ML20085F547

AO AO-27-74:on 741130,released to Plant Vent Stack from Waste Gas Surge Tank.Caused by Waste Gas Surge Tank Leakage, Due to Leaky Relief Valve.Valve Repaired & Pressure Tested

ML20085F547
Person / Time
Site:	Palisades
Issue date:	12/12/1974
From:	Sewell R CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
To:	US ATOMIC ENERGY COMMISSION (AEC)
Shared Package
ML20085F551	List: ML20085F547 ML20085F552
References
AO-27-74, NUDOCS 8308220429
Download: ML20085F547 (3)

Text

. -. , ,. ,

()

. (v;  %)

f 6'  % Consumers Et

%vp Power 4 company h, [ hth l General Offices: 212 West Michigan Aver'uu. Jackson, Mic higan 492O1. Area Code 517 7H8-OSSO December 12, 1974 S  %

9) b qL;\yl' [

                                                       ,th          J                  9 CEO13 l9745 Z-              Re:     Docket 50-255 Directorate of Licensing                        81 "         a         Uf          License DPR-20 USAtomicEnergyCommission(b               \A        7 Palisades Plant Washington, DC 20545                      \ S';s " " "- a    [j                    A0-27-74 y        7    , 3s Gentlemen:                                      '< C Attached is Abnormal Occurrence Report A0-27-74. This report concerns a release of radioactive gas that did not meet the limiting conditions of operation which require a minimum 15-day holdup time.

This release was caused by a leaking pressure relief valve. An analysis of the release rate showed it to be about one-one thousandth of the allowable limit. Yours very truly, F Ralph B. Sewell (Signed) DAB / map Ralph B. Sewell Nuclear Licensing Administrator CC: JGKeppler, USAEC l l l 8308220429 741212 DR ADOCK 05000 y l 12Gl9-( 2h COPYsSENT REGIONL

                                                                                                                        $   m

C 0'

d. v ABNORMAL OCCURRENCE REPORT Palisades Plant Docket 50-255

1. Report No: A0-27-74, Docket 50-255 2A. Report Date: December 12, 197h 2B. Occurrence Date: November 30,1974 (classified an A0 on December 2,1974)

3. Facility: Palisades Plant, Covert, Michigan

4. Identification of Occurrence: Gas release to plant vent stack from the vaste gas surge tank (WGST).

5 Conditions Prior to Occurrence: The reactor was in the shutdown condition.

6. Description of Occurrence: During preparations to degas the primary system, the following sequence of events took place: At 1020 hours, operators lined up a vaste gas decay tank for operation; a vaste gas compressor was bump-started to test system operability. The operator noticed that the tank he placed in service (T-68A) had in-creased from 0-47 psig and that a second tank (T-101B) had decreased from 75-50 psig. Simultaneously, at 1020 hours, the control room vented the volume control tank (VCT). At 1022 hours, the control room received a " gas radiation monitor hi-rad" alarm. The operator checked the alarm and found it to be the stack-gas selected isotope. The monitor vas thought to have spikef. erroneously and-the alarm was cleared. At 1024 hours, the control operator received a second alarm on the stack-gas selected isotope monitor and immediately stopped venting the VCT.

7 Apparent Cause: WGST relief valve leakage. A pressure test of the WGST, following the abnormal occurrence, showed that the relief valve was leaking.

8. Analysis of Occurrence: As the VCT was being filled, the cover gas was vented to the WGST (a normal operation). During this operation, the pressure relief valve on the WGST malfunctioned causing the gas to be vented to the plant stack. The vented gas was detected by the stack-gas single isotope montitor. Calculations showed that approximately 3 0 cubic meters of gas were released to the plant stack.

Investigation of the drop in pressure in T-101B revealed that the inlet valve was open which resulted in an attempted pressure equali-zation between T-101B and T-68A. A calculation of the gas transfer { showed a discrepancy of approxinately 1 3 cubic meters for which we i could not account. Either tank pressure instrumentation inaccuracy l' l l l l  ! l I

l e- .e er o

                                                                               ~

O 2 or loss to the WGST through the compressor could account for this , discrepancy. Because it is known that gas can leak back through the compressor, it is assumed that the " lost" gas was released with the gas from the VCT venting. Calculations were made as to the activity that might have been released. These results are shown below: Loss From Tank 101B: Xe 'I * ****

• 133 I

131 6.67x10[mCireleased. Fraction of Allowed Rate: 2 74 x 10 Loss From WGST (as a result of VCT vent): Xe133 .00 x 10 4 mci released. I g 4.23 x 10 f mci released. Co g 2 97 x 10 I mci released. Fraction of Allowed Rate: 1.12 x 10-3 9 Corrective Action: The primary cause of losing the gas was the faulty WGST relief valve, which has since teen repaired and pressure tested. This is a rubber seated relief valve e.nd the O-ring was distorted. That gas which is assumed (above) to have backed out of T-101B would have had to leak back through a compressor check valve. These check

                                 ' valves are a known problem and because of this the compressors are normally left in manual with the WGIyf inlet valves shut. A modifi-cation is nearing completion which will interlock the tank inlet valves with compressor starting to prevent back leakage.

10. Failure Data: There is no record of previous failures. This relief valve was a Farris Type 26FA10-141.

4 i , _ _ _ _ _ _ _ _ _ - _ _ _ - - _ - - - - . - _ _ - - - _ - - - _ - - - - - - _ - - - - - - _ - .}}