Reliability Analysis of the Collapse of Corroded Submarine Pipelines Subjected to Bending Moment

With the increasing development of oil exploration in the oceans, the demand for submarine pipelines has risen considerably, which implies significant environmental risks in the event of structural failure, resulting in oil leaks and consequent billion-dollar environmental damage. The integrity of these structures is primarily compromised by external pressures and bending moments, especially in areas of high curvature between platforms and extraction points. These bending moments, as identified in the investigations of Bjørset (2000), decrease the pipelines resistance to collapse due to the progressive ovalization of the sections, reducing the moment of inertia. Additionally, corrosion defects, which decrease the thickness of the pipeline walls, are prevalent and significantly impact the resistance to collapse, requiring precise evaluation in simulations (CABRAL, 2007). For the modeling of this mechanical phenomenon, the Finite Element Method (FEM) has shown promising results, being chosen to calculate the collapse pressures by combining the effects of bending moments and corrosion defects (MOTTA, 2020). However, the simulation process faces challenges due to uncertainties associated with measurement errors, equipment calibration, and manufacturing faults. These uncertainties are generally overlooked in traditional deterministic methods but can be addressed through reliability analyses that use statistical concepts to establish failure functions and assess failure probability, resulting in a more robust and reliable analysis (MOTTA, 2020). These analyses are rarely discussed in the existing literature, suggesting a need for more well-founded and detailed approaches. The results of this study, with the application of reliability techniques, are expected to provide safer insights into the influence of bending moments on the resistance of submarine pipelines. Furthermore, it is intended to compare deterministic equations with simulation results, proposing, if possible, adjustments to these equations to optimize the accuracy of the provided results.