Liquefaction Hazard Assessment of Pumice Rich Sands – Case Study of Rangipo Substation
S. Anwar, G. Blakeley & Z. Rana
AECOM, Auckland, New Zealand.
ABSTRACT
For any proposed development site, it is standard geotechnical practice to assess the foundation soils for their strength and load-bearing capacity. In seismically active regions, it is also customary to evaluate the liquefaction hazard of soils situated below the groundwater table. When assessing soils that are not well represented in empirical liquefaction databases such as those with moderate plasticity, gravel, sand–silt mixtures, or soils containing significant carbonate (shell) or pumice content, a different approach is required to accurately determine their cyclic strength or resistance. In simplified penetration data-based methods, the soil strength is typically estimated indirectly from in-situ penetration tests such as the Standard Penetration Test (SPT), Cone Penetration Test (CPT), or Dilatometer Test (DMT). Alternatively, it is often recommended to assess soil strength directly using shear wave velocity (Vs) obtained from surface geophysical testing, particularly for materials such as clean gravels or soils where penetration tests cannot be performed effectively or tend to underestimate strength, as in the case of carbonate or pumice-rich sands. Pumice-rich soils are widely distributed throughout the central North Island of New Zealand, particularly within the Taupō Volcanic Zone (TVZ). The Rangipo Substation (RPO) site, located within the TVZ, is underlain by a thick, saturated deposit of pumice-rich sands. When the RPO site was initially assessed for liquefaction hazard using simplified penetration-based methods, the analyses predicted significant liquefaction potential. However, since these methods were originally developed for hard-grained sands rather than crushable pumiceous sands, their applicability to the RPO soils was uncertain. To verify the validity of the simplified approach, laboratory cyclic triaxial testing was undertaken on representative samples from the site. The laboratory test results demonstrated that the cyclic resistance ratio (CRR) calculated using simplified penetration-based methods was substantially underestimated for the crushable pumice-rich sands.