ML20205T335
| ML20205T335 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 06/09/1986 |
| From: | Hukill H GENERAL PUBLIC UTILITIES CORP. |
| To: | Stolz J Office of Nuclear Reactor Regulation |
| References | |
| 5211-86-2103, NUDOCS 8606130111 | |
| Download: ML20205T335 (6) | |
Text
~
e GPU Nuclear gglgf 100 Interpace Parkway Parsippany New Jersey 07054 201 263-6500 TELEX 136-482 Writer's Direct Dial Number:
June 9, 1986 5211-86-2103 Office of Nuclear Reactor Regulation Attn:
J. F. Stolz, Director PWR Projects Directorate No. 6 Division of Licensing U.S. Nuclear Regulatory Comission Washington, DC 20555
Dear Mr. Stolz:
Three Mile Island Nuclear Station Unit 1 (TMI-1)
Operating License No. DPR-50 Docket No. 50-289 Fuel Resintering This letter confirms discussions between NRC and GPU Nuclear personnel on April 22,1986 concerning the status of fuel resinter test results for TMI-l inventory fuel to be used in future cycles beginning with Cycle 6.
Attached is a sumary of the subject discussions.
If any additional information is required, please contact us.
Sincerely, H.D.l_* kill lu
(
Director, TMI-1 DJD:HDH:jh 3428f Attachment cc:
J. Thoma i
R. Conte i
D. Fieno 8606130111 e60609 PDR ADOCK 05000289 P
PDR D)
GPU Nuclear is a part of the General Public Utilities System 1
4 ATTACHMENT 1
BACKGROUND Resinter testing is required to confirm the assumptions used in application of the approv'ed B&W model for predicting in-reactor fuel densification. The tests are usually perfonned on pellet samples selected randomly from each production lot during fabrication. For two batches of TMI fuel (TMI-1, Batch 8 (91F) and TFil-2, Batch 4 (21B)) sampling and testing were not done during pellet manufacture, and rod loading and assembly of the fuel bundles were complete'd without resinter tests. Fortions of both of these batches will be used in Cycle 6 and future cycles.
RESINTER PROGRAM To determine the densification characteristics of the two batches a modified sampling plan was developed to produce resinter ddta frne fuel pellets removed fYom the previously assembled fuel rods.
1 For standard B&W resinter sampling, about twenty pellets are randomly selected from each fuel pellet lot. A pellet lot is the quantity of material that is 1
blended, milled and pelletized together. A fuel reload batch is typically composed of eight to ten pellet lots.
Densific tion variability within a lot i
car < be" quantified by statistically examining the densification behavior of the i
resinter pellets.
I
\\
(.
l l
[
3428f
. Such totally random pellet sampling was not practical for the two TMI fuel batches since the assemblies were already completed.
Thus, a sampling method was implemented that would still provide a reasonable representation of the pellets within the fuel lot.
The sampling plan was as follows:
1.
Randomly select and remove two fuel rods composed of pellets from a given lot.
2.
Randomly select twenty pellets from each rod for standard resinter testing.
(The samples from each rod were kept separate to allow the detennination of the density change variance between the samples).
3.
Perfonn (1) and (2) above for each pellet lot.
This sampling plan yielded two independent fuel pellet sample populations for each lot. The fuel bundle disassembly and reassembly work and the resinter tests were done at the B&W Commercial Nuclear Fuel Plant (CNFP).
The selected fuel rods were replaced in the fuel assembly with rods having similar enrichments (2.83 w/o U235 replacing 2.85/2.95 w/o U235).
Impact of the replacement rods on the fuel cycle and fuel perfonnance analyses was evaluated and found to be negligible. The assemblies were fully recertified for use.
3428f
. To minimize the necessary number of fuel assemblies to be torn down, loading records were reviewed to find assemblies containing rods from two or more pellet lots. This resulted in the selection of five bundles from 21B and three from 91F which provided all lots for which no data existed. Sample fuel rods were selected from the outer two rows which contain half of all the rods in an assembly. This allowed the guide tubes to remain intact with the upper end fitting avoiding additional tear-down.
EVALUATION The resinter test data was evaluated using the B&W MAB0M code which performs a statistical assessment of the data and determines the predicted densification potential for the various analysis applications.
The average pellet densification results provide inputs for:
1 1.
Creep Collapse Analyses (maximum densification) t 2.
End-of-Life Fuel Rod Pressure Analyses (nominal densification) l l
l The individual pellet densification results provide inputs for:
l 1.
Linear Heat Rate to Melt Analyses (maximum densification) l i
2.
LOCA Initialization (maximum densification) l l
The predicted in-reactor densification in each case was compared to the densification assumed in the generic fuel performance analyses applicable to TMI-1.
3428f j
- _ _ ~
e RESULTS The densification characteristics for both batches of fuel were shown to be acceptable. The existing generic fuel thermal analyses for TMI-l remain applicable and bounding.
Statistically, tests for equality of variances within pellet lots were found to be generally consistent with the hypothesis of equal variances indicating that fabrication process factors affecting fuel densification behavior were under control. The modified sampling plan was judged to have yielded fuel pellet sample populations and, subsequently, densification characteristics representative of each of the two fuel batches. Also, the results were consistent with the densification potential of other reload fuel pellets made at B&W CNFP during the same period.
Thus, the sampling plan for resinter testing of the two TMI-l batches, compared to the standard sampling technique for production fuel, shows an equivalent degree of densification and provides a representative comparison between the two samples per fuel pellet lot.
The batch statistics, as determined by the MAB0M model for predicting densification potential, have shown that margin remains between the two batches and the assumptions for LOCA initialization and fuel melt limit in the existing analyses. Based on stored energy equivalency the peaking margin is 3428f
I
. about 0.5 KW/FT. Therefore, it is concluded that application of the densification characteristics for the two TMI-1 fuel batches will result in conservative fuel perfomance predictions. Resinter data for all other fuel pellet lots contained in the two batches, including fuel pellet lots contained in the replacement fuel rods, were previously obtained using standard B&W sampling methods and also found to be acceptable.
i 3428f
.