Cladding Rupture, Swelling & Coolant Blockage as a Result of a Reactor Accident (Generic Letter 79-69)

GL79069  

                                UNITED STATES   
                        NUCLEAR REGULATORY COMMISSION   
                           WASHINGTON, D. C. 20555 

                              December 20, 1970 


ALL POWER REACTOR LICENSEES 


Gentlemen:  

On November 1, 1979, we met with reactor fuel vendors, some plant licensees, 
and other interested parties to discuss recently developed staff views on 
cladding rupture, swelling, and coolant blockage that could result from 
reactor accidents. Based on a preliminary evaluation of the correlations 
being developed, the NRC staff determined that parts of the ECCS models 
might be non-conservative in this area and therefore might not be in 
compliance with Appendix K of 10 CFR 50. We understand from the discussions 
at the meeting, confirmed by vendor letters of November 2, 1979, that 
differences between the present models and our preliminary correlations are 
either small within the limited range of applicability or that these 
differences do not produce large changes in peak cladding temperature. In 
either case, we now understand that such differences should not affect 
compliance with the temperature limit specified in 10 CFR 50.46 for licensed 
operating reactors. On this basis it might be expected that the effect on 
plants under review would also be small, but that effect has not yet been 
determined.

We are transmitting herewith our draft report on cladding rupture 
temperature, strain, and resulting assembly flow blockage; this report 
provides a full discussion of the information we presented at the November 
1, 1979 meeting. Should you desire to comment on this draft report, 
"Cladding Swelling and Rupture Models for LOCA Analysis," please provide us 
with written comments by January 9, 1980. Because we are requesting a 
critique that will help ensure the technical quality of the final report, we 
request that the focus of your comments be technical in nature. You may also 
wish to address the topics for discussion that we have sent to the fuel 
vendors and which are set forth in the enclosure to this letter. We also 
welcome comments and suggestions on clarity so that the final report can be 
made complete and explicit on correlation derivation and use.

                                   Sincerely,



                                   Darrell G. Eisenhut, Acting Director
                                   Division of Operating Reactors
                                   Office of Nuclear Reactor Regulation

Enclosures:
1.Draft Report
2.Topics for Discussion
.

                            Topics for Discussion

1.   Confirm that the Zircaloy cladding models displayed in Section 4.0 and 
     which are referenced in Section 5.0 are the models that are used in 
     your licensing LOCA analyses. Confirm that your models have been 
     displayed accurately (i.e., to within +/-5%). If you are unable to 
     respond affirmatively to the above requests, provide the appropriate 
     references and describe the discrepancies.

2.   The location, magnitude, and shapes of superplastic strain peaks and 
     low-ductility valleys cannot be determined precisely from prototypical 
     rod burst tests because there are too few such experiments with enough 
     controlled variables. Do you have any information that would suggest 
     altering the shapes and magnitudes of the strain and blockage 
     correlation curves?

3.   Most of the recent (since 1974) prototypical data were supported by 
     public funds and are publically available. It therefore appears 
     practical and beneficial to develop standardized rupture temperature, 
     strain, and blockage curves. The curves in the report (or modifications 
     that we might make) could serve as an interim licensing standard, and 
     an industry standards committee could develop revised curves based on 
     present and future research results.

4.   It may be appropriate to require that approved vendor cladding models 
     be revised to conform with the correlations that will appear in the 
     final version of the report. If your present models are in agreement 
     with, or conservatively overpredict, the NRC correlations over the 
     range of temperature and stress of interest, and if you wish not to 
     change your present curves your ECCS model revision could simply 
     consist of explicit limits on the range of applicability of your 
     correlations.

5.   The alpha-plus-beta strain and blockage "valleys" portray a real 
     phenomenon, but the exact location of the very steep sides of the 
     valley may be unknowable for real LOCA conditions. Sensitivity analyses 
     could be done to account for uncertainties in the location of the 
     curves and in prediction of the rupture temperature and stress, but 
     this would have the effect of narrowing the allowable calculated valley 
     and creating a pseudo singularity in the analysis. It might be better 
     for the licensing analysis to be insensitive to this feature.

6.   The on-going NRC research program has produced data over, a wide range 
     of conditions. Based on discussions with those performing licensing 
     LOCA analyses, it appears that the actual range of interest may be 
     quite narrow, and that the future program could be beneficially focused 
     on a narrower range.