IR 05000361/1997015
| ML20210T330 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 09/10/1997 |
| From: | NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| To: | |
| Shared Package | |
| ML20210T281 | List: |
| References | |
| 50-361-97-15, 50-362-97-15, NUDOCS 9709150058 | |
| Download: ML20210T330 (66) | |
Text
.
.
MGLQMfilL2 U.S. NUCLEAR REGULATORY COMMISSION
REGION IV
Docket Nos.: 50 301 50-362 License Nos.: NPF 10 I NPF 15
>
Report No.: 50 301/97 15 50 302/97 15 Licensee: Southern California Edison Co.
Facility: San Onofra Nuclear Generating Station, Units 2 and 3 Location: 5000 S. Pacific Coast Hwy.
San Clemente, California Dates: June 30 through September 2,1997 Inspector: 1. Barnes, Technical Assistant Approved By: Arthur lil, Director Division of Reactor Safety ATTACHMENTS:
Attachment 1: Supplemental Information.
Attachment 2: Letter, Mr. G. T. Gibson to Mr. A. , dated July 22,1997.
Attachment 3: Licensee information presented at August 21,1997, open meeting.
Attachment 4: Licensee information presented at August 21,1997, open meeting.
9709150058 970910 PDR ADOCK 05000361 G PM
.
.
EXECUTIVE SUMMARY San Onofre Nuclear Generating Station, Units 2 and 3 NRC Inspection Report 50 361/97 15;50 362/97 15 Maintenance
- The San Onofre Nuclear Generating Station has experienced, as of the end of the
'
Cycle 9 refueling outages, cracking in 4 Unit 2 and 14 Unit 3 reactor coolant system inconel 600 nozzle penetrations. The available information suggested that l the mechanism for the cracking, with the exception of one nozzle, was primary l water stress corrosion cracking (Section E1.1).
_q
- Two heats of nozzle material, NX7630 and 9294, have exhibited confirmed cracking in four individual nozzle penetrations, indicating a high susceptibility of these heats to primary water _ stress corrosion cracking, in addition, three other
_
i Unit 3 Heat NX7630 nozzles were replaced during the Cycle 9 outage because ;
evaluation of inspection results indicated they were potentially cracked (Section E1.1).
- A violation was identified pertaining to the failure to consider the identification in 1995 of through wall cracking in four Unit 3 reactor coolant system nozzle penetrations, as of July 10,1996, before placing the Unit 3 reactor coolant system in a 10 CFR 50.65(a)(2) category. Licensee staff did not identify the four failures during performance history review and, as a result, also did not make the required entry of the information into the reactor coolant system maintenance order review matrix (Section M1.1).
Inconsistencies were noted in the Maintenance Rule program procedures with respect to the definition of what constituted a functional failure. For the reactor coolant system, the Maintenance Rule system description required a system loss for a failure to be considered a functional failure; whereas, Procedures SO123 XIV 5.3.2 and SO123 XV 5.3 discussed functional failures, respectively, as (1) the f ailure must have a system or train effect, and (2) the failure causes an inability of a system or train to perform one or more of its intended functions (Section M1.1).
Enaineerina ._
- The licensee engineering program requirements for inconel 600 reactor coolant system nozzle penetrations, which were contained in Document 90022. Revision 1, dated February 9,1995, stated that a progressive replacement program, based on susceptibility ranking, was the most important of the defined activities.
Document 90022, Revision 1, also stated, however, that there were no current-planned nozzle replacement activities due to the development of an in-house cepability to perform external nozzle repairs (Section E1.1).
.
.
3-
Actions to qualify the in house capability to perform external no:zle repairs were not initiated until the third quarter of 1996. No information was provided, as of the I June 30 through July 3,1997, onsite inspection, that would indicate a progressive replacement plan for nozzles had been approved for implementation (Section E1.1).
