Sacrificial Materials for SFR Severe Accident Mitigation

Proceedings of ICAPP ‘10 San Diego, CA, USA, June 13-17, 2010 Paper 10092

Sacrificial Materials for SFR Severe Accident Mitigation Christophe Journeau1, Kamila Plevacova1, Gérald Rimpault2, Sandra Poumerouly2 1

CEA, DEN, STRI, Cadarache, F-13108 St Paul lez Durance, France CEA, DEN, SPRC, Cadarache, F-13108 St Paul lez Durance, France Tel: +33(0)4 42 25 41 21, Fax: +33 (0)4 42 25 77 88, E-mail:[email protected] 2

Abstract – In case of postulated Sodium Fast Reactor severe accidents, the core could melt and form a mixture called corium. In this event, sacrificial materials could then be used to reduce the heat load to the retention structure and to avoid the criticality risk. This approach was applied in the past with, for instance, depleted uranium oxide used as sacrificial material in the SNR300 ex-vessel core catcher. A review of sacrificial material candidates has been conducted, considering both neutron absorber and diluent materials. This review was initially based on criteria related to thermophysical and chemical thermodynamic properties (melting and boiling temperature, ability to form a liquid solution with molten fuel…). Neutronic calculations have been done for some generic configurations in order to estimate the reactivity decrease due to the mixing of corium with some different materials. Materials such as aluminium oxide, uranium oxide and hafnium or europium oxides will be presented in more details and their relative advantages discussed.

I. INTRODUCTION In the postulated case of a Sodium Fast Reactor severe accident, the core would melt and form a mixture called corium. In the case of material recompaction, an energetic criticality event could lead to the so-called Core Disruptive Accident1. In order to mitigate this risk, even if its probability is very low, several engineered solutions have been proposed: part of the molten core can be rapidly evacuated as e.g. in the FAIDUS concept2 or by diluting the corium3. In the latter case, two options exist: either the corium is mixed with a (relatively large) amount of diluent material, or it is mixed with a smaller quantity of neutron absorbers, in order to achieve the same loss of reactivity. Such sacrificial material could be mixed with the molten fuel during the primary or transition phase, as in some designs proposed for CAPRA3 or in the late phases in a core catcher, as for the uranium dioxide installed in SNR3004 core retention device. In this last case, sacrificial material shall have both a role with respect to criticality prevention and a role as a thermal barrier. In this paper we propose the general characteristics that sacrificial materials should verify for Sodium Fast Reactors, as it had been described elsewhere for VVER 1000 core-catcher5 or EPR™ temporary retention in reactor pit6. Then some of the candidate materials will be described and neutronic calculations will be shown in order

to assess the requested amount of material to be mixed with the molten core to guarantee criticality-free scenarios. II. EXPECTED MATERIAL CHARACTERISTICS A sacrificial material must fulfill a large number of criteria, the importance of which may vary depending on the chosen severe accident mitigation strategy (introduction in the core region, in an in-vessel core catcher, in an exvessel core catcher, …). During normal reactor operation, • The material shall remain chemically stable and shall only be slightly activated by the neutron flux (depending on its storage position); • The material shall be compatible with coolant material (sodium); During a severe accident sequence, • the mixing of sacrificial material with corium shall guarantee subcriticality (if this mitigation strategy is chosen); • the mixing of sacrificial material with corium must remain stable, therefore formation of a liquid solution (without miscibility gap) between sacrificial and fissile materials is a preferred option;


