T52B-04
Induced earthquake magnitudes are as large as (statistically) expected

Friday, 18 December 2015: 11:05
302 (Moscone South)
Nicholas van der Elst, USGS Earthquake Science Center, Pasadena, CA, United States, Morgan T Page, USGS Pasadena Field Office, Pasadena, CA, United States, Deborah A Weiser, University of California Los Angeles, Los Angeles, CA, United States, Thomas Goebel, California Institute of Technology, Seismological Laboratory, Pasadena, CA, United States and S. Mehran Hosseini, University of Southern California, Los Angeles, CA, United States
Abstract:
Key questions with implications for seismic hazard and industry practice are how large injection-induced earthquakes can be, and whether their maximum size is smaller than for similarly located tectonic earthquakes. Deterministic limits on induced earthquake magnitudes have been proposed based on the size of the reservoir or the volume of fluid injected. McGarr (JGR 2014) showed that for earthquakes confined to the reservoir and triggered by pore-pressure increase, the maximum moment should be limited to the product of the shear modulus G and total injected volume ΔV. However, if induced earthquakes occur on tectonic faults oriented favorably with respect to the tectonic stress field, then they may be limited only by the regional tectonics and connectivity of the fault network, with an absolute maximum magnitude that is notoriously difficult to constrain.

A common approach for tectonic earthquakes is to use the magnitude-frequency distribution of smaller earthquakes to forecast the largest earthquake expected in some time period. In this study, we show that the largest magnitudes observed at fluid injection sites are consistent with the sampling statistics of the Gutenberg-Richter (GR) distribution for tectonic earthquakes, with no assumption of an intrinsic upper bound. The GR law implies that the largest observed earthquake in a sample should scale with the log of the total number induced. We find that the maximum magnitudes at most sites are consistent with this scaling, and that maximum magnitude increases with log ΔV. We find little in the size distribution to distinguish induced from tectonic earthquakes. That being said, the probabilistic estimate exceeds the deterministic GΔV cap only for expected magnitudes larger than ~M6, making a definitive test of the models unlikely in the near future. In the meantime, however, it may be prudent to treat the hazard from induced earthquakes with the same probabilistic machinery used for tectonic earthquakes.

<< Previous Abstract | Next Abstract >>