For subsea pipeline UHB design, it is essential to establish appropriate UHB load factors by undertaking a structural reliability analysis based on a target acceptable probability of failure. The current practice is to setup the acceptable probability of failure according to the best industrial practice coded in DNV OS F101 and ISO16708 but regardless of trench performance, imperfection details and inspection results.
This paper is intended to discuss the relationship between the UHB failure probability for the whole pipeline and the trench performance and imperfection frequency statistics. A detailed statistic and reliability analysis is undertaken to address some disparity in the current practice. It is shown that in order to achieve a specific reliability level required for the entire pipeline, it is paramount to calibrate the target probability of failure based on the survey data. The common practice for UHB is shown un-conservative giving rise to an insufficient safety level and non-compliant OOS design. An improved approach is outlined to address the critical issues allowing for a robust OOS assessment by means of case studies.