ERMSAR COMET S2 5

Slide1: 

Bertrand SPINDLER, CEA, Grenoble Kresna ATKHEN, EDF, Villeurbanne Michel CRANGA, IRSN, Cadarache Jerzy FOIT, FZK, Karlsruhe Monica GARCIA MARTIN, UPM, Madrid Werner SCHMIDT, AREVA, Erlangen Tuomo SEVON, VTT, Espoo Claus SPENGLER, GRS, Cologne Simulation of Molten Corium Concrete Interaction in a Stratified Configuration: the COMET-L2-L3 Benchmark

Late phase of MCCI: stratified configuration: 

Late phase of MCCI: stratified configuration Molten Corium Concrete Interaction (MCCI) In some scenarios of Severe Accident, corium is assumed to spread over the concrete basemat of the reactor pit Ablation of the concrete occurs, with complex phenomena of thermohydraulics and physico-chemistry: MCCI Investigations concerning MCCI are still going on Late phase of MCCI Decrease of the ablation rate due to the decrease of the residual power the increase of the heat transfer areas Decrease of the gas flow rate issued from the concrete decomposition Decrease of the oxide phase density due to light oxides from the concrete decomposition Stratified configuration is expected Metal phase at the bottom (mainly Fe, Cr, Ni) Oxide phase at the top (mainly UO2, ZrO2, SiO2, CaO, Al2O3)

Late phase of MCCI: stratified configuration: 

Late phase of MCCI: stratified configuration Stratified configuration Main uncertainty: heat transfer between the two layers Consequence: axial and radial ablation rates not well known Few experimental programs The BETA test at FZK with a large test matrix New tests at FZK: COMET-L2 and COMET-L3 The COMET-L2-L3 benchmark COMET-L2 as for post-test simulation COMET-L3 for blind test simulation

The COMET-L2-L3 benchmark: 

The COMET-L2-L3 benchmark Frame SARNET WP 11.2: Molten corium concrete/ceramic interaction Participants CEA, AREVA, EDF, FZK, GRS, IRSN, UPM, VTT Schedule COMET-L2: March to August 2006 COMET-L3: September 2006 to January 2007 Only 1.5 month delay at the end Planned final meeting canceled

COMET-L2, -L3 tests: 

COMET-L2, -L3 tests MCCI tests at FZK Stratified oxide-metal configuration with metal at the bottom Input power in the metal layer (induction heating) No input power in the oxide layer COMET-L2 February 2005 Used for post-test simulations COMET-L3 November 2005 More oxide, higher heat flux, than COMET-L2 Water aspersion after a first period of dry erosion Used for blind simulations

Slide6: 

Scheme of the facility and of the concrete test section COMET-L2, -L3 tests

Slide7: 

Scheme of the thermocouples instrumentation in the plane NW-SE COMET-L2, -L3 tests

Slide8: 

COMET-L2 test 430 kg metal: 90 % Fe, 10 % Ni 35 kg oxide: 56 % Al2O3, 44 % CaO Mean power: 200 kW Initial temperature: 2023 K Power off after 1015 s

Slide9: 

COMET-L3 test 425 kg metal: 90 % Fe, 10 % Ni 211 kg oxide: 56 % Al2O3, 44 % CaO Mean power: 220 kW Initial temperature: 1940 K Top flooding at 800 s Power off after 1878 s

Slide10: 

COMET-L2, -L3 tests Initial period of about 100 s until end of initial overheat, with isotropic ablation Steady state regime with faster axial ablation rate Agreement with the results of the BETA tests COMET-L3: low influence of flooding

COMET-L2, -L3 benchmark: 

COMET-L2, -L3 benchmark Participants and code AREVA with COSACO CEA with TOLBIAC-ICB (base case and modifications) EDF with TOLBIAC-ICB FZK with WECHSL GRS with MEDICIS and with WEX IRSN with MEDICIS (base case and modifications) UPM with MELCOR (COMET-L3 only) VTT with MELCOR Same input data Models depending of the codes

