Towards improved code-based performance objectives for liquefaction hazard analysis

Abstract

Ground failure due to liquefaction in loose sand deposits poses substantial risks to the built environment, and has caused significant damage in past earthquakes to a wide range of infrastructure. Advances in liquefaction hazard analysis in practice have largely stagnated in recent years; the state of practice remains rooted in simplified procedures that ignore considerable uncertainties in liquefaction phenomena, and are largely conditional on single-return period ground motions. As a result, they lack any sort of liquefaction-specific design criteria or performance objective. Presented herein is a roadmap for using probabilistic liquefaction hazard analysis (PLHA) to address many of these limitations and improve liquefaction design guidelines. PLHA incorporates hazard contributions from the full ground motion hazard space in conjunction with probabilistic liquefaction models, to produce hazard curves for various types of liquefaction-related demands. In this study, PLHA is utilized to assess the current, implied liquefaction design levels at 76 study sites throughout the U.S. using ASCE 7 guidelines, by computing effective return periods of liquefaction factor of safety FSL, and liquefaction potential index LPI. The results indicate broad inconsistencies in these design levels across different parts of the U.S, with return periods varying from about 300 years in deterministically-capped parts of California, to nearly 3,000 years on the Pacific Northwest coast and in the Charleston Fault zone region. These results are also used to inform potential strategies for establishing consistent, liquefaction-specific design objectives in the future, based on return period averaging methods that weight the importance of a study site according to both the population and relative liquefaction hazard level.