ML19309A323
| ML19309A323 | |
| Person / Time | |
|---|---|
| Site: | Rancho Seco, Arkansas Nuclear |
| Issue date: | 08/03/1973 |
| From: | SACRAMENTO MUNICIPAL UTILITY DISTRICT |
| To: | |
| Shared Package | |
| ML19309A313 | List: |
| References | |
| NUDOCS 8003270752 | |
| Download: ML19309A323 (7) | |
Text
.
EVALUATION OF ENRICHMENT CROSSOVER V
1.
Introduction This report documents the events which occurred at th'e Babcock & Wilcox Commercial Nut. ear Fuel Plant (CNFP) that resulted a crossover of a small number of fuel pellets between fuel being processed for the Sacramento Municipal Utilities District l
(Rancho Seco Corc I) and the Arkansas Power and Light Company (Arkansas Nuclear One, Core I).
The detailed evaluation that follows describes the circumstances that led to the crossover of a total'of nine foreign pellets into Rancho Seco Batch I fuel. These foreign pellets which have been replaced with the proper pellets were found as a result of visual inspection of fuel pellets and neutron scanning of loaded fuel rods.
O e source of the enrichment crossover has been traced to the design of two roller micrometer (R/M) installations. The R/M*s were not a part of the normal process line, but were only used to sort, by diameter, those pellets that were increased in density by resintering of already manufactured pellets. The design of the R/M's ellowed the fuel pellets to become lodged in locations inaccessible for visual inspection and clean-out at enrichment changes, while permitting them to fall out into subsequently processed material.
Af ter the R/M had been identified as the source of the enrichment crossover, all the production records were reviewed to identify the pellet lots, and fuel rods Joa:!ed from those pellet lots, hat could have been affected by this operation.
All pellets that had not yet been loaded into fuel rods were examined visually for the proper enrichment code stamp, and those fuel rods that had already been Asembled were examined by neutron scanning.
In two separate quality assurance 8003270 v
w
audits, District personnel have verified that there has been an accurate re-O construction of the events that occurred and that all of the fuel subject to -
possible enrichment crossover has been identified and has received one hundred percent inspection, either visual or by neutron scanning. As a result, Rancho Seco fuel does not presently contain any foreign pellets.
2.
Evaluation In late 1972, B&W began processing the Rancho Seco and Arkansas fuel. After discovery of the fuel densification phenomenon at the R. E. Ginna plant and sub-sequent evaluation of information, it was decided that it would be prudent to in-crease the nominal density of the Rancho Seco and Arkansas fuel from 92.57. theo-I retical density (T.D.) to 95-967. T.D.
At the time of this decision, rod loading of Batch I (2.067.) Arkansas fuel was virtually complete, Batch 2 (2.727.) Arkansas fuel
- I pellets were in production at Kerr-McGee, and Batch 3 (3.057.) Arkansas fuel pellets deliveries by Kerr-McGee were approximately 507. complete. Deliveries of Batch I (2.017.) Rancho Seco fuel pellets by NUMEC were approximately 507, complete, Batch 2 (2.677.) Rancho Seco fuel pellets were in production at NUMEC, and production had not yet begun on Batch 3 (3.007.) Rancho Seco fuel. Approximately 13 metric tons of 3.057. Arkansas fuel and approximately 13 metric tons of 2.017. Rancho Seco fuel were then in storage at the CNFP. NUMEC and Kerr-McGee were. instructed to sinter
~~
the balance of the fuel for these contracts to the higher density and to resinter all 92.57. T.D. fuel prior to shipment to the CNFP. B&W decided to instiall a sintering furnace at the CNFP to resinter that fuel already on hand. Since the 92.57. T.D. fuel pellets would shrink to varying degrees upon resintering, it was necessary to incorporate a means of sorting the resintered fuel by diameter. The roller-micrc=eter (R/M) was selected for this sorting. The sumary of resintering requirements for Rancho Seco and Arkansas fuel is shown in Table I.
It was not G
.necessary 'for CNFP to sort the Arkansas fuel resintered at Kerr-McGee because tie two lots of 2.72% Arkansas fuel were centerless ground to final diameter subsequent to resinter and - the nine lots of 3.05%' Arkansas fuel resintered at Kerr-McGee were also sorted at Kerr-McGee. Resintering of-the 3.05% Arkansas fuel at the CNFP 4
4 began December 4, -1972, and continued to February 7,1973, at which time resintering of the 2.01% Rancho Seco fuel cannenced. Resintering of the 2.01% Rancho Seco fuel was completed March 17, 1973. The first R/M station became operational December 7, 1972. A second R/M station was installed on December 11, 1972 Roller-micrometer operations wsre concluded on March 17, 1973.
During the loading of Lots 1 and 2 of 2.01% Rancho Seco fuel on March 16, 1973, a 2.06% Arkansas pellet was visually detected due to obvious differences in color, texture, and length. CNFP records 'show that on man"h 16 a number of 2.06%
Arkansas pellets were exposed to the pelled loading room environment for 1-1/2 4
,<1ours to determine their rate of moisture pickup. Since that was the only exposure of 2.01% Rancho Seco fuel to 2.06% Arkansas fuel, then a 2.06% Arkansas pellet must have crossed over into the Rancho Seco fuel at this time.
