Practical use of defect assessment procedures for industrial component integrity assessment

Abstract

The practical use of defect assessment procedures for industrial component integrity assessment is described through two practical examples in this paper. In the first example, the procedure is used to perform low temperature fitness-for-service (FFS) analysis of a longitudinal seam welded vessel manufactured from a duplex stainless steel. The impact energy obtained from the Charpy impact test performed on the weld of the vessel was found lower than the value of the minimum impact energy criteria given by the British Standard BS 5500. Charpy impact energy and fracture toughness correlation and failure assessment diagram (FAD) methodology were used in this FFS analysis to determine the fracture resistance of the vessel. For the assumed defect size used in the assessment, the weld was found to meet the fracture resistance criteria and therefore would still be fit-for-purpose. The FAD analysis was, however, repeated using the J-values obtained from the CTOD test to gain better confidence as the Charpy impact test does not provide direct measurement of the fracture toughness. The FAD was later used to determine the critical surface crack length which was then represented as a function of crack depth. The Charpy impact energy correlation method was found to be more conservative than the method of evaluation using the CTOD J-values. The lack of side wall fusion (LOF) which is a typical defect in this type of vessel would usually be influenced by the size/diameter of the electrode wire used in the weld and the number of runs. In this case, the fracture resistance of the 15mm thick vessel with the longitudinal seam weld should be adequate if less than 5mm diameter of electrode wire is used.The second example illustrates defect assessment of a high temperature plant component. A defect was found in the high pressure final superheater header. A defect assessment incorporating FAD on the header showed that the defect was non-critical. This led to the need to perform creep and fatigue crack growth calculations and remaining life assessment in order to determine the mitigation plan for the engineers. The deterministic approach, which mainly considers the worst case scenario, suggested that the remaining life of the header was approximately 4.5 years. Probabilistic analysis showed that the component could still be fit for service up to 6 years. This will allow the engineers to mitigate a more efficient plan with a decision to either repair or replace the header and when. The use of probabilistic lifing methodologies and algorithms could consequently bring considerable financial benefits to the plant owners/operators e.g. by avoiding premature component repair or replacement. Nevertheless, it would be in the management’s interest to avoid a forced outage. The recommendation would be that the component could still be fit to use until their next minor outage (in 3 years). From then, the remedy option would be to grind down to a depth of suspected crack, excavate a small area/surrounding which might be affected, followed by correct and regular monitoring, or alternatively replace the component.