Event-based probabilistic liquefaction hazard analysis for defining soil acceptance criteria

Abstract

In professional practice liquefaction susceptibility has been assessed through different methods, frequently aimed at establishing a safety factor against liquefaction, or its occurrence probability, for an earthquake with known magnitude and location, along with a well-characterized soil profile. These conventional analyses allow estimating the liquefaction probability for only one event, such as the Maximum Credible Earthquake (MCE), or the life safety limit state associated with a fixed return period. Therefore, it is not possible to estimate the occurrence frequency of liquefaction because there is only a vague link between the design earthquake and its occurrence frequency and because, in most cases, there are many earthquakes, besides the design earthquake, which could contribute to liquefaction with a non-negligible probability. This paper shows an update of the Probabilistic Liquefaction Hazard Analysis (PLHA) that has been used in the past, reframing it into a contemporary context, mainly taken from Probabilistic Seismic Hazard Analysis (PSHA), indicating how a rigorous PLHA can be performed to determine how frequently liquefaction will occur and how that procedure can be easily integrated into a modern PSHA code, namely R-CRISIS. The probability of liquefaction needs the estimation of earthquake acceleration levels at the surface of the soil profile and, therefore, this paper presents a comprehensive approach to consider the nonlinear behavior of the soil in the estimation of amplification factors. Besides this, we propose a soil-characteristics acceptance criterion, coherent with those typically used in other earthquake engineering areas. Finally, the paper proposes some approximations that, under certain conditions, allow having reasonable estimations of the annual frequencies of liquefaction occurrence in a soil profile, without performing a complete PLHA.