C.E. Dickerman

Treat sodium loop experiments on performance of unbonded, unirradiated EBR-II Mark I fuel elements

Abstract Description of sodium loop experiments performed in the Transient Reactor Test (TREAT) Facility specifically to provide guidance for analyses on the behavior of an element which either has no sodium thermal bond between fuel and cladding or which, due to some unspecified defect, loses the bond upon reactor startup. Transient heat transfer calculations of the experiments were performed utilizing an idealized model of meltdown in which fuel slumping against the cladding was assumed to occur instantaneously along the length of the pin when the fuel surface at the axial midplane reached 1050°C. The result of TREAT experiments and of the calculations imply that although fuel melting may accompany a loss-of-bond type failure in an EBR-II driver fuel element, the short term consequences are acceptable from the viewpoint of reactor safety. Comparison of results from calculations and experiment indicate that the relatively simple modelling used represents a conservative description of the temperature rise in the cladding following fuel-cladding contact resulting from the fuel melting.

Summary and analysis of treat sodium loop experiments on behavior of single EBR-II Mark I pins under accident conditions

Abstract The study of fast reactor safety is concerned with a number of areas, among which the possibility of core meltdown is of major interest. A number of factors might be involved in such an accident; this article deals with a study, experimentally and by computation, of the mechanisms of fuel failure and associated movement of fuel and coolant. These two phenomena could lead to extension of any initial fuel failure and thus would probably be the factors responsible for a serious meltdown accident. Fuel failure experiments have been conducted in instrumented sodium loops, which, when installed in the TREAT reactor, provide a simulated reactor environment. These loops provide for setting conditions of varying temperature and flow of sodium coolant past the fuel elements under test; and the power burst of the TREAT reactor, into which the loops are inserted for testing, can be adjusted to specified intensities. The results of testing a number of sample fuel elements under various conditions of power level and coolant flow are presented and analyzed here.

Transient behavior of “high-swelling” EBR-11 Mark-1A driver fuel in treat

Abstract Transient studies on “high-swelling” Mark-1A driver fuel indicated that temperatures at clad failure, even with irradiated elements, were not significantly lower than clad failure temperatures for normal low-swelling fuel. These temperatures represent thresholds far in excess of the temperatures conceivable for postulated overpower or loss of coolant conditions in EBR-11. In addition to high-speed color photography and cladding temperature measurements, the test capsule was outfitted with electromagnetic motion transducers to follow pre-failure fuel movements. This instrumentation proved to be a powerful tool, both in the determination of the prompt negative feedback reactivity available from fuel expansion, and the description of clad failure phenomena. The nature of cladding failure was similar to observations on normal fuel. Two possible mechanisms for the release of stress from the restricted expansion of sodium trapped between the fuel andcladding appear to be either (1) motion of fuel axially to break the weak interaction forces between the fuel and cladding, or (2) compression of the highly swollen fuel.

Power pulse meltdown experiments performed in the mark I TREAT sodium loop on clusters of 7 EBR-II Mark I type pins

Abstract Two experiments were performed on fuel failure and post-failure movements of fuel and coolant under power excursions. Test samples consisted of seven-pin clusters in instrumented integral sodium loops inserted into the TREAT reactor. One cluster was run with sodium flow and the other with pump power off (transient-induced flow could occur). Results agreed with the predictions of a simple calculational model of pin failure which had been tested previously in less elaborate experiments. Extensive post-failure movements of molten fuel occurred. Although sharp pressure spikes were recorded their magnitude is within a range that can be accommodated by reasonable engineering designs.

Recent results from treat tests on fuel, cladding and coolant motion☆

Transient in-pile tests in the TREAT reactor are showing dispersive fuel movement and substantiate use of mechanistic modeling of such movements in Liquid Metal Fast Breeder Reactor (LMFBR) safety analyses. Test conditions are selected to mock-up the thermal histories calculated for hypothetical LMFBR accidents, and include use of previously irradiated fuel with various fuel structures. These test data are used in analyzing the consequences of LMFBR fuel failure under a range of hypothetical accident conditions and possible accident scenarios, as well as guiding descriptions of the extended fuel motion leading into the transition phase between the immediate post-accident configuration and the final stable, coolable configuration. Status of these experiments as of Spring 1974 was reviewed in three papers at the 1974 American Nuclear Society topical meeting on fast reactor safety: one on studies of transient overpower accidents (Rothman et al., 1974), one on loss-offlow accident studies directed toward behaviour of irradiated fuel (Deitrich et al., 1974), and one on loss-of-flow accident studies with fresh fuel using direct simulation of key axial LMFBR dimensions (Grolmes et aL, 1974). Since those reports, new data on cladding and fuel movement have been obtained from post-mortem examinations on those tests. Also, additional experiments have been run with the primary aims of clarifying the phenomena producing fuel motion, and of assessing the role of cladding motion as it might limit or otherwise affect fuel movement.