- An apparent violation of Criterion XVI of Appendix B to 10 CFR Part 50 was identified pertaining to the failure to implement actions to preclude recurrence of failures of Heat NX7630 nozzle penetrations (Section E1.1).
O
.
BetLPatch This inspection was performed in response to the identification during the Units 2 and 3 Cycle 9 refueling outages of recurrinti through wall cracking in reactor coolant system inconel 600 nozzle penetrations.
Summarv of Plant Statn Units 2 and 3 were, respectively, in a maintenance outage and refueling outage during the June 30 through July 3,1997, onsite portion of the inspection. Both units were at
,
approximately 100 percent power during the final onsite review performed August 4 8, I
1997.
II. Maintena,ma M1 Conduct of Maintenance An assessment of 10 CFR 50.65 (Maintenance Rule) implementation was conducted during August 4 8,1997, with respect to recurring through-wall cracking in reactor coolant system inconel 600 nozzle penetrations. Performance of this assessment was decided upon during in office review (see Section E1.1 for additional information) following the initial June 30 through July 3,1997, onsite inspection.
M1.1 Maintenance Rule implementation for Reactor Coolant System inconel 600 Nozzig Components a. Inspection Scone (92903)
The inspector reviewed Maintenance Rule program procedures, established performance criteria for the reactor coolant system, the Maintenance Rule system analysis report for the reactor coolant system, and it. formation provided to management regarding reactor coolant system trends and performance, b. Observations and Findinas The inspector noted from review of the functional criteria for risk significant systems contained in Procedure SO123 XIV 5.3.2, " Determination of Maintenance Rule Performance Criteria," Revision 0, Temporary Change Notice 01, that the applicable criteria for dispositioning the reactor coolant system from 10 CFR 50.65(a)(2) to 10 CFR 50.65(a)(1) were:
.
.
.
5-
Less than 98 percent system availability in the previou1 four quarters
Two, or more, repetitive functional failurcs in the previous 36 months
Three, or moro, random and unrelatad functional failures in tne previous 36 months
A footnote to the table containing the functional criteria identified that all functional i
f ailures must have a train or system effect to be counted against the applicable functional criteria.
I The inspector observed from review of failure history (see Table 1 in Section E1.1 below) that four functional failures (i.e., through wall cracks) had occurred in Unit 3 inconel 600 nozzle penetrations in 1995. It, thus, appeared that this failure history exceeded the licensee's defined limit of two repetitive functional failures in the previous 36 months. The licensee had, however, placed the Unit 3 reactor coolant system in a 10 CFR 50.65(a)(2) category, and not 10 CFR 50.65(a)(1), as of July 10,1996, the offective date for implementation of the Maintenance Rule.
Licensee staff informed the inspector that nozzle penetration cracking had not been treated in the past as a functional failure, if detected during either approach to or in a refueling outage. Such cracking was, however, considered a functional f ailure if detected while the unit was at power or in the process of return to power.
The inspector expressed disagreement to licensee personnel regarding this classification process, in that it appeared to the inspector to be both inconsistent and to focus primarily on unit unavailability considerations. Licensee staff respo' Jed to the inspector that the established performance criteria in the Maintenanco Rule system analysis report for the reactor coolant system (for comparable functional failures among similar critical components) was two cr more failures, which caused a loss of the system within 36 months. The inspecto?
informed licensee staff that this criterion appeared inconsistent with the definitions for functional f ailure contained, respectively, in Procedure SO123 XIV 5.3.2, Revision 0 through Temporary Change Notice 01, and Procedure SO123-XV 5.3,
" Maintenance Rule Program implementation," Revision 0 through Temporary Change Notice 0-4. Use of the terminology " loss of reactor coolant system" was also considered by the inspector to create the potential for interpretation differences between licensee staff on what circumstances represented a system loss.
