ML20247D556

Augmented Insp Team Rept 50-293/89-80 on 890413-15 & 17-26. No Violations Noted.Major Areas Inspected:Causes,Safety Implications & Util Response to 890412 RCIC Sys Suction Piping Overpressurization

ML20247D556
Person / Time
Site:	Pilgrim
Issue date:	05/08/1989
From:	Eugene Kelly NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:
Shared Package
ML20247D552	List: ML20247D548 ML20247D556
References
50-293-89-80, NUDOCS 8905250468
Download: ML20247D556 (122)

See also: IR 05000293/1989080

Text

{{#Wiki_filter:U.S. NUCLEAR REGULATORY COMMISSION REGION I Report No.: 50-293/89-80 - Docket No.: 50-293 Priority Category C -- License No.: DPR-35 Licensee: Boston Edison Company 800 Boylston Street Plymouth, Massachusetts 02199 Facility Name: Pilgrim Nuclear Power Station Insr > .cion At: Plymouth Massachusetts Inspection Cor. ducted: April 13-19, 1989 Inspectors: Joseph A. Golla, Team Member, RI Theodore A. Easlick, Team Member, RI Tae K. Kim, Team Member, RI Daniel G. Mcdonald, Jr. , Team Member, NRR Lambros Lois, Team Membe NRR ! Approved By: / Eugene M elly, Team Le r, RI /Date Inspection Summary: An Executive Summary can be found in Report Section 1.0, and recommendations for corrective action are made in Section 2.3. Potential enforcement issues are being considered separate from this inspection. The AIT was conducted onsite from April 13-15, 1989, and in the Rer,1on I office

during April 17-26, 1989. The scope and composition of the AIT are described l in Sections 2.1 and 2.2. The AIT Charter is included as Attachment I to this report. A synopsis of the results of the licensee's internal Oversight Commit- tee investigation of the April 12, 1989, RCIC Pressurization Event is included as Attachment 2 to this report. l Qsagg $$hu# G i

, .I 1 l TABLE OF CONTENTS 4 4 ] Page i _ ~ ] 1.0 EXECUTIVE EUMMARY......................................... 1 l l 2.0 INTRODUCTION.............................................. 4 j ' 2.1 Scopo of Inspection.................................. 4 2.2 Team Composition..................................... 5 ! 2.3 Summary of Recommendations........................... 5 ' 3.0 EVENT DESCRIPTION.......................................... 8 3.1 Initial Conditions and Background.................... 8 i 3.2 Conduct of RCIC System Logic Test.................... 9 ' 3.3 Leakage and Termination of Event..................... 10 3.4 Recove ry o f RC I C Sy stem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 , 3.5 Plant Shutdown on April 16, 1989..................... 11 l 3.6 Similarity to 1983 HPCI Event........................ 11 4.0 OPERATIONS AND HUMAN FACT 0RS.............................. 14 4.1 Background........................................... 14 4.2 Tagging Errors....................................... 14 4.3 Independent Verification............................. 16 4.4 Conduct of Operations................................ 17 4.5 Training............................................. 18 4.6 NWE and STA Ro1es.................................... 18 4.7 E0P-4 Entry Condition................................ 19 4.8 Conclusions............. ............................ 20 5.0 RCIC SYSTEM PERFORMANCE AND TEST HISTORY.................. 21 5.1 Background........................................... 21 l ' 5.2 Adequacy of Testing.................................. 21 5.2.1 Mechanical Exercise........................ 21 5.2.2 Leak Testing............................... 22 5.3 Operability Verification............................. 22 5.4 Logic System Functional Testing...................... 23 5.5 Analogous HPCI System Testing........................ 24 5.6 Itolation Valve Performance.......................... 25 l l 5.6.1 Leakage Test History........... ........... 25 5.6.2 Containment Isolation Issues............... 27 5.6.3 Retest of CK-1301-50....................... 28 i .. _ _ _ - _ _ - _ _ _ _ - _

- - __- - ._ _ _ _ _. _ _- Table of Contents (Continued) Page 5.7 Check Valve 1301-50 Design............................ 29 5.7.1 Design Description......................... 29 5.7.2 Modification to Remove Air Operator........ 30 5.7.3 Ma i n t e n a n c e H i s to ry . . . . . . . . . . . . . . . . . . . . . . . . 31 5.7.4 Furmanite Application...................... 31 5.7.5 Failure Mode............................... 33 5.8 Conclusions.......................................... 34 6.0 EFFECTS ON RCIC EQUIPMENT AND SYSTEM RECOVERY............. 36 6.1 Background........................................... 36 ! 6.2 Suction Pressure Switch.............. ............... 36 i 6.3 Pump Inspections / Repairs............................. 37 j 6.4 Environmental Room Effects........................... 37 l 6.5 Suction Pipino....................................... 37 6.5.1 Weld Examinations.......................... 37 6.5.2 Repairs.................................... 37 6.6. Hanger Walkdown...................................... 38 6.7 Water Level in RCIC Quadrant................... ..... 38 6.8 Relief Valve bench Test.............................. 39 6.9 Conclusions....... .................................. 39 7 . 0 RADI O LOG I CA L I M PACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 7.1 Background........................................... 40 7.2 Immediate Surveys end Samples........................ 40 7.2.1 Airborne Hazards........................... 40 7.2.2 Decontamination and Exposures............. 41 7.2.3 Sampling...................... ............ 42 I 7.3 Conclusions................ ......................... 42 i l 8.0 OVERPRESSURIZATION IMPLICATIONS........................... 43 l 1 8.1 Background............... .. .............. ......... 43 ' 8.2 Thermal Hydraulic Conditions of the Discharge Line... 43 8.3 Discharge Capacity of the Suction Thermal Relief Valve.............................................. 44 8.4 RCIC System Low Pressure Piping...................... 44 ii . _ _ _ _ _ _ _ _ - _ _ _ - _ - _ _ _ _ _ _ _ _ -

____-- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . . i

Table of Contents (Continued) { i i Page ! 8.4.1 Design Limits.............................. 45 , 8.4.2 Pressure and Hydrodynamic Pulse Loading.... 45 i 8.4.3 Thermal Loading................... ........ 46 i 8.4.4 Hangers and Supports....................... 46 8.4.5 Pipe Inspection and Testing................ 46 8.5 Environmental Impact................................. 47 8.5.1 Extent of Leakage and Condensation......... 47 8.5.2 Temperature Alarms and Settings............ 49 8.5.3 Environmental Qualification Design......... 49 8.6 PRA Assumptions............,......................... 49 8.7 Event V Evaluation................................... 50 8.8 Conclusions.............. ........................... 51 9.0 ORGANIZATIONAL RESPONSE........... . ..................... 53 9.1 Background........................................... 53 9.2 Immediate Shift Response............................. 53 9.3 Management Critiques................................. 54 9.3.1 Event Critique............................. 54 9.3.2 Investigation Teams........................ 54 9.3.3 Management Oversight and Assessment.. 56 ..... 9.4 Peer Review Process.................................. 56 9.5 Corrective Actions................................... 57 9.6 Deportability Considerations......................... 58 9.7 Conclusion on Quality and Scope of BECo Response..... 60 10.0 MANAGEMENT MEETINGS....................................... 62 iii _ _ - _ _ _ _ _ _ _ _ _ _ _ -

-- -. - _-... _ _ _ - _ _ . _ _ _ _ __ - _ _ - ._ _ _ _ _ - . _ . . _ _ _ - _ - _ _ _ _ _ _ _ - _ _ _ - - _ - Table of Contents (Continued)- Page APPENDIX A - Proximate Causes of April 12, 1989 RCIC . 0verpressurization........................... .. A-1- ~ APPENDIX B - Principals Contacts............................... B .1 - ' APPENDIX C - Detailed Sequence of Events....................... -C-1 APPENDIX D - RCIC P&ID.. ...................................... D-1 . APPENDIX ~E - Photographs (A through L).................-......... E-1 > APPENDIX F .RCIC Quadrant Area P1an........................... F-1 APPENDIX-G -Acronyms.......................................... G-1 ' APPENDIX H - RCIC Discharge Piping Isometric Drawings.......... H-1 APPENDIX I - RCIC Suction Piping............................... I-1- . APPENDIX J -'RCIC System Lineups; Normal, Test and Faulted..... J-1 APPENDIX K - RCIC Suction Pipe Relief Valve PSV-31 Weld Repair.......................................... K-1 APPENDIX L - Check Valve 1301-50 Vendor Prints and Sketches.... L-1 ATTACHMENT 1 - AIT Charter ATTACHMENT 2.- Boston Edison Company Investigation Results, Oversight Committee Report ATTACHMENT 3 - Pilgrim Licensee Event Report (LER) 83-048 . 1 iv _ _ _ _ - _ _ - _ - _ _ - - _ _ _ _ _ _ .

_ . _ _ _ - _ _ _ - _ _ _ I { !

I i l 1.0 EXECUTIVE SUMMARY l The Pilgrim Station experienced a pressurization transient condition of ' the Reactor Core Isolation Cooling NCIC) system at 8:45 a.m., on April 12, 1989, during the conduct of a RCIC logic system functional test. The duration of the transient was approximately two minutes, and involved a spill of approximately 100 gallons of fluid which was stagnant water from the RCIC discharge piping (upstream of a check valve) and possibly a small amount of higher energy feedwater and reactor water cleanup fluid (downstream of the check valve). The transient resulted in RCIC quadrant space temperatures that did not exceed 160 F, and was expeditiously ter- minated by control room operators. Therefore, the safety significance of this event was minimal. The licensee concluded, and the Team subsequently concurred, that this event was not reportable per 10 CFR Part 50.72 because it did not represent a significant degradation of a principal safety barrier. l There were eleven proximate causes (listed in Appendix A to this report) I identified by the Team which, considered collectively, potentially repre- sent a broader and more fundamental question regarding the licensee's defense in depth in the conduct of operations. Significant weaknesses at several working levels in the licensee's operational organization were revealed, and exposed potential root causes related to: Shift supervision's overall awareness and control of surveillance -- testing. Licensed operator cognizance of plant status and equipment conditions. -- -- Procedural adhe ence and validation. Independent verification of tagouts. -- Attention to detail and work quality. -- The RCIC pressurization transient was similar to a September 1983 over- pressurization of high pressure coolant injection (HPCI) system suction piping, although the root causes of the 1983 HPCI event were different (i.e., poor communications and a different kind of check valve failure). The HPCI and RCIC check valves were replaced in kind as a result of the 1983 event. The April 12,1989, RCIC event was evaluated in light of WASH-1400 Event V (inte" system LOCA) considerations and concluded by the Team to not be a signif, cant precursor. This is because of (a) the inter- face with feedwater and reactor water cleanup (RWCU) fluid which is sepa- rated from reactor coolant by an additional barrier (i.e., the inboard feedwater isolation valve), as well as (b) the isolability of the event by virtue of two available motor-operated valves on the RCIC discharge pip- ing. Also, the low pressure RCIC suction piping adequately withstood conservatively calculated hydrodynamic loads in excess of minimum Code allowable values. _ _ _-__ _ __ ._

_ _ _ _ - _ _ _ - _ _ _ _ 2 The effects on equipment from the overpressurization were thoroughly evaluated and found to be such that future RCIC operability was not adversely impacted. RCIC system low pressure piping and supports were conservatively analyzed and concluded to not be degraded. The RCIC pump and turbine were similarly unaffected but will nonetheless be functionally tested prior to power opration. An indication was detected by non- destructive examination of a piping socket weld attachment for the low pressure RCIC suction piping thermal relief valve. However, the direction and orientation of the indication is not what would otherwise be expected as a direct result of the overpressure event, and probably existed prior to the event. The only outstanding question as of the end of the inspec- tion was the proper leak-tightness of the RCIC discharge check valve at hot operating conditions. The check valve was found to be binding from three previous applications of Furmanite sealant during the period of March 1985 to April 1986, that impeded free movement of the valve disc, and created a situation where subsequent routine freedom of movement checks could have left the disc off of the seat prior to the April 12, 1989 event. This check valve is one of two immediate barriers to high-low pressure interfacing protection for i RCIC. This type of check valve is one of two currently installed and operational at Pilgrim (and apparently unique to the Pilgrim plant); the other is in the similar HPCI discharge line which was also inspected, -sith no Furmanite or binding found. The previous application of Furmanite was in accordance with a station program which did consider the use of the sealant and its effect upon RCS chemistry and the potential for IGSCC but did not employ pre-determined sealant volume control. The subsequent permanent repair of the RCIC check valve in August 1987 failed to ade- quately remove all of the Furmanite from the valve's internal bushing and actuator shaft. The unremoved Furmanite impeded free swing of the disc and possibly left the disc 10-15 degrees off of its seat during the last IST forward flow exercise (during cold shutdown) on January 13, 1989. Although a RCIC flow test was conducted on April 3,1989 as part of reactor startup, actual RCIC flow through the check valve never occurred since its original installation in October 1984. The check valve may never have returned te a fully seated condition after January 13, 1989 until the event on April 12, 1989. The Team concluded that the licensee's internal response to the event was ' well organized, properly paced and focused, and comprehensive in its identification of root causes. Corrective actions in the short term were judged to be appropriate, although broader more long-term corrective actions remain to be assessed by both the Boston Edison Company (BECo) Peer Review processes in place as well as the NRC's Restart Panel, in light of the human factors concerns identified by the Team and licensee. ___ _ - - _ _ _ _ - - _ _ _ _ - _ _ _ - _ _ _ _ _ - _ _ _ _ _ _ -

_ _ _ _ _ _ _ - _ _ _ _ , 3 The licensee's cooperation with the Augmented Inspection Team (AIT) was excellent, allowing for the evaluation of a large amount of technical information to be well understond in the time allotted to the Team. One deficiency noted by the Team was the difficulty in reconstruction of accurate post-event times and details due to the unavailability of process computer data because of the incomplete operational status of their Emerg- ency Plant Information Computer (EPIC). However, this was coinpensated for in part by an effective Human Performance Evaluation fystem which, al- though not yet formalized, was well executed following this event by reliance on personnel interviews. Precise times during this event were not critical due to short event duration, but these would be more critical if the event was more prolonged and com,) lex. The reactor remained at 25's power, with the RCIC system declared inoper- able following the April 12, 1989 event, until a planned shutdown was begun on April 16, 1989, when the licensee concluded that RCIC could not be returned to service prior to the 7-day LCO because of the indeterminate state of the check valve. The Team identified one potentially generic concern associated with the location of the RCIC discharge check valve. Because of the design con- figuration of the RCIC check valve, CK 1301-50, located downstream of the normally closed motor-operated discharge valve, MO 1301-49 (49 valve),the intervening fluid trapped between these valves is expected to heat up and pressurize, with a potential to recreate the April 12, 1989 event by either: (a) inadvertent opening of the 49 valve, or (b) initiation of the RCIC system with normal opening of the 49 valve which may present to the RCIC pump a backpressure which would compete with pump flow / head develop- ment. The licensee's engir,eering organization is reviewing this concern for potential corrective action at Pilgrim. l 1 I l l

-_ - - _ _ _ - _ - _ _ _ _ - _ _ _ _ - _ - _ _ - _ _ _ - _ _ - _ - __ _ .-- -_ - - - _ _ _ _ _ _ _ - l l L 4 L l 2.0 INTRODUCTION 2.1 Scope of Inspection The AIT was chartered on April 13, 1989 (Attachment 1).to review the causes, safety- implications, and licensee's response to the April 12,1989 Reactor Core Isolation Cooling (RCIC) system suction l piping overpressurization. The AIT was concluded with an exit meet- ing held in the Region I office on April 19, 1989. The Team prepared a detailed chronology of events (Section 3 and Appendix C), performed a comprehensive evaluation of the licensee's' conduct of operations and human factors errors which directly . led. to the event (Section 4), and assessed the effect of the event on RCIC system equipment (Sections 6 and 8.4). The Team also independently determined the safety significance of the event in light of WASH-1400 Event V Intersystem LOCA potential (Section 8.7), including a care- ful . review for deportability (Section 9.6) and comparison of this event to a similar HPCI event in 1983 (Section 3.6). The AIT assessed the maintenance and test history of RCIC and HPCI system valves, ~particularly the discharge check valves (Section 5). Proxi- mate causes were deduced by the Team (Appendix A). Finally, the Team judged the scope and quality of the BECo response to the April 12, 1989 event via their formation of an Oversight Committee to direct three separate investigatory teams in search of causal factors and proposals for corrective actions (Section 9 and Attachment' 2). The Team arrived onsite on April 13, 1989 and remained onsite through April 15, 1989. Information was later exchanged and a preliminary presentation made by the licensee in the Region I office at the end of the inspection on April 19, 1989, followed by an exit meeting on that same date. Additional relevant information concerning: final stress calculations; disassembly, inspection and repairs to the RCIC and HPCI system discharge check valves; and, thermal hydraulic con- sioerations in the RCIC discharge piping, was evaluated in the ensu- ing week after the exit (April 20-27,1989), and the AIT inspection was then considered complete. Principals contacted during the course of the inspection are listed in Appendix B. l - - - _ __ __ - - _____ . _ _ __ _ _ - - _ _ _.

_ - - _ -_-_ -. _ .__. _ __ _ ._ __ _ _ _ _ _ - __ _ _ - _ _ _ _ _ - _ _ _ _ _ . 5' 2.2 Team Composition The AIT was led by a Region I Division 'of Reactor Projects Section Chief,_ and was composed of five full time Team Members including: the Pilgrim Resident Inspector, the . NRR Project Manager an NRR Thermal Hydraulics Specialist,' a Region _I Operator Licensing Examiner and a Region I Test Programs Specialist. The AIT was also augmented y by the remaining members of the Pilgrim resident inspection staff as well as representatives from the NF.C's Pilgrim Restart Team who pro- vided coverage of site activities. prior to 'and subsequent to the AIT's departure from the site. During conduct of the AIT, the Team Leader took direction from the NRC's Deputy Regional Administrator for Region I, and remained separate from NRC Restart Panel considerations or. decisions. 2.3- Summary of Recommendations The . consensus of _ the Team was that BECo's operating staff responded well immediately following the event, and that BECo Oversight Commit- tee and Management Critiques capably recon,tructed events, causal factors and made sound corrective action proposals. The Team's con- cerns for the human factors issues which led to the errors that caused the April 12, 1989 event, resulted in the following principal recommendations: Consider a Pilgrim Station policy aimed towards retaining the -- NWE in the main control room following an event, until assurance (with the help of an STA) is reached that stable plant condi- tions, E0P's and EAL's are fully understood and appreciated. Organize and fit pre-evolution test briefs into routine shift -- activities involving complex surveillance testing as an added layer of defense in-depth towards safe operations. -- Consistently stress true verification of equipment manipulations (tagging), and independently monitor and assess the effective- ness of this fundamental precept of operation. Human-factor test procedures with appropriate bold precautions, -- consistency between component titles locally (i.e., at the device or breaker) and at control panels, and use that termin- ology unambiguously on tagouts. Also, stress strict adherence to test procedures and proper use of procedure change notices (PCN's) when discrepancies are found. - _ _ _ _ _ _ _ - _ - - - _

_ _ _ _ _ _ _ _ _ _ - _ ___ ____________ __ _ _ _ _ - _ _ _ _ - _ _ _ _ _ _ _ 6 Stress detailed logkeeping and formalize the Human Performance -- Evaluation System (HPES) method utilized by the post-event Over- sight Committee, particularly while EPIC operability is pending, to enable accurate reconstruction of event data. Also, return the inoperable channel of EPIC to service as soon as practicable. Use tags at the main control switches (as well as locally at the -- devices), when tagging equipment for test or maintenance. Establish clear administrative controls as to the role of -- licensed operators in the conduct of surveillance testing. Regarding the Team's conclusions with respect to Pilgrim Station's design vulnerabilities (in light of Event V) RCIC system which can be contributed to the April 12, 1989 transient, the following .ecom- mendations were made: -- Revise alarm response procedures to explic';1y describe the "Possible Cause" of high suction pressure as potential inter- facing system pressurization. Log suction pressure as part of routine operator rounds. -- Resolve the appropriate space temperature alarm settings in the -- RCIC and other Reactor Building quadrants, and make these con- sistent with entry conditions to E0P-4. Also, routinely log space temperatures using back panel recorders. Add explicit precautions to the HPCI and RCIC LSFT procedures to -- warn against simultaneous opening of both motor-operated dis- charge valves, and its consequence as experienced in September 1983 (HPCI) and April 1989 (RCIC), in the context of Event V considerations. -- Use volume control pre-estimates when applying Furmanite seal- ant, and track multiple instances of Furmanite application (to the same component) prior to component disassembly and repair. -- Make the April 12, 1989 RCIC event, including BECo's corrective actions, available on the INPO Notepad Network. l - - _ _ _ _ _ _ _

. f 1- ( i; . 7 L Finally, with respect to calculational and analytical issues, -the Team recommends that BECo: Notify NRC representatives of .any substantive changes in pipe -- stress calculations as these are formalized, independently ver- , ified and approved by NEO, and make these final calculations available to Region'I for information. Review the design of HPCI and RCIC, particularly on the fluid -- discharge side, and investigate system operation and logic .in . light of the expected performance of the discharge check valves. L- Consider the feasibility of logic changes (interlocks or per- missives) to motor-operated discharge valves, or the addition of a check valve at the pump discharge, .in consideration of back pressures experienced' during normal starting transients where full HPCI or RCIC injection into the feedwater system is expected. Following the ~ completion of such a review, which would be expected to ' be completed within the next two to three - months, the licensee should inform NRC Region I, of the results and conclusions. The . Team recognizes that many of the above recommendations were already adopted by the BECo Oversight Committee. l l _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

,____ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 8 l 3.0 EVENT DESCRIPTION i The following narrative provides a broad overview of the activities and ' actions which occurred at the Pilgrim Nuclear Power Station (PNPS) prior to, during, and subsequent to an event which resulted in an overpressure transient in the suction piping of the RCIC system during performance of a required system test. The transient also resulted in tiie lifting of a pressure relief valve on the suction piping releasing water. The water was released into the RCIC quadrant area drain system, backflushing con- taminants into the general area. Subsequent to the event, five workers' shoes and one of those worker's hands were slightly contaminated. The shoes and hand were promptly decontaminated. , l A detailed Sequence of Events and Actions for the first day of the tran- sient are provided in Appendix C. Specific aspects and details of the event which have technical and safety significance are provided in other sections of this report. Sketches of the RCIC system in both the normal and test lineups (Appendix D) are included in this Section for immediate reference. 3.1 Initial Conditions and Background On the morning of April 12, 1989, the PNPS was operating with the reactor at about 25 percent of rated power and the main turbine generator providing 125 MW of electrical power to the distribution grid. The reactor pressure was 950 psig and the reactor coolant temperature at 530 F. The PNPS is currently in a Power Ascension Program after an extended outage. The NRC has a Restart Staff onsite which is monitoring the Power Ascension Program. The PNPS Technical Specifications require a logic system functional test (LSFT) of the RCIC system every six months. The RCIC system's design basis is to provide makeup water to the reactor as part of planned operation for periods when the normal heat sinks are not available and during certain transient conditions. RCIC provides capability for reactor core cooling if the HPCI system, which is an Engineered Safety Feature (ESF), is unavailable. However, the RCIC system is not taken credit for in the safety analysis of design bases events and is therefore not an ESF system. The purpose of the RCIC LSFT is to de~ mtrate that, when a preset high water level in the reactor is en. . xd, the RCIC turbine pump will trip but subsequently, will automatically restart when reactor water level reaches a preset low level. The performance of the test is the responsibility of the Instruments- tion and Control (I&C) technicians and requires operations support both inside and outside of the contro? room for pretest readiness and performance of the test. - - - _ _ - _ - _ _ _ - _ - - - _ _ _ _ _ _

h I 7 t S L 1 N S - E E O 10 O RO L 3 T I 1 SS ,RO PP D PT L E O TR P E f7 R O m U N S S E O P P P V2 > P L 6 O. O C 1 U \\ G3 - 1 0 S Y Y 3 L L 1 L L A A M M I 1 S 3 P R R - 1 0 0 O O 0 3 N N 1 1 h/ g>L - V 6 \\ J / 2 g'7 \\ML Y 1 g' 0 / 3 - AP 1 EP TU SS 2 E 2 T - AE 1 NA*"/ SG

6

'c" / c 1 i ER^ \\ 2 'R DO NT'\\ W f1 - 1 R OS C 0 A )(- 3 1 0 C 6 O T - \\/ 1 2 R EG 3 0 6 E - S E yJ v 1 1 0 RN T 3 A L1 AEA 8 BD SP J 4 3 1 5 - N NM 0 - O E U 3 1 E v_9 0 C D P 1 3 N L1 y O J C R ME ,9 - D E 4- - E V ,0 E T - - 1 - 3 E A - W L - _ T 1 AP - S V U Y LE 50 1 Ai 0 - S AN 8 1 - A I 1 3 '2 - C ML 0 1 6 2 - R U 1 V 6 - C ~C ICO W . / [# 7 N R R M6 A 85-

lll1i

g k

, '

' . . 16 g ' I- N g' [o O t I g I 7 1 S L 1 N S O E E O RO L I T 1 S PT PP D L S RO E O TR P E H U N S f7 RO S E O T P P yV > P L 6 P os O C 1 g0 /S\\ U \\ ov - 1 Y Y I 3 L L / L L 1 A A M M I 1 S 3 R R P - 10 O O M3 N N 1 1 NT g3 E k - N y 6 U B A 2 7 Y k 10 L / 3 AP 1 EP TU SS E 22 T - AE 10 SG K gg3 g Oc* NA N r g e 1 ER A 3j 2 n N DOT W 1 O - NT 1 R I OS C r S C A O 1 0 S L 6 E O T - RO 1 \\/ g1 R 2 0 PP E 3 r P S E h 1 U V T E1 S EA A 8 SP 4 3 y1 5 E - N NM ml 0 T - O E U p3 g0 AE 1 C D P 1 3 SG K N P_ 1 yN A N O 1 ER A C 9 R DOT M mk 4 D E NT - y0 E T OS 1 E T 3 E A C 1/FW\\ T S P \\ S E U Y T E 05 S DN 1 /, - 8 1 EI - 0 fid I L 3 1 C R 0 1 2 - E u I 1 V U V c _ <C IC QL w V Yf - R EA R gaA RV s mg ,lll ' l

