ML18054A609

LER 86-017-01:on 860410,unidentified Primary Coolant Sys Leakage Calculated to Be in Excess of Tech Spec Limit.Caused by Relief Valve Not Fully Reseated & Reactor Head Vent Sys Valves Not Fully Seated.Valve repaired.W/890323 Ltr

ML18054A609
Person / Time
Site:	Palisades
Issue date:	03/23/1989
From:	Johnson B, Kozup C CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
LER-86-017, NUDOCS 8903290111
Download: ML18054A609 (9)

Text

---18.&31 U.L lllUC\.IAll lllQU\.ATOll'I' 0 1*1a.t """'°"10 0. llO. J1I0-41CM LICENSEE EVENT REPORT ILEA) .. ,. .... , ,,,, .. PACILIT'I

...._ Ill PALISADES NUCLEAR PLAHT l oac:alT -II CZI r-* ,,. O 15 I 0 1.0 IO 1215 15 1 loF 01 8 TITU I., PRIMARY COOL.AJifT SYSTE...M LEAKAGE GREATER THAN 1 GPM UNIDENTIFIED IYIMT OATI I* u11-.-111m lllPOllT DATI 171 OTMlll PACIUTIU -.vm

• MOM'H QAY YIAll PACll.ITY

........ OOCXIT NUMellllll N/A o 14 i I o e 6 s I 6 -o I 1 11 -o 11 o I 3 3 s 19 N/A Ol'lllATI*

nt* lllPOllT IS IUmflTTIO l'UMUANT TO THI lllQUllllMINTI OP 10 CPll §: 10-. -er,,,_ ., .,,,. fel ..... 1111 11111D1 *1 N *.C1G11 -*.a111 ** ,1ta111111or1 -----....... -+--t ---"°"'" I -...C.lllllD 0 I 91 8 -ao_....1111111 IO.Jllolm

• .711111111*1 n.71*1 72.711el ---I0.21101121

..... "'-"""" ..... --t Ii*= -I0.7Jlall21Ctll

--OTMlll

__ ,,, r .. r. NlfC ,_ .......

...... ICtlltorl

• • • 1*11111"' I0.7Jlolllllll

-I0.7Jloll21111

--I0.7'41112111111 LICl ... 1 CONTACT POii TMll Liii 1111 IO.'JllllllllllllllAI

,.., I0.71191111111111111

• IO.nllll2lh1J AlllA COOi C S Kozup, Technical Engineer, Palisades Plant 61116 716141 -18191113 c:m9'\.ITI ONI LINI POii UCM COll9'0NINT PAILUlll OllClllllO IN nt* lllPOllT Ila! MANUPAC. TVlllll CAUH SYITIM COWONINT CAUll l'flTIM COWONINT x c 1 B P 1 I p 1 x A1.B 2 t 01 I I I I I I I I -..P\.lllllNTAL llll'OllT lxPICTIO 11'1 --,

llf ..... -IX,.,C'TIO

$UllllO/ON DATii IXl'ICTIO su11111a1011 OATI 1191 NllCP--111-131 Abs.tract On April 10, 1986 with the Plant at 98 percent power, unidentified primary coolant system leakage was calculated to be 1.25 gpm, which is in excess of the Technical Specifications iimit of 1.0 gpm. Following confirmation of leakage in the Containment Building, the Plant was placed in.the hot shutdown condition.

Leakage was identified from two sources: relief valve RV-2006, which protects letdown piping, had not fully reseated and was leaking to the quench tank and ultimately to the containment floor pit. Internal damage in the valve inhibited its disc from properly seating. reactor head vent system valves PRV-1067, PRV-1068 and PRV-1072 were also not fully seated and were leaking to the containment floor pit. the cause is unknown at this time. RV-2006 was repaired and returned to service. The reactor head vent valves were cycled, resulting in the downstream valve properly reseating.

The failure of RV-2006 has been attributed to excessive cycling due to high letdown system pressure.

