IR 05000272/1990011

Insp Repts 50-272/90-11,50-311/90-11 & 50-354/90-08 on 900317-0430.Violation Noted.Major Areas Inspected:Operations Radiological Controls,Maint & Surveillance Testing,Emergency Preparedness,Security & Engineering & Technical Support

ML18095A237
Person / Time
Site:	Salem, Hope Creek
Issue date:	05/21/1990
From:	Swetland P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:
Shared Package
ML18095A235	List: IR 05000272/1990011 ML18095A234 ML18095A236
References
50-272-90-11, 50-311-90-11, 50-354-90-11, NUDOCS 9006040334
Download: ML18095A237 (41)

Text

Report No License No Licensee:

Facilities:

Dates:

Inspectors:

Approved:

U. S. NUCLEAR REGULATORY COMMISSION

REGION I

50-272/90-11 50-311/90-11 50-354/90-08 DPR-70 DPR-75 NPF-57 Public Service Electric and Gas Company P. 0. Box 236 Hancocks Bridge, New Jersey _ 08038 Salem Nuclear Generating Station Hope Creek Nuclear Generating Station March 17, 1990 - April 30, 1990 Thomas P. Johnson, Senior Resident Inspector David K. Allsopp, Resident Inspector Stephen M. Pindale, Resident Inspector Stephen T. Barr, Resident Inspector Glenn M. Tracy, Reactor Engineer Daniel T. Moy, Reactor Engineer Roy K. Mathew, Reac r Eng"neer 2A Inspection Summary:

Inspection 50-272/90-11; 50-311/90-11; 50-354/90-08 on March 17, 1990 -

'*

April 30, _!990 Areas Inspected:

Resident safety inspection of the following areas:

operations, radiological controls, maintenance & surveillance testing, emergency preparedness, security, engineering/technical support, safety assessment/quality verification, and licensee event report and open item foll owu Results:

The inspectors identified one violatton for th~ Salem station.,There were two Hope Creek licensee identified, non~cited violation An executive summary fo.11 ow '~0060403::::4 f'DR ADOCK u

.

900524 O~':i00027:2

!='DC

'. EXECUTIVE SUMMAR I DETAIL. SUMMARY OF OPERATION.1 Sa 1 em Un it.2 Salem Unit.3 Hope Cree. OPERATIONS.

rAB!£- OF eONTENTS -***

..

-

...

Page 1. *'.,:;.,,:*

3

3

4 Inspection Activities

....

.......

-~

2.2 Inspection Finding & Stgnificant Plant Event~-.

2.. Sa 1 em...

Hope Cree.

RADIOLOGICAL CONTROLS...

• , Inspection Activities....

3.2 Inspection Findings & Review of Event.. Sa 1 em......

Hope Creek.... MAINTENANCE/SURVEILLANCE TESTING 4.1 Maintenance Inspection Activities 4.2 Survei_llance Testing Inspection Activity.*

4.3 Inspection Findings.

4.. Salem...

Hope Cree.

EMERGENCY PREPAREDNESS. Inspection Activity 5.2 Inspection Findings SECURITY........ Inspection Activity 6.2 Inspection Findings

..

.,_,~

• .

• 4

18

18

19

21

23

24

24

25

. 25.

-Table of-co-ntents (Continued)---- ENGINEERING/TECHNICAL SUPPOR.1 Sa 1 em...

7.2 Hope Cree. SAFETY ASSESSMENT/QUALITY VERIFICATION 8.1 Sa 1 em............

. _

8.2 Hope Creek............

8.3 Individual Plant Examinations..

• - Page -

25

32

34 35 LICENSEE EVENT REPORTS (LER), PERIODIC & SPECIAL REPORTS, AND OPEN ITEM FOLLOWUP 37 LERs & Report.2 Open Item.

EXIT INTERVIEW.1 Resident..

10.2 Specialis.*-.

..

37

38

.~~

..

-,_,*;.*_,.

EXECUTIVE SUMMARY Salem Inspection Reports 50-272/90-1-1;.50-311/90-11 Hope Creek Inspection Report 50-354/9D-08 March 17, 1990 - April 30; 1990 Operations (Modules 71707, 60710, 93702)

Salem:

Numerous operational events occurred during the perio causes are listed in the following table:

Apparent Date(s)

March 27, 1990 March 28, 1990 April 3, 1990 April 6, 1990 April 9, 1990 April 10-20, 1990 Event Unit 1 shutdown -

inoperable safeguards equipment cabinet Unit 1 main steam line isolation Unit 1 reactor protec-tion system actuation Unit 1 feedwater isolations Unit 1 reactor trip Containment radiation monitor spikes Cause(s)

equipment failure spike due to air in lines; poor procedure operator error; poor oversight erratic steam dump operation steam generator feed pump trip due to poor maintenance poor procedure and method for setting background Licensed operator response to these events was good and in accordance with procedure Unit 2 refueling and outage operations were adequately performe Hope Creek:

A marsh fire resulted in a unit scra Operator response was consistent with procedure Reactor feed pump and.reactor protection system motor generator set trips were ad~quately responded to by the operator q

• ,, ___.

..

-

2:

Radio-1 ogica-1 -Cont~o 1-s --(Modu-le- -7-1701)---- _:_ - ----- - - -- -- -- --- -

Salem:

The* licensee responded ade*quately to high rad,iatfon in the Unitl auxiliary building due to a reactor coolant system crud burs Hope Creek:

The licensee identified two technical specification violations:

(1) failure to adhere to locked high radiation area door requirements; and (2)

failure to include an inoperable radwaste effluent monitor in the semiannual

-

effluent repor *--

Maintenance/Surveillance (Modules 61726, 62703)

Salem:

A surveillance test was performed on the wrong trai Poor maintenance on the steam generator feed pumps resulted in a Unit 1 tri Hope Cieek:

Maintenance and surveillance activities were effectivjly performe Emergency Preparedness (Module 71707)

Unusual events declared at Hope Creek and Salem were timely and consistent with emergency plan requirement Security (Module 71707, 92709)*

Strike contingency plans for a po~sible labor action were adequat Engineering/Technical Support (~odules 71707, 90713, 37700)

Salem:

The root caus*e for MSIV slow closure times remains unknow The licensee made a 10 _CFR Part 21 report regarding defective keys in Limitorque valve operators. -

Hope Creek:

Licensee response to and corrective actions for the electrical transient and scram were thorough and aggressiv Licensee actions were adequate with regard to an allegation regarding a breaker failure in 198 Safety Assessment/Assurance of Quality (Modules 71707, 40500, 30703)

Salem:

Failure to perform 2 year procedure reviews and take adequate corrective action associated with these overdue reviews is a violatio Significant Event Response Team (SERT) reviews of the two reactor trips on Unit 1 were timely and thorough.

Hope Creek:. Licensee has taken actions to reduce t,he b.a~klog of overdue 2 yea.r,

procedure review SERT review of the Hope Creek scram was also timely and thoroug,.. __ ;.*.:.-,

,.;'* *

... *

~.:." '

,_

De ta i1 s SUMMARY OF OPERATIONS

.

~ -. *..

,. - '

• -~

1.1 Salem Unit 1 Salem Unit 1 began the report period operating at 92%, limited by an

• -.~,...

..

inoperable heater drain pump (HOP).

Operation continued at 92%

except for minor load reductions (to 80%) due to solar magnettc disturbances (SMDs) on March 20 and 25, 199 The No. 11 HOP was -

repaired and the unit achieved full power on March 26, 1990. - On March 27, 1990, a unit shutdown to Mode'3 (Hot Standby) commenced due to an inoperable emergency diesel generator load sequence The unit was subsequently shutdown to Mode 5 (Cold Shutdown) on March 28, 1990 due to continuing sequencer problem Repair~activities were subsequently completed and a unit startup commenced on March 30, 199 The unit entered Mode 3 on March 31, 199 Due to reactor coolant system leakage into the No. 12 cold leg accumulator, the licensee entered Mode 4 (Hot Shutdown) to perform leak rate tests ~n several check valve Following the valve tests and resolution o~:

the leakage concerns, the unit entered Mode 3 on April 3, 199 Later that day, an automatic reactor protection system actuation occurred due to operator erro The reactor was made critical on April 4, 1990 and main steam isolation valve testing commence The unit was brought on-line on April 7, 1990 and reactor power was increased to 90% (for steam flow transmitter calibration) on April 8, 199 On April 9, 1990, the reactor automatically shutdown due to an equipment problem in the No. 12 steam generator feed pump governo On April 11, 1990, while in Mode 3, the licensee identified that the calculated flow.rate for one of the two intermediate head safety injection pumps for each unJt was greater than the maximum design value of 650 gp The unit was then placed in Mode 5 on April 13, 1990 so that a full flow discharge and flow balance test could be performe The intermediate head safety injection pumps were teste However, further flow distribution and pump capacity discrepancies were identified when the high head charging pumps were tested, and the unit remained in Mode 5 until the end of the inspection perio Salem Unit 2 Salem Unit 2 began the report period operating at full power, and continued with the exception of load reductions to 60% on March 20 and 25, 1990 due to SMD On March.31, 1990, the unit was shutdown and commenced its fifth refueling outage, The outage is scheduled

. -.

,..

for 55 day Mod~_5_was-entered on A~ril 1, 199 Mode 6 (Refueling)

was entered on April ;16. 199 The core offload was completed on -

April 24, 199..1-

• ':,* *.*

,_,

1.3 Hope Creek The unit began the report period at 98%1Mith power limited.:,;as*'a result of the 11 lC 11 and "2C 11 feedwater heaters being isolated due to,a leak in the "2C" feedwater drain coole The Hope Creek.. untt experienced an eight day forced outage due to a seram caused by ca fire in the surrounding marshes (see section __ 2.2).

During the outage, the drain cooler was repaired, and the unit returned to 100%

powe The unit remained operational throughout the remainder of the inspection perio Power reductions occurred to accommodate maintenance and testing, and*-also as a precaution for solar magnetic disturbance.

