Axial Offset Anomaly Modeling Studies on EDF Cores

Abstract

The Axial Offset Anomaly phenomenon, commonly called AOA, is one of the possible consequences of the undesirable presence of deposits on the nuclear fuel. AOA appears in PWR cores as abnormal distribution of the power, as compared to the design reference values. If the amplitude of the phenomenon becomes significant, it may lead to additional constraints in operating the reactor. Several factors contribute to the root cause of AOA. The state-of-the art knowledge relies on the fact that deviation should appear when the conditions of temperature, boron and corrosion product concentrations are appropriate to form crud on the fuel surface, that is thick enough to allow precipitation of lithium borates. Then, the neutron capture is locally enhanced by the additional presence of boron, leading to local flux depression and redistribution of the power all over the core. The experience feedback on AOA is rather consequent, mainly in the United States. On the EDF nuclear fleet, excepted two notable cases in the mid 90’s, the phenomenon remained limited to a few, low amplitude observations on the 1300 MW type cores. To perform AOA studies, EDF has developed its own numerical tool based on: • the neutron kinetics COCCINELLE code for power distribution computations, • the thermal-hydraulics THYC code, • the dedicated BOA code to evaluate, at the core scale, crud deposition and boron loading. The EDF software COCCINELLE and THYC are commonly used for core design and safety analysis. For AOA studies, they provide a 3D, best-estimate representation of the clean core. Then the thermal-hydraulic data are used by the BOA code as boundary conditions to determine, in both single-phase flow and sub-cooled nucleate boiling, how and where deposits form on the fuel surface. In the standard approach, the total mass of trapped boron is compared to a threshold defining the AOA risk limit. In a prospective approach, an advanced COCCINELLE, THYC and BOA coupling is proposed to account for the power distribution changes all along the cycle, instead of using the clean core data. Numerical simulations of an AOA cycle of a 900 MW core show that this feedback effect has a visible effect on the final axial-offset prediction.