Inspection sampling of pitting corrosion

The objective of this paper is to estimate the sample size required to assess the maximum pit size over an entire asset that is subject to pitting corrosion.

The Bayesian updating approach and the extreme value method are integrated into a single proposed framework to estimate the sample size. The extreme value method is used to predict the maximum pit size by extrapolating the inspected area (sample) to the entire asset's area. The Bayesian updating approach is used to update prior information obtained from the previous inspection once newly obtained information is available from the current inspection.

The estimated sample size ensures that the predicted maximum pit size over the entire asset is within an acceptable margin of error at a specified confidence level.

Using the proposed framework, two closed-form formulae are developed for calculating the sample size required to assess pitting corrosion. The two formulae address the two situations when prior information is either available or unavailable. These formulae are compared to the classical method for calculating the sample size through a case study of a pipeline subject to pitting corrosion. The comparison shows that the sample size estimated with the classical method is not large enough to assess the maximum pit size over the entire pipeline.