OPERATIONS (71707, 93702, 60710).2 2. Inspection Activities The inspectors verified that the facilities were operated safely aTid in conformance with regulatory requirement Public Service Electric and Gas (PSE&G) Company management control was evaluated by direct observation of activities, tours of the facilities, interviews and discussions with personnel, independent verification of safety system status and Technical Specification Limiting Conditions for Operation, and review of facility record These inspection activities were conducted in accordance with NRC inspection procedures 71707, 93702 and 6071 The inspectors performed normal and back shift inspection, including deep backshift inspection as follows:

Unit Ins~ection Hours Dates Salem 10:00 p.m. - midnight 3/19/90 3:00 p.m. - 8:25 /31/90 Hope Creek 10: 00 p.m. - midnight 3/19/90 9:30 a.m. - 8:15 /25/90 Inspection Findings and Significant Plant Events Salem Engineered Safety Features (ESF) Actuations Caused by Radiation Monitoring Systems (RMS)

Several ESF actuations occurred during this inspection period initiated by the Unit 1 and Unit 2~MS. *In each case, the licensee adequately responded to the event, acknowledged the isolations, rep~ired or restored the' RMS instrument as approp-riate, made an emergency notification *system (ENS) call and promptly informed the resi~ent.insp~ctor.

. '.*

. *". -.. '.~.,. '

.. :*; -*

The first event occurred at~nit ~ on April 3, 1990 at 1:45 a.m., when a containment ventilation isolation (CV!} resulted _

after an auxiliary operator inadvertently"bumped an electri'cal;;_ :.'

breaker that in turn, actuated,..the., 2R4J.. radiation monitor This was a personnel error and wtll be submitted to-the NRC as a*

separate licensee event report. * *

The following tabulation.summarizes primary containment particulate and noble gas ESF actuations:

Date Time Unit Rad Monitor 4/10/90 9:38 R12A 4/15/90 3:00 R12A 4/15/90 10: 59 R12A 4/16/90 10: 59

2R11A

, /17/90 5:15 R12A 4/19/90 8: 10 R12A 4/20/90 12:50 R11A The above ESF actuations for Unit 2 were related to resetting the trip setpoint to 2 times background as required by Technical Specifications Table 3.3-The licensee concluded that a random spike would periodically exceed the trip s~tpoint. A review of background level determinations was undertake The licensee redefined backgro~nd countrate based on expected values and standard deviatio The instruments were recalibrated and retested satisfactoril The inspector reviewed the licensee 1s process and had no further question At 11:00 a.m. on April 30, 1990, a control room ventilation occurred when radiation monitor lRlB spiked inadvertentl.

Unit 1 Shutdown Required by Technical Specifications On March 27, 1990, a Unit 1 shutdown was initiated as required by Technical Specifications (TSs) due to an inoperable lA safeguards equipment control (SEC) cabinet subsyste The SEC provides 4 kV vital bus lA load shedding and sequencing functions during accident condition At 9:30 a.m., following the completi~n of a surveillance test for the lA SEC, Procedure No. M38, 11SEC System Sequencer Step Timing Test 11, an automatic actuation of the lA SEC occurred for no apparent reaso The lA SEC.cab.;inet,door had been closed and the test was completed 3-4 minutes prior to the spurious actuation. The techni~ians'had left the work are The SEC actuation caused the-norr-vita*l**loads to be shed from the 11A

..,..,.,..

,.. :.;:--

.*. ~--r**.

.:.:

.. '.-.

• vital bus and -caused the-actuation-6f--safety equipment such--a"s-the 11A 11 emergency di e.se 1 generator {started, but did not load),

No. 11 residual he.at removal pump,and. No. 11 safety. injection pum The reactor was operati,r1g at,100% power and the.re was __ no safety injection flow to the reactor* coolant system as a result of the actuatio The equipment that actuated was subsequently restored to a normal standby condition by plant___operators and the lA SEC was declared inoperable as of 9:30 a;m. on March 27, 199 Technical Specification (TS) 3.3.2.1 requires that with one SEC inoperable, the unit must be placed in Mode 3 (Hot Standby)

within 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and in Mode 5 (Cold Shutdown) within the following 30 hour3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br /> The licensee initiated troubleshooting activities immediately. However, since a root cause was not immediately apparent and the SEC was to remain inoperable, a unit shutdown to Mode ~was initiated at 10:40 The licensee notified the NRC via the ENS of the initiation of the unit shutdown required by TSs and of the Engineered Safety Features (ESF) actuations in accordance with lOCFRS0.72 reporting requirement Maintenance personnel replaced the chassis for the lA SEC with a spare chassi While returning the SEC to service, a second SEC actuation occurred at 11:39 a.m. on March 27, 199 Most of the same equipment was actuated as before except that certain equipment such as service wate~ system valves did not actuate because maintenance technicians had immediately reset the SEC during the process of placing the SEC in service per procedure M3B requirement The second actuation occurred while at 75%

reactor powe This second ESF actuation was also reported to the NRC Via ENS as require The inspector observed portions of the licensee's troubleshooting activities. A troubleshooting action plan was developed by the responsible system engineers. Activities included reinstalling the original chassis following a satisfactory bench test, re-placing _the 15-volt logic power supply, inspecting selected relays, checking/repairing various SEC connectors and connecting a chart recorder to various SEC component A source of electrical noise was detected on the chart recorde While investigating the source of this noise, the installed original chassis sustained a 11 hard failure 11 on March 28) 199 That is, the local SEC auto test fault light illuminate Since the appropriate ESF equip-ment and SEC. output were tagged out_ of service for the trouble-shooting, no--additional ESF actuations resulte The Jicnsee

.

did not know :why-the *second ESF *actuation occurred with :the spare chassis installed. A.third chassis was installed following satisfactory performance of surveillance procedureM3B in.an energized bench test rac *.**:

.***.

• • -.~~-.

..

-'. ~*

.

• .. *~:...,.:.

• - During the 1 icen see 1-s--troub 1-eshoot-i ng *act i-vi ti es-,--'a-- root--cause-- -~*--: --- --- ----- to the event was not ~onfirme On January 29, 1990 (NRC * *

_

Inspection Report No. 50-272/90-04),- a similar lA SEC problem.,,.,;:..,*; had occurre However, there were.. no.ESF actuations--associated.... *

with that even *

Following maintenance activities on March 28, 1990, the chart recorder was connected to *monitor the SEC to assis-t in identi-fying the root cause in the event an additional failure were to occu The recorder installation was classified as a temporary modification, and was reviewed by the Station Operations Review Committee on 1March 29, 199 The inspector attended the meeting~_

~nd no deficiencies were note The inspector participated in a ~~

conference call with Region I on March 30, 199 The license~

explained their actions and their determin'ation of SEC operabi*-lity.-~

The inspector concluded that the li~ensee took the appropriate actions to comply with the TS requirements. Although the licensee was not able to reproduce the event or positively identify the root cause, the potential suspect components were ins~ected, tested, or replace Additionally, a chart recorder was temporarily connected to assist in SEC monitorin The inspector had no further questions at this tim Removed equipment from the SEC, including the spare chassis that was subsequently removed, has been sent to the manufacturer for testing and analysi The licensee has indicated that the SECs for each unit will be upgraded during future scheduled refueling outage period Unit I Main Steamline Isolations On March 28, 1990, while in Mode 4 (Hot Shutdown), a partial main st~am line isolation (SLI) occurred at Unit 1 during surveillance testing at 6:35 A second complete SLI occurred at 8:54 p.m., also during surveillance testin *

Instrument an~ Control (I&C) technicians were in the process of performing sensor calibrations to the steam generator (SG) flow transmitter The No. 14 steam flow channel was out of service for calibration when the No. 12 steam flow channel momentarily spiked hig This action satisfied the two out of four logic which caused an SLI for all four steam line The main steam isolation valves (MSIVs) were previously closed, but the MSIV bypass valves were ope Only two bypass valves closed, those associated with loops 13 and 14, whi.le. the remaining bypass valves (11 and 12) remained ope This is not the expected performance for the bypass valve The partial SLI did not result in a plant operaMona*l transient, however, the.reason *'for.;.

the partial actuation was_,not,known immediately by station

~;~

personne The ESF actuation was rported the to NRC via ENS-.fo accordance with 10CFR50.72 reporting require~ent '>>~**' ;

. *... ;._;,,

-.

~**** *,.* *; *:. *

.B

--* - --. -*- Tne -urrfFcorrEfnued 'its*--sliLi"td6wrf to *Mode-* 5-(Cold *sh-Litaa*~n)" due*--to:~* /:.- ---. -*-

a diesel g~nerator sequencer problem::(Section 2.2.1.A). -*lhe

.

.

1 ;*censee elected* to... continue *_with*-*-stearn **flow transmitter

-.

..-r~:::-.*-; <**

calibration However, with,thecNo.... <11*~.. cStearn flow channel o.ut of service for its calibration, another channel spike occurred~.

(No. 14) at 8:54 The secohd isolation resulted in the

..

automatic closing of the remaining MSIV bypass valves (11 and 12').

-:..>:

_,,, *-*

On March 29, 1990, the licensee conducted troubleshooting*

. -- -

• .:.-

activities, including the injecti6n of input s~ikes at the p6inf*

where steam flow for -the No. 12 channel would normally be sense The input spikes varied from approximate1y 10,-to 45 mill i seconc;l_s __ _

(ms).

The results indicated that *fhe number of bypass MSIVs that closed changed as the spike duration increased.. Fo~-~xamplf:!_~*~

an 8 ms spike resulted in two valves receiving a close s-igna::r;--a-**~-

10 ms spike caused one, and a 44 ms spike resulted in a 11 four-e:

valves receiving a close signa The licen~ee therefore concluded.;~~

that the partial SLI was due to a spike that had not sealed~in due to its extremely short duratio *

Further troubleshooting activities indicated that in both SLI events, the channel that spiked had just been returned to service following its sensor calibratio The licensee concluded that the most probable cause of the spikes was due to air trapped in the sensing legs during calibratio The air bubbles rising in the legs would expand as the pressure head above them decreased, displacing larger amounts of water as the bubbles approached the_

condensate pot The licensee postulated that if air was trapped in the low pressure leg, a momentary hi'gh differential pressure signal c6uld re~ult when the bubble reached the surface.*- The action of the bubble breaking when it reached the surface and the increased differential pressure due to level loss in the leg could cause the channel to momentarily spike into the alarm stat The licensee typically performs the steam flow calibrations at either steady state full power operation or in Mode 5 (Cold Shutdown).

