Rocky Mountain Section - 59th Annual Meeting (7–9 May 2007)

Paper No. 4
Presentation Time: 9:10 AM

PROBABILISTIC SEISMIC HAZARD STUDIES IN THE ROCKY MOUNTAIN WEST


ANDERSON, Larry W., Seismotectonics and Geophysics Group, Bureau of Reclamation, Denver Federal Center, P.O. Box 25007, Denver, CO 80225, landerson@usbr.gov

The Bureau of Reclamation, with responsibility for over 300 dams in the western United States, has been conducting detailed site-specific seismic hazard studies for over 30 years. While early studies were deterministic in scope and methodology, studies over the last 10 years have been essentially entirely probabilistic. This change in approach has been brought about because Reclamation has embraced a risk-based approach to dam safety evaluations, as well as changes in the state-of-the-practice and the general acceptance of probabilistic seismic hazard analyses (PSHA) within the engineering community.

Reclamation has recently completed PSHAs for several facilities in the Rocky Mountain west, including Canyon Ferry Dam in west-central Montana. Like many Reclamation facilities, initial studies in the 1980's and 1990's identified the primary hazard as coming from “random” or “background” earthquakes (M 6 to 6 ½ events) because nearby faults either were not considered “active” under deterministic criteria or were believed to have low slip rates (< 0.1 mm/yr). The study for Canyon Ferry Dam points out the benefits of site-specific PSHAs and the need for site-specific data.

Initial scoping-level studies concluded that the nearby Canyon Ferry fault probably had no evidence for latest Pleistocene activity; hence, the fault was not considered a significant earthquake source. Later field studies identified fault scarps on what appeared to be latest Pleistocene deposits. A paleoseismic trench was excavated near the central portion of the fault that provided important information on the slip rate, recurrence, and slip per event for the fault. These data indicate a total dip-slip displacement of approximately 9 m occurred over a 55 kyr period between about 68 ka and 13 ka, and a long term late Quaternary slip rate of 0.16 mm/yr (0.13-0.2 mm/yr). More importantly, there are strong indications of clustering of seismic events which indicate short term rates of 0.54 mm/yr (0.35-0.91 mm/yr). Without detailed study, the seismic hazard at Canyon Ferry Dam would have been dominated by the random earthquake, ground motions would have fairly low, and the risk underestimated. The detailed, site specific data indicate that at return periods of a few thousand years or more, the hazard is controlled by the Canyon Ferry Fault.