The inspector ascertained from review of Procedure SO123 XV 5.3, Attachment 1, Revision 0 through Temporary Change Notice 0 4, that the licensee had defined the term " function" as " task performed by a system / train or component" and the term
- functional f ailure" as "f ailure that results in the inability of a system or train to perform one or more of its functions." The inspector agreed with these definitions and compared them against the system functions described in the Maintenanco Rule system analysis report for the reactor coolant system. One of the defined system functions was to " serve as a barrier to the direct release of radioactive materials."
.
.
6-Licensco staff were informed that through-wall cracks in the reactor coolant system pressure boundary were considered to be a functional f ailure in accordance with this definition, which was viewed by the inspector to be one of the commonly accepted basic functions of the rebetor coolant system.
l The inspector was provided during the August 4 B,1997, portion of the inspection I with a copy of two in process action requests, 970701233 and 970701234, which '
were initiated on July 23,1997. Action Request 970701233 Identified that the Unit 3 reactor coolant system exceeded its reliability performance criteria of less l than three functional failures and two repeat functional failures for the 12 quarters i
ending June 30,1997. The action request further stated that four repeat functional failures were observed on the hot leg during April 1997 due to inconel nozzle leakage indications, and that this observation reflected a more conservative failure criteria, which considered leakage indications during reactor downpowering as functional failures, Action Request 970701234 identified that the Unit 3 reactor coolant system was the significant contributor to the Unit 3 unplanned capability loss factor falling to meet its plant level criteria threshold of less than or equal to 2 percent, During review of the licensee Maintenance Rule program procedures, the inspector noted that paragraph 6.6.3 in Procedure SO123 XIV 5.3.3, " Preparation of-Maintenance Rule System Analysis Report," Revision 0, Temporary Change Notice 01, required each functionally critical component to be listed in the system analysis report. Functionally critical components were defined by paragraph 6.6.1 of Procedure SO123 XIV 5.3.3, Revision 0, Temporary Change Notice 0-1, as those components whose failure would directly cause a failure of the system or train, or substantially impair a system function. The inspector noted that reactor coolant system pressure boundary components were not listed in the reactor coolant
- system analysis report. Licenseo staff informed the inspector that the passive components were not included, because reliance was being placed on other existing programs to monitor performance (i.e., inservice inspection program and visual inspection program for nozzle penetrations). The inspector noted that the Maintenance Rule program procedures did not uniformly identify when other existing programs were being relied upon to monitor performance, or articulate what use would be made of the information generated. The inspector considered this subject to be one where program procedures could be strengthened.
The inspector observed during review of the maintenance order review matrix for the Unit 3 reactor coolant system that the 1995 through wall cracking in four Inconel 600 nozzle penetrations had not been entered in the matrix. The inspector ascertained from review of Procedure SO123 XV 5.3, Revision 0. Temporary Change Notice 0-4, that paragraph 6.5.1 indicated (with respect to monitoring risk significant structures, systems, and components) that site technical services reviews site specific informational sources on a routine basis to capture functional f ailures associated with the structures, systems, and components. Paragraph 6.5.1 also stated (in sub paragraph 6.5.1.3) that site technical services was to record the
.
.
action request and corrective maintenance order numbers and/or the event type and as much of the known information as possible for the failed structure, system, or component in the maintenance order review matrix. The inspector also noted that paragraph 6.9 in Procedure SO123 XIV 5.3.3, Treparation of Maintenance Rule System Analysis Report," Revision 0, Temporary Change Notice 01, required that each corrective maintenance order examined be listed in the maintenance order review matrix.
The i ,.ictor questioned licensee staff about the reasoris for the 1995 Unit 3 reacts coolant nozzle penetration cracking not being included in the maintenance ordet review matrix. Ucensee staff informed the inspector that they became aware of the oversight in mid July 1997. Initial review of the circumstances identified that nozzle repairs had been accomplished using construction work orders and not corrective maintenance orders. The screening of failure history that was performed
,as part of the Mr.intenance Rule program activities did not include construction work orders, since it was assumed by staff that the repairs were accomplished using corrective maintenance orders. Accordingly, the 1995 failures were not identified in the maintenance order review matrix. Licensee staff informed the inspector that an-action request was initiated to: (1) evaluate the circumstances, including human performance deficiencies, that resulted in the non-inclusion of the 1995 failures in the maintenance order review matrix for the reactor coolant system; and (2) provide corrective action recommendations. The inspector did not specifically review this action request.
