Detection of Active Corrosion From a Comparison of ILI Runs

Abstract

Repeated in-line inspections (ILI) of transmission pipelines have been used for many years to estimate corrosion rates. However, the calculation of a corrosion rate from a direct comparison of ILI anomalies is often dominated by the ILI measurement error. As an alternative to assessing a corrosion rate, it may be possible to use repeated in-line inspections to simply detect the presence of active corrosion. This paper presents the application of various statistical measures to detect active corrosion with a high-level of confidence. From a pipeline integrity management perspective, this method will enable the operator to address each location where there is a high probability of active corrosion. Furthermore, despite there being no explicit calculation of corrosion rates, the advantage of the method is that it can yield an upper bound on the corrosion rate of anomalies left unexcavated on the pipeline.