Validation of the Severe Accident Module of the EUCLID/V2 Integral Code on the Basis of Experiments on a Failure of Simulators of Single Fuel Rods and Fuel Assemblies

The development of computer codes for modeling accidents in a reactor unit requires validation of the models built into these codes. In this work, the EUCLID/V2 integrated code developed at IBRAE RAS was validated as applied to the simulation of severe accidents with a failure of the core of liquid-metal cooled fast breeder reactors (LMFBR), against experiments on melting of the cladding of fuel-rod simulators carried out at the Institute of Thermophysics, Siberian Branch, Russian Academy of Sciences (IT SB RAS), and SCARABEE BE + 3 experiments performed at the Commissariat à l’Energie Atomique (CEA) in France. The investigations performed at IT SB RAS included measurements of the cladding surface temperatures without liquid-metal cooling of the fuel-rod simulator, which is typical for accidents involving an instantaneous blockage of the flow section in the fuel assembly (FA) or with loss-of-coolant for type BN-1200M reactor units (RUs). To create such conditions, experiments with fuel rods were carried out in an argon atmosphere at room temperature (25°C) and a pressure of approximately 10 5 Pa, and the surface temperature of the fuel-rod simulator was recorded with a pyrometer. In France, the SCARABEE BE + 3 series experiments were carried out in the SCARABEE reactor to study the consequences of a hypothetical accident with a complete instantaneous blockage of the flow cross-section in a sodium-cooled fast reactor. To determine the effect of uncertainty in the initial data, diversified calculations were made. The validation was done by comparing the predictions with the experimental values of temperatures in the range between 500 to 1800 K (experiments of IT SB RAS). The maximum calculation error did not exceed 200 K. For the experiments in the SCARABEE reactor, it was not greater than 88 K for the fuel-rod claddings and 100 K the coolant. The obtained data will be used to estimate uncertainty in the predictions by the models of severe accidents with thermal destruction of fuel rods in fast reactors.