Probabilistic fracture mechanics analysis of reactor vessels with low upper-shelf fracture toughness

A class of submerged-arc welds used in fabricating early reactor vessels has relatively high copper contents. Studies have shown that when such vessels are irradiated, the copper contributes to lowering the Charpy upper-shelf energy level. To address this concern, 10CFR50, Appendix G requires a fracture mechanics analysis to demonstrate an adequate margin of safety for continued service. The B and W Owners Group (B and WOG) has been accumulating J-resistance fracture toughness data for these weld metals. Based on a mathematical model derived from this B and WOG data base, the first Appendix G analysis was performed. Another important issue affecting reactor vessel integrity is pressurized thermal shock (PIS) transients. In the early 1980s, probabilistic fracture mechanics analyses were performed on a reactor vessel to determine the probability of failure under postulated accident scenarios. Results of such analyses were used by the Nuclear Regulatory Commission (NRC) to establish the screening criteria for assessing reactor vessel integrity under PTS transient loads. This paper addresses the effect of low upper-shelf toughness on the probability of failure of reactor vessels under PTS loads. Probabilistic fracture mechanics codes were modified to include the low upper-shelf toughness model used in a reference and a series of analyses was performed using plant-specific material conditions and realistic PTS scenarios. The results indicate that low upper-shelf toughness has an insignificant effect on the probability of reactor vessel failures. This is mostly due to PTS transients being susceptible to crack initiation at low temperatures and not affected by upper-shelf fracture toughness