Resolution of Generic Safety Issues: Issue 170: Fuel Damage Criteria for High Burnup Fuel (Rev. 2) ( NUREG-0933, Main Report with Supplements 1–34 )
Following the TMI-2 accident, the NRC converted its fuel behavior research program into a severe accident research program and, consequently, no further confirmatory work on fuel damage criteria was pursued. However, some work of this type was conducted by foreign agencies and, in March 1994, the NRC received the results from a reactivity test in the French Cabri test reactor which showed multiple brittle failures at a relatively low energy of 30 calories/gram (cal/gm) in the cladding of a commercial fuel rod with a burnup of 63 gigawatt-days/ton (GWd/t); dispersal of finely divided fuel particles was also observed. Test results from the Japanese Nuclear Safety Research Reactor (NSRR) and the Russian Impulse Graphite Reactor (IGR) also appeared to indicate reduced failure thresholds and fuel dispersal. The corresponding failure threshold used by the NRC for some similar situations is 170 cal/gm and no loss of fuel is assumed below 280 cal/gm.
Licensee requests for higher burnup fuel limits in operating reactors resulted in the issuance of an NRR request1595 for an RES assessment of fuel damage thresholds for reactivity transients. Confirmatory analysis of fuel conditions required RES to request experimental data from the French test.1596 Information Notice 94-641597 was issued and the Commission was informed1598 of the staff's actions. Following staff review of the Cabri test data, an assessment1599 of its safety significance was forwarded to the Commission. High burnup fuel behavior was the subject of Session 11 of the Twenty-Second Water Reactor Safety Information Meeting, the transactions of which were documented in NUREG/CP-0139.1600
Changes in fuel pellets and cladding occur at high burnups that appear to reduce fuel resistance to damage. Two of these changes are: (1) a reduction in cladding ductility that results from neutron damage and internal hydriding associated with oxidation; and (2) the formation of a very fine grain structure in the fuel pellets and the accumulation of microscopic fission gas bubbles on the grain boundaries. While the underlying processes that cause these changes have been known for many years, the extent and effects of these changes were not realized until recently.
Fuel damage criteria for LOCAs are also brought into question at very high burnups. To avoid fragmentation during quenching, 10 CFR 50.46 requires that the peak cladding temperature not exceed 2200F and that the total oxidation of the cladding shall nowhere exceed 0.17 times the total cladding thickness before oxidation. Reduced ductility at high burnup due to internal hydriding may affect the validity of these limits. Further, oxidation during normal operation to burnups above 50 GWd/t has been observed to be on the order of 15% of the cladding thickness, leaving little margin for additional oxidation during an accident.
GDC 10 requires that specified acceptable fuel design limits are not to be exceeded during normal operation, including anticipated operational occurrences. Such limits are specified by applicants and approved by NRC, rather than being prescribed. The limits are intended to keep the fuel rod cladding from leaking, thus protecting its function as a fission product barrier; a commonly used value is a 1% limit on cladding strain. Ductility of cladding from fuel with burnups around 60 GWd/t is found to be reduced by a factor on the order of 5, compared with unirradiated cladding, with observed uniform strains of 1% or less at the time of rupture. It is thus unlikely that a 1% strain limit is providing the protection desired for high burnup fuel. As discussed below, the safety significance of these fuel damage criteria varies depending on the type of event.
Reactivity Transients: These transients can lead to pressure pulses in the coolant, loss of coolable fuel geometry, and releases of radionuclides. The absence of pressure pulses, implied by the present criteria, is probably not affected by high burnups; while fine particulates may be formed, high temperatures (above 300 cal/gm) required to generate pulses should not occur in high burnup fuel by an even wider margin. Fuel dispersal constitutes loss of coolable geometry and may occur with low energy depositions at high burnup; however, the worst reactivity transients are very localized, and localized loss of geometry should not lead to a core-melt. The lower threshold for release of gap activity plus the dispersal of particulate fuel at high burnup would increase plant activity and public dose, and the dispersal of fuel would alter the character of these Chapter 15 transients.
LOCAs: Excessive oxidation of cladding (i.e., more than 17%) on high burnup fuel could, in principle, lead to loss of coolable geometry from hydraulic loads (from the ECCS) without ever experiencing high temperatures during a blowdown. On the other hand, additional oxidation may not occur during the transient regardless of the amount of initial oxidation present on high burnup fuel. Transient temperatures for these accidents are very sensitive to power level, and even marginally lower power levels in high burnup fuel might keep transient temperatures sufficiently low to avoid further oxidation. If either the criteria or the safety analyses fail to provide a margin to loss of coolable geometry, then a core-melt could result from these design basis accidents.
Specified Acceptable Fuel Design Limits: Failure of the 1% strain limit or any other "specified acceptable fuel design limit" to provide the assurance that is assumed would result in plant releases or public doses that are not permitted during normal operation. None of these situations, however, should lead to core damage.
Resolution of this issue could be accomplished by updating the existing burnup-independent criteria to include the effects of burnup, or to develop substitute criteria, as appropriate. Updated criteria could be incorporated in revisions to 10 CFR 50.46, Regulatory Guide 1.77,1594 and SRP11 Section 4.2, as necessary. Implementation of the resolution would require a screening of certain approved licensing topical reports and reloads that were reviewed previously to permit reactor operation to high burnups.
Existing and emerging data, largely from foreign sources, are expected to be adequate for the criteria revisions envisioned. At the time the issue was evaluated in May 1995, NRC programs to obtain and analyze these data were being planned or were in place; no major new testing programs were anticipated. The provisions in 10 CFR 50.46 that are in question were controversial when originally established and changes to this regulation will be avoided unless absolutely necessary; it is possible that the existing criteria can continue to be used as long as careful attention is given to initial oxidation and method of analysis for high burnup fuel.
Significant changes in exposure of plant operating staff are not expected during normal operation. Although the specified acceptable fuel design limits are probably not providing the protection intended, it is believed that licensees are employing more stringent measures that are not derived from the licensing safety analysis, e.g., power maneuvering restrictions and barrier fuel designs are being used to reduce fuel failures, which the 1% strain limit would not prevent. On the other hand, reductions would be expected in exposure of plant operating staff following a major transient or accident. These reductions could be accomplished by changes in operating conditions or fuel designs such that fewer fuel failures would occur during accidents and attendant fuel dispersal would be avoided.
At the time this issue was identified, an action plan for resolving it had been developed by the staff and presented to the Commission. Thus, resolution was planned and in progress and the issue was considered nearly-resolved in January 1995. It was later given a HIGH priority ranking in SECY-98-166.1718 The impact of a license renewal period of 20 years was to be considered in the resolution of the issue.
The staff performed an evaluation of data collected and confirmed that the use of fuel up to the existing limits did not pose safety problems. Confirmatory research with industry cooperation was expected to refine the staff's understanding of issues that may arise from additional increases in burnups. Thus, the issue was RESOLVED with no new requirements.1778