ML20101M149
| ML20101M149 | |
| Person / Time | |
|---|---|
| Site: | U.S. Geological Survey |
| Issue date: | 06/24/1992 |
| From: | Smithmd B INTERIOR, DEPT. OF, GEOLOGICAL SURVEY |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9207080081 | |
| Download: ML20101M149 (4) | |
Text
<
}p m,
Q% >,,g
/
.i~
Ng Uilited States Deixtrtilleilt of tile liitei' ion r
oe Gl'OI.OGICAl. $l'R VIT ILOX 25016 M. $._373_
I)ENVER Il DFR.\\l. CFN I ER 1)ENVER, (:()l ORAD () 80223 1% Rt Pl i K} llit Irp June 24, 1992 Docket No. 50-274 License No. R-113 U.S.
Nuclear Regulatory Commission Document Control Desk Washington, D.C.
20555
Dear Sirs:
SUBJECT:
REPORT OF FUEL ELEMENT CLADDING LEAK AT USGS TRIGA Attached is a written report with details concerning a fuel element cladding leak that we experienced during June 1-16.
The presence of fission product activity in the reactor water and facility air was positively identified on June 11.
This event was reported by telephone on June 11 and followed by subsequent telepi ne updates.
The USGS research reactor returned to routine operation on June 17, 1992.
If you have further questions regarding this report, please contact Tim DeBey at (303) 236-4726.
Sincerely '
y
) f L"y i
David B.
Smith Reactor Administrator Copy to:
US Nuclear Regulatory Commission ATTN: Mr.
L.J.
Callan 611 Ryan Plaza Dr, Suite 400 Arlington, TX 76011 Tim DeBey 6
9207080081 926654
/
PDR ADOCK 05000274 i
R ppg J
FUEL ELEMENT CLADDING LEAK - USGS - JUNE 1992 A fuel element cladding leak was detected at the USGS TRIGA reactor facility by observing increased readings on the facility's continuous air monitor (CAM).
The first abnormal indication was detected on June 9; however, gamma analysis of the CAM filters determined that the majority of the activity was caused by natural decay products of uranium and thorium.
No fission products were identified at this time.
The indications were not typical of a fuel element clad failure since all CAM increases occurred after reactor shutdowns, approximately 15 to 45 minutes following the rod insertions.
Another CAM increase on June 10 confirmed that a problem existed, but it was not until early on June 11 that the fission products rubidium and cesium were positively identified on the CAM filter and krypton and xenon were also found in the reactor water.
A search for the source of the fission products was started on June 11.
On that same date, the Reactor Administrator, Reactor Committee chairman, Nuclear Regulatory Commission (both region and headquarters), and General Atomics personnel were notified of the problem.
An attempt to set up for_ pulsing operations to enhance leak detection capability revealed that the digital control console had a fault that would not allow pulsing to be performed.
A review of the CAM strip chart recording showed that a small increase after shutdown had first occurred on June 1 with a 100 cpm increase in a typical background level of 300 to 400 cpm.
The released fission products were identified as Krypton-88, Rubidium-88, Xenon isotopes, and Cesium-138.
The particulate isotopes (Rubidium and Cesium) were detected on the CAM filters and all identified fission product isotopes were detected in water samples.
All samples and filters were analyzed on a high purity germanium detector.
The longest lived isotope was Xenon-133, with a half-life of 5.25 days.
The maximum personnel dose equivalent from fission products was evaluated to be less than 0.2 mrem for the period from 6/1/92 through 6/16/92.
After discussions with other personnel, the fission product source was postulated to be in the poison section of a fuel-followed control rod, where thermal expansion from gamma heating after a_ shutdown would cause the fission product gases to leak.
The regulating rod was 4uspected since it was installed in December, 1991.
Vatious tests were performed by going to full power for up to 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />, then dropping at least one control rod and observing the CAM response.
These tests seemed to confirm that the reg. rod was leaking by showing increased CAM readings when the reg. rod was dropped.
