Consultant's Meeting on Integral Data in Nuclear Data Evaluation

Consultants' Meeting on Integral Data in Nuclear Data Evaluation

14-17 November 2017, IAEA, Vienna

Objectives

The purpose of the Consultants' Meeting is to discuss the validity and scope of the use of integral data in the process of the assembly of evaluated nuclear reaction data files.

Background

Integral quantities can usually be measured much more accurately than differential nuclear data, so it is tempting to use such data to tune evaluations to improve integral performance. Integral data imply spectrum-averaged cross sections or cross section ratios, kinetic parameters, leakage spectra, scattered-neutron yields, multiplication factor, etc. Experience from the analysis of criticality benchmarks indicates that data adjustments on such benchmarks are not unique. An obvious example is fairly good performance of all major libraries in predicting the criticality of major benchmark cases in spite of relatively large differences in the cross sections, where good performance is achieved mainly due to compensating effects. Similarly, blind application of data adjustment techniques can lead to data files that apparently perform well in selected integral benchmarks, but violate the uncertainties in experimentally measured differential nuclear data. In general, integral benchmarks can be used for data validation, but should be avoided for the adjustment of general-purpose libraries. Obviously, one needs to be very careful if integral data are to be used in the evaluation process. The main questions that need to be answered relates to the proper use of different categories of integral or quasi-differential data.
Suitable candidates that are directly usable in the evaluation process may include:
• Spectrum-averaged cross sections in reference neutron fields.
• Reaction rate ratio measurements in well-characterised irradiation facilities.
• Etc.

Benchmarks limited in use for library validation or for generating specific application libraries may include:
• Quasi-differential measurements of reaction-rate spectra; for example the RPI scattering benchmark, measured leakage spectra from spheres or through slabs, the Grenoble lead slowing-down spectrometer measurements, etc.
• Criticality benchmarks.
• Complex reactor benchmarks; for example, power distribution, isotopic composition with burnup, etc.

Scope

At present we wish to focus on the first category of integral benchmark experiments. The tasks to be addressed are the following:
• List the data sets that could be used in the evaluation process of general purpose libraries.
• Define acceptability criteria.
• Propose the procedures for compiling and archiving the data.

Expected Output

• Document summarizing the recommendations on the use of integral data in the nuclear reaction data evaluation process, including the definition of criteria for acceptability and procedures for compilation of such data.
• Recommendations for additional integral measurements to be performed.

Outcome

The expected long-term outcome is the improvement in the reliability and enhanced performance of evaluated nuclear data files in modelling integral experiments and a reduction of the uncertainties in the calculated integral quantities due to nuclear data.