The failure of PRV-1067, PRV-1068 and PRV-1072 have been attributed to metal shavings found in the valve seat area. 8903290111 890323 ADOCK 05000255 PDC LER 86017-01-1101 I I I NRC Fo.rm ..... 19-<!31 LICENSEE EVENT REPORT (LERI TEXT CONTINUATION U.S. NUCLEAI! REGULATORY COMMllllON APPROVED OMB NO. 3150--'11().t EXPIRES. 8131 /85 FACILITY NAME 111 DOCKET NUIMEA 121 LEI! NUMHI! l!ll ,AGE 131 P.A..LISADES NUCL:t::P_"ii

?:SA.HI' 0 1 s I 0 I 0 I 0 I 2 j 5 I 5 816 -0 11 17 -0 I l 0 12 OF 0 I 8 TEXT 1/f now_.. io ,.,,.-, u* -NllC Form .m.4'1! 1171 Description On April 10, 1986 at 0452 hours0.00523 days <br />0.126 hours <br />7.473545e-4 weeks <br />1.71986e-4 months <br />, the results of the daily primary coolant system (PCS) [ABl leakage calculation indicated unidentified leakage at 1.25 gpm. At the time, the Plant was operating at approximately 98 percent full power in its fourteenth continuous day of power operation.

With unidentified PCS leakage in excess of 1 gpm, the Plant entered a six hour LCO, per Palisades Technical Specification (TS) 3. l.5(a). Throughout the period preceding the April 10, 1986 event, PCS unidentified leakage was consistently calculated to be below the TS limit. By April 9, 1986 the calculation showed approximately 0.5 gpm unidentified leakage, which was approximately equivalent to unidentified leakage noted in previous operating periods. The following day's (April 10, 1986) calculation, showed a step change to 1.25 gpm, initiating the event. At 0753 approxirnxately three hours into the event, Operations personnel commenced a power reduction after confirmation of system leakage to the containment

[NH] floor. An Unusual Event declaration was made at 0815 with all required notifications completed by 0828. Plant shutdown continued, culminating with the reaction being placed in the hot shutdown condition at 1353, April 10, 1986. The following sequence describes the event as observed by the operators:

April 2, 1986 April 3, 1986 April 4, 1986 April 5, 1986 April 6, 1986 April 7, 1986 April 8, 1986 1933 April 9, 1986 PCS unidentified leakage calculated to be 0.215 gpm PCS unidentified leakage calculated to be 0.166 gpm PCS unidentified leakage calculated to be 0.227 gpm PCS unidentified leakage calculated to be 0.387 gpm PCS unidentified leakage calculated to be 0.34 gpm PCS unidentified leakage calculated to be 0.338 gpm PCS unidentified leakage calculated to be 0.349 gpm Auxiliary operator reports some minor unquantifiable leakage from the area near the letdown orifices.

0505 PCS unidentified leakage was calculated to be 0.519 gpm. The Shift Supervisor noted in the logbook entry that PCS unidentified leakage appeared to be NAC FO,._M 3e&A 19-831 LER 86017-01-1101 NRC For"' -A 19-831 LICENSEE EVENT REPORT (LERI TEXT CONTINUATION U.S. NUCLEAR llEQULATOllY COMMllllON APPROVED OMI NO. 3150-41114 EXPIRES: 81311115 FACILITY NAME 111 OOCKET NUMHR 121 LEiii NUMallll Ill PAOI I I PALISADES NUCLEAR PLANT o 1 s I o I o I o I 2 15 I 5 8 16 -o 11 I 7 -OJ 1 o 13 OF o I 8 TEXT l/f,,,.,...,;...

ii,.,,..,_, --NlfC Fonn .-,.*111171 steadily increasing.

Operators became more attentive to potential leakage sources. 1245 Area around letdown orifices was inspected.

April 10, 1986 0452 0515 0640 0700 0710 0753 0815 0828 1353 1800 NflllC 388.A 19-831 LER 86017-01-LIOl Observed leakage was determined to be insignificant.