_ _ - _ . - _ - _ _ _ - _ _ _ _ - _ _ _ _ _ _ _ - _ _ - _ - _ _ _ _ _ _ - - _ _ _ _ _ - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ L 1 ! 9 3.2 Conduct of RCIC System Logic Tesj About 75 30. a.m. on, April 12, 1989, the lead I&C technician responsi- , L ble for performing the RCIC system LSFT reported to the Nuclear Oper- ations Supervisor (N05) and Nuclear Watch Engineer (NWE) in the con- trol room to receive permission to perform the test and coordinate the ' test activities. The I&C technician. filled out the tagout sheets- and tags required for establishing a vcive lineup for '5e test. The NWE approved the tagouts, the NOS reviewed the tagoW and the NOS had the control . room operator verify or align the eight RCIC system valves specified by the procedure, of which seven were to be tagged out of service. The lead I&C technician sent two other technicians to the cable spreading area where RCIC system logic relay panels are located. Two operators were assigned to. perform tagout of seven RCIC system valves in the reactor building motor control center where the electrical circuit breakers for the valves were located. The operators, accom- panied by the lead I&C technician, proceeded to the reactor building and performed the tagout by: (I) dividing the seven tags; (2) per- forming the tagouts; and (3) then each verifying the other's work. The two operators and the lead I&C technician returned to the control room. The operators signed the tagout as being completed and the lead I&C technician signed the tagout sheet as being accepted by I&C. The tagout sheet was then signed by the NOS indicating that the test could proceed. Six of the seven tagouts were incorrect; one due to a procedure error and five resulting from not positioning the electrical circuit breakers (CB) properly. The procedure error also resulted in removal of power from another RCIC system valve which was not in accordance with the test procedure. As the result of the procedure error and not positioning the electrical CBs correctly, the position indica- tions on the control board in the control room were incorrect for the seven RCIC system valves. The lead I&C technician established communications with the I&C tech- nicians at the logic panels in the cable spreading room and coordi- nated the test with them and the control room operator. The control room operator initiated the RCIC turbine trip button, verified that the steam to the turbire was isolated, and reset the logic for the auto isolation signal reset. The automatic restart signal was then simulated at the logic cabinets by the I&C technicians. - _ _ _ _ _ - _ _ _ _ _ _ .

g 7 1 S L 1 N S - E E O 10 O RO t 3 I T 1 S P P v7,1 L S P E O U ,{P RO S P P y > 6 P o 1 U \\ c - 1 S\\ 03 D 1 - E N S 1 m3 E O - P L 1 0 0 0 3 O C 1 1 N/ . g3 i - 6 V \\/ , s7 / k 2 2 \\M L / Y 1 , 'd - AP 1 EP TU SS 2 E 2 T - AE SG*/ b8 S" NA / c ' E R "^ h 2 DO - NT'\\ w f1 - 1 n OS c ^ ][ 0 3 C O @ 1 T \\/ 4 3 1 2 R E 0 E 1 1 S E y 0 RN T 3 A 1 AEA 8 BD SP 4 3 N NM gA1 5 - - 0 O E U gD3 1 E 0 C D P 1 N r_' 3 1 O 1d de C 9 R v S gA4 DE ei in M F A gD1 E T ed - c 3 E A s e 0 O W r e r E W t i v ou 1 anq T P2 - l 1 v 2 e U1 E 1D r6 r S EP 0 os R 1.E f 1 a 5 Y NE U0M 8 1 n0 l' 3 - o3m 1 T 0 A - S L S D 1.R I i 1 o 1 t 0 2 ado 1 i 2 - r E2 U 1 v H cn - CE N CO idal o 2 F C _ <c ( rA V n0 - t 2 E O .R W /h/ - I

i6n1

CL H R M. E Enop - - A R A Te0c 1 R VWP8P M6 8 O p3 nt 5 NO1 is - '

_ _ _ - - - _ _ _ _ _ i 1 - l 10 3.5 Leakage and Termination of-Event t The . RCIC . turbine did not s: art because the trip / throttle valve had been closed by previous control room operator action. However, the valves, which were incorrectly left with their electrical power con- nected, opened. The discharge side of .the RCIC system connects to the high pressure Reactor Water Clean -Up (RWCU) system which is ! ' slightly over 1000 psi' at about 400 F. The . incorrect opening of the two valves in the high pressure RCIC discharge piping resulted in a pressure transient in the low pressure RCIC system suction piping. As the result, several alarms were initiated in ' the control room following the RCIC high pressure alarm and this lifting of a relief valve in the suction piping. These alarms included RCIC turbine bearing high pressure, two floor leakage. alarms and two smoke detec- tor alarms. Steam and hot mist, after the opening of the relief valve, initiated the smoke detectors. The NOS instructed the control room operator to terminate the test and isolate the RCIC system. The NWE and the two operators left the control room area and proceeded to the vicinity of the .RCIC quadrant to investigate the source of the alarms. 3.4 Recovery of RCIC System The NWE and the two operators observed vapor rising from the RCIC- quadrant and "B" RHR areas and notified Health Physics (HP). A Radiation Protection (RP) technician arrived shortly and initiated radiological control activities including radiation sampling and measurements. The RP technician informed the NWE of water on the floor of the RCIC quadrant. The NWE called the NOS in the control room to have him open floor drains in the RCIC area and the "B" RHR area. While waiting for results from the radiation surveys, the NWE and operators checked the tagouts noting the errors. The NWE direc- ted the operators to remove the tags and return the RCIC system valve circuit breakers to their normal position. The RP technician cleared the RCIC area for entry. The NWE and two operators wearing boots, shoe covers, gloves and lab coats entered the area. The NWE noted that the floor was wet, indications of floor drain " blowback" (or rust-like residue) were seen over much of the area, and portions of the suction piping were cool to the touch. However, the discharge piping, RCIC pump casing, and a portion of the suction piping (including the pressure relief valve) were hot to the touch. He also noted that these components, including the turbine and motor operator, were dry and had a film of rust. The NWE's observations led him to believe there was a problem with the check valve between the high pressure hot water in the RWCU system and the two valves in the discharge piping of the RCIC system. _ - - _. ___ _-_. __ __ . _ _ _ - _ _ _ _ _ _ - _ _ _ _ _ _ - _ _ - _ _ _ _ _ _ _ _ _ _ - _ - _ _ - -

.__ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - _ _ 11' The NWE left the area and returned to the control room to : inform management of his observations. Based on the overpressure transient, which occurred at 8:45 a.m., the -subsequent observations of the NWE and discussions with site operational management, the RCIC system was declared inoperable at 10:00 a.m. and a seven day Limiting Condition for Operation-(LCO) was entered per the Pilgrim . Technical Specifica- tions (TS). The HPCI system subsequently was tested and verified to be operable. 3.5 Plant Shutdown on April 16, 1989 A plant shutdown was begun. on April 16, 1989 -' four days after the overpressure transient which caused the RCIC system inoperability - and the plant was placed in a Cold Shutdown condition by April 17,1989. The shutdown, during which an Unusual Event was declared because the RCIC system was projected to be not capable of return to service in the 7 days allotted by the TS LCO, is described in Section 9.6. 3.6 Similarity to 1983 HPCI Event On September 29, 1983, while conoucting HPCI system logic surveil- lance tests, an overpressure transient occurred .in the low pressure piping of the HPCI system. The Team reviewed the - Licensee Event Report, LER 83-48, dated September 30, 1983 which provided details of the event. The cause of the event was personnel. error resulting from poor communications while conducting more than one surveillance test on the HPCI system at the same time. As the result of the communi- cation problem, both motor operated valves in the HPCI discharge line opened and the check valve was partially open resulting in the over- pressure transient. Shortly after the event, the control room oper- ator attempted to perform an operability test on the HPCI system. The auxiliar.9 oil pump was started and the turbine stop and control valves opened. The stop valve could not be tripped from the control room and the HPCI system was declared inoperable and the required Technical Specification actions initiated. As reported in NRC Inspection Report 50-293/83-19, . dated October 23, 1983, the licensee's investigation determined that the root cause was verbal miscommunication between the control room oper- ator and an I&C technician. The immediate actions taken by the licensee were to hold a critique with the operators and I&C tech- nicians, and initiate administrative actions to ensure strict compli- ance with surveillance precedure action steps. Followup actions included: system and component inspections; replacement of a failed m ___i_..________._ _ _ _ _ . _ _ _ _ _ _ l

- _ _ _ _ - _ _ - _ - _ - _ - 12 ,- plant seal -condenser gasket and failed turbine -stop valve trip solenoid; check valve corrosion repairs; and, analyses. to determine the status of the low pressure piping. The HPCI . system was declared operable on October 2,1983, and, as stated above, transient analyses l were initiated to ensure long-term operability of the low pressure piping. The 'HPCI system injection check valve was determined to be partially open during the 1983 HPCI transient due to the valve stem being rusted to the actuator device. This condition was determined to have possibly held the check valve disc in a partially open ' position dur- ing normal operation, with no differential pressure across it, but to not.cause the salve to bind. The corrosion problem was repaired, and power was reduced to perform HPCI system operability testing. The check valve had a history of leakage which, following this corrosion experience, caused the licensee to replace the valve in kind as part of a Pilgrim Valve Betterment Program. NRC Inspection Report 50-293/ 83-23, issued in December 1983, concluded that the licensee had ade- quately reviewed and. reported the circumstances surrounding the event. Although there are similarities in the September 1983 and April 1989 transients, there are also distinct differences. The 1983 transient occurred because two tests were being conducted due to communication problems between the control room operators and an If technician. No errors in the two test . procedures being implemented were identi- fied. The more recent 1989 transient was primarily the combined result of a procedure error, improper tagouts, and incorrect inde- pendent verification. The communication between the control room operator and I&C technicians did not contribute to the April 1989 RCIC system overpressure transient. There was only one surveillance being performed on the RCIC system at the time and the Team found no evidence to indicate that the surveillance procedure's action steps, other than those associated with tagging of components, were not being adhered to. The HPCI system was declared inoperable in 1983 and an LCO was , entered due to the failure of the stop valve solenoid. The RCIC l system was declared inoperable in 1989 as the result of the uncer- tainty relating to the operability of the check valve. The NRC was promptly notified of the 1983 HPCI system overpressure transient within 24 hours, with written followup as required by plant TS. The TS reporting requirement (governed at the time by a previous . 10 CFR 50.72 Rule) was "...a failure or malfunction of one or more i components which prevent or could have prevented, by itself, the ful- fillment of the functional requirements of the system (s) used to cope with accidents analyzed in the Safety Analyses Report (SAR)." The reporting requirements were subsequently changed by Amendment No. 88 to the Pilgrim Operating License dated August 14, 1985. _ _ _ _ _ _ _ - _ _ _ _ - - _ _ _ _

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ [- ) l .. ' .13 ' .i '( The' deportability of the April 1989 RCIC system overpressure: trans- 1 ient is addressed in Section 9.6 of this report,'wherein it is noted that the RCIC. system is not considered as a safety-related system credited to. cope with accidents analyzed in the SAR. The Team . concluded that, although there are similarities in the two. events, they were not the result of the same ' root cause. The licen- - see's corrective actions as a result of the 1983 HPCI system trans- ient were adequate, and subsequently followed-up and documented in . NRC inspection reports as indicated above. , _ _ _ _ _ _ _ _ . _ _ _ _ . _ _ . _ _ _ _ _ . _ _ .l

. _ _ _ _ _ _ _ _ - _ - - . n' e L- 14 r E - 4.0 OPERATIONS AND HUMAN FACTORS 4.1 Background During the performance of. Procedure Number 8M.2-2.10.11.1, RCIC High Water Level Turbine Trip / Auto Restart Logic Test, both injection line motor-operator valves were inadvertently simultaneously opened . and allowed to communicate with a small (22-foot) section of low pressure ~ RCIC . system piping. This resulted in a lifting of the suction line pressure relief valve, which relieved into the RCIC quadrant (or space). Control room personnel were alerted to this event by a RCIC quadrant smoke- alarm energizing as a result of the steam and water- vapor in the room. The following sections of the report will detail the operations and human factors that contributed to this event. The. Team also maintained contact with and attended meetings of the licen- see's investigatory review group formed af ter the event - to recon- struct human factors and performance evaluation. 4.2 Tagging Errors The Logic System Functional Test (LSFT) for the RCIC Trip / Auto Restart Logic required that eight (8) RCIC valves be positioned closed, by the reactor operator (RO), at the start of the test. Seven (7) of those valves were also required to have their respective DC breakers tagged in the following positions: Valve' Breaker Breaker position M0-1301-60 D784 Closed MO-1301-61 0751 Closed M0-1301-62 D794 Closed M0-1301-49 D744 Open MO-1301-48 D771 Open M0-1301-26 0764 Open M0-1301-22 D754 Open A review of this tagging operation revealed six (6) of the seven (7) tags were improperly applied at motor control center (MCC) D7. These tagging errors resulted in incorrect breaker alignments. The first tagging error occurred when the LSFT procedure incorrectly listed the breaker for injection valve MD-1301-49 as D744, and not the correct breaker which was D774. The D744 breaker powers the RCIC steam line isolation valve, MO-1301-17, a normally-open containment isolation valve. This procedural error resulted in discharge injection valve MD-1301-49 not being deenergized at the start of the test. The error _ _ _ _ _ _ - _ __ __ _

_ _ _ _ _ . _ - _ _ _ - _ - - , . 15 also caused steam supply isolation valve M0-1301-17 to be disabled j" open. The second tagging error occurred when the breaker for dis- charge valve M0-1301-48 was left closed and energized, contrary to the desired position on the tagout sheet. These first two err. ors . left. the two series motor-operated discharge valves (48 and 49)- energized. The third and fourth tagging errors occurred when the breakers for condensate' storage tank (CST) suction valve MO-1301-22 and torusu suction valve M0-1301-26 were left in the' closed / energized position, . contrary to the tagout sheet's desired position of open/deenergized. The ~ fifth and sixth tagging errors occurred when the breakers for minimum flow valve MD-1301-60 and cooling water valve M0-1301-62 were opened /deenergized. Contrary to the tagout ~ sheet's desired position- of closed / energized and tagged in that position. Finally, .the breaker for steam line . isolation . valve MO-1301-17 was erroneously- left open/deenergized and tagged in that position; this valve was not intended to and should not have been affected by this test. An inspection of.the tagout sheet and the red blocking tags used in this tagging operation indicated that only the breaker numbers were written on the blocking tags. The actual component name did ' not appear on the tag,. although the description of the component (valve number) did appear on the tagout sheet. The tagout sheet did not call for the valve control switches in the control room for the aforementioned valves to be red-tagged as apparently required by the tagging Procedure No. 1.4.5. Step 2.3.13 of that procedures states: "... controls for the equipment being isolated should bear a red tag to indicate that'the equipment has been taken out of service." A visual inspection of the MCC-D7 revealed no obvious human factors problems with identifying the breaker number or valve number and description. All breakers were clearly labeled, and the Team subse- quently had no problems identifying breaker position (either open or. closed), via the breaker positioning handle. However, actual com- ponent descriptions as labeled on the breaker cubicle did not, in all cases, agree fully with titles used in the LSFT procedures for both RCIC and HPCI, as discussed in Section 5.3. As a direct result of the tagging errors, the test status of six (6) RCIC system valves was incorrect prior to the start of the LSFT. The M0-1301-48 and MD-1301-49 valves were left susceptible to a simul- taneous opening on an initiation test signal, which subsequently did occur when the RCIC system LSFT was begun on April 12, 1989. - - _ _ ___ _ _ _ _ _ _ _ __ _ -- _ _ _ - _ _ _ _ _ __ _--_-- ___ -_- __- -

-- _ _ - . _ _ _ - - - - .{ l

ll q 16 { i , , f, I , ~

,- 4.3 Independent Verification An inspection of the verification process, with respect to the tag- ging operation stated above, indicated a significant breakdown in y implementation. Three levels of independent verification were potentially in place and available to detect. and correct the tagging errors that occurred. The reactor operator (a licensed _ individual who was' fulfilling' a non- licensed role at the time) was ~given the tagout sheet by the Nuclear Operations Supervisor (N05) who instructed him to hang the tags per the tagout sheet. . The NOS also informed the R0 that a second verif- ication was required on this tagout and assigned a second person (non-licensed equipment operator or E0) to the tagging operation. The two operators. then proceeded to MCC-D7 to perform the .tagout. This was, contrary to the Tagging Training Module No. 0-NL-04-02-02 which states: ". . .a verifier may not accompany the individual who is perform- ing the isolation or restoration." The tags were then divided between the two ' operators ' and applied. At that point, the operators swapped roles and verified each others' work. This was the first level of verification that was' ineffective, and fL essence not done at all. The Instrument and Control (I&C) technician who initially requested and filled out the tagout sheet also accompanied the operators when they proceeded to MCC-07, and was present at the time the blocking tags were applied. The three then proceeded back to the control room at which time the I&C technician signed Item No. 4 of the tagout sheet which reads: " Isolated Reviewed / Inspected and Accepted By." The Tagging Procedure 1.4.5, Step 6.4.10 states: "The supervisor who is in charge of the work for which the iso- lation is made, shall review the physical isolation and tagging in the field prior to beginning work. The supervisor shall sign the Isolation Reviewed / Inspected block on the tagout sheet." This physical inspection did not take place as required, in spite of the fact that the I&C technician was present at the D7 MCC at the time the tags were applied. This was the second level of verifica- tion that was ineffective and in fact not done. 1

__ _ _ _ . - - - - _ - _ - _ _ - _ _ _ - _ _ _ - _ -17 4 & L The final level of verification should have ' been for the control board indication for RCIC in the control room. ' These indications were not observed by the reactor operator, (RO) (or' the I&C tech- nician)' at the panel prior to test initiation. The. results' of the tagging errors would have provided the R0 with the following indica- tions of a problem: (1) M0-1301-48, 49, 22 and . 26 all- had then " closed" position indicator lights energized when all of the valve indication for' those valve should_have been extinguished; (2) M0-1301-17.(steam 'line isolation valve) indicating lights were . extinguished when the "open" position indicator light should have been energized; (3) MO-1301-17 valve indication ~ is mimicked on the vertical section 'of 903 panel and had no position lights energized; and, (4) M0-1301-60 and .62 indicating lights. were. extinguished when the " closed" position indicator lights should have' been energized. The LSFT procedure requires the RO to perform certain steps and con- 3- trol switch manipulation at the RCIC system panel just prior to the: ~ initiation of the relays to be tested. This should have provided him with the opportunity to verify system status before the event, as well as the I&C technician 'who was in communication with and direc- ting the R0 and recording test data'in the control room. ~4.4 Conduct'of' Operations , The I&C technician received permission to start the test from the ~ Nuclear Watch Engineer (NWE) early in the shift (between 7:30 and 8:00 a.m.). The technician then received permission to tag the valves in the LFST procedure from the NOS. At no point prior to the commencement of the test was a pre-evaluation briefing conducted by the operating shift. This is contrary to the Conduct of Operations Procedure 1.3.34, Step 6.10 which states: "Certain complex or infrequently performed activities warrant a pre-evaluation briefing. Conduct a pre-evaluation brief for j the following event: ...(B) an event which may result in chal- " lenges to safety systems if improperly conducted (such as Logic System Tests)." The watchstanding R0 and the I&C technician were performing this LSFT j test for the first time. The two operators involved in the tagging j operation stated they were unaware of the reason for the tagout, and i were only following the instructions on the tagout sheet (i.e., they f did not know the tagouts were to support a RCIC system LSFT).

3 4 l . < - m

- _ _ _ _ ___ _ _ _ -. _ . , 18~ The' inspectors reviewed the LSFT procedure and found no explicit pre- caution to warn against the possibility of simultaneous opening of the injection valves and the overpressurization that may occur. This precaution should have, in the Team's estimation, been implemented.in light of a similar event during.a HPCI system LSFT conducted in 1983 (LER 83-48) as described in Section 3.6 of this report. The -inspectors concluded that the shift's response after the event occurred was proper, timely and led to a' rapid mitigation of the casually. Their interpretation and actions taken as a result of what appeared to be an unrelated smoke alarm led to a rapid isolation of the RCIC system and termination of the overpressure transient within ih to 2 minutes. 4.5 Training Tagging procedures training is incorporated in the Non-Licensed Nuclear Plant Operator Initial Qualification Program (NLNPO) . The 0-NL-04-02-02 training module on tagging was reviewed and found to be adequate, with explicit instructions for' independent verification. Satisfactory completion of the NLNP0 qualification training program is a prerequisite .for enrollment into a reactor operator license training program. Training on tM tagging process is ~ also incor- porated into the requalification training program for both licensed and non-licensed operators. On-watch-training was conducted in August 1988 and presented to all operators. The purpose of the training was to inform the staff of changes to the tagout procedures and associated auxiliary procedures in order to emphasize the changes as they impact operations. The names of the two operators involved in the tagging event appeared on an attendance sheet dated August 17, 1988. Therefore, those two operators received adequate training on tagging and verification. 4.6 NWE and STA Roles During the review of the sequence of events it became evident to the Team that, early in the event, the NWE had left the control rcom to proceed to the RCIC quadrant. It is difficult to pinpoint the exact time at which he left due to the lack of a process computer, which would provide specific valve manipulation and relay actuation times. An important note is that the NWE had left the control room before he could make a determination of whether or not all entry conditions for - _- -___-- __ _-__---__----__-_ _ _ ---

- _ _ _ _ _ _ _ _ _ _ _ - . - _ _ _ = _ - _ - - . --_ -_ _ . _ - _ - ___ _. __ ___ 4 19 Secondary Containment Control, emergency operating procedure- (E0P) number 4 were satisfied. This was evident since the entry condition of 1-inch or more water level in'a quadrant can only be determined by visual inspection of the room. The NWE did, however state that he

remained in 'the control room long enough to verify the RCIC system

isolation before he proceeded to the RCIC system quadrant. The. Team concluded that,.while aware of plant status and stability, entry con- ditions for E0P-4 could and should have been fully verified before he- left-the control-room. The Shift Technical Advisor (STA) was not in the control room at the time of 'the event, but was contacted and returned after the RCIC system was isolated. This was an appropriate situation, although the STA's role in advising the NWE regarding E0P's could not be fulfilled. 4. 7. E0P-4 Entry Condition The entry conditions for the emergency operating procedure (EOP) Secondary Containment Control, E0P-4 with respect to the RCIC . quad- rant are 1-inch water level and/or 105 F ' quadrant air temperature. - A review of the control room indications of these parameters revealed two alarm windows which read "S.W. Quadrant Floor Leakage Alarm" and " Steam ' Leakage Hi Area Temperature." The leakage detection in the room is a ball-float recessed in the floor to indicate water in- leakage. The alarm response card (ARC) instructs the R0 to dispatch someone to the room .for a visual inspection of the water level. The area temperature alarm setpoint is set at 175 F, which is 70 F above the entry condition of 105 F. This fact was presented by the Team to t e Operations Manager who stated that the licensee's engineering staff would review this issue. During the event, the NWE observed the RCIC system room cooler cycling on and off. The operating setpoint for this cooler' is at between 90 F and 95 F. This would indicate the area temperature dur- ing the event remained less than 95 F. With respect to water level in the quadrant, it was difficult to determine the level due to the presence of steam and water vapor in the room. The licensee's post- - event review indicated that approximately 100 gallons of water were released in the quadrant. Water marks (on walls, support baseplates, and other equipment pedestals) in the room indicated that probably less than 1-inch of water overall was present as a result of the relief valve actuation. The entry into E0P-4 becomes a moot point in light of the brief time period for the event and considering the NWE _ _ - - - _ - - _ _______ _ -_ _ _ - - _ -

20 immediately and appropriately instructed the control room operator to open radwaste valves and drain the quadrant. The. inspector reviewed a statement, from a conditional licensed RO in the control room at the time of the event, who stated that he referred to the E0P-4 flow chart and discussed this with the RO at the panel when the event occurred. 4.8 Conclusions The tagging process is a fundamental basis for the safe conduct of operation. This event demonstrated significant weaknesses, at several working levels in the operations staff r ot only in tagging but also in the verification and backup to that verification. Oper- ators' performance prior to this event and in conduct of the testing indicated that they were not aware of their responsibility with regard to safe tagging practice and its need to validate actions. If any one of the levels of verification had been properly accomplished, this event probably would not have occurred. The licensee's policy and longstanding philosophy with respect to the NWE's leaving the main control room was, in the case of this event, effective in that it provided for assessment of the status of the RCIC system and restoration of the system and plant to a safe condi- tion. However, the policy needs to be reconsidered in light of a more serious esent (e.g., check valve stuck open or an isolable leak from the low pressure portions of the system). Also, methods of monitoring temperatures in the RCIC quadrant and other spaces in the Reactor Building need to be reconsidered and correlated with alarm response and emergency operating procedures. Finally, consideration needs to be given to tagging control switches at main control room panels as well as at breakers during maintenance or testing. l l

_ - _ _ _ _ _ - 21 5.0 RCIC SYSTEM PERFORMANCE AND TEST HISTORY 5.1 Background The Team reviewed the conduct of inservice testing (IST), local leak rate testing (LLRT), and operability flow rate testing as applied to the RCIC system and its components. This inspection assessed the licensee's ability to perform technically sound component and system te stirig . In addition to RCIC, analogous HPCI system testing was reviewed by the Team in light of a similar overpressurization of that system in 1983. The RCIC pump injection check valve (1301-50) design and maintenance history were reviewed by the Team to determine its contribution to. the April 12, 1989 overpressure transient. The effect of previous applications of Furmanite to the RCIC check valve is addressed in Section 5.7.4, as is a modification to that valve performed in 1984 (Section 5.7.2). Finally, the Team closely followed the progress of the licensee's investigation of the event, particularly one of the three working groups setup to specifically address the cause, status and restora- tion of RCIC system discharge check valve. The Team attended the two daily meetings of the licensee's investigatory group, and maintained regular contact with group members and the group leader. 5.2 Adequacy of Testing The Team reviewed IST procedures and discussed inservice testing with cognizant test personnel. 5.2.1 Mechanical Exercise Procedure No. 8.I.11.9, RCIC Injection Check Valve Cold Shutdown Operability, is intended to perform forward flow exercising of the RCIC injection check valve per ASME Code requirements for normally closed valves. A mechanical exerciser external to the valve is utilized to apply the needed torque to cycle the valve's disc open for a freedom of movement check. The maximum allowable torque is based upon vendor ( Anchor / Darling) limitations for preventing ' damage to the valve, as well as Code requirements. ASME Code Section XI requires that the acceptable torque be based upon the most limiting criteria within Code Section IWV-3522(b). The actual acceptance criterion (56 foot- pounds) is derived from 200*4 of the observed torque required to actually exercise the valve when new and in good operating condition. ._-__ _ _ - _ - _ _ _ - _ -

- _. . _ - _ _ - _ _ _ - _ _ _ - _ _ _ _ - __--- 22 The exercise test is performed by -attaching a calibrated torque wrench to the exercising nut and moving . the disc through a prescribed rotation. The inspectors verified that the procedure provides adequate detail to ensure 'a proper test, given the design of the exercising ' mechanism , ' inside the valve. The Team also determined that the origin of the test acceptance criterion was valid, and determined in strict accordance with the Code. 5.2.2 Leak Testing Procedure No. 8.5.5.7, Hydrodynamic Test for Measuring Leakage Thru RCIC System, guidance provides for performing hydrodynamic testing to determine leakage through the RCIC injection check valve and motor-operated discharge injec- tion valve M0-1301-49. Both valves serve as a normally closed pressure interface between low pressure RCIC pump suction piping and high pressure injection p , .ag down- stream. The valves are hydrodynamically ' tested for water leakage to minimize the probability of intersystem leakage of reactor coolant. The 1301-49 valve additionally serves as a containment isolation valve and, as such, is air tested in accordance with 10 CFR 50, Appendix J. The c hydrotest is conducted with a test rig and valve alignment' _ to collect and accurately measure leakage in a graduated container. The test is conducted for a minimum 5-minute length of time in order to verify stable conditions. The test method and acceptance criterion of 2 gpm satisfies the requirements of the applicable version of Section XI of the ASME Code. The Team similarly reviewed the licensee's methods of con- tainment isolation valve local leak rate testing (LLRT). , This was found to be performed in accordance with appli- cable Appendix J requirements and industry practice; with discrepancies noted. Leak test results are listed in Section 5.6.2. 5.3 Operability Verification The Team reviewed recent results of Procedure No. 8.5.5.1, RCIC Pump Operability Flow Rate and Valve Test at 1000 psig. This test is designed to perform operability verification of the system, and 1.s performed in three parts in order to satisfy Technical Specifications and IST requirements. The first part is a monthly pump flow rate test with data *aken for turbine speed, pump flow rate and discharge ____ - - __ - _________--__ _ ___-__ _ ___-_______ ___ _-.

