Obtaining Corrosion Growth Rates From Repeat In-Line Inspection Runs and Dealing With the Measurement Uncertainties

Abstract

The ability to accurately determine the rate of corrosion growth along a pipeline is an essential input into a number of key integrity management decisions. For example, corrosion rates are needed to predict pipeline reliability (probability of failure and/or probability of exceedance) as a function of time, to identify the need for and timing of field investigations and/or repairs and to determine optimum re-inspection intervals to name just a few applications. As more and more pipelines are now being inspected using intelligent in-line inspection (ILI) tools for a second or even third or fourth time, pipeline operators require reliable guidelines for comparing repeat ILI data sets to obtain valid corrosion growth rates. Because of the measurement uncertainties associated with corrosion size estimated from a single ILI run, the corrosion growth rate calculated from consecutive ILI runs has a degree of uncertainty that needs to be considered in determining valid and accurate corrosion growth rates. The ratio between the measured corrosion growth and the measurement error is an important parameter in determining a meaningful distribution of the corrosion growth rate either when performing defect to defect comparisons or when comparing the defect populations in pipeline segments. When this ratio is small the associated uncertainty can be too large to make meaningful probabilistic inferences. As the ratio increases, the effect of measurement uncertainty becomes more manageable, allowing growth rate distributions to be calculated with reasonable confidence. This paper describes an approach to define the probability distribution of corrosion growth rates as a function of a simple parameter that characterizes the ratio between the ILI-observed corrosion growth and the ILI measurement error. This approach has been developed as part of an ongoing PRCI-sponsored research project to produce procedures for determining and validating corrosion growth rates from repeat ILI runs. The paper also provides examples using sample data from repeat ILI runs showing the application of these procedures, the treatment of measurement uncertainty, the resulting corrosion growth rate information that can be obtained and the associated level of confidence in the results.