Letdown orifices were trimmed to the desired flow rate. The results of the daily PCS leakrate calculation indicated PCS unidentified leakage at 1.25 gpm. TS LCO 3.1.S(a) was entered due to unidentified leakage greater than 1 gpm. Operators began isolation of potential leakage sources. Based on recent operating problems, control valve CV-3069 [BQ;TSV] in the safety injection tank [BQ;TK] fill-and-drain header and the three-way diversion valve, CV-2056 [CB;20], in the chemical and volume control system (CVCS) [CV]°, were isolated.

An auxiliary operator was dispatched into the Containment Building to search for evidence of PCS leakage. PCS letdown was isolated.

Auxiliary operator located PCS leakage inside containment, but could not specifically identify the source. A power reduction was commenced from 98 percent power due to the confirmation of system*leakage.

Unusual Event declared.

Unusual Event notifications were completed.

The reactor [AB;RCT] was placed in hot shutdown condition.

A second entry into the Containment Building was made. The observed leakage path was determined to be through open manual valve 1060F PC [AB;V]. Observation of the area determined that the labels on manual valves 1060F PC and 1060G PC were switched.

Consequently, previous valve line-ups had inappropriately opened 1060F PC, while 1060G PC was in the closed position when it should have been

---

---

NRC Form 31i1A 19-831 LICENSEE EVENT REPORT (LERI TEXT CONTINUATION U.I. NUCLEAll llEGULATOAY COMMllllOfll APPROVED OMI NO.

EXPIRES: 8/31185 FACILITY NAME 111 DOCKET NU .. Ell 121 Liii Ill PAOI Ill PALISADES NUCLEAR PL.ANT 0 l5IOIOI0!2.!5j5 8j6 -01117 -Oj l 014 OFO TEXT l/f,,,.,,.

-* ,..,,.,,.,, ---NltC Form .m4'111171 2000 2130 2200 2307 April 11, 1986 0208 0445 0630 0930 NlllC FOlllM lee.A 19-831 LER 86017-01-LIOl open. With 1060F PC open, the quench tank [CA;TK] was capable of relieving through this line to the Containment Building.

The system was realigned to the correct closing 1060F PC and opening 1060G PC. A confirmatory three hour duration PCS leakrate was initiated.

RV-2006 was isolated.

Over a 30 minute period following isolation, the downstream temperature from RV-2006 was noted to decrease by 25 degrees F, indicating that it _had indeed been a leak path from the PCS. The results of the three hour PCS leakrate calculation indicated unidentified leakage to be 0.25 gpm, confirming that RV-2006 was leaking to the quench tank and subsequently, to the Containment Building.

With acceptable unidentified leakage, the LCO on PCS leakage was exited. The Plant secured from the Unusual Event'." A containment entry was made verifying that 1060F PC and 1060G PC were*now in their correct position.

AdditionaJly, the tags on the valves were

• exchanged.

The results of a confirmatory leakrate calculation indlcated PCS unidentified leakage at 0.551 gpin. The relative accuracy of this calculation was questionable, however, due to large level changes in both the volume control tank and pressurizer

[AB;PZR].

At approximately the s*ame time, the quench tank level was noted to be increasing.

Investigation into the level increase was initiated.

Operators commenced another three hour PCS leakrate calculation.

Since the quench tank level steadily increased after manual valves 1060F PC and 1060G PC were realigned, it appeared that RV-2006 was not the entire leakage source. The results of the three hour PCS leakrate calculation showed unidentified leakage at 0.513 gpm. Several containment entries were made during

. NRC Form .....19-631 LICENSEE EVENT REPORT (LERI TEXT CONTINUATION U.I. NUCl.EAR REOUl.ATORY COMMllllON APPROVED OMll "10. 3150-'11°'

EXPIRES: 8/31185 FACll.ITY NAME Ill DOCKET NUMllER 121 LER NUMalll Ill PALISADES NUCLEA.E PL.Al"J'r TEXT (If,,,...-*,..._, ---NflC "-

1900 2200 2340 April 12, 1986 0155 0400 the day to examine the PCS components in systematic deta'il with results showing no visible leakage. A new three hour leakrate was commenced.

