Quantifying Nuclear Data Uncertainty from Scattering Angular Distributions for Fluence Studies

The development of advanced reactors – such as Lead-cooled Fast Reactors (LFRs) – together with the enhancement of the performance of the current fleet demands highly reliable assessments of nuclear data uncertainties. Traditionally, such uncertainties are primarily studied in the context of criticality safety and, to a lesser extent, in waste management. However, other essential aspects—such as the aging and integrity of ex-core components exposed to neutron irradiation—are also significantly impacted by nuclear data.

Of particular relevance is the accurate estimation of neutron fluence or displacements per atom (dpa) on components like the core barrel. These estimations influence design decisions regarding neutron leakage, which links directly to material degradation and component lifetime.

Quantifying how nuclear data uncertainties affect the neutron flux is therefore of direct importance. Covariance data are available for several nuclear data types, including cross sections, prompt neutron multiplicities, and secondary energy and angular distributions. Tools like SANDY, in combination with Monte Carlo particle transport solvers such as SERPENT-2 or MCNP6, allow for stochastic propagation of most of these uncertainties—except, until now, for angular distributions.

Prior studies have identified that in compact systems sensitive to lead – particularly involving isotopes such as 206Pb, 207Pb, 208Pb – the elastic scattering reactions above 100 keV, and the corresponding angular distributions of scattered neutrons, significantly influence the neutron leakage behavior. In addition, angular distributions for structural materials (e.g., 56Fe, and possibly Cr and Ni in steel alloys) also contribute notably to neutron transport beyond the core, and to dpa calculations in ex-core regions. Furthermore, magnesium (24Mg), if used as a reflector, introduces additional uncertainty.

The main goals of the project are:

• To develop a methodology for propagating uncertainties from elastic scattering angular distributions.
• To benchmark this methodology by assessing the impact on dpa calculations for the reactor core barrel in representative systems.
• To compare the relative importance of angular distribution uncertainties against other nuclear data uncertainty sources (e.g., cross sections).

The project is expected to yield:

• A modified or extended version of SANDY capable of sampling elastic scattering angular distribution covariances, or, alternatively, a robust dataset of perturbed files for UQ studies.

• A quantitative evaluation of the contribution of angular distribution uncertainties to critical metrics such as fast neutron flux and dpa at the core barrel.

• A comparative analysis showing the importance of angular distribution uncertainties relative to other nuclear data uncertainties.

• More robust design bases for future LFRs.

• Possibly, a peer-reviewed publication presenting the methodology and results.