. Upon mechanical damage, the high ductility of salt formations ensures that porosity and permeability evolve to very low values over time (e.g. Hunsche and Hampel, 1999). In addition, fluid-assisted diffusive mass transfer processes, such as dissolution-precipitation and recrystallization processes, further reduce pore-connectivity and permeability, ensuring healing and sealing (Houben et al., 2013). This self-healing capacity could potentially make rock salt caverns ideal storage locations for compressed air, hydrogen, natural gas and nuclear waste. However, damage induced during salt mining may impact cavern sealing integrity, forming a permeable zone in which porosity will increase due to microcracking. The aim of this project is to investigate crack evolution and healing properties to predict the timescale at which cracks will close again, and by which processes, to guarantee the safety of the storage. Microstructure analysis on intact and damaged rock salt provides an overview of the different deformation mechanisms that might affect rock formations. The starting material is characterized by a dilatant behaviour in which intergranular cracks and triple junctions are dominant; while at high confining pressure salt rock undergoes non-dilatant behaviour at in which recrystallization processes might modify the entire microstructure. The material deformed at low confining pressure provides information about the transitional phase between dilatant and non-dilatant behaviour, showing both dislocation structures and recrystallized grains over relict of old ones. Measurements of seismic velocities under compression and time-lapse X-ray tomography are used to monitor the microcracks evolution and healing. Compression experiments on Beberthal sandstone are done as control experiment to assess if crack density and fracture network can be quantified by wave velocities measurements under triaxial compression and then used on salt rock. The experiment represents a great tool to analyse crack evolution, and consequently healing, by monitoring the change in seismic velocities and therefore an in-depth analysis is suggested, but could not be executed due to time-constraints. CT-scanning imaging are done on compacted aggregates of granular rock salt to monitor cracks and investigate its healing, without inducing damage to the sample. Prior to the experiment, a theoretical model is derived in this study to predict the healing behaviour of microcracks that contain a fluid film at the crack surface. A comparison between the samples scanned at different days show differences in crack size and evolution, confirming that healing processes occurred during the time. The quantitative crack measurements on sample VD14 (2 and 3 days exposure of cracks to brine) are eventually compared with the theoretical data on the progress of crack healing obtained using different values of diffusion coefficient and thickness of the fluid film. In some cases quantitative results show lower values of crack healing then the theoretical values obtained using 1,96E-22 as value for DCδ/a (found using D= 10E-9 m2/s and δ= 280E-9 m from Houben et al., 2013) and are more similar to the ones obtained using a value of 7,00E-25 for DCδ/a (found using D= 10E-18 m2/s and δ= 1E-9 m from Koelemeijer et al., 2012); this is probably due to a different thickness of the fluid film at the crack surface. Further research is suggested to estimate and predict with more precision the evolution of cracks in salt rock.