COMET-L2 post test simulations: 

COMET-L2 post test simulations Metal temperature versus time (no measurements for comparison) power off initial period with overheat Large dispersion (150 K), but 6 results between 1750 and 1780 K

COMET-L2 post test simulations: 

COMET-L2 post test simulations Oxide temperature versus time (no measurements for comparison) Large dispersion: 450 K at 1000 s

COMET-L2 post test simulations: 

COMET-L2 post test simulations Axial ablation depth versus time initial period with overheat steady state regime Experiment: no symetry Large dispersion in the initial period Similar ablation rate in the steady state regime Maximum ablation depth underestimated

COMET-L2 post test simulations: 

COMET-L2 post test simulations Radial ablation depth versus time Radial ablation depth overestimated

COMET-L2 post test simulations: 

COMET-L2 post test simulations Final shape of the cavity

COMET-L3 blind simulations: 

COMET-L3 blind simulations initial period with overheat power off Metal temperature versus time (no measurements for comparison) Lower dispersion compared to COMET-L2

COMET-L3 blind simulations: 

COMET-L3 blind simulations Oxide temperature versus time Lower dispersion compared to COMET-L2

COMET-L3 blind simulations: 

COMET-L3 blind simulations Top surface temperature versus time, with measurement top flooding Before flooding dispersion 800 K measurements in between the calculations After flooding dispersion 1500 K only one code at water temperature

COMET-L3 blind simulations: 

COMET-L3 blind simulations Heat flux density from metal to oxide layer versus time Initial phase: positive or negative Steady state before flooding: positive After flooding: positive or negative

COMET-L3 blind simulations: 

COMET-L3 blind simulations Heat flux density at the top surface Flooding effet very different depending on the code top flooding

COMET-L3 blind simulations: 

COMET-L3 blind simulations Cumulated hydrogen production versus time Factor 5 between the final minimum and maximum results

COMET-L3 blind simulations: 

COMET-L3 blind simulations Axial ablation depth versus time Less dispersion than for COMET-L2

COMET-L3 blind simulations: 

COMET-L3 blind simulations Radial ablation depth versus time Overestimation, or in agreement with the measurements

COMET-L3 blind simulations: 

COMET-L3 blind simulations Final shape of the cavity

Overview of the codes and models: 

Overview of the codes and models COSACO by AREVA Crust formation and solidification in the pool for oxide Coupling with CHEMAPP for physico-chemistry Heat transfer with slag layer for metal Isotropic heat flux distribution MEDICIS in ASTEC by IRSN Pool-crust interface temperature between solidus and liquidus IRSN: close to liquidus; GRS: solidus Heat transfer with slag layer Greene correlation for heat transfer between the two layers Multiplying factor for radial heat transfer (IRSN) MELCOR by Sandia National Laboratory Pool-crust interface temperature is solidus temperature Heat transfer with slag layer Greene correlation for heat transfer between the two layers

Overview of the codes and models: 

Overview of the codes and models TOLBIAC-ICB by CEA Phase segregation model with pool-crust interfacial temperature equal to liquidus temperature Coupling with GEMINI code for physico-chemistry Reference: isotropic heat flux distribution Multiplying factor for radial heat transfer (COMET-L2-L3) WECHSL by FZK Heat transfer with film or bubble or transition model Crusts at the interfaces Heat transfer between the two layers with a correlation by Haberstroh and Reinders modified for gas percolation WEX in ASTEC by GRS Modified version of WECHSL Different empirical fitting of the heat transfer models

Summary : 

Summary Large scatter of the code results for the different variables Large scatter for the same code by different users with different models Very different behavior of the heat transfer between the two layers MCCI phenomena still not well understood Results specific to the COMET-L2-L3 configuration ? Consequences of these uncertainties on reactor cases ? Next step for an answer to theses questions new benchmark proposed in the frame of SARNET for reactor cases