ERMSAR 2012, Cologne March 21 – 23, 2012 Main results of the ISTC Project #3876 “Thermo- Hydraulics of U-Zr-O Molten Pool under Oxidising Conditions in Multi-Scale Approach (Crucible - Bundle - Reactor Scales)”, (THOMAS) M.S. Veshchunov, V.V. Chudanov, A.E. Aksenova, A.V. Boldyrev, V.A. Pervichko, V.E. Shestak Nuclear Safety Institute (IBRAE) Russian Academy of Sciences

ERMSAR 2012, Cologne March 21 – 23, 2012 ISTC Project THOMAS General Information Duration:October 2008 — December 2011 Leading Institution:IBRAE (Moscow) Collaborators:KIT (Karlsruhe) ITU (Karlsruhe) IRSN (Cadarache) CEA (Cadarache) IVS (Trnava)

ERMSAR 2012, Cologne March 21 – 23, Project Objectives On the base of analysis of available test data from small and large scale experiments, to develop a mechanistic description of U-Zr-O molten pool behaviour in oxidising conditions For this purpose, to carry out a tight coupling of the two advanced numerical tools developed within the previous ISTC Project #2936: the SVECHA physico-chemical (molten pool oxidation) model and the 2D thermo-hydraulic code CONV This will allow extension of thermal hydraulic consideration of oxidised melt from small scales (crucible tests) up to a large scale (reactor pressure vessel), including an intermediate scale corresponding to molten pools in bundle tests

ERMSAR 2012, Cologne March 21 – 23, Calculates: U-Zr-O melt composition and average temperature U-Zr-O melt oxidation and bulk ceramic precipitates formation UO 2 pellet dissolution (Zr,U)O 2 peripheral crust thickness and temperature distribution Melt blockage relocation Validated against: FZK crucible tests CORA (melt relocation) Phebus FPT 0&1 (molten pool oxidation) 2-d Model for Molten Pool Oxidation and Relocation Previous ISTC Project #2936

ERMSAR 2012, Cologne March 21 – 23, d Model for Molten Pool Oxidation Verification against FZK Crucible Tests on ZrO 2 dissolution by molten Zr Corrosion-erosion mechanism: depending on oxygen flux matches at the solid/melt interface, the peripheral oxide layer (crust) can grow (“corrosion”) or dissolve by corium melt (“erosion”).

ERMSAR 2012, Cologne March 21 – 23, Heat flux matches Heat balance Mass flux matches Mass balances 2-d Molten Pool Oxidation Model Melt Boundary layer Oxide crust VS FeO Additional layers Supplied with additional layers (Vessel Steel, FeO) and corresponding flux matches at new interfaces for consideration of vessel wall corrosion kinetics +

ERMSAR 2012, Cologne March 21 – 23, Vessel Steel (VS) Oxidation Model Parabolic correlation from METCOR tests (15 Kh2NMFA vessel steel) under unlimited steam supply “Oxygen starvation” regime: oxygen flux J ox through the crust becomes rate controlling during relatively long period of interactions T=1273 K J ox has to be calculated from the solution of the oxygen diffusion problem in the multi-layered system

ERMSAR 2012, Cologne March 21 – 23, U-Zr-O Corium Melt - Steel Oxidation Model (HTLQ) HTLQ calculates the solid phases thicknesses, temperatures and O fluxes HTLQ has to be coupled with 2-D thermo-hydraulic code through the heat flux and oxygen mass flux at the melt-solid interface C, T r T melt T int T ox TSTS T out COCO C O (t,r) FeO Boundary mesh Heat flux Oxygen flux (U,Zr)O 2 C U, C Zr Melt bulk J dif

ERMSAR 2012, Cologne March 21 – 23, T ox/crust < T eut  1600 K Oxide thickness  Corrosion depth T ox/crust > T eut  1600 K Oxide thickness (  200 μm) << Corrosion depth (  5mm) METCOR observations

ERMSAR 2012, Cologne March 21 – 23, “Flowering” mechanism Low temperaturesHigh temperatures Melt Crust FeO Initial position of Fe boundary Compressive stresses Tensile stresses Cracks and tears Melt Eutectics VS Formation of FeO/crust eutectic melt and its extrusion (or “drainage”) through the crust Melt Crust FeO VS

