Wall thinning in pressure retaining components, especially due to the flow-accelerated corrosion, has been a significant factor affecting the safety and unplanned system downtimes. On the other hand, overestimating the impact of possible wall thinning often leads to unnecessary or expensive inspections and replacements. The simplified or quick (short-cut) methods of analysis and prediction often lack the requisite degree of accuracy and confidence. This paper presents a few techniques for better analysis of the wall thinning data to address these issues. These techniques make use of the statistical methods, pattern recognition, and optimization to perform a robust data filtering and thinning rate estimation that accounts for measurement uncertainty. The techniques are discussed with application to a large database and an inspection program. The impact of these analytical improvements is presented in comparison with results of the simplified method of analysis. The results include both the margin on remaining life and the projected wall thinning rates, with implications for inspections.

This content is only available via PDF.
You do not currently have access to this content.