The Belgian Nuclear Research Centre is embarking the construction of MYRRHA, the first accelerator driven research reactor in the world. MYRRHA will demonstrate transmutation of spent nuclear fuel as a promising solution to the problem of radioactive waste disposal and management. MYRRHA will be cooled by the eutectic alloy lead bismuth (LBE).
In LBE coolant, oxidized corrosion products together with lead oxide (PbO) are the most important sources of macroscopic amounts of solid impurities. Simple or complex mixed oxides of corrosion products may form; a prominent example is magnetite (Fe3O4). These solid impurities when are transported along with the coolant, may accumulate in narrow channels and, on the long term, lead to reduced flow and cooling. As opposed to water-cooled nuclear reactors where a loss of coolant is the main accident trigger, these LBE-specific processes may culminate into a so-called loss of flow accident. Physicochemical interactions of corrosion products in the liquid coolant including the mechanisms that govern solids formation and deposition onto surfaces are currently insufficiently understood to allow accurate predictions outside the ranges covered in specific experimental campaigns.
Using an interdisciplinary approach involving specific experimental campaigns and multi-scale CFD modeling, it is the objective of the present PhD thesis to reach a level of scientific insight at which accurate predictions of these phenomena are within reach.
More specifically, the detailed mechanisms of corrosion product oxide nucleation, growth and deposition at the wall should be studied first experimentally, in state-of-the-art lab-scale autoclaves containing LBE and equipped with electrochemical probes for measuring the changes of the dissolved impurities content in LBE (mainly oxygen).The measurements should provide the necessary information to build a mathematical model.
Using the so-called population balance method, the developed nucleation and growth model should be then implemented in computational fluid dynamics (CFD) codes to describe particle-flow interaction in non-isothermal flowing LBE systems. A successful model should be able, for a given corrosion product source term and macroscopic thermal-hydraulic environment, to predict precipitation rates and particle size distributions as well as deposition rates at each location in the MYRRHA reactor, in order to be in a position to reliably assess potential safety issues that may result from these processes.