It was concluded on August 28,1997, following review by an NRC panel, that a violation of 10 CFR 50.65(a)(1) and 10 CFR 50.65(a)(2) occurred. As of July 10, 1996, the time when the licensee elected not to monitor the Unit 3 reactor coolant system against licensee-established goals pursuant to 10 CFR 50.65(a)(1), the licensee f ailed to demonstrate that the condition of the Unit 3 reactor coolant system was being effectively controlled through the performance of appropriate preventive maintenance. Specifically, the licensee failed to consider Unit 3 reactor coolant system functional failures, which occurred previous to July 10,1996, before placing the Unit 3 reactor. coolant system under Section (a)(2). Through wall
_
cracking was identified in four Unit 3 reactor coolant system nozzle penetrations during 1995. This cracking, which represented multiple failures to meet one of the licensee defined reactor coolant system functions (i.e., serve as a barrier to the direct release of radioactive materials) was not evaluated against the established
- functional criteria (50 362/9715 02),
c. Conclusions A violation was identified pertaining to the failure to consider the identification in 1995 of through wall cracking in four Unit 3 reactor coolant system nozzle penetrations, as of July 10,1996, before placing the Unit 3 resctor coolant system-
..i a 10 CFR 50.65(a)(2) category. Licensee staff did not identify the four failures-during performance history review andi as a result, also did not make the required entry of the information into the reactor coolant system maintenance order review matrix. . inconsistencies were noted in the Maintenance Rule program procedures
.
.
with respect to the definition of what constituted a functional f ailure. In the case of the reoctor coolant system, the Maintenance Rule system description required a l reactor coolant system loss for a failure to be considered a functional failure; whereas, Procedures SO123 XIV 5.3.2 and SO123 XV.S.3 discussed functional-failures, respectively, in the context of: (1) the failure must have a system or train effect, and (2) the failure causes an inability of a system or train to perform one or more of its intended functions.
111. Enainstring E1 Conduct of Engineering i
E1.1 Review of Stress Corrosion Crackino in Reactor Coolant System inconel 600 Nozzle Comognants :
a. Inspection Scone (92EQ3J !
The inspector reviewed: (1) the licensee history of identified stress corrosion i cracking in Units 2 and 3 reactor coolant system inconel 600 nozzle components; (2) primary water halide and sulfate history; (3) licensee evaluations of NRC generic- !
communications pertaining to stress corrosion cracking of Inconel 600 alloys, '
(4) licensee actions taken to determine root cause; and (5) engineering program ,
requirements for identification and resolution of cracking problems in reactor coolant '
system nozzle penetrations.
b. Observations and Findinas inconel 600 Nozzle Penetration Stress Corrosion Crackina History The history of identified stress corrosion cracking in Units 2 and 3 Inconel 600 reactor coolant system nozzle penetrations, as of the end of the Units 2 and 3 Cycle 9 refueling outages is summarized in Table 1. Table 2 provides a listing both of the material heat numbers, which have exhibited stress corrosion cracking, and the respective numbers of nozzles from these heat numbers that have been replaced and remain in service.
The first inconel 600 nozzle penetration leakage (pressurizer steam space) occurred in Unit 3 in February 1986 after limited operational service (i.e.,10,200 hours0.00231 days <br />0.0556 hours <br />3.306878e-4 weeks <br />7.61e-5 months <br />).
The nozzle was removed and an inconel 600 replacement installed using the original internal J weld design. The licensee preventively replaced the remaining four installed Units-2 and 3 nozzles that had been manufactured from material from the .
f ailed heat (i.e., Heat ' 54318), af ter a subsequent root cause analysis concluded that the failure was caused by primary water stress corrosion cracking (see
'"Licenseo Actions to Determine Root Cause" below for additional information).