The calibrations performed on March 28, 1990, were performed in Mode 4 while ~pproaching Mode The licensee stated that the transient 'condition of the plant may have contributed to the even The inspector monitored the licensee's actions associated with the even No deficiencies were identified~ The inspector

~articipated in a~conference catl with Region I on March 30, 1990:

During this call, *the licensee discussed their actions taken a*nd... :.

what assurance they_had that the SU logic was operabl The

.

licensee reviewed tne'~'*calibration procedures and determined that ~"*

changes to ensure the'*instrument is* backfi 11 ed are appropr.i at '*.*

to prevent the accumulat*ton,.of.air in.. the sensing lines.. and thereby prevent recurring incident The inspector had no

. - '=-,~

,* _.,. :

~-.. -....

I I further questions at this time and concluded that the lfce~see 1 s troubleshooting activities and proposed corrective actions;were acceptabl.,

.,..

• • *

Unit 1 Reactor Coolant System Leikage During Plant Startup

-* .** ~-.,....... >* ~*

On April 1, 1990, while in.Mode: 3 (Hot Standby-) and preparing for startup from the Marcb 27, 1990 shutdown, Unit 1 experienced reactor coolant system (RCS) leakage into the No..12 cold leg

accumulato The leakage was estimated to be about 1 gpm and possibly through several check valve The flow path for the -

leakage was not know The accumulator isolation valve was closed in order to minimize accumulator in-leakage and borated

-=

water volume dilution, and the appropriate Technical Speci fi cati*on -

(TS) Limiting Condition for Operation was.entered at 9:18.a~m..

The licensee commenced a unit cooldown to Mode 4 (Hcit Shutdown)

to find the source of the leak rate by performing check valve integrity surveillance test SP(0)4.4.6.:3, 11 Emergency Core Cooling System - ECCS Subsystems.

Mode 4 was entered at 2:20 p.m. and the TS Action requirement was terminate The licensee then performed leak rate tests for several check valves as directed by the surveillance procedur None were identified as having leakage rates in excess of acceptance criteria, and the previously identified RCS leakage had abate The licensee attributed this to possible seating of the affected check valves during the testing proces The inspector observed portions of the licensee 1s troubleshooting and testing activities and no deficiencies were identifie The unit was returned to Mode 3 at 1:30 a.m. on April 3, 1990 to continue with startup preparation No similar leakage problems were subsequently encountere Unit 1 Reactor Protection System Actuation On April 3, 1990, a Unit 1 automatic reactor protection system (RPS) actuation occurred while in Mode 3 (Hot Standby) due to low-low steam generator (SG) water leve The reactor was subcritical with the shutdown bank control rods fully withdraw Plant operators were in the process of transferring steam control from the No. 11 to the No.. 12 SG atmospheric relief valv The transfer caused an increased steaming rate on No. 12 S However, the feedwater flow that was being supplied by the auxiliary feedwater (AFW) system, was not increased sufficiently by the reactor operator-to prevent the.w.ater level in the No. J.2...SG reaching its low-low reactor trip setpoint (16%).

The licensee notified the NRG of this event via ENS in accordance with

~**-

10CFR50.72 reporting requirement *--.,-

• -,-'

.;' ~'

• The licensee initiated a Significant Event Review Team-fSERT) to review the circumstances surroundinithe RPS actuation. 'The SERT i dent ifi ed the root cause of tbe,.event to be poor -Judgement..,*, *.,

on the part of the licensed operator combined with weak command and control on* the part of the licerised senior reactor operator (supervisor).

The licensee *stated *that the reactor operator failed to establish favorabJe conditions on No. 12 SG before initiating the atmospheric dump valve transfe Specifically, SG water level was at about 28% and had indicated a downward tren Normal water level is 33% with a 5% operating ban The-licensee also concluded that the senior reactor operator who was in the control room observing the a~tivities, had failed to order a more appropriate operator respons The AFW flow to the No. 12 SG was being manually controlled via a loop flow control valv ~onsole indicators available to the operators included AFW flow control valve demand and actual flo Although zero flow indicated on the AFW flow indicator and SG level was decreasing, the operator maintained a relatively constant valve position, with only minor open valve demand Both the licensed operator and supervisor recognized that, for unknown reasons, the loop 12 flow indicator did not indicate any flow until demand was near 20%.

A work request was previously initiated on March 28, 1990 to address this same problem; that no AFW flow is indicated during low flow conditi6n This was unique to loop 12 and the operators incorrectly believed that they were adequately providing sufficient AFW flow to the No. 12 S The Station Operations Review Committee (SORC) performed a root cause determination for the event similar to that of the SER The inspector attended the SERT debriefing and SORC meeting on April 4, 199 During a followup review of this event, the inspector concluded that the operation of the No. 12 AFW flow control system may have significantly impacted operator response to this event in an adverse manne The inspector reviewed the licensee's SERT report, and found that the abnormal response of the No. 12 AFW flow controller was considered to be a contributor to the level control problem The licensee had initiated additional efforts to troubleshoot the control system, and had subsequently replaced a circuit card that corrected the proble Previous troubleshooting activities had not identified any operational problems and did not result in timely resolution of the flow indication discrepancie This was the first un*it startup since increasing the low-low"SG reactor trip setpoint (.on both units) from 8.5% to 16% to

..

support a recently revised.. se,tpoi.nt methodolog The operators on shift at the time of the event were aware of the increased * *

setpoint Licensee*corrective actions for this event included briefing all oncoming crews of the event and the-importance of establishing favorable plant conditions, and -0f the necessity.:to respond to operabl~ ~antral room indi£ations~

~he licensee also directed that only experienced licensed operators, specifically trained at the simulator with the T6% trips, are to.be at the feedwater system ~ontrols ~uring plant-startup The inspector concluded ~that the SERT comp 1 eted a thorough and complete review of the even Prior attempts by maintenance personnel to resolve problems with the AFW flow controller circuit were ineffectiv Additional troubleshooting activities were conducted on the No. 12 AFW flow control circuit. A square root extractor was subsequently replaced which resolved the

control room AFW flow indication problems on.April 6, 1~90. The inspector had no further quest i ans at this tim *

Unit 1 Feedwater Isolation During Plant Startup On April 6, 1990, during~ plant startup, a feedwater isolation (FWI) occurred at Unit 1 with the reactor operating at 7% power;*

While controlling steam pressure using the turbine steam dump system valve 12TB10 exhibited erratic behavio The valve quickly opened and then reclosed, resulting in a severe water level transient in the No. 12 steam generator (SG).

The level increased to the steam generator high-high setpoint of 67%

(narrow range) and the FWI actuate The FWI isolated feedwater flow to the SG This ESF actuation was reported to the NRC via ENS in accordance with lOCFRS0.72 reporting requirement The programmed SG water level for the existing plant conditions was 33% narrow rang Operations personnel indicated that level was being maintained slightly higher than 33% just prior to the event, thereby decreasing the operating margin to the FW The licensee performed a packing adjustment on steam dump valve 12TB10 and lubricated the valve stem before resuming startup activitie The unit was.subsequently placed on-line on April 7, 1990, and reached 90% power (for steam flow transmitter calibrations) the following mornin The inspector reviewed the licensee acttvities associated with this event and no additional deficiencies were identifie : ~-...

Unit 1 Reactor rrii:~n Apri1 9, 1990 On April 9, 1990~ a ~ni~ 1 reactor trip occurred from about 90%

reactor power due*;to*'low'""low steam generator (SG) water level on the No. 12 S The No. lZ steam driven SG feedwater pump drove*

to idle speed due to separation of the turbine governor servo- * ~

motor linkag The servo-motor amplifies the signal output from

• '.. *.,,.

-

--:;:

the governor linkag~*to the turbin~'s steam admtssion valve assembly. Dperato.rs responded*'.to the transient by i-niti.ating a rapid load reducti6n and plac~ng~the rod control system in* the automatic mode of *operatio The low-low SG water level was reached on the No. 12 SG, resulting in the automatic rea~ctor tri The inspector.responded*te the *control room fo*llowing the tri The licensee reported the ev1ant to the NRC via ENS in accordance with 10CFR50. 72 reporting ::requirements..

The licensee assembled a Signifi~a~t Event Response Team (SERT)

to independently assess the tri During licensee followup of the event, it was determined that there was an app~rent abnormal rod control system (RCS) response when the operators placed the RCS in the *automatic mode of operatio In that mode, a slower than normal insertion rate*of the rods was *suspected by the operator The licensee.performed extensive ~esting activities, including functionally testing the rod control syste No si gni fi cant/relevant ope rat i ona 1 defi-ci enci es were detected by the 1 i cen se Si nee the unit was subseq*uent ly shutdown to Mode 5 the licensee continued with troubleshooting efforts. At the end of the inspection period, abnormalities related to the operator's observations of the rod control system were not identified.

During SG feedwater pump investigation and repair activities, the licensee identified that a pin bushing in the linkage assembly was missin In addition, the lock nut that previously maintained the proper linkage connection was found to be incorrectly installed such that the locking side of the nut (flat side) was installed backward The last governor/linkage alignment was found to be conducted in 1986 as determined by a review of historical work order dat The licensee determined that four repairs had been.accomplished to the governor in 1989, however, none of them should have directly affected the above linkag NRC Inspection 50-272/90-200 provides further followup to these maintenance related issue The licensee conducted a Stati~n Operations Review Committee (SORC) meeting on April 10, 199 The inspector attended the meetin The startup issues discussed were the repair of the No. 12 feedwater pump linkage, inspection of the No. 11 feedwater pump linkage, investigation of~a potentially abnormal rod control~system response, review of a*~ost-event problem identified ~egardlnif erratic operation of an intermediate range neutron detector, ~nd repair o(,an atmospheric dump valve controller that had experienced a minor automatic control problem about 10 minutes~following the trip. All problems were repaired/resolved'by the licensee.*

..,.....

