Numerical modeling of oxygen mass transfer in the MYRRHA system

The control of dissolved oxygen concentration is essential for the use of lead-bismuth eutectic (LBE) as primary coolant for nuclear systems. The control of the oxygen concentration in the LBE coolant of MYRRHA needs to be achieved with a robust oxygen supply system. The supplied oxygen will be consumed by the oxidation of structural materials and to a lesser extent by oxidation of nuclear reaction products. The oxygen concentration in the system can be maintained within a targeted range by balancing the oxygen supply and consumption rates. The careful designing process of a successful oxygen control system requires accurate numerical models for the prediction of oxygen mass transfer and distribution in LBE. Computational Fluid Dynamics (CFD) models of oxygen mass transfer in LBE, based on lead oxide (PbO) dissolution and steels oxidation, have been developed in this work. The CFD models have provided important inputs towards the development of an oxygen control system based on PbO mass exchanger technology. The performances of different mass exchanger configurations, in terms of the mass transfer coefficient, have been successfully predicted by the simulations, which were afterwards validated by experimental measurements performed at the DELTA and CRAFT loops. The validated model was used to develop a Sherwood number correlation for design of innovative mass exchangers for oxygen control in LBE. The model and the correlation gave rise to the efficient design of a new venturi-type PbO MX, which is currently operating as predicted by simulations in the MEXICO loop, at SCK•CEN. The oxygen concentration has been controlled accurately in the MEXICO loop with the aforementioned mass exchanger, without significant occurrence of PbO poisoning by dissolved impurities in LBE. The hydrochemical CFD model has also shown to be an invaluable tool while developing a 19-pin fuel bundle simulation, in which the competing processes of diffusion, convection and surface reaction (oxidation by steels) were solved simultaneously. This simulation provided a full mapping of the oxygen concentration, enabling the identification of critical areas of the core in terms of corrosion. The simulations enabled one also to define the required oxygen concentration in the bulk of LBE to provide enough oxygen at the interface to maintain a stable oxide layer with a minimal dissolution of metal. Furthermore, it is of crucial importance to identify rate-limiting factors and determine representative experimental conditions for the corrosion research programme. In particular, the simulations showed that the local oxygen concentration can vary significantly from the bulk oxygen concentration. A maximum cladding temperature of 460 °C might be acceptable if the inlet bulk oxygen concentration in LBE is higher than 1·10-6 wt.%. The steel's surface should be oxidized prior to start up. After the system has undergone a sufficient time period at 1·10-6 wt. % below 400 °C, the temperature can be increased up to a maximum cladding value of 460 °C. On the other hand, in the case of a bulk oxygen concentration of 1·10-7 wt.%, the maximum cladding temperature should be limited below 400 °C. Pre-oxidation at lower temperature is equally required for this condition.