PCS leakrate was determined to be 0.294 gpm unidentified.

Operators reopened 1060F PC. Leakage was identified through open 1060F PC, originating from the reactor head vent system through valves PRV-1067 and/or PRV-1068 and also through PRV-1072.

PRV-1072 was cycled several times, resulting in no visible leakage through 1060F PC. With PRV-1072 apparently sealed, both PRV-1067 and PRV-1068 were also cycled. Following verification that PRV-1072 was sealed, manual valve 1060F PC was The quench tank level remained stable following the cycling of PRV-1072, confirming the reactor head vent system as a second leakage path. The results of a three hour PCS leakrate calculation unidentified leakage at 0.139 gpm. With extremely low unidentified leakage, the reactor was taken critical for return to power. Evaluation And Corrective Actions The excess PCS unidentified leakage resulted from two distinct system failures involving valves which did not properly reseat after operation; both pressure letdown relief valve RV-2006 and reactor head vent system valves PRV-1067 and/or PRV-1068 and PRV-1072.

While the mislabled and consequently mispositioned manual valves 1060F PC and 1060G PC were not a true source of PCS leakage, this condition was responsible for precluding both actual sources of leakage from being directed to and maintained in the quench tank where a corresponding level increase would have been readily detectable.

A. Pressure Letdown Relief Valve RV-2006 Pressure letdown relief valve RV-2006 protects the intermediate pressure letdown piping and letdown heat exchanger

[CB;HX] from overpressure.

The valve is a Farris Engineering, 2600 Series safety-relief valve, Serial All. Pressure surges in the letdown piping as the result of trimming the letdown orifices on the N .. C FO"M 368.& 19-631 LER 86017-01-LIOl PAOI 131 NRC Form 311A 19-331 FACILITY NAME 111 LICENSEE EVENT REPORT (LERI TEXT CONTINUATION DOCKET NUMeEll (21 U.I. NUCLEAll llEQULATOllY COMMllllON APPROVED DMll NO. 3150--0104 EXPIRES: 8/31 /85 LEll NUll9111 (II 'AQI 1:11 PALISADES lJUCLBA."R PLANT TEXT llr ,_. -ii,.,,.,..,,, --NftC Form.-.)/

1171 previous day (April 9, 1986, 1245 hours0.0144 days <br />0.346 hours <br />0.00206 weeks <br />4.737225e-4 months <br />) caused RV-2006 to li.ft and relieve to the quench tank several times. Following a lift, the valve disc did not properly reseat. From this point, RV-2006 was a continuous source of leakage, responsible for the step increase of 0.7 gpm unidentified leakage. this leakage would normally have been detected as a level increase in the quench tank. However, an open vent path existed from the quench tank to the Containment Building through manual valve 1060F PC, which was open rather than closed. This situation masked most leakage from RV-2006 into the quench tank since both leakage volume and pressure were vented from the quench tank. Evaluation into the root cause has determined that RV-2006 should not have been challenged as it was during the orifice trimming evolution.

Pressure surges in the intermediate pressure letdown piping should have been anticipated and transmitted to pressure indicating controller PIC-0202 (Fischer Porter Model 53EL3000)

[CB;PIC] to control one of two identical backpressure regulation valves (CV-2012 or CV-2122) maintaining the intermediate letdown pressure below the lift setting of RV-2006. The inability of the backpressure regulation circuitry to properly anticipate and respond to such pressure increases in the intermediate letdown piping was responsible for frequent lifting of RV-2006; likely accelerating its ultimate failure. After an evaluation to coordinate the response time of the backpressure circuitry with the letdown orifice settings, the intermediate pressure control circuitry was adjusted to minimize and maintain pressure less than the 600 psi setpoint for RV-2006. This action has resulted in proper operation of the letdown system. Subsequent disassembly of RV-2006 revealed that the valve's bellows were distorted.

