Test reactor embrittlement database

Several industry databases of surveillance test data have been maintained over the years. For example, during the 1990s EPRI maintained the Power Reactor Embrittlement Program (PREP) database (EPRI, 1996), which was developed under joint EPRI and CRIEPI (Central Research Institute of the Electric Power Industry) sponsorship. Later, EPRI sponsored the RPVDATA database program (EPRI, 2000) as a successor to PREP. The industry databases have faced greater sustainability issues than the NRC databases because of limited funding. 

The Russian predictive embrittlement method is based primarily on test reactor data and utilizes the shift in CVN impact energy properties essentially at the 47 J temperature (PNAE, 1989). The effects of Cu, P and O (E > 0.5 MeV) are included in the method. A new local approach called the Unified Curve (similar to the Master Curve method) for assessing vessel integrity has recently has been added for structural analysis purposes (Margolin et ai, 2007). Additionally, an IAEA activity on embrittlement prediction for WWER-440 RPVs was completed (IAEA, 2005), with the published IAEA report providing recommended guidelines based on a larger database than previously available. 

In France, initially, US Regulatory Guide 1.99, Revision 1 was recommended. Then several irradiations were made in test reactors with split melts varying the level of residual elements. From these data, Guionnet and co-workers produced several equations. These trend equations included the effect of known embrittling elements but also included the effect of elements such as P, Sn and N. An analysis of a larger database was made by Brillaud et al., and from this evaluation the so-called FIS and FIM formulae were derived. 

There are four main types of reactors in France CPO (900MWe), CPY (900MWe), P4 and P 4 (1300MWe) and N4 (1400MWe). The surveillance data from six CPO reactors and 26 CPY reactors as well as the test reactor data, including those obtained after the development of the FIS model, were used for the calibration of a new embrittlement correlation method. The surveillance database includes 139 pieces of base metal data, 130 weld metal data, 131 heat-affected zone (HAZ) data and 93 standard reference material (SRM) data, while the test reactor database contains 36 pieces of... 

Original trend curves given in the Russian Code were constructed on the basis of test results obtained after irradiation within the material qualification programmes in experimental reactors. Thus, their lead factor (relation between neutron flux in the experiment and in the RPV) was more than 100, and such a large factor is not allowed by current standards. Moreover, the database of results was relatively small at that time and irradiation embrittlement of 15Kh2NMFA-type steel was studied only on steePwelds with nickel content lower than 1.5 mass %, as this was the original intention for this type of steel. 

---