It is also true that this was the only sample of a different, enrichment present in the loading room during loading of Rancho Seco fuel. Loading of Rancho Seco Lots I and 2 was halted immediately. Rancho Seco Lot 4 was loaded March 19, 1973 pending com-
. _ l pletion of the examination of Lots I and 2 A decision was then made that no further resintered 2.01% Rancho Seco fuel would be loaded until each pellet was i
virually examined to verify the enrichment code. An inspection area was established, personnel were briefed regarding. the problem, and instructions were given to reject l
any pellet whose enrichment code letter was not "H".
In addition, arrangements were made at this time to rent a neutron scanner for inspection of loaded fuel rods, 1
t I
e
~,,
- +, -
r
r N
l Beginning on March 20,1973, all 2.01% Rancho 'Seco fuel was loaded only af ter visual
/' inspection for the enrichment code. ~ During visual inspection of the 2.01% Rancho 1
.Seco fuel, the following foreign pellets vasa: identified:
' Date Foreign Pellet i
2.67% Rancho Seco
. March 20, 1973 March 22, 1973 3.05% Arkansas March 26, 1973 2.67% Rancho Seco March 26, 1973 3.05% Arkansas i
An investigation was made to determine the point, or points, at which the enrichment crossovers could have occurred. Since all ' foreign. pellets were detected within large quantities of a second enrichaent, the problem seemed to logically be-
~
associated with the clean-up b'etween enrichment changes. A detailed equipment review
~
indicated the possibility of pellet entrapment in the R/M which might not be discovered during a normal enrichment change over inspection. A similar possibility was not identifiable in any other equipment. To evaluate this theory a R/H was inspected to simulate preparation for an enrichment change over. Scrap pellets were then i
processed by the standard procedure using a production operator. The R/M was then cleaned out and inspected in the usual manner. After operating the R/M without fuel
~
for a chort whilc. a fuel pellet dropped to the table'.
Subsequent dismantling of
..I the R/M disclosed several pellets in hidden locations.
1 From productiot records it was established that fuel rods containing R/M sorted i
pellets from certin lots were free from suspicion of enrichment crossover. -
1.
Lots 9,10, and 11 of the 2.01% Rancho Seco fuel were the first lots sorted at the R/:1 station installed December 11, 1972.
,x 2
Lots 31 and 32 of the 2.67% Rancho Seco fuel were the first lots somted at the R/M, station installed December. 7,1972.
s e l'
p 3.-
Lots 3,'5,.6, 7, and 8 of the 2.01% Rancho Seco fuel were 100% visually inspected prior' to fuel rod loading.
Portions of Lots 1 and 2 of the 2.01k' Rancho Seco fuel were 100% visually 4.
inspected prior.to fuel rod loading.
Therefore, the rods containing portions of Lots 1 and 2 which were not visually inspected and all of Lot 4 were designated for neutron scanning.
The fuel rod scanner became operational on May 7,1973, and scanning of all the fuel,
rods from seven assemblies containing suspeet" fuel from Lots 1, 2, and 4 began.
As a result of scanning the 863 suspect fuel rods, a single 3.05% Arkansas pellet was found in each of four rods. The 284 rods from Lot 11 and the 309 rods that had received previous visual inspection were confirmed as being free of foreign pellets.
i 3J Corrective Actions The enrichment crossover which occurred was due to the design of a type of process kuipment. Actions taken by B&W to prevent a recurrence of this situation are as.
follows:
1.
The design of the R/M which allowed pellets to become lodged in locations inaccessible for inspection and cleanout has been corrected by equipment modifications.
2.
Procedures relating to fuel pellet enrichment control have been reviewed in depth. These procedures are being strengthened where necessary to provide more comprehensive processing, control and documentation, prior to further operations.
3.
Procedures related to enrichment change over have been thoroughly reviewed and revised to provide fully independent and redundant inspections of facilities prior to processing a different enrichment.
4 O.
4.
Conclusions The evaluation presented in Section 2 of this report provides both an explanation of the mechanism by which the enrichment crossover occurred and describes the extensive investigation BG has conducted to assure that Rancho Seco fuel contains no foreign pellets. The corrective actions delineated in Section 3 will prevent a recurrence of the situation disclosed in this report.
O O
4 0
m 4
i l
1 4
e 4
D, 4
0
.O.
O 4
l e,
,,y e-
.-a
-,e-, -..,.. - - -..,,
,)
)
s.
d d
d d
ogs n
n n
n nnt o
o o
o in y
y y
y hre e
e e
e t em b
b l
l b
b it e l
l wnr d
d A
A d
d ii n
n n
n ssu a
a a
a t eq ore 2
3 3
0 L
R 1
3 6
5 i?
ed 9
b er 1
2 2
4 ro 1
3 6
oed e
e u
t t n u
d n
n d
r ne r
n o
o n
h siV h
a N
N a
t t s t
. oet 1
1 1
LRA 9
3 6
4 1
4 1
I ed 0
M be 8
4 rP M
oeF u
e e
e e
u ttN r
n n
n n
r T
nC h
o o
o o
h si t
N N
N N
t t st oeA 1
1 LP 3
t*n e,
me hdco H
L M
J S
N iC rn E
1 7
0 6
2 5
0 6
0 0
7 0
2 2
3 2
2 3
S A
S N
D A
U K
M R
S A
4 1!
4 l
l, 441