_ - _ - _ - - - _ - _ _ _-_ _ _____ - _ _- -_____ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ - _ - - _ _ _ - - _ _ _ _ _ - . 23 pressure, and valve stroke times. The second part is a quarterly test with additional data .taken including pump vibration and other hydraulic data. The third part of the procedure is a once 'per cycle cold quick start of the RCIC system in the standby lineup which takes suction from and returns (via a line) to the CST. The Team noted that the last time the RCIC pump was run,. prior to the April- 12, 1989 Logic System Functional Test, was for a cold quick start on April 3, 1989. Test records indicated acceptable performance on that- date. The licensee's investigation team that concentrated on the check valve's . performance indicated. that the April 3,1989 test may have caused a 1500 psi pressure transient (from normal RCIC discharge pressure) to be passed through the closed M0-1301.49 valve, which could possibly have affected the CK 1301-50 check valve downstream. The Team was presented no evidence or supporting analyses to. prove or disapprove this postulated effect. In addition to the above, Procedure No. 8.A.17, RCIC System Integrity. Surveillance, is run to provide operations personnel with a method.of verifying the integrity of the RCIC system outside of containment. This test is normally run in conjunction with the pump operability flow rate and valve test. While this surveillance is being per- formed, piping and equipment is inspected for leakage in the pump room, the torus room, the traversing incore probe (TIP) room, and the steam tunnel. The acceptance criterion for this test is simply. that no leakage is found, and a maintenance request is submitted for any observed leakage. This test was last successfully performed on April 3,.1989. 5.4 Logic System Functional Testing The Team reviewed test procedure No. 8.M.2-2.10.11.1, RCIC High Water Level Turbine Trip / Auto Restart Logic Test, which is a logic system functional test (LSFT). It is the events surrounding the conduct of this test on April 13, 1989, which are the subject of this AIT. Section 4.0 of this report contains an explanation of the tagging errors which were the proximate cause preceding the overpressurizing of the RCIC system suction piping. An evaluation of the purpose for this test is given below. The test initiates the RCIC system start and trip logic, but does not result in actual run of the pump / turbine or developed flow. The test is run to demonstrate control logic which, under design conditions, trips the turbine on high reactor water level. ,

-_ _ _ - _ . p 24 The turbine trip signal on reactor high water level closes the fol- lowing valves which cause the turbine to stop operation: MO-1301-61 Steam Supply to Turbine M0-1301-60 Minimum Flow Line MD-1301-62 Cooling Water Supply The logic also permits the turbine to restart automatically-without manually resetting the trip throttle valve when a reactor vessel low water level signal reoccurs. The LSFT performed prior to the April 13, 1989 test was successfully conducted on October 5,1988, with no discrepancies noted. The Team questioned, given an erroneous breaker identification which remained uncorrected in the procedure for motor-operated valve 1301-49 (D744 instead of D774 for MO 1301-49), as to how the test was successfully run on at least two occasions in the past. The licensee indicated- that test personnel were aware of the procedural error, but that they tagged the correct breaker although no record of such as a test deficiency was found nor was a correction to the procedure made. This failure to. adhere to test procedures or pursue changes where necessa ry, represents a significant weakness in the testing process. 5.5 Analogous HPCI System Testing The inspectors reviewed Proccdure No. 8.M.2-2.10,4-2A-1, HPCI Initia- tion Logic /High Water Level Trip Reset Test, which is analogous in purpose to the RCIC system LSFT. This test was last performed successfully on October 9,1988. An error had existed in this test which, although found and documented on October 9, 1988,- remained uncorrected as of the time of the AIT. The procedure calls for closure of the D832 breaker to M0-2301-35 which is the pump #2 suc- tion / suppression chamber valve. This breaker, however, does not exist. The correct breaker for MO-2301-35 is D824. This error was documented as a test discrepancy, and it was noted that a procedural change notice (PCN) was written to correct it. The licensee indi- cated that test personnel were nonetheless aware of the error and knew which breaker was the correct one to tag for the MO-2301-35 valve. The Team walked down the remaining valves / breakers for the HPCI system LSFT and satisfactorily verified proper correlation between devices. However, the referenced PCN could not be found during the AIT, and the official record copy of the HPCI LSFT still remained the previous revision with the same error uncorrected. The licensee took prompt action, once identified by the Team, to correct the HPCI system test procedure. The fact that the HPCI system LSFT was conducted in the past, with this error, is another example of the procedural weakness noted above in the testing process. -____ __- -- - - _ _ _ . - _ _ _ _ - - _ - _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ - - _ _ _ _ - _ _ _ _ _ _ - - -

_ _ _ _ _ _ _ ! [ ! l- l 25 5.6 Isolation Valve Performance l 1 5.6.1 Leakage Test History The Team reviewed leakage test results for the RCIC system injection check valve CK-1301-50, and for containment isolation valve (CIV) MO 1301-49 and the. inboard feedwater check valve 6-58A which is a containment isolation valve (CIV). The following table shows leakage results for the valves over the past 3 to 5 years. The Team noted that the check valve has been essentially leak tight since its installation in August 1984. It was also noted by the Team that the feedwater check valve 6-58A has a history of high leakage but that this is an industry-wide problem with feedwater check valves. - _ _ _ _ _ _ _ _ - _ _ _ _ - _ _ _ - _ _ _ _ _ _ - -

_ _ _ _ 26 Leakage Test Results Check Valve Valve No. Test Date Test Medium Results Comment 1301-50 9/5/84 Air 22 SLM Acceptance test after installation 1301-50 11/23/84 Air 15 SLM After adjustment 1301-50 11/24/84 Air 0 SLM After adjustment 1301-50 8/19/87 Water 0 GPM 1301-50 6/15/88 Water 0 GPM Motor-Operated Discharge Valve 1301-49 7/2/86 Air 0.71 SLM 1301-49 8/18/87 Air 0.1 SLM 1301-49 8/19/87 Water 0 GPM 1301-49 6/10/88 Air 0.1 SLM 1301-49 6/15/88 Water 0 GPM Feedwater CIV 6-58A 6/26/86 Air Failed Could not hold test pressure 6-58A 6/30/86 Air 178 SLM After maintenance 6-58A 12/2/86 Air 0.1 SLM After maintenance 6-58A 5/19/87 Air Failed Could not hold test pressure 6-58A 6/1/87 Air 0.1 SLM After maintenance 6-58A 6/9/88 Air 0.1 SLM NOTES: GPM - Gallons (Water) per Minute SLM - Standard Liters (Air) per Minute ! 1 - - - - - - _ -- - - - - - - - - - - - - - - _ _ _ _ _

. _ _ _ - _ _ _ _ _ _ _ _ - _ _ _ L l 1 27 .. 5.6.2 Containment Isolation Issues The Team noted that an open NRC inspection item from'1983 regarding the containment isolation function'of RCIC system check valve CL-1301-50 had existed. The valve,' along. with numerous others, was identified as part of unresolved item (UNR) 50-293/83-03-05 associated with inconsistencies regarding its isolation function. These inconsistencies appeared in various tables in Technical Specifications, the FSAR, and leak test procedures. The unresolved item was updated in a 1987 inspection report (50-293/87-38) and subsequently closed in a 1988 inspection report (50-293/ 88-19). The Team noted that an inconsistency remains in the case of the RCIC discharge check valve CK 1301-50, in that it is still listed in the FSAR as a containment / reactor isolation valve but is not tested (e.g., Appendix J LLRT) as such by procedures. The Team discussed this with licensee repre- sentatives ' who stated that the valve serves a pressure isolation function, but is not considered to be a CIV. A leakage limit of 2 gpm is imposed via the IST program and is intended (as stated in BECo's response to NRC Generic Letter 87-06) to prevent overpressurization of the low pressure. suction piping of the RCIC system, and thus mini- mizes the likelihood of an intersystem LOCA. The valve was designated a pressure isolation valve in 1987; previous to that time, it did not serve as a CIV. The inclusion or removal of CK 1301-50 will be considered by the licensee during future FSAR revisions. The Team reviewed IST stroke time testing results for the MO-1301-48 and MD-1301-49 pump discharge valves from December 1985 to March 1989. Stroke testing of power oper- ated valves is required to be done quarterly. The limiting value of full-stroke time of each power operated valve is to be specified by the Owner per ASME Section XI. BECo has specified the limiting opening and closure time for both the 1301-48 and 1301-49 valves to be 20 seconds. In all cases since December 1985, stroke times for both valves ranged from 12-15 seconds and generally consistent in per- formance. Deviations in opening and closure time for each valve has resulted in MWR's on several occasions. Most - _ _ _ _ _ - _

-- - _- - - -- . ' n 28 notable was a February 3,1987 stroke test of' the 1301-491 valve. This test indicated a deviation of approximately 4v seconds. The valve .then received maintenance and the 'fol-r lowing official test on March 16, 1988, showed that' opening > and closure times were' .. identical at 14.5 ' seconds. .The valve had not undergone ' testing on the usual quarterly schedule because of the extended plant outage. 0ther than - deviations in opening'and closure times from test to test the valves have performed satisfactorily. 5.6.3 Retest'of Check Valve 1301-50 After - the overpressure transient of the RCIC system' on April 12, 1989, the licensee devised a . special test in- tended- to aid in the evaluation of the status of CK-1301- 50. The test was qualitative in nature, and designed to measure the pressure increase upstream of the check valve with downstream feedwater system pressure greater than or equal to 1000 psig. This pressure increase was expected to be realized if the check valve were to leak by under the test conditions on April 15 and 16,1989, (the reactor was at 25% power in hot standby condition), or if the valve (as suspected) was. not fully seated upon initiation of the test. The RCIC system was lined up such that a pressure increase could be measured at a test connection upstream of the check valve. With the feedwater and Reactor. Water Cleanup- (RWCU) systems at rated pressure and hot temperatures, the MO 1301-49 valve was cycled open to allow the unimpeded backpressure transient to reach the instrument located aownstream of the closed MO-1301-48 valve. The M0-1301-49 valve was ' then closed and the process repeated several times. Pressure and temperature data were taken continu- ously during the test. The test results were as follows: a rapid pressure increase (less than one second) was meas- ured in +.he upstream side of CK-1301-50. This would indi- cate leakage through the check valve. However, a subse- quent cold hydrodynamic leak test (the normal test de- scribed in Section 5.2.2) was performed on Ap:-il 18, 1989 with the reactor shutdown, cold and depressurized, and > which showed essentially no leakage past the check valve. It is therefore possible that the cold hydrodynamic leak test performed under the requirements of the IST program may not provide the same indication of the leak tight in- tegrity of the check valve under hot operating conditions. ! - _ - _ _ _ . - _ _ _ - - _ _ - - - - - -_.

- _ - _ _ _ _ _ _ _ _ _ .___. h 29 The licensee was investigating this apparent anomaly be- tween hot and cold leak tightness as of the end of the inspection. The Team questioned the apparent indeterminant state of check valve CK-1301-50, which showed little or no leakege under cold pressurized conditions with no solid leg of water between CK-1301-50 and MO 1301-49 and a leak- off path provided during the test). Yet, under hot opera- ting conditions, such as during April 15 and 16, 1989 test- ing, the solid water volume between the 50 and 48 valves was rapidly pressurized (when 49 was opened). As of the week following the AIT exit meeting on April 19, 1989, the licensee's Nuclear Engineering Department (NED) represen- tatives had preliminary concluded that this was, indeed, the normal expected behavior of an operable RCIC system injection check valve CK-1301-50. This was explained by the licensee as follows: Even for a leak-tight CK-1301-50 valve, hot RWCU sys- -- tom fluid would eventually heat and expand the con- tained solid water volume between valves 50 and 49. Note that most of the piping from check valve 50 (in the steam tunnel) back to the normally closed MO- 1301-49 valve is an insulated (lagged) 60-foot hor- izontal run of piping which was found after the April 12, 1989 event to have a temperature on the order of 300 F. The expanded volume would " crack open" the check valve -- and allow pressure equalization across the disc which would also rapidly (instantaneously) pressurize the contained fluid space between the 50 and 49 valves. i Only opening the MO 1301-49 valve, and allowance for -- dynamic flow and hence depressurization would perman- ently and positively seat check valve CK-1301-50. 5.7 Check Valve 1301-50 Design 5.7.1 Design Description The RCIC system injection check valve CK-1301-50 is a 4-inch carbon steel exercisable swing check valve manufac- tured by the Anchor / Darling Valve Company. The present valve was installed in 1984 when its predecessor, an air 1 operated swing check valve, was removed and replaced (see ] Section 5.7.2). The valve is installed horizontally, 4 _______-.___-----__-_--___---_-_-_-_____-_-__--_-_-_.-_-_-_--_--_---___----------_---.-_-_--_._J

. _ - _ _ _ _ _ _ _ _ _ _ 30 - c approximately 120 11neal feet of pipe (and after two.MOV's). ~ downstream of the RCIC pump discharge. Its disc is mounted. ' .c. 5 degrees off of: true; vertical by design, and is mechan- ically exercised during the cold shutdown condition'. The! valve employs metal to metal seals at the ; disc / seat inter- - face, and eight packing rings and one lantern ring along the actuator shaft. The valve.' is. designed for seismic l Cl as s . 1, per the 1974. ASME Code' Section III, with an approximate weight of 160 pounds. l ., The valve is regularly exercised at cold shutdown by attaching a calibrated torque wrench to a square wrench nut on ' the. actuator shaft. Turning the wrench nut counter- clockwise forces the disc into the seat and turning the nut clockwise 90 degrees allows the disc to swing free. Con- tinuing to turn the wrench nut clockwise then opens the disc. The maximum allowable wrench nut torque per vendor recommendation is 75 ft-lbs. The licensee has specified a maximum allowable torque of 56 ft-lbs. for inservice. test-- ing purposes, with a maximum allowable leak rate of.2 gpm: ' of. water. The valve's design is unique to the Pilgrim Station, and the only other valve of similar kind is cur- rently installed and operational .in the HPCI system's pump discharge line as 8-inch check valve 2301-7. The HPCI sys- tem's injection check valve, which unlike RCIC CK-1301-50 has double resilient (soft) seats, has also'shown good leak tight integrity when tested routinely under cold shutdown conditions. 5.7.2 Modification to Remove Air Operator Check valve CK-1301-50 was replaced in kind by modification 83-29 in August 1984. An older (prior to 1984) different kind of air-operated check valve (A01301-50) had experi- enced high seat and hinge pin leakage due to feedwater backpressure. Since the valve's replacement was not air actuated, its 1-inch bypass valve (AO-1301-71) used for ' valve disc actuation testing was no longer required and also removed. The currently installed regular swing check valve was chosen by BEco for ALARA and human factors considerations. By the use of a regular swing check valve, worker exposure during maintenance of the air operator, solenoid valve and position switches was eliminated. Worker exposure to manually test the valve was estimated at approximately 10 _ - _ _ - _ - _ - _ - _ -

_ _ _ - _- l 31 minutes each time operability of the- new regular swing l check valve is determined. This is considered tzr be muctr- l less than the exposure required to ma'ntafrrthe: abover ' accessories of the old air-operated (AO) valve. i The l . replacement of the A0 valve with the regular swing.. checks valve was considered a human factors improvement for the control room operators because -of the elimination of several indicating lights associated with the position. switches for the valve. The new valve' has also demon- strated better leak tight integrity. 5.7.3. Maintenance History .The Team investigated the maintenance history of the new (since 1984) check valve CK-1301-50. The only maintenance < history available to the Team, performed on the valve since- acceptance testing in October 1984, was a job that included repacking the valve ' and . replacement of the pressure seal gasket and actuator shaft plug gasket. This_ work was per- formed in August 1987. Also performed at that time was a replacement of a Furmanite fitting for the valve with a 31/8 inch gland plug. Mechanical exercising of the valve after .this maintenance indicated that 20 ft-lbs. were required to forward flow exercise (open) the valve. This- met the established . acceptance criteria that the ' valve operate at less than 56 ft-lbs. Leak testing of the valve at- that time with both air and water indicated zero leakage. s- Previous maintenance was discovered after the exit for the AIT which indicated that Furmanite leak sealant had been applied in March 1985 which was not adequately removed or cleaned from the valve during the August 1987 permanent repair. 5.7.4 Furmanite Application The licensee has instituted a program for Furmanite leak repairs that includes generatior, of a maintenance request to permanently repair a valve for which Furmanite has been applied as a temporary modification, including removal of all traces of the sealant. Such a maintenance action was initiated (MR 86-13-24) on August 6,1986, for check valve CK-1301-50. _ _ - _ - - _ - _ _ _ _ _ _ _ _ - - _

- __- -- - - - - - _ _ _ - - _ - _ _ - _ _ _ _ - _ _ _ - __ - -_ - _ . av ^

32

The original Furmanite was evaluated in a BEco safety eval- ' uation that referenced GE: Service = Information Letter (SIL) - No. '305 originally issued in .0ctober 1979, regarding - the: . potential introduction: of contaminants into ~ theireactor coolant system. However, while chemistry concerns were not' an issue in this case, the potential binding ; from over- application' of Furmanite was not addressed in ' the safety , evaluation. Further, subsequent maintenance in August 1987 ~ ' failed to removal all of the Furmanite which' was found lodged between the valve body and shaft bushing on ~ April 21, 1989. 'l GE SIL No. 305 recommends chemical analyses to minimize leakable chlorides and contaminant concentrations below maximums of 500 ppm chlorides, and . 700 ppm sulfides and heavy metals. The Furmanite is recommended to be applied only as a temporary fix for short periods of time (approxi- mately 90 - days), and then only primarily for locations l where excessive injection would not introduce Furmanite directly into reactor systems. Otherwise, calculations and procedural controls should prevent potential excessive injection by estimating and accounting for total injection volumes based on clearances involved. l The temporary modi i:ation approved an March 8,1985, pro- posed drilling of a 1/8-inch diameter hole into the RCIC CK 1301-50 hinge pin bracket to inject a. controlled volume of Furmanite into .the packing chamber to seal gasket leetage. The injection chambers were then-plugged with Furmanita and i a threaded adapter installed. .However, the safety e'+alua- l tion for. the temporary Furmanite repair / modification. only addressed the potential effects of the sealant (and the new l hole and plug) on the structural integrity of the valve's p; essure boundary. The packing gland was not considered pressure-retaining, and . hence a non-structural appurten- ance. Therefore, the evaluation . concluded that the Furmanits process, "...will have no adverse effect on valve operability or pressure integrity." l [ Relevant industry guidance on Furmanite (EPRI Report NO- ' 3111, Testing and Evaluation of On-Line Leak Sealing Methods, issued in May 1983), as well as the GE SIL No. 305, indicate that other important considerations not l employed by the BECo evaluation of RCIC check valve Furmanite modifications need to be addressed such as: l l f t Y _ _ - - . _ _ _ _ _ _ _ _ - _ _ _

_ - _ _ _ - -. ._ -_ . ._ -- -33 -- Consideration of Furmanite repairs for only short-term 'use (less than.one outage); i .j Consideration of the added mass of repair .fiktures -- with regard to seismic loads. on valves and piping systems; Certified analysis of . sealants used -to minimize- the -- risk of stress ' corrosion - cracking due to- corrosive ions leached from ' the' sealant compound during pro- longed exposure to leakage and high temperature; and, Knowledge of total sealant volume injected (i.e., pre- -- ' determined volume control) to prevent excessive injection and/or sealant extrusion into the ' primary system. Although Safety Evaluation 1795 was initially prepared for the first application of "three sticks" of Furmanite under maintenance request .(MR) 85-13-4 on March 9,1985,' two sub- sequent re-injections of Furmanite were applied on November 17, 1985 with the reactor at 70% power (MR 85-651,. quantity unspecified) and again on April 1,1986 with the , reactor at 50% power (MR 86-236, "two sticks") because of continued leakage at the hinge pin cover. This leakage resulted' in steam issuing from the packing. The licensee used the original evaluation for all three injections, and referred to a material release inspection report (MRIR) No. 85-135 for chemical composition of the compound (Lot #107 and 134, good through February 1989). The engineering disposition was for these temporary modifications /repairr to be rem'oved at -refueling outage ' RF0-7, begun in April 1986. 5.7.5 Failure Mode An additional engineering evaluation was documented in a March 8,1985 NED memorandum to support safety evaluation No.1795, allowing use of Procedure M-885084 to implement the repairs. However, the engineering evaluation stipu- lated that "... Maintenance shall exercise control over the volume of Furmanite injected to ensure that only the amount necessary to seal the valve is injected." That condition was apparently not strictly met, in that the sealant became more than just an extension of the existing gasket, and in fact did have an adverse effect on check valve operability. I e_ _ _ _ _ _ _ .

