Quenching of Melt Layers by Bottom Injection of Water in the COMET Core-Catcher Concept

1. Quenching of Melt Layers by Bottom Injection of Water in the COMET Core-Catcher Concept J. J. Foit, M. Bürger, Ch. Journeau , H. Alsmeyer, W. Tromm Forschungszentrum Karlsruhe IKET Karlsruhe, Germany

2. Outline: • Introduction • Limitations by heat conduction • Cooling by bottom flooding: Experiments and models • Conclusions

3. 1. Introduction: The Problem of Coolability Requirements for ex-vessel melt cooling • No penetration of basement • Safe inclusion of corium and fission products • Safe extraction of decay heat, short- and long-term To be considered: different failure modes of RPV, various melt compositions, and release in different pours Exclusion of high pressure RPV failure by depressurization of primary system: Collection of melt in the reactor cavityor in the spreading compartment

4. Characteristics of ex-vessel corium melts • Up to: 120 t oxide (UO2, ZrO2, …) 80 t metal (steel, Zr, …) • Melt heights: • up to 90 cm in reactor cavity of 6 m  • some 40 cm for larger spreading areas (mainly future power plants) • Initial temperatures up to 3000 K • Decay heat in the melt: • 30 MW initially • 10 MW after days

5. Criteria for ex-vessel melt cooling “European Utility Requirements” on melt stabilisation for future reactors: “…The Designer shall prove that corium would relocate in the reactor cavity in a coolable configuration. As an alternative the Designer may include a Corium Collecting and Cooling Device (CCCD), usually called a core-catcher, provided with a corium cooling system. This system shall not have any active component inside the containment.“

6. Different Options for Melt Cooling under Consideration • Indirect cooling: • Externally cooled crucible, possibly with dilution of melt by sacrificial material Direct water contact: • Wet cavity • Top flooding • Bottom flooding • How far can requirements be fulfilled? • What is applicable in present / future reactors?

7. 2. Limitations by Heat Conduction Typical values for NPP:  qVol = 0.8 MW/m3 (2 h after start of accident) k = 2 W/ (m . K) (oxidic melt) Heat conduction • Corium Cooling from Top and Bottom • Tmax = 2800 K • T1 = T2 = 400 K h  22 cm Height for complete solidification parabolic temperature profile Tmax – T1 = qVol / 2 k  h2 For complete solidification: • Corium Cooling from Top • T2 < 1600 K to avoid concrete ablation • qdown 0 h  8 cm Height to exclude concrete erosion

8. Consequences for Corium Cooling • Low heat conductivity of oxide materials requires shallow layers to extract decay heat by conduction only. • Dilution of melt by sacrificial material or concrete reduces power density, but increases melt height and. Net consequence is reduced coolability. • Cooling under reactor typical conditions requires fragmentation of melt to reduce compact melt layers, if complete solidification shall be achieved. Otherwise long persistence of partly liquid melts.

9. 3. The COMET Concept Based on fragmentation of corium and porosity formation: • After erosion of a sacrificial concrete layer, the melt is passively flooded from the bottom by injection of coolant water. Sacrificial concrete layer allows melt accumulation, initial cooling and preconditioning (solidification range, viscosity) of the melt for subsequent flooding. Advantages: • fast cool-down and complete solidification of the melt. • porous structure of the melt allows a safe, permanent long term cooling. • The structures in the lower containment and the basement remain cold and intact. Disadvantage: • Fast release of steam during the quenching process which results in a steam pressurisation of the containment (condensation would subsequently reduce the steam pressure).

10. 3.1 COMET Variants a) The first variant uses an array of plastic tubes, embedded in a horizontal concrete layer (Fig. 1); connected to a water reservoir. b) The second variant uses a layer of porous, water filled concrete (CometPCA) from which flow channels protrude into the layer of sacrificial concrete (Fig. 2). The porosity of the concrete can be adjusted to yield an appropriate coolant water flow into the melt. The water-filled porous concrete layer provides high resistance against downward melt attack by. Fig. 1: Water injection through channels. Fig. 2: CometPCA: Water injection from porous concrete layer.

11. COMET Experiments Melt: • Thermite generated melt. Admixture of approx. 35 wt% CaO reduces Tsol of the oxide to about 1670 K. Also the viscosity of the melt is decreased and is comparable with that of a corium melt upon the admixture of the sacrificial concrete. The heavier metal melt is layered below the oxide melt (expected after admixture of major concrete constituents to the UO2/ZrO2 part of corium melt. • The melt is poured onto the cooling facility. Scale: • 92 cm diameter circular section of the large cooling facility in a reactor. • The geometrical heights of the melt, the temperatures of the melt are representative. Long-term heat fluxes that have to be removed from the inductively heated metal melt are in the range of 450 kW/m2 (decay power level shortly after melt release).