The distortion was apparently sufficient to inhibit the valve's disc from fully reseating.

Following necessary repairs, RV-2006 was placed back in operation.

The reliability of RV-2006 was evaluated to determine whether system operation be enhanced by valve replacement with a different type of valve. During this evaluation a review of operating histories of R\'-2006 revealed that the valve was a high maintenance item after the valve had cycled eight to ten times. However, reliability of different valve types indicated similar performance after repeated cycling. Therefore, changing valve design was felt to be imprudent.

B. Reactor Head Vent Valves PRV-1067, PRV-1068 and PRV-1072 The reactor head vent system, installed at Palisades in response to NUREG-0737, Item II.B.1, incorporates a Target Rock (Model 80 B-001) solenoid operated pilot valve system. The system's vent path to the quench tank utilizes two of the valves, PRV-1068 and PRV-1067 in parallel, followed by a third valve, PRV-1072.

The downstream path from PRV-1072 is normally aligned directly to the quench tank. A N"'C FO"-' 388A 19-831 LER 86017-01-LIOl Form llilA 19-831 FACILITY NAME 111 LICENSEE EVENT REPORT (LERI TEXT CONTINUATION DOCKET NUMHFI 121 U.I. NUCLEAlll lllEGULATOlllY COMMllllON APPROVED OMD NO.

EXPIRES: 8/31185 LElll Cll PAO! 131 o 1 s Io Io Io I 21 5 15 81 6 _ 01117 -o 1 1 o I 7 OF o 18 T£XT llr mon -ii,.,,.-, u* -NlfC Fomt NlllC FOPIM l88A path to the Containment Building floor pit is normally isolated by manual valve 1060F PC. When the Plant was being returned to service on March 24, 1986, following a short maintenance outage, operators had become aware that valve PRV-1067 and/or PRV-1068 either was or had for a time, not fully seated as evidenced by pressure indication immediately downstream of these valves which showed nearly full PCS system pressure.

It was also recognized, however, that PRV-1072 was fully seated and holding pressure.

Consequently, on March 24, 1986 operators cycled PRV-1072 to vent off the pressure between the upstream and downstream valves. Since there was no level increase in the quench tank, PRV-1067, PRV-1068 and PRV-1072 were incorrectly assumed to have all fully reseated.

The reason for the incorrect assumption was that manual valve 1060F PC was not closed as it should have been. Because it was open, operators were unaware that PRV-1067 and/or PRV-1068 along with . PRV-1072 had all failed to fully reseat. From this point, the reactor head vent system was a continuous leakage source to the Containment Building floor pit through open manual valve 1060F PC. Upon realignment of manual valves 1060F PC and 1060G PC. Upon realignment of manual valves 1060F PC and 1060G PC, this leakage was directed to the quench tank, which facilitated the identification of this leakage source. Target Rock valve systems of the type utilized at Palisades have been known to exhibit a common mode problem involving the spurious opening of the downstream solenoid operated pilot valves caused by the pressure pulse which is initiated from the opening of the upstream valves. The problem observed at Palisades appears to be unrelated to the common mode problem of unintentional lifting. the downstream valve PRV-1072 did not spuriously open. It was operated to the open position and for reasons unknown at the time of the event, failed to fully reseat. The common mode nroblem concerns a temporary unintentional lift until upstream and downstream pressures equalize.

At Palisades, PRV-1072 remained parti2lly unseated until it was again cycled on April 11, 1986 at approximately 2200 hours0.0255 days <br />0.611 hours <br />0.00364 weeks <br />8.371e-4 months <br />. PRV-1067 and/or PRV-1068 remain partially unseated at this time. For the interim period, a caution tag had been hung on the key switch for PRV-1072 stating that the valve is not to be cycled without the permission of the Shift Supervisor and warning of the potential PCS leakage path from the PCS should PRV-1072 be cycled and fail to fully r*eseat. While it is not expected that justification will exist for cycling PRV-1072, any system leakage through this path, whether the valve is cycled or not, will now evidence itself as a level increase in the quench tank. All six vent valves, PRV-1067, 1068, 1069, 1070, 1071 and 1072 were disassembled and repaired during a 1086 maintenance outage. During repairs, small metal shavings were found in the valve internals.

