Under the guidance of the Working Party on Nuclear Criticality Safety (WPNCS), this subgroup will explore burn-up credit loading curves for irradiated PWR fuel through a simplified demonstration for nuclear criticality safety application.

Criticality safety evaluation (CSE) methodologies, including their validation bases, may have different levels of conservatism and comprehensiveness, depending on the application type. Some methodologies can be based on “generic” nuclear criticality safety (NCS) criteria, while for others “case-specific” criteria can be defined. Such case-specific criteria may, for instance, involve nuclear data (ND) adjusted to fit benchmark measurements, or an adjusted calculation bias and related uncertainty for an application system 𝑘_eff. Several approaches, either frequentist or Bayesian statistics or even a combination of these can be used, applying benchmark measurements.

An optimal level of complexity of criticality safety evaluation methodologies can be an open question, in particular in the context of potential industrial applications such as nuclear criticality safety for final disposal canisters. For example, the use of a single generic upper subcritical limit (USL) would noticeably simplify the derivation of the “loading curves” for used nuclear fuel. On the other hand, case-specific approaches such as a burn-up-dependent upper subcritical limit could be more cost-efficient by reducing the number of required disposal canisters, or through cost savings on the canister design. The effects of such potential methodological differences on the loading curves, in terms of required minimal burn-up are not obvious.

The verification and validation of 𝑘_eff calculations together with the comparison of criticality calculations using different codes and nuclear data libraries as well as burn-up credit related studies have been in the focus of WPNCS activities since its creation in 1997. The EG UACSA Phases I and V exercises were focused on the 𝑘_eff bias and its uncertainty for test applications. The SG11 goal was to perform a clean comparison of the bias-corrected 𝑘_eff values and their uncertainties for the specified test cases, between different methodologies and validation approaches (Bayes, Maximum Likelihood, Trending Analysis, etc.). The studies listed above required the use of appropriate tools and methodologies which, however, did not necessarily represent methodologies applied or being developed for realistic practical applications.

The comparison of existing NCS criteria is, for the first time, now a subject for analysis at the WPNCS. Such a study is very useful, in particular, to assess the level of maturity and reliability of the BUC methodologies for applications related to such important and complex tasks as NCS for final disposal canisters loaded with used nuclear fuel (UNF). To address the above questions, the WPNCS SG13 activity has been launched and related exercise specifications have been developed. Making this study as an international exercise under the auspices of the NEA WPNCS provides an opportunity to collect solutions from NEA member countries and different organisations, including industrial, research and Technical Safety Organisations (TSO).

The objective of the SG13 exercise is to imitate a direct comparison of existing NCS approaches as applied to final disposal canisters loaded with UNF, using a well-defined and simplified pseudo-application case. For this purpose, the calculation needs and potential sources of participants’ implementation variability should be minimised as much as possible. Thus, the study is limited to:

• Using a 3D quarter symmetry sector model representing four identical pressurised water reactor (PWR) fuel assemblies (FAs) immersed in water and surrounded by a water reflector;
• Considering a limited number of initial enrichments of fresh FAs;
• Establishing of calculation biases and their uncertainties, based on calculations of experimental benchmarks, optionally supported and complemented by estimated nuclear data uncertainties of the 𝑘_eff results;
• Using a fixed set of credited actinides and fission products;
• Specifying a set of validation benchmarks to define USL;
• Leaving depletion calculations outside the benchmark exercise, and providing the UNF compositions vs. BU (for zero decay time).

The participants’ task is to determine the BU values at which the model 𝑘_eff meets the NCS criteria for each initial enrichment, following their own methodologies. The NCS criteria are to be defined by the participants individually, whereby a common administrative margin for subcriticality equal to 5% (D𝑘_eff) would then be used.

Several “reference” solutions can be provided by a participant, if they are considered equally relevant for the specified exercise. Otherwise, it is expected that a single solution is defined as a “reference” one, leaving the possibility to provide other “optional” and complementary variants that could be useful for verification and sensitivity studies. On this basis, the results labelled as “reference” solutions by the participants will be evaluated, with every single reference loading curve having equal weight.

An output of the study would inform and support practical applications for the safe transportation and final disposal of UNF, by enabling a better understanding of important aspects of the loading curve determination methodology, such as the USL definition, and by providing further insight into the expected methodological uncertainties.

Contact

Coordinator: A. Vasiliev (Switzerland)

Monitor: A. Hoefer (Germany)

NEA Secretariat: J.-F. Martin

Members' working area

WPNCS SG13 working area (password protected | reminder)