Embrittlement trend curves (ETCs) are used to estimate the magnitude of neutron irradiation embrittlement as a function of both exposure (fluence, flux, temperature, …) and composition (copper, nickel, manganese, phosphorus, silicon, …) variables. ETCs provide information needed to assess the structural integrity of operating nuclear reactors, and to determine their suitability for continued safe operation. ETCs may use any of a number of different metrics (e.g., ΔT41J, ΔYS, ΔTo) to quantify the magnitude of embrittlement, and may use data from a number of different sources (e.g., commercial power reactors, material test reactors). Past efforts on ETC development in the United States have used data drawn from domestic licensees. For example, the ETCs in Regulatory Guide 1.99 Revision 2 and 10 CFR 50.61a were based on ΔT41J data from Charpy specimens tested as part of licensee surveillance programs [RG199R2 Regulatory Guide 1.99, “Radiation Embrittlement of Reactor Vessel Materials,” Rev. 2, U.S. Nuclear Regulatory Commission, Washington, DC, May 1988 and 10CFR5061a Title 10, Section 50.61a, of the Code of Federal Regulations, “Alternative Fracture Toughness Requirements for Protection against Pressurized Thermal Shock Events,” promulgated Jan 4, 2010.]. While this approach has addressed past needs well, future needs such as power uprates, license extensions to 60 years and beyond, and the use of low copper materials in new reactors produce future operating conditions for the U.S. reactor fleet that may differ from past experience, suggesting that data from sources other than licensee surveillance programs may be needed. While data for these conditions is currently available, it is mostly quantified using embrittlement metrics other than ΔT41J, and it arises mostly from sources other than the U.S. surveillance program. This paper draws together embrittlement data expressed in terms of ΔT41J and ΔYS from a wide variety of data sources as a first step in examining future embrittlement trends. This evidence supports development of a “wide range” ETC based on a collection of over 2500 data. The paper includes an assessment of how well this ETC models the whole database, as well as significant data subsets. Comparisons presented herein indicate that a single algebraic model, denoted WR-C(5), represents reasonably well both the trends evident in the data overall as well as trends exhibited by the following four data subsets: ΔT30 data from the U.S. surveillance program, ΔT30 data from non-U.S. surveillance programs, ΔT30 data from test reactor irradiations, and ΔYS data from test reactor irradiations. Of particular importance, the WR-C(5) model indicates the existence of trends in high fluence data (Φ > 2−3 × 1019 n/cm2, E > 1 MeV) that are not as apparent in the U.S. surveillance data due to the limited quantity of ΔT30 data measured at high fluence in that dataset. Additionally, WR-C(5) models well the trends in both test and power reactor data despite the fact it has not term to account for flux. While the appropriateness of using data from such a variety of sources to inform a trend curve is debated by the technical community, it is suggested that one appropriate use of the WR-C(5) trend curve may include the design irradiation studies to validate or refute the findings presented herein. Additionally, WR-C(5) could be used, along with other information (e.g., other trend curves, theoretical expectations, plant-specific data, etc.), as one tool to predict irradiation trends pending the availability of confirmatory data in the high fluence régime.
