Fault Rupture Assessments for High-Pressure Pipelines in the Southern San Francisco Bay Area, California

The San Andreas, Hayward, and Calaveras faults are major active faults that traverse the San Francisco Bay area in northern California, and may produce surface rupture during large earthquakes. We assessed the entire Pacific Gas & Electric Company natural gas transmission system in northern California, and identified several locations where primary pipelines cross these faults. The goal of this effort was to develop reasonable measures for mitigating fault-rupture hazards during the occurrence of various earthquake scenarios. Because fault creep (e.g., slow, progressive movement in the absence of large earthquakes) occurs at the pipeline fault crossings, we developed an innovative approach that accounts for the reduction in expected surface displacement, as a result of fault creep, during a large earthquake. In addition, we used recently developed data on the distribution of displacement across fault zones to provide likely scenarios of the seismic demand on each pipeline. Our overall approach involves (1) identifying primary, high-hazard fault crossings throughout the pipeline system, (2) delineating the location, width, and orientation of the active fault zone at specific fault-crossing sites, (3) characterizing the likely amount, direction, and distribution of expected surface fault displacement at these sites, (4) evaluating geotechnical soil conditions at the fault crossings, (5) modeling pipeline response, and (6) developing mitigation measures. At specific fault crossings, we documented fault locations, widths, and orientations on the basis of detailed field mapping and exploratory trenching. We estimated fault displacements based on expected earthquake magnitude, and then adjusted these values to account for the effects of fault creep at the ground surface. Fault creep decreases the amount of expected surface fault rupture, such that sites having high creep rates are expected to experience proportionally less surface displacement during a large earthquake. Lastly, we modeled the expected amount of surface offset to reflect the distribution of offset across the fault zone, based on data from historical surface ruptures throughout the world. Where specific fault crossings contain a single primary fault strand, we estimated that 85% of the total surface offset occurs on the main fault and the remainder occurs as secondary deformation. At sites where the pipeline crosses multiple active fault strands in a broad zone, we consider complex rupture distributions. Using this approach yields realistic, appropriately conservative estimates of surface displacement for assessing seismic demands on the pipelines.Copyright © 2004 by ASME