Implications of Quench Bundle Test Simulations for Plant Applications With MELCOR Code

Abstract

Reflooding of an uncovered and overheated core is one of the most expected measures during the early phase of a severe accident progression. The capability of analytical tools to model correctly processes during independent sub-phases of this accident progression is on different levels, and the most concentrated effort is focused on the study of fuel cladding quenching topic. The main objective of the Forschungszentrum Karlsruhe (Germany) Quench experimental program is to investigate the hydrogen generation during the reflooding of overheated core. The Nuclear Research Institute Rez (hereafter NRI) contributes to this program by the analytical simulation of the Quench bundle experiments with the MELCOR code mainly and ICARE2 code as well since 2000. The NRI analyses covered the Quench-01, Quench-03 and Quench-06 with version MELCOR 1.8.5 (including reflood model), and Quench-01 and Quench-11 tests with the latest version MELCOR 1.8.6. The tests Quench-01 and Quench-06 were characteristic of the lower reflooding onset temperature (1900–2050 K), and the tests Quench-03 and Quench-11 had high reflooding onset temperature (> 2350 K) with the fast heat-up phase before reflooding. The integral code MELCOR is capable to sufficiently predict the heat up and reflooding phases for the tests with the lower onset temperature, but the Quench-03 test with high onset temperature was not predicted correctly starting in the heat-up phase and also the hydrogen generation during bundle reflooding was significantly underpredicted. The interpretation of the Quench-03 test with ICARE2 code confirmed the MELCOR results, but its more detail description enabled to identify a cause. The underestimation of the bundle temperatures is influenced by the shroud insulation behaviour, which could be interpreted as the reduction of heat losses through shroud and simultaneously more intensive heat-up of a bundle. The phenomenon, which causes this behaviour, is not yet known, and is not important for plant applications, although it is relevant for the test interpretation. Also the interpretation of the Quench-11 showed an importance of correct modeling of the bundle shroud, which is in the MELCOR 1.8.6 simulated by newly implemented independent component. The Quench-11 test analysis showed simplified failure criteria for the cases when the shroud is not supported by the former components, which are assumed in the basic MELCOR model fundamentals. Concerning the plant application of the MELCOR code for the scenario with the overheated core reflooding, the user has to give attention on the correct treatment of heat losses through shroud and its degradation modeling, mainly if the shroud supporting formers absent in the core periphery design. Hydrogen generation during reflooding is predicted well for lower onset temperatures (below 2100 K) and slightly underpredicted for higher onset temperatures. It is important to take these code capabilities into account because a reflooding of the reactor core will consist of varied reflooding onset temperatures.