Enhancing the Passive Monitoring of the Rock Damage Process

Abstract

Abstract This study explores the effectiveness of double-difference event location methods in detecting early signs of cracking, leading to leaks from structurally trapped CO2. Leaks carry the potential to escalate into bursts, highlighting the importance of proactive pore-pressure management measures. While the double-difference method is frequently applied for precise identification of fluid-induced seismic events, its reliance on a constant velocity model poses limitations, especially in dynamic environments potentially undergoing fracturing. Consequently, integrating velocity variations within a fractured medium into time-lapse seismic localization remains a significant challenge. In this study, we investigate how point-in-time average velocity changes influence the accuracy of double-difference event location. We analyze acoustic emissions (AEs) data gathered from laboratory experiments to explore this relationship. We performed a triaxial mechanical test on Berea sandstone, monitoring stress, strain, acoustic emissions (AEs), and variations in acoustic velocity as damage accumulated over time. AE event locations were determined using P-wave arrival times and magnitudes derived from source-receiver distances. The test revealed a significant surge in AE activity during the stages of damage progression, characterized by crack propagation ultimately leading to failure. Conventional double-difference technique, which assumes constant velocity, led to clustered distributions of acoustic emissions (AEs) that inaccurately represented fracture geometry. However, integrating point-in-time velocity changes improved the alignment of AE distributions with fracture geometry, unveiling a clear time-lapse imaging of crack initiation, propagation, and coalescence along the primary stress direction. This study highlights the crucial role of temporal velocity variations in ensuring precise event mapping, particularly in the monitoring of CO2 leaks. It proposes that average velocity fluctuations be monitored during injection or estimated from rock physics models for continuous and cost-effective monitoring. Ultimately, integrating time-lapse velocity changes improves our capacity to monitor crack progression and mitigate risks associated with CO2 leakage.