1. 1 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France About the active role played by the UO 2 oxidation on irradiated fuel collapse temperature M. Barrachin (IRSN) P.Y. Chevalier, B. Cheynet, E. Fischer (THERMODATA/INPG/CNRS)

2. 2 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Background (1)  In case of a severe accident, one challenge of the safety analysis is to evaluate the amount of materials in the lower plenum and the composition of the molten pool.  This depends on the degradation scenarios, in particular on the interaction between UO 2 fuel and Zry cladding.  Large spectrum of conditions : temperatures between 900 and 2800 K and atmosphere from highly oxidising to reducing.  Expected different values for the fuel collapse temperature (SA code parameter), function of these different conditions.

3. 3 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Background (2) Two extreme situations  Highly reducing atmosphere Zry extracts oxygen from UO 2 fuel and can dissolve it  Interaction solid-solid (1273-1973 K) (Hofmann-1984)  Interaction liquid Zry-UO 2 solid (> 2023 K) (Olander-1994-96) Expected fuel collapse temperature ~1000 K below fuel melting temperature  Highly oxidising atmosphere (this presentation) Zry quickly oxidised, transformation in ZrO 2 In this situation, no UO 2 fuel reduction  Interaction ZrO 2 solid-UO 2 solid (UO 2 -ZrO 2 phase diagram,Lamberston-1953) Expected fuel collapse temperature > 2800 K, i.e. slightly below fuel melting temperature

4. 4 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Experimental observations VERCORS tests UO 2 irradiated fuel (from 38 GWd/tU to 70 GWd/tU) Complete oxidation of the cladding at low temperature : complete transformation Zr  ZrO 2 After this period, high temperature phase in oxidising atmosphere Detection of the fuel collapse by gamma signal  There was evidence of fuel collapse temperature at about 2500-2600 K It means that the observed fuel collapse temperature is 300 K below than the expected fuel collapse temperature (2800 K) SO WHY ?

5. 5 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Proposed interpretation for the reduction of fuel collapse temperature (1) VERCORS tests (Pontillon et al.-2005) They mentioned the effect of the burn-up (BU), i.e. the UO 2 -Fission Products interactions MATPRO correlation melting Temperature of UO 2 (BU=50 Gwd/tU) reduced by 200 K NOT CONSISTENT WITH THE ANALYSIS OF THE EXPERIMENTAL DATA Experimental data (apparently contradictory) -Christensen (1964) melting temperature decrease with BU -Yamanouchi (1970) melting T. of UO 2 (30 Gwd/tU) = melting T. UO 2 (fresh) Christensen technique = tungsten crucible, possible interaction, composition change during the measurement  unsignificant up to 50 GWd/tU (on the basis of the Yamanouchi’s results)

6. 6 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Proposed interpretation for the reduction of fuel collapse temperature (2) VERCORS tests Effect of non fully oxidised cladding during the low temperature plateau  could be a satisfactory explanation Difficult to go further in the VERCORS test analysis : no PIE examination  only qualitative interpretation.  Another explanation is possible on the basis of the PHEBUS FPT1 post-mortem examinations BUT No consistent with the VERCORS RT4 test observation : UO 2 -ZrO 2 debris bed initial configuration, It means no cladding pre-oxidation, highly oxidising atmosphere,  fuel collapse temperature at 2500 K.

7. 7 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France FPT1 PHEBUS test Main events  Fuel bundle : 1 m high, 18 irradiated rods (23 GWd/t), 2 instrumented fresh fuel rods, 2 zircaloy grids (0.24 and 0.76 m), SIC control rod (steel clad)  Oxidising atmosphere (P=2.2 atmospheres)  Main degradation events :  Cladding burst ~5600-5800 s at T  1100 K (inner rods)  Rupture of control rod ~9690 s at T=1623 K (steel-Zr interaction) : RELOCATION OF METALLIC MATERIALS TOWARDS THE BOTTOM OF THE TEST SECTION  Oxidation period (11060 s-13200 s) : LARGE RELOCATIONS OF FUEL MATERIALS DUE TO THE INTERACTION BETWEEN MOLTEN ZIRCALOY CLADDING/UO2  High temperature period (11060 s-13200 s) : PROGRESSIVE RELOCATION OF UO 2 /ZrO 2 MIXTURES (as in VERCORS TESTS) from 15380 s for 0.4-0.6 m elevation  Complete formation of the molten pool at 16900 s between 0.16 and 0.23 m elevations

8. 8 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France FPT1 PHEBUS test Post-mortem examinations 607 mm 473 mm 1 m The composition of irradiated fuel remnant was measured after the test (U 0.86 Zr 0.12 Fe 0.01 )O 2.42 Interaction Fuel/Cladding Interaction Fuel/steam (U 0.99 Zr 0.01 )O 2.23 Interaction Fuel/steam Impact of the fuel oxidation on the fuel collapse temperature ? PROGRESSIVE RELOCATION OF UO 2 /ZrO 2 MIXTURES FROM 0.4-0.6 m

