SCEC Award Number 11191 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title What is the effect of induced seismicity on tectonic faults?
Investigator(s)
Name Organization
Emily E. Brodsky University of California, Santa Cruz
Other Participants Lia Lajoie
SCEC Priorities A4, A6, A9 SCEC Groups SHRA, EFP, FARM
Report Due Date 02/29/2012 Date Report Submitted N/A
Project Abstract
 Geothermal power generation commonly induces seismicity. These induced earthquakes have aftershocks, which can potentially occur on other faults in the region. Therefore, it is possible that induced seismicity can trigger tectonic events. The goal of this project was to begin to assess this possibility by quantifying the aftershocks on the neighboring faults from earthquakes in a major geothermal field in Southern California. SCEC3 funded initial work on this study in 2011. During this year, beginning graduate student Lia Lajoie analyzed publicly available seismicity and production data. She constructed her own ETAS inversion code, benchmarked against previous work in the area, and collected geothermal data, including through an in-person visit at the Salton Sea Geothermal Plant. The major accomplishments thus far are: (1) identification of earthquake interactions within the Salton Sea Geothermal field by modeling seismicity in a 10 km radius with an Epidemic-Type Aftershock Sequence (ETAS) model, and (2) preliminary testing of a modified ETAS model with background rate proportional to the fluid injection. This initial success of an ETAS inversion is important in itself. The result shows that aftershock behavior in the field is not readily distinguishable from other regions. Therefore, our initial assessment is that secondary earthquake triggering by induced seismicity is a realistic possibility.
Intellectual Merit Establishing the circumstances under which fault interact provides key information on the types and magnitudes of stresses that initiate earthquakes. This information is some one of the observational conerstones to evaluating earthquake predictability based on first-principle approaches.
Broader Impacts Evaluation of the hazard from induced seismicity has historically been limited to the direct effects of pumping and therefore presumes that the induced earthquakes are bounded by the size of the reservoir. However in natural seismicity, earthquakes trigger other earthquakes through a combination of static and dynamic stresses. Triggered seismicity that extends beyond the region of pumping is not necessarily limited by the size of the faults or fractures adjacent to the wells. Therefore, the resulting earthquakes could potentially be much larger than the previously proposed maximum sizes of induced earthquakes. This societally significant consequence of geothermal power exploitation has not been systematically explored and may be particularly important for a geothermal resource abutting a major fault system

In the Salton Sea, we can evaluate these possibilities by examining the relationship between the induced earthquakes and the adjacent San Andreas system. Plants operate within a few kilometers of the Brawley seismic zone, between the Southern San Andreas and the Imperial Fault. The edge of the Salton Sea geothermal field is 22 kilometers from Bombay Beach, which has been modeled as a possible initiation point of a large San Andreas event.
Exemplary Figure Figure 1. Salton Sea Geothermal Field cumulative seismicity and injection and production volumes. (a) Gross production and injection volumes based on monthly, field-wide aggregates. (b) Net volume (production – injection). The sharp step in seismicity in Aug. 2005 is the previously identified creep transient (Lohman and McGuire, 2007). The net volume change peaks in July 2005. Earthquake locations are from the ANSS catalog 1984-Oct. 2010 with M>2.5 within 15 km radius of the Del Ranch plant in the Salton Sea Geothermal Field. Volume equivalent of water at atmospheric pressure and temperature is calculated based on reported mass and an assumed density of 1000 kg/m3.