_-- - - __ __ _ _ _ - _ __ _ _ _ _ ._ _ _ _ . . _ _ . ( 34 As stated earlier, the MR 86-13-24 initiated in August 1986 and performed in August 1987 to permanently repair and repack the valve failed to completely remove the Furmanite which had migrated past the packing into the valve, and onto the face of the disc, actuator shaft bushing, and plug. This ultimately caused binding and impeded free movement of the check valve. The complete valve internals were removed and inspected on April 21, 1989, in the presence of an Anchor-Darling repre- sentative. Disassembly and repair under MR 89-13-34 was completed on April 24, 1989, and a succe:sful hydrodynamic pressure test showed leak tightness (held at 1055 120 psig and cold 83 F conditions for 5 minutes) at less than 0.01 gpm. The similar HPCI system check valve was also inspec- ted. No evidence of Furmanite or records of its use were found. The disc was observed to swing freely, and its stellite seat was being examined for minor indications (subsequently dispositioned "use as is") discovered on the disc seating surface. The licensee was searching mainten- ance records to identify other valves which had Furmanite applied, and had placed all Furmanite repairs on hold pend- ing reconsideration of their sealant program. 5.8 Conclusions The Team found adequately implemented IST, specifically as applied to RCIC check valve CK-1301-50. Leak tightness of check valve CK-1301-50 in the cold hydrotest condi- tion was good; however, seemingly contradictory results were found when tested under hot conditions with solid water in the intervening i space between M0-1301-49 and the check valve. This is preliminarily explained by the expected conditions whereby: (a) the trapped fluid is heated and expanded; (b) the check valve unseats; (c) pressure equalizes across the disc; and, (d) the fluid volume trapped up to the MO 1301-49 valve is quickly pressurized to steady state condi- tions of approximately 300 F and 1000 psig. The Team's concern is for the resultent opening of the normally-closed 49 valve either inadvertently or during RCIC system initiation. In the former case, an overpressurization and release of contained water (and flashing) would be expected - even though the check valve should firmly seat as the piping depressurizes and flow down is established back toward the RCIC quadrant space similar to the event that occurred on - April 12, 1989. In the latter case, potential system equipment or , piping damage could be postulated, depending upon the relative times for backpressure pulses felt at the low suction side versus the developing pump head / flow. The licensee's engineering representa- tives were aware of the Team's concerns, and were in the process of addressing these questions as of the issuance of this AIT report. - - -- -

?. 35- . The concerns - for. past performance of the RCIC system LSFT 'using a procedure with a known error, which was only surfaced as a . result of the' April 12, ~1989 event, raises _ questions: regarding adherence to h test procedures and,_ for the case of the. analogous HPCI system LSFT,- the use and. incorporation of procedure change. notices (PCNs). -Maintenance history for the RCIC system - injection check - valve CK- ~ 1301-50 ' exposed a weakness in the use and control of Furmanite as a leak sealant, wh_ich was a contributing cause to the April 12, 1989 transient and the valve's performance, l .: _--___:______-_. ---

. _ _ . . ._ - _ _. _ _ - - ________ _ _ - 36 , I l , 6.0 EFFECTS ON RCIC EQUIPMENT AND SYSTEM REC 0VERY

6.1 Background <

!

During the licensee's initial critique of the event, it wastdeter- mined that a number of specific actions would be necessary to iden- tify the effects of the transient on the RCIC system. The Team reviewed the licensee's efforts in this area, and performed calcula- tions and inspections in parallel with the licensee. The results of L the licensee's efforts and the AIT's review are documented below. 6.2 Suction Pressure Switch I At the outset of the event, two alarms were received in the control ' room which were driven from RCIC system instrumentation. The turbine thrust bearing high temperature alarm, 13-TS-1301-1 (setpoint 180 F) J was received in the control room and remained in the alarmed condi- tion. The licensee conducted Procedure No. 8.E.13, RCIC System 1 Instruments Calibration, to verify operability of the instrument and determined that the instrument was operable and had not been damaged during the event. The licensee believes that the high temperature alarm was initiated by hot water spraying on the instrument, not by bearing overheating. The bases for this conclusion are documented in Section 8.5.2 of this report. The RCIC system suction high pressure alarm,13-PS-1360-21 (setpoint 70 psig), was also received at the onset of this event. The alarm ! remained actuated after the event was terminated even though local it.dication showed that the suction pressure had returned to normal. Testing of the pressure switch showed it to be inoperative and subse- quent disassembly of the switch showed that the pressure sensing device, (a Bourdon tube), had been damaged by the pressure transient. The device manufacturer has ,arovided information that the sensing device would undergo permanent deformation at pressures above 400 psig and that the bourdon tube would burst above 500 psig. The device was deformed but did not burst. The device is a dual purpose switch which provides low suction pressure protection for the RCIC

system pump and a high pressure alarm. A replacement switch is not { readily available, so the licensee has made a design change to pro- vide these functions with two separate switches. The licensee has determined through its reviews that these were the only instruments that may have been damaged in the event. t ) _ - _ _ _ - - 1

- - _ - _ . _ _ _ _ _ _ . 37 6.3 P' ump Inspection / Repairs ' The licensee pursued .a number of different' issues' in' assessing 'the operability of the RCIC system turbine and pump. It must be assumed that the pump and turbine turned backward for some period oft time because of the back flow of hot water and steam through the pump. Reverse . rotation of the equipment had the possibility of causing the following problems: thrust bearing damage, failure of the turbine governor drive gear assembly, and pump impeller wear. Additionally, the licensee was concerned that the rapid heating of the pump inter- nals may have caused wearing of the impeller or casing. The licensee reviewed available documentation and discussed the event r with equipment vendors to identify any additional concerns with the equipment. No additional concerns were identified. The licensee replaced the lubricating oil in both the turbine and pump and had oil samples analyzed onsite for water and particulate matter. The oil samples showed no evidence of water intrusion or bearing material. This information coupled with the free rotation of the equipment led the licensee to conclude that no pump or turbine , damage occurred. The licensee- will conduct pump performance and vibration testing to confirm those conclusions during the next startup. The licensee disassembled and inspected the governor drive gear assembly 'in response to the turbine vendor's recommendation. The drive gear assembly was found to be in excellent condition. The tur- bine vendor believes, based on industry experience, that the com- ponents of greatest concern were the governor drive gears and based on the inspection results, oil analysis, and physical inspection of the equipment, the licensee concluded that no damage occurred. 6.4 Environmental Room Effects Refer to Section 8.5 for a discussion of the effects of steam and temperature on environmentally qualified equipment in the RCIC quadrant. 6.5 Suction Piping 6.5.1 Weld Examinations In ' response to the stress calculation results, the licen- see's Nuclear Engineering Department recommended that the licensee conduct non-destructive examination (NDE) of six welds in the RCIC system suction piping. These examina- tions were conducted and with one exception were negative. The licensee did identi fy one linear indication on the _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _

N: s. 9.- L 38 suction relief' valve, PSV 1301-31, inlet pipe. The NDE method used to detect the indication was dye penetrant. The indication. was approximately h inch in . length ,.and because ' of its condition, i.e., rusting, the . licensee believes-it existed prior to the event. The licensee pre- L pared a work. package to repair. the indication ~ but becausej l of -the depth, was unable. to do so. The flawed "sockolet"

was cut out and replaced,. and will be sent out for metal- ! lurgical failure analysis. < 6.5.2 ,Repai rs The Team evaluated the linear indication in the'"sockolet" connection ~ of the valve No. PSV-1301-31. Based on the oscillation of the indication, and the results of examina- tion by the licensee, which was communicated to Region I staff in a telephone conversation, .it appears that the indication is not the result of the overpressurization transient suffered by the RCIC system. The examination indicated a plane and very close joint (i e., relatively smooth joint surfaces) which is normally associated with manufacturing defects similar to' a common forging -anomaly; rather than a ductile tear, with measurable, separation which normally is an indication of plastic tearing failure. Furthermore, the indication extends from the top of the. "sockolet" towards the pipe, which is not consistent with the mechanics of crack formation due to overpressurization in the pipe to which the "sockolet" was attached L(refer to Appendix K and Appendix E, Photograph B). 6.6 Hanger Walkdown The licensee conducted a detailed walkdown of the RCIC system using the system as-built and . isometric drawings to determine if the pressure and thermal shock had' damaged any portion of the piping or piping support syt. tem. The results of this walkdown were negative. 6.7 Water Level in RCIC Quadrant Refer to Sections 3.3, 4.7 and 8.5 for discussions regarding observed water level in the space immediately following the event. _ - - _ _ _ _ _ _ _ _ - _

_ _ _ _ _ _ _ _ _ - . [ l 39 6.8 Relief Valve Bench Test The licensee has removed suction relief valve PSV 1301-31 from the hCIC system and has conducted bench testing to determine the set- point. The results of that testing showed that the valve relieved at 91 psig, which was slightly lower than its 100 psig design setting, but nonetheless satisfactory. 6.9 Conclusions The Team has concluded that the licensee's approach to determining the effect of the transient on the RCIC system equipment was well thought out and comprehensive in nature. Observation by Team members of work in the field as well as attendance during the licensee's i planning process confirmed that the planning and performance of the activities was in accordance with industry standards and regulations. The primary focus of the licensee's effort was to accurately assess the condition of the RCIC system equipment and to effect any neces- sary repairs in a thorough and timely manner. I

--- _ - _ _ _ - - _ _ _ _ _ - - - _ - .. . _ - - - _ _ _ _ _ _ _ _ _ - - _ _ - _ _ _ __ . _ _ _ 40 .; 7.0 RADIOLOGICAL IMPACT - - 7.1. Background a ~ + - The Team assessed the licensee's immediate - response and subsequent m actions of their radiological protection organization. On the day of .the event (and prior to the arrival of the AIT onsite on April 13, 1989), the Pilgrim NRC resident inspectio'n staff. provided for immediate onsite followup of the radiological impact of the RCIC overpressure transient and spill of approximate 100 gallons of fluid, some of which flashed to steam and contaminated the RCIC quadrant- space. Immediate response by licensee operators and Health Physics personnel resulted in five minor exposures. The significance, extent and control of this contamination is described below, and an evalua- tion by the Team of the licensee's response is provided. 7.2.~Immediate Surveys and Samples 7.2.1 Airborne Hazards The radiological protection organization's response to the event was prompt and appropriate. The watch radiation pro- tection (RP) technician was notified at about 8:45 a.m. of a " steam leak" in the RCIC quadrant by an equipment oper- ator. Within five minutes, the RP technician arrived at the scene and started air sampling, smears, and a radiation survey in the mezzanine level of the RCIC quadrant. A nor- mal radiation level of 1 to 2 mrem /hr was noted. The tech- nician observed an estimated 2 to 3 inches of waterton the lower level floor, as well as water vapor in the area. The RCIC quadrant was roped off immediately at the top of the stairs, and posted as a contaminated and airborne activity area pending analysis of the survey results. The RP tech- nician informed the Nuclear Watch Engineer (NWE), who pre- viously observed steam from the RCIC and "B" RHR quadrant spaces, that there was approximately 2 to 3 inches of water on the RCIC quadrant lower level floor. The NWE subse- quently ordered the Nuclear Operations Supervisor to open the RCIC quadrant and "B" RHR quadrant space floor drains to drain the water. The air sample was counted by gemma spectroscopy which in- dicated 0.08 of maximum permissible concentmtion (MPC). Smears indicated less than 1000 dpm. Tha csntinuous air monitor at the entrance to the RCIC quadrant indicated i 5 x 10 " mci /cc. A supervisor reviewed two survey results ' and moved the radiological boundary from the top of the stairs (elevation 23') to the bottom of the stairs (eleva- , l tion -17'). I

- , 1 I 41 The NWE, two operators, and the RP supervisor entered the RCIC quadrant lower -level (elevation -17') for investiga- ti on .' By this' time, the water had drained from thei RCIC . quadrant. A gross qualitative' smear on the quadrant floor was taken by the RP supervisor; the smear read 60 mrad /hr. Subsequent surveys indicated contaminationLlevels' up to 40 mrad /hr on smears and 0.0012 MPC. At about 10:00 a.m. , an ALARA engineer reported to the RP- -"pervisor that there were localized water spots and some eu:t residue material near the floor drains in the "B" RHR

urdrant. An RP technician was immediately dispatched for

e survey of the area. The technician noted that the ' water spots did not cross' previously established contamination barriers in the "B" quadrant. The survey results indicated contamination levels up to 28 mrad /hr on smears and 0.02 MPC airborne at the lower level (-17'). Contamination levels up to 20,000 dpm were found on the stairs. The top of the stairs for the "B" RHR quadrant was posted as a con- taminated area. The RP technicians then surveyed the 23 foot level of the reactor building and noted no detectable contamination. Supervisors then also inspected the control rod drive (CRD) quadrant, "A" RHR quadrant, and torus com- partment and found no evidence of floor drains backing up. 7.2.2 Decontamination and Exposures Five licensee individuals were slightly contaminated as a result of the event and the immediate response to the- event. Six additional contaminations occurred as a result of the subsequent cleanup efforts in the RCIC quadrant but were not evaluated by the AIT. Two individuals who were in the RCIC quadrant mezzanine floor received contamination of up to 2000 dpm/cm2 on their shoes. The mezzanine level had been cleared for entry after contamination surveys showed no detectable contamination. Subsequently, the licensee moved the contaminated area boundary back up to the top of the RCIC quadrant stairs and performed additional survey of the area. Two individuals who were in the "B" RHR quadrant at the -17 foot elevation received contamination up to 1000 dpm/cm2 on their shoes. Another individual who was also in that area received contamination up to 5000 dpm/cm2 on his hands and shoes. Personnel contamination was removed by soap and water. The licensee moved the contaminated area boundary back up to the top of the "B" RHR quadrant stairs and performed additional surveys of the area. i L__ _ __ _ _____ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ . _ _ _ _ _ _ _ _ _ _ ____ _ _

. _ _ _ _ _ _ _

The licensee also surveyed the 23-foot elevation of the reactor building for the possible spread of contamination; the surveys showed no detectable contamination. 7.2.3 Sampling No samples of water were taken by the licensee since the water in the RCIC quadrant floor had quickly drained to the. floor drains. Spectroscopic analysis of smears identified typical floor drain " blowback"; namely cobalt-60 and cesium-37 isotopes. 7.3 Conclusions The radiological protection organization's response to the event was prompt, efficient, and thorough. Five workers were slightly contam- inated and were promptly decontaminated. There was no offsite release ' of radioactivity, all 100 gallons were contained with the floor drain system inside of the Reactor Building, and the event was therefore judged to be of minor radiological significance. 1 _ _ _ _ _ _ _ _ _ _ _ - _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ . _______-______--_____A

-__ _ . Reactor Bu!khng Level: -17'6" Chemical RadWaste Sumps RHR & Core Spray instr. Racks l , HPCI Instrumentation Stai s RHR Loop "B" ' ~ ~ AHR Pumps Heat Exchanger l CAD Pumps . g , H&V Unit L.8 d ' "D . llI' l

I cooiers '

, , . , _ _ , Drain Valve ~ - glll ". O '

l Hatch Above .gN' I "# Pit m Escape _g l = Shaft l - . eS ay Pump HPCIPump - gl CRO ". y Dump SUCtton , . Top of the Torus structure Filters = n HPCI Gland 'J N <

SealCond - U g ! Auxilian - O-- . HPCI Turb L/O Cooler Lubs Oil Pump tu_ Drywell l 1 2 E l ..y.. .........y......... .......... ....................... .. g .. .................... .............. ' ! ' , s 2 [ I ~ . j -- Elevator , y~ .

Manhole l Stairs ll ) RadWaste t ' = Cont o< lgll [ I = RHR anc CS ' RCIC Pump Keep Fill Manifold m ) hy

i RCIC Turb ". A - Core Soray i ' 3 Pump i N llhl '$

f " RCIC 'l m Condenser = k RHR Pumps i F//// /////////I/////// ///// //W// Stairs RHR Loop "A" Stairs RCIC QUADRANT Heat Exchanger (SW corner of(Turbine steg) Rx. Bldg.) AsR and Core Spray instr Racks ! -S- ' l _ _ _ _ _ - _ _ _ _ _ - _ _ _

____-. _ _ . - - - _ - _ - - _ - - - -- -_ . _ - - _ _ . -_ J 43 ' ,{ , ; 8.0 OVERPRESSURIZATION. IMPLICATIONS- 8.1 Background l The information in this section was obtained from discussions with BECo personnel at the' site Technical Support Center and is based upon

instrument readings, P&ID or isometric drawings, the results of .{ . licensee analyses, and interviews with plant personne'1. Precise data-

were not available due to an inoperable plant computer; therefore, whenever ^ possible, an attempt was made to corroborate and quantify the infor- mation or verify its reasonableness. In many instances, reconstruc- tion and evaluation of event parameters (temperature, pressure flow, etc.) were ~ based 'on engineering judgement by the licensee and the Team. The'following topics are addressed in this section: 1 a. Thermal hydraulic conditions of the discharge line; b. Discharge capacity of the suction relief valve; c. RCIC system low pressure piping design and stress calculations; d. Environmental impact in RCIC quadrant space; e. Probabilistic Risk Assessment (PRA) assumptions for RCIC system; and, f. Event V (high-low pressure interface system LOCA) evaluation. .. 8.2 Thermal Hydraulic Conditions of the Discharge Line The elevation and the length of the piping from check valve CK-1301- 50 to thermal relief valve pSV 1301-31, including suction side pip-- ing, are shown in Appendix I, which is included for convenience in Section 8.2. The pipe from check valve CK-1301-50 to motor-operated valve M0-1301-49 is size 4-inch specification DB-13, with correspond- ing design parameters of 1,600 psi and 562 F, respectively. . The total length of this piping segment is 75 feet, of which 65 feet are located in the pipe tunnel, and are horizontal and insulated. The remaining 10 feet are located outside of the pipe tunnel, vertical, and not insulated. - _ _ - -__ - _- -

A( - . 5 c t. s - e q, e 1 h c, m 0 c c - u 3 l 1 rn o eo V &ti 0

nt 1 e98 ec p44cu i - - s (6 P11 p . 9 00mo g e3 3ut t1 1 p k' a r mo ooe 9, it ttt x n . o098 ee r544cm p- - - u p111 pl A000mo 333uV f 111 p - l mmmma - . 0 . p oooot . 5 D rrrro ." E - FFFFT 0) E 1 . hj., 1 g 3 5 , . M Mi G' O2 I - a5 0 tj 13 " ~ _ %.(%o# 9 _ n _ ' j 4%t/' i . ~ k g, . e . egra hcc 9 W o si o4 ir 2 - R .

Dt M1 . no 1 [+o k. { 0 C '# _ . e dm 3 _ o o, _ . as T + _ . I _ n - oe it _ _ ti . cs uo . , _ Sp m _ g 6 _ po 0 mC

%s% 2 u ' a Pg - r

b

n p Ci I p i _ s . Ci . . RP _ 3 t _ 5 . O F . D_ _ 0 p _ M _ _ 1 o( _ . 3 . r _ / + _ _ _ _ 8 - - 4 - _ o1 _ 0 . , o* 4- 3 , g 1 _ O g _ - - . l l

- - . - - _ _ 44 Due to normal leakage past CK-1301-50'or conduction through the disc over an extended period of time, pressurization (on either side of CK-1301-50) is expected. Even. with virtually no leakage, past the check valve heatup and expansion of the trapped fluid between.CK-50 and M0-49 will ultimately crack open and equalize pressure across the disc- of the check valve. Thus,' pressure in this isolated pipe seg- ment prior to the event should have been approximately 1,000 psig. The temperature was measured after the event with contact pyrometers.. and found to. range from .approximately 380 F close to CK-1301-50 to 80 F at the end of the insulated portion. The temperature gradient was determined by the licensee to be roughly 5. degrees per linear L foot. As discussed in Section 8.8, Conclusions, and Section 1.0, L Executive Summary, a potentially generic concern was expressed by the Team associated -with the location of the injection check valve and the inevitable heatup and pressurization of the fluid in the dis- charge piping. The piping between MD-1301-49 to M0-1301-48 is size 4-inch'specifica- l tion DB-13 designed for 1415 psig and 170 F, and is normally water solid at about 30 psig and ambient temperature. The total length of i this segment is 58 -feet with a net vertical drop of 23 feet. The discharge side of the RCIC system pump (to M0-1301-48) is specifica- tion EB-13 pipe designed to 1415 psi at 170 F, sums a total length of- 39. feet, and it is normally filled with water at about 30 psig and ambient temperature. 8.3 Discharge Capacity of the Suction Thermal Relief Valve The suction side of the RCIC system pump piping is si7.e 6-inch (line specification HD-13) designed to 80 psig at 170 F, and equipped with a thermal relief valve PSV-1301-31, orificed at 0.11 inch 2 and a dis- charge capacity of 10 gpm at'100 psig and 30 gpm at 900 psig. The relief valve is piped directly to the floor drain system, and sized (flow-wise) to accommodate the drain capacity. Its pressure setpoint is set to protect the pipe from exceeding design pressure. Pressure gauge 1360-20 (Appendix E, Photograph C) is mounted on the suction piping. Manufacturer's information indicates that the sensor would deform at 130 psig, but has a rupture design strength ( f 400 psig. The fact that this gauge did in fact deform, but did not break, demonstrates that the pressure at the suction side exceeded 130 psig but most likely not 400 psig. 8.4 RCIC System Low Pressure Piping The relief valve has a choke flow capacity of approximately 50 gpm (at backpressure on the order of 500-1000 psig), and is designed for a slow thermal expansion of trapped suction fluid which heats up from 40 to 140 F. 4 _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ ________m____.._______..____________.___-- w

, _ _ _ . _ _ _ _ _ _ - - - - h.'.- A 5 7 45 , 8. 4.1 - . Design Limits L The _' low . pressure RCIC_ system suction piping is 6-inch ' Schedule 40 ' A-106B carbon steel 'with a minimum allowable ' stress of 15,000 psi, a minimum required yield strength' of 35,000 psig and an expected ultimate strength . of 60,000 psi. The system _ design. conditions for this section of suction , piping (on which ' relief valve PSV-31 is mounted) are 80 > psig and 170 F. The following stress calculations .were performed by the licensee and evaluated by the ' Team in , l; light of margin to yield strength which, although no certi- fied material test reports (CMTR's) could be found, is most ' c probably in the actual range of 40-42,000 psi for typical carbon steel of ASTM A-1068 grade based on engineering experience. :The RCIC systern suction piping was originally designed to - Code B31.1 and, in lieu of CMTRs, had been accepted via Certificate of Compliance as an acceptable l method'of material certification. - 8.4.2 Pressure and Hydrodynamic Pulse Loading

The opening of motor-operated valves M0-1301-49 and MO- ! 1301-48 eliminated the. differential pressure of ~ 950 psig. The initial loading is a pressure wave which rapidly pro- pagates along the pipe and eventually is reflected from the disc of check valve CK-1301-23. The initial value of the ' pulse is the pressure difference in the pipe spaces sepa- rated by M0-1301-49 (i.e. , about 950 psig). However, the pulse height attenuates, as it propagates through different equipment, pipe bends, tees, etc.

'The fact that pressure gauge PI-1301-20 did_not rupture is an indication that the pressure at that point was less than 400 psig. The loading resulting from the pressure pulse is l of very short duration, and would tend to affect structures- l: transverse to its path such as valve discs or elbows. A variable longer-term pressurization (for the duration of the transient) of about 400'psig and at temperatures up to 300 F will also be present. This pressure will decay from its original value as the relief valve allows leakage. Licensee analyses using methods described in the Pilgrim FSAR that account for dead weight, steady pressure and hydrodynamic pulse, conservatively demonstrated that the maximum stress experienced by the low pressure RCIC system suction piping was below allowable limits. -

- _ _ _

46 8.4.3 Thermal Loading The suction as well as the discharge side of RCIC turbine piping were subjected to thermal shock loads due to the temperature difference in the discharge. Such loading affects the piping as well as the supports. Licensee analyses assumed a conservative temperature of 300 F in combination with the dead weight and the static pressure loads; the results showed that the pressures due to this type of loading is a small fraction of the yield stress. It should also be noted that this type of loading is more important for the thick walled discharge piping rather than the thin walled suction side. However, the discharge side ir designed for such temperature differences. Therefore, thermal stress loads were evaluated to be well within the design limits. 8.4-4 Hangers and Supports The pressure pulse, thermal shock and heating of the piping to temperatures beyond design resulted in potentially excessive loading applied to pipe supports. However, licensee analyses demonstrated that these loads were well within pipe support design limits. 8.4.5 Pipe Inspection and Testing In addition to the results of the analyses, the licensee (and Team members) performed visual inspection of the pip- ing and pipe supports. No observable deformation or dis- coloration was found. In addition, the licensee performed a liquid penetrant and found a half-inch crack in the base weld of the suction side relief valve support PSV-1301-31, depicted in the Appendix K sketch. This flaw could not have been caused by any of the loadings due to this event, thus, it was concluded that it must have pre-existed. The licensee cut the entire relief valve support out for destructive examination. A new support assembly will be installed. Finally, the licensee performed a magnetic particle examination of the suction piping .vhich did not indicate any flaws. From the above discussion, the Team concluded that the suction side of the RCIC system pump did not sustain damage and thus, could be safely returned to service.