9. 9 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France FPT1 PHEBUS test Fuel oxidation experimental evidences  Measurements : oxygen contents measured by EPMA : not reliable  Molybdenum FP absent in the 5-metal FP precipitates (Mo-Ru-Tc-Rh-Pd)  oxidation of the molybdenum during the test

10. 10 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France FPT1 PHEBUS test Fuel oxidation calculations Thermodynamic evaluations Kinetic approach (Dubourg-2005) TMI-2 post mortem analyses : O/M=2.14 (Bottomley-1989) O/M=2.08 O/M=2.11 PO2 (2673 K, 2.2 bars)

11. 11 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Impact of fuel oxidation on melting UO 2 temperature O-U phase diagram (atmospheric pressure) Liquid FCC G hyperstoichiometric hypostoichiometric U3O8U3O8 U 3 O 8 UO 2+x +G UO 2+x +G Liquid -Transition fixed by the shape of the liquidus of UO 2+x at high temperature -Up to very recently, only experimental data of Latta was available (1970), W contamination. -Different published values 3077 K (Chevalier-2002) 2873 K (Roth-1981) 2700 K (Guéneau-2002)

12. 12 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Impact of fuel oxidation on melting UO 2 temperature New Manara’s data on liquidus-solidus on UO 2+x (2005)  Laser heating enabling fast melting and freezing  Container-less conditions  High pressure to prevent non-congruent Evaporation  Thermal arrest method

13. 13 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France PhasesThermodynamic modelExperimental information Reference liquid phase UO 2+x solid solution (FCC) associate model (U,O,UO 2 ) 3-Sublattice model (U) 1 (O, □) 2 (O, □) 1 Liquidus-solidus RTlnPO 2 = f(O/U) Manara (UO 2+x ) Latta (UO 2-x ) Chevalier-2002 U 3 O 8 UO 3 U 4 O 9 G(T) Cp,  Hf, S Chevalier-2002 Gas phase O,O 2,O 3,UO 3,UO 2,OU,U G(T)COACH databaseCheynet-2002 Impact of fuel oxidation on melting UO 2 temperature New thermodynamic modelling of the U-O phase diagram at high temperature (1)

14. 14 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Impact of fuel oxidation on melting UO 2 temperature New thermodynamic modelling of the U-O phase diagram at high temperature (2) Atmospheric pressure High pressure 1atm : UO 2+x +G Liquid 2694 K 2atm : UO 2+x +G Liquid ~2600 K

15. 15 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Impact of fuel oxidation on fuel collapse temperature Modelling of the U-O-Zr phase diagram at high temperature (1) Rich oxygen part of the U-O-Zr phase diagram 1atm : Liquid at T>2500 K 2atm : Liquid at T>2400 K

16. 16 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Impact of fuel oxidation on interaction UO 2+x /ZrO 2 Modelling of the U-O-Zr phase diagram at high temperature (2)

17. 17 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Impact of fuel oxidation on interaction UO 2+x /ZrO 2 Modelling of the U-O-Zr phase diagram at high temperature (3)

18. 18 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Fuel collapse temperature Come back to the PHEBUS tests CompositionT L calc(K)T S calc(K) (U 0.88 Zr 0.12 )O 2.000 30803020 (U 0.87 Zr 0.12 Fe 0.01 )O 2.00 30602860 Calculation of impact of structurals materials : relatively limited Stoichiometric compositions  In agreement with the experimental measurements of Ronchi (2002) on (U,Zr)O 2 (23Gwd/tU)  In agreement with the experimental data of Uetsuka (1993) on (U,Zr,Fe)O 2 (TMI2 core simulating material)

19. 19 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Fuel collapse temperature Come back to the PHEBUS tests Calculation of impact of oxidation : significant Hyperstoichiometric compositions CompositionT L calc(K)T S calc(K) (U 0.88 Zr 0.12 )O 2.08-2.11 2980-29602760-2660 (U 0.87 Zr 0.12 Fe 0.01 ) O 2.08-2.11 2960-29202560-2460

20. 20 International VERCORS Seminar, October 15-16th, 2007 – Gréoux les Bains, France Conclusions  Experimental evidences of fuel collapse temperature at 2500-2600 K in oxidising conditions (PHEBUS FP FPT0 and FPT1 tests, VERCORS tests).  New experimental data on liquidus/solidus on UO 2+x was produced by Manara (ITU), more precise than the past one by Latta.  New thermodynamic modelling of U-O phase diagram, and U-O-Zr phase diagram taking into account these new data.  On the basis of this new modelling, the oxidation of fuel could quantitatively explain the observed low fuel collapse temperatures.  Evidence of lower fuel collapse temperature in oxidising conditions (VERCORS HT2) than in reducing conditions (VERCORS HT1 and HT3).