ECA of Girth Welds for Deepwater HPHT Flowline with Lateral Buckle Mitigation

Abstract

Abstract An Engineering Critical Assessment (ECA) for reeled subsea high-pressure and high-temperature (HPHT) flowlines includes fatigue and fracture assessments of the girth welds for the reeling, installation, and in-service phases. The in-service ECA phase requires input from an in-place assessment of the flowline under different loads. The fracture loads are often provided in terms of longitudinal peak strain to minimize inaccuracies due to different tensile properties used in the in-place and fracture assessments. Typically, a stress-strain curve with higher yield strength and ultimate tensile strength is used in the ECA (even under load control), so using strain to transfer loads from an in-place analysis model to ECA guarantees conservatism. However, applying an incorrect strain definition, such as the total strain (sum of elastic, plastic, and thermal strains) instead of the total mechanical strain (sum of elastic and plastic strains) as per industry standards, can result in incorrect flaw acceptance criteria. For an HPHT flowline with lateral buckling (LB), the thermal strain along the line is influenced by the temperature distribution. Where axial displacement is restricted, this thermal strain will induce axial forces, and thus mechanical strains and stresses. Different patterns of mechanical strains may be induced, depending on the boundary conditions. It is difficult to clearly explain the relationship between thermal strain and mechanical strain for a flowline with lateral buckling. To demonstrate this relationship, a simplified one-dimensional flowline under different boundary conditions and frictional resistances is first studied, followed by a finite element (FE) analysis of a flowline with a sleeper under lateral buckling, to show the relationship between stresses and various strains. The analysis results show that the longitudinal mechanical strain is induced by the thermal strain when restricting the pipeline movement by means of boundary conditions, i.e., frictional resistance between the flowline and seabed, or other externally applied loads. The effect of the thermal strain is already captured in the mechanical strain, so does not need to be added again before converting the longitudinal strain into the longitudinal stress for use in an ECA. To summarize, this paper demonstrates by means of practical ECA and other examples that the longitudinal mechanical strain should be converted into the longitudinal stress when accounting for in-service fracture events in the ECA. Instead of the longitudinal total (elastic plus plastic plus thermal) strain provided by FEA software, the longitudinal mechanical strain (elastic plus plastic) should be used in the fracture assessment, which is part of an in-place ECA. Adding thermal strain into the mechanical strain again before converting to stress will lead to incorrect results and can lead to unnecessary weld cut-outs and repairs.