A New Probabilistic Methodology for Undertaking External Corrosion Direct Assessment

External Corrosion Direct Assessment (ECDA) has now become acknowledged, by the Office of Pipeline Safety (OPS) in North America, as a viable alternative to both in-line inspection (ILI) and the hydrostatic pressure test for the purpose of managing the integrity of high pressure pipelines. Accordingly an ECDA standard is now in existence. The essence of ECDA is to use indirect above ground survey techniques to locate the presence of coating and corrosion defects and then to investigate some of the indications directly by making excavations. However, one of the problems of above ground survey techniques is that they do not locate all defects and are susceptible to false indication. This means that the defects will not be present at all indications and that some defects will be missed. In view of the limitations of above ground survey techniques the ECDA standard requires that at least two complimentary survey techniques should be used. The selected survey techniques will depend on the nature of a particular ‘ECDA segment’, taking account of the surface characteristics. However, in many situations the surveys will include a coating survey and a corrosion survey. In general the outcome from these two surveys will be N_H locations at which just the coating survey gives an indication, N_C locations at which just the corrosion survey gives an indication and N_HC locations at which both surveys give an indication. This paper presents a new probabilistic methodology for estimating the distributions of the actual numbers of coating and corrosion defects, taking account of the outcomes of the surveys and the probabilities of detection and false indication of both techniques. The method also shows how the probabilities of detection and false indication are updated depending on what is found during the excavations and the distributions of the numbers of remaining corrosion and coating defects are subsequently modified. Based on a prescribed repair criterion the analysis is used to determine the probability that at least one remaining corrosion defect will exceed the repair criteria. As excavations are sequentially performed the probability naturally reduces. The attainment of an acceptably low probability is used as a trigger to terminate the excavation programme. A detailed description of the development of the method is given in this paper and the application is illustrated through a simple numerical example. A description of how the method is used to build a Direct Assessment module for a pipeline integrity management system is described in an accompanying paper.