ML20198S083

Ack Receipt of Informing NRC of Steps Taken to Correct Violations Noted in Insp Rept 50-219/97-06.Actions Will Be Examined During Future Insp of Licensed Program

ML20198S083
Person / Time
Site:	Oyster Creek
Issue date:	01/12/1998
From:	Eselgroth P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:	Roche M GENERAL PUBLIC UTILITIES CORP.
References
50-219-97-06, 50-219-97-6, NUDOCS 9801260060
Download: ML20198S083 (2)

See also: IR 05000219/1997006

Text

{{#Wiki_filter:.. - 6 January 12, 1998 _ ' Mr. Michael B. Roche Vice President and Director GPU Nuclear Corporation Oyster Creek Nuclear Generating Station P.O.__ Box 388 Forked River, New Jersey 08731 SUBJECT: INSPECTION REPORT NO. 50-219/97-06- REPLY Dear Mr. Roche: This letter refers to your December 17,1997, correspondence, in response to our November 17,1997, letter. Thank you for informing us of the corrective and preventive actions documented in your . letter. These actions will be examined during a future inspection of your licensed program. Your cooperation with us is appreciated. Sincerely, k & Peter W. Eselgroth, Chief Project Branch 7 . Division of Reactor Projects Docket No. 50-219 cc: G, Busch, Manager, Site Licensing, Oyster Creek . M. Laggart, Manager, Corporate Licensing State of New Jersey

< A ' + , k> 126006o ,e0ita ' $. ADOCK:0SOOO319. ffflll Ijl. . . Arg - ~ ~ .

. .. . . - . . . L * :. 'P '* . ! Mr. Michael B. Roche 2 > Qistributioni Region I Docket Room (with concurrences) - Nuclear Safety Information Center (NSIC) . NRC Resident inspector PUBLIC - ' W. Axelson, DRA (irs) P. Eselgroth, DRP D. Haverkamp, DRP N. Perry, DRP -T. Kenny,DRS- A. Linde, DRP W. Dean, OEDO P. Milano, NRR/PD l-3 R. Eaton, NRR/PD 1-3 . R. Correia, NRR , . F. Talbot, NRR -DOCDESK- Inspection Program Branch, NRR (IPAS) 4 . . DOCUMENT NAME: G:\\ BRANCH 7\\0C9706. REP

T3 recolw a copy W this document, Indicate in the box

V = Copy without attachment / err!osure

"E" = Copy with attachment / enclosure . 'N' = No . copy' 0FFICE- RI/DRP- Rl/J)Pf I / l l l- , NAME- JSchoppy / % Ju Pf91H4roth DATE 01/99/98 -/ 0;.// t-/98 - 01/ /98 01/ /98 01/ /98 '"- 0FFICIAL RECORD COPY _ 4 4 -

. . . .... .. - . ., - - .. _. . , = . , = . -, , . ':. GPU Nuclear,Inc. U.s. Route M South : Post Office Box 388 - NUCLEAN ,,,i,4 n,,,, na oe733.o3 e , Tel 609 9714000 December 17, 1997 6730-97-2282 U. S. Nuclear Regulatory Commission - Attn: Document Control Desk , Washington DC 20555 Dear Sir: , Subject: Oyster Creek Nuclear Generating Station i Docket No. 50-219 Inspection Report 50-219/97-06 Reply to Notices of Violation By letter dated November 17,1997, the NRC docketed the results of a Pre-Decisional Enforcement Conference held with GPU Nuclear, Inc. on September 30,1997. That letter contair.ed six notices of violation. The attachment to this letter provides the requisite reply. If any additional information or assistance is required, please contact Mr. John Rogers of my staff at 609.971.4893. Very truly yours, s - - Michael B. Roche Vice President and Director Oyster Creek MBR/JJR -cc: Administrator, Region I: , ~ NRC Project Manager ' Senior Resident Inspector - ! $5 $U - 5.6 ' ) (f . '

_. . . - - _ _ _ . _ _ . _ _ . _ _ . _ _ . _ - - . _ _ _ , , . .c . ] Attachment I z l > Area I; Violations Related to Design Controh k , "The Code of Federal Regulations,10 CFR 50, Appendix B, Criterion 111, requires, in part, , that measures shall be established to assure that applicable regulatory requirements'and the .

