ILI Performance Verification and Assessment Using Statistical Hypothesis Testing

Following an inspection, API 1163 recommends that operators verify the accuracy of the ILI measurements. This paper examines the performance of an ILI tool from two separate perspectives. The first question is whether the ILI tool’s performance meets expectations and contractual requirements. The second question is to what level of accuracy the ILI tool can be relied upon for making integrity related decisions. This paper develops a method of determining the number of excavations and the number of anomalies to be investigated in the field to verify and assess the accuracy of the ILI reported depths. The verification and assessment of the ILI accuracy are two separate questions, and each is addressed as a hypothesis testing procedure. The first hypothesis states that the tool meets expected and contractual standards. That hypothesis is tested against the excavation data. Its acceptance means that the excavation data is consistent with the expected accuracy of the tool, but it does not specifically verify that accuracy. A second hypothesis states that the tool fails to meet some level accuracy as stated by a tolerance and certainty level. That hypothesis is constructed so that when it is tested against the excavation data it is rejected. Its rejection means that the tool exceeds the stated level of accuracy with a high degree of confidence.Copyright © 2006 by ASME