Seismogenic response to fluid injection operations in Oklahoma and California: Implications for crustal stresses

Thursday, 17 December 2015
Poster Hall (Moscone South)
Thomas Goebel, California Institute of Technology, Seismological Laboratory, Pasadena, CA, United States and Fred Aminzadeh, University of Southern California, Los Angeles, CA, United States
The seismogenic response to induced pressure changes provides insight into the proximity to failure of faults close to injection sites. Here, we examine possible seismicity rate changes in response to wastewater disposal and enhanced oil recovery operations in hydrocarbon basins in California and Oklahoma. We test whether a statistically significant rate increase exists within these areas and determine the corresponding timing and location based on nonparametric modeling of background seismicity rates. Annual injection volumes increased monotonically since ~2001 in California and ~1998 in Oklahoma. While OK experienced a recent surge in seismic activity which exceeded the 95% confidence limit of a stationary Poisson process in ~2010, seismicity in CA showed no increase in background rates between 1980 and 2014. A systematic analysis of frequency-magnitude-distributions (FMDs) of likely induced earthquakes in OK indicates that FMDs are depleted in large-magnitude events. Seismicity in CA hydrocarbon basins, on the other hand, shows Gutenberg-Richter type FMDs and b~1. Moreover, the earthquakes and injection operations occur preferably in distinct areas in CA whereas in OK earthquakes occur closer to injection wells than expected from a random uniform process. To test whether injection operations may be responsible for the strongly different seismicity characteristics in CA and OK, we compare overall well density, wellhead pressures, peak and cumulative rates as well as injection depths. We find that average injection rates, pressures and volumes are comparable between CA and OK and that injection occurs on average 0.5 km deeper in CA than in OK. Thus, the here tested operational parameters can not easily explain the vastly different seismogenic response to injection operations in CA and OK, and may only be of secondary importance for the resulting earthquake activity. The potential to induce earthquakes by fluid injection operations is likely controlled by the specific geologic setting and stress state on nearby faults.