Submarine pipelines can get damaged in various ways. An important failure mode is collapse. In the deep sea, the hydrostatic pressure is very large and the threat of said failure exists. There are prospects for pipelines in water depths of up to—at least—3,700 m. A pipeline requires a very thick wall to cope with such, hostile environment. This means that D/t, the diameter-to-wall-thickness ratio, must be low.
The analytical limit-state equations included in DNV-ST-F101, a prevailing design standard for submarine pipeline systems, do not work well for pipe with extremely thick walls. Collapse-pressure predictions are increasingly conservative for lower D/t. By following the standard, pipelines operating in ultra-deep water may be over-dimensioned. This can culminate in manufacturing requirements that adversely affect project economics or threaten the technical feasibility altogether.
This paper proposes an improved analytical framework for calculating the collapse pressure. Like the DNV model, the proposed model reflects the fundamental underlying and interacting physics of material yielding and geometric instability. It has been found that the collapse resistance of pipe with extremely thick walls (roughly D/t ≤ 15) is underestimated when using the DNV formulation. On the other hand, the capacity of pipe with slightly thinner walls (about 15 < D/t ≤ 40) could be overpredicted. The latter may occur if cold-formed line-pipe joints have not been subjected to some form of heat treatment after forming—for instance, via a coating process.
To improve consistency between the predictive model and finite- element analysis, the analytical model has been extended to account for behaviours that were not yet included: triaxiality of stress and plastic bifurcation. The result is an elegant set of equations that can be used to evaluate the collapse pressure. Although the focus of this study was thick-walled pipe, the improvements lead to better collapse-pressure predictions for pipe with thinner walls, too.