ERMSAR 2012, Cologne March 21 – 23, METCOR tests Eutectic T  1600 K

ERMSAR 2012, Cologne March 21 – 23, HTLQ Numerical Calculations T S 1600 K Steam VS TSTS Growth of corrosion depth with time

ERMSAR 2012, Cologne March 21 – 23, Comparison with on-line measurements Calculation runsOn-line measurements #1 (Low temperature) Oxide/crust temp.: T S  1200 K Heat flux: F  1 MW/m 2 Corrosion rate: R  0.1 mm/h MCP-2 test (regime 2) Oxide/crust temp.: T S  1220 K Heat flux: F  0.81 MW/m 2 Corrosion rate: R  0.13 mm/h #2 (High temperature) Oxide/crust temp.: T S  1800 K Heat flux: F  1.5 MW/m 2 Corrosion rate: R  5 mm/h MCP-2 test (regime 9) Oxide/crust temp.: T S  1640 K Heat flux: F  1.2 MW/m 2 Corrosion rate: R  5.8 mm/h

ERMSAR 2012, Cologne March 21 – 23, Model predictions Variation of vessel steel wall thickness with time at T S > 1600 K Outer surface temperature: 273 K Heat flux from melt: 2×10 5 W  m -2 Outer surface temperature: 273 K Oxygen flux from melt: 0.5 mole  m -2  s -1

ERMSAR 2012, Cologne March 21 – 23, Modification of CONV2D Initialization of the initial and boundary conditions CONV2d: Base calculation cycle Heat conductivity block (convection + diffusion ) Hydrodynamics block (calculation of the velocities and pressure) Output of the results (2d- temperature fields, heat fluxes, 1d characteristics) Transformation of data to format of the melt-steel oxidation 1-D module melt-steel oxidation 1-D module Transformation of data to CONV2D format Modification of the boundary conditions block for O 2 Development of advection-diffusion block for O 2 Adaptation for two type of geometries: reactor case & experimental facility Adaptation of turbulence model for the O 2 transfer in the reactor case A set of the output parameters was extended - Modified blocks - New blocks

ERMSAR 2012, Cologne March 21 – 23, Steel Vessel Crust Melt LAVA Test Apparatus at FZK PRV CONV2D Adaptation to Different Geometries 90  45  00

ERMSAR 2012, Cologne March 21 – 23, t = 5 s t = 35 s Temperature Calculation Results Small-scale (crucible) Oxygen

ERMSAR 2012, Cologne March 21 – 23, Calculation Results Large-scale (RPV) r = m Initial melt temperature: 2773 K Heat source: 1 MW/m 3 Initial oxygen concentration: 10 5 mole m -3 Oxygen flux to the melt: 1 mole m -2 s -1 Outer wall surface temperature: 100ºC (at interface) (near wall)

ERMSAR 2012, Cologne March 21 – 23, Conclusions (1) The model for U-Zr-O molten pool oxidation (developed within the previous ISTC Project #2936) was upgraded for simultaneous consideration of vessel steel (VS) corrosion by the corium melt; on this base, the new oxidation/corrosion module HTLQ was developed The model allows interpretation of the METCOR tests observations and qualitatively describes VS corrosion kinetics observed in low- and high- temperature regimes The oxygen advection-diffusion block in the thermal-hydraulic code CONV2D was developed; adaptation of the turbulence model of CONV2D for solving the oxygen transport problem in the reactor case was carried out

ERMSAR 2012, Cologne March 21 – 23, Conclusions (2) The oxidation/corrosion module HTLQ was implemented in CONV2D; the coupled code was thoroughly tested and verified, and then applied to simulation of corium retention after melt relocation into RPV Calculation results indicate that in-vessel retention by cooling the outside vessel wall with water might be ineffective, owing to physico- chemical dissolution of solid ceramic crust (at melt/wall interface) that prevents vessel walls from direct physico-chemical attack of the corium melt, and wall thinning (melting through) in the lack of crust up to a few cm This important conclusion suggests further, more thorough investigation of the crust physico-chemical stability under conditions of oxidized corium convection in RPV with residual heat generation in the melt and external wall cooling by water