•

the presence of sacrificial material on a core catcher shall provide a protection against jet impact and guarantee a grace delay before decay heat removal is needed – for instance, an endothermic reaction of corium with sacrificial material would slow the corium progression, whereas it would be detrimental to a rapid mixing; • the sacrificial material shall reduce the thermal load on the core catcher and its support; • the sacrificial material shall remain compatible with sodium up to its boiling point (~900°C); • the presence of the sacrificial material shall not have any adverse effect on the release of (low volatile) fission products. In the long-term post-accidental phases, the mixture of corium and sacrificial material shall form a stable solid having satisfactory mechanical properties. As for any component of an industrial reactor, considerations of cost and availability shall also be taken into account. These requirements can be translated in term of material properties: • The material boiling point shall not be reached during the considered sequences. Either the material is to be used only in a low temperature debris bed configuration and boiling temperatures above 1400°C are sufficient, or liquid corium-sacrificial material pools are to be considered and the boiling temperature shall be above 2800°C. • The material fusion temperature shall be above the maximum expected temperature in the design basis operations. For in-vessel material, a minimum melting temperature of 900°C is suggested. • The time requested to melt the sacrificial material can be assessed in function of the volume heat capacity and the volume heat of fusion, since in practice, the volume available for sacrificial material will be set by design constraints. • A low thermal conductivity will be preferred in order to create a large thermal resistance between corium and structures. • Density relative to molten MOX fuel is an important parameter: a light material will promote convection in a molten layer, whereas denser material shall be less prone to convection. In any case, it is preferred that

sacrificial material is denser than sodium (900 kg.m-3) to prevent the risk of floatability.

III. REVIEW OF SOME CANDIDATE MATERIALS II.A. Absorbing Materials Boron carbide (B4C) is used as absorber in most Fast Reactors. It has been proposed for use in fusible shutdown devices9. Due to its chemical nature, boron carbide will not form a chemical solution with (U,Pu)O2 . Thermodynamic calculations with GEMINI2 and the NUCLEA database10 have indicated that in this system, boron will essentially mix with the metallic phase, potentially reducing some uranium oxide to uranium metal, but that most of the molten fuel will not form a boron-containing solution11. This result has been supported by ongoing experiments. There is therefore the risk of phase separation leading to different relocation of the metallic phase (containing most of boron) and the oxide (containing most of the fissile material) and it will thus be hard to prove that criticality can be excluded in all probable scenarios. This is why oxides, which will form a liquid solution with uranium and plutonium dioxides, will be considered in the remainder of this study. Dalle Donne et al.7 have promoted the use of sodium metaborate: NaBO2 or (Na2OB2O3) for a SFR core catcher. Gou et al. have similarly proposed the use of a lead borate glass, Pb2B2O5 or (2PbOB2O3), as core catcher sacrificial material. Table 1 lists some the properties of these candidate materials. TABLE 1 Some properties of sodium and lead borates (from References 7,

Solid density at room temperature (kg/m3) Expension coefficient (K-1) Melting range(°C) Boiling point (°C) Specific Heat (J.kg-1.K-1)

(Na2OB2O3) 2464 (1950 at Tfusion) 4.40 10-4

(2PbOB2O3) 6700

966 1434 1060

4971-520 1472


The boiling point is rather low, of the same order of magnitude than the melting point of steel. These materials can only be used in core catchers and in configurations in which the corium always remains below 1400°C (thus as a debris bed). At these temperatures, dissolution kinetics are expected to be slow (solubility of UO2 in sodium metaborate has been estimated by Dalle Dionne et al. around 0.03 g/cm².h at 1100°C). Other absorbing oxides have a higher melting and boiling point. For instance, dysprosium titanate (Dy2TiO5), which is used in MIR, RBMK and VVER-1000 V-412 control rods14 has a melting range between 1752 and 1870°C. According to Pannerseelvam and al.15, its density (at room temperature) is of 6900 kg.m-3 (with a coefficient of linear extension between 0.6 and 1.10-5 K-1) and its heat capacity is around 200 J.K-1.mol-1. The eutectic between UO2 and TiO2 lies at a temperature16 of 1645°C, therefore it is expected that dissolution starts at a temperature below 1645°C. Pannerselvam and al.15 have proposed to use gadolinium titanate as a replacement of dysprosium titanate. Risovany and al.14 have suggested to use dysprosium hafnates (DyxHfyO3x+2y) which have a higher density and in which the relative amount of hafnium and dysprosium absorbers can be optimized between 25 and 75 mol%. Maschek and Struwe17 have suggested the use of europium oxide as an accident mitigation material. Table 2 lists some properties of these oxides. TABLE 2 Some properties of Dy2HfO5, Gd2TiO5 and Eu2O3 From references 14, 15, 18, 19 and 20. Solid density at room temperature (103 kg/m3) Melting range(°C) Boiling point (°C) Eutectic with UO2 (°C)

