Influence of manufacturing uncertainty treatment on neutronics uncertainty analysis for lattice physics parameters in PWR-UO₂ fuel assembly

ABSTRACT

This study discussed the practical perspective on the neutronics uncertainty analysis related to manufacturing uncertainties by identifying discrepancies in analysis results due to manufacturing uncertainty treatment. First, the probability density function given for manufacturing uncertainties was focused on. The random sampling-based uncertainty analysis was performed with the normal and uniform probability distributions, using data from Phase I of the OECD/NEA/NSC LWR-UAM benchmark for PWR-UO₂ fuel assembly. The results showed that the uncertainty of lattice physics parameters was smaller with the normal distribution, which is explained that broader sampling results were given by the uniform distribution. Then, the influence of manufacturing uncertainty treatment was investigated under the simple assumption, with or without the correlations between the manufacturing uncertainties, considering the total mass of fuel materials and the diameter of fuel rods. With such correlations, the uncertainty analysis results generally became small since the manufacturing uncertainties were modified to be small by calculating the covariance matrix for the manufacturing parameters. Consequently, clarification of the correlation data in the manufacturing process or introduction of a conservative assumption for the correlation condition related to manufacturing parameters is necessary for the appropriate neutronics uncertainty analysis.

KEYWORDS:

• Manufacturing uncertainty
• LWR-UAM benchmark
• uncertainty analysis
• random sampling
• normal distribution
• uniform distribution

Acknowledgments

The author would like to thank SAKAI Tomohiro in the Nuclear Regulation Authority of Japan for the problem presentation regarding the neutronics uncertainty analysis due to manufacturing uncertainty, such as the correlation between the manufacturing uncertainties.

Disclosure statement

No potential conflict of interest was reported by the author(s).