Th~ _i_nspe~tor __ reviewed'the lic_ensee~s post-trip-activities,--

including a review_of SERT and SORC effe-ctivene-s The inspector concluded ;,,that. the root.c-ause,,-0.fa,the event can be attributed to the-SG feedwa~er.pump_m.issin_g.bushing and the incorrectly installed locking nut.,The SERT determined that the root cause wa:s,-unknown*'a;nd-the SORC':-identified the root cause of the trip to be,SG-feedwater pump linkage separatio Further review by the.. li censee determined the root cause to be poor maintenance due to inadequatetcorrective and:~reventive maintenance procedure The proposed corrective actions were adequate, including *inspecting-similarly configured component The inspector will continue to monitor the effectiveness of licensee event evaluations, with particu_lar attention to root cause determination and documentatio *'*-* -

~- ::

---

,__

~-. --

. ;:...-..*.. Salem Unit 2 Midloop Operation On April 10, 1990, in order to perform maintenance on the steam generators, Salem Unit 2 entered midloop operation In midloop operations, the reactor coolant level is lowered to the midpoint of the reactor vessel hot and cold leg nozzle In this state of reduced inventory, additional instrumentation and monitoriTig of reactor water level is required to ensure proper core coverage and coolin Licensee procedures require thermocouples to be used to monitor core exit temperatures and intermediate leg loop flow differential pressure cells be used to measure water leve Vessel water level is also measured visually by use of transparent tubing connected to an intermediate leg loop_

drai Temperature and level alarm set points are adjusted to provide early indication of a loss of cooling or a decrease in coolant inventor Additional monitoring and logging of these parameters is required as wel Shortly after the reactor vessel water level had been established at the midloop point, the inspector toured the plant in order to review the licensee's control The inspector determined that the required instrumentation had_been installed and all monitored parameters were indicating in the safe rang Through discussions with the control room opera*tors, the.. inspector found the operators knowledgeable of ~resent plant c~nditfons, the indications available to them, and of the procedures to be followed if core cooling were to be los The inspector also reviewed the control room-logs and found* them to be complete and satisfactor Based on his tour, the inspector concluded that midloop operations had been reached and wa.s being ma.i,n:tai,ned in a safe and proper manne Salem Unit 2 Def.ue'ling --

---

'

.

The Sa 1 em Un it 2 reactor,,.core was offloaded into the spent fuel....,, -:

pool beginning April 20 and defueling was completed on April 24,.

199 The inspector veriffod that _reactor operators were

• ,....

• . A.

• .

,*; -*

knowledgeable of defueling acttvitie Contractor personnel in charge of core offload were also interviewe The bundle pull sheets were checked and the core status board was veriHed,_:to,-be *

accurat Appropriate refueling procedures and technical specifications were also reviewe No unacceptable conditfons were note Salem Licensed Operator Staffing At 8:10 p.m., on March 17, 1990, due to a family emergency, an unexpected absence occurred for the Unit 2 reactor operator (RO).

Unit 2 was operating at full powe The licensee responded by calling in a replacement RO who arrived a~9:45 The inspector verified that these actions were in

~-

accordance with Technical Specification Table 6.2-During the inspection period, the licensee added 6 ROs to shift rotation who recently passed their licensee examinatio These RO additions have added one operator for each of the 5 operating shift The licensee now has an extra RO for each shif Hope Creek Reactor Scram on March 19, 1990 At 6:50 p.m., on March 19,1990, the Hope Creek unit experienced a low reactor vessel water level scram which was caused by the loss of all feedwater and condensate pumps in response to an electrical transien Level decreased to less than -38 inches and the high pressure coolant injection and reactor core isolation cooling systems automatically started to recover 1eve1.

In accordance with station procedures, an Unusual Event was declared at Hope Creek from 7:00 p.m. to 7:25 p.m. due to the high pressure coolant injection initiation and injectio The electrical transient occurred when an offsite marsh fire produced a phase to phase short in the 500 KV Deans transmission line leaving the Salem switchyar As a followup to the event, the-licensee conducted a Significant Event Response Team (SERT)

and had the Nuclear Department electrical engineering staff investigate various plant response A second Unusual Event was declared from 7:35 p.m. to 8:35 p.m. for the entire Artificial Island Compl~x due to a fire lasting more than 10 minutes which resulted in a mode change at Hope Cree Salem Unit 1 was operating at 92% and Salem Unit 2 was operating at 100% prior t~

the inciden The event had no impact on power operation of either Salem uni The.. marsh fire was fought by both the,-onsite *

and offsite local (Lower ~lloways Creek) fire departments~ The -- *

...,*.-

fire was ex ti ngui shed by some counter burning and eventually by *

a rain storm which occurred about ~0:30 The licensee elected to keep H~pe Creek shut down for approximately one week to repair "the,lC feedwater heate ~ : ;:.

'

On March 26, 1990, the engineering staff-presented its conclusions to the Station Operations Review Committee (SORC)

prior to the restart of the reacto The team concluded that the plant responded as designed with the exception of a~ operator aid indication in the control room-for feed pump statu The electrical transient caused a 50%.voltage dip on the 500 KV line for approximately 4 cycles. A low voltage tri~ of the 7.2 KV and 4.16 KV busses did not occur due to the large induction loads on these busses which tended to maintain voltage during the momentary di The voltage drop was passed on to lower voltage 120 VAC control circuit The degraded voltage condition on the 120 VAC busses produced.false process control input signals for the condensate pumps.. Specifically, the discharge valves of the primary condensate pumps indicated closed, and the secondary condensate pumps received a low lube oil pressure signa These false signals tripped all condensate pumps which, in turn, tripped the feedwater pumps and produced the low reattor water level conditio The test results and conclusions of the engineering team were reviewed and concurred upon by both the licensee's SORC and SER Following SORC review of the root causes of the event, the plant restarted on March 26, 199 The inspectors responded to the station on March 19, 199 Both Salem and Hope Creek control rooms were toure The inspector verified that the Hope Creek unit was stabilized in the hot shutdown conditio The shift operators were interviewed, control room logs and chart recorders were reviewed, and sequence of events printouts were examine The inspector concluded that immediate licensee actions were appropriat The inspector also toured the fire area and interviewed security and fire fighting personne *

Further followup included a review of GETARS printouts, the post trip review procedure, the LER, and incident report (Also see sections 7.2.A and 9.1).

-..;,

..,.-*- -..

The inspector observed the primary containment closeout of the drywel 1 which appeared orderly and. ready for reactor restar The inspectors observed portions of the reactor startup

• '.~*-

• • '******

including:

• '*****

OP-GP.ZZ-002 Primq._ry.Containment

• c1 oseout

• OP-IO. ZZ-0003 OP-SO.AE-0001

Startup "from Cold Shutdown to-Rated Po~er Feedwater,System Operation Criticality was achieved at 4:14:a.m. on ~arch 27,1990, and the plant returned to 100% power operation on March 28, 199.

. Reactor Feed Pump Trip On April 3, 1990, the 118 11 reactor feed pump turbine trippe Consequently, reactor water level dropped to the 31 inch level before the 11A 11 and 11 C 11 pumps increased speed to account for the 1 oss of 118 11 pum Reactor water 1eve1 was restored and, in fact, reached a maximum at a level of 39 inches before it stabilized at the normal level of 35 inche Upon investigation, the senior nuclear shift supervisor determined that, at the time of the feed pump trip, an instrumentation and control technician was performing the channel 11 C 11 reactor pressure vessel narrow range level 8 trip surveillanc The licensee organized a fact-finding team to determine a root cause for the even The team concluded the most likely cause of the event was a spurious trip of an additional 118 11 feed pump trip relay while the technician was conducting the surveillance on the 118 11 feed pump trip relay associated with the 11 C 11 level 8 trip channel. If two relays had been in the tripped condition, the necessary logic coincidence would have been satisfied, and the 118 11 feed pump turbine would have been trippe As a conservative measure, the licensee replaced both 118 11 feed pump trip relay units involved in the event and, as a precautionary step, initiated a design change to extend the test jacks used in the logic cabinet in order to preclude the possibility of any future personnel erro The inspector reviewed the associated incident report and discussed the event with the Hope Creek maintenance manager, a member of the fact-finding tea The inspector concluded that the licensee had responded in a conservative manner and that no unresolved safety issues remai Containment Isolation Valve Inadvertent Closure

.During the morning of April 9, 1990~ a control room operator noted that valve 88-SV-4311 had inadvertently close The alarm chronolog showed thaf*the valve (reactor recirculation sample outboard containment isolatton valve) had been* shut fQr approximately 15 minutes.. The operator attempted to reopen the valve from the control room with no succes The

~ *..

• instrumentation and control, and electrical maintenance*

department was notified in order to repair the proble U~on '

investigation, the licensee determined***that*the control power for the valve had been de-energized during maintenance on_valve BC-SV-F079A, the residual heat removal loop 11A 11 sample valve. The control power fuse for the 4311 valve had been incorrectly wired at the terminal board common,to the control power for the*:

two valve The hot side of the control power circuit for the 4311 had been wired in a series circuit with the fuse for F079 When the F079A fuse was pul 1 ed, 4311 was de-energized and subsequently close Maintenance department inspection determined that the w1r1ng problem at the terminal board had existed since the terminal board was originally wire The fuse for F079A was pulled in*

conjunction with a 5-year environmental qualification (EQ)

rebuild of the same valv This was the first time such work was done in association with the faulty terminal boar There were no precursors to this even The licensee inspected selected similar circuits and no deficiencies were foun The inspector reviewed the applicable electrical drawings and determined that they were correc The terminal board has since been rewired correctly, and after discussing the event with the senior electrical maintenance supervisor, the inspector concluded that the licensee's response and corrective actions were appropriate and sufficien The inspector had no further question Reactor Protection System (RPS) Motor Generator Set Trip At 7:40 a.m. on April 18, 1990 while at 100% reactor power, Hope Creek had an engineered safety feature (ESF) actuation due to the loss of power to the B RPS Motor Generator (MG) Se This resulted in the subsequent loss of power to RPS channels B and D, and a half scram condition. Additionally, the Nuclear Steam Supply Shutoff System (NSSSS) isolated the reactor water cleanup system (RWCU), the reactor recirculation sampling system and the main steam supply drain valve The cause of the loss of power to the RPS MG set was a ground fault on Motor Control Center (MCC) 008-58 This non-safety related MCC feeds the RPS motor generator, as well as other smaller loads including: radwaste pumps, two welding receptacles, and hydraulic pumps associated with the cooling system isolation valve The MCC was entirely lost as well as all of the above mentioned load Power was immediately restored to B RPS by placing "the *selector switch to the alternate power suppl The RPS logic was rese~-

and systems associated with the NSSSS were unisolated within two minutes of the even The licensee made an ENS cal. -.