Dy2HfO5 10

Gd2TiO5 ~7

Eu2O3 8

TABLE 3 Redox potentials at room temperature 21 Redox couple Pb2+/Pb Hf4+/Hf Ti4+/Ti Gd3+/Gd Dy3+/Dy Na/Na+ Eu3+/Eu

Potential -0.126 V -1.55 V –1,63 V -2.28 V – 2,29 V -2.71 V -2.8 V

II.B. Diluent Materials The choice of a sacrificial material has been restricted to an oxide, in order to reach a satisfactory solubility between this material and the fissile fuel. In order to have a boiling temperature above 3000°C and a melting temperature below 3000°C, the candidate materials can be restricted to UO2, MgO, ZrO2, HfO2, CeO2, Al2O3, La2O3, Eu2O3, Gd2O3, TiO, Dy2O3, Sc2O3, Cr2O3, BeO, SrO, Y2O3, Nb2O3 and their compounds. The last 3 oxides (SrO, Y2O3, Nb2O3) have been later discarded because of their too small specific and latent heats while BeO has a too high thermal conductivity21 (~100 W.m-1.K-1). Concerning the compatibility with sodium, Singh22 indicates that liquid sodium should be compatible with La2O3, MgO, ZrO2, UO2, Al2O3 and TiO2. The formation free enthalpies21 for HfO2 and Sc2O3 indicate that they also should be compatible with sodium. Depleted uranium oxide and alumina are two major candidate sacrificial materials, if pure oxides are to be considered. TABLE 4 Some properties21, 23 of UO2 and Al2O3

2450-2570 >3000 Unknown

1775-1790 >3000 3000 2115°C

In the absence of data on the stability of these oxides in sodium, redox potentials21 are considered. All the considered oxides have more electronegative than the Na/Na+ couple (at least at room temperature) and should be stable, even if Singh et al.13 have signaled a poor stability of titanium dioxide with sodium at oxygen concentrations below 1 ppb. Europium oxide stability would have also to be verified since the europium couple is more electronegative than the sodium couple.

Solid density at room temperature (kg/m3) Melting point(°C) Boiling point (°C) Eutectic with fuel (°C) Volume Specific Heat (J.m-3.K-1)

UO2 10 960

Al2O3 3 970

2878 4100 (2878)

2072 2980 1915

2.6

3.1

The use of mixed oxide is a potential way to achieve a lower melting temperature, and a lower eutectic temperature for the sacrificial material – fuel system.


For instance, while aluminium and hafnium oxides melt respectively at 2042 and 2850°C, they have a eutectic composition24 (67 mol% Al2O3) melting at 1890°C. Data on some candidate mixed oxides are listed in Table 5. TABLE 5

Reactivity insertion with 20% of sacrificial material mixed with the destroyed core (R= positive reactivity insertion in the absence of diluent)