- design basis, as defined in i 50.2 and as specified in the license application - for those . ~ - structures, systems, and components to which this appendix applies, are correctly translated . into specifications, drawings, procedures, and instructions. Design control measures shall be ~ ' applied to compatibility of materials." . Notice of Violation 1 A: - 2" Contrary to the above, in October 1993, the vendor specification for the emergency service Lwater (ESW) pump bowl assemblies was changed from cast iron to stainless steel without > adequately applying design control measures for use in considering the compatibility of j materials specified for other ESW pump components. - As a result, in November 1993, the "C" , ' and ,"D" ESW pumps'were installed with cast iron top case flanges and stainless steel bowl assemblies, and the materials in these components were incompatible in that they created the - ' opportunity for accelerated galvanic corrosion. Subsequently, the "C". ESW pump was found to be inoperable on July 31,1997, because of a broken coupling (top case flange) between the pump bowl and the discharge head." , . GPUN Reply to Violation IA: 1 , GPUN acknowledges that a violation occurred and offers the following hdditional information: - . Cast iron was used in the original design of the Emergency Service Water Pumps. In 1985, . the vendor started to recommend material changes to the design of the pump which replaced 3 the cast iron column pipe with stainless steel.- GPUN approved the recommendations. This ^ created an opportunity for accelerated galvanic corrosion. - In 1992, the vendor rebuilt the pump which was eventually installed in the 'C' location to the vendor's approved Bill of Materials which had cast iron top, case and head flanges, in 1993, the vendor rebuilt three pumps under 10 CFR 50 and 10 CFR 21 programs. These . pumps were to be upgraded to'all stainless steel. Due to a request for expediting, the vendor k

supplied one of the three pumps with a cast iron head flange (supplied in error by GPUN) and :

reused the existing top case flange (without advising GPUN),LThis pump was eventually . i ~placed into the 'D' location. - During this time period,? a GPUN Engineering Evaluation was - . performed for the stainless steel bowl assemblies, but did not evaluate the head flange or top Lease assemblies.1 Stainless steel parts were delivered to the warehouse as spares. Existing cast - iron parts.were scheduled to be scrapped as the stainless steel parts arrived ' ' $ , . ~ - . - - - - , ..l , . - . - . . - - - ' 1 '

. - - - , _ - _ . . . _ .. _ _ _ _ . _ . _ . . _ _ _ . _ _ _ s -<..

. ,

.. . , ' - - ~6730-97-2282- Attachment I . Page 2 4 d - Thet "C" and "D". pumps were replaced with the rebuilt stainless steel pumps with cast iron - , ' (flanges, in 1995 GPUN verified that no inappropriate cast iron ESW pump spare parts were in the warehouse. . , On July 31,1997, the "C" ESW pump failed during~'a routine surveillance test. > ' - Root Cause: The immediate cause of the failure was inadequate design and configuration control by both the vendor and GPUN. This allowed a material incompatibility to occur with the resultant , galvanic corrosion accelerated by the galvanic cell. S The root cause of this event was a lack of knowledge and understanding on the part of the

GPUN reviewer on the impact of material compatibility. A contributory cause was the lack of-

, procedural guidance regarding material compatibility. Corrective Actions: Short Term The three remaining pumps were inspected for a common mode problem. The 'D' ESW pump was found to have a cast iron top case flange. Both the 'C' and 'D' pumps were declared . inoperable. They were replaced with pumps containing stainless steel parts. The 'A' and _'B' pumps were verified to cotitain only stainless steel parts. The warehouse was searched and no unacceptable ESW parts were found. Long Term The vendor was informed of the compatibility concern and was requested to expedite his & investigation and corrective action processes. The vemier issued a 10 CFR 21 report. The Oyster Creek modification _and alternative replacement procedures were upgraded to include material compatibility considerations. A new engineering design change evaluation for the f stainless steel ESW pumps was issued. . , , Engineering memorandums which had directed the warehouse to dispose of material since . - 1985 will be identified and reviewed. Action will be taken to ensure proper implementation. ' . 1 ' Training will be provided to appropriate engineering personnel. These actions are presently ~ - projected for completion by the end of December 1997. t - Date When Full CampHance 'was. Achieved: . , Full compliance was achieved in August 1997_when the two ESW pumps were replaced with.- ' L pumpsicontaining compatible parts. , . ' _ T i -- j \\ ' - . , ? ,.

