SITE-SPECIFIC SEISMIC HAZARD ANALYSIS OF DEL VALLE DAM AND THE IMPACT OF LOCAL FAULT SOURCES

We performed a site-specific probabilistic seismic hazard analysis (PSHA) of Del Valle Dam in Alameda County, California. This facility, which is owned by the California Department of Water Resources (DWR), provides storage for the South Bay Aqueduct, a branch of the California State Water Project. The Del Valle Dam region includes numerous active faults that accommodate movement of the San Andreas fault system and growth of the Diablo Range. The seismic source model used in the PSHA included representations of known active faults and background earthquakes within 100 km of the damsite and was based on prior seismic hazard analyses performed for critical infrastructure in the region as well as recent “consensus” seismic hazard studies performed state-wide. The seismic sources that contribute most to hazard at PGA and 0.2 sec spectral acceleration (SA) are the well-studied Calaveras-Hayward fault system (at shorter return periods) and the lesser-known Verona-Williams fault zone (at longer return periods). Other top contributors at PGA and high frequencies are the widely recognized Greenville, Las Positas, and Mt. Diablo thrust fault sources. Important for the site-specific PSHA is the inclusion of lesser-known fault sources in the direct vicinity of the damsite. The Verona-Williams fault zone, which is not in statewide and national source models (including ERF-2023), is an east-dipping reverse fault that underlies Del Valle Dam and the southern portion of the densely populated Livermore Valley. DWR-funded studies interpret the fault to branch from the Greenville fault at or near the base of the seismogenic crust, and its slip rate is estimated to be on the order of 0.3 mm/yr. Comparisons of our site-specific PSHA results with the 2018 USGS National Seismic Hazard Maps showed that our site-specific hazard is 19 to 31% and 9 to 20% higher than the USGS results for PGA and 1.0 sec SA, respectively. A hazard sensitivity demonstrated that the differences are primarily due to the source characterization, and the clearest difference is the presence in our model of the Verona-Williams fault source. From this and other projects, we find that the inclusion of nearby fault sources, even in cases of low slip rate faults surrounded by higher slip rate faults, is important for seismic hazard analysis of critical facilities.