Lawrence Berkeley National Laboratory Lawrence Berkeley National Laboratory

Transcription

1 Lawrence Berkeley National Laboratory Lawrence Berkeley National Laboratory Title CHEMICAL ASPECTS OF PELLET-CLADDING INTERACTION IN LIGHT WATER REACTOR FUEL ELEMENTS Permalink Author Olander, D.R. Publication Date escholarship.org Powered by the California Digital Library University of California

2 LBL DE CHEMICAL ASPECTS OF PELLET-CLADDING INTERACTION IN LiaiT-WATER-REACTOR FUEL ELEMENTS by D. R. Olander Materials and Molecular Research Division of the Lawrence Berkeley Laboratory and the Department of Nuclear Engineering University of California Berkeley Berkeley, California This work was supported by the Director, Office of Energy Research, Office of Basic Energy Sciences, Materials Sciences Division of the U.S. Department of Energy under contract # W-7405-ENG-48 and by the Office of Nuclear Waste Isolation.

3 In contrast to the extensive literature on the mechanical aspects of pellet-cladding interaction(pci) in light water reactor fuel elements, the chemical features of this phenomenon are so poorly understood that there is still disagreement concerning the chemical agent responsible. Since the earliest work by Roseribaum, Davies and Pon(l), laboratory and in-reactor experiments designed to elucidate the mechanism of PCI fuel rod failures have concentrated almost exclusively on iodine. The assumption that this is the responsible chemical agent is contained in models of PCI which have been constructed for incorporation into fuel performance codes (2). The evidence implicating iodine is circumstantial, being based primarily upon the volatility and significant fission yield of this element and on tha microstructural similarity of the failed Zircaloy specimens exposed to iodine in laboratory stress corrosion cracking (SCC) tests to cladding failures by PCI(2-6). The fission yield of cesium is many times larger than that of iodine and because of the great stability of the compound Csl, essentially all of the iodine inside an intact fuel rod is contained in this form. However, in laboratory tests, Csl is not an SCC agent for Zircaloy. In order to demonstrate that elemental iodine is responsible for Zircaloy SCC inside a fuel pin, one of the following conditions needs to be met. First, although Csl is the dominant iodine-bearing species in irradiated fuel, a very small partial pressure of elemental iodine is predicted thermodynamically. Cubiocciotti et. al. determined that Csl reaction with the fuel produces an iodine pressure of the -1-

4 -19 " order of 10 atm(7), Gotzmann calculates an iodine pressure of 10" atm at 600 K for the reaction of cesium iodide with fission product molybdenum in stoichiometric urania(8). If iodine is the cause of fuel rod PCI failures, it must be active as an SCC agent to Zircaloy at extremely low pressure. Using kinetic Aeory to calculate the impingement rate of a gas at 10" atm on a surface, assuming unit sticking probability and estimating the projected area 02 of an iodine atom to be 12 A, three months would be required for a monolayer of iodine to form on the inner surface of the cladding of a fuel element. This is a formidable kinetic limitation to promotion of Zircaloy SCC by iodine at a pressure lower than that achievable in any ultra high vacuum system. Alternatively, it is possible for iodine to be transported in the gas phase in a cladding crack by the Csl/Cs couple, in much the same way that oxygen can be redistributed in mixed oxide fuel by CCU/CO or H-O/H, gas couples. However, when examined quantitatively (9), this mechanism founders on the very small pressure of Cs(g) which is thermodynamically permitted to coexist with 10 -Q atm 1(g) and 10 ' atm Csl(g) (the vapor pressure at 600 K) while satisfying the equilibrium Cs(g) + 1(g) = Csl(g) (K = 3 x atm" 1 at 600 K). These considerations lead to the question of the experimentallydetermined minimum iodine pressure for Zircaloy SCC. unfortunately, the iodine pressure is rarely measured; most experimental studies have simply exposed metal specimens to known quantities of iodine in a mechanical device for producing the desired stress (e.g. a closed, -2-

5 argon-pressurized tube). This has led to usage of the unfortunate and thermodynamically meaningless iodine concentration unit of mg/cm. Given the great excess of zirconium over iodine in a typical tubeburst test, the most likely equilibrium state (if it were achieved) is the two solid phase system Zr + Zrl, which, coincidentally, supports an equilibrium iodine pressure of x. 10 atm. However, there is no way of proving that equilibrium has been achieved inside a sealed Zircaloy tube containing iodine, or that kinetic limitations permit gettering the gas to this level of pressure. All that can be said of the iodine pressure in sealed, tube tests is that it lies between the initial pressure of iodine (typically 10 atm) and the equilibrium value of 10 atm. In the only experiment in which Zircaloy SCC has been induced in the presence of a controlled iodine pressure, the minimum required for failure was found to be 10 atm(10). The second possibility for justifying the role of iodine in PCI is based on the observation of iodine release from gamma-irradiated cesium iodide crystals(ll). From this fact, radiolysis of cesium iodide has been proposed as a mechanism of liberating sufficient elemental iodine to produce stress corrosion cracking of the cladding of light water reactor fuel elements(7). However, attempts to verify this hypothesis by in-reactor and laboratory tests have been unsuccessful(10,12-14). Finally, elemental iodine could be responsible for PCI if, despite the thermodynamic imperative, Cs and I formed separately by fission are kinetically restricted in forming Csl. Such behavior appears -3-

