Seal Leakage Analysis: CO 2 Diffusion, Seismic Response, and Mineralogical Changes in Sandstone-Shale Systems

The integrity of low-permeability seals is critical for long-term CO 2 storage, yet most assessments emphasize advective leakage along faults, fractures, or wells and largely neglect molecular diffusion into the seal. Here, we combine core-scale diffusion experiments, ultrasonic monitoring, and post-experiment imaging to quantify CO 2 diffusion in a Berea sandstone–Niobrara shale system and evaluate its implications for storage security and seal performance. Brine-saturated cylindrical cores were exposed to CO 2 at 22°C and 7 MPa in a manometric sorption setup, where pressure decay and ultrasonic waveforms were recorded continuously. Pressure–time data yield effective diffusivities in the sandstone of D ^ (1.8 – 5.5) × 10 –11 m 2 /s and intrinsic (molecular) diffusivities D (0.70 – 2.1) × 10 –9 m 2 /s, whereas stacked-test shale values of D̂ (1.8 – 5.5) × 10 –11 m 2 /s and D (3.6 – 11) × 10 –9 m 2 /s represent upper bounds on purely diffusive transport in the seal. Waveform analysis shows relative velocity changes (Δ v / v) of 0.1 – 1% along with a 75% decrease in waveform energy, indicating that attenuation is a more sensitive indicator of CO 2 sorption than velocity alone. SEM–EDS imaging shows calcite dissolution, NaCl precipitation, and mineral transfer at the sandstone–shale interface. These results demonstrate that diffusion-driven CO 2 sorption, coupled with geochemical reactions, can measurably modify seal properties over time and should be incorporated into reservoir-scale conformance and risk assessments. By quantifying CO 2 diffusion into the caprock and nearby formations, this study strengthens predictive modeling for long-term storage and increases the reliability of CO 2 storage security assessments. This integrated laboratory study provides, to our knowledge, one of the first combined diffusion–seismic–mineralogical characterizations of a sandstone–shale reservoir–seal pair under CO 2 exposure and yields quantitative parameters for long-term storage modeling and monitoring design.