Reliability of Five Axial Pile Capacity Design Methods â€“ A Case Study-Paper

Paper

Pdf Summary

The study evaluates the reliability of five methods for estimating the axial capacity of offshore piles: API and four CPT-based methods (Fugro, ICP, NGI and UWA). Using a case study of a three-legged North Sea jacket supported by 39 m piles in alternating sand and clay layers, the authors combine statistical models of soil properties and loads with method uncertainty derived from a database of pile load tests. First-order reliability method (FORM) analyses estimate annual failure probabilities and calibrate resistance factors for target probabilities of failure of 10⁻³, 10⁻⁴ and 10⁻⁵.<br /><br />Calculated reliability varied substantially among methods and between the two approaches used to represent method uncertainty. API produced the lowest estimated pile capacities and the highest failure probabilities, reflecting its greater method uncertainty. Overall, the CPT-based methods had lower uncertainty than API and were at least as reliable in this case.<br /><br />For the CPT-based methods, calibrated resistance factors using a 100-year representative environmental load ranged from 1.09–1.74 for a 10⁻³ target failure probability, 1.28–2.14 for 10⁻⁴, and 1.47–2.55 for 10⁻⁵. Using a 50-year load required factors about 7–8% higher to achieve the same reliability.<br /><br />The authors conclude that CPT-based methods may support more economical pile designs, but emphasize that the results apply only to this case study. Broader calibration requires multiple case studies and consideration of factors such as load redistribution within pile groups.

Keywords

offshore piles

axial capacity

reliability analysis

FORM

CPT-based methods

API method

resistance factors

pile load tests

North Sea jacket

failure probability