6 to occur in fuel which has been artificially doped with cesium and iodine(ls) but experiments on irradiated fuel rods suggest release as a nonvolatile iodide(16). However, there have been no experiments on irradiated U0 2 in which the released fission product mixture was mass-analyzed to detennine the chemical form of iodine. If, as seems likely, iodine is not responsible for cladding SCC, many other fission products cause Itl-type failures in Zircaloy. Of these, cadmium is a promising candidate(10, 17). Contrary to the controversy concerning the chemistry of iodine inside a fuel pin, the elemental state of cadmium is indisputable. This element is often dismissed as a possible cause of PCI failures because of its low fission yield (^ 0.1%). However, it is a st rai^ht-forward matter to show that if all of the fission product cadmium in a 1 cm ID. fuel pin with 90 i smeared density was released to the void space, the cadmium pressure would attain the vapor pressure at 623 K fo.25 Torr) after a burnup of 0.15 MlVd/tnnne. Even if a substantia] fraction of the cadmium is not released from the fuel, there clearly is enough of this element to exceed the minimum pressure for SCC f^ 10 arm, rcf. 1.0), particularly for fuel burnups of tens of thousands of MKd/tonne. Acknowledgment This work was supported by the Director, Office of Energy Research, Office of Basic Energy Sciences, Materials Sciences Division of the U.S. Department of Energy under contract 1 W-"?.105-r-f-.'C.-48 and by the Office of Nuclear Waste Isolation. -4-

7 REFERENCES 1. H. S. Roseribaum, J. H. Davies and J. 0. Pon, "Interaction of Iodine with Zircaloy-2", GEAP (1966). 2. A. K. Miller, H. Ocken and A. Tasooji, J. Nucl. Mater. 99_, 254 (1981). 3. C. C. Busby, R. P. Tucker, and J. E. McCauley, J. Nucl. Mater., 55_ (1975) M. Peehs, H. Stehle, and E. Steinberg, "Out-of-Pile Testing of Iodine Stress Corrosion Cracking in Zircaloy Tubing in Relation to the Pellet-Cladding Interaction Phenomenon", ASTM STP-631, (1979) B. Cox and J. C. Wood, "Iodine Induced Cracking of Zircaloy Fuel Cladding", Proc. Conf. of the Corrosion Division of the Electrochemical Soc., Corrosion Problems in Energy Conversion and Generation, C. S. Tedman, Ed. ( F. Garzarolli, R. von Jan and H. Stehle, At. Energy Rev. 17 (1979) D. Cubicciotti, R. L. Jones, and B. C. Syrett, "Chemical Aspects of Iodine-Induced Stress Corrosion Cracking of Zircaloy," ASTM Filth Conference on "Zirconium in Industry," ii Gotzmann, "Evolution of Iodine Potentials in LWR Fuel Pods Relevant to PCI Failures", paper presented at the ANS/ENS Topical Meeting on Reactor Safety Aspects of Fuel Behavior, Sun Valley, Idaho, August 2-6 (1981).

8 9. D. R. Olander, "Is PCI an Iodine SCC Phenomenon?", submitted to.1. Nucl. Mater. 10. S. 11. Shann and D. R. Olander, "Stress Corrosion Cracking of Zircoloy by Cadmium, Iodine and Metal Iodides", submitted to J. Nucl. Mater. 11. D. Cubicciotti and J. H. Davis, Nucl. Sci. Eng., 60_ (1976) J. H. Davies, F. T. Frydenbo and M. C. Adamson, J. Nucl. Mater. 80 (1979) J. H. Davies, J. S. Armijo, and R. A. Proebstle, "In Pile Fuel Stress Corrosion Tests," American Ceramic Society Meeting, San Francisco, November J. C. Wood an"3 D. C. Hardy, "Characteristics and Interpretation of Power Ramping Defects," AMS Topical Meeting on Light Water Fuel Performance, M. Peeh;:, R. Manzel, W. Schweighofer, W. Haas,!;. Haas and R. Wurtz, J. Nucl. Mater., 97 (1981) D. 0. Campbell and A. P. Malinauskas, Nucl. Technol. 53 (1981)