- , . , i, ' 47~ _. '8.5 Environmental Impact' 8.5.1 Extent of Leakage and Condens'ation The -Team examined all available physical evidence in order . to form a conclusion regarding the extent of. the -leakage and, consequently, the' thermal and pressure shock to which the low pressure RCIC system piping-was subjected. The -licensee estimated the extent of the release to be approximately.100 gallons. There is no precise evidence for either the duration of the release nor the pressure ' variation during the release. The licensee's time estimate for. the event was about 2 minutes, 'and pressure was esti- mated to peak at between 400 psig to 130 psig as evidenced by the damage of the suction side pressure gauge. Under those conditions, an upper value of the release would be 80. gallons. .However, markings on the RCIC quadrant walls- indicate'that about 1 to 2 inches of water collected on the floor. This indicates over 100 gallons of leakage if the - total leakage were to have' covered the entire floor (which, because of the location and type of leakage, and the. floor drains, probably did not). The operator who arrived first on the scene, observed steam rising in the RCIC quadrant stairwell. This suggests that some of the expelled water . flashed to steam. .The piping water temperature distribution measured three days after the event suggests that the entire 125 gallon contents , (from the check valve, upstream and back to the RCIC pump and suction piping) had to have been expelled for the 300 F water to' reach the relief valve and flash to steam. This conclusion is also supported by the fact that the ceiling- mounted smoke alarm (about 30 feet high above the quadrant floor) was activated. The equipment arrangement in the. ~ RCIC quadrant would favor steam rise through the stairwell and condensation on cooler surrounding surfaces in the room. Thus for steam to reach the ceiling, there had to have been significant release; however, this is difficult to quantify because of the indeterminable performance of CK-1301-50 (thought to have been initially off of its seat). ! - _ _ _ _ _ _ _ --__-__-_- __. __- . _ _ _ _ _ -

_ -- _ _- _ - _ - _ - _ ._ - _ _ ._ _ _ - _ _ __ - _ - - -- -- i, f i '48- E In considering the relie'f capacity of' PSV-1301-31, < the duration of/ the release Emay have been longer than the original estimate of two minutes. Check valve CK-1301-50 .is gravity-assisted and . responds to ' flow;. _ therefore, some 400 F RWCU system water must have flowed through before the , valve was seated to check (i.e., stop) the' flow. Examina- tion of the valve after the event was inconclusive regard ' ing the valve's performance, although .it was most likely off of its seat by is to is inch prior to the . event (See Section 5). Thus, the transient may also' have been ter- . minated by the closing of M0-1301-49 and M0-1301-48 val ve s .- On the other hand, the steam release did not' raise RCIC- quadrant space' temperature above about 165 F, and this con- clusion is based on the following: (a) With the stairwell being the natural ventilation air flow means' for the RCIC quadrant, it would require several hundred gallons of superheated water to sig-- nificantly raise space temperature. There is no evi- dence that this happened; (b) The fire alarm mounted on the ceiling and set at.165 F did not activate; (c) A similar fire alarm also set at 165 F mounted in the stairwell did not activate; i- (d) The RP technician who approached the RCIC quadrant space about 10-15 minutes after the release found- space ventilation equipment cycling on and off. which occurs at ' temperatures between 95-100 F. Had. the space heated to well 'above 100 F, the area probably could not have been entered at that point. (e) The RCIC turbine thrust bearing oil temperature alarm set at 170 F was activated. The bearing oil container was doused with water and was probably heated from the outside, in addition to the frictional heating due to the turbine's backward spinning. However, the junc- tion box which contains the thermometer leads was also doused with water and/or steam. It is likely that the leads were shorted, thus, this alarm did not represent true space temperature. ! . . ._ _-. . - _ . - _ - _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ - - _ _ _ _ _ _ - . . _ - _ _ - _ _ _ - - - _ _ _ _ - - _ _ _

' a 49 The Team concluded that some small amount (on the order of one to two' gallons) of 300 F-400 F liquid from the vicinity of.CK-1301-50 must have been released;'thus, the piping was. 1 in all likelihood subjected to thermal shock at 300'F and a q pressure. pulse which peaked at certain points 'in the RCIC ' system initially as high as 1,000 psig. 8.5.2 . Temperature Alarms and Settings As noted previously, there were five instrument / alarm indi- i cations relevant to this transient, i.e., (a) the ceiling mounted fire / smoke alarm; (b) the wall mounted stairwell fire / temperature alarm set at 165 F; (c) the ceiling mounted fire / temperature alarm set at 165 F; (d) the tem - perature . alarm is the turbine ' thrust bearing oil set at 160 F; and, (e) the suction side pressure gauge set at 100 psig. As noted in the preceding discussion, the space temperature did not exceed 160 F. 8.5.3 Environmental Qualification Design Inspection of the RCIC quadrant revealed that a cable tray crosses this space at about 2-3 feet below the ceiling. The cable serve the equipment in the RCIC quadrant but also contain HPCI cabling which are environmentally qualified to 350 F. As we have noted in the preceding, the . space tem- perature did not rise .above 160 F which is well below the qualification temperature. Inspection of the equipment in the RCIC revealed no effect due to degraded environment. Review of equipment qualification sheets showed that, for example, the steam isolation valves, the EQ envelop was at '

242 F. 8.6 PRA Assumptions r In terms of core damage probability, the absence of a functional RCIC system (total unavailability) increases that probability by a factor , of 25. Conversely, the presence of a fully available RCIC system lowers the core damage probability by about 50%. These numbers were not available in absolute probability terms, and were obtained from BECo's Individual Plant Examination (IPE) report under NRC Generic Letter 88-20 which is presently in preparation. The Team concluded that a functional RCIC is important to risk and reactor safety, l although not formally credited as an Engineering Safety Feature (ESF) in design basis analyses. ! , - - _ _ - - - _ _ . _ . - ~ . _ - _ _ _ - _ . - - - _ - _ - _ _ _ _ _ _ _ _ - - - . - __A

- _ . - _ _ - _ _ _ _ _ _ _ _ _ - _ _ . - _ _ _ _ _ - _ - 4 1 , l 50 (l- . i 8.7 Event V Evaluation. Events involving leakage of high pressure (contaminated) water, raise Event . V - concerns, i.e., high-low ~ pressure interfacing ' system LOCA- with failure of mitigating capability. Although examinations of check valve CK-1301-50 were inconclusive regarding its performance during the event, the Team conservatively assumed that, for. purposes of initially assessing the transient as 'a potential Event V concern, CK-1301-50 failed - (i .e. , it leaked at .a rate higher than its minimum acceptable limit of 2 gpm). In view of- the inadvertent opening of MO-1301-48 and M0-1301-49, and the ~- pressurization lifting of relief valve PSV-1301-31, the path to ,...;i- low pressure interfacing LOCA would be through inbcard' feedwater _ isolation valve CK-6-58A. The following table summarizes those. com- ponents, and their status and response. to the . April 12, '1989 < RCIC system pressurization transient: Challenged Initially Mitigation During Failed or Capability Component Transient Unavailable Available L CK-6-58A No No. Yes l CK-1301-50 Yes Yes Yes** l MO-1301-49 Yes Yes Yes ' MO-1301-48 Yes Yes Yes , PSV-1301-31 Yes Yes* Yes ! Low Pressure Yes No Yes l HD-13 Piping

• Failed in the sense that fluid was released from the low press-

ure suction piping; although this thermal relief valve'did pro- vide de pressurization mitigation.

• • Assumed to seat after the initial moments of' the transient,

although initially degraded. Therefore, three barriers to reactor water leakage were isolable and two of them were closed by the operator, i .e. , M0-1301-49 and MO- 1301-48. Even with CK-1301-50 failed, the leakage was not reactor water. Therefore, the Team concluded that the event was not an actual Event V release. The Team evaluated the RCIC system overpressure transient in order to determine whether it was a potentially significant precursor to Event V using existing definitions in NUREG-1272 (Volume 2, No. 1), Accident Precursor Sequence Program. , '______E ___.-__..__m- - - _ . - - - . - - _ _ . . - _ - - - - _ - - . _ - - - - - - - - - - - - - - - - - - - - - - - --- - - - -

---

__ _ -_ l 51 l- The NRC Office for Analysis and Evaluation of Operational Data (AE00) accident precursor program defines a precursor event as one in which " ...any of the following occurs at power or at demand": one or more plant safety systems fail or are found failed; ~- L ' the redundancy of two or more safety systems are found degraded; -- .an initiator occurs. -- The licensee had previously committed that the Pilgrim Station com- plies with the conclusions and recommendations of NUREG/CR-5124, " Interfacing System LOCA: Boiling Water Reactors" (NRC Gener.ic Issue No. 105), specifically regarding the level of-testing of air-operated pressure isolation check valves, to reduce the risk of interfacing systems'LOCA. 8.8 Conclusions The Team evaluated the thermal-hydraulic conditions of the RCIC system piping involved in the transient, the resultant pipe loadings, the environmental impact in the RCIC quadrant and the Event V poten- , l tial of the April 12, 1989 RCIC system overpressure transient. l The Team concluded that: , The leakage was most likely larger than the capacity of the pip- -- ing from CK-1301-50 to pSV-1301-31, thus, some water from the L RWCU must have flowed by CK-1301-50 before it was properly seated; The resultant hydrodynamic pressure pulse was attenuated below -- 400 psig when it acted on the suction side of the RCIC piping, as evidenced by the condition at pressure indicator 1360-20; The pressure and thermal loadings experienced during this -- transient were superimposed to dead weight as well as static pressure, and the resultant stress was found to be well within the limits of yield stress and the operability li nit. Visual inspection of the pipes and pipe supports, revealed that the pipes did not sustain any damage and thus, can be safely returned to service; - - _ . _ - _ _ _ _ - _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ - - - - _ _ _ _ _ _ _ _ _ - _ _ - _ _ _ _ _ -

_ - - ___ _ _ , 52 RCIC space temperatures remained below 150 F; thus, there were -- no adverse environmental effects in the RCIC space; The 100 gallons of leakage which probably included some amount -- of RWCU water (but no reactor coolant) was not an Event V as defined in WASH-1400, Reactor Safety Study. Similarly, applying the NRC AEOD's definition, the event did not constitute a significant Event V precursor; and, A potentially generic concern exists for the design location of -- check valve 1301-50, and its effect upon the expected heatup and pre surization of fluid trapped between the normally closed MD-1301-49 valve and the check valve. The concern is for the hydrodynamic effects on the RCIC system during either: (a) in- advertent opening of MD-1301-49; or (b) normal initiation of the RCIC system which simultaneously opens the MD-1301-49 valve along with admitting motive steam to the RCIC pump turbine as initial flow is developed. The licensee's engineering organiza- tion is reviewing this concern for potential corrective action at Pilgrim. l l l l 1 - _ _ _ _ _ - - _ _ _ .

- _ . _ , _ _ _ _ - _ -_ - _ _ _ - _ - _ _ _ , 53'

' 9.0 ORGANIZATIONAL RESPONSE .9.1. Background The -Team evaluated the licensee's: (a) immediate on-shift response to the RCIC - system pressurization; (b) the initial management cri- tiques of the event held on the same day (April 12, 1989); (c) the formation of an Oversight Committee to guide separate Investigatory Groups focused on human factors, RCIC system equipment restoration, and check valve CK-1301-50; (d) subsequent management briefings with the AIT, including an April 19, 1989 meeting held in the NRC Region I office; and (e) the fir.a1 Oversight Committee's Report (Attachment 2). Corrective actions were evaluated, including deportability con-- siderations (Section 9.6), and the overall scope and quality of. BEco's response (long and short-term) was assessed. ' 9.2 Immediate Shift Response The Team determined that the immediate on-shift response to the event was appropriate and timely. The control room was alertad by six alarms in rapid succession at about 8:44 a.m. focusing the R0, NOS, and the NWE's attention to the RCIC system; the alarms included RCIC high suction' pressure, thrust bearing' high temperature, two floor drain leakage and two system quadrant smoke alarms. The R0, as directed by the NOS, isolated the RCIC system steam (MO-1301-16) and discharge (M0-1301-48) (M0-1301-49) valves. The NOS terminated sur- ve111ance test 8.M.2-2.10.11 and directed I&C to restore the. system to normal The NOS also dispatched an equipment operator to the RCIC quadrant space for investigation. The NWE left the control room within two minutes after the alarms were received and proceeded to the RCIC quadrant space. To summarize those Sections, at approximately 9:30 a.m. , RP tech- nicians cleared the RCIC quadrant for access and the NWE, two oper- ators, and a Health Physics supervisor entered the area and observed the following: Wet floor with rust over much of the area; -- RCIC suction lines from the CST and the torus were cool to the -- touch; RCIC discharge lines, pump casing, and suction lineup for the -- suction relief valve were hot to the touch; RCIC turbine, motor operator, pump and lines were dry and -- covered with a thin film of rust. , i i

r. ! 54 The NWE then returned -to the control' room and briefed plant management. In reviewing ,immediate on-shift response to the event, the -Team identified a potential concern regarding the Watch Engineer leavin0 the control room immediately following the' event. This is addressed in detail in Section 4.6. The Team. concluded that the on-shift operations personnel appropri-- ately responded by. isolating the source of the break and opening.the RCIC quadrant floor' drains. - 9.3 Management Critiques 9.3.1 Event ~ Critique At 11:00 a.m. on April 12, 1989, the Operations Section Manager. conducted a critique of the pressurization trans- ~ient, to establish facts and determine cause. The critique was atter;ded by the various individuals involved and the NRC resident . inspectors. The Team determined that an' appropriate level of licensee management and individuals originally- involved in the event were present at .the critique to establish critical facts and determine valid. causes. The licensee's initial critique concluded that personnel' errors and a , e. dural error (Surveillance Test Procedure 8.M.2-2.10.11) had caused the- event. The critique identi- fied a potential weakness in that operators missed at least two opportunities to identify the errors. Control room valve position discrepancies and valve nomenclature (writ- ten on the valve tagout sheets) were also not. utilized in the double verification check. -The critique identified that incorrect tagging of RCIC valve MD-1301-17, with the valve in the "open" position, caused the plant to be in - technical noncompliance with the primary containment pro- visions of Technical Specifications for a'oout one hour due to a lack of redundancy. However, in-line 1.t lation valve MD-1301-16 was available to isolate a steam iine break, if necessary. l

_ _ _ _ _ _ _ _ _ ___ -

. _ - _ _ - _ _ _- - - -- - - - - _ - _ _ _ _ _ - - _ _ _ E [ , 55 . The licensee concluded that additional information -was necessary to determine operability of the RCIC system,. including the ~ injection check valve CK-1301-50, and' to evaluate the training and the competence of the operators involved in the event. The pressurization transient' did not affect the public health and safety, but could~ have affected personnel safety 'if anyone was near the RCIC system suction line relief valve at the time of. the -pressurization. The Team reviewed the . draft event critique report and determined that the critique was held in a timely manner , and its investigation was thorough. Critiques conclusions were judged to be valid, based on the facts gathered. As

an immediate corrective action following the' critique, 1 operations management directed each on-shift NWE to discuss the event and the results of the critique with the opera- tions personnel during a preshift briefing. ' 9.3.2 Investigation Teams In accordance with the recommendations from the event critique, the li.censee's managen;ent formed an ad-hoc event investigation team under the Operations Section Manager to I further evaluate root causes, equipment operability and human factors implications leading to the event. The licensee's Investigation Team consisted of an Investigation Oversight Group, the RCIC System Operability Evaluation Group, and the Human Performance Evaluation Group. The 22 full time group members included representatives from Operatiens, Maintenance, Radiological Protection, Station Systems Engineering . Group, Training Department, and Engineering Department. The Oversight Group coordinated the licensee's activities, provided resources, and interfaced with the station manage- ment. The RCIC System Operability Evaluation Group focused on determining the effect on operability of the following RCIC subsystems: check valve ' CK-1301-50; low pressure suction piping; repair / replacement of the damaged suction pressure switch; and inspection of other pump and turbine components. The Human Performance Evaluation Group focused on restructuring . detailed facts- by: interviews related to the event, determination of root causes; and, evaluation of human factors involved in the event. _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _. . _ _

_. --. - - - _ - i 56 ! 1 Team members attended' the Investigation Team briefings _(twice a day at 11:00 a.m. and 5:00 p.m.) and also observed individual group activities. Detail . observations of the Investigation's Team activities are discussed in pertinent sections of this report.- The Investigation Oversight Com- mittee's report is also incit.ded with this. AIT report as Attachment ~2. The Team determined that the licensee's investigation was organized and well structured with broad-based discipline- and experience. The licensee's scope of investigation was clearly defined, and each Group's evaluation processes were . thorough and_ conservative. Overall, the Investigation and Team approach reflected a well-conceived and effective approach by BECo management. 9.3.3 Management Oversight and Assessment On April 13, 1989, the licensee's management held a Manage- L ment Osersight and Assessment - Team (MO&AT) meeting to- review the event and to formulate management directives to the station organization. The MO&AT is chaired by the Senior Vice President - Nuclear, and the members include Station Director, Director of Engineering, Special Projects Director, Plant Manager, Engineering Department Manager, and QA Department Manager. The MO&AT reviewed the critique of the event and approved the proposed charter for the Event Investigation Group. The MO&AT reactivated the on- shift peer-evaluator program on April 13, 1989, to monitor personnel performance. The MO&AT also directed the QA Department to increase QA' surveillance on equipment tag- ging and independent verifications and focused the licen- see's peer evaluations on specific proximate causes of the i pressurization transient. I The peer evaluators held regular debriefings with audited organizations to discuss identified strengths and weaknesses. 9.4 Peer Review Process The licensee had previously implemented a formal Peer Evaluation Program of routine personnel performance monitoring during their Power Ascension Testing Program. The individuals selected for the peer evaluator program are selected from the onsite organization, receiving training on performance monitoring techniques and are = _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .- __ j

_ _ _ _ _ - _ _ _ 57 assigned to monitor specific activities. The peer evaluator program provides around-the-clock monitoring of operational activities as well as routine audits of other areas of facility activities. Peer evaluators hold regular debriefings with audited organizations to discuss identified strengths and weaknesses. In light of the RCIC system pressurization transient on April 12,1989, the licensee's management directed the peer evalua- tors to specifically focus upon tagouts in process, particularly equipment tagging and independent verifications thereof. 9.5 Corrective Actions The licensee instituted several immediate corrective actions (within the first few days after April 12, 1989) which included: Immediately declared RCIC system inoperable and conducted

surveillance. Made Night Order instructions (same day as event) on proper tag-

ging and independent verification and excellence of Plant Operations. Trained all on-watch crews on new procedure changes.

Initiated survey on the role and responsibility of nuclear

plant operators during independent verification. Initiated training on results, to be factored into requalification training. Assigned an additional Reactor Operator (RO) to control room

during all safety-related system testing. The R0 will be stationed in front of panels for systems being tested. Removed two operators involved with event from all on-watch

duties. Replaced RO on shift with another R0, (disciplinary action pending). Stopped al' I&C testing after the AIT identified an existing

known nametag error in HPCI surveillance procedure; all I&C sur- veillances field-walked for equipment nametag accuracy prior to j performance. j i Required all isolation tagout sheets to be stamped by the !

Administrative Assistant in the control room to ensure latest i ! revision prior to issue. l l l l i 'I - -

_ _ _ _ , l l 58 l- Require all entries on tagout sheets to be doubled spaced prior-

• .

to issuance, to help preclude transposing errors. Issue written test to all operators within one week to measure

understanding and retention of training _on tagging double verification. , Require tagging and primary containment isolation system scen-

, arios be included in all simulator training sessions for an indefinite period of time. Perform random observation of tagging and surveillance testing

with both operations and I&C' personnel. These corrective actions were: (a) verified in place as appropriate, by the AIT while on site; (b) described to the AIT during .the BECo presentation in the Region I office on April 19,1989, at the con- clusion of the inspection; (c) reiterated in the licensee's April 21, 1989 letter to the NRC regarding restart commitments; and, -finally, (d) described in the Oversight Committee Report dated April 24, 1989 (Attachment 2 to the AIT Report). The Attachment 2 memorandum, in addition to many_ of .the aforemen- tioned short-term actions, described longer-term commitments (beyond- restart of _the plant) by the licensee which address evaluations of potential design modifications such as insulation removal, additional events / drains and post-work leak-stop validations, re-validation of- I&C surveillance procedures, and other formal recommendations for improvement or investigation (RFI Nos. 89-537 through 548). The Team concluded, subject to_ its own recommendations in Section 2.3, that the licensee's proposed or already implemented actions were compre- hensive and sufficiently broad and focused so as to strengthen their defense-in-depth for the conduct of testing and operations. I 9.6 Deportability Considerations Requirements for notification to the NRC for events, both emergency and non-emergency, which occur at operating nuclear power plants are provided in the Code of Federal Regulations, Parts 50.72 and 50.73, Part 50.72 req' ires notification to the NRC Operations respectively. u Center, via the Emergency Notification System (ENS), within one or four hours when certain classification criteria in 10 CFR 50.72 are met. _ _ _ _ _ _ _ _ _ - _ _ - - _ _ _ - _ - _ _ _ _ - _ _ _ _ _ _ _ _ _ - - _ _ _ _ - - - - _ _ . _ _ _ _ _ _ - _ _ _ _ _ - -

_ _ - - - _ _ . _ _ - 4 - 59 The licensee determined that the RCIC overpressure transient did not fall within any of the four emergency classes specified in its approved Emergency Plan. The lowest classification is an Unusual Event which indicates a degradation in the level of safety of tb_ plant. This classification includes conditions which require plant shutdown in accordance with .the plant Technical Specifications or other than . normal : plant shutdowns. It further assumes no ' releases of radioactive materials requiring offsite . response or monitoring are expected unless further degradation of safety systems occur. The licensee determined that the event, as discussed in detail in this report, did not rewit in. a degraded level of safety or in a plant condition, (including its principal safety barriers), degraded to a point that an Unusual Event should be declared. The licensee determined that the event was an Off-Normal event and notifications' were made in accordance with -its Nuclear Operations Procedure (N0P) 88-A.2, "Non-Emergency Notification of BECo/PNPS Management." The procedure identifies personnel to be notified including the Station Manager. It also requires the NRC Resident Inspector be notified as well as Commonwealth and local officials. The guidance provided in the procedure for declaring an Of.f-Normal Event includes contamination of a clean area, personnel contamination and events which require increased attention by the Station Manager. It also includes a caution that, at anytime a determination is made that the event should be upgraded to an emergency classifica+, ion, to immediately do so. In the Team's independent assessment of the licensee's decision on classifying the RCIC system pressurization and transient and its subsequent deportability, consideration was given to: (a) the dura- tion of the transient; (b) the capability to isolate the RCIC system; (c) intact low pressure piping and components; (d) the extent and duration of any compromise to containment : solation; and, (e) the integrity of the reactor coolant system pressure boundary. When the NRC noticed its new reporting requirements (in Federal Register, 48 FR 39039, dated August 29,1983), guidance was provided which recognized that licensee.; may use engineering judgement and experience in determining the classification and deportability of an event. The Team concluded that the licensee's initial assessment was performed in a timely fashion, the status of the RCIC system was understood, the 7-day LC0 was entered on the RCIC system in accord- ! l ance with the plant Technical Specifications, and the level of over- l all plant safety was well understood. Therefore, the licensee's classification of "Off-Normal," which included notifying the resident inspectors, and the determination that the event did not require prompt (one hour) notification of the NRC via the ENS was reasonable. L - - _ _-______ _ __

- _ __ _-_ _ _ _ _ ____ - - _ ._. - _ - -- $ 60 The Team also assessed the licensee's consideration of the four-hour ENS notifications required by '10 CFR ~ 50.72. The licensee, sometime later in the day after- the transient ' occurred on the morning of April 12, 1989, reached the decision to issue a press release and notification was subsequently made at. 5:00 p.m. via the ENS. Prior to the ' ENS notification, the Commonwealth, local officials. and the ' NRC . Public Affairs Officers, (both in' Headquarters and Region I), were notified. The decision to issue a press release was conserva- tive in that the plant was currently. in a controlled Power Ascension Program, and the RCIC system pressurization transient had posed no public health or safety concern. Four days later, the licensee informed the NC Operations Center via the ENS phone on the morning of April 16, 1989, that the plant would be shut down for a planned maintenance outage to work on the RCIC sy; tem turbine main steam leaks and nuclear monitors. The licensee also indicated that a press release related to the outage was planned. Subsequently, in the early afternoon of the same day, a management decision was made to exit the 7-day LC0 and declare the RCIC system inoperable, resulting in a plant shutdown being required by the Tech- nical Specifications. An Unusual Event was declared at 1:40 p.m. on April 16,.1989. The decision to declare the. RCIC system inoperable was based on the testing results (detailed in Section 5 of this AIT Report) and ongoing evaluations to determine the operability of the RCIC check valve. Three days remained in the seven-day LC0; however, the dr. cision to declare the system inoperable was based on the lack of assurance that RCIC system operability could be restored before l the LCO would expire. The licensee also made a preliminary deter- mination that a Licensee Event Report (LER) would be provided within 30 days to the NRC pursuant to 10 CFR 50.73(a)(2)(1)(A), ". . .the completion of any nuclear plant shutdown required by the plant's Technical Specifications." The Teams' conclusion that the licensee's reporting of the event was appropriate and reasonable is based on the Teams' understanding of the event as detailed in this report. The licensee's initial cri- tique, immediate actions, investigative process and followup actions support their overall judgement in determining that the level of plant safety was not significantly degraded, t 9.7 Conclusion on Quality a- ope of Boston Edison Company's Response The Team determined that the immediate on-shift response to the RCIC system pressurization transient was generally appropriate and timely. The licensee held an expeditious and thorough critique of the event to establish facts and determine causes. _ _ _ _ _ _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _

, O 61 ' T The licensee's team approach to the event' investigation was effec- tive. The licensee's Investigation Group ~ was a. well-structured organization,. with broad-based discipline and a clear scope of investigation. Senior BECo' management involvement was visible ttiroughout the licen- see's investigation. The licensee clearly understood the details of the event, although. parts of the : sequence were difficult to recon- struct due to the inoperable plant' process computer (EPIC). Overall, the licensee's investigation process was properly paced, raising appropriate questions and leading'towards the correct paths in deter- mining root causes. Throughout the AIT, the licensee was very cooperative and responsive to the Team's requests. F . B ' . ____.--i

- _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ . - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ . _ _ _ _ _ - _ _ _ _ - _ _ _ 62 10.,0 MANAGEMENT MEETINGS During onsite inspections on April 13-15, 1989, the AIT met regularly with Station Management, including the Senior Vice President for Nuclear Oper- ations and the President of Boston Edison Company (BEco). Following a presentation of BECo preliminary investigation findings and technical details, an exit meeting was held in the Region I office on i April 19, 1989, with BECo Managers including, Ken Highfill, Bob Grazio and ~ Ed Kraft. Phone calls were held during the time frame of April 20-26,1989 to further clarify information relevant to the AIT's findings. ! ! , f l _- _ - )

