RADIATION DOSE TO REACTOR PRESSURE VESSEL AND INTERNALS

• evaluate the accuracy of the calculation methods used in NEA Member countries for predicting long-term radiation doses to reactor pressure vessels and internal structures; • identify points for improvement and validate the performance of improved methods for fluence calculations; • initiate a study on the modelling of radiation-induced damage in metals. This report fulfils the first and second objectives of the Task Force by providing a critical discussion of the most recently published literature on computational methods of reactor dosimetry, and a detailed overview of the computational techniques currently used in the dosimetry programs of NEA Member countries. The analysis presented suggests additional factors whose importance should be considered in modelling these phenomena. Proposals are made for further work on these problems. Adequate knowledge of cumulative radiation doses, or fluence, is an essential step in developing and validating more effective models for prediction of radiation damage to reactor components. The present report reviews the computation techniques for calculating neutron/gamma radiation doses to reactor components, and describes in considerable detail the methods presently used in NEA Member countries for computing long-term cumulative dose rates. Although the median of results reported in national calculations appears to lie within 20 per cent difference between calculations and measurements, significantly higher and lower values are also reported. Moreover, the numbers reported are difficult to compare, since each country has its own methodology including different reactors, computer codes, nuclear data sets and measurement procedures. On the basis of these country reports, the working group concludes that no firm judgement can be formed on the current international level of accuracy in pressure vessel fluence calculations. To identify the range of differences between calculations and measurements, the NEA/NSC Task Force on Computing Radiation Dose and Modelling of Radiation-induced Degradation of Reactor Components (TFRDD) is conducting an international " blind " intercomparison exercise. The varying methodologies are applied to predicting dose rates in the Belgian Venus test reactor, calculating in both two and three dimensions for comparison with measured data. This comparison of methods for dosimetry calculations should lead to consensus on: • The level of accuracy of methods currently used in the NEA Member countries in calculating radiation dose to reactor components; • The relative merits of different calculation methods; • Possible improvements to these methods; • The advantages of fully 3-D methods over 2-D methods. Additional problem areas were identified during the review of recent literature: …