Vapor Explosions in a One-Dimensional Large Scale Geometry with Simulant Melts (NUREG/CR-6623)
On this page:
Download complete document
Manuscript Completed: October 1999
Date Published: October 1999
H.S. Park, R. Chapman, M.L. Corradini
University of Wisconsin - Madison
1500 Engineering Drive
Madison, WI 53706
S. Basu, NRC Project Manager
Division of Systems Analysis and Regulatory Effectiveness
Office of Nuclear Regulatory Research
U.S. Nuclear Regulatory Commission
Washington, DC 20555-0001
NRC Job Code W6183
when molten fuel is generated and contacts residual water coolant within the reactor vessel or below in the containment reactor cavity. The experimental objectives for this work were to obtain well-characterized data for the explosion propagation/escalation phases, while systematically investigating the effect of a comprehensive set of initial and boundary conditions on the explosion energetics; i.e., trigger strength, fuel mass, composition and temperature, coolant mass, viscosity and temperature and system constraint. First, a vapor explosion apparatus, WFCI, was developed which allowed for well-characterized explosion data and demonstrated reproducible explosions with a simulant fuel. Second, the explosion energetics was examined as a function of varying initial and boundary conditions for this tin simulant. Finally, the simulant fuel was changed to a molten iron-oxide, which was more prototypic of the actual molten fuel compositions and explosion energetics were reexamined and FCI were found to be quite weak.
With respect to reactor safety issues, this experimental work has quite important safety implications. First, this work has provided clear evidence of the reproducibility of vapor explosion energetics for a controlled set of initial and boundary conditions. This suggests empirically that this phenomenon is predictable if one can establish and control the initial and boundary conditions. Second, the experiments demonstrate that geometric scaling can be properly specified; e.g., a rigid radial constraint for one-dimensional tests is conservative for energetics when compared to full-scale, while the axial constraint scale factor from test to prototype needs to be unity to preserve energetics. Finally and most importantly, the data suggests that once the fuel-coolant initial conditions are within an envelope for triggered events, the energetics is much less than thermodynamic, apparently due to the small amount of fuel that participates in an explosion timescale. And this envelope of triggerability is much smaller for a simulant molten oxide with low superheat, such as molten iron-oxide in our tests (corium in the KROTOS tests). This suggests that material scaling for reactor safety issues must preserve the same fuel composition and superheat from the test to the prototype.
The current work has limited data at larger scales with more prototypic molten oxides. It is recommended that further tests could be carried out under these conditions to empirically verify our findings. Models developed from our analysis can also be used to analyze these experiments. Finally, it is known that the mixing conditions determine the envelope of explosivity for the vapor explosion. Thus, it is of fundamental interest to better measure the mixture local conditions just prior to the explosion to correlate with the explosion energetics; i.e., void fraction profiles, fuel volume fractions and mixing diameters. Our future work in vapor explosion research is specifically targeted toward this purpose.