资源环境科技发展态势分析平台

Induced seismicity models that do not account for the effects of flow boundary conditions and uncertainty in initial stresses on the seismicity rate fail to predict the seismicity observed in real-world injection sites. We present a model for the probabilistic assessment of induced seismicity that incorporates in situ stresses, pressure, and reservoir boundary conditions. The model considers a randomly distributed network of faults undergoing frictional failure due to pressure-induced effective stresses. A semianalytical method is used to capture the effect of flow boundary conditions on pressure and seismicity rate based on pressure diffusion and Mohr-Coulomb failure. We compare the model predictions with induced seismicity observations in Paradox Valley and German Continental Deep Drilling Program. We demonstrate that differences in the flow boundary conditions can drastically change the seismic response to injection. The strongest difference is observed for a closed boundary reservoir for which the induced seismicity rate increases exponentially.

Plain Language Summary Earthquakes triggered at Rocky Mountain Arsenal and Rangely, Colorado, are classic examples of fluid-induced seismicity. The recent increase in seismicity rate from the background rate in the Central United States starting in 2009 brought much attention to induced seismicity associated with wastewater disposal. The relationship between injection rate or injected volume and the rate of induced seismicity is observed to be complex with significant deviation from a linear relationship. We explore plausible explanations for the nonlinearity observed in the relationship between the cumulative number of seismic events and injected volume. We develop a model for probabilistic assessment of induced earthquakes that incorporates geomechanical analysis of the state of stress, pressure diffusion, and different reservoir boundary conditions. We validate our model using the data from the German Continental Deep Drilling Program 2004-2005 injection experiment and Paradox Valley 1996-2004 saltwater disposal. We propose that the acceleration in seismicity rate at the German Continental Deep Drilling Program can partially be explained by the effects of reservoir boundaries in addition to changes in injection rates. We conclude that a physics-based, probabilistic framework of induced seismicity can successfully capture the site-specific relationships among injection parameters, reservoir parameters, and expected seismicity rates.