19-1131 LER 86017-01-LIOl

..

NRC For"' -A U.I. NUCLEAI! l!EGULAT_ORY COMMIDION 19-8JI LICENSEE EVENT REPORT (LERI TEXT CONTINUATION APPRO\IED OMll NO.

EXPIRES: 8/31185 FACILITY NAMI 111 DOCKET NUMIEI! 12) LEI!

Ill PALISADES NUCLEAR PLMJT 0 I 5 I 0 I 0 I 0 12 I 5 15 81. 6 -0111 7 -0 11 0 I 8 OF . 0 ,s TEXT l/f ,_. -* ,.,,.-. -..--NM:"-.-.**11171 These shavings are believed to be the reason for past valve leakage. A review of Target Rock valve reliability and applicability in the head vent system revealed that the application was appropriate and that any valve of this design is subject to failure if debris is present. Following valve repair, the reactor head vent system was flushed to remove remaining metal shavings.

C. Misaligned Manual Valves 1060F PC and 1060G PC Manual valves 1060F PC and 1060G PC were installed during the 1981 refueling outage. At this time valve labels were correctly hung on the two valves using a drawing from the modification package. Subsequently, when the piping and instrument drawing (P&ID) was updated to include this modification, the valve numbers were transposed on the drawing. Consequently, when the system was aligned using a checklist made up from the P&ID; the position of manual valves 1060F PC and 1060G PC became interchanged.

The valve positions and labels were changed to match controlled Plant P&IDs on April 10 and 11, 1986. The personnel error in transposing the valve labels occurred in the drawing* revision process, but cannot specifically be identified.

This occurrence is considered to be isolated.

The existing review process is generally considered adequate to preclude this type of problem. Analysis Of The Event Unidentified leakage in excess of 1 gpm necessitated the completion of a Plant shutdown which requires this event to be reported per iOCFR50.73(a)(2)(i)(A).

Although the unidentified PCS leakrate exceeded TS limits, no threat to public health or safety resulted.

The maximum leakage was far below the charging capacity of a single charging pump (approximately 40 gpm). The Plant was shutdown in an orderly fashion in advance of the time allotted to be shutdown by TS 3.1.S(a).

Additional Information Previous occurrences of PCS unidentified leakage were reported in Licensee Event Reports 85022, 84012, 84013, 84016, 84024 and 84025. None of the previous events concerned the vaives involved in this particular occurrence.

NlllC FO,_M l&eA LER 86017-01-LIOl I

'* consumers Power POWERING MICHlliAN'S PROliRESS General Offices: 1945 West Parnall Road, Jackson, Ml 49201 * (517) 788-0550 March 23, 1989 Nuclear Regulatory Commission Document Control Desk Washington, DC 20555 DOCKET 50-255 -LICENSE DPR-20 -PALISADES PLANT -LICENSEE EVENT REPORT 86-017-01

-PRIMARY COOLANT SYSTEM LEAKAGE GREATER THAN l GPM UNIDENTIFIED Licensee Event Report (LER) 86-017-01 (Primary Coolant System Leakage Greater Than 1 PGM Unidentified) is attached.

This revision is being submittal to report additional information regarding valve failure. The outstanding commitment to provide a revis:Lon to this LER was identified in the Palisades SALP 8 report dated August 15, 1988. The March 23, 1989 due date was established with the Palisades NRC Resident Inspector.

This event was reportable to the NRC per 10CFR50.73(a)(2)(i).

Brian D Johnson Staff Licensing Engineer CC Administrator, Region III, USNRC NRC Resident Inspector

-Palisades Attachment OC0389-0103-NL02