Characterization of salt deposit layer growth and prediction of cladding temperature of heated rod bundles under long-term seawater injection and pool boiling conditions

Growth and geometry of salt deposit layers were expressed as a product of the shape function and Matsuoka's kinetic law (Matsuoka, 1991) where the growth factor is expressed by the Arrhenius law. Model parameters of the kinetic law and the shape function were quantified based on digitized image data of deposit layers acquired in the abovementioned precipitation tests based on the X-Ray CT scan. Combining with the axial onedimensional two-phase flow model, the effective heat transfer coefficient of the cladding surface was evaluated taking into account growth of deposit layers and two-phase flow regimes. Finally, cladding temperature responses were evaluated based on the energy balance. Final geometry of salt deposit layers was digitized under a wide range of pool boiling conditions based on image data acquired in precipitation tests employing 5 × 5 short length mockup bundles. Based on these data, the shape function of each test case was reduced. It was confirmed that Matsuoka's kinetic law could predict a total mass increase of deposit layers with a good accuracy by specifying a constant in the Arrhenius law as nearly equivalent values for all test cases. From a comparison of measured and predicted surface temperatures at the exit of a heated length of the center rod, it can be interpreted that temperature increased gradually after the salt concentration reached the saturation point, then temperature began to increase steeply when a gap formed by neighboring rods was closed and isolated chimney flow paths surrounding the center heater rod were formed. Another important finding is that there was a certain margin of time until chimneys were formed after reaching the saturated salt concentration, which indicates feasibility of a practical fast running evaluation model based on decay heat, average salt concentration and boundary conditions that can be estimated to a certain degree of confidence level even under accident conditions.