Assessment of Existing SPT–CPT Correlations Using a New Zealand Database

Abstract

Using a database of approximately 900 data pairs from 230 sites of co-located cone penetration test (CPT) soundings and boreholes with standard penetration tests (SPT) in the South Island of New Zealand, this paper assesses the applicability of a range of existing SPT–CPT correlations. Simple linear SPT–CPT correlations for different soil types as well as correlations based on the soil behavior type index (Ic) from CPT data were assessed. The measured SPT N60 data were poorly fit by all the correlations, with under- and over-predictions greater than 50% for between 40% and 55% of the database. The 95% confidence interval of the bias between predicted and measured SPT N60 values was large across all variables. The bias varied significantly as a function of CPT tip resistance and sleeve friction, with underestimates at low values dominated by fine-grained soils (Ic>2.05), and overestimates at high values dominated by silty sand to dense sand (Ic≤2.05). There was no clear effect of the different drilling methods on the bias observed across all parameters. These results highlight the significant uncertainty involved in the application of these correlations in geotechnical design and assessments.