---. ,

, ,. , ....+,..-;.-,, , '

__ _ _ . _ _ _. _ _._ . _ . ._ _ . _ _ _ _ . _ _ . , _ _ __

. :. m- Specifically, in April 1989, the licensce modified the RWCU system procedures such that ! operators periodically opened V-16 2 at full reactor pressure (1020 psig) in order to fill'and

' pressurize the RWCU system. However, the design input of 125 psid differential pressure used in the design basis calculations that determined the capabilities of V-16-2 failed to consider that V-16-2 might be required to close against full reactor pressure. Therefore, there was no assurance that V-16-2 would be able to close to perform its required isolation function . against full reactor pressure." lGPUN Reply to Violation 1B: GPUN acknowledges that a violation occurred and offers the following additional information: Lin 1984, V-16-2 was replaced with a valve designed for 1125 psid. In November 1988, the system was modified to place a one inch line downstream _of V-16-2 (a six inch valve). The operating procedure was then revised to allow for filling and pressurizing at power through V-16-2. Valve thrust requirements at this time were based on Torrey Pines calculations which used the design pressure of the valve as the differential pressure. - In February 1992, the valve differential pressure calculation was incorrectly changed based on the assumption that the valve was not opened at elevated reactor pressures. In March 1997, the design input omission was self identified during expert panel review while establishing relative risk ranking for the interim Periodic Verification Valve Test Program.- Root Cause: The immediate cause of the violation was that an incorrect design input was used in the design - basis calculation and an inadequate design verification was performed. .The root cause of the p occurrence was that there was a lack of thoroughness on the part of the preparers and U reviewers, Additionally, no formal training was given to emphasize the need of the preparer .and design verifier to assure calculation inputs were correct. sCorrective Action: - p immediata' t CAdministrative controls were put in place tb prevent the opening of V-16-2. The operating . , iprocedure for the RWCU system was revised to allow for filling and pressurizing with a > , . iparallel valve (V-16-14). .- . . - '-' . p ._ , . _ . . - l . , , m - l S. . a.: ., l y l^f = , r _ - - - _ -_ _ . - _ _ - _ - .__

, . . .. . ,. -- - . . . . - . - - . ..- , - - . . . - . - . . _ - .- . . - . - - ' ^ - . --, . . , - i, - ' C 6730-97-2282

- Attachment I ~ Page 4 f Shcrt Term

In March 1997, the existing design Ap calculations for GL 89-10 were reviewed for any--.

_! . similar occurrences.- Two drywell spray and two torus cooling valves were identified which - t could have been operated differently than the CL 89-10 assumptions. ' The EOPs did not .; . . preclude the operators from cycling the valves with the Containment Spray pumps running. - It : was subsequently' concluded that the valves were ooerable.- ' ' a Training emphasizing the need for the preparer and the design verifier to assure that

[ calculation inputs are correct was given to the appropriate personnel. Long Term jl A GL 89-10 thrust calculation for V-16-2 was completed to account for a high energy line L > break,; In September 1997, the valve electrical circuitry was modified and procedures revised " _

to allow operation at power. The differential pressure calculation for all GL 89-10 valves' is

t . being revised to enhance our ability to use the performance prediction model. Design inputs , F were subject to multidiscipline review. An update to the GL 89-10 Supplement 3 response , was submitted to the NRC on November 5,1997.

ic - Date When Full Compliance was Achieved: r , Full compliance was achieved March 12,1997 when the administrative controls were put in place to prevent the opening of V-16-2.

