Liquefaction induced ground damage in the Canterbury earthquakes: predictions vs. reality

Predicting the severity of liquefaction induced ground damage is the key outcome from a liquefaction assessment. Ishihara (1985) developed a simple and logical method to predict the occurrence of sand boils, ground fissures and other features of liquefaction damage. This paper examines the effectiveness of Ishihara’s method in predicting the liquefaction damage that occurred in the recent Canterbury earthquakes.

The applicability of the method to complex multi-layered soil profiles encountered in Canterbury is addressed, and a procedure for calculating an equivalent liquefied layer thickness in such profiles is proposed. This procedure is applied to a large number of sites with CPT data in Christchurch, combining simplified liquefaction analyses with detailed observations of ground damage following three large earthquakes.

Ishihara’s method was generally successful in predicting the occurrence of liquefaction damage, and is considered appropriate for simplified assessments of liquefaction hazard in layered Canterbury soils. However, the Ishihara method may unconservatively predict no damage for sites where thin liquefiable layers are present within the upper 3m of the soil profile; observations from the Canterbury earthquakes indicate that significant damage may occur in these cases. The data from Canterbury suggests that once liquefaction is triggered the occurrence of liquefaction damage is primarily dependent on the proximity of the liquefied layer to the ground surface, not the ground shaking intensity. Finally, it was observed that a crust thickness greater than 3.5-4m was sufficient to prevent liquefaction induced damage, even with significant thickness of liquefied soil below.