These replacements were performed during the Unit 2 Cycle 4 outage (pressurizer water space nozzle) and Unit 3 Cycles 3 and 4 outages (two pressurizer steam space and one pressurizer water space nozzles). All replacements utilized Inconel 600 nozzles and the original intemal J weld design,
.
.
Table 1 Units 2 and 3 Nozzle Penetration Failure History as of Cycle 9 Unit Failure Failed Hours' Heat YS'
Year Nozzle Penetration (k) Number ksi 2 1992 Pressurizer 2FO 01012 62 NX4411 38.5 1992 Pressurizer 2FO 01014 62 NX4411 38.5 1993 Hot Leg 2PDT00781 72 NX7630 38.5 1997 Pressurizer 2TE0101 63 K248 37.9 mummmmmmmemummmmmme mummmmmmmumnummmmmmummammmmmme
- )1 m -
3 1986 Pressurizer 3FO-01012 10.2 54318 60.9 1992 Pressurizer 3FO 01012 43 NXO571 51.6 1992 Pressurizer 3FO 0101 1 37 94758 53.0 1992 Pressurizer 3FO 01014 50 NX7630 38.5 1995 Pressurizer 3FO 0101 1 8 Weld N/A 1995 Pressurizer 3FO 01013 8 Weld N/A 1995 Hot Leg 3PDT0979-4 73 NX7630' 38.5 1995 Hot Leg 3TE01221 73 9294f 35.1 1997 Hot Leg 3TE0'i221 88 9294 35.1 1997 Hot Leg 3TE0121X2 88 9294* 35.1 1997 Cold Leg 3TE0121Y23 88 :9294L 35.1 1997 Hot Leg 3TWO138E 88 NX9915 37.8 1997 Hot Leg 3PDT0979 3 88 NX7630 38.5 1997 Hot Leg 3TWO138A 88 NX9915 37.8 1 Service hours (x 1000) at time of discovery of cracking.
2 Material yield strength.
3 Failure caused by f atigue and not primary water stress corrosion cracking.
.
.
'
'
During the Unit 3 Cycle 6 Refueling Outage in February 1992, evidence of leakage l was found in one pressurizer steam space nozzle, with two other steam space nozzles also identified te be cracked. The affected nozzles included the replacement (Heat NXO571) for the 1986 failed nozzle, the final remaining original steam space nozzle (Heat NX7630), and one of the Cycle 3 preventive replacements (Heat 94758). As a result of this identification, all four Unit 3 pressurizer steam space nozzles were replaced. Inconel 690 material was selected for the replacement nozzles due to it having greater resistance than inconel 600 to primary water stress corrosion cracking, inconel 82 filler material (which is a filler material normally used for welding inconel 600) was used for the Internal J welds for the internal J Welds, because filler materials with comparable composition to the inconel 690 base material had not yet received approval from the ASME Code.
In March 1992, evidence of leakage was detected during a shutdown in two Unit 2 pressurizer steam space nozzles (Heat NX4411). PartialInconel 690 nozzles were used to temporarily replace the failed Inconel 600 nozzles, with the nozzle attachment weld made to an inconel weld buildup on the exterior of the pressurizer.
During the subsequent 1993 Unit 2 Cycle 7 refueling outage, all four pressurizer steam space nozzles were replaced with fullinconel 690 nozzles using an internal J weld design and comparable composition inconel 52 filler material.
During the Unit 2 Cycle 7 Refueling Outage in June 1993, the first instance of leakage at a hot leg nozzle (i.e.,2PDT09781) was detected. This nozzle had been manufactured from Heat NX7630, thus, representing the second f ailure of this material heat in Units 2 and 3. The nozzle was replaced with a partialinconel 690 nozzle, with the attachment weld made to a weld buildup pad on the piping outside diameter surf ace. This replacement was the first licensee use of a partial Inconel 690 nozzle with an external weld as a permanent design change.
