Statistical Evaluation for Understating the Characteristics of Potentially Undetected Defects in Metallic Pipelines

Inline pipeline inspections provide actionable information for pipeline managers to ensure reliability of their water assets. Ideally, inline inspection tools provide 100% coverage of the inspected asset; however; in practice, environmental elements or tool limitations may prevent full inspection coverage of the subject asset, therefore resulting in incomplete condition data. This leads to decreased confidence in the actionability of those inspection results. To improve upon the confidence in insights gained from incomplete inline inspection coverage, this paper presents a statistical approach to maximize and leverage the inspection data to better understand and quantify potentially undetected defects. It then applies this approach to a 900-mm (36-in.) ductile iron pipeline. Leveraging the inspection data, a series of statistical methods have been used to provide insights and extend past the perceived limitations in that data, for a better understanding of potentially undetected defects. These insights can then be used to plan a more targeted approach to manage the asset, for example, by informing recommended next steps for data collection.