Evaluation of in-vessel corium coolability by water penetration into Corium-to-vessel gap and debris

Abstract

In this thesis, studies on corium-to-vessel gap formation, water penetration into the gap, and tool development of gap cooling heat transfer were performed. Based on these studies, transient in-vessel corium coolability and heat load of the vessel were analyzed considering water penetration into the gap and particulate/cracks of fractured crust. Firstly, this study proposed the initial gap formation mechanism based on the Inverse-Leidenfrost effect. Applying lubrication theory with the vapor formation produced in the gap, we proposed two gap-formation models considering the pressure field of the vapor flow under the conditions of the flat and spherical gaps: flat gap model and spherical gap model. With the initial gap formation model, we included the effects of thermal deformation and thermal fracture of the crust for estimating transient gap thickness. It turns out that the current gap formation model much better predicted the experimental gap size data. Secondly, the water penetration rate into the gap is limited by the flooding phenomenon. We proposed four key ideas for the classification and the modeling of flooding in a narrow channel

characteristic length of a narrow channel instead of the equivalent-hydraulic diameter, the concept of a unit-cell, the separated direction of two-phase flow in a narrow channel, and the concept of hyperbolicity breaking/singular points in the 1D two-fluid model. We classified 330 flooding data into two categories depending on the entrance and exit geometry. The RMSEs of the current models for narrow rectangular and narrow large-diameter annuli were 38.25% and 31.10%, respectively. It turned out that the proposed model can apply to various narrow channel-types: a rectangular, an annular, and an annulus with a large diameter. Thirdly, a tool, COCOA-GAP (COre COolability Analysis tool in the GAP) for gap cooling analysis was developed for predicting the thermal behavior of four media (melt, crust, gap, and vessel). We derived a 1D analytic solution for the thermal behavior of the crust layer considering solidification, decay heat, and convective boundary condition. Moreover, we solved the 1D/2D heat conduction of the spherical vessel with the three-regime quenching model: wetted, precursory, and dried regimes. The water front location is identified when the temperature of the wetting front location just reaches the Leidenfrost temperature. In the case in which the penetrating water limited by CCFL may not be completely evaporated, we assumed that water overflowing beyond the wetting front is allowed and the precursory region exists until the overflowing water is completely evaporated. Upstream of the water front, the radiation, and convective heat are transferred from the crust outer surface to the water film flowing down the inner surface of the vessel, while downstream of the water front, the heat is transferred from the crust outer surface to the inner surface of the vessel as well as the overflowing water. Through the extensive validation against the gap cooling experiments with the large-scaled melt mass to 360kg (LMP200 experiments) as well as the small-scaled melt mass of 30, 50 kg of melt (LAVA and ALPHA experiments), we introduced the correction factors which account for the uncertainties of the degree of local contacts between the melt and reactor vessel and the effect of the solidified debris penetration in the gap on CCFL. It turned out that these correction factors were consistently used for the reactor-scale gap cooling analysis. The COCOA-GAP tool was further extended through integrating with tools of the fuel-coolant interaction analysis and water penetration into particulate debris bed. With the unified tool, in-vessel corium coolability of 19 tons corium-TMI-2 reactor vessel was analyzed. As a result, we successfully simulated the gap cooling and 1400K of a hot spot and showed the integrity of the vessel under the TMI-2 accident conditions. Based on the TMI-2 calculation, we quantitatively identified the considerable amount of heat removal by the water penetration while the existing in-vessel corium cooling analysis tools do not consider the water penetration into the gap and the thermally-fractured top crust. It was shown that the heat removal from the upper and lower crust was increased by 2.2 and 1.3 times than the ones predicted with the assumption of no water penetration, respectively. Although the heat removal rate from the bottom crust is less than the one of the top crust, we confirmed that the water penetration into the gap played an essential role in maintaining the thermal integrity of the reactor vessel. Moreover, at the given decay heat level, we proposed the safety limits for maximum corium mass at which the peak temperature of the reactor vessel does not reach its melting point.