05000311/LER-1993-008, :on 930527,control Rod 1SA3 Withdrew to Approx 15 Steps from Fully Inserted in Response to Manual Insertion Command Due to Failure of Integrated Circuit Chips.Emergency License Amend Requested on 930617

:on 930527,control Rod 1SA3 Withdrew to Approx 15 Steps from Fully Inserted in Response to Manual Insertion Command Due to Failure of Integrated Circuit Chips.Emergency License Amend Requested on 930617

ML18100A581
Person / Time
Site:	Salem
Issue date:	08/26/1993
From:	Pastva M Public Service Enterprise Group
To:
Shared Package
ML18100A580	List: 05000311/LER-1993-008, :on 930527,control Rod 1SA3 Withdrew to Approx 15 Steps from Fully Inserted in Response to Manual Insertion Command Due to Failure of Integrated Circuit Chips.Emergency License Amend Requested on 930617
References
LER-93-008, LER-93-8, NUDOCS 9309030268
Download: ML18100A581 (6)

LER-1993-008, on 930527,control Rod 1SA3 Withdrew to Approx 15 Steps from Fully Inserted in Response to Manual Insertion Command Due to Failure of Integrated Circuit Chips.Emergency License Amend Requested on 930617

Event date:
Report date:
Reporting criterion:	10 CFR 50.73(a)(2)(v)(A), Loss of Safety Function - Shutdown the Reactor
3111993008R00 - NRC Website

text

{{#Wiki_filter:NRC FORM366 16-891 U.S. NUCLEAR REGULATORY COMMISSION APPROVED OMB NO. 3150-0104 LICENSEE EVENT REPORT (LERI EXPIRES: 4/30/92 ESTIMATED BURDEN PER RESPONSE TO COMPLY WTH THIS

• INFORMATION COLLECTION REQUEST: 50.0 HRS. FORWARD COMMENTS REGARDING BURDEN ESTIMATE TO THE RECORDS AND REPORTS MANAGEMENT BRANCH IP-530). U.S. NUCLEAR REGULATORY COMMISSION, WASHINGTON, DC 20555, AND TO THE PAPERWORK REDUCTION PROJECT (3150-0104). OFFICE OF MANAGEMENT AND BUDGET, WASHINGTON, DC 20503.

FACILITY NAME (1) I DOCKET NUMBER (21 I PAGE 131 Salem Generating Station - Unit 2 o 15 I o Io I o 13 I 11 l 1 !oF. 01 6 TITLE 141 Rod Control System Design Basis Concern (both Salem Units). EVENT DATE (5) LEA NUMBER (61 REPORT DATE (71 OTHER FACILITIES INVOLVED IBI

• MONTH DAY YEAR YEAR J:t SE~~~~~~AL t? ~~~~~~ MONTH DAY YEAR FACILITY NAMES DOCKET NUMBERISI Salem Unit 1 o 1 5 I o I o I o I 2 I 71 2 o I 5 21'1 9 3 9 I 3 -

o I o I s - q 1 o 1 s 2 I 6 91 3 OPERATING MODE IBI 3 THIS REPORT IS SUBMITTED PURSUANT TO THE R!,.QUIREMENTS OF 10 CFR §: (Ch*ck on* or more of the following) (111 '20.402lbl 20.405lcl 50.73(*112llivl 73.71lbl POWER

• I LEVEL 1101 0 I 010 20.405(*111 llil 20...05{*111 lllil 50.3Slcll11 50.38lcll21
• x - 50.73l*ll2llvl 73.i11cl 50.731*11211viil OTHER IS1H1cify in Abstr*ct btJ!ow and in Text, NRC F.orm 20.4051*111 llilil 50,731*112llil 50,73l*ll21lviiil IAI 366AI 20 *..051*111 llivl 50.73(*112lliil 50.731*112llviiillBI 20.4051*111 IM 50.731*112lliiil 50.731*112llxl LICENSEE CONTACT FOR THIS LEA 1121 NAME TELEPHONE NUMBER AREA CODE M. J. Pastva, Jr. - LER Coordinator*

6 10 I 9 3 13 I 9 1 - 15 I 1 16 15 COMPLETE ONE LINE FOR EACH COMPONENT FAILURE DESCRIBED IN THIS REPORT 1131

CAUSE

SYSTEM COMPONENT MANUFAC* TUR ER MANUFAC* TUR ER x A1A E1C I B1D w I 112 I 0 y

=:.. ::::,:::::::::::::.. ::~:::::.. :::::.

