Project Description
The locations of microseismic events are used to assess the evolving underground state of stress during CO_{2} storage operations. Uncertainty in subsurface structure contributes significantly to error in estimated microseismic event locations. This project will adapt Lawrence Livermore National Laboratoryâ€™s Bayes location (Bayesloc) algorithm to more accurately predict the microseismic event location needed to inform studies about rock strain and fracture that may occur during CO_{2} injection. At the core of the Bayesloc is a statistical model that links observed data to unobserved parameters. The statistical model consists of three main components: (1) a prior probability model for the source parameters in 3D location and time, (2) a statistical model for the correction to the underground model (i.e., the travel-time corrections), and (3) a statistical model for the error in the observed data (e.g., the spread of the arrival-time residuals). In the field, seismic sensors, indicated by the blue circles in the figure, are used to measure microseismic events and infer location. The microBayesloc algorithm produces an estimate of location uncertainty, an essential feature to help distinguish real from phantom faults. MicroBayesloc determined microseismic event location uncertainty estimates are indicated by the red ellipsoid clouds.
Project Benefits
This project will develop a fundamentally better approach to geological site characterization and early hazard detection. It combines innovative techniques for analyzing microseismic data with a physics-based inversion model to forecast microseismic event locations. This project will devise fast-running methodologies that will allow field operators to respond quickly to changing subsurface conditions.
Contact Information
Click to view Presentations, Papers, and Publications