During the Unit 3 Cycle 8 refueling outage in June 1995, evidence of leakage was found in two of the four pressurizer steam space nozzle penetrations that had been installed in 1992. The licensee information indicated that the leakage resulted from the presence of cracks at it.e J weld to buttering interface. No specific information was seen by the inspector that would confirm whether the cracking was related to welding deficiencies, during the 1992 installation, or was caused by primary water stress corrosion cracking. All four Unit 3 pressurizer steam space ne'zles were again replaced with full Inconel 690 nozzles, using internal J. welds and Inconel 52 comparable composition filler material, in addition to the pressurizer nozzle cracking detected during this outage, two instances of leakage at hot leg nozzles were identified (i.e., Nozzles 3PDT0979-4 and 3TE01221). Hot leg Nozzle 3PDT0979 4 was manufactured from Heat NX7630, thus, representing the third failure of this material heat in Units 2 and 3. Nozzle 3TE01221 was manuf actured from Heat 9294, which was the first failure in either unit of this material heat.
.
11-In March 1997, a steam plume was observed emitting fr .n the Unit 2 pressurizer niid shell instrument Nozzle 2TE 0101. At the time of detection, Unit 2 was in .
'
Mode 4 and in the process of retuming to power from the Cycle 9 refueling outege.
The failed nozzle (Heat K240) was a 1987 Inconel 600 preventive replacement of l the odginal f ailure heat, Heat 54318. The nozzle was replaced with a partial inconel 690 nozzle, with the attachment weld made to an external weld pad.
During the Unit 3 Cycle 9 refueling outage in April 1997, inspection of the reactor coolant system nozzle penetrations iound evidence of leakage in five nozzles. Four
!
of the nozzles (i.e.,3PDT0979-3,3TE0121X2,3TE0122 2, and 3TWO138E) were hot leg nozzles, with the remaining f ailure, Nozzle 3TE0121Y2, located in the cold leg Nozzle 3PDT0979 3 was ascertained by the inspector to have been manufactured from Heat NX7630, thus, representing the fourth failure of this material heat in Units 2 and 3. (See below in " Engineering Program Requirements for identification and Resolution of Cracking Problems in Reactor Coolant Systern Nozzle Penetrations," for additional discussion on the recurring failures of Heat NX7630). The cold leg nozzle,3TE0121Y2, and two of the hot leg nozzles, 3TE0121X2 and 3TE0122 2, were manuf actured from Heat 9294, thus, bringing the total f ailures of this material heat to four, allin Unit 3. The inspector considered a f ailure mechanism of primary water stress corrosion cracking to be unlikely for the cold leg nozzle due to the low service temperature (i.e., approximately 553' F). No
, information was received from the licensee that specifically addressed the probable f ailure mechanism for this nozzle. During review of licensee root cause evaluations (see " Licensee Actions to Determine Root Cause," below), the inspector learned, however, that another cold leg nozzle penetration.was evaluated for root cause after being found to be leaking in 1996 through the threads of a plug that had been installed in the nozzle (3TE9179 3) thermowell. This evaluation determined that tne thermowell had failed by fatigue, with the failure related to improper installation of the thermowell in the nozzle during vessel fabrication. The fif th nozzle showing evidence of leakage was hot leg Nozzle 3TE0138E, This nozzle was manufactured from Heat NX9915 and was the first failure from this material heat. All five nozzles were replaced with partialInconel 690 nozzles, with the nozzle attachment welds made io an external weld pad.
The inspector was informed by telephone on July 7,1997, subsequent to the onsite inspection, that a walkdown inspection detected leakage in Unit 3 hot lug Nozzle 3TWO138A, The walkdown inspection was performed while the reactor coolant system was at 350 psia and the unit in the initial stages of preparation for return to power following the conclusion of the refueling outage.