r

Notice of Violatlun 1C: " Contrary to the above, in 1994, the licensee failed to take adequate corrective measures to , assure that specifications for the 4160 volt vital bus degraded voltage relay (DV_R) setpoints .were adequate to ensure that the onsite distribution system, was capable of providing L acceptable voltage under worst case station electric load and grid voltages as specified 16 the - ~ Final Safety _ Analysis Report (FSAR), Sections 8.3.1.2.a and 8.2.2.g. Specifically, the , . degraded grid analysis,-' performed in 1994, failed to consider that the startup transformer voltage regulators could lower bus voltage. The analysis did not consider the design of the b voltage regulatorst specifically the regulator response time and setpoints, in selection of the - ' . , - - DVR setpoints.EAs a result, on August 1,1997, when the output of the Oyster Creek unit was 11ost due to a' manual reactor scram, the startup transformer voltage regulators. failed to supply . , Lsufficient voltage to the vital busses to preclude an undervoltage'(UV) condition. Thic UV. 4 - , l condition lasted for greater than'10 seconds which caused the DVRs to trip (and not reset) ,

1. which de-energized the vital buses and started the emergency. diesel generators contrary to the

s . ' , edesign bases," - = a < >,x _- r- -e re r e r e + e-, , .- - +wn, r--,, -- -..n. . , . - n -+- - . . -

. . _ . _ . . . _ ._ _ _ . _ . _ . __ _ . _ _ _._ ___ __ _ _ _ _ . - [[. _

i

6730-97 2282

[ 7 j. - Attachment I l Page 5

d7 GPUN Reply to Violation ICt - '

GPUN ahknowledges that a violation occurred and offers the following additional information.

_

The separation of the 4160v vital buses from the preferred offsite power source occurred due - , to the actuation of the DVR sensors and sustained undervoltage condition for 10 i 1 ' seconds. -

' The relays operated as designed to protect the plant's. electrical distribution system, however, the separation occurred above the lower analyzed voltage limits for the off site power source. -

Root Causes:- l The 1994 degraded grid analysis utilized incorrect design input assumptions regarding the

control and operation of the Induction Regulators. The fact that the Induction Regulators could ! . ilower bus voltage was not considered. The Design Verification did not identify this oversight. Additionally, there was a lack of thoroughness on the part of the preparers and reviewers. ~ No

formal training was given to emphasize the need of the preparer and design verifier to assure calculation inputs were correct. Finally, management expectations were not fully 4 communicated. Cornetive Actions: ' . - Prior to Restart i A voltage band was established to ensure that the existing degraded grid design bases were satisfied, accounting for the Induction Regulators. The switchyard tour frequency was . = increased from weekly to daily. The operator logs were revised to identify the acceptable - _ voltage regulator bands. Future operability determinations for the Startup Transformers will , include the Induction Regulators. The event was reviewed with the licensed operators. , - Onnoine Corrective Actions . The degraded grid analysis for Oyster Creek is being reviewed by an' independent industry - expert to ensure that the methods of the analysis are consistent with the regulations and > industry standards, and that the' analytical assumptions are valid. This review is presently projected to complete by the end of December 1997, i qThe degraded grid analysis for_ Oyster Creek will be updr.ted to include the Induction = Regulatorst any issues identified by the. independent consultant, and further benchmarking a against the actual plant response. The design inputs will be reviewed by a rnultidisciplinary panel.LNew degraded grid voltage sensing relays which are not susceptible to harmonics are ,- _ e being evaluated. . . - ( These actions are presently scheduled for completion prior to restart from the 17R refueling , outage. , 1 + l 4 i .a y N w w

y = v -,- r w , .-oe

_ _ _-

l.

i 6730-97-2282 ~ Atta;hment I Page 6 = In recognition that the root cause of incorrect design input assumptions could be widespread and affect other calculations performed by GPUN, a one time review of calculations on 1 ey parameters performed by GPUN will be conducted. The review project commenced in November 1997 with the distribution of the ident19ed calculations to the responsible i engineering sections. Phase I will complete in June 1998 with a multidiscipline verification panel review of the design input adequacy / thoroughness of these calculations. At that time, a decision will be made concerning which (if any) calculations wiu receive a more in-depth review by verification teams. Finally, training was completed on the calculation process / procedure to ensure the ongoing accuracy of calculation inputs, i i

