Abstract:
Liquefaction and associated lateral spreading during the 1987 MW 6.5 Edgecumbe earthquake caused severe damage within parts of the Whakatane Township proximal to the Whakatane River and along the modern coastline in areas underlain by recent fluvial and marine sediments. The existence of an extensive CPT dataset has allowed CPT based liquefaction assessment methods to be applied across the township and to be compared with the liquefaction manifestations. To encompass the uncertainty in ground motion intensity and groundwater depth during the earthquake, three models for each were developed. Ground motion intensity models were developed through modification of ground motion prediction equations to fit strong motion recordings in the wider region. Groundwater models were created by adjusting groundwater recordings based on a relationship between the river level data (which were available for 1987) and seasonal fluctuations in piezometer recordings. Using the conditions defined by the aforementioned models, the CPT based liquefaction assessment performed well in areas with surface manifestation, with significant manifestation predicted. However, these assessment methods also suggested that much of the Whakatane Central Business District should have experienced significant liquefaction manifestation, even though nothing was observed in this area following the Edgecumbe earthquake. This study then extended beyond the standard CPT based methodologies to try and explain the reason behind over prediction in some areas of the township, focussing on both areas with and without surface manifestation. Through paleo-liquefaction trenching, the presence of subsurface liquefaction features was found to align with surface manifestation and also allowed characterisation of the subsurface sediment. Additional CPTs complemented by direct push crosshole shear and compression wave velocity (VsVp) testing were also performed to understand shear stiffness of the deposits and the depth to saturation. Partially saturated soils below the modelled water tables and the influence of interbedded soils on CPT data were found to be likely contributors to the over prediction, particularly in the southern CBD. Lab testing undertaken to identify an appropriate Ic-FC correlation generally aligned well with the recommended one, however further lab testing is required to confirm this. The effect of pumiceous soils, grain angularity and sand particle size on CPT based liquefaction assessment methods were identified as additional potential contributors to over prediction, however further research and lab analyses is required to confirm and quantify these. Overall, the over prediction of liquefaction manifestation in Whakatane is found to be due to a complex contribution of a number of factors which vary on a site to site basis.