Nozzle 3TWO138A was manufactured from Heat NX9915, thus, representing the second f ailure (both in Unit 3) of this material heat. Licensee evaluation of prior walkdown inspectiore results from the Cycle 9 refueling outage identified that a deposit had been previously observed associated with Nozzle STWO138A, for which isotopic analysis had not confirmed the presence of an active leak. This evaluation of Cycle 9 results also identified three other nozzles, all manufa:tured from Heat NX7630, with deposits present and similar inspection results to
_ __ _
.
.
12-Nozzle 3rWO138A. As a result of this review, the licensen made the determination t& u. pair the th;ee potentially cracked norzies in additiot, to Norrle 3TWO138A.
The in:cector did not specifi e:Ir verify the repair method used; however, the norile locaisas v>ould nurmally dictate use of a part;alInconel 690 nozzle and the I attachroent v, eld made to a>1 ceraal weld ped.
Table 2 Units 2 and 3 NozzM Penetration Replacement History as of the end of Cycle 9 I,
Failure Heat Failure Oty. No. Replaced No. Remaining Unit 2 Unit 3 Unit 2 Unit 3 Unit 2 Unit 3
. av 54318 0 1 1 4 0 0
_
l 94758 0 1 0 4 0 0 NX4411 2 0 2 0 2 O NX7630 1 3 1 6' 9 /
NXO5'/1 0 1 0 1 0 0 9294 0 4' O 4 5 18 K248 1 0 1 0 0 0 NX99 5 0 l 2 0 2 12 5 1- Three NX7630 nozzles were repaired during the Unit 3 Cycle 9 outage because of visual inspection similarities to a NX9915 nozzle that was found to be leaking during plant restart from Cycle 9.
2- Failure mechanism of one of four cracked 9294 nozzles was f atigue and not primary water stress corrosion cracking.
Primarv Water Halid_e and Gulf ate @lgty The inspector compared licensee Units 2 and 3 historical primary water chemistry data against the requirements of Licensee Controlled Specification 3.4.101,
" Chemistry," Revision 1, and Procedure S01231111.123, " Units 2/3 Chemical Control of Primary Plant and Related Systems," Revision 30 and Temporary Change Notice TCN 30-1. In general, the dhta showed excellent conformance to the specification and procedure raquirements. The inspector noted no halide or dissolved oxygen data that would suggest a possible relationship between primary water chemistry history and the observed incidence of reactor coolant system nozzle cracking. The inspector also reviewed available sulf ate chemistry information
.
.
- 13-against the criteria contained in Electric Power Research Instituto Document TR 105714. "PWR Primary Water Chemistry Guidelines," Revision 3. No elevated sulf ate values, which are a known contributor to intergranulcr stress corrosion cracking of inconel 000 alloys, were seen during this review. The inspector concluded that the sulf ato data provided by the licensee did not indicate any correlation with the observed incidence of reactor coolant system nozzle cracking.
Licensee Eva.lpation of Generic Communications Pertaining _to Stress Corroll0D l .Qrackina of Inconel 60Q The inspector requested licensee evaluations of generic communications pertaining to stress corrosion cracking of Inconel 600 materials. The licensee provided in response a number of evaluations of industry communications and an evaluation of Information Notice 9010, " Primary Water Stress Corrosion Cracking (PWSCC) of inconel 600." The majority of the evaluations pertained to degradation issues related to reactor vessel components; therefore, the inspector restricted review to the evaluations of Information Notice 90-10 and Combustion Engineering Technical Bulletin 89 06, * Pressurizer Heater Sleevo Leakage." The evaluations provided i limited infortnation, essentially indicating compliance with Combustion Engineering recommendations and participation in and reliance on Combustion Engineering !