_ _ _ _ _ _ _ _ __ _ . 1 1 l APPENDIX A Pilgrim Augmented Inspection Team (AIT) 1 Proximate Causes of April 12, 1989 RCIC Overpressurization .. 1. RCIC Logic System Functional' Test (LSFT), Procedure 8M-2-2.10.11.1, con- tained an error which resulted ir; de-energization of the M0-17 steam sup- ply in. an open position. Note that one RCIC discharge valve was never tagged; the turbine steam supply valve was incorrectly de-energized instead. 2. Tags specified the breaker numbers but not the valve equipment name/ description. .3. Tags were applied incorrectly; 6 of 7 were tagged, but breakers were in- correctly left either de-energized or energized, although the. valves were properly positioned by licensed personnel in the main control room. 4. Independent verifications were signed-off by two operators, but not inde- pendently, and failed to detect the incorrect tagged errors in 3 above. '5. I&C " review / inspect" (in the field) and acceptance of tagouts missed the errors in cause 3 above because the acceptance was not conducted per pro- cedure. 6. Lack of a Shift Pre-evolution Brief (in accordance with Conduct of Opera- tions Procedure 1.3.34, Step 6.10(1)G) by the Nuclear Watch Engineer which is required to be conducted as close to performance of the test as possible. 7. Licensed supervision (N05) approval to " start work" (or test) at 0837 hours on April 12, didn't confirm proper RCIC test alignment and therefore missed the valve misalignments on the RCIC control board / panel. 8. Licensed operator (RO) cognizance of plant status (as he initiates turbine trip push button and isolation reset, as his part in the test), fails to recognize 7 indications (i .e. , panel lights, including an inoperable con- tainment isolation valve M0-17) of misaligned RCIC equipment. 9. Lack of precautions in the RCIC LSFT procedure regarding Event V and the hazard of simultaneous opening of RCIC discharge valves 48 and 49 (Ref: NRC Inspection Report 85-11). 1 A-1 .. - - . . _ _ _ _ _ _ _ _ _ _ _ _ _ . - - ________

. _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - Appendix A - Proximate Causes of 4/12/89 RCIC Overpressurization 10. Procedure validation processes missed the error in the LSFT procedure; also, the procedure hasn't been " human factored" to date. 11. RCIC discharge CV 1301-50 leaked by in the hot condition. The check valve was in an indeterminate state (as to fully seated) until disassembly on April 21; internal binding was found which was attributed to previous Furmaniting in March 1985 that was not fully removed by permanent repairs done in August 1987. The valve was, in all likelihood, approximately 10 to 15 degrees off its seat at the time of the April 12 event. A-2 _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ - _ _ - _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ - _ _ _ _ _ _ _ _ - _ _ _ _ - _ - _ _ _ _ _. . _ _ _ - _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ - - _ _ _

_ ___ - ____- < APPENDIX B Principals Contacted Boston Edison Company Jack Alexander, Operations Manager Ralph Bird, Senior Vice President - Nuclear William Clancy, Systems Engineering Division Manager John Dwyer, Operatior.s Senior QC Engineer Robert Fairbank, Design Section Manager Robert Grazio, Regulatory Affairs Section Manager Paul Hamilton, Compliance Division Manager Ken Highfill, Pilgrim Station Director Janet Kelly, Compliance Engineer Edward Kraft, Acting Plant Manager Richard Mattos, RCIC/ Systems Engineer < Leon Olivier, Chief Operating Engineer Michael Pyle, Human Performance Evaluation System Coordinator Elaine Robinson, Manager, Nuclear Information , Jeffrey Rogers, S&SA Division Manager Richard Schifone, QA Surveillance Division Manager Ronald Sherry, Maintenance Manager James S ery, Technical Section Manager i Paul Smitn, Chief Technical Engineer ' i I B-1 IL____ _ _ _ _ __. \\

_--_ _ - _ _ _ - _ _ _ _ _ _ - _ _ _ _ - _ _ _ - _ - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ . APPENDIX C Detailed Sequence of Events The 'following Sequence of Events was developed from test procedures, tagout sheets, plant logs, personnel interviews, personnel statements, drawings, and observations of NRC inspectors who were onsite during the incident. The plant computer was not available at the time of the event to provide a record of the various. system parameters or alarms. There are obvious variations in the times recorded in the different source documents and' the recollection of the person- nel . involved. Where this was' critical to understanding the event, the most reliable source or combination of sources was used by the Team to analyze the event. The timing was further narrowed by using the times indicated on com- puter printouts of personnel key card access in and out of areas of the plant prior to and during the. event. Where log entries. were not clear, specific details not included, or brief - statements provided, they have been supplemented with commentary' for clarity. !: C-1 ___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ - _ _ _ .

__ . _ _ _ _ - _ _ - _ _ _ _ _ _ _ _ _ _ _ , . From Licensee's Investigatory Team ? ~ Fc. E.O. #2, I/C #1, and NOS o 1/C Tech #1 discussed { \\s,igned tagging sheet h surveillance test with NWE & NOS o s @ . IM Tech #2 & #3 pushed and S 1/C Tech #1 prepares tagging sheet U releahed relays 13A-K1 (Step #12) E $ \\ 1/C Tech #1 presented tagout to e Ei MO-130(-49 and MO-1301-48 9 NWE for review o 3i, opened N g, NWE permission to start a . RCIC suction \\igh pressure alarm $ surveillance 8.M.2-2.10.11.1 g 2o received '* i ' g 1/C Tech #1 discussed testing O . RCIC suction reliefhqlve PSV 31 evolution with 1/C #2. (n -* @ lifted N 2 s o 1/C Tech #1 Initiates Procedure $ $ . RCIC turbine bearing hi-te'mp alarm b 8.M.2-2.10.11.1 In Control Room r3 received s {g \\ E.O. #1 Receives tagout and og . Smoke Alarm RCIC Ouad area \\ tags from NOS <m received % . l g N Hanging verification of tags "o h.te Storage Closed Open Tank - Supply I MO-1301-12 Closed Closed i C-3 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . - _ _ _ _ _ _ _ ___. _ - - _ _ _ _ _ . . - - - _ - - - - --_-____.__u

_ _ _ . _ .. ? '\\' ' Appendix C . Detailed Sequence of Events Date and Time ~ Event / Activity Source i E 8:12 a.m. The I&C technician leaves'the control room and Key Cards sends two other I&C technicians to the cable Interviews spreading room where the logic relay panels, 930 and 933, are located to perform. required TP 1 logic verifications and actuations. 8:15 'a.m. The I&C technician returns to the control room. Interviews

to coordinate.the test activities. Key Cards- 8:22 a.m. The NOS assigns two operators to perform the Interviews required tagouts and circuit breaker positioning Key Cards in. the reactor building at motor control center ., (MCC) 0-7. - The two operators leave the control room with the I&C technician. 8:23 a.m. The two operators and I&C technician enter the Tagout Sheet- reactor building. The. operators were required Key Cards - to tag and position the circuit breakers in MCC-D7 as follows: Valve Breaker Position MD-1301-60 D784 Closed (Energized) M0-1301-61 D751 Closed M0-1301-62 D794 Closed M0-1301-49 0744 Open (Deenergized) M0-1301-48 D771 Open M0-1301-26 D764 Open MO-1301-22 D754 Open The two operators divide the tags, which Tagout Sheet identified the circuit breakers and required Statements position (open or closed), but did not include Interviews ,, the valve identification, the operators performed the tagging and circuit breaker positioning. Each operator then independently verified the other's work. The required and "as left" status of the seven valve tags and breaker positions were: C-4 - _ _ _ __

_ _ _ _ _ . _ _ _ - _ - - . - - _ - _ - , . I J Appendix C.- Detailed Sequence of Events -j d 'Date and Time Event / Activity Source a Breaker Position Valve Breaker Required .As-Left < M0-1301-60. D784 Closed Open j MO-1301-61 D751 Closed Closed a MD-1301-62. 0794 Closed- Open

• M0-1301-49

0744 Open Open M0-1301-48 D771 Open Closed M0-1301-26 D764 Open Closed MD-1301-22- D754 Open C1osed

• Procedure error. Circuit breaker D744 supplies

power to.MO-1301-17, which is a normally open steam admission valve for the RCIC system and also required for containment-isolation. This circuit breaker, 0744, should have remained closed. Circuit breaker D774 which supplies + power to M0-1301-49_was.not tagged and remained in the closed (energized) position, which was incorrect for the' test. 8:30 a.m. The two operators and the I&C technician Statements leave the reactor building and return to Interviews the control room. Key Cards 8:37 a.m. The operators sign the tagout sheet indicating Tagout Procedure the RCIC system has been isolated per the tagout Statements sheets. The I&C technician signs the tagout Interviews sheet indicating that the isolation'had been reviewed, inspected,.and accepted by I&C. The NOS signs the tagout sheet indicating the test could proceed per the TP. Control board indications of the valve position Interviews on Panel 904 were incorrect, as the result of Statements the procedure error and incorrect circuit Drawings breaker positioning, as follows: C-5

_ ___ - - - _- _ _-_______ _-. . _ _ _ . i ,g,- ,

Appendix CL- Detailed Sequence of Events ' Date and Time Event / Activity Source

• Position Indication

Valve Breaker Incorrect Correct ' M0-1301-60 D784 . Lights Off On MO-1301-62 0751 Lights Off On M0-1301-17 0744 Lights Off On MD-1301-49 D774 Lights On Off MO-1301-48 D771 Lights On Off MO-1301-26 0764 Lights On Off M0-1301-22 D754 Lights On Off

• When circuit breakers are opened and power is not

available to the equipment, all position lights lose power. The redundant position indication on vertical panel 903, which shows containment isolation valve positions for M0-1301-17, was incorrect (lights out). 8:37 a.m. The control room. operator performs.the steps required by the test procedure and the I&C technician'in the control room coordinates with the I&C technicians performing the relay logic steps required by the TP. The control room operator initiates the RCIC turbine trip push button and verifies that the " RCIC trip / throttle valve is closed, then resets the auto isolation signal reset for both logic . channels, A and B, per the TP. The events / The I&C technicians in the cable spreading room Test Procedure activities simulate RCIC cuto restart signal by exercising Interviews which follow logic relays. Statements occurred within 2 or 3 minutes prior to the termination of the test at approximately 8:45 a.m. C-6

,. -_ Appendix C - Detailed Sequence of Events D_ ate and Time Event / Activity Source a The two RCIC discharge valves, M0-1301-48 and Test Procedure M0-1301-49, improperly open due to circuit Drawings breakers not being opened (deenergized). The supply valve from the Condensate Storage Tank M0-1301-22, improperly opens for the same reason. Although the circuit breaker f,or the suppression pool alternate supply valve was not opened, the valve remained closed. This valve does not receive an automatic initiation signal. The opening of MD-1301-48 and 49 and the Interviews partially opened CV-1301-50 initiates an Statements overpressure transient in the RCIC suction piping. The following alarms were received: RCIC Suction High Pressure RCIC Quadrant Smoke Alarm RCIC Turbine Bearing High Temperature Southwest Quadrant Floor Leakage The RCIC pressure relief valve in the suction Interviews piping, PSV-1301-31 lifts. Statements Although the exact sequence cannot be determined Interviews after the initiation of the automatic restart Statements signal, interviews indicate the RCIC Suction High Pressure Alarm was first. The opening at PSV-1301-31, then occurred followed by the other alarms. Hot steam or mist from the relief valve initiated the smoke detectors. The licensee's further interviews of control room personnel indicates two additional alarms, Smoke Alarm and Floor Leakage from the North West Quad, were also received. 8:45 a.m. The NOS directs the control room operators to Interviews terminate the test and isolate the RCIC system. Statements The operator isolates the system closing the Steam Admission valve, MOV-1301-61, the Discharge valves, MOV-1301-48 and 49, and the Condensate Storage Tank valve, MOV-1301-22. The NWE and one of the operators who performed Interviews the tagging leaves the control room and Key Cards proceed to the RCIC quad area in the reactor building. C-7 - _ _ -- _ -

_ _ _ _ _ _ _ - - tr' i' Appendix C - Detailed Sequence of- Events Date and Time' Event /Activiti Source 8:48 a.m. NWE. enters the' reactor building. Key Card 8:52 a.m. .The operator enters the reactor building. Key Card 8:53'a.'m. The other operator who performed the tagging Key Card leaves the control room. 8:55 a.m. The second ' operator enters the reactor building. . Key Card Health Physics is notified via telephone by one Interviews of the operators that vapors are rising from Statements the RCIC quadrant area. The NWE also notes vapor from the "B" RHR quadrant. 8:58 a.m. A Radiation Protection (RP) technician is- Interviews dispatched to the RCIC quad area. The RP Statements initiates radiological control activities Key Cards by starting air sampling, takes. smears and performs radiation surveys. The RP then ropes off and posts the area. See Section 7.0, Radiological' Impact, for furthe'r details. The RP technician informs the NWE of the water on the RCIC floor. The NWE then calls the NOS in the control room and requests the floor . drains in both the RCIC and "B" RHR quadrants be ' opened. The NWE and operators check the tagouts at the motor control center while waiting for the results of the radiation surveys and noted the tagging errors. '

The NWE directs the operators to clear the tagouts and return the seven RCIC valve circuit breakers to their normal alignment. Health Physics clears the RCIC quadrant for access. The NWE and the two operators, wearing boots, shoe covers, gloves and lab coats, enter a the area accompanied by an HP supervisor. l l I 1 . ) C-8 _ . _ ' - . _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . . _ _ _ _ . _ _ _ . _ _ _ _ _ _ - - . _ _ _ _ - - . _ _ - - - _ _ _ _

, _ - - _ ----_- - - p , , , .

Appe;, dix C - Detailed Sequence of'

_* '

Events

, So'urce Date and Time Event / Activity ' The NWE notes.the following: Interviews Statements- The floor is wet and rust residue is over much of the area. The RCIC suction piping from the Condensate , - Storage Tank and Suppression Pool were cool to the touch. RCIC discharge piping, pump casing and suction' piping to check valve, CK-1301-23, including pressure relief valve PSV-1301-31,'were hot to the touch. These system components, including the turbine and motor operator were also dry and had a thin film of rust. 9:08 a.m. The two operators leave the reactor building ' Key Cards Interviews and return the tagouts which were removed. 9:11 a.m. The two operators enter the control room. Interviews Key Cards 9:15 a.m. The NWE leaves the reactor building to return Interviews - to the control room and inform management of Statements: the event. Key Cards 9:22 a.m. The NWE enters the control room and informs management. 10:00 a.m. The RCIC system is declared inoperable and Control Room Technical Specification 3.5.D, Limiting Log Condition for Operation (LCO), is entered NRC Inspector which permits power operation for seven days provided that the High Pressure Coolant Injection (HPCI) system is operable during the LC0 period. HPCI surveillance tests started as required by Control Room Test Procedure 8.5.5.5, RCIC System Inoperable. Log C-9 ..

- _ _ _ _ _ _ _ - _ - - _ _ _ _ _ Appendix C -' Detailed-Sequence of' Events ' , Date and Time Event / Activity Source 10:05 p.m. Kv/E tagouts were placed on the following RCIC Operators Log , components: Circuit Valve Breaker Valve Function Position Position- M0-1301-61 Steam Admission Closed' Open ~ MO-1301-49 Discharge Valve . Closed Open 10:15 a.m. Started Surveillance Test, Procedure 8.5.4.4, Control Room I' HPCI Valve Operability Log ' NRC Inspector 10:20 a.m. Off-Normal (not an EP classification) Event L Declared. 10:35 a.m. Massachusetts Department of Public Health Notification . notified of Off-Normal Event (not an EP Check Lists classification). ' 10:45 a.m. An LCO, A-89-44, is written for the RCIC Control Room system which was declared inoperable at Log 10:00 a.m. 11:00 a.m. Licensee Critique Meeting of the RCIC. event Interviews' held. NRC Inspectors Control Room Log 11:15 a.m. Operators dispatched to the HPCI quadrant to Interviews support HPCI testing. NRC Inspectors Control Room Log 11:45 a.m. Started Test Surveillance Procedure 8.5.4.1, Control Room -l HPCI Pump Operability Flow Rate and Valve Log

Test at 1000 psig and 8.A.15, HPCI System Interviews l Integrity Surveillance. NRC Inspectors ' 11:53 a.m. HPCI turbine driven pump started as required by Control Room Surveillance Test Procedure 8.5.4.1. Log Interviews C-10 _ _____ ______________._ __- -_ _ _ __ _ - _ - - _-_

p- . , , pry 5 , ~ 1 Appendix C - Detailed Sequence of Events Dats and Time Event / Activity Source -12:10 p.m. HPCI turbine secured. Control' Room Log. Interviews 12:15 p.m. Completed Surveillance Test Procedure 8.5.4.4. Control Room Log -Interviews 12:30 p.m. Completed Test Surveillance Procedure 8.5.4.1. Control Room HPCI System Operable. Log Interviews' NRC Inspectors - 12:30 p.m. Local Town Officials and Massachusetts Civil Notification thru Defense' Agency notified of Off-Normal Event. Check List 1:30 p.m. 2:15 p.m. Failure and Malfunction (F&M) Report, F&M No. F&M No. 89-1 .89-151 for'RCIC discharge check valve, NRC Inspectors CV-1301-50, is written by the Shift Technical Control Room . Advisor (STA). Log 2:30 p.m. NRC Public Affairs Office in Headquarters Interviews notified of anticipated news release. Statements 2:35 p.m. NRC Region I Public Affairs Officer notified Interviews of anticipated news release. Statements 5:25 p.m. Emergency Notification System (ENS) telephone Control Room call to NRC to report a press release relating Log to the RCIC event. Call in accordance with NRC Inspectors 10 CFR 50.72(b)(2)(vi). 5:45 p.m. J&M No. 89-152 documenting ENS notification F&M No. 89-152 written. Control Room Log C-11 - _ _ _ _ _ _ _ _ _ _ _

- _ _ _ _ _ _ _ - _ _ _ q ~u T L_'. M '~' ~ ~ *^ ' - c.-a es . < J APPENDIX 0 " ' ' " " I!!EfB$rf$ '{ Pilgrim Station RCIC P&ID s- N a .,u.. . w . , , . m m: Y i

3

a a _ 1 . I "*- 9 4,_ 'c --- % , M, , y .cr . . . 4 . wis at: a .u: a ,, YJ C.'//l.!f Gs~ c M- (c D1U ~~-- -y e. .- r m wa , w ,,~ O% T % m s. u ... w -N "' *~' " L p; f h:a ds' ~ , r ;.s

.W h,, - .m' _ .d_ _ .?" iN. e , . . - n . r* T- mN y*I- - "fi- -- - ~' o :4 M;;* ; * 'C: lL e

• • • ws. e .a . .N 'T

S'** J

J 1 G l1

• L ao

mr . . 'y = ' ' _d g,6 --- - " ' ' ' " nc .c't n 3p . l

~ ur et n ~ M . \\ ,o..s / -- v [ '. ,

Lf, l p [n.~ . .nd Q*" j - m - ,

mo- ams - g , gn .m - , J

-

Ve L "i r" 5 . 5 /'if, 5 e - 4 !g~' I s- 1 " > . . . -i- pi -- -- 1 l { 4 I W , , g " T;..'- : , h .rs P'.'", %': ---,% [ s ,gfg , - P- -t . $, j 5 ' l'*fe-** h L f . . . ..t oe b*( . ,. .q - - - , . 8I

• * * * *

I i~ f/ 4 e p,p . . , } l

• a et

&,os;g 's* V ,6 3 y, ggy '..'h_ , sit ec s re-

• '

-l $P h ;3 ~

• p M, % d $ sj k.... -

30, xl .["" 1, E ' w ,e - w .i e . , .. s , . ' ' ' ' j 1 _ -- t6 .. s' .. y } 3, g,. Q at e s es. -t ' e ac. ew - .. g

• Q<

.?. -5= ' .,y * m~.h '. . .. w , ,, ,, w e DC ,, p- oc h h , . - . . . 3 h-

• -

, p n a . . . . 4 / e, u w,',,[. % !s[ - , -- . ue. ., . <. . , l ' u. . y*o r 4':', I, e .., l ,, i , , , - i q l PIPING DESIGN Line Station Material Pipe Specification Temp / Press Temp / Press Schedule HA-26 100 F/30 psig 500 F/150 psig 10 HE-26 100 F/80 psig 500 F/150 psig 40 i HD-13 170*F/80 psig 500 F/150 psig 40 EB-13 170*F/1415 psig 850 F/600 psig 80 DB-13 562 F/1600 psig 850 F/900 psig 120 DL-6 562 F/1600 psig 850 F/900 psig 120 NOTE: Class breaks are circled 1 D-1 1 i

. ._ _ _ _ _ . - _ _ . - _ _ _ _ - _ _ _ - - _ _ _ _ ._ ._, i APPENDIX E LIST OF PHOTOGRAPHS A. LOW PRESSURE SUCTION PIPING B. PSV-31 RELIEF VALVE C. DAMAGED PRESSURE GAUGE 1J01-20 D. FLOOR SUPPORTS E. BREAKER 771 (VALVE M0-1301-46) F. BREAKER 774 (VALVE M0-1301-49) G. D744 (RCIC STEAM ADMISSION 17 VALVE) H. D771 AND D774 WITH BOTH CUBICLE DOORS OPEN 1. SQUARE NUT AND EXERCISE FITTING 1 J. EXTENDED BUSHING SIDE K. DISC AND INTERVALS L. LEAK OFF FITTING AND EXTERNAL EXERCISE SHAFT l I ' ______._._________._._-_..._._________________._____;

._ PhUlublOWild A & b RCIC PIPING AND QUADRANT (PRIOR TO DECONTAMINATION) y, 1 > l i ,

. , l - m c, - ,i

- 1 . , ,, .; , ,

4 ' ~ s LOW PRESSURE SUCTION PIPING "A" f - }; , ,, v' b .. m . f.;h ..,y . : Jo , f,:e i ., ; . ' .. y. (! ,'[,] ; al,, f .. k. . , ' ' '- a . , . "k%WW, ,p~ h .]': 7,y - - < - .n l n L. __ ' 6 \\ . - - _. \\ . .

' 9: ' .! . , .

.- . . - - ~ . , s. .-

? . , s. . e. '

' f l -^ ' ' A >b ' . . .. Y,, . fi$;1 . - . s .:.;. w-af\\ ,! . . - s , . N . . ? E ' %

.% a .

, - 4*

- . , . ;,. ' 3.- . e 44a - . " , . . .. . ..;. - +- . . - . ' g , , ' ,q_. ' ^ % y t$t ** !' . ' 4 j: ' ' . . ;f.

,7 '

s . e. . . . s, s .. . sg .y ,3 f . ,- . . ) PSV-31 RELIEF VALVE "B"

Pil0TOGRAPilS C & D RCIC PIPING AND QUADRANT'(PRIOR TO DECONTAMINATION) '. a 1.s

tt

. , .$ - , ,. l$W 1 1 . I 2.z," i gt. . ggyg&s,,iop,;- ,n,Ag,. vi .j~t 1 , , r;

s jos . . a , . :

y ' t; , ,a . . y , . )> . .

y g. , , 9J;q ,4 . , . . . ; ;, , - M, ,. , . . d, ;,, . w~ ,. o .;p , . 3.- . u;- .u , f,t;> - g2iM. e , Er!"' f . . e . n . D , f qs.gu . .:'- , ' i , g-f. - .: vm . . , .. . .. ' ' " Y ,;ll i 'k, .I , ' ' g, , a ; , , - .- c . ,, DAMAGED PRESSURE GAUGE 1301-20 "C" . . - ~

4g.

. . - , 4

• • ; s

. y -

r . . ,- . , - s , g- .

,.

!, , f, p.

. . - . - ' -. - . . ,. ., . . . , , ' l 'n -,s. ~~ -

p. -

, . , ([,Q ' ..

, % ,7 9 - . . [ . f. .$ . d..--. ,. . , - , , - .- s . _' 29 .M I3 ^ , " ' _ v .- . FLOOR SUPPORTS "D" -

PHOTOGRAPHS E-H RCIC 07 MOTOR Col 4 TROL CEt4TER CutslCLES 7 m -- .. . , . , . ! , . , . . - . % ., Q ' .. s'g 1 ., BREAKER 771 BREAKER 774 a "E" (VALVE MO-1301-48) (VALVE M0-1301-49) ugu ,

-

, [k '- " ldi - 2# Q~ ' j z 3 , ..f'f; Q

?hl

' . ~ - "l y i . . l'!: ' . M! ',j,,J in , A . ! i ,, k .[e , "G" D744 (f,1CIC, STEAM ADMisslora D771 & Dfj4 WIJii 80TH ugu , , . ,

-__ _ _ _ _ _ _ - _ . _ _ _ PHOTOGRAPHS I & J ! RCIC CHECK VALVE MD-1301-50 1 i . . . k

. _ . . i , - . - . , . s - . , .- . . . ,, ^ . i. i '

y . , _ <r .

. , . SQUARE NUT AND EXERCISE FITTING "l"

. , . . - o . , .. gfg.? .. . . . . , . . . , - O _ .- . .. .. s .. - _ .- - _ .- , ... _ .,b ' : ;

._

^ l . i

y " ' c., 4fi Q " , , .tl- ._ , , .. . .< -

- u ' ;s!c ' * , , ' - .e ,( 2 4 .m ,. 4 .; 3. . ... e _ . d' f , f y ' > , %,. . ic. . [" , ' ~ , ;- f ,., ,. . ; . . ,/ ' ' ' , , , . ~fh ,*' ~a'b , ,? k[: ','(- . ,Y Y) ^ -: I. n - c.; .1 , , , . .p.g . , . . ~ ' " ., . ..w . .w.. . .. I-

'.,l

...a....,' 1;e ' - ,..] ,.ry - w,. p _, ., - - , t. ; EXTENDED BUSHING SIDE "J" . . .

