No abstract

Basalt formations are often fractured, which poses the risk of CO 2 leakage from a geological CO 2 storage reservoir. Despite their fractured nature, basalts are considered for large-scale CO 2 injection and storage due to large potential storage capacity and rapid carbon mineralization (Matter et al., 2016;McGrail et al., 2016;Xiong et al., 2018). The advantage of fractures is the connectivity to a larger pore network adjacent to fracture surfaces, which provide a large mineral reactive surface area for chemical reactions, and may play a dominant role in changing the porosity of basaltic rock. Mineral precipitation in fractures and connected pores could act as a self-sealing process in the case of a CO 2 leakage from a CO 2 storage reservoir, thus reducing the risk of the longterm leakage of CO 2 . Mineral precipitation in basalt fractures has been observed in natural analogue systems where hydrothermal fluids have migrated upwards through basalts over geological time (Eggleton et al., 1987;Goldberg et al., 2008;Rogers et al., 2006). However, a range of factors including the flow velocity of the CO 2 -enriched plume, fluid chemistry, pressure and temperature conditions, and the fracture aperture size need to be considered in order to predict the rate and extent of the potential self-sealing of fractures (Brunet et al., 2016).Several laboratory studies have been performed to understand the influence of transport limitations on the extent of mineral formation and pore network alterations in natural basalt cores when exposed to acidic CO 2 -rich fluids under relevant geologic storage conditions (

No abstract

We consider the axisymmetric displacement of an ambient fluid by a second input fluid of lower density and lower viscosity in a horizontal porous layer. If the two fluids have been segregated vertically by buoyancy, then the flow becomes self-similar with the input fluid preferentially flowing near the upper boundary. We show that this axisymmetric self-similar flow is stable to angular-dependent perturbations for any viscosity ratio. The Saffman–Taylor instability is suppressed due to the buoyancy segregation of the fluids. The radial extent of the segregated flow is inversely proportional to the viscosity ratio. This horizontal extension of the intrusion eliminates the discontinuity in the pressure gradient between the fluids associated with the viscosity contrast. Hence at late times, viscous fingering is shut down even for arbitrarily small density differences. The stability is confirmed through numerical integration of a coupled problem for the interface shape and the pressure gradient, and through complementary asymptotic analysis, which predicts the decay rate for each mode. The results are extended to anisotropic and vertically heterogeneous layers. The interface may have relatively steep shock-like regions, but the flow is always stable when the fluids have been segregated by buoyancy, as in a uniform layer.

Micro-computed tomography (micro-CT) is increasingly utilized to image the pore network and to derive petrophysical properties in combination with modelling software. The effect of micro-CT image resolution and size on the accuracy of the derived petrophysical properties is addressed in this study using a relatively homogenous sandstone and a heterogenous, highly porous bioclastic limestone. Standard laboratory procedures including NMR (nuclear magnetic resonance) analysis, micro-CT analysis at different image resolutions and sizes and pore-scale flow simulations were used to determine and compare petrophysical properties. NMR-derived pore-size distribution (PSD) was comparable to the micro-CT-derived PSD at a resolution of 7 µm for both the rock types. Porosity was higher using the water saturation method as compared to the NMR method in both rocks. The resolution did not show a significant effect on the porosity of the homogeneous sandstone, but porosity in the heterogeneous limestone varies depending on the location of the sub-sample. The transport regime in the sandstone was derived by simulations and changed with the resolution of the micro-CT image. The transport regime in the sandstone was advection-dominated at higher image resolution and diffusion-dominated when using a lower image resolution. In contrast, advection was the dominant transport regime for the limestone based on simulations using higher and lower image resolutions. Simulation-derived permeability for a 400 Voxel3 image at 7 µm resolution in the Berea sandstone matched laboratory results, although local heterogeneity within the rock plays an integral role in the permeability estimation within the sub-sampled images. The simulation-derived permeability was highly variable in the Mount Gambier limestone depending on the image size and resolution with the closest value to a laboratory result simulated with an image resolution of 2.5 µm and a size of 300 Voxel3. Overall, the study demonstrates the need to decide on micro-CT parameters depending on the type of petrophysical property of interest and the degree of heterogeneity within the rock types.