Volume 9, Issue 4 (April 2012)
Path Forward for Dosimetry Cross Sections
In the 1980’s the dosimetry community embraced the need for a high fidelity quantification of uncertainty in nuclear data used for dosimetry applications. This led to the adoption of energy-dependent covariance matrices as the accepted manner of quantifying the uncertainty data. The trend for the dosimetry community to require high fidelity treatment of uncertainty estimates has continued to the current time where requirements on nuclear data are codified in standards such as ASTM E 1018. This paper surveys the current state of the dosimetry cross sections and investigates the quality of the current dosimetry cross section evaluations by examining calculated-to-experimental ratios in neutron benchmark fields. In recent years more nuclear-related technical areas are placing an emphasis on uncertainty quantification. With the availability of model-based cross sections and covariance matrices produced by nuclear data codes, some nuclear-related communities are considering the role these covariance matrices should play. While funding within the dosimetry community for cross section evaluations has been very meager, other areas, such as the solar-related astrophysics community and the US Nuclear Criticality Safety Program, have been supporting research in the area of neutron cross sections. The Cross Section Evaluation Working Group (CSEWG) is responsible for the creation and maintenance of the ENDF/B library which has been the mainstay for the reactor dosimetry community. Given the new trends in cross section evaluations, this paper explores the path forward for the US nuclear reactor dosimetry community and its use of the ENDF/B cross-sections. The major concern is maintenance of the sufficiency and accuracy of the uncertainty estimate when used for dosimetry applications. The two major areas of deficiency in the proposed ENDF/B approach are: 1) the use of unrelated covariance matrices in ENDF/B evaluations and 2) the lack of “due consideration” of experimental data in the evaluation.