Difficulties in assessment of liquefaction from SPT

Peter W. Mitchell

The difficulties in assessing liquefaction potential from the results of Standard Penetration Tests (SPTs) are illustrated by a case study. The method of Youd et al. (2001) is used to predict the potential for liquefaction. The important parameters for use in a ‘design earthquake’, namely peak ground acceleration and earthquake magnitude, have been derived for the site of a large industrial complex in Port Adelaide, South Australia. An assessment of the potential for liquefaction was required, as the site contained loose saturated sands and a loss of foundation support would have considerable implications for footing and retention wall designs. Questions in the assessment arose when the factor of safety against liquefaction determined from the results of SPT differed significantly from those determined from cone penetration tests (CPTs). Based on the SPT results, the site was predicted to be unstable against liquefaction for the design earthquake event, but the CPT assessment showed that the site had an adequate factor of safety. An investigation into the differences between results on this site and for a number of neighbouring sites revealed that the SPT N values were generally lower than values expected from published correlations with CPT profiles. A discussion of possible reasons for the disparities is presented within this paper and it is concluded that assessments based on SPT results alone are over-conservative for the ground conditions considered. Assessments using CPT results will provide a better assessment of liquefaction potential, as the CPT results are more reliable. The industrial site was therefore considered to have an adequate factor of safety against liquefaction for the design earthquake, resulting in considerable construction savings.