S. W. Hong

Triggered Steam Explosions with the Corium Melts of Various Compositions in a Two-Dimensional Interaction Vessel in the TROI Facility

Abstract Three triggered steam explosion experiments were performed in the TROI facility with a two-dimensional interaction vessel of 0.6-m diameter. The melt compositions were pure zirconia (ZrO2), 70:30 (UO2:ZrO2 = 70:30 wt%) eutectic corium, and 50:50 noneutectic corium. All tests were performed in a 1.0-m-deep water pool under atmospheric pressure. The water temperature was maintained at room temperature. The melt mass released to the water pool was ~10 kg for each test. The tests with pure zirconia and 70:30 corium resulted in triggered steam explosions, while the test with 50:50 corium did not. However, a weak trace of a steam spike was detected with 50:50 corium with a fairly long delay time (~0.1 s) after an external triggering. The explosion efficiency was estimated from the dynamic load and dynamic pressure. The explosion efficiency was calculated to be 0.1% for zirconia and 0.04% for 70:30 corium. The explosivity of corium material was found to be rather low, compared to the simulant material (alumina, ~3%).

AN INVESTIGATION OF THE PARTICLE SIZE RESPONSES FOR VARIOUS FUEL-COOLANT INTERACTIONS IN THE TROI EXPERIMENTS

The TROI tests were analyzed in view of the particle size responses for various types of fuel-coolant interactions. This can provide an understanding about the relationship among the initial conditions, mixing, and explosion. First, several findings from the TROI experiments were considered. The results of the fuel-coolant interactions (FCIs) were dependent on the composition of the corium, the water depth, and the water area in the TROI experiments. Then, the difference between the explosive FCI and nonexplosive FCI was defined by comparing the final particle size. This analysis indicates that the explosive FCI resulted in a large amount of fine particles and a small amount of big particles. With this, the mixing size of the particles to participate in the steam explosion and the fine particle size produced from the steam explosion could be defined in the TROI test. And then, the parametric effects on the particle size were analyzed using the nonexplosive TROI tests. We note that the explosive test results cannot provide information on the mixing process. This analysis on the particle size response indicates that the explosive system includes large-sized particles to participate in the steam explosion, but the nonexplosive system includes less large-sized particles and more fine-sized particles. These particle size responses should be considered during a reactor safety analysis because the particle size will be an important parameter for explaining a steam explosion occurrence or steam explosion strength.

Results of the triggered steam explosions from the TROI experiment

Triggered steam explosion experiments have been carried out in the TROI facilities to investigate the energetics of the steam explosions. Two types of corium melt were used as a melt. One was eutectic corium at 70:30 wt% (UO2:ZrO2), and the other was corium at 80:20 wt%. The diameter of the water pool was 0.6 m, and the depth was varied from 0.67 to 1.3 m. An external trigger (PETN, 1.0 g) was applied just before contact of the melt and the bottom of the interaction vessel, which is believed to be the time of a possible spontaneous triggering. The external trigger led to triggered steam explosions in all the experiments. In the experiments with 70:30 corium, the maximum recorded dynamic pressure and the dynamic load were 17.0 MPa and 360 kN, respectively. Meanwhile, in the experiment with 80:20 corium, the maximum dynamic pressure and the dynamic load reached 7.7 MPa and 200 kN, respectively. The energetics obtained from the triggered steam explosion tests with 70:30 corium were greater than those from the triggered experiment with 80:20 corium. The strength of a triggered steam explosion was found to depend on the composition of the corium.

On the Fuel and Coolant Interaction Behavior of Partially Oxidized Corium

To simulate a fuel and coolant interaction phenomenon during a postulated severe accident in a nuclear reactor, a series of experiments were performed using a partially oxidized corium, which is a mixture of UO2, ZrO2, Zr, and stainless steel. The composition of the melt was chosen such that a separation of the oxidic liquid from the metallic liquid occurred due to the existence of a miscibility gap. A melting and solidifying experiment and two fuel and coolant interaction experiments to explore the possibility of an energetic steam explosion were performed in the TROI facility. The placement of a metal-rich layer consisting of U, Fe, and ZrO2 beneath the oxidic corium layer due to the existence of a miscibility gap was observed in the melting and solidifying experiment. An energetic steam explosion with a propagation of the dynamic pressure wave was observed in one test out of the two tests. The physical and chemical analyses were performed for the corium particles collected after the experiments. It is shown that U, Zr, and Fe formed a heterogeneous mixture and the morphology was in irregular shape with many pores at nonuniform sizes. In the case of nonenergetic interaction, where the melt temperature was lower than the energetic case, the mean particle size was bigger than that of the energetic case, and the melt-water interaction resulted in a substantial amount of hydrogen gas generation, while the amount of hydrogen gas generation was negligible in the case with an energetic steam explosion.