HfTiO2

Gd2TiO5 CaHfTi2O

7 400

-

7 000

5 600

2070

19802200 -

1790

>1380

2920

2715

7

III. NEUTRONIC CALCULATIONS In order to assess the efficiency of various candidate sacrificial materials, neutronic computations have been performed with the ECCO/ERANOS code system29 in RZ transport using the reference computational route with the ERALIB1 nuclear data library at 1968 energy groups. The so-called SFRv2b30 core has been considered in this calculation. It contains 80 t of fuel and 27 t of steel. In this study unirradiated fuel is considered. It is not expected that the consideration of irradiated fuel would significantly affect its results. This calculation will consider a worst-case scenario: melting of the whole core, sodium voidage, melting of the steel cladding and hexagonal cans, relocation to a stratified configuration with an oxidic layer (fuel) below a steel layer. Due to the change in configuration, the material temperatures are also modified: the fuel is heated from 1227°C in nominal configuration to 2700°C. Let us note R, reactivity insertion computed for this scenario without the adjunction of sacrificial material. Table 6 presents the calculation results for various sacrificial materials (assuming here that 20vol% of sacrificial material has been mixed with the whole fuel inventory and that metal has stratified above the fuel). Europium oxide is the best absorbant material, in accordance with the results of Maschek and Struwe17. Then gadolinium and hafnium oxides are the second best (but with more than 3 times less anti-reactivity insertion). But europium oxide electronegativity ( Table 1) at room

HfO2 Y2O3 La2O3 CeO2 Eu2O3 Gd2O3 depleted UO2

Reactivity insertion - 0.8 R + 0.4 R + 0.3 R + 0.4 R - 2.7 R - 0.8 R + 0.1 R

Ranking 2nd 6th 5th 7th 1st 2nd 4th

The necessary amount of europium oxide to prevent criticality has been estimated at 6.5%. This would correspond to a mass of 4t Eu2O3. For depleted UO2, it appears that 32% (33t corresponding to axial blankets of about 50 cm) would have to be mixed with the core to prevent recriticality. These figures indicate that significant masses of sacrificial materials need to be introduced in the core in order to prevent criticality in the case of total core relocation with segregation of the metallic melt. Hence a combination of partial fuel ejection with the introduction of sacrificial material has been studied since it reduces the volume of the absorber material to introduce in the core. For both europium oxide (Fig. 1) and uranium oxide (Fig. 2), the relationship between the ejected fuel mass and the sacrificial material mass needed to prevent recriticality are affine.

4 Absorbant mass (t)

GdAlO3

2470

TABLE 6

Material

Some properties of mixed oxides From references, 15, 21 and 25.

Solid density at room temperature (kg/m3) Melting point(°C) Volume Specific Heat (J.m-3.K-1)

temperature yields doubt on the compatibility of this oxide with sodium.

3

2 1

0 0

10

20 30 Ejected Fuel mass (t)

40

Fig. 1. Amount of Eu2O3 needed to prevent recriticality in case of partial fuel ejection


ACKNOWLEDGMENTS This work was partly performed within a collaborative agreement between CEA, EDF and AREVA. Part of the presented research benefitted from a financial support from the EURATOM 7th Framework Program [FP7/ 2007-2011] under the grant agreement n° 232658 (CP-ESFR - Collaborative Project on European Sodium fast reactor). This paper represents the views of the authors and not necessarily that of all their partners.

Depleted UO2 assed mass (t)

30

20

10

0

0

10

20 Ejected Fuel (t)

30

REFERENCES

40

Fig. 2 Amount of depleted UO2 needed to prevent recriticality in case of partial fuel ejection In the absence of sacrificial material, 42 t of materials have to be ejected in the assumed configuration. If we assume a 10% ejection of the fuel material, 5.6 vol% of europium oxide, or about 25 vol% of uranium oxide shall be mixed to the remaining fuel for avoiding recriticality. These values are much more reasonable. It confirms that only a combination of provisions would possibly avoid re-criticality.

1.

H. YAMANO, Y. TOBITA, S. FUJITA and W. MASCHEK, “First 3D Calculation of core disruptive accident in a large-scale sodium-cooled fast reactor”, Ann. Nucl. Ener., 36, 3, 337-343 (2009).

2.

