In late summer 2005, the U.S. pressurized water reactor (PWR) fleet imposed mandatory inspection requirements upon itself to address the challenge posed by primary water stress corrosion cracking (PWSCC) in PWR reactor coolant system (RCS) dissimilar metal (DM) piping butt welds. Under this program, the highest temperature, and thus most susceptible, locations have been addressed first. The set of highest temperature locations comprises the DM piping butt welds on the pressurizer. Within three years of promulgating the requirements, all pressurizer locations will have been inspected and nearly 90% of these locations will have been mitigated. In October 2006, several indications of circumferential flaws were reported in the pressurizer nozzles at Wolf Creek. These indications raised questions about the need to accelerate refueling outages or take mid-cycle outages at other plants. In order to address these concerns, an industry effort was undertaken to evaluate the viability of detection of leakage from a through-wall flaw in an operating plant to preclude the potential for rupture of pressurizer nozzle DM welds given the potential concern about growing circumferential stress corrosion cracks. Previous calculations of growth of PWSCC in Alloy 600 wrought materials and Alloy 82/182 weld metal materials have assumed an idealized crack shape, typically a semi-ellipse characterized by a length-to-depth aspect ratio. A key aspect of the industry effort involved developing an advanced finite-element analysis (FEA) methodology for predicting crack growth when loading conditions do not lead to a semi-elliptical flaw shape. The work also investigated an extensive crack growth sensitivity matrix to cover geometry, load, and fabrication factors, as well as the uncertainty in key modeling parameters including the effect of multiple flaw initiation sites in a single weld. Other key activities included detailed welding residual stress simulations covering the subject welds, development of a conservative crack stability calculation methodology, development of a leak rate calculation procedure using existing software tools (EPRI PICEP and NRC SQUIRT), and verification and validation studies. This paper will describe the study undertaken to model growth of circumferential weld cracks and its application to a group of nine PWRs with regard to implementation of the industry inspection and mitigation program [1]. The paper will also explore implementation progress of the industry program as the three-year mark approaches, as well as industry actions to support completion of baseline DM weld examinations.
Probabilistic and deterministic investigation on single crack growth in dissimilar metal welds of a piping system