~::*;,;..

Ground isolation procedures were commence All MCC 008-582 loads were meggered with no discrep~~cies identifie No welding receptacles associated with,-thi:s,;MCC were being utilized at the ti me of the even **..,*.

The breaker in the MCC was replaced, and the power _supply tcrtbe B RPS bus was shifted back to the MG~et without inciden Inspection and testing of the breake~-that had tripped on the ground fault was conducted by the Hope Creek electrical maintenance and electrical engineering staf No problems wer~

identifie The licensee concluded that the cause of the event was a spurious trip of the solid state tri~ device (SST) in the

~

breake In discussing the matter with an electrical maintenance supervisor and a senior staff engineer, the inspector learned a similar SST had experienced the same type of trip approximately a year ag That SST was returned to the vendor, General Electric, for further testing and evaluation, and Hope Creek Engineering is still awaiting the result The licensee has deferred further testing of the second SST and breaker pending the arrival of additional information from General Electri The inspector concluded that the licensee acted prudently in replacing the MCC breaker and its SST following the spurious trip and, now that a second similar trip has occurred, believes that a resolution of spurious SST trips is being more aggressively pursued by the Engineering Departmen The inspector will follow up on this event when additional information is availabl.

RADIO LOG I CAL CONTROLS ( 71707, 93702) Inspection Activities PSE&G's conformance with the radiological protection program was verified on a periodic basi These inspection activities were conducted in accordance with NRC inspection procedure 7170.2 Inspection Findings and Review of Events 3. Salem High Radiation Area in Unit 1 Auxiliary Building During the morning meeting on April'-16, 1990, the inspector learned that radioacttve crud (Cobalt-58) was deposited in the safety injection (SI) system piping in._Unit 1 auxiliary building on April 14-15, 1990 during SI pump testin The inspector verified that the licensee controlled the affected areas as high

3. *

radiation area Tours of the areas were conducted on April 16 and 17, 199 The inspector reviewed:survey data, the 5pecific **

radiation work permit, and interviewed.~adiation protection~and chemistry personne The licensee initiated actions to flush the p1p1ng. These actions were successful to reduce the radiation level The inspector concluded that licensee actions were appropriate in*

response to this even Hope Creek High Radiation Area Doors Radiation protection technicians at Hope Creek check all locked high radiation area doors once per shift to ensure that the doors are indeed locke On the morning of March 15, 1990, the radiation protection technician performing this check found the door to the 6C feedwater heater room unlocke The door was closed, but because the radiation level in the room is greater than 1 Rem/hour when the plant is at power the door is required by Technical Specification 6.12.2 to be kept locked.. The radi-ation protection technician immediately locked the door from the outside using his master key.

Upon investigation, the licensee determined that the turbine building 137 1 elevation master key had been checked out and returned by an equipment operator earlier the same morning.* The radiation protection supervisor on shift notified the resident inspector and began an investigation into the cause of the even When questioned, the equipment operator stated that as he exited the room the latch on the door stuck, but when he turned the inner knob the latch unstuck and he was able to shut and lock the doo In order to determine the cause of the latch sticking, the licensee replaced the latch on the door and had the original latch dismantle Inspection of the door latch internals revealed a small burr that might have caused the latch to jam, giving the impression ~hat the door was indeed locked, but the licensee concluded that it was most probable that the method the equipment operator used to check the door was not sufficient to reveal the door was not properly locke As part of the investigation of the event, the Radiation Protection Department reviewed all exposure records for the morning of the event to ensure that no one had made an inadvertent entry into the roo The review showed no abnQrmal exposures and it was determined that the high radiation area had not been entered while the door was unlocke *

As a result of this event..,the licensee initiated the following corrective actions:

.:..

>

• ** ; *-.~~;;

• . :-***-**

~

.

• *.

.. :.

1*** *

The equipment operator's self monitor radiation protection qualification was suspended pending requalification and an interview with the radiation protection engineer and"the radiation protection superviso *

Each time a locked high radiaticin area door key is checked*

• out and returned, an independent verification of the door being locked is required to be performed.by a second operato The locked high radiation area log has been changed to

~~ *

require an initial by the radiation protection technician

• .,

who checked the doors locked during the normal shift round The inspector tracked the progress of the investigation of the event and reviewed the findings as they develope The licensee's response to the event was aggressive, and the conclusions reached were accurat The inspector concluded that the corrective actions tak~n were sufficient to prevent a recurrence of the even This is a licensee identified violation of Technical Specification 6.12.2 (50-354/90-08-01).

Containment Tour The Hope Creek drywell was inspected on March 26, 199 No unacceptable conditions were note Radwaste Effluent Monitor The licensee reported that the radwaste effluent radiation monitor was out of service for more than 30 day The monitor was operable but its isolation functions were bypasse This was not reported in the plant's semi-annual radioactive effluent release report as required by Technical Specification (TS) 3.3.7.1 Apparently personnel making up the report looked at the component as being in operation but they did not look to see if it was fully fun ct i ona The reportabi 1 ity of this event was discussed with the NRC on April 24, 1990, and the licensee concluded that this event was reportable in accordance with 10 CFR 50.72 requirements at 2:00 p.m. on April 25, 199 Liquid releases were made to the radwaste system when the radiation monitor was not fully operabl In these instances, two independent samples were analyzed of the releases made to the radwaste system and two independentc people :verified the.~.velease. -

rates of each discharge to the radwaste system which is in accordance with TS 3:3.7.1 The inspector reviewed the even~~and concluded this was a licensee identified violation of TS 3.3.7.10 (50-354/90-08-02).

• *

MAINTENANCE/SURVEILLANCE TESTING *(62_703, *6172'6)

4.1 Maintenance Inspection Activity The inspectors observed selected maintenance activities on safety-related equipment to ascerta i n**that these activities were conducted in accordance with approved.procedures, Technical Specifications, and appropriate. industriBl.cbdes and standard These inspections were conducted in accordance with NRC inspection procedure 6270 Portions of the following activities were observed by the inspector:

Unit Salem Hope Creek Work Request/Order (WR/WO)

or Procedure WO 900327193 SC.IC-GP.ZZ-006(Q)

WR 0088183 WO 900106105 M6G/WO 900413098 WO 900423104 WO 900424168 WO 900315108 Description Replace lA SEC chassis with spare Troubleshoot lA SEC Investigate reason for partial steamline isolation Installation of sequence of events recorder per DCP 2EC-2272 No. 12 charging pump repair electrical systems troubleshooting Scram solenoid pilot valve replacement 118 11 main steam line radiation monitor power supply replacement 11 C 11 circulating water pump replacement With the exception of poor maintenance on No. 12 steam generator feed pump (see section 2.2.1.G and NRC Inspection 50-272/90-200), the maintenance activities inspected were effective with respect to meeting the safety objectives of the maintenance progra *.

.,:**

4.2 Surveillance Testing Inspection Activity

..

~-.

,. '

The inspectors performed detailed technical-procedure reviews,-c

':,

,,,

witnessed in-progress survei 11 ance testing, and reviewed complete.

surveillanc~ package The inspectors verified that the surveillari~e tests were performed in accordance with Technical Specifications,, *

approved procedures, and NRC regulation These inspection activit~es were conducted in accordance with NRC inspection procedure 6172 The following surveillance tests were reviewed, with portions witnessed by the inspector:.,

Unit Procedure N Salem 1 OP-TEMP-9013-1 Salem 1 SP(0)4.7. Salem 1 SP(0)4.4. Salem 1 lIC-18.1.010 Salem 2 OP-ST.SJ-0013(Q)

Hope Creek OT-ST.AC-002 Hope Creek lC-TR.SM-010 Hope Creek OP-ST.AC-002 Hope Creek RE-ST.BF-001 Hope Creek OP-ST.BJ-001

,.. *Test Main Steam Isolation Valves - Fast Closure Test Main Steam Isolation

-- - -*.

