Tomography and rock deformation

One of the critical challenges in Carbon Capture and Storage (CCS) is to ensure that the CO2 is stored safely in the reservoir over a long time. The approach is to monitor the stability based on seismic surveys that measure the change in seismic wave propagation as the amount of stored CO2 increases. This is a well-established tool to monitor high CO2 concentrations. To also access the regime of low CO2 concentrations, imaging of two-phase fluid saturations and understanding the effect on the mechanical behavior on the laboratory scale is the key to ensure precise CO2 monitoring. Low-frequency measurements give direct access to elastic properties within the seismic frequency band, which can be directly compared to seismic surveys. Additional µCT imaging can validate the actual saturations and give insight into dispersion mechanisms on the pore scale. 

To perform laboratory measurements at seismic frequencies and simultaneous capture processes inside the rock, we built a CT transparent triaxial stress apparatus. This allows for the validation and calibration of rock physics models for partially saturated media utilized for geological sequestration. The sample is saturated with water, and confining and pore pressure are increased. Then, CO2-saturated water is flown in the sample, which is stepwise depressurized so that part of the CO2 changes from liquid to gaseous CO2. Rock mechanical properties are determined at low frequencies (0.5 – 150 Hz) and strain amplitudes from 10-7 – 10-5, measuring Young's modulus, Poisson's ratio and attenuation at seismic frequencies and ultrasonic P- and S-wave velocities. In addition,  µCT-imaging of the pore structure and fluids is performed. The results show differences in elastic properties between a dry, partially and fully saturated sample and the effect of partial CO2-gas saturation at seismic and ultrasonic frequencies. P-wave velocities are reduced when free CO2 gas is in the system. The change is more significant at seismic frequencies, which leads to an increased dispersion.

The apparatus allows for 3D pore-scale imaging of CO2 in a fluid and gaseous phase while measuring frequency-dependent attenuation. Identifying fluid-solid interactions and estimating saturation via µCT-imaging is particularly critical when only a small amount of gas is present. Therefore, our combined approach allows precise determination of elastic properties at seismic frequencies for different CO2 saturations to establish a rock physics model calibrated to partial CO2 saturations to ensure safe CCS injection and monitoring.

Under repeated loading and unloading, the mechanical response of brittle rock exhibits pronounced hysteresis characteristic which is related to its fundamental damage behavior. Quantitative analysis based on 4D X-ray Computed Tomography (XCT), and Digital Volume Correlation (DVC) techniques were conducted on a fine-grained granite to get insights of damage process and strain filed evolution along loading/unloading and unloading/reloading hysteresis loops. Based on DVC analysis, the bulk axial, volumetric and von Mises strains were calculated. Then, the evolution of strain fields and strain population distributions were analyzed to investigate the evolving strain process and identified the relationship between bulk strain and local strain distribution. Finally, the mechanism of hysteresis loop behavior was investigated by analyzing the interaction between volumetric strain and voids evolution, and the strain concentration zones along the failure faults. The results showed that the bulk volumetric and lateral strains could better reflect the damage evolution compared to axial strain fine-grained granite. The dilation developed at the onset of loading and expanded with loading. During unloading, the dilation volume was increased slightly, and a non-uniform strain release process was observed: the strain release of contraction volumes was prior to the dilation volumes. Further damage was found in unloading/reloading hysteresis. Despite the tensile mechanism dominated the damage evolution pattern approaching to the failure, the shear-tensile mechanism was also involved in the process of damage evolution and final failure. Based on analysis, it is speculated that the development of damage unlocked the behavior of hysteresis: the difference in between sliding, and reverse-sliding of shear-tensile fractures was considered as the primary mechanism of the non-uniform strain release path. The dynamic interaction between opening of perpendicular voids and the reverse-sliding of shear fractures was considered as the damage mechanism during unloading.

In this study, we present experimental investigations on a low-porosity bioclastic calcarenite from the Cotiella Basin in the Spanish Pyrenees. Our objective is to elucidate the dominant failure mechanisms during the laboratory reactivation of natural deformation bands oriented at various angles to the maximum principal stress (σ1) direction. Triaxial compression experiments were conducted at the I12-JEEP beamline of the Diamond Light Source, UK, using a modified version of the Mjolnir cell employed by Cartwright-Taylor et al. (2022). Moreover, 4D (space and time-resolved) X-ray computed tomography images were acquired at 8 μm3 voxel size resolution during the triaxial compression tests (10 MPa to 30 MPa confining pressure).
Our mechanical data show that the presence of natural deformation features within the tested samples influences the material's strength. When comparing intact samples of the host rock under the same confining pressures, we observed that these samples exhibit higher peak stresses as opposed to those containing natural deformation features. Our research reveals that new deformation bands are formed as the angle (θ) between the deformation bands and σ1 increases. In this low-porosity carbonate, the reactivation of pre-existing deformation bands only occurs when their dipping angles are close to 70o.
To investigate the spatial and temporal relationships among naturally occurring and laboratory-induced deformation bands and fractures, we employed time-resolved X-ray CT and Digital Volume Correlation (Figure 1). Utilizing the SPAM open-source software (Stamati et al., 2020), we calculated the volumetric and shear strain fields. The orientation of the laboratory-induced failure planes is influenced by the orientation, width, and presence (or absence) of porosity along the length of the pre-existing natural bands. Additionally, the pre-existing secondary deformation features may contribute to additional mechanical damage, which can either facilitate the development or divert the newly formed failure planes.
In summary, our findings emphasize that the presence of natural deformation features weakens the material. We also observe that the reactivation of pre-existing bands occurs primarily at dipping angles near 70o in this low-porosity carbonate. The use of advanced imaging techniques and the SPAM software have allowed us to explore the relationships between the naturally occurring and laboratory-induced deformation features, highlighting the influence of orientation, width, and porosity on the orientation of failure planes. Finally, the presence of pre-existing deformation features triggers additional mechanical damage, affecting the development and direction of new failure planes.





Figure 1: Volumetric and shear strain fields calculated using the SPAM software during the loading of sample 25, which hosts a porous deformation band favourably oriented with respect to the σ1. Below the histogram which shows the intervals selected for Digital Volume Correlation (DVC), the images depict the volumetric and shear strain fields for these intervals, while at the bottom there are the shear and volumetric strain fields at failure, overlapping the equivalent x-ray CT vertical cross-section. The volumetric strain fields initially show lower strains within the deformation band, however, as the loading progresses compactant strains develop within the favourably oriented pre-existing deformation band. The shear strain fields show higher shear strain within the deformation band.