FEA of Burst Pressure in Subsea Pipelines With Generalised External Corrosion

Vasco Nuno Pereira Moura

Key information

Authors:

Vasco Nuno Pereira Moura (Vasco Nuno Pereira Moura)

Supervisors:

Baiqiao Chen (Baiqiao Chen)

Published in

December 12, 2025

Abstract

This study proposes a combined finite element and probabilistic framework for predicting burst pressure reduction in steel pipelines affected by random generalised corrosion. A three-dimensional FEA model was developed in ANSYS MAPDL and coupled with a MATLAB interactive automation tool called PipeGen to generate probabilistically defined corrosion profiles and perform large sets of simulations. The corrosion profiles follow lognormal distributions described by the mean corrosion depth and its degree of dispersion, allowing a realistic representation of surface irregularity. The model was validated against experimental data from the literature with errors below 4%. A total of 580 simulations covering multiple combinations of mean corrosion depth and its level of dispersion were executed, each combination was repeated ten times to quantify uncertainty within a 95% confidence interval. Results show that the average corrosion depth induces an approximately linear loss of burst capacity, while the dispersion in corrosion depth has nonlinear effects and increases the prediction uncertainty. Two quadratic surfaces were fitted to the normalised burst pressure data, producing empirical equations that estimates residual strength from inspection parameters. The proposed approaches enable rapid integration of in-line inspection data into integrity assessments, offering a more straightforward alternative to existing deterministic codes. Although calibrated for a single geometry and steel grade, these results provide the first steps in the development of a predictive model for burst pressure capacity in pipelines with random generalised corrosion, which requires further generalisation to a broader range of pipeline configurations and steel grades.