For management and mitigation of liquefaction hazards on a regional basis, it is generally desirable to evaluate liquefaction hazards in terms of annual probability of liquefaction (APL). In this study, an approach that combines the knowledge-based, clustered partitioning technique with the Hasofer-Lind reliability method is developed for estimating the annual liquefaction probability. Because it is generally difficult to validate the computed annual liquefaction probability, the results obtained from a modified version of the Davis and Berrill's energy dissipation model are used as a reference. The two models are examined with borehole data in the Yuanlin, Taiwan area that were investigated shortly after the 1999 Chi-Chi Earthquake. Results of the analysis reveal that annual probabilities of liquefaction estimated by the two models are consistent with each other and both deemed reasonable.
- Annual probability of liquefaction
- Energy dissipation model
- Knowledge-based clustered partitioning
- Reliability index
- Standard penetration test