Pil0T0 GRAPHS K & L RCIC CHECK VALVE MO-1301-50 ya >- J t i '!j l . ".'y 'h 3 }\\~ , , >

y

M . . . A. s- n a ,; ?., ,y V:yl$ngi% .. ' -

<

c;4 c ef# I M e 4 ,', g{, g;;;1 g > % f ' ; f y h,, " . ' / ~,.!g.t.h,p. g:J : '

,. [p,[,'i'h; ; ;, NB:.,jN' ' y ' ~ ~~ s .. ,, y n " ' - ' . ,.. ' , - c 1 7 !' ,',[ 9 , P 4;,v,Nio4)D;y . . , .;p *i : : +4

s

-- YM , '; 1A ,;- 3 '* ': } ' 'I' O w. .m o., ,, DISC AND INTERVALS "K" ' - . . , g, r .) g;. car u- ) - ., , ,, ,- i - 4 t~~- ., p 8*~;e -:fi;gt- .. , < ,GLhk* yy , a5 < .g . [ 7.,.4,j - 9 , , 9. Yd,6 ! . N )*; ' ,y 1

' + , <[ ?; ' I %

f )p '- [4 ' / 4 L 4 W Y k t ., b( "A s .Y JG z .$. g? . [g/j l} / t, N , j. - ? j.,- .j , , t - > > m .j.q)':[l4 /, e[.,,.n,1 2 ' . t.<

T ..g., ) & g,r

a; - f 47) 19 e, 'L[:'. 'I [I;p; ; f s. ,, If f ad ' ' iN - , p. , ,. . q he - cy - , , , a . 3-[y . .. g- g -f ,, lEAK 0FF FITTING ANo EXTERNAL EXERCISE SHAFT "L"

_ ______ -_____ -_-_ APPENDIX F Reactor Building Level: -17'6" Chemical RadWaste J Sumps RHR & Core Spray instr. Racks i' HPCI instrumentation Stai s RHR Pumps Hea E cha g r CRD Pumps Stairs H&V Unit L8 Q' " ' I Cool 2rs D . lgI' ' = '

, . c_, glll l l Hatch Above

A ' '-' prain valve C BV l l =O . Pit w l I Escape ,_.g l " * = Shaft - l j , . e Sp ay Pump ll l CRD " HPCIPump - .

Pump Suction . Top of the Torus 6tructure Filters i e l HPCI Gland J

l SzalCond 7 U g Auxiliary 0- . HPCI Ture UO Coote- Lube Oil Pump Drywell e 1 E l l --W--

---

d---------

-- E --

---

-

---

---

--------------

I 1 , 5 k ~ j -- Elevator y Vanhole - -- Stairs n,c g,,,, l ) Control . = RHR and CS l l llgl - Core Spray ' RCIC Pump Keep Fill Manifold m h, . =

RCIC Turo Pump . RCIC ". $ D////ssrshisisissisis//$sl /miskf Condenser RHR Pumps = ' <//// Stairs RCIC Quadrant n s a toop -4- Stairs (SW Corner of Rx Bldg) sea, g ,cnang , (Turbine Bldg ) RHR and Core Spray Instr Racks ! _g_

F-1 _ _ _ _ _ _ - _ - _ _ _ - _ - _

_ _ _ - _ _ _ _ _ _ _ _ - _ _ _ ' (v k , APPENDIX G Acronyms -AIT . Augmented Inspection Team ALARA As Low As Reasonable Achievable A0 Air-0perated . ARC Alarm Response Card ASME American Society of Mechanical Engineers BECo Boston Edison Company CB Circuit Breakers CIV Containment-Isolation Valve CK Check Valve- CRD Control Rod Drive CST. Condensate Storate Tank DPM Disintegrations Per Minute EAL Emergency Action Level E0 Equipment' Operator E0P- Emergency Operating Procedure EPIC Emergency Plant Information Center ESF Engineered Safety Feature F Degrees, Fahrenheit GPM Gallons Per Minute HP Health Physics HPCI High Pressure Coolant Injection HPES Human Performance Evaluation. System I&C. Instrumentation and Control IGSCC Intergranular Stress Corrosion Cracking .INPO Institute of Nuclear Power Operations IPE Individual Plant Examination , IST In-Service Testing LCO Limiting Condition for Operation LER Licensee Event Report LLRT Local Leak Rate Testing LOCA Loss of Coolant Accident - LSFT Logic System Functional Test MCC Motor Control Center MO&AT Management Oversight and Assessment Team MOV Motor Operated Valve MPC Maximum Permissible Concentrations MW Megawatts NDE Non-Destructive Examination NED Nuclear Engineering Department NLNPO Non-Licensed Nuclear Plant Operator N00 Nuclear Operations Department NOS Nuclear Operations Supervisor NWE Nuclear Watch Engineer < G-I 4 _ _ . _ . _ _ _ _ _ _ _ _ _ - _ _ . _ _ _ . _ _ . _ . _ _ _ _ _ . _ . _ _ _ _ _ _ _ _ _

_ _ _ - - _ _ _ _ . ____ ___ ___ i L Appendix G.--Acronyms PAP. Power Ascension Program PCN Procedure Change Notice. PDCR Plant Design Change Request P&ID Piping and Instrument Drawing PNPS Pilgrim Nuclear Power Station PSIG Pounds Per Square Inch, Gauge RCIC. Reactor Core Isolation Cooling RCS Reactor. Coolant System RHR- Residual Heat Removal R0 Reactor Operator RP Radiation Protection RPS Reactor Protection System RWCU Reactor Water Cleanup SAR Safety' Analyses Report SIL Service Information Letter SLM Standard Liters (Air) Per Minute STA Shift Technical Advisor TIP Traversing Incore Probe TS Technical Specification- TSC Technical Support Center . i a G-2 - _ _ _ _ . - - _ _ - _ __

& s .- e e 1 h m 0 c u 3 l 1 rn o eo V &ti P11 p . [<#, , nt e98ec p44cu i - - s , 00mo o33ut t1 1p a r .~ mo ooe it ttt g x n- ' o098 ee r544cm p- - - u p111 pl A000mo 333uV 111 p -

l mmmma 0 -

1. .O

oooot rrrro 5 - FFFFT 0) 1 3 1 S s,s% _ " ~ [o% % .

1 egra h ., cc 9 .a s3 o4 w .. w . . y P a ., Mt u0 . - ,W.h .s c y. 3 1 o ,% g T 3

1. g

- t ' 5 ' g y

• • '

,, +< M' m 6 - 0 U h* ,, ' (' n 2 8 - n 1 p 'l i t N ' 3 ' 5 M O 9 M ( ,, O r 3 0 1 / 9 . , e. t x) e )- . . , _ c _ [O7 ;<,. , ,s, _ o R R _ _ g _ _ _ N _ _ x" _ _ _ l

APR 26 '89, 12:00 P10 . . . . , APF: 26 '99 10 43 BOSTON ED2 SON P.10 l j 1 ' S y I f l l g I

. I fit 3J,U' hi :l a ~ I ' a ? " - 4ffi jij I 1 l Syd Z '? ' -*=qd l L ji pi

1. 4 ffh 4 j1h

3Jjh i !!hi til f 1 ~ jjf 3:g $f dI ' 5 I d i i u m+tm" . oe a min, .

- '; 6 J i t 4,( \\lf i i!Ill e R G o _f. !% % 0 f 9 % h% li q [,ytTE-l sh A ar i s, ~ dy5 . , j Y. \\ 9/ \\ l ,}j r , ' -% '~ g - f{kg fk4 @\\/. .l e 3 _[ /a;s . !f9 'V4%

• ,, #

/ ij ~ 3 ,4 g* f p* \\ W @ R .'*A. , s , . k N _m v ' . '/ ,? e \\33W,g . ls g

• 1

1, 'S g / I'b I .b \\ /' \\ @? ,($ \\ "4 /~ t ' ./ v' Ep A, 6/ >

pLlg

\\v , , "2 ,4, S ? Y . 9 % I l " - $d p e i_s

. .2 d' v ea~s -u g l'H E - ~ \\,/i g!s ' ,,': . - '\\ d+{g !

H-2 ._.

_ - - _ _ _ _ _ j dd J!y y'l dif ' .j s Il jhlp ~

' . . E .S $ l J ? . k !I !'f [f.) \\, g 7 1 , f N 1t,1 91' \\p - ' kCc l l,? l lal lI i 11 , ! ' +-w r ; , -,. . , , , , .. g g a

%."- '/. I

. ,; .4 llf 7 e. y g p ,' Sy % < M U -- ,. .' . _ lli 2 g l l l}g ~ . 7 - . _ . . , 3 3 A !, / ,, ,., . . n .L 1. c

-

- . 8 1 u . .t (h [* I s [,b 3

• 4E8

li , + e.s. . e; t. j. . t . -, n - ,..;y.$' .~ - .. ! .. 'q .,' .. g/ j E - .. d t. sN' f "y r#g&(\\ ss . - >g. s ., . _ - h '@ ~g , @

1. 's /

9 '" 9 /7 , - . ~ ~ ' ' " - 8 3 ? 7 ea l @s, q !$f h ! 'O I'. I d _" . , M ~' ' ' - - d) /// ' ., ' 4 %

s 4 , , . //. II i . . . . . . . _ . . . H-3 - _ _ - _ _ _ .

P13 1 APR 26 '89 12:02.. . . - ,3 . e 26'i 10:50 oH EDISON p,13 - l i . . - Iy.$s yl ey y , . h Q, lI % . l T,u ~y .-- - - ~b - j ,,,,

. i 3@ % 1 I st r i g c, 1: g e e. M.. .( , ' s, , ,.. @her , b , '\\ g,\\ . 7 < $ .- T' \\ Q i d sp,i c . i[PQ:, M,- i \\j y , , - . c--d .. -4 M 'f. 4 .. , ' \\' (h 4 }\\,g 7 ' \\ <' s a j $ .! $ i '\\ hI -# 1\\'Q/

,@ \\ v g. - , s y I ! l i

, '

1. t

' ,x NV p w ,[,* \\ ..f{I, @,f g4 . j'pd ~* 3, ) - ,,

• [[' T4

"'Il$[l l llll . , . x m\\ ' a\\ , , m . "\\ / @ s. N e E !r.1 ) !\\ \\ s

,

. 2 . ' . . ,o..s 'MW"Fi"' l

[ e 5y

' !; n a a 1, (/ H-4 ' LJ.#f6 j

"~P=:23- _, __ ] _-___ - _ - _ - , , . , - - _ I.' ' ~ .I I7( J IC , IC, APPENDIX I ~ pf, }[ .. . .'!. - tQ h c ' ' x .}. - , ,. , C ' DOSTON EDISON CO u . >., , ' y; . -*

. QUALITY ASSURANCE ' , , , . . 3.t ,# - c a,.. a= ( S. .. c. ,5 vm m u.nue j se i . . - ' . g +a . < e ., .7nfL = h: ..,c.

• . .. . , . .. ,) ;' -f. *

}. 5o' *gs . /} y p [1 \\ t . . ~ r \\ ' , ' . om . ' 4-4/L7L- - r. n - , , , , , . .

<-

1 - M A 6 ' 'c !' .' r ., . ' . . ' . r,o # Qp ON T105 L5.0. NAS BEEN SN0P AND FitLD WIL85 . ._ .

.y:// q. ' i . . .? ,' g [ . . J g INSP5(TES AND C0'MPLf ' . ' ' '19 J: . g .

c y

g.fi.

% ; 4" CN W3M U.S.A.L*1G1.1.0-1967 t, jl# A u R g s. fT - $zs.*,Of~R. h . Yg .- g f)5 ) . a/. . e- 4 h.?g. v v.a.u] (- e-$" . 5 8 .

, ..

, 7 9 ~~ ' ,ik,,h C'c;p\\f /.cf N - . . - %, ., c. '-

?,- . -- .. N '4% % <* 9 .-, , l ,f gh _ff , ,v. u, -

1. ^*

. '/ '\\ ~ ,%k %s Q s.N~ f gn / ~.v - Ne , ,s . . s .e ~y' . - . N k ,gC . s h* '~ ' i $ % f .c v -/ ~ g3.2v% . l ^.s. - o , 4-; r&p 's' 4 ,' .". -- ,e , ,, y <>..o e % V l c. &. y, , g# , . 1 - .v + %s. ., , - M, < 9, '.7 k.9 /,a ,e r sQ.. e~-A% 4/- 7 . .c % ' p / .% 1 ;. . > - g re . %' w -a.- ?$h,;'han-

• ? % . /. h f ,~ % f ' Y, sp

.. - .. .s N ,> . ~ . . . , T < ' - . W gy* 4 ~'v 4. , -

Av/

%,_- < - - .- / d* ' i

---

.-.--.;

V . _ oa wg - - 8rss . - ,' h. % vn it . . n ~ .- - n u.u. = .w x g.e - m... ~ u. s. , .m _._ _ . ww r >.,i ! . 2mm x a m .., w .a, START-UP S , = db/ dC . sa.ed,628,.5ML S, , K N ~ ' Y Y. W ? ' ,- _ .sek:nd.,irsf cri aves -- - --. 50 18 W. A'. , 'PR A IC 5, G PC SCH 40. BotE

t .o os sva.o ; ee spreo r'A L - SPOOLS * JoosC MATf244 \\1LW5 .' 2 5,/44:,7 - sporp inst aw r 'l f WHs2. /* "' *4 Mi:":?, ',fia ,, a bre vs. 2sgasi.u): -- - sca 1-s f. c ino se rs tr.o s sic arrie c e

• "'o 'os '* """"m*"-
• 2

' .

• 3

' ' N'70 kov'4* Fees 4 F.v ' . ss ~ : . 3 94eaccrb, 4 HA M tes , G" CNCCt* y/ M23f(tJot.ty) 2 n, 7,, g g , , , , , ,. .;, . . , _ _ " ' ' ~ ~.: i y "1:

• ~;;.c a,,g.g~->#p- C

.

• 6.,.

s c,a rr vn.' t n G. set.?o) G z i- s', . . , '~ . H-is 1 r s . - c~ cuccg ya. ts6 (t)of.gg) + . ., . .Q",%* f','s."$ $~.? . . . . -

• Alf H -

V. '.f~.u..f, . ' ~~ . .

;, .

, . ._,4[ : .$'u:t t if. 5%"J'"4'kDtwQS?q,sae U.;.: ?; % M L:Q*;* -]. p n --:u _.)'yw " . ' g . .. . _ g , - .a . I-1

_ - _ _ - _ _ _

O f I 7 1 S L N S 1 O E O - 1 M R O I L yR P P D 1 S P S P E E O T U N S f7 RO S E O P P 6 > P L P 0 N O C 1 U \\ 0 - 10 S Y Y 3 L L 1 L L A A M M I 1 S 3 P R R - 1 0 0 O O 0 3 N N 1 1 N"/ g>A - y 6 p 2 g22 Y 1 g' 0 L / 3 - AP 1 EP TU SS 2 E 2 T - AE 1 SG K 5 'cN c NA N 1 ER A 2 'n DOT W f1 - NT 1 R OS C 0 h )[- 3 1 0 C 6 O T - 1 \\/ 2 R E 0 6 3 E v_J - '1 S E 1 V gg 0 T 3 S PyE 1 A A 8 J 4 3 1 5 - N NM - 0 O E U 3 E9 0 1 C D P 1 3 V_J N L1 V O C R ME ,9 D E 4- E V )0 E T 1 3 E A \\ L T 1 AP \\ S V U Y LE 50 1 I - S AN 8 V1 - 0 A I 1 ML 3 0 2 C 1 6 2 - R U 1 v 6 - C _<c ICO W y 7 N , - R R - <- , ' !

'f I 7 1 S L N S 1 0 O E E O - 1 RO L 3 T I I PP D Pi 1 S L rRO S P E E O TRH U N S f7 R O S E O T P P v > P L Gg% 6 o O C P 1 U \\ - Y Y 1 S\\ 0 L L 3 / L L 1 A A M M I 1 S 3 P R R 0 0 O O - 1 0 3 N N 1 1 I / - hT / g >L EN V 6 J B 2 R 7 Y U 1 0 L / 3 - AP 1 EP TU SS 2 E 2 T - AE 0 1 SG K gg3 Oc" c NA N rE1 i ER A 3j 2 R DOT W 1- > NT 1 R OS O r A C O 1 06- T 1 \\/ 2 R E 0 6 t'3 E - 1 S E 1 V g0 7d T 7c 3 A 1 A 8 SP J 4 35 - N NM ' 10 - O E U >3 1 E 0 D P 1 t'3 C N y 1 O lJ C R M ,9 4 D E - E >0 E T 1 E A T 3 /F W T S P 1 SE U Y T E M S DN 8 1 A' 3 - 1 - 0 A EI 1 L 0 2 Y'6 R 2 - CI U 1 V 1 6 E - C _ <C U V I y C QL W R M6 A R EA - 85 RV - Lb

7 1 S L S 1 N E O - 1 O M R O I 1 S rP P P L S M P - E O U V RO S P P V > 6 P O 1 U \\ G - 10 S 3 D 1 E N S i 1 s 3 E O - e P L 10 = O C 3 i 1 / - . g3 1 6 V N/ f 3 2 , 27 " \\M L / k Y 1 , 'd - AP 1 EP TU SS 2 E 2 T - AE 0 1 SG" ER" M1 E ** - NA c K 2 DO'^\\ w k1 - NT n 1 [0 OS

3 C O^ ] 1 @

T 1 \\/ 4 2 E 0 R 3 gZ E 11 y 1 S E E0 RN T X3 A 1 8 AEA 4 3 BD SP N NM gd1 5 - - 0 O E U gD3 E 0 1 D P L'3 1 C N 1 O 1J de 9 R v C in S 4 D E ei - c M F A E T ed 1 s e 3 E A 0 O W r e r E W t i v ou 1 l T P2 anq U1 E 1. D v2re r6 S EP R 1.E 1 0 os f 1 a 5 Y NE U0M n0 1 />0 o3m - 8 1 - T i1 o S L S D 1.R I 1 3 t 0 ado 1 r 2 - icn E2 u 1 V dal CE NC O o V 2 F C _ <C n0r t CL EO.R w

i6

I / En n1 A H R M. E R Te0 co 1 O p3 ' R VWP8P NO1 in - _ - c !

. _ - _ _ _ _ _ _ _ _ APPENDIX K RCIC Suction Pipe. Relief Valve PSV-31 Weld Repairs . . . E E pgg-tac t. B / ~ sacrion Retter I . E ' _K'E 1 _ _ iI] usata 1/

~ f axnuonso ' vaa- Tr . . . . . N - F74 cmc Daare JJc A Y9-9Y/

. . K-1 _ - - _ - - - _ - - - - - -

o .. i a < ... l. " . l APPENDIX L . . L . . M; 4m g sa_. - ' ' ' " " I 'n ammi- Check Valve 1301-50 X- - " ' , = i e ame Vendor Prints . ,,, y < s : === == em a II E 6 /Y/ . I i I r IIIS M10 418100 0003 4 f - I & b M NAL WWE1 IWe 08Lup 8tAfD h ' I 1 I I RBW ABW=S ! Ell / , il i IS T MTABIWI IRI5-W

• 8

I 8 IB T IETA805t 0155>19 i i --'

,C=- ! j. g I .I 8 .f3 Or ,u. WW '111 M BST j 'Isi - == , 6 4 illL i 11 11 13 &ag Wigt l --*

. ii EN T EET. WWFT PLERI WW SS m j Q $ % in - am --- := y 3 seemnas amen ___ ,,,, .. ',8 - appes M . I SCM tas Pter - -- -- , ,, l 4 FLNW WOS- 7 l* ' W N IEF Agg- qq lft M SWT S30813 m bete i 4 B05T Svm 183= Il 3 . ,sup ones igg. , s.n ess*= e e isa ene.e o i e im g3,333 R334F3 c , ses on.see , 3 . _ gapp 334 a.n i i i == 1I 4 I In3 GPLAAB i RRSS-Se .le,e l"I2 see . imesi uns ases e. ses ,, iuse une aan al +s. i W 1 IIIE san & i d-go I see onec mean a: i sie i i m ilNisRTER estPT sav 4 i gyggggg , I . sur m, aus aus 30 MN ndWENsena 4L Q3 ME FIlWI 50 E. PIPE 33 p.Scsan Pim ili Fase .N g., jy M8 .B88 PLAIDn ll Kim .Sf.m.L. P 305 -v . .. . . u - u M 33 appgWit sin a .3r1 seg FIM CAP 1 e i5 ,sk,,,E, ,,, espePLATE , m,.

ass i til see INT *L.m. i.g g . = . nincA .,, . 888E 8bilPNITE. Perf 333 8'T 81118 8 6,uggg% see IInt Mitadgs gun , peigg ggg , , . A MT*L agle Ales eu. puurt Mc mee M INT *L gas agW. dWED 8#83 &#93 FWF 833, m 48W 70 AWE 3. 8SI egyrgs

• • eusm4.S..,3,TAs seggyg g ,s,.agu

- , ma 8. tNILVE IS AEW CLASS I PER 8914 ABE SECTION E 4 59-15 13 ggg, O FIM 19 ISIR S. attu maest 1975 JEEDen 8EB.8444. M , m W pg . , , , . , . ....n...u..= . t't,.m. .=.,g;,g 1.w. .. . .m u& .= . . . . . EEV. Times _ puurv3 5. WILWE TO E IIST418D IN SEERIIENf% AIJ18 le *.' :'s,a_s 3;p,g,p:;r,=.g, , . . .l .e i.- . umn . , ,, .,.,,, . i. A ic . .. n ,u == . - . , , , , , . . . . - was one e m. ric=" e =n r . A. -=... nsc 3,.es 33 , , , , , , , , , , , , , , , , . E U m~,n.7, 8

% 1. itsessa saman un conne- ' ~ =+w . . .n. .AT ' ' sismase asumsm aan cumesses sua S1SS 38E TO SBlas FIEE. esurname w nses immen ' sen assess ePSS DISC. y,y es.pm,.v,pra g g.- y m,s- -

.. . n , ' I seus = m et.Las i . g=gi$rh 8" s. asesen amortsee asso pois At carr

i

.*l'::' E""~~*

IO_ (p. som ses.+ess l -* . - - - , , I .' i s. serr useums mis .e ecampusen ' NT BE 995 A .m""' .m . _, __8st= ~ ~ ~ - s ' , S . FMF 055 NSW OHer44 add ___V_

1. ffed# AF 065 4t3 W$d A#+ Mad +//* 4 3

l *) N

• • • • T a i i ' s I e ! * 1 * ' ' ? *

i n 1 n n a.a I i .2 a -_. _ .m s ., . . . . o- . . n. , f Ef#MJ BAft tag,1y, 1Mn h nt E .. I i' aluusarnasues watyt Gb 4fausef5PWif. PA. F101 _.. P1Lents sinYtes es, ese t isnt se. s suseum untsme careert"- minist ca. -- :-- -- simi s. I ""'"#4irri'fasi .efd- I'0 . - 3 DECO DWG _ O M132Ci N -- - H vaveenseamy _ _ _ '4"-992 HELD ENDS

CARBON STEEL EXERCISABLE SHING CECK VALVE ,-ns ,jwe;m.: 7 - - m ,. tp 5 7!El .. +-i- n.- ~ $+7 l , , . .R'_ i o. - . , . hMbI . + L-1 - _ _ _ _ _ _ - _ - _ _ - _

. - . - _ _ _ _ _ - - _ - _ l l ! r , .- i A- - - l ~ . _ _ , I ,.gg.a2f i F'27 I '-fh, // s,/// ' .// 1--a.a " ,, . -- - . ,---- l ,, h E S t "plo - _. esur a , 3 at!N 4,. " -i g!!!! '15 l , ,",, N y -- - -

190 WIDI A W CTil i C-C 230 W 120 PIPE yQ 7 g,g g 30 - p ._ n _ . _ . l WE'D Poseeduse 313 47 D****st Peeen

188 l l c., M w*e m _, 812 I I

s ., ..., , _VM . , . 444+y9 2 809 _ . 1 . -- cV, 7 tc _ 8 , , 180 h SSt -- f, $ --

b

k. . 7 . . '/ , ____ 7 . 'i, l l - ' - $l-

.-

• /

, ECTION D-9 ii l:

L.J _ 5 DIA 2 7 12 , g FOR DISASEDeLY . 1_a " oD " s s ,,,M M E

• Y$

N ,. , l E " &- = p - Os - oom ,

" ' kS ll ~= rrrra _ ,

_

Y

• • ' " ____

s . - - y %s i %. um/ k m /a 1 ...,,i.. >" - POR INFORr.in i - ' . - .4 - ONLY t -- 1 . . -i ... -i O,. - . , . _ .._.. .... . . . - --~ - , _. , l L-2

..=.a..=. . . m- - - . . ..: -..^.....z..=... -v- w .n - . . .. ) ., . . . . .. .* m- Attachement I- .

. . . Sheet 1 of 1 . 1

QlECK VALVE RCIC CK 1301-50 I . TOP VIEH - ' ! - . . - 4 ! - @tacu nut e' f L X t.-1/2" so ,, x t . ' ~ , , =.a s = . - lr i -l s'7 in ' _I . - -

' . . l 'N I ' . -</ r$, --6 . / -- , m .N ' / \\ Npg.}* - / / r+ y - - ~~- , _ / N I 1 l' ' ~ /A P / s - g 5 s y ' .- l\\\\Y k

s.i i i v \\ , 1 I -- ) . I l ' 1 I < LJ l . -

I TP89-39 Rev. O Page 10 of 10 L-3 _ _ _ _ _ , _ , ,, ,m.- . _s. . , . , . . . _ . . . , _ . . _ _ . _ ... . - . . . . . _ . . . , . . . . _. ., .

.- - u ,, , - ~

'#"* '(HGCkVkLVf lECI'5'O OlSC- SffAFT bfl M MENT , 4 . 1 - - 1 ' . - ( i 1

3. , ' - . . l l l .

. . ' ' oftr> < 4 . g l: ~( % MYWAY ' \\ ' . , s s , - -\\ . . i~stf. ',% l s \\ig-g N tY0TO - \\ s s. ,r \\ - PostinN SkoJAl IS THC n s\\ s PRCC s%nNCr ORICivtA nM. ' i ~ ' , rQ ~ r ~IN -ints pert 1t*U UALYCCAAI l *s\\ *~ ~ OMWr CPut 'sHkax(rt( 70$ . v -. , , , ee g \\ ' - . r- f ' 0~ > \\ , .. I . . ' ) . . . . r 1 > . - y s . , .

