Application of secondary and residual stresses to the assessment of the structural integrity of nuclear power-generating plant

Abstract

Magnox nuclear power stations were built in the 1960s to design codes that, in general, required weldments to be subject to a post-weld heat treatment to remove residual stresses. Implicit in this was the assumption that the heat treatment reduced the stresses significantly such that as stated in the codes “stresses caused by fabrication and welding are practically annulled”. However, it has since been realised that the stresses remaining, although small, could still be significant when incorporated into the subsequently developed failure avoidance methodologies such as R6. Moreover, either at the time of construction or during the operating life, repairs are undertaken to remove manufacturing or service-induced defects. These repairs can be put into service with or without a post-weld heat treatment. As a consequence of a paucity of data for the two- and three-dimensional distribution of the magnitude of these stresses, extremely conservative values of stresses have been adopted to ensure that the plant is secure against the design intent throughout the service life. In this paper, the requirements of the failure-avoidance methodology R6 Revision 4 are briefly reviewed with respect to the categorisation of secondary and residual stresses and the application of the three approaches for determining the as-welded residual stress distribution at room temperature. These three levels comprise, Level 1, simple estimates, Level 2, bounding profiles, and Level 3, detailed evaluation. Examples are presented where knowledge of the residual stresses has been an important component of the overall integrity assessment. The first relates to multi-pass weldments in superheater headers fabricated from a ferritic steel and the second to the weldments in the standpipes, both at Magnox power stations with concrete pressure vessels. Although in these cases the weldments had been subject to a post-weld heat treatment, the remaining residual stresses presented a significant challenge to the perceived structural integrity. Finally, the future requirements for the residual stresses to be incorporated into structural integrity assessments of ferritic steel nuclear power plant components and structures will be discussed. This will be by reference to various weldment geometries and heat treatment cycles including repair welds and fillet attachment welds. In particular, there is a need for improved measurement techniques and analytical/computer models to provide more realistic three-dimensional stress distributions for a range of weldment geometries. In addition, there is a need to revisit the criteria adopted to establish the initiation fracture toughness of ductile ferritic steels with respect to the attendant redistribution of residual stresses.