In this paper the accuracy of DNV-OS-F101 Appendix A has been investigated and discussed. The crack driving force (CDF) for 4 different flaw sizes in a OD 277 × WT 15.5 mm pipe subjected to bending into the plastic regime have been investigated by 3D FE analyses. The same flaw sizes have been applied in the Appendix A of DNV-OSF101 (2007) procedure and the crack driving forces have been compared with the results of the FE analyses. The analyses do not include internal pressure and are, hence, most relevant for pipeline installation methods introducing large plastic strains as for instance the reeling method. The actual accuracy of the ECA procedure specified in DNV-OS-F101 (2007) has not been quantified because this will require that all the physics are calculated correctly and that reliability analyses are performed. However, certain "worst case assumptions" and the combination of several "worst case assumptions" and which consequences this may have on the fracture reliability is discussed. Finally, a new simplified procedure for performing more accurate prediction of the crack driving force based on DNV-OS-F101 Appendix A is presented.
Requirement for performing fracture mechanics assessments (engineering critical assessment, ECA) in order to establish maximum allowable flaw sizes and AUT weld defects acceptance criteria has been stated in DNV-OS-F101 since the 2000 rules. The procedure specified in the 2000 rules was not very specific and the results were dependent on how the designer defined the input parameters. However, in the 2007 rules a new appendix, Appendix A, "Structural Integrity of Pipeline Girth Welds in Offshore Pipelines" was introduced and both the methodology and the requirements to how to determine and chose the input parameters were specified in detail. Hence, ECA's performed in accordance with DNV-OS-F101 should be more uniform than ECA's performed in accordance with DNV-OS-F101.
