Mineral Spatial Heterogeneity Constrains Permeability Evolution in a Limestone Fracture

Due to the inherent low matrix permeability of caprock formations, any leakage of CO₂ through these formations is likely to occur through fractures or existing wellbores. It is therefore paramount to leakage risk assessments that we understand and are able to predict permeability evolution along these pathways during leakage of CO₂ acidified brine. Resolving and accounting for spatial variations in mineralogy along these leakage pathways may improve the predictive capabilities of models used to estimate the permeability evolution through reactive caprock fractures. In this study the permeability evolution of a single dolomitic limestone fracture is estimated using a 2-D steady-state fluid flow model. The initial aperture field of the fracture is estimated via X-ray computed tomography of a fractured limestone core. Changes in local fracture aperture are evaluated for two different theoretical dissolution scenarios with equivalent increases in total fracture void volume. In the first case, dissolution is assumed to occur uniformly throughout the entire fracture such that all local apertures experience the same increase in aperture. The second case employs a novel combination of X-ray computed micro-tomography and energy dispersive spectroscopy to identify two-dimensional mineral spatial heterogeneity along the fracture pathway. This mineral identification procedure allows for translation of 3-D mineral spatial information into a 2-D map of fracture surface mineralogy. Previous experimental results have demonstrated that the largest increases in fracture aperture will occur at locations where calcite is the dominant mineral exposed at the fracture surface. Using this knowledge, all dissolution in the second scenario is assumed to occur only at locations where calcite is identified to exist at the fracture surface. For a doubling in fracture volume, permeability was overestimated by as much as 40% when dissolution is assumed to occur uniformly along the fracture. This overestimation is due to the existence of bands of less reactive minerals perpendicular to the direction of flow, which are only accounted for in the second model scenario. These bands restrict flow along the fracture, thereby serving to control fracture permeability. Although this result is specific to the particular fracture analyzed in this study, comparison of the two model scenarios highlights the importance of accurately accounting for spatial variability of dissolution along a fracture when predicting permeability evolution.