Y. TOBITA, H. YAMANO and I. SATO, “Analytical study on the elimination of severe recriticalities in large scale LMFBRs with enhancement of fuel discharge”, Nucl. Eng. Des., 238, 1, 57-65 (2008).

3.

W. MASCHEK, M. FLAD, G. ARNECKE and P. LO PINTO, “Mitigation of core disruptive accident energetics in burner cores”, Progr. Nucl. Ener, 32, 3-4, 639-646 (1998).

4.

H.J. FRIEDRICH, “SNR 300 Tank external core retention device design and philosophy behind it”, Proc. Post Accident heat Removal Information Exchange meeting, SNL report SAND-75-0497 (1975).

5.

V.V. GUSAROV, V.I. AL’MYASHEV, S.V. BESHTA, V.B. KHABENSKII, Yu.P. UDALOV and V.S. GRANOVSKI, “Sacrificial materials for Safety Systems of Nuclear power Plants: A New Class of Functional Materials”, Therm. Eng., 48,9, 721-724 (2001).

6.

M. NIE, Temporary Melt Retention in the reactor Pit of the European pressurized Water Reactor (EPR), Doctoral Thesis, University of Stuttgart, Germany (2005).

7.

M. DALLE DONNE, S. DORNER, G. FIEG, G. SCHUHMACHER and H. WERLE, “Development work for Fast Reactor Core Catchers on the basis of Sodium Borates”, Proc. Int. Mtg. Fast Reactor Safety Technol., Seattle, Wash. (1979).

IV. CONCLUSIONS Depending on the chosen severe accident mitigation strategy, the requirement for sacrificial materials can be determined. In order to guarantee that the sacrificial material will remain in all cases well mixed with the fuel, the use of oxidic materials, forming solutions with uranium and plutonium oxides is recommended. Several candidate materials have been listed: europium, gadolinium and hafnium oxides for the neutron absorbers and aluminum or depleted uranium oxides, for the diluents, have been preselected. Neutronic calculations indicate that significant masses of sacrificial materials must be mixed with the core fuel (from 4 t in case of Eu2O3 to more than 30 t in case of depleted UO2 compared to the 80 t of a typical SFR core) in order to prevent criticality in the case of total core relocation with segregation of the metallic melt. A combination of partial fuel ejection with the introduction of sacrificial material reduces the volume of materials to introduce in the core. Further R&D is necessary to assess the thermochemistry of the candidate materials and their interaction with corium as well as with sodium.


8.

P.F. GOU, C.D. SAWYER, H.A. UPTON and S.S. KHORANA, Core catcher cooling by heat pipe, US Patent 6,353,651 (1999).

9.

G. LEMERCIER, Nuclear Fuel Assembly with an integrated Passive Safety Device, US Patent 5,333,156 (1994).

10. S. BAKARDJIEVA, M. BARRACHIN, S. BECHTA, D. BOTTOMLEY, L. BRISSONNEAU, B. CHEYNET, E. FISCHER, C. JOURNEAU, M. KISELOVA, L. MEZENTSEVA, P. PILUSO and T. WISS, “Improvement of the European thermodynamic database NUCLEA”, Progr. Nucl. Ener., 52, 84–96 (2010). 11. K. PLEVACOVA, C. JOURNEAU and P. PILUSO, “Mixing of molten fuel with absorber materials to prevent recriticality in a postulated SFR severe accident”, FISA’2009, Conference on "EU Research and Training in Reactor Systems", Prague, Czech Republic (2009). 12. M.C. WEINBERG, G.L. SMITH and G.F. NEILSON, “Glass formation and Crystallization of High Lead Content PbO-B2O3 Compositions”, Am. Ceram. Soc. Bull.., 65, 1502-1505 (1986). 13. N. SINGH, K.J. SINGH, K. SINGH, and H. SINGH, “Comparative Study of lead borate and bismuth lead borate glass systems as as gamma radiation shielding materials”, Nucl. Instr. Meth. Phys. Res.I, B 225, 305309 (2004). 14. V.D. RISOVANY, E.E. VARLASHOVA and D.N. SUSLOV, “Dysprosium titanate as absorber material for control rods”, J. Nucl. Mater., 281, 84-89 (2000). 15. G. PANNEERSELVAM, R. VENKATA KRISHNAN, M.P. ANTONY, K. NAGARAJAN, T. VASUDEVAN and P.R. VASUDEVA RAO, “Thermophysical measurements on dysprosium and gadolinium titanates”, J. Nucl. Mater., 327, 220-225 (2004). 16. G. ARTHUR and D. SCOTT, “The effect of TiO2 addition on the sintering of UO2”, Trans. Brit. Ceram. Soc., 63, 8, 417-426 (1964). 17. W. MASCHEK and D. STRUWE, “Accident analyses and passive measures reducing the consequence of a core-melt in CAPRA/CADRA reactor cores”, Nucl. Eng. Des., 202, 311-324 (2000).

