Use of an Offset in Assessing Radiation Embrittlement Data and Predictive Methods

Analyses of reactor pressure vessel (RPV) surveillance data from Charpy V-notch shift results coupled with our latest knowledge of the mechanisms of radiation embrittlement have led to new predictive correlations/models that have a strong technical underpinning. In this paper we examine how well the new CRIEPI embrittlement predicts US RPV surveillance data. Secondly, we note that within the US surveillance data sets there are indications that the data may follow the same form as the predictive models, but the data may be offset by a constant amount (either positive or negative) from the predictive values. This offset can be attributed in some cases to inadequate baseline data. In other cases, there does not appear to be a constant offset, or such an offset is hidden by data scatter. This paper also reviews the potential use of an offset adjustment and focuses on several surveillance datasets for comparisons.