Fines-content effects on liquefaction hazard evaluation for infrastructure in Christchurch, New Zealand

Abstract

To assess soil liquefaction hazards for civil infrastructure, several competing liquefaction evaluation procedures (LEPs) are used to estimate the potential for liquefaction triggering, often for use in a liquefaction potential index (LPI) framework. However, due to the relatively uncertain effects of fines-content (FC) on liquefaction behavior, LPI hazard assessments may be less accurate at sites with high FC. Accordingly, this study investigates “fines-content effects” on the accuracy of LPI hazard assessment during the 2010–2011 Canterbury Earthquake Sequence (CES). These effects are resolved into: (1) criteria based on the soil-behavior-type index (Ic) for identifying liquefaction-susceptible soils; (2) FC-corrections inherent to each LEP, used to adjust liquefaction resistance for the presence of fines; and (3) the potential for non-liquefied, high-FC soils to inhibit liquefaction manifestation. This investigation is performed using 7000 liquefaction case studies from the CES, wherein LPI hazard assessments computed with the Robertson and Wride [50], Moss et al. [41], and Idriss and Boulanger [30] LEPs are compared to field observations. For the assessed dataset, LPI hazard assessments were significantly and uniformly less accurate at sites with silty and clayey soil mixtures. For these sites, the existing LPI framework has inherent limitations, such that all LEPs produce erroneous hazard assessments. In particular, the capacity of plastic soils to inhibit liquefaction manifestation by affecting pore pressure development and redistribution should be further evaluated.