Currently, there exist various models that predict the burst capacity of a pipeline containing corrosion defects. Recent studies have indicated that these models tend to be overly conservative for long corrosion defects. This paper, based on a PRCI-sponsored study, aims at minimizing this conservatism through a series of steps. First, different definitions for long corrosion defects prevalent in the literature were examined and compared, and the most suitable criterion was implemented. Next, three existing burst pressure models for general corrosion defects were identified and evaluated: ASME B31G-modified, a model developed at C-FER and a model developed at the University of Waterloo. The suitability of these models for long corrosion defects was assessed using a database of 50 full-scale burst test specimens containing natural long corrosion defects. Finally, based on this evaluation, the most apposite burst pressure prediction model for long corrosion defects was selected and a corresponding model error factor was derived.

This content is only available via PDF.
You do not currently have access to this content.