← Back to Sessions

National Liquefaction Loss Database and Fragility Functions

Session: Seismic Hazard Analysis for Critical Infrastructure [Poster]

Type: Poster

Date: 4/19/2021

Presentation Time: 03:45 PM Pacific

Description: 

Liquefaction is a secondary hazard that occurs during earthquakes and can cause severe damage to overlying infrastructure. As a result, liquefaction can be a significant contributor to loss in earthquakes as observed during the 1964 Alaska event. A geospatial liquefaction model developed by Zhu et al. 2017 and implemented by the USGS on the ground failure tab of the earthquake overview page is used to estimate liquefaction spatial extent (LSE) after an earthquake. The model estimates LSE using two shaking parameters and four globally available environmental parameters.

The total areal extent over which soil is expected to liquefy in an earthquake is calculated for each event (TLSE) and evaluated against observed liquefaction (Rashidian and Baise, 2020). The USGS Pager system utilizes a slightly different algorithm to calculate a variation of TLSE referred to as “aggregate liquefaction hazard”, abbreviated as Htot. The USGS Pager system also calculates “aggregate liquefaction population exposure”, abbreviated as PopExp. However, neither the geospatial liquefaction model nor the USGS Pager system currently predicts infrastructure or economic loss due to liquefaction.

We present a liquefaction loss database of 12 United States events. Each entry in the database is labeled by an infrastructure category: transportation, utilities, and buildings. When possible, liquefaction loss is associated with a spatial location. This database is used to relate economic loss by infrastructure category to Htot and PopExp. We also present fragility functions for each infrastructure category using the national loss database and damage state thresholds defined by damage costs.

Additionally, we build on work by Bird and Bommer (2004), a global dataset of liquefaction damage states for 50 events, by expanding the dataset to 86 events and developing fragility functions for these damage states. These fragility functions provide probabilities of liquefaction causing minor/moderate damage or major damage relative to the overall event’s damage based on one of two excitation measures: Htot or PopExp.

Presenting Author: Alexander A. Chansky

Student Presenter: Yes

Authors