Valves - Emergency Close Time Response Emergency Core Cooling System Subsystems Solid State Protection System Train A, Reactor Trip Breaker Undervoltage Coil and Automatic Shunt Trip Test Emergency Core Cooling System Flow Verification Test Turbine Control Valve and Stop Valve Test Time response testing of reactor vessel level Rosemount transmitter Turbine stop valve testing Control rod drive scram time determination Monthly high pressure coo 1 ant injection

. _,-

--;system flowpath verificat.ion-'

__ ;....

23 *.*

With the exception stated *below (4.3;J.A), the surveillance testing activities inspected were effective with respect to meeting-the-~

safety objectives of the.,surveillance.,tes:ting progra.3 Inspection Findings 4. Salem Surveillance Test Performed on Wrong Train On March 31, 1990, an Instrument and Control (I&C) technician performing a Unit 1 operability surv.eil 1 ance test of the 11A

reactor trip breaker (RTB) undervoltage coil and automatic shunt trips inadvertently performed a portion of cthe test on the 118

RTB Solid State Protection System (SSPS) cabine The unit was in Mode 3 (Hot Standby) at the tim In accordance with the surveillance procedure, No. lIC-18.1.010*,

11SSPS Train A, RTB Undervoltage Coil and Automatic Shunt Trip Test 11, the technician entered the Train 11 B 11 cabinet to place the multiplexer test switch to normal, however, when proceeding with the test, the technician reentered Train 11 B" instead of Train 11A 11 *

Operators subsequently closed the 11A 11 RTB per the procedure instruction Several steps later, when the technician released the Block Shunt Trip pushbutton (associated with the 11 B 11 RTB), the operators and technician noted that the 11A 11 RTB did not ope It was then identified that the I&C technician had performed that portion of the test in the opposite trai Both Trains 11A

and 11 B 11 were immediately returned to normal and the test was terminated. *Emergency safeguards were not disabled at any time during the tes The inspector reviewed this event and identified several concern The surveillance procedure is provided with a specific precaution, which states that it is extremely important that the doors to only one train at a time are open, to preclude the introduction of test conditions into both train That precaution was not followed by the technician, thereby contributing to him entering the wrong trai Also, this test requires that Quality Control (QC) shall be present for the performance of the procedur The licensee used an experienced Hope Creek inspector (due to increased resource demand for the Unit 2 refueling outage), however,.he had spent very little time at Salem and was unfamiliar with the syst~m and the procedur Although there was not a sp~cific hold point for that evolution, the QC inspector failed to identjfy,.that. the.wrong train.was entered or that both doors had remained ope The third concern identified by the inspector was that the recently instituted peer review was not implemented by *the 1 i cense In September -

• ,* >*"

1989, the licensee instttuted.,the peer review of critical steps.*

for testing of sensitive safety related systems in an effort.to

• 4. preclude reactor trips and ESF actuation The licensee informed the inspector that they had intended the peer review*to occur primarily at power operations, but.... commated to further review their practice for non-power operation Following the event, I&C perso~ne1 implemented a procedure change which provided a caution -statement *to ensure that the steps are performed on the appropriate trai Similar procedures were also reviewed and changed*as necessar The inspector concluded that this event occurred as a result of technician and QC inspector inattention to detai Increased supervisory oversight may be appropriate in -preventing recurring event *

Hope Creek No noteworthy findings were identified.. EMERGENCY PREPAREDNESS (71707, 93702).2 Inspection Activity The inspector reviewed PSE&G's conformance with 10CFR50.47 regarding implementation of the emergency plan and procedure In addition, licensee event notifications and reporting requirements per 10CFR50.72 and 73 were reviewe Inspection Findings Unusu~l Events Unusual Events declared at Hope Creek and Salem due to a marsh fire and Hope Creek scram were consistent with emergency plan requirements (see section 2.2.2.A). Hope Creek Emergency Preparedness Drill On April 26, 1990, the PSE&G Emergency Preparedness Department conducted a training drill for the Hope Creek statio The drill was not pre-staged and included activation of the Emergency Operations Facility and Emergency News Center offsite and the Technical Support Center (TSC), Operations Support Center (DSC)

and Control Point (CP) onsit The Hope Creek simulator in the offsite training center was utilized in place of the Hope Creek control roo The inspector observed portions of the drill conducted at the onsite locations and, at the conclusion of the drill, discussed the -results with both drill participants and referee The insp~ttor*.noted that the drill was well controlled and that good communicati_oos exist~d. between the various drill

..

~

'

~--.<*

-*~~.

location Based on the responses received from the parttctpant and referee interviews, the inspector concluded that the'drirl*

had been a worthwhile training exercise, for those involve..,._, * SECURITY (71707, 92709).1 Inspection Activity PSE&G 1 s conformance with the security program was verified on a periodic basis, including the adequacy of-staffing, entry control, alarm stations, and physical boundarie These inspection activities were conducted in accordance with NRC.inspection procedures 71707, 9270.2 Inspection Findings Strike Contingency Plans The licensee was notified of pos~ible labor actions by site contractor The licensee initiated st~ike contingency plan including the activation of an alternate security gat Other plans were adopted by the licensee and reviewed by the inspecto Both Salem and Hope Creek operations planned for control room coverage if the labor actions were to occu The inspector reviewed the auxiliary security guard house operation during the period and no unacceptable conditions were note No strike or work interruption actually occurre ENGINEERING/TECHNICAL SUPPORT (37700)

7.1 Salem Safety Injection Pump Flow In Excess of Design Requirements On April 11, 1990, with Unit 1 in Mode 3 (Hot Standby) and Unit 2 in Mode 5 (Cold Shutdown), the licensee identified that the calculated flow rate for one of the two intermediate head safety injection pumps for each unit was greater than the 650 gpm maximum specified in Technical Specification A shutdown was initiated on Unit 1 and Mode 5 was reached on April 13, 199 Both units plan* to correct the condition by performing a full flow discharge and flow balance test while in Mode 5 prior to startu This event is discussed in detail in NRC Special Inspection Report No. 50-272/90-12 and 50-311/90-1 Main Steam Isolation Valve (MSIV) Closure Times

-..

Technical Specifications (TSs) require ~hat the MSIVs be

-~-

demonstrated operab 1 e by verifying full closure within five seconds every 92 days unless the unit is on line. If an MSIV-is

• ---~.. ~-'*.

• • .

slow it r~mains closed until the caus~ has been correcte The licensee routinely tests the MSIVs during plant startup. :***on October 14, 1989, during a Unitc2 controlled shutdown~ the licensee elected to perform the test to~preclude subsequent startup delays. However, 3 of 4 MSIVs failed to meet the 5-second stroke closure time'~fiteria. *(See NRC Inspecti~n 50-311/89-19).

Salem Units 1 and 2 utilize main steam isolation valves (MSIVs)

manufactured by Hopkinsons (distributed in the U.S. by Atwood &

Morrill).

The valves are reverse acting double disk gate valves, with two integral operating pistons and cylinder Emergency fast-closure of the valves is accomplished by using the force of steam pressure acting on a 1 ower steam.cylinder, while the upper electrohydraulic cylinder acts as a snubbe During norma 1 operation, the, 1 ower steam cylinder, which is divided into two chambers, has equal steam pressure in each of the two chambers because of an equalizing orifice.. A drain tube is also provided in each dividing plate for drainage of condensation from the upper to lower steam chambe Each MSIV has two air operated dump valves connected to the upper chambe~

of the steam cylinde Dump valve position is controlled by~

solenoid valve, located in the air supply line to each valv The solenoid valves allow air pressure to hold the dump valve in the closed position unless an MSIV emergency fast closure signal is receive Upon receipt of an MSIV fast closure signal, steam evacuates from the upper chamber through the two dump valve Since the high pressure steam cannot make up to the upper chamber through the small equalizing orifice as fast as it is exhau~ted, the resulting differential pressure closes the MSI The licensee believes the slow closure problem to be attributable to condensation buildup in the upper steam chamber thereby creating a hydraulic lock when the condensation is discharged through the dump valve Only one other U.S. plant (D.C. Cook)

and several French plants use the same type valve and they have experienced similar problem It was suspected that the conden-sation was allowed to buildup and not drain from the upper

.'**

• . ~ '".....,* *.*.

chamber during sustained operating period When the surveillance test is normally run during unit startup, closure times are typically less than 5 second The licensee believes this is primarily due to draining of the condensation while in outage condition An NRC Region I Waiver of Compliance was granted for Unit 2 on

. _

March 30, 1990, to allow one additional MSIV closure for each valve if the first closure time was between 5 to 8 seconds for the testing scheduled for March 31, 1990.

~

.,_

.~*-*..

---

'27 On March 31, 1990, during a Unit ~2-'shutdown for it-s fifth refueling outage,' a similar fast-tlosure_test was performed to obtain data relat.ive to the slow,_,,,closure.times.. One MSIV:-.was**:.-.-

initially closed -in less than 5 seconds, while the remaining three were between*<5 to 8 sec't>>nd Subsequent fast closure times for these.3~MSIVs were l~ss than 5 second Unit 1 was shutdown to Mode 5 due to equipment problems on March 27, 1990 (see section 2.2.1.B).

An NRC Headquarters Temporar.v.:.,

Waiver of Compliance and Emergency Technical Specif-ication Amendment request was submitted by the licensee to.allow similar Unit 1 MSIV testing.... The licensee,. expected to per_form the initial startup surveillance test, then al lo~ a 14 hour1.62037e-4 days <br />0.00389 hours <br />2.314815e-5 weeks <br />5.327e-6 months <br /> soak time while ~n Mode 2 and perform testing to confirm whether the slow closure phenomenon recurred.' However, on April 5,_1990, during the first Unit 1 MSIV test, the~closure time~is-l.32 second The test was immediately terminate The licensee:-*

remained in Mode 2 pending resolution of the related technical concerns and issuance of a TS Temporary Waiver of Compliance by the NR The waiver was issued on April 15, 1990 allowing MSIV closures time up to 8 seconds for the current operating cycl Subsequent testing was performed, and each Unit 1 MSIV closure time was greater than 5 seconds but less than 8 second On retesting, each MSIV closed in les~ than 5 second The resolu-tion of this slow MSIV closure issue, including root cause, remains open pending NRC licensing action Service Water System Motor-Operated Valve (MOV) Shaft-To-Pinion Key Failure While investigating a failure of a Limitorque operator at Unit 1 and 2, the licensee determined that shaft keys between the motors and the pinion gears in six service water (SW) MOVs were damaged (one) or had sheared (five).

These SW valves (Nos. 11, 12, 21, 22SW20; and 1, 2SW26) were unique in that they are quick shutting MOVs with high torque in conjunction with a high gear ratio operator in order to close within 10 second PSE&G found that vendor-supplied Woodruff ASTM-1018 keys installed in the six SW MOVs (three per unit) were sheared and wedged between the motor shafts and the pinion gears at the key-slot The keys apparently failed due to impact loading and possibly too soft a key materia The wedging allows valve operation under limited conditions while-~isguising possible unreliable operation under high-torque condition The valve units involved use hi~b~speed (3600 rpm) SMB-0 Limitorque operators with 25~foot~pounds of torque operatin~ 30-inch *

Jamesbury butterf1y Nalves.through ~ converter hea The licensee had replacd the original GO-second shutting valves

.

-~**

'

!-.