I I I I I. I I x A IA E IC I Bl D W 1* 112 I 0 y I I I I I I I SUPPLEMENTAL REPORT EXPECTED 1141 MONTH DAY YEAR ---, YES (If yes. complot* EXPECTED SUBMISSION DATE) ~NO EXPECTED SUBMISSION DATE 1151 I I I ABSTRACT (Limit to 1400 spaces, i.tJ., approximtJttJ/y fiftetm single-spac~ typewritten lines) 116) During Unit 2 Reactor startup activities, following the unit's 7th refuel outage, it was determined on 6/4/93 that a postulated single failure concern existed where failure of one Rod Control System slave cycler decoder card, in conjunction with a rod motion command signal, may cause an unplanned Rod Control Cluster Assembly (RCCA) withdrawal.. At 1734 hours, all control rods were inserted, the Reactor trip breakers were opened, and the Unit *was stabilized in MODE 3. On 5/27/93, at 1844 hours, rod 1SA3 had withdrawn approximately 15 step~ from fully inserted following a manual insertion command. Rod control power was then deenergized to fully insert the rod. The RCS single failure concern is attributed to RCS design. 1SA3 withdrew as the result of inappropriate current orders to the RCCA. Integrated circuit chips on two slave cycler decoder cards had failed due to the relay* driver circuit card connector Pin No. 4 not making electrical contact with the surge suppression diode. Pin No. 4 was repaired.and the slave cycler cards were replaced to restore operability of rod 1SA3. An additional corrective action was installation of suppression diodes on the rod step counters of the RCS circuitry, of each unit, to mitigate consequences of an open or bad connection on the relay driver circuit card connector pin No. 4. All Unit 2 RCS logic cards.were replaced and satisfactorily tested and all RCS Power Cabinet cards were pulled, visually inspected, and retested satisfactorily. On 6/29/93 Unit 2 was taken critical. 9309030268 -930.826 1*.i PDR ADDCK

• 05000311 NRC Form 366 16-891 S

PDR ~ I

• I SUPPLEMENTAL LICENSEE EVENT REPORT (LER) TEXT CONTINUATION Salem Generating station Unit 2 DOCKET NUMBER 50003*11

PLANT AND SYSTEM IDENTIFICATION

Westinghouse

- Pressurized Water Reactor.

LER NUMBER 93-008-01 PAGE 2 of 6 Energy Industry Identification System (EIIS) codes and component function identifiers are identified in the text as-{xx/xx} IDENTIFICATION OF OCCURRENCES: Rod Control System Design Basis Concern (both Salem Units) Event Date.: 5/27/93 Discovery Date: 6/4/93 Supplement Report Date: 8/26/93 The original report was initiated by Incident Report No. 93-263. This event is reportable in accordance with 10CFR50.73(a) (2) (v) (A). CONDITIONS PRIOR TO OCCURRENCE: 5/27/93: Unit 1 - Mode. 1 - Reactor Power 100% - Unit Load 1145 MWe 5/27/93: Unit 2 - Mode 3 Reactor Power 0% - Unit Load -o-MWe; Unit 2 Reactor startup activities in progress following completion of. the unit's seventh refueling outage. DESCRIPTION OF OCCURRENCE: on May 27, 1993, at 1844 hours, _control rod 1SA3 withdrew to approximately 15 steps from fully inserted in response to a manual insertion command (rod full out is 228 steps). Rod control power was deenergized to fully insert the rod. on June 4, 1993 at approximately 1700 hours, investigation determined that a single failure in the Rod Control System (RCS) could possibly result in a single rod withdrawal. event (applicable to both Salem Units). All control rods on Unit 2 were inserted at 1734 hours, the Reactor trip breakers were opened, and the Unit was stabilized in MODE 3. A Justification for Continued Operation (JCO) of Unit 1, with the RCS in manual control, was provided to the NRC on June 8, 1993 (reference evaluation S-C-RCS-EEE-0819). The Nuclear Regulatory commission (NRC) was notified of this event in accordance with the requirements of Code of Federal Regulations 10CFR50.72(b) (1) (ii) (B). The identified failure is conservatively postulated to be a single failure. The Updated Final Safety Analysis Report (UFSAR) states that only multiple failures would cause the withdrawal of a single Rod Cluster Control Assembly (RCCA). This condition in which a single RCCA withdrew results in an Unreviewed r* 1