Date When Full Compliance Was Achieved: i Full compliance was achieved prior to restart when the voltage band for the Induction Regulators was established and the operator logs revised to identify acceptable readings. j Area 2, Violation Related to Inoperable CRD Pumps: Notice of Violation 2 " Oyster Creek Technical Specifications 3.4.D.1, requires that the control rod drive (CRD) hydraulic system shall be operable when the reactor water temperature is above 212 degrees F. " Contrary to the above, for a substantial period of time prior to August 1,1997, with reactor water temperature above 212 degrees F, both CRD pumps were inoperable in that both CRD pumps tripped and could no: be restarted remotely when they attempted to load onto the vital buses. The anti-puraping feature of the CRD pump breakers prevented the breakers from closing after a trip signal with a start signal continuously supplied. The CRD pumps are designed to autorratically load onto the vital busses following a loss of offsite power." GPUN Reply To Violation 2 GPUN acknowledges that a violation occurred and offers the following additional information: During the August 1,1997 event, both CRD pumps failed to load onto the vital buses. Both the CRD pump start and undervoltage trip signals were concurrently present. This caused the breaker to trip and lockout due to the anti-pumping feature in the breaker logic.

. _

< w

. 6730-97-2282 Attachment I Page 7 The last ten years of operating history were reviewed. No similar occurrence when both CRD pumps failed to start was identified. Prior to this event, a vital bus was lost and re-energized by a diesel with the CRD pump receiving an autostart signal a total of five times. In four out of the five cases, the CRD pump started. Root Cause The root cause of this event was a lack of understanding of the original plant design for the CRD breaker logic. This lack of understanding led to an inadequate testing of the breaker and its control logic. Corrective Actions: Short Term Site personnel started a pump locally at the breaker. Prior to Restart The CRD pump breakers and UV devices were successfully tested. The pump start logic was modift:d to require a 60 second delay for all diesel loading sequences. The timer modification was successfully tested. The remaining DG loading circuits were reviewed. No additional automatic starting problems existed. Lone Term The CRD circuitry was evaluated under the methodologies of GL 9o41. The evaluation correctly identified that the immediate start circuit had not been tested. No other problems were identified. The remainder of the GL 96-01 circuit reviews are presently projected for completion by the end of December 1997. Date When Full Compliance Was Achieved: Full compliance was achieved on August 3,1997 when the CRD pump was manually started. Area 3, Violations Related to Inadequate Corrective Actions: "The Code of Federal Regulations,10 CFR 50, Appendix B, Criterion XVI (Corrective Action), requires that measures shall be established to assure that conditions adverse to quality, such as failures, malfunctions, deficiencies, deviations, defective material and equipment, and nonconformances are promptly identified and corrected in the case of significant conditions adverse to quality, the measures shall assure that the cause of the condition is determined and corrective action taken to preclude repetition.

B n- r

6730-97-2282--

' ,* " Attachment I ' Page_8 Notice of Violation 3At " Contrary to the above, between January 22,1997, and July 9,1997, the licensee failed to ' . ' take timely and effective corrective action to preclude repetition of a significant condition - adverse to quality.that resulted in bo'h trains of the Standby Gas Treatment (SBGT) System

being inoperable.=

"Specifically, the licensee failed to ensure that technicians assigned to calibrate the reactor building (RB) ventilation exhaust radiation monitors on July 9,1997, had received remedial logarithmic scale training. Training should have been provided as corrective action'for a - previous event on January 22,1997, which resulted in miscalibration of the RB ventilation exhaust radiation monitors. . However, all of the technicians that calibrated the radiation monitors on July 9,.1997, had not received the training and the technicians' supervisor failed

Lto ensure that the technicians were properly trained. As a result, the RB ventilation exhaust radiation monitor setpoints were set non-conservatively high which would have delayed initiation of the SBOT System." GPUN Reply to Violation 3A: i GPUN acknowledges that a violation occurred and offers the following additional information: ' Following the January 22,1997, occurrence, the corrective actions focused on Human ' ' Performance (the meter was misread), it was determined at that time that administrative " controls or engineerinF changes were not necessary. Following.the July 9,1997, occurrence, the corrective actions focused on administrative _ controls and engineering changes. 'Ihe corrective actions were also expanded to other - instrumentation involving logarithmic scales. . _ Root Cause: . The immediate cause of both occurrences was initially determined to be a lack of both proper self checking and proper peer checking._ lt was subsequently determined that instrument & 4 Control Technicians experienced difficulty reading logarithmic scale instruments as evidenced - -.tiy initial l training sessions, and the instrument set point was not'at a discrete marking on meter = face tesulting in a human factors issue, fAlso, the technicians were not trained on the'new . surveillance procedure prior to its implementation. - Additionally, the technicians assigned to