' *.. ~ _;

• .:*,*

with the subject 10-second ~huttin~ ~~lves during a recent upgrade progra This type of valve* is only used for higti_-speed isolation of non-safety SW cool.ing,,c,following an accident. -

A four hour ENS call was made on Apfil 19, 199 A lDCFR part 21 notification was made on April 27, 199 Licensee corrective actions include an intention to replace the key material, cycle the MDV and check for wear, coordinate efforts with the vendor and periodically inspect each key durj*ng refueling outage period The inspector participated in discussions with the licensee and vendors and notified NRC headquarters of the_

potential generic implicatio Charging Pump Casing Cracking During rotor replacement for Unit 1 No. 12 CCP, the licensee identified cracks/indications in the pump casin The CCPs at Salem are Pacific pumps supplied by Dresser, model 2-1/2 11 RL Type IJ eleven stage centrifuga Since 1980, the licensee has been performing inspections of the pump casing stainless steel clad material based on vendor and Westinghouse informatio Recent inspections on No. 12 CCP have found numerous cracks and indications, some of which are through the clad material and into the carbon steel casin The licensee believes the cause of cracking to be as follows:

corrosion caused by boric acid attack on the carbon steel casing after the stainless steel cladding was cracked due to fatigue caused by differential expansion of the dissimilar metal The licensee attempted repairing the cracks in the No. 12 CC However, non-destructive examiriation determined the cracks to be less than minimum wall thickness and current plans are to changeout the pump casin The remaining Unit 1 and Unit 2 pumps are being reviewed by the license Final disposition will be reviewed in subsequent inspection.2 Hope Creek Evaluation of the Electrical Transient The Hope Creek unit scrammed on March 19, 1990 due to an electrical transient offsite. (See section 2.2.2.A) The following discusses an evaluation of that transient and the on-site effect Three independent off site power.SOJ,Jrces supply the Hope.Creek uni One source is the Salem-Hope Creek 500kV tie lin The other two sources are.. from.. the Kenney Switching Station and the New Freedom Switching Station. The 500kV system supplies the preferred power for the _.plant.via the J3.8kV switchyard,ring... bu.s_**.;.,

... 29 This supplies. both**class IE and non-class IE loads during p 1 ant startup~ normaJ operations, shutdown and post-shutdown..

Even though Sa 1 em.and Hope Creek have separate* SOOkV swrtch yards, they are electrically. interconnected through the SOOkV gri Any heavy fault in the SOOkV system would be felt in the nearby di stri'bution syste' The SOOkV.grid stability and analysis of critical faults are discussed in the Hope Creek Final Safety Analysts.. Repor.,

The SOOkV 11 Deans 11 * ~ine leaving the Salem Generating Station*:

experienced a phase B to phase C short which was seen and wa~

properly cleared in approximately 4 cycles by the carrier

~.

protection relayin The magnitude of the voltage transient at the SOOkV level was recorded in-the plant oscillograph, and verified by the;hand calculation to be 0;512 per unit, on the affected phases to neutr~l. This tran~ient ~as felt-throughout the medium and low voltage distribution system at both Salem units and the Hope Creek uni The 4.16kV and 7.2kV buses did not see the entire transient since large induction motor loads were contributing voltages to the syste The lack of undervoltage targets and alarms, and the oscillograph readings on these buses indicated that the voltage was approximately -80%

of the normal voltag The voltage did not dip below the under voltage trip set point of 70% of bus nominal voltag However, the 120V AC systems experienced the full voltage drop caused by this transien This fact was verified by the undervoltage relay alarms at both station During this event, none of the electrical buses lost its powe However, several non-safety loads were tripped as a result of the transien The only safety related load affected by this transient was the control room area chille These loads are discussed in the following paragraph The licensee's engineering staff reviewed this event and concluded that the root cause of the plant trip was due to the loss of the condensate/feedwater system resulting from the undervoltage condition in the 120V AC interruptible power system caused by the SOOkV transien *

The inspector reviewed the licensee's engineering evaluation of the event, Significant Event Response leam's (SERT) evaluations, the chronology of the event, various elec~rical loop drawings and logic diagrams, oscillograph readings, hand calculations and computer chrofiofog showing plant parameters and sequence of event The.purpose of the evaluation was to assure that the electrical eq.ujpment performed as designed and to assure that the design.was :ade.quate. *Based on the review, the inspector noted that licensee's analyses of *the 480 volt and high voltage*

motor trips were r.easonable.. and they.were substantiated through

30'

bench test The tripping of the 480 volt motors was due to,~the seal-in dropouts and contactor coil ~ropouts_

Cutler-Hammer (CH) 42 Contactor Coi.ls are.provided by the motor control'manufacturer within each br~~ker tubicl These loads are manually operated and are maintatned in the required condition by means of a circuit sealing-arrangement, i.e., relay contact in parallel with momentary initiating -contac Although the relay which provides the sea 1-i n feature -is not an under-vo l tage relay, it is affected by momentary undervoltages and deenergizes under degraded conditions. *The voltage level at which the relay deenergizes is unpredi~table~ It is-controlled by the manufacturer 1 s toleranc A bench testing of these relays showed that the relay dropout time and dropout voltage were consistently within the range specified by the manufacturer: A full dropout occurred between 50 and 60 VAC in 30 milliseconds.-~

Cutler-Hammer type D26M pilot relays are also usd in the starting

~~

~ircuit of some of the loads as interposing relays, to assist in picking up the CH 42 contactor coil Bench testing of this relay showed a full dropout between 30 and 45 VAC in 30 milli-second Therefore, the voltage reduction in the 120 VAC control circuit caused the seal-in circuit to open and tripped various motor load The undervoltage was also experienced in the Salem units, but none of the motors were tripped due to the absence of seal-in (maintained contact) circuit desig The failure of the 4.16kV and 7.2kV primary condensate pumps (A/B/CP102) and secondary condensate pumps (A/B/CP137) was due to the temporary degradati_on in 120 VAC interruptible power to the discharge valve limit switches and low lube oil pressure trip signals in the motor trip logic cabinet These circuits are normally energized and the loss or interruption of the interrogation power to this signal immediately causes the logic to change state resulting in a pump tri The cause of the reactor feed pump trips was due to the loss of the condensate pump The only unexplainable event that was not fully analyzed by the licensee was the failure of the condensate pump

indication to flash in the control room showing a non-commanded trip conditio The licensee 1s preliminary study indicated that a possible ground loop existed during the transien Further analysis is planned for the next refueling outag The inspector determined that the lack of flashing of the indication lights is not a concern since the trip condition of the pumps was indicated as a solid ligh The root cause of the pump trips was thoroughly analyzed by the licensee and conclud.ed.,... to be the 120 VAC interruptible power voltage dro This was further clarified by the field*tests.

ll

• The inspector noted_:that the feedwater/condensate system worked as designe However, the inspector,~mphas~zed the need"fo~a complete review.of the feedwater/condensate control system to

avoid any further pump trips due to a similar-electrical*

transien *

The control room area chiller (1AK400) was the only safety related load that was affected by this transien The normally energized seal-in control circuit was dropped out in approximately 2 cycles during* this voltage di This load is supposed to be shed du~ing a loss ~f offsite power scenario, and then sequence back onto the vital. bus fed from~the diese Therefore, the interruption of this load is not a safety concer Several other loads were tripped as a result of the normal plant response to the process signals and in some cases, due to deenergization of the process control relay These events are normal and no abnormal conditions were note However, unrelated malfunctions of some instruments were note They were. analyzed and corrected by the license The licensee's Tower group inspected the affected transmission lines and confirmed that no damages resulted from the faul During the review, the inspector observed that the voltage information from the oscillograph was very hard to interpret accuratel The licensee stated that they are planning to install a computerized analysis system to study the electrical transient The inspector had no further questions at this tim In conclusion, the licensee responded promptly and effectively in dealing with the transien The root cause of the event was properly identifie Analyses were descriptive and thoroug The undervoltage condition existing during the transient was not outside the design basi The loss of feedwater and loss of offsite power scenarios were analyzed in the accident analysi The control room operators responded appropriately and the plant engineering staff and the SERT team responded effectively to identify and to assure that no adverse conditions existed before returning the plant to operatio Motor Control Center Breaker Failure (Allegation RI-A-90-0026)

The NRC received an allegation concerning.aw electrical circuit breaker failure at Ho.pe Creek during construction on January 12, 198 The alleger stated that the breaker failed during testing due to an apparent defect resulting in* injury to a worker and damage to the breaker.<.and. ass.ociated.motor control center (MCC)..

..

The breaker (10-8-232-023) was for main steam stop va1ve 1AB-HV-3631 The alleger also stated concerns regarding breaker coordination issues and inadequacies associated with,the *

licensee 1s corrective action The inspector confirmed that such an. event did occu~.

The~~~

inspector reviewed the event by discussions with the licensee --

and by reviewing the following documents:

electrical schemattcs~

piping drawing, component data forms, ~on-conformance significant-deficiency report (SOR) No. AB-0024, equipment troubleshooting form number GWP-MO.ZZ-001, and work order number 85-1-14-11. -

The licensee concluded that the cause of the breaker failure was a ground fault on the line side of one phase (A).

In addition, the licensee identified that undersized breaker thermal overloads were used in the breake The most probable scen~rio was that the breaker overloads overheated and exploded, resulting in grounding the A phase line side to the breaker bucke This action caused the upstream feeder breaker from the load center to trip on overcurren The licensee also concluded that this breaker. action was consistent with desig Licensee corrective actions included:

initiating the SOR documenting the event and failure of the breaker, troubleshooting the failure using approved construction procedures, ieplacing the breaker with a new one with correctly sized overload devices, retesting the breaker satisfactoril The inspector concluded that licensee actions taken during startup and construction activities in 1985 were consistent with procedure The inspector had no further questions at this tim The inspector concluded that the allegation addressed an event that did occu However, licensee corrective actions appear to have been appropriat This allegation is considered close.

