Evaluation of the ASTM E900 ΔT41J Prediction Equation in Light of New Data

Abstract

Predicting the future embrittlement state of operating nuclear power reactors is a critical part of their safe life management. Guidance on this topic traditionally comes from various sources, including ASTM standards such as ASTM E900, Standard Guide for Predicting Radiation-Induced Transition Temperature Shift in Reactor Vessel Materials. This paper summarizes ongoing efforts within ASTM to evaluate the Charpy transition temperature shift (ΔT41J) equation contained within ASTM E900-15 in light of the 195 new data records made available since the 2015 update. A simple statistical analysis of how well the trends in these data conform to the E900-15 predictions was performed as well as a more comprehensive bootstrap evaluation to ensure that our conclusions are robust across sample size and not a consequence of the particular 195 newly available records. Both evaluations demonstrated that, especially for the variables nickel and manganese, a statistically significant variation exists between the available data and E900-15 ΔT41J predictions. However, the bootstrap analysis demonstrated that these discrepancies in how E900-15 represents certain variable trends may be approximately offsetting, and, in any event, are less than the ΔT41J measurement error. Thus, although these results suggest that a better embrittlement trend curve (ETC) than E900-15 exists in principle, they also demonstrate that the continued use of E900-15 is not expected to produce errors in ΔT41J predictions of any practical significance. Therefore, we believe that there is no pressing need to update the E900-15 ETC.