Experimental studies on local fault (LF) accidents in fast breeder reactors have been performed in many countries because LFs have been historically considered as one of the possible causes of severe accidents. Comprehensive and consistent interpretations of in-pile and out-of-pile experiments related to LF were arrived at in this study based on state-of-the-art review and data analysis techniques. Safety margins for a hypothetical local overpower accident, which was evaluated as a LF accident in the licensing document of the construction permit for a prototype fast breeder reactor called Monju, were also studied. Based on comprehensive interpretations of the latest experimental database, including those performed after the permission of Monju construction, it was clarified that the evaluation of the hypothetical local overpower accident in the Monju licensing was sufficiently conservative. Furthermore, it incorporated adequate safety margins in terms of failure thresholds of the fuel pin, molten fuel ejection, fuel sweep-out behavior after molten fuel ejection, and pin-to-pin failure propagation. Moreover, these comprehensive interpretations are valid and applicable to the safety evaluation of LF accidents of other fast breeder reactors with various fuel and core designs.

No abstract

In a fast breeder reactor, a slow power ramp accident could lead to a local melting of the fuel depending on design and assumptions. If we assume cladding failure in addition to the fuel melting, possibility of fuel ejection from the pin and subsequent consequences of such fuel ejection become important for safety consideration. In order to answer this question, the RB1 and RB2 tests in the CABRI-RAFT program (from 1996 to 2002) were performed under the collaboration between the French ‘Institut de Radioprotection et de Suˆrete´ Nucle´aire’ (IRSN) and JNC. These tests were aiming at a study on impact of fuel pin failure under an overpower condition corresponding to a control-rod withdrawal-type incident leading to fuel melting. Using a special technique, combination of through-cladding failure and fuel melting was realized. In the RB1 test, fuel ejection was prevented under a limited fuel melting condition. On the other hand, significant fuel melting was applied in the RB2 test so as to get the fuel ejection, thereby obtaining information on the fuel ejection behavior. Interpretation for these tests through the detailed experimental data evaluation and the analysis with PAPAS-2S and PHYSURAC codes are performed in this study. Through this study, it is indicated that molten fuel ejection can be prevented with the low smear density fuel as far as the fuel melting is not large for a slit-type cladding defect. Fuel ejection becomes possible in the case of significant fuel melting with a very thin solid fuel shell surrounding the molten fuel cavity. However, the rapidness of the fuel ejection with the low smear density fuel is less pronounced when compared with that of the high smear density fuel. It should be noted that the coolant channel basically kept its wetness by liquid sodium in the RB2 test in the early part of post fuel ejection phase. This fact is encouraging us concerning coolability of the ejected fuel. It is also confirmed that there is considerable DN (Delayed Neutron) -precursor release into the coolant flow already before fuel ejection. This result is useful for evaluation of anomaly detection with DN signal observation.

In the CABRI-FAST experimental program, four in-pile tests were performed with slow-power-ramptype transient-overpower conditions (called hereafter as ''slow TOP'') to study transient fuel pin behavior under inadvertent control-rod-withdrawal-type events in liquid-metal-cooled fast breeder reactors. The slow TOP test with a preirradiated solid-pellet fuel pin under a power ramp rate of approximately 3%Po/s was realized as a comparatory test against an existing test in the CABRI-2 program where approximately 1%Po/s was adopted with the same type of fuel pin. In spite of the different power ramp rates, the evaluated fuel thermal conditions at the observed failure time are quite similar. Three slow TOP tests with the preirradiated annular fuel resulted in no pin failure showing a high failure threshold. Based on posttest examination data and a theoretical evaluation, it was concluded that intrapin free spaces, such as central hole, macroscopic cracks, and fuel-cladding gap, effectively mitigated the fuel cladding mechanical interaction. It was also clarified that cavity pressurization became effective only in the case of a very large amount of fuel melting. These CABRI-FAST slow TOP tests, in combination with the existing CABRI and TREAT tests, provided an extended slow TOP test database under various fuel and transient conditions.

In the CABRI-FAST experimental program, four in-pile tests were performed with slow-power-ramptype transient-overpower conditions (called hereafter as ''slow TOP'') to study transient fuel pin behavior under inadvertent control-rod-withdrawal-type events in liquid-metal-cooled fast breeder reactors. The slow TOP test with a preirradiated solid-pellet fuel pin under a power ramp rate of approximately 3%Po/s was realized as a comparatory test against an existing test in the CABRI-2 program where approximately 1%Po/s was adopted with the same type of fuel pin. In spite of the different power ramp rates, the evaluated fuel thermal conditions at the observed failure time are quite similar. Three slow TOP tests with the preirradiated annular fuel resulted in no pin failure showing a high failure threshold. Based on posttest examination data and a theoretical evaluation, it was concluded that intrapin free spaces, such as central hole, macroscopic cracks, and fuel-cladding gap, effectively mitigated the fuel cladding mechanical interaction. It was also clarified that cavity pressurization became effective only in the case of a very large amount of fuel melting. These CABRI-FAST slow TOP tests, in combination with the existing CABRI and TREAT tests, provided an extended slow TOP test database under various fuel and transient conditions.