SAFETY ASSESSMENT/QUALITY VERIFICATION. (40500, 71707)

- * 8.1 Sal em Two Year Procedure R~views The inspector evaluated {unresoJved item 90-80-001) the status of two year Salem procedure reviews and determined that

...

approximately 27% of the station* s 2922 affecte-d.procedures were currently overdu This includes overdue procetlures *from the -- *

following departments:

~ *

Operations Instrumentation & Controls Maintenance 401 out of 953

'380 out of 1718 30 out of 251

• .1;.

The inspector determined that the overdue procedure backlo~-was~

approximately 10% in August 1989, when management reallocated

-~

procedure reviewers to support the.procedure upgrade program (PUP).

The PUP was implemented in.June 1989, to improve the

~:*

quality of Salem's procedure The. PUP has prioritized its upgrade ~ffort based on the relative strength of procedures as determined by the user departmen Although this ensures the weakest procedures receive the most immediate review, it con-tributes to the ovetdue backlog as weak procedures within their two year review are worked before overdue procedure In order to expedite the PUP progress, resources were transferred from the two year review process to support PUP activitie The large backlog of overdue procedures indicates that the management decision made in August 1989, regarding resource allocation did not adequately assess the negative consequences of the real-locatio This corrective action taken in response to poor procedure quality exacerbated the backlog of biennial procedure reviews as required by Technical Specification 6.8 and Administ-rative Procedure No. 32 Step 5. Although the licensee identified this violation, it is being cited because of the ineff~ctive corrective action that was implemente Also, the licensee made no attempt for formal notification to or relief from the NRC (50-272/90-11-01).

Due to the -large backlog of overdue procedure reviews, Salem has recently implemented the following interim compensatory measures:*

Complete a full two-year review (per procedure TI-10) and all additional research required by the PUP project for all procedures not reviewed within five years by April 20, 199 For all procedures exercised via a work order (principally maintenance and I&C):

• Perform a review of all current advance change notices (ACNs) and revision requests outstanding by April 20, 199 Any that are judged to be technically significant or that could significantly impact the proper use of the procedure will be identift~d~and the procedure change A four week look-ahead report will be generated each wee For all procedures overdue for the two year review, a complete two year review will be done prior to the procedure being use *

For all procedures exercised without a work order (principally operations procedures):

Perform a review of all current ACN 1 s and rev151on requests outstanding by April 20, 199 Any that are judged to be technically significant or that could significantly impact the proper use of the procedure will be identified and the procedure change A complete two year review will be completed by July 31, 199 By the end of the inspection period the number of overdue procedures had been reduced to 246 for Operations, 288 for I&C, and 2 Maintenance procedure Hope Creek Biennial Procedure Review Backlog Followup (Closed) 50-354/89-80-08; NRC Inspection Report No. 50-354/89-80 issued a violation for having a significant backlog of procedures that were overdue for their 24 month revie The review is required by procedure SA-AP.ZZ-032(Q),

11 Review and Approval of Station Procedures and Procedure Revisions 11 *

The NRC special maintenance team inspection report at Hope Creek dated February.7, 1990, stated that approximately 50% of mechanical maintenance procedures and approximately 40% of instrumentation and control (I&C) and electrical maintenance procedures were overdue for their biennial review~

By a letter dated March 9, 1990, PSE&G responded to the Notice of Violation and committed to adding two additional procedure writers to the permanent staff along with six consultants, in order to eliminate the procedure review backlog by June 199 During the inspection period, the inspector met with the Hope Creek Technical Manager and the technical engineer managing the *

procedure review program to ascertain the progress that had been*

made in reducing the identified backlo The inspector was informed that the technical revtew;staf~.had been supplemented and upon re viewing the weekly _progress.. report issued by* the Technical Department, determined that both the mechani~al procedure and the I&C and el~ctrical pro~edure backlogs-had been reduced to approximately 5% eac The Technical Manager informed the inspector that the present program wi 11 be kept.: in place until the backlog is eliminated and, in fact, until the procedure review program is a month or two ahead of schedul The purpose of working ahead is to account for events, such as refueling outages, that traditionally delay the procedure review proces The inspector found the licensee 1 s program effective, and no inadequacies were identifie Based on the licensee 1 s response to the violation, corrective actions taken, the significant reduction in the backlog, and plans to eliminate the backlog, the violation is considered close Individual Plant Examinations for Severe Accident Vulnerabilities -- NRC Generic Letter 88-20 (Salem)

On November 23, 1988, the NRC issued Generic Letter (GL) 88-20 to request individual Plant Examination (IPE) for severe accident vulnerabilities from all licensee The general purpose of this examination is to (1) develop severe accident behavior, (2)

understand the most likely severe accident sequences that could occur, (3) gain quantitative understanding of the overall probabiliti~s of core damage and fission product releases, and (4)

reduce the core damage and fission product releases by modifying hardware and operating procedures that are intended to prevent or mitigate severe accident The NRC issued NUREG 1335 in August 1989 to provide specific guidance for IP Supplement No. 1 to GL 88-20 was issued on August 29, 1989 to announce the issuance of NUREG 133 GL 88-20 and its supplement requested the licensee to submit their proposed program for completing IPE to the NRC within 60 days of the publication of NUREG 1335 and address the followi~g:

Identify the method and approach selected for performing the IP Describe the method to be used for the examination, and*

Identify the milestones and schedules for performing the IPE and submitting ihe final.results to the NRC.

Additionally, the GL requested th~ licensee.to complete the IPE and submit the fi na 1 report within 3 :years of ethe *issuance of NU REG 1335~-

The 1icensee 1 s proposed program in* response.. to GL 88-20 was.submitted to the NRC on October 31, 198 As stated in this proposed program, the licensee uses NUREG /CR 2300 (PRA Procedures Guide) to develop the Salem IP NUREG /CR 2300 is one of the methods considered adequate in GL 88-20 for performance of IP The final submittal for Salem Units 1 and 2 IPE is presently scheduled for September 1, 199.,.

At the time of the inspection, the licensee 1s staff had developed a-detailed IPE Program Plan for management revie This plan describes a s~ries of actions and a schedule for a comprehensive IPE program at PSE& The general scope of this program is to deveJop a risk model based on current as-built configuration of both Salem Units and to increase the awareness of PRA concepts within station op~ration The p 1 an a 1 so has provision to deve 1 op and document a computer risk,

model for predicting the core damage frequency and the ability of

'~ *-

• ~.,

-~**...... :

containment to mitigate accident sequences at the Salem Units.

• *The

• • • • • ':{

results from the study will be used for future plant betterment activities related to Technical Specification improvement, design*

change package (DCP) review and screening, reliability centered maintenance (RCM) and support for licensing in resolution of technical issue Licensee had completed a preliminary Level 1 PRA in October 198 The results of this study were being refined using plant specific data at the time of this inspectio The licensee 1 s IPE activities are primarily carried out under the cognizance of the risk assessment grou This group has a full time s~pervisor and an authorized staff of six personne The expertise of this group is further augmented by the use of contractor personne At the time of this inspection, the risk assessment group consisted of three licensee employees and two contractor The licensee was actively recruiting to fill the three vacant positions by the end of 199 The licensee also arranged a contractor to conduct a PRA workshop in June 1990 for its employee The contractor personnel, the group supervisor and one employee of the Risk Assessment group were familiar with details of the activities related to GL 88-2 The other two personnel were recently assigned to this group and were,.gatni ng.exper.ti se by performing analyses as directe '"***.. **_;...

..

..

/J

This review was primarily to determine the. status activities in response to GL 88-2 As* such, the conclusions made in. this report are preliminar assessment of the licensee actions.in response to will be made separatel of licensee observations and Formal NRC the above NRC GL LICENSEE EVENT REPORTS (LER), PERIODIC AND SPECIAL REPORTS, AND OPEN ITEM FOLLOWUP (90712, 90713, 92700) LERs and Reports PSE&G submitted the following licensee event reports and, special and periodic reports, which were reviewed for accuracy and the adequacy of the evaluation:

Salem and Hope Creek Monthly Operating Reports for March 199 Hope Creek LER LER 90-03 9.2 Open Items concerns a reactor scram which resulted from transmission line faults caused by an offsite marsh fire on March 19, 199 The event is discussed in paragraph 2.2.2.A of this repor The cause of the event was determined to be ionized air and carbon ash generated by the fire which caused a flashover of two phases of an offsite 500 KV transmission lin The subsequent low voltage transient was propagated through Hope Creek 1s electrical distribu~ion syst~m and, at the 120 VAC level, caused a trip of the condensate pump The loss of the condensate pumps caused the reactor feed pumps to trip, resulting in a low reactor water level and a subsequent reactor scra The licensee 1s corrective actions were reviewed and determined to be appropriate ahd satisfactor This is the first such event at either Hope Creek or Salem, and the licensee intends to try and reduce the potential for marsh fires around the site and determine any improvements which would enhance electrical system reliability during voltage transient The following previous inspection items were followed up during this i.nspection and are tabulated below fo.r. cross reference purpose Hope Creek

• section Status 354/89-80-08

. '"8.2.. A Closed

.-~. -

Salem Section Status 272/90-80-07 1 EXIT INTERVIEW (30703)

8. C'l osed 10.1 Resident 10. 2 The inspectors met with Mr. L. K. Miller and Mr. J. J. Hagan and other PSE&G personnel periodically and at the end of the inspection report period to summarize the scope and findings of their inspection activitie Based on Region I review and discussions with PSE&G, it was determined that this report does not contain information subject to 10 CFR 2 restriction NRC Commissioner Rogers visited Salem and Hope Creek Generating Stations on March 26 and 27, 1990, respectivel The visit included plant tours and interviews with selected station and NRC personnel_

The Commissioner was accompanied by a NRC Region I management

representative and a technical assistan Specialist Inspection Reporting Date(s)

Subject Report N Inspector 3/29-23/90 Requalification 272/90-09; Examinations 311/90-09 Hughes 4/9-27/90 Maintenance Team 272/90-200; 311/90-200 Ball 3/17-23/90 Operator Licensing 354/90-05 Wa 1 ker Examinations 4/16-20/90 Transportation and 354/90-09 Furia Solid Radwaste