SUPPLEMENTAL LICENSEE EVENT REPORT (LER) TEXT CONTINUATION Salem Generating Station Unit 2 DOCKET NUMBER 5000311 DESCRIPTION OF OCCURRENCE: (cont'd) LER NUMBER 93-008-01 PAGE 3 of 6 Safety Question, per 10CFR50.59. Public Ser~ice Electric and Gas {PSE&G) submitted an Emergency License Amendment to the operating License (for both Salem Units) to address the postulated RCS single failure. ANALYSIS OF OCCURRENCE: The RCS {AA} is used to withdraw control rods for reactor startup and to cqntrol reactor power during power operation. It consists of one Logic c,abinet, five Power Cabinets, and one Direct current "(DC) Hold Cabinet: The Logic Cabinet translates manually initiated or automatic commands into signals required by the Power Cabinets to step the banks of Shutdown and Control rod assemblies. This cabinet contains power supply assemblies and processes logic commands required for rod movements. - The Power Cabinets provide DC power.pulses to drive the Control Rod Drive Mechanisms {CRDMs) by converting three-phase alternating current {AC) power to DC power and applying it to the CROM magnetic coils. The DC Hold Cabinet is used to supply power to the stationary .gripper coi_ls of one group when required by Power Cabinet maintenance. Westinghouse was contracted for full RCS refurbishment service during

- the Unit 2 seventh refueling/maintenance outage to avoid aging-related RCS circuit card failures.

Wes'tinghouse supervised removal and testing of RCS logic cabinet and power cabinet printed c~rcuit cards, and their return to service. Numerous card problems were repaired, including suspect solder joints, arced and pitted terminals, and bad resistors. On May 25, 1993,* the PSE&G Controls Group satisfactorily -completed Individual Rod Position Indication {!RPI) calibrations and Control Rod Drop testing. At 2300 hours (same day), following.completion of prerequisite testing, initial reactor startup commenced. After encountering various RCS card failures, all control rods were inserted. _The cards were replaced or repaired, car¢l edge connectors adjusted, and rod testing was completed satisfactorily. It was during these repairs to the RCS that Pin No. 4 of the relay driver was repaired. bn May 27, 1993 at 1837 hours, reactor startup commenced. At 1844 hours, when-Shutdown Bank A was withdrawn to 20 steps, the Individual Rod Position Indicators (IRPis) did not indicate rod movement. A rod insertion signal was applied to the bank. As the

SUPPLEMENTAL LICENSEE EVENT REPORT (LER) TEXT CONTINUATION Salem* Generating Station Unit 2 ANALYSIS OF OCCURRENCE: DOCKET NUMBER 5000311 (cont'd) LER NUMBER 93-008-01 PAGE 4 of 6 l group step counter indication decreased from*20 too steps, the IRPI for rod 1SA3 indicated a rod withdrawal to 15 steps. The position of 1SA3 was verified and rod control power fuses w*ere removed to reinsert (drop) the rod. Troubleshooting revealed the withdrawal of 1SA3 RCCA was caused by failure of integrated circuit chips on two. RCS slave 1* cycler decoder cards. This failure resulted in inappropriate current orders being supplied to and simultaneously energizing the RCCA lift, movable, and stationary coils of all eight rods in Shutdown Bank A. The sequence of these current orders was such that the most probable result wo~ld be outward rod motion for all rods in the bank.

However, only 1SA3 withdrew. The*two logic cards (slave cycler decod~rs) were replaced with new, tested_ cards.

On May 28, 1993, a* functional check of.the Logic Cabinet was. satisfactorily completed. In addition, a recorder was connected to monitor the RCS 15 VDC power supplies. Station management obtained Westinghouse assurance that the May*21th withdrawal of 1SA3 had not damaged the rod-control rod drive mechanism. Post-maintenance and operability testing confirmed that the RCS was ready for reactor start-up, which commenced,at 1512 hours. At 1812 hours, Cont.rol Bank c, Group*l rods dropped and.the Reactor was tripped manually. The reactor t~ip of May 28, 1993 is reported in LER 311/93~001~00. Ori June 2, 1993, at 2053 hours, following satisfactory RCS post~maintenance testing and Operations surveillance testing, a reactor startup commenced. At 2338 hours (same day) a Pulse to*Analog (P/A) converter read zero for control Rod Banks B and D. Troubleshooting revealed a faulty supervisory data logging card with a failed integrated circuit chip and at 0347 hours (next day) all 'control rods were fully inserted. These problems were resolved ahd on Jurie 3, 1993 at 0620 hours, Reactor start-up commenced-and at 1359 hours criticality was achieved. On June 4, 1993, at approximately 1700 hours, it was determined that a design basis concern exists where a single RCS failure could result in the withdrawal of a single RCCA. This resulted in an Unreviewed Safety Question in accordance with 10CFR50.59. At 1705 hours, a Unit 2 Reactor shutdown commenced and at 1734 hours the Reactor Trip Breakers were opened. The Unit was stabilized in MODE 3. A Unit 1 Justification for Continued Operation (jco) with the.RCS in manual control was provided to the NRC for review (reference letter evaluation S-C-RCS-EEE~0819 dated June 8, 1993). Following a reactor trip on June 8, 1993, Unit 1 was subsequently maintained in MODE 3 (Hot Standby) pending resolution of the design basis concern. On June 20, 1993, following restart authorization, Unit 1 was subsequently synchronized to the grid. The Unit 1 Reactor trip is rep~rted in LER 272/93~011-00.