the task'had not'all received the logarithmic scale training. Finally, management failed to

implement corrective actions in a timely fashion. l N % .C. , - - -

, s t -

. .. . y,.* , - 6730-97-2282 Attachment I Page 9 Corrective Actions: Refresher training was provided to all Instrument and Control personnel on reading' logarithmic scales, self-checking and peer checking practices. The setpoints were changed to a discrete marking on the meter face. The setpoints on all logarithmic scale instruments were reviewed for similar conditions. - Surveillance procedures were revised to address human factor concerns as appropriate. Instrument & Control technicians and supervisors were given additional - training on conducting pre-job briefings. Finally, the monitoring of corrective action tracking has been enhanced. Date When Full Compliance Was Achieved: Full compliance was achieved on September 23,1997, when the refresher training for the Instrument and Control Technicians was completed. Notice of Violation 3B: " Contrary to the above, between September 13,1996, and August 3,1997, the licensee failed to take effective corrective action to preclude repetitio.n of a significant condition adverse to quality that removed the low suction pressure protection feature from the Shutdown Cooling (SDC) pumps, and resulted in inadvertent SDC pump trips. The SDC pump trips could have resulted in a loss of SDC. Specifically, the licensee failed to assure that corrective actions to revise procedures and provide training were adequate to ensure that the suction pressure switches were not isolated during SDC pump operation following similar events on ~ September 13,1996, and April 24,1997. On April 24,1997, and August 3,1997, a shutdown cooling pump inadvertently tripped because the suction pressure switches were isolated." GPUN Reply to Violation 3B: GPUN acknowledges that a violation occurred, and offers the following additional information: For the September 13,1996, discovery (no pump trip involved), all of the instrument root isolation valves were found closed while lining up system in preparation for startup. Several additional inconsistencies were identified with the System Operating Procedure. A revision to .the System Operating Procedure was initiated. For the April 24,1997, occurrence (pump trip involved), the operators intentionally left all the instrument root isolation valves closed waiting for system flow to stabilize. Additionally, the operators failed to communicate this decision (valve position) to the oncoming shift. The. word - selection in this revision led to procedural misinterpretation. A second revision to the System '

Operating Procedure was initiated.

. E.* ' .

l p' 1

6730-97-2282 L ' Attachment I- ' , -. , Page 10

For the August 3,1997, occurrence (pump trip involved), all of the instrument root isolation '

] valves were found closed. It was identified that the valves were left closed following shutdown ! _ - of the system using the revision of the System Operating Procedure that had been identified as - deficient following the Ssptember 1996 discovery. Upon discovering these valves closed, the inconsistency between the procedure revisions was identified and the valves were' appropriately

- positioned (opened) in accordance with the current revision (initiated after the September 1996 _ discovery)'of the System Operating Procedure. -l - Root Causet ! .- .- . . . -

The immediate cause of the pump trips was that the actual valve lineup was consistent with that

'

required by the old procedure revision, but it was different from that required in the new - ! l procedure revision. It was further determined that the procedural revisions were examples of Operator Work - Arounds. Reliance on operators to properly manipulate valves was inappropriate to - compensate for system design deficiencies. Additionally, the change management process was found to be weak. The new procedure revision did not cause the valves to be repositioned to

reflect the new requirements.

Corrective Actions: Interim Corrective Actions The System Operating Procedure was revised to include specific valve lineups when starting up ! or securing the SDC System. This event was reviewed with all Operations Department - personnel. - Existing Operator Work Arounds were evaluated for similar situations and engineering solutions are being pursued, as appropriate. The monitoung of corrective action - tracking has been enhanced. ! - Lona Term Actions ! L Appropriate personnel are evaluating potential engineering alternates to eliminate the necessity- l for valve manipulations. The Procedure Control Process is be_ing revised to include a " flag" 4 for identifying any procedure changes that require valve lineups. g

Date When Full Compliance Was Achieved: ' Full compliance was~ achieved 'on May 25,1997, when the operating procedure was Lappropriately revised.' ' _ 1 ~ , A, _ - , _ t i s 4 ' , i ' _1 "' - - }}