12. COMET–H test series Nine experiments were performed under following conditions: • Initial melt temperature of about 1900 °C. • Variation of the water supply pressure and the height of the melt  coolability limit: hmelt ≤ 50cm, p>0.1 bar. • The presence of unoxidised zircalloy. • A possible occurrence of inhomogeneous downward concrete erosion (melting plugs of different length, COMET-H2.1 and H2.2) A ceramic base layer was introduced to improve the stability. Fig. 3: Porously solidified melt from COMET-H 2.2 experiment. • Complete quenching and solidification within approx. 1300s. • Safe cooling, i. e. a long-term removal of the 300 kW decay heat has been achieved.

13. CometPCAexperiments Four performed experiments demonstrated an efficient cooling: • Melts up to 50 cm height can be safely cooled with an overpressure of the coolant water of 0.2 bar. Fig. 4: Heat removal by coolant evaporation in CometPCA-H4 experiment • Complete quenching and solidification within approx. 15 min. • Safe cooling, i. e. a long-term removal of the 300 kW decay (230 W/kg) heat has been achieved.

14. VULCANO VW-U1 experiment • Performed at CEA Cadarache (France) to validate the PCA concept with prototypic corium and simulated decay heat in a D=25 cm COMET cooling device. • Melt: 40 kg melt (45 wt.%UO2, 19.3wt.% ZrO2, 19.6wt.% SiO2, wt.5.3% FeOx, 0.7wt.% CaO and 0.1wt.% Al2O3 ) generated in the VULCANO plasma arc furnace. • Initial temperature above 2000 K. • The inductive heating power varied from 10 to 30 kW (250-750 W/kg). • Erosion of the 1 cm sacrificial concrete layer in 57 s. • The melt was quenched within less than 20 minutes

15. Integral model in WABE code: WABE-COMET • The crucial process for a successful cooling is sufficient breakup of the compact corium layer and the formation of a porous structure. • the underlying mechanisms for porosity formation are very complex as demonstrated by the experimental experience. Model assumptions: • Rapid lateral expansion of the interaction between melt and injected water due to strong local pressure buildup with subsequent strong expansion leads to enhanced fragmentation. • From a laterally extended steam production region, the produced steam may flow upwards through channels • creating an interconnected porosity in the upper region. • Freezing stops both processes. • Local porosity formation rates depend viscosity. Fig. 5: Porosity formation.

16. WABE-COMET application to VULCANO VW-U1 experiment Pre calculations: Comparison of measured and calculated steam mass flow rate and total mass (Fig.6): • Slightly higher final amount of steam of about 18 kg (experiment: 15 kg) released in only 10 min (calculation) compared to 25 min in the experiment. • Possible explanation: non simultaneous opening of the two inlets in contrary to the simulation. • Post-test calculations: • Consider non simultaneous. • different friction in the porous concrete, especially in the inlet region. Fig. 6: Experimental findings and WABE results.

17. Post calculations: Fig. 7: Comparison of measured and calculated steam mass flow rate and total mass. New results with increased friction in porous concrete (simulated by reduced porosity from 0.3 to 0.2).

18. Fig.9: Temperature evolution Fig. 8: Porosity formation.

19. Conclusions The performed series of experiments have shown the high efficiency and reliability of the bottom flooding COMET and CometPCA concepts: (i) Up to 50 cm high oxide plus metal melts are safely arrested and cooled through bottom flooding with 0.2 bar overpressure of the coolant water. (ii) The flooding rate of the coolant water is about 2 kg/(m2s), and results in a transient high cooling rate of some 3 MW/m2 which is about one order of magnitude above the decay power level. (iii) The dominant process for the highly efficient heat removal is the fragmentation of the melt by evaporation of the injected water which creates open porosities and large surfaces for heat transfer from the melt. Based on these results, a reliable design and operation of either COMET or CometPCA cooling facility in nuclear power plants seem to be possible. • Essential processes of porosity formation and quenching appear to be understood and modelled with WABE-COMET. Key features of the model: • A heuristic correlation approach on local porosity formation, which is directly proportional to the local overpressure. • Temperature dependent melt viscosity is included. It hinders porosity formation. • Freezing is taken into account via a temperature criterion. It finally stops porosity formation.