A Practical Summation Method of Tolerances for Misalignment and SCF Calculation for Girth Welds

Abstract In conventional fatigue analysis of girth welds of risers and pipelines, misalignment with corresponding estimates of stress concentration factors need consideration. This is most important for cracks starting from the weld toe. The three most important sources for the misalignment are: i) diameter tolerances, ii) out-of-roundness (or ovality) tolerances and iii) thickness tolerances. In fatigue analysis of subsea wellhead systems, the most fatigue utilized girth welds are usually between machined components such as the wellhead or conductor housing and extension pipes or between two pipes. In this context, we have seen different industry practices on combining misalignment from the different sources of tolerances, leading to different stress concentration factors and correspondingly different fatigue lives. Both simple and somewhat complex combination methods can be rather conservative, resulting in high stress concentration factors and consequently, unacceptable low fatigue lives. There appears to be a lack of detailed guidance for combining these tolerances, and different engineers make different evaluations. We see the need for a common methodology, and in this paper, we present a practical approach, based on geometrical considerations and simplified statistical summation methods, to assess the resulting misalignment from tolerances. The applicable stress concentration factor is then calculated as a function of the misalignment using a standard formula. The method is applicable for specific tolerances provided in the relevant fabrication specification or standard. It can also be used in combination with measured so-called high/low values from the outside or inside of the pipe. Application of the proposed methodology is demonstrated for fabrication tolerances provided in API 5L, and example results are calculated and presented as a function of pipe diameter for different pipe wall thicknesses. Results are computed for girth welds connecting a pipe to a machined component and for pipe-to-pipe connections. Different tolerances for seamless and welded pipes as given in API 5L are considered. Intermediate results are provided, such that it is transparent to see the contribution to the total misalignment from the different tolerance sources. It is likely that the method leads to somewhat lower stress concentration factors than what are most commonly used today by the industry for wellhead fatigue analyses. Although some conservatism is removed, the method is still judged to be on the conservative side which is also supported with reference to a large database of actual measured misalignments generally being less than 1 mm (OMAE2018-78408), indicating that the actual misalignment is generally well within typical fabrication requirements.