Statistical study of localized internal corrosion defects in oil and gas pipelines through sampling inspection

Historical failure data of corroded oil and gas pipelines helps estimate risk and reliability using statistical models if the distribution of the characteristics of corrosion defects is known. This research presents a field survey of corrosion defect characteristics in Mexican oil and gas pipelines. Because of the different conditions that pipelines typically undergo and the different fluids that they transmit, it was found that corrosion defect depth has a bimodal behavior. This unusual characteristic is a reason to study the cause of this conduct in more detail. The fluid properties are of great importance to characterize statistically the internal corrosion damage. A Bayesian methodology allows the statistical estimation of the distribution for the depth of internal corrosion defects using limited information obtained from "unpiggable" pipelines.