Hydrogen Mitigation Strategy of the APR1400 Nuclear Power Plant for a Hypothetical Station Blackout Accident

In order to analyze the hydrogen distribution during a hypothetical station blackout accident in the Korean next-generation Advanced Power Reactor 1400 (APR1400) containment, the three-dimensional computational fluid dynamics code GASFLOW was used. The source of the hydrogen and steam for the GASFLOW analysis was obtained from a MAAP calculation. The discharged water, steam, and hydrogen from the pressurizer are released into the water of the in-containment refueling water storage tank (IRWST). Most of the discharged steam is condensed in the IRWST water because of its subcooling, and dry hydrogen is released into the free volume of the IRWST; finally, it goes out to the annular compartment above the IRWST through the vent holes. From the GASFLOW analysis, it was found that the gas mixture in the IRWST becomes quickly nonflammable by oxygen starvation but the hydrogen is accumulated in the annular compartment because of the narrow ventilation gap between the operating deck and containment wall when the igniters installed in the IRWST are not operated. When the igniters installed in the APR1400 were turned on, a short period of burning occurred in the IRWST, and then the flame was extinguished by the oxygen starvation in the IRWST. The unburned hydrogen was released into the annular compartment and went up to the dome because no igniters are installed around the annular compartment in the base design of the APR1400. From this result, it could be concluded that the control of the hydrogen concentration is difficult for the base design. In this study design modifications are proposed and evaluated with GASFLOW in view of the hydrogen mitigation strategy.

Triggered Steam Explosions with Corium Melts of Various Compositions in a Narrow Interaction Vessel in the TROI Facility

Abstract Three triggered steam explosion experiments using corium melts of various compositions were performed in the TROI facility. The interaction vessel was 0.3 m in diameter. The melt compositions were 70:30 (UO2:ZrO2) corium, pure zirconia, and partially oxidized corium (UO2:ZrO2:Zr:SS = 53.91:23.09:12.00:11.00 in weight percent). The test with 70:30 corium was performed with a 0.95-m-deep water pool under an elevated pressure of 0.205 MPa, while the others were performed with a 1.3-m-deep water pool under atmospheric pressure. The water temperature was maintained at room temperature. The melt mass released to the water pool was ˜10 kg for each test. The test with 70:30 corium resulted in a triggered steam explosion, considering the long duration of the dynamic pressure and the large amount of fine debris. The dynamic pressure trace from the steam explosion seemed to be superimposed on that from the external trigger. The test with pure zirconia led to multiple spontaneous steam explosions before any external triggering. The zirconia melt confirmed its explosivity. The spontaneous steam explosion with pure zirconia seems not to be affected by the water depth and diameter of the interaction vessel. The test with partially oxidized corium also resulted in a spontaneous steam explosion before an external triggering. These results are different from the previous TROI tests with 80:20 corium in a narrow interaction vessel of 0.3-m diameter, in which no spontaneous steam explosions occurred. The geometry of the interaction vessel used in these tests does not seem to influence the occurrence of a steam explosion, but the corium composition does affect the triggerability of it.

Fuel-Coolant Interaction Test Results Under Different Cavity Conditions

Abstract Some advanced reactors adapt the in-vessel corium retention concept by cooing the outside wall of the reactor vessel in severe accidents. If a reactor vessel failure happens in this case, the molten corium in the reactor vessel is directly injected into the water in the reactor cavity without the process of a free fall. Experiments using ZrO2 and molten corium to simulate the conditions in which the reactor vessel is fully flooded were recently carried out at the Test for Real cOrium Interaction with water (TROI) experimental facility, and the results are compared with the data produced under conditions in which the reactor vessel is partially flooded. It was observed that the melt front velocity in the water under submerged reactor conditions is much faster than that under partially flooded reactor cavity conditions, and a large bubble was observed at the surface of the mixing zone under submerged reactor conditions. Accordingly, it is estimated that the breakup of the melt jet in the water during the fuel-coolant interaction (FCI) test under submerged reactor conditions would be different than that of the FCI test under partially flooded reactor cavity conditions.

The Triggerability and Explosion Potentials of Reactor Core Melt at Fuel Coolant Interactions

The Test for Real cOrium Interaction with water (TROI) experiments have been performed to reveal unsolved issues of a steam explosion using real core material at the Korea Atomic Energy Research Institute. One of the findings from the TROI experiments is that the results of a fuel coolant interaction (FCI) are strongly dependent on the composition of corium, which is composed of UO2, ZrO2, Zr, and steel. The TROI tests were analyzed in view of a particle size response for various types of fuel coolant explosions. This can provide an understanding about the relationship between an initial condition, the mixing, and the explosion. The particle size distribution data from the TROI tests and a single-particle film boiling model were used for all these analyses. The difference between a quenched FCI and an explosive FCI was defined by comparing the final particle size. This analysis indicates that an explosive FCI resulted in a large amount of fine particles and in a small amount of large-sized particles. With this, the mixing size of the particles that participate in the steam explosion and the fine-particle size produced from a steam explosion can be defined in the TROI test. The particle size distributions of the quenched TROI tests were then considered. We note that the explosive test results cannot provide information on the mixing process. This analysis on the particle size indicates that a self-triggered system includes large-sized particles to participate in a steam explosion, but a non-self-triggered system includes smaller-sized particles and more fine-sized particles. Finally, the explosion potentials of the quenched TROI tests were compared to each other. Thus, the single-particle film boiling model based on the particle size distribution provides the explanation for the explosion behaviors of a variety of melts.