Calibration of load and resistance factors for the limit state-based assessment of dent-gouge defects on energy pipelines

Abstract

Buried oil and gas pipelines are sometimes subjected to mechanical damage resulting from the impact with foreign objects such as excavation equipment or rocks. A severe form of the mechanical damage is the so-called dent-gouge. A dent is a permanent inward deformation of the pipe wall, whereas a gouge is a localized metal loss on the external surface of the pipe wall. In a dent-gouge, the gouge is typically located at the deepest point of the dent. The pressure containment capacity, i.e. burst capacity, of a pipeline containing a dent-gouge flaw can be markedly lower than that of the pipeline in the pristine condition. Dent-gouges are detected and sized by inline inspection tools that are widely used in the pipeline industry as part of the pipeline integrity management program. However, geometric properties of the dent-gouge reported by the inspection tool contain non-negligible measurement errors, which contribute to uncertainties in the fitness-for-service (FFS) assessments of dent-gouges. The material properties of the pipe steel, in particular the material toughness, are also sources of significant uncertainties in the assessment. Finally, FFS assessment models commonly employed in the pipeline industry are generally of empirical or semi-empirical nature, and associated with high model uncertainties. A limit state-based FFS assessment framework for dent-gouges is proposed in the present study to address the above-mentioned uncertainties. This framework allows pipeline engineers to identify dent-gouges that require interventions with a consistent level of implied failure probability. To this end, the Gaussian process regression is first employed to improve the predictive accuracy of a dent-gouge burst capacity model well known in the pipeline industry such that the model uncertainty is reduced to a satisfactory level. The limit state-based assessment is then formulated by assigning resistance factors to key variables in the improved dent-gouge capacity model and a load factor to the internal pressure of the pipeline. The values of the load and resistance factors are further calibrated based on pre-defined allowable failure probability levels and a large number of analysis cases with wide ranges of pipe geometric and material properties, sizes of dent-gouges and utilization factors under internal pressure. The limit state assessment framework developed in the present study can be applied to other types flaws on pipeline such as metal-loss defects and surface cracks.