Investigating Impact of Local Hydrogeology and Tectonics on Physics-based Induced Earthquake Forecast Models
Abstract
Injecting large volume of brine, a byproduct of oil and gas production, into deep rock formations, is linked to the elevated seismicity in the US Midwest. Several models are proposed in the literature to forecast induced earthquakes knowing injection history. Zhai et al. 2019 propose a physics-based model that computes pore pressure and poroelastic stresses using a coupled poroelastic model and employ a seismicity rate model derived from the rate-state-friction law to predict induced seismic hazards in Oklahoma successfully. In this model, local hydrogeology such as rock permeability, porosity, and saturation level controls the process of fluid diffusion, which is spatially variable. Moreover, the fault orientations near the injection wells may affect the distribution, size, and timing of the induced earthquakes. Here, we perform series of thorough sensitivity tests using a sophisticated finite element model (FEM) to investigate how induced seismic hazards vary as a function of different rock permeability distributions, fault orientation, basement-reservoir boundary topography, and injection schedules. The FEMs would be built to mimic the realistic poroelastic environments of the Oklahoma and Kansas sites. We further present results from an optimization scheme that yields an injection schedule that minimizes the induced seismic hazard while maximizing injected volume. Such models enable a safer practice, benefiting both the general public and the energy industry.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.S23E0682T
- Keywords:
-
- 7209 Earthquake dynamics;
- SEISMOLOGY;
- 7223 Earthquake interaction;
- forecasting;
- and prediction;
- SEISMOLOGY;
- 7230 Seismicity and tectonics;
- SEISMOLOGY;
- 8168 Stresses: general;
- TECTONOPHYSICS