L-4 . . _ _ - _ _ _ _ _ _ - . . _ _ _ _ ~ _ _ _ _ _ _ - _ _ _ - - - . . i . ...- . . a c . . ca ' -

- _ _ _ - _ - - _ _ _ _ _ . /177 l' 347.2 # Z CHccn vntvc tsot-so Pos>Troas _.. . .. . . .. . . _ . _ _ . _ _ _ .. .. PCet740Al 5 8.. . . . h.ht 'ttM 5..._ . . . . . .. . .. . . . }ptTMN. 5..I _ i . - _ . . . . - . . _ . . . _ . _ . .- . - - - _ . _ _ _ . _ . . . . . . . . ... . l \\, ' ,. . Ifcas. .. . . _ . . . .. #c':a*6 . . . . . . , . . .._. _ .. . .__ . . . _ _ - . _ . . _ - _ _ _ . _ . . . . . ._......- .. . . . . . . _ . . . . . . _ . . _ ... . .. OStitmo .I % VAwr Osc../.S..r..d...'ht')~/.a Suist5- As_i7 tan) M.. .c. .hN. Re.t.n..te.'T. o l . _ _ . . . . . . . . . ._ . . .. T b Y d o ft F) f M t 7.t w o. _.. _.. _ ._ _.._. . ... -._..._ ...... _....._.__. _ _..._ _ _.___._..._. . . - - . . _ . - . . . . _ _ _ . - . . . . . . . _ . _ . . . . . . . . _ __ .._ . . fMsTwi E: VAiVE 06c Hns activ hMWLt4U.J EXEEC/ SED CWSdC. IN YMS . _ . . . . b 2'7t @ ..'TH [ . VMME'. 0/.S C. 6 Nd? 3G~~ 70. 0%. __._. . .. . . . . . _ . . . . . . . _ . . _ . . . . . . . - . . ... . . . . . . . . _ . _ _ . . . . . b % TotJ 5 : %2- Orse /S ^/ YNE d!&J MSittwo @M ne7 MYwAY '44Wr41stn> j ' i . .~ VALVE ca e wru-. Remw *P.e9. . . . . . _ . _ _ _ . . . . . _ . . . _ . _ . _ . . . . _ . . t j _., ... . . . .. . . . . . . . _ _ ._ . . . . . . . . ..___..... ... . . $ - , i 4 .. AS Gwb casa receu : _ . _._ , . t 1 ..

- . .. __. __ . .._ . . . . .. . , Qo'-ts , o 1 4AS$ tuts, l'hyC /S M/ 7tt'ECibSC'D fort 7/M. i L-5 _ - - _ - - - _ - - _ .- m o

_ _ _ . _ _ _ _ _ _ _ ___ '

' #*% ATTACHMENT 1

k UNITED STATES f S NUCLEAR REGULATORY COMMISSION ' REGION I g ***** e 476 ALLENDALE ROAD KING OF PRUSSIA. PENNSYLVANIA 19406 APR 131989 MEMORANDUM FOR: William F. Kane, Director Division of Reactor Projects FROM: William T. Russell Regional Administrator SUBJECT: AUGMENTED INSPECTION TEAM (AIT) - RCIC LOW PRESSURE PIPING PRESSURIZATION AT PILGRIM NUCLEAR POWER STATION i You are directed to perform a prompt Augmented Inspection Team (AIT) review of the causes, safety implications, and associated licensee actions which led to the release of steam and water into the reactor core isolation cooling (RCIC) compartment on April 12, 1989. The inspection shall be conducted in accordance with NRC Manual Chapter 0513, Part III, and additional instructions in this memorandum. ! DRP is assigned to conduct this inspection and Gene Kelly is designated as the Team Leader. The Team will also include participants from the Division of Reactor Safety and from NRR. OBJECTIVE ! The general objectives of this AIT are to: ! a. Conduct a timely, thorough, and systematic review of the circumstances surrounding the April 12, 1989 event; b. Collect, analyze, and document all relevant data and factual information ! to determine the causes, conditions, and circumstances pertaining to the event; c. Assess the safety significance of the event and communicate to Regional management the facts and safety concerns related to the problems identi- fied; and to d. Evaluate the adequacy of the licensee's internal review of the event. SCOPE OF THE INSPECTION < The AIT response should identify and document the relevant facts and determine l the probable causes of the event. It should also critically examine the l licensee's response to the event. The Team Leader will develop and implement a specific, detailed inspection plan addressing this event upon his arrival onsite. l _ - _ _ - _ _ _ - _ _ _ _ _ _ _ _ _ - - _ _ _ _ _ _ _ _ - ._. - - -

_ - - _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ L. yc i ,

, -;. Memorandum for William F. Kane 2 " APR131989 As a minimum, the AIT should: Develop a chronology of the event;- a. 'j b. Determine the root cause(s) of the event; 9 c. Determine the scope and quality' of licensee's internal review of the < event, including its review of the event for deportability; d. Determine the impact of this event on the RCIC system and its components; Identify if this event was a potential Event V precursor; and, e; f. Develop 'a history of operability problems and testing results associated with the HPCI/RCIC check valves. SCHEDULE The AIT shall be dispatched to Pilgrim Nuclear - Power Station no later ' than April 13, 1989, and shall remain there as long as necessary to accomplish the objectives of this inspection. It is expected that this will take no longer than three working days. A written report on'this inspection will be provided to me by May 15, 1989. TEAM COMPOSITION The assigned Team members are as follows: i Team Leader Eugene Kelly, RI - Team Members- Tae Kim, Resident Inspector - Joseph'Golla, RI - Theodore Easlick, R1 - Daniel Mcdonald, NRR - Lambros Lois, NRR - ame_ >n . ei 1 d ator ] CC: B. Boger, NRR R. Wessman, NRR T. Martin, DRS S. Collins, DRP i J. Wiggins, DRP R. Blough, DRP iTehm1 Members %e<a 1

-_ ___ _ _ ATTACHMENT 2 0FFICENEMORAhDUM Boston Edison Company To: J. Alexander From: H.S. Clancy/ M I Record Type A4.08 I P.D. Smith Date: April 24, 1989 Dept. Doc. TCH89-73 Non-Safety Related Subject: Oversight Committee Report on the RCIC Pressurization Event of April 12, 1989 - . Distribution: i D. DeVries H. Riggs R. Mattos J. Rogers S. Bernat M. Pyle Attachments: (1) HPES Team Report (2) Check Valve Team Report (3) ERM 89-312 & 325 Providing NED Evaluation of Impact on RCIC System (4) Review of Event Safety Significance (5) Draft Critique of the Event (6) Team Structure (7) Appendix of RFI's and ESR Puroose: The purpose of this memo is to collate and transmit the individual committee reports regarding the subject event. Attached are the final reports with short term corrective actions to be complete prior to startup and long term corrective actions incorporated into existing tracking programs. Backaround: During performance of a RCIC initiation logic test (8.H.2-2.10.11.1 LSFT) a combination of personnel errors, procedural deficiencies and a malfunction of the CK-1301-50 valve led to an overpressurization of the RCIC suction piping. A critique was completed to capture the immediately available evidence (Attachment 5). A series of 3 teams were created (see Attachment 6) to investigate this event. The Human Performance Evaluation System (HPES) Team was chartered with using a comprehensive INPO developed system to independently establish root cause and recommend corrective measures as a validation of the critique. The Check Valve Team was chartered with guiding the investigation of the performance of the CK-1301-50 valve and recommending corrective actions. The Oversight Team was made responsible for guiding the efforts of the aforementioned teams and maintaining appropriate perspective to ensure the event was thoroughly investigated. _ _ - _ _ - _ _ - _ _ _ _ _ - _ _ - _

- . _ _ - -

Oversight Connittee Report on the RCIC Pressurization Event of April 12, 1989 (continued) Page 2 of 6 Discussion: The HPES evaluation is complete. The report is included in this memo as Attachment 1. The report identifies four key causal factors:

Equipment tagging and second verification procedures were violated in performing the tagout. Procedural and administrative controls did not ensure independent

hanging and second verification of danger tags. The procedure validation process did not implement complete

validation of existing procedures. Additionally, the practice of l ' procedure validation by a job supervisor concurrent with the normal workload detracts from adequacy of the validation. Individuals with sufficient plant knowledge did not conduct

pre-evolution briefings to ensure that the full impact of the surveillance on the plant was understood by the key participants. The HPES Team recommended the following short term actions be completed prior to unit restart to address these concerns: Revise Procedure 8.M.2-2.10.11.1, RCIC High Water Level Turbine

Trip / Auto Restart Logic Test, Attachment A, Step 9.d to reflect Breaker D-774 as the correct isolation for MO-1301-49. (Action Complete) Revise the Danger Tag process to include placement of the device

number and equipment description on the Danger Tag consistent with , the information specified on the tagout sheet. (Action complete) Revise Procedure 1.4.5 (PNPS Tagging Procedure) to define the

requirements for what constitutes an independent verification. (Action complete) Provide specific instructions to the operators on how to perform an

independent verification. (Action Complete per revision to 1.4.5 and on watch training) Revise Procedure 8.M.2-2.10.11.1 to incorporate documented

verification of control panel indication to be consistent with the required system lineup after the isolation is performed. (Action Complete) Revise Procedure 8.M.2-2.10.11.1 to incorporate a signoff documenting

a pre-evolution briefing based on the criteria stated in Procedure 1.3.34, Conduct of Operations, Step 6.10. (Action Complete) Perform Procedure 8.H.2-2.10.11.1 prior to startup. (Planned)

_ _ _ - _ _ _ _ _ _ _ _ _ _ _

_ _ - - _ - _ _ _ _ _ _ .___ _ ___ ___-_______ _ _ Oversight Committee Report on the RCIC Pressurization Event of April 12, 1989 (continued) Page 3 of 6

Revise Procedure 1.3.34, Conduct of Operations, Step 6.10 to read: "Certain complex or infrequently performed activities warrant a pre-evolution briefing. The NOS shall perform a pre-evolution briefing with the key participants as close to the beginning of the I evolution as possible. Key participants unable to attend the group

briefing shall be briefed by the NOS prior to their involvement in the evolution. [1] Pre-evolution briefings shall be conducted for the following l evolutions:" ...(text from existing procedure) (Procedure 1.3.34 revised) The immediate actions recommended by the HPES Team are complete. The HPES Report also recommends longer term action items which will be sequenced once the outage is completed. The most significant of these items is the revalidation of I&C surveillance procedures prior to next use. The Check Valve Team utilized a Kepner-Treghoe like process to postulate potential check valve ma1 performance scenarios and developed appropriate methods to test these hypothesis. Attachment 2 summarizes the results of the CK-1301-50 testing team approach to date. Two temporary procedures were written and executed to extract data / evidence on-line. These tests indicated the check valve disk was probably intact, apparently not restrained by the exercise arm and apparently near/on the seat, the actual i leak tightness of the valve could not be determined on line. Nonetheless,

this information did serve to rule out 3 of the 5 scenarios postulated as described above. Additional testing has been performed to determine the actual leak tightness and as found condition of the valve now that the unit is in cold shutdown. Test 8.5.5.7 (Hydrostatic Backleakage) was sequenced before any other work to ensure an accurate as found leak rate is determined prior to disturbing any evidence. The valve was essentially leaktight with an "as-found" leakage rate of 0.0008 gpm. This test sequence was revised to open and inspect the valve after leak tightness measurements, irrespective of the results, to ensure the most thorough investigation possible. When the valve was opened, the disk was found to be in the closed position. However, significant resistance to disk motion was noted. The resistance was sufficient to prevent the valve from closing when the disk was released from the maximum open travel position. In addition, it was noted the disk would jump a few degrees open when the exercise arm was rotated to the closed and then back to the neutral position. When the internals were disassembled, foreign material appearing to be "Furmanite" residue was found on the hinge shaft / pin, disk arm and bushing. _-- __-_ _ ____ _ __ _- __ _ _ _ _ _

_ _ _ _ _ - - - _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ . - _ _ - _ _ _ _ _ . _ -- -. _ _ - _ - - _ _ _ _ - _ hversight Committee Report on the RCIC Pressurization Event of April 12, 1989 .centinued) lage 4 of G A record search revealed that this check valve packing gland had been Furmanited during the previous operating cycle, and per the process the The Check material was removed from the gland area during this outage. Valve Team and the valve vendor representative determined that this material in the shaft bushing to actuator shaft plug interface caused adequate friction to prevent the disk from closing without a force greater The same material (when present on the actuator than its own weight. shaft) would also account for the disk leaving the fully closed seated position when the exercise arm was returned to the neutral position. L Therefore, it is believed that the Furmanite residue caused the valve to hang slightly open following its cold shutdown exercise in January 1989. This probably resulted in the formation of an outboard hot water slug which flashed when depressurized. This flashing phenomena, in conjunction with whatever momentary backflow occurred before the valve checked . resulted in.the coolant discharge to the RCIC Quadrant. The similar check vala in HPCI (CK-2301-7) was opened and inspected. It was free to close consistent with the fact that it shows no sign of past furmaniting. Some cracking of the stellite surface of the disk was noted and is being dispositioned under the NCR process. The internals of the CK-1301-50 valve were replaced (except the disk which A preliminary records search has not identified any was reused). ' additional cases of check valves that have been Furmanited remaining -inservice (CK-2301-7 was Furmanited prior to replacement in RFO-6 and was Furmanited prior to retirement in place via cut and cap in CK-1001-64Additional consideration of the potential impact on internal RFO-7). components and the designation of appropriate confirmatory postwork testing-will be included in future leak stop (i.e., Fprmanite) applications. The Check Valve Evaluation Team recommended the following short term actions be completed prior to restart: A Temporary Procedure 89-39.

Implement procedural changes to ensure the isolation valve (1301-49

or 2301.-8) is closed after any RCIC or HPCI injection. (In progress) Perform a temporary procedure to demonstrate the RCIC injection check

valve forward flos capability and verification of closure following a vessel injection. (TP 89-39 ready to run) Long term recommendations have been requested in ESR 89-334 to evaluate; pipe insulation removal, appropriate design?, addition of vents / drains, prescription of post work testing / inspection following leak stop injections. !

___ __ i Oversight Committee Report on the RCIC Pressurization Event of April 12, 1989 (continued) i Page 5 of 6 A critical analysis of personnel performance during the event indicates j that there are other non-causal issues which warrant management attention. Control Room Operator attentiveness to changes to control j panel indication and Operator sensitivity to containment isolation valve status are issues which require improvement. Long term corrective action is required to address both issues. Control Room Design Review (CRDR) input would be helpful in addressing sensitivity to both. Simulator exercises need to be developed to reinforce Operator alertness. In the short term, procedural changes to require tagging Control Room panel switches when intentionally disabling equipment in the plant will facilitate Operator recognition of inappropriate position indication changes. A second dedicated R.O. will be provided to the Control Room crew during surveillance testing. The Nuclear Engineering Department was requested to evaluate the potential impact of the pressurization event on the RCIC System. Attachment 3 provides the results of this analysis. The analysis shows that the discharge piping system met the operability limits of IEB-79-14 and that the suction piping exceeded the limits by approximately 20%. Therefore, NED proposed a number of walkdown inspections of the system as well as specific NDE. The walkdowns are complete with no new additional findings beyond the instrumentation damage noted immediately following the event (PI-1360-20 and PS-1360-21 apparently overranged). The NDE is complete without abnormality except a 1/2" linear indication on the piping to PSV-31. This indication appears to have come from a weld buildup used during initial construction. The indication was ground out, rewelded and confirmed acceptable by radiography. Replacement of the damaged instruments were completed through the HR process. PS-1360-21 is obsolete (Robert-Shaw) and was replaced using Static-0-Ring equipment under PDC 89-24. NED was also asked to review the safety significance of this event with an eye towards Event-V or " Inter-System LOCA (NASH 1400)". Attachment 4 provides a summary of this review to date. The PNPS event should not be considered indicative of an Inter System LOCA (ISL) since an ISL by definition involves the discharge of reactor coolant through a non isolable inter system / Low pressure piping path outside containment. In the Pilgrim event, 3 means of isolation definitely existed (Feedwater Check Valve 58A, and RCIC HOV HO-1301-48 & 49. In addition, CK-1301-50 remained functional and constituted a fourth barrier or isolation means once the valve had checked. It should be noted that redundancy for isolating the RCIC Steam Line Containment Penetration was lost momentarily when the breaker for HO-1301-17 was tagged open with the valve open. HO-1301-16 remained operable and available to isolate this penetration in the event of a steam I line break. Fortunately, no such demand / transient occurred and the breaker to H0-1301-17 was promptly reclosed when the error was found.

Oversight Committee Report en the RCIC Pressurization Event Gf April 12, 1989 (continued) Page 4 of 6 A record search revealed that this check valve packing gland had been Furmanited during the previous operating cycle, and per the process the

material was removed from the gland area during this outage. The Check Valve Team and the valve vendor representative determined that this material in the shaft bushing to actuator shaft plug interface caused adequate friction to prevent the disk from closing without a force greater than its own weight. The same material (when present on the actuator ! shaft) would also account for the disk leaving the fully closed seated i position when the exercise arm was returned to the neutral position. Therefore, it is believed that the Furmanite residue caused the valve to hang slightly open following its cold shutdown exercise in January 1989. This probably resulted in the formation of an outboard hot water slug which flashed when depressurized. This flashing phenomena, in conjunction with whatever momentary backflow occurred before the valve checked, resulted in the coolant discharge to the RCIC Quadrant. The similar check valve in HPCI (CK-2301-7) was opened and inspected. It was free to close consistent with the fact that it shows no sign of past formaniting. Some cracking of the stellite surface of the disk was noted and is being dispositioned under the NCR process. The internals of the CK-1301-50 valve were replaced (except the disk which was reused). A preliminary records search has not identified any additional cases of check valves that have been Furmanited remaining inservice (CK-2301-7 was Furmanited prior to replacement in RFO-6 and CK-1001-64 was Furmanited prior to retirement in place via cut and cap in RFO-7). Additional consideration of the potential impact on internal components and the designation of appropriate confirmatory postwork testing will be included in future leak stop (i.e., Furmanite) applications. The Check Valve Evaluation Team recommended the following short term actions be completed prior to restart:

A Temporary Procedure 89-39.

Implement procedural changes to ensure the isolation valve (1301-49 or 2301-8) is closed after any RCIC or HPCI injection. (In progress)

Perform a temporary procedure to demonstrate the RCIC injection check valve forward flow capability and verification of closure following a vessel injection. (TP 89-39 ready to run) Long term recommendations have been requested in ESR 89-334 to evaluate; pipe insulation removal, appropriate design?, addition of vents / drains, prescription of post work testing / inspection following leak stop ' injections. I - - _ - _ _

- _ _ _ _ - _ _ - _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ - _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ _ _ _ - _ _ _ Oversight Committee Report en the RCIC Pressurization Event of April 12, 1989 (continued) Page 6 of 6 Conclusion and Action: Although each of these conditions has been addressed and there was no apparent risk to the health and safety of the public, the errors and programmatic deficiencies noted could have caused significantly greater problems under other circumstances. Therefore, this event should continue to be treated as significant. , The analysis is complete; corrective actions identified, immediate < corrective actions are complete or in progress, long term actions will be tracked (Attachment 7) and an effectiveness evaluation performed. - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ . _ - _ _ _ _ _ _ _ _ _ _ - _ _ _ - - _ _ _ _ - _ _ _ - - _ _ _ _ _ _ - _ _ _ _ _ __

, - ____ _ _ _ - _ - _ - _ _ _ - _ -- - j 3::sToN EDISON COLPANY ATTACHMENT 3 ! l 1 PROMPT REPORTABLE OCCURRENCE . TELECOPY MESSAGE ' TO . (215)i337-5324 ! DATE/ TIME: 9/30/83 / J 00 [4 - Telephone Number ' l l l ~ .- l . _ To : - l _ j FROM: Pilgrim Nucisar Power Station Regional Administrator, Region 1 RFDil, Rocky Hill Road i 17. S. lNuh Regulatory Commission Plymouth MA 02360 l 631 Park Avtaue l King of Prussia, PA 19406 \\ l' i i Docket Number 50-293 ' License DPR-35 Assigned LER Number _83-048/01X-0 Reported per T.S. Section 6.9.B. I .e , Event Deser'iptient On 9/29/83. while conductine a HPCI lorir nuvvo471nnen. m EPCI high suction pressure alarm and HpCT erea_ smoke __deterrne mia m were received in the control room. Investigation indicated an apparent over-pressurization of HPCI piping by feedwater had occurred. Cause and Corrective Action: At this time, the most probable cause is a leaking in-line check valve in ! he HPCI pump discharge line. HPCI was declared inoperable, t redundant' system surveillance initiated and investigation to determine root t, I < i cause commenced. Facility Status: 96 % Thermal MW , c) Reutine Startup g) Shutdown d) Routine Shutdown __ h) Rafueling ~ I e) Steady State x _ i) Other _ _ _ f) Loud Changing j) Not Applicable A written follevup report vill be sent within two weeks. NRC Person Ectified Resident insnector Prepared by C. G. Whitney f \\ ll l 4 ces Document Control Desk U.S. Nuclear Regulatory Commission Washington. D.C. 20555 . . , f4 ' StandardBECo.LERjDistribution il i;

, !. l

• I

' ! ! l l ll ! _ _ _ _ _ - . - - _ _ _ _ _- 1

_ - . _ _ _ - _ _ - _, - - _ _ _ .- - . - _ _ - = - . - - _ . . . - , <- - 50STON RDIgDN COMPANY 7 PILGRIM NUCLEAR POWER STATION ! i DOCKET No. 50-293 g y .; Attachment to LER 83-048/017-0 . I b ! On 9/29/83, during steady state operation, while conducting surveillance for . "HPCI steam Supply Isolation Valve Logie" and "NPCI Injection Valve Logic" controlir m annunciators were observed for "HPCI Turbine High Suetion Pressure" and "HPCI Turbine Cil Cooler Discharge Hi 011 Temperature". While the control room operator was alleviating the high suction pressure condition by openinglvalves in the HPCI test return line (~1 minute) the following alarms were mise observed: "250V Power Battery Cround", "23' El. Rx. illdg. West" (smoke alarm) and "HPCI Turbine Room" (smoke alarm). In- E vestigations into the cause of these alarms were immediately initiated with particular attention.to the smoke alarms. Thesmoklealarmawerefoundtohavebeeninitiatedbyvaporsfromheatedse ' i '

I of non-lagged HPCI suction pipe and the battery ground caused by water from a ruptured gland seal condensar gasket spraying a limit switch. All alarms were clearedtin a short time with the exception of the oil cooler discharge high oil temperat'ure annunciator. II Following these initial' investigations, a HPCI operability test was initiated. During the initial steps in the surveillance, the turbine stop valve failed to close ona remote nignal and at thi.e point HPCI was declared inoperable, re- dundant system surveillance initiated and the NRC notified. Further investigation indicated that a feedwater pressure transient had occurred to the HPCI suction piping through a partially open injection check valve (2301-7) whe. both HPCI pump discharge valves (2301-8 and 2301-9).were inadvertently valved open during the aforementioned logic tests. It was postulated that, if feedwater pressure was allowed through the open check valve (2301-7) and through both the;2301-8 and 2301-9 discharge valves through the pump and into the suction pipinr., the pump might not be capabis of producing its design pressure in the required time. At this time a Prompt Rcport was prepared and issued. An analysis was performed which conservatively assumed that the entire piping system had been subjected to 1100 peig. It was determined that none of the pip D g exceeded yield but had stayed in the elastic ragion. In addition to this analysis.,a system walkdown showed no visible damage to piping or supports. Therefore, it was concluded that the HPCI piping and supports were operable. Root cause of this event has been determined to be a combination of three items. 1) A failea coil for the turbine stop valve control unit which caused the HPCI system to belc'aclared inoperable. (It is believed this failure would have been found at the next scheduled HPCI surveillance.) 2) A miscommunication between station personnel which allowed both discharge valves to be opened at the samel time; 3) A' partially opened check valve which allowed a build-up of feedvater. pressure beyond its seatwhich created the instantaneous pressure transient-to flow through the inadvertently opened discharge valves. !! ! + The following corrective actions have taken place: 1) The coil was replaced in-kind and the stop valve returned to sarvice; 2) An investigation was conducted at the requot of the jVice President-Operations. The investigation revealed that root cause of the incident was verbal miscommunication between the control room operator landanI&Ctechnician. Instructions for verbal communications were not followed'. These instructions are being implemented at the station. 3) The suspect valve had been scheduled for replacement during the station wide valve betterment program.l l ! i i ' page 1 of 2 - _ _ _ _ _ _ - _ _ _ _ _ - _ _ _ _ _ _

_ ______ _ _ _ _ _ _ _ __ . .. ! Sul, equent investigations have determined that the check valve was partially op 3 , due to pressura equalization across the seat. When the 2301-8 valve va> opened the pressure spike occurred. The operator vented this pressure and, the check valve closed. An approved temporary procedure to test the valve was pertormed which monitored any build-up of pressure in the piping between:the check valve and the 2307-8 discharge valve. After approximately 16 hours no pressure build-up was detected. In addition, the postulated event previously described would not occur because, on a HPCI system initiation, as the pump begins to turn the 2301-8 valve is opening and t.he minimum flow bypass valve is also opening, therefore, if a pressore build-up is present downstream of the:2301-8 valve.it would be relieved through the bypass, the pump is then capable;of attaining its design function and the system operable. 7 l i , page 2 of 2 i }}