18. A. V. SHEVCHENKO, L. M. LOPATO and L. V. NAZARENKO, “Interaction of HfO2 with oxides of samarium, gadolinium, terbium and dysprosium at high temperatures”, Inorg. Mater., 20, 1615-1618 (1984). 19. R. J. BEALS, J. H. HANDWERK and B. J. WRONA, “Behavior of Urania-Rare Earth Oxides at High Temperatures”, J. Am. Ceram. Soc., 52, 578-581 (1969). 20. J. L. WARING and S. J. SCHNEIDER, “Phase Equilibrium Relationships in the system Gd2O3-TiO2” J. Res. Natl. Bur. Stand., A 69, 255-261 (1965). 21. D.R. LIDE, Handbook of Physics and Chemistry, CRC (2001). 22. R.N. SINGH, “Compatibilities of ceramics with Liquid Na and Li”, J. Amer. Ceram. Soc., 59, 3-4, 112-115 (1976). 23. S.M. LANG, F.P. KNUDSEN, C.L. FILLMORE and R.S. ROTH, “High Temperature Reactions of Uranium Oxide with Various Metal Oxides”, Nat. Bur. Stand. Circ. (US), 568, 1-32 (1958). 24. L.M. LOPATO, A.V. SHEVCHENKO and G.I. GERASIMOV, Inorg. Mater., 12, 9, 1331-1334 (1976). 25. P. WU and A.D. PELTON, “Coupled thermodynamicphase diagram assessment of the rare earth oxidealuminium oxide binary systems”, J. Alloys Compounds, 179, 259-287 (1992). 26. J. COUTURES, J.P. COUTURES, H.W. KING, A.M. LEJUS, P. NICHOLSON, P. ODIER and V. LOC, High Temp. – High Pres., 13, 1, 97-103 (1981). 27. I. BARDEZ, P. LOISEAU, D. CAURANT, N. BAFFIER and C. FILLET, “Etude de céramiques et de vitrocéramiques à base de zirconolite-Hf (CaHfTi2O7) destinées au confinement de déchets nucléaires de haute activité”, Proc. Matériaux 2002 (2002). 28. R. STEVENS, B.K. HOM, J. BOERIO-GOATES, B.F. WOODFIELD, R. PUTNAM, J. GUTIEREZ and A. NEVROTSKY, “Molar heat capacities and thermodynamic functions of CaHfTi2O7 (cr) and CaZr0.26Hf0.74Ti2O7 (cr) ”, J. Chem. Thermo., 33, 14411455 (2000).


29. G. RIMPAULT et al, “The ERANOS code and data system for fast reactor and neutronics analysis”, Proceedings of PHYSOR 2002, Seoul, Korea (2002). 30. P. SCIORA et al., “A break even oxide fuel core an innovative French Sodium Cooled Fast Reactor : neutronics studies results ”, Proceedings of GLOBAL 2009, Paris, France (2009).