Statistical approach to inspection planning and integrity assessment
Authors:
- F KhanFKF Khan
- R HowardRHR Howard
- 1 January 2007
- journal article
- Published by British Institute of Non-Destructive Testing (BINDT) in Insight - Non-Destructive Testing and Condition Monitoring
- Vol. 49 (1) , 26-36
Abstract
Process industries are threatened by leaks and failures causing serious consequences. To avoid such incidents integrity inspections are carried out by means of nondestructive testing methods. Typically, these methods are applied to sample areas, which are later extrapolated over the whole area of equipment and also used in estimating many important parameters such as degradation rate, residual life, inspection intervals, etc. Statistical methods to improve the estimation of degradation rate or to extrapolate the data have been suggested for over four decades; however, they have been applied only in a few isolated cases. These statistical methods, when combined with reliability methods, offer a potential for better use of inspection results and for the prediction of the probability of future leaks and the remaining useful life of the equipment. As observed by the UK HSE (2002), the application of these methods is not common, largely because the use of statistics requires specialist knowledge, and because no standard reference exists. Addressing the latter issue, HSE has recently proposed broad guidelines for the use of statistics in the analysis of corrosion inspection samples. The present work is an attempt to transform these broad guidelines into a practical working method. This paper presents a simplified practical approach for the use of statistical tools for inspection planning and integrity assessment. The study is focused on corrosion related material degradation of piping on an offshore production facility. The application of the approach is demonstrated using seven years' past inspection data. The case studies presented herein illustrate the benefits to be gained by applying well-established statistical methods to the analysis of inspection data, amongst which are: quantification of inspection findings, the ability to specify the extent of inspection required for the defined level of confidence, and the use of limited inspection data to infer the condition of inspected areas under similar exposure conditions to those inspected. Further, this approach can be integrated with riskbased inspection and integrity assessment methods thereby improving the value of these assessments.Keywords
This publication has 2 references indexed in Scilit:
- Statistical Analysis of Extreme Values1997
- Extrapolation of Extreme Pit Depths in Space and TimeJournal of the Electrochemical Society, 1990