Uncertainty in the Assessment of Corroded Pipelines

Pipeline corrosion is uncertain by nature. It depends on several uncontrolled parameters, such as the transporting fluid, the surrounding soil, pipeline geometry, and material strength. As mentioned in the previous chapter, pipeline operators can monitor corrosion at the inner/outer wall using ILI measurements every 2 to 6 years. However, these inspections are subject to uncertain measurements as well as how the corrosion will evolve between consecutive inspections. Uncertainties include the degradation model, the location of the defects, and the local uncertainties of the inspection tool. The degradation uncertainties are related to the lack of knowledge (epistemic uncertainty) of how each defect evolves. Also, local variations of features affect the degradation process, e.g., material properties, stress, temperature, or pressure. The corrosion degradation process changes with time, and finally, the data comes from imperfect inspection results. ILI measurements may hide existing defects that did not fulfill the detection requirement of the inspection tool. Also, the reported defects can have inaccurate locations and sizes, depending on the implemented PIG tool. This chapter focuses on the local inspection uncertainties and those associated with the temporal modeling of the degradation corrosion. The corrosion prediction is required to support further decisions to maintain adequate pipeline integrity. ILI measurements indicate the pipeline condition, which allows updating the corrosion predictions between inspections and identifying further critical pipeline segments. Different parameters such as soil aggressiveness, operating parameters, or the fluid would affect the obtained predictions. This chapter introduces the uncertainties associated with ILI inspections and the prediction of the degradation process.