A new and openly accessible database of tests on piles driven in sands Available to Purchase

This paper reviews the critical need to develop better load test databases for piles driven in sand before reporting on the significantly augmented, openly accessible, Zhejiang University/Imperial College London (ZJU-ICL) database. Key quality parameters, the population of current entries and the reporting format are described before offering preliminary results obtained from comparisons between axial capacities calculated by various predictive approaches and site measurements. The results confirm that the offshore industry-standard ‘Main Text’ American Petroleum Institute RP2GEO procedures are less reliable and have larger coefficients of variation than alternative cone penetration test (CPT) methods, among which the ICP (Imperial College piles) and UWA (University of Western Australia) procedures appear to give the least bias and scatter. It is also shown that the ‘simplified’ ICP variant proposed by some practitioners is over-conservative and that its use could be discontinued. The new pile capacity and stiffness database offers broad scope for evaluating potential prediction biases relating to a wide range of soil and pile parameters. The submission of further high-quality tests for inclusion in regularly updated versions of the ZJU-ICL database is encouraged.