The reg. rod was replaced on June 12 and then the reactor was operated at steady state for two hours and scrammed.
This produced a small increase on the CAM (~250 cpm) which caused more confusion.
4 More steady state testing was performed on June 15 with various y
combinations of control rod insertions from full power.
None of these tests produced a significant increase in the CAM response, indicating that the problem may have been cured by the reg. rod replacement.
On June 16, the reactor was operated at steady state for about 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> and scrammed (pulsing capability was still not available).
This scram produced a net increase of about 1200 cpm on the CAM.
When this CAM increase was first detected, a visual examination of the reactor tank located small bubbles coming from an instrumented fuel element in the G-1 grid position.
These bubbles were being emitted at a rate of about two or three per minute.
After the CAM indicated that no more fission products were being released into the reactor room atmosphere, the instrumented fuel element was removed from the core and replaced with a standard element.
The instrumented element (Serial number 5667) was originally installed in the core in May, 1969, and had been in the core for about 14,500 MWH of operation.
This element is currently in storage at the side of the reactor tank.
After removal of the instrumented element, the reactor was then j
operated at full power for an hour and scrammed.
No CAM increase was seen.
Another full power operation for about 4.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> was followed by a scram that also produced no change in the CAM reading.
The original regulating rod was returned to the core and routine reactor operations were resumed on 6/17/92.
No evidence of fission product leakage has been seen since that time.
The cladding defect is evidently configured such that it was open only during fuel element cooldown after a reactor shutdown.
The characteristic of leaking fission products only after a shutdown is counter to our previous experience with fuel cladding failures and caused confusion during the search effort.
ISOTOPE DATA FOR FISSION PRODUCTS DETECTED Isotope Half-life Appx.B, Table I Appx.B, Table II Kr-88 2.8 hr 1E-6 2E-8 Rb-88 17.8 min 1E-6 3E-8 Xe-133 5.25 days 1E-5 3E-7 Xe-135 9.1 hr 4E-6 1E-7 Xe-138 17.5 min 1E-6 3E-8 Cs-138 32.3 min 1E-6 3E-8 Analyses of reactor tank water samples showed a peak fission product activity of 3.46 pCi per liter, with a total tank water inventory of about 30,500 liters.
__ _- -__L
__m a
i The largest fission product release occurred on-June 16 when the 16.02 uCi emitted gave a daily-averaged concentration in the reactor room air of 4.9E-10 uCi/cc.
This is more-than a factor of 2000 below the limit of 10CFR20 Appendix B, Table I, and more than a factor of 40 below the limit of 10CFR20 Appendix B, Table II.
The following table provides a synopsis of the fission product leak indications on the CAM.
All activities are based on the release occurring over a 30 minute duration.
It can be seen that the fission product releases do not correspond directly to the amount of reactor operation for the day.
Date Operate Net increase Daily release in microcuries MWH on CAM (particulate)
(noble gas)
(total)
June 1 8
100 cpm 0.12 1.25 1.37 June 2 8
100 cpm 0.12 1.25 1.37 June 3 4
0 cpm 0
0 0
June 4 6
150 cpm 0.17 1,77 1.94 June 8 0
150 cpm 0.17 1.77 1.94 June 9 8
500 cpm 0.58 6.06 6.64 June 10 6.5 600 cpm 0.70 7.31 8.01 June 11 5.8 1050 cpm
- 1.22 12.74 13.96 June 12 2
250 cpm 0.29 3.03 3.32 June 15 5.2 0 cpm 0
0 0
June 16 8
1200 cpm
- 1.40 14.62 16.02 June 17 6.5 0 cpm 0
0 0
June 18 10 0 cpm 0
0 0
June 22 8
0 cpm 0
0 0
June 23 4
0 cpm 0
0 0
TOTALS 4.77 49.8 54.6
- indicates that the air was exhausted through a HEPA filter I
1 1
.-