I i *. SUPPLEMENTAL LICENSEE EVENT REPORT (LER) TEXT CONTINUATION. Salem Generating Station Unit 2 APPARENT CAUSE OF OCCURRENCE: DOCKET NUMBER 5000311 LER NUMBER 93-008-01 PAGE 5 of 6 The apparent root cause of the postulated RCS single.failure design_ basis concern is "Design, Manufacturing, Construction/Installation" inadequacy, as per NUREG-1022. Subsequent testing and evaluation has demonstrated that failure of one slave cycler decoder card of an RCCA in conjunction with a rod motion command signal may cause an unplanned RCCA withdrawal. 1SA3 withdrew due to failure of integrated circuit chips*on _the 22AC and 22BD slave cycler decoder cards {AA/ECBD}. This resulted in

• inappropriate current orders being supplied to the RCCA operating.

coils causing the rod lift, movable, and stationary coils to energize at the same time. Although the current orders produced did not replicate those requi'red for a. normal rod withdrawal, the form of t_he resulting current orders was such that the most probable result would. be outward rod motion. However, only 1SA3 withdrew. The reason why on1y*1sA3 withdrew has been attributed to manufacturing tolerances. The Westinghouse owners Group is continuing to pursue resolution of this issue;; Failure of the integrated circuit (IC) chips resulted from the relay driver circuit card not making electrical contact with the surge suppression diode in the circuit. This allowed the back electromagnetic field from the RCS step counters to apply voltage transients to the components of the Logic Cabinet's circuit cards which caused failure of the cards' IC chips. The poor electrical connection was caused by a spread pin (no. 4) on the logic card. connector. The diodes suppress counter-electromotive force (CEMF) from the collapsing field coil of the group step demand indicators PRIOR OCCURRENCES: This is the first occurrence of a single control rod withdrawal at either Salem Unit.

SAFETY SIGNIFICANCE

This event did not affect the health and safety of the public. The. current Licensing basis, as described in UFSAR sections 4.3 and 15.3.5, assumes that only multiple failures would cause the withdrawal of a single Rod Cluster Control Assembly (RCCA). PSE&G now considers this event as an American Nuclear Standards Institute (ANSI) Condition II "FAULTS OF MODERATE FREQUENCY" event rather than a Condition III "INFREQUENT FAULTS" event. The Amendment/JCO issued June 17, 1993, discusses this in detail. For the current fuel cycles on each unit, analysis indicates two adjacent rods withdrawn from Control Rod Bank D (one from each group) is more limiting than one rod withdrawn from Control Bank D, which is addressed by UFSAR section 15.3.5. This increase in probability

• t 9*

SUPPLEMENTAL LICENSEE EVENT REPORT {LER) TEXT CONTINUATION Salem Generating Station Unit 2

SAFETY SIGNIFICANCE

(cont'd) DOCKET NUMBER 5000311 LER NUMBER 93-008-01 PAGE 6 of 6 resulted in an Unreviewed Safety Question, in accordance with 10CFR50.59. However, by taking.credit for the available generic thermal margins, the.Departure from Nucleate Boiling Ratio (DNBR) limit is still met. on June 17, 1993, Public Service *Electric and Gas Company (PSE&G), requested an Emergency Amendment to the Operating License, for both Salem Units, to address the potential RCS single failure analysis.

CORRECTIVE ACTION

_On June 17, 1993, Public Service Electric and Gas Company (PSE&G); requested an Emergency License Amendment to the Operating License/Justification.for Continued Operation, for both Salem Units, to address the postulated RCS single failure (reference NLR-N93098). 'The RCS Logic Cabinet slave cycler stationary decoder card,* Westinghouse Part No. 3359C62G02, and the slave cycler ll,lovable card, Westinghouse Part No. 3359C62G03, were replaced to restore operability* of rod 1SA3. Additional suppression diodes were installed on the group step demand indicators (the source of the CEMF) of the RCS circuitry of each Unit. This action eliminates the consequences of an open or bad . connection on the relay driver circuit card connector pin no. 4. All RCS logic cards on Unit 2 were replaced and satisfactorily tested. In addition, all RCS Power Cabinet cards were pulled, visually inspected, and satisfactorily retested. Additional immediate corrective actions to this event, as committed to on June 18, 1993 in a meeting at NRC Region I, have been implemented. PSE&G Nuclear Fuels will incorporate this new Condition II event into the fuel.reload safety analysis for each unit. The UFSAR will be reviewed and revised as appropriate in reference to this event. Recommendations resulting from the Westinghouse owners Group evaluation of this event wiil be evaluated. MJPJ:pc SORC Mtg. 93-081 U. General Manager - Salem Operations l }}