Owners Group program activities for addressing the degradation. ;
Licensee Actions to Determine Root Cause The licensee furnished two root cause reports in response to a request from the inspector for information on root cause analyst that had been performed in response to identified through wall cracking in inconel 600 reactor coolant system nozzle penetrations. '
Report RCE 92 019, " SONGS 3 Pressurizer LevelInstrument Nozzle Leakage," dated June 19,1992, pertained to the 1992 identification of through wall cracking in one Unit 3 pressurizer steam space nozzle penetration and cracking in two others. The cracks were revealed by liquid penetrant inspection to be located in the vicinity of the J weld. All four steam space nozzles were removed and the remnants subjected to laboratory examination. The removat process destroyed the cracked portions of the nozzle penetrations, thus, limiting the scope of laboratory examinations that could be performed.
Visual examination of the inside diameter surfaces of the four removed nozzles identified rough surf ace irregularities introduced during machining and f abrication of thn aozzles. Metallographic examination found a cold worked layer at the inside diameter surf ace of the nozzles that had been created during machining, with an average depth of 0.002 inches, in the nozzle that had exhibited leakage (Heat 94758) a shallow crack was found coincident with the cold worked zone.
The examination concluded for this nozzle that the in service through wall cracking
.
.
..
14-was mostly due to primary water stress corrosion cracking and perhaps initiated by the cold worked inside diameter r,urface layer. The inspector noted that the metallographic examination datected intergranular cracking in the remaining portion of one of the other two cracked nozzles (i.e., Heat NX7630), which was mostly due to primary water stress corrosion cracking. The Heat NX7630 nozzle microstructure also exhibited carbide segregation in bands, which suggested to the licensee investigator that the cracks found by liquid penetrant inspection in the vicinity of the J weld may have initiated from a high density carbide area. The inspector considered that the presence of shallow intergranular cracking (remote from the more highly _ stressed material adjacent to the J. weld) indicated that the Heat NX7630 material was highly susceptible to primary water stress corrosion cracking. The laboratory results were, thus, considered to provide some explanation for the subsequent failure history of this heat.
Report RCE 97 001, "RTD 3TE 9179 3 Thermowell Cracking," dated May 28, 1997, pertained to the September 1996 detection of a leak through the threads of a plug that had been installed in a thermowell at the Resistance Temperature Detector 3TE9179 3 location. Upon cutting the weld that joins the thermowell to the nozzle penetration, the licensee found that only the upper part of the thermowell was present and the tubular section (3/8 inches diameter,9.87 inches long) was missing. The remaining portion of the thermowell was submitted for root cause evaluation. Laboratory examination identified the failure mode as fatigue fracture due to bending. Rubbing damage was also present on the bottom of the large diameter section of the thermowell, indicating contact with the no;zle. . As a result of the design drawings specif) og a 1/8 inch gap between the nozzle and the bottom of the large section of .he thermowell, the licensee concluded that the thermowell was instal led improperly during original plant construction. The licensee's review identified that the particular '1ozzle location had a history of difficulty installing resistance temperature detectors. The licensee postulated that the most likely root cause was that manipulation of the resistance temperature detector in the thermowell during previous refueling outages created cold work of the inconel at the sectional transition and initiated a crack. The crack further propagated via fatigue, as a result of flow passing across the protruding thermowell, resulting in the final fracture. The inspector concurred with the licensee root cause evaluation, with the exception of manipulation of a resistance temperature detector in a thermowell being expected to initiate a crack at the section transition. The inspector considered that the stresses originating from the improper thermowell installation, in conjunction with plastic deformation at the thermowell section transition caused by resistance temperatute detector manipulation, could be expected to initiate and propagate a fatigue crack. Overalllicensee actions to bound the scopb of the problem were considered satisf actory.
.
.
15-Enaineerina Proaram Requir9mpnts for identification and Resolution of Crackinn Problems in Reactor Coolant System Norrlo Penetrations The inspector reviewed Document 90022 " Susceptibility of Reactor Coolant System Alloy 600 Nozzles to Primary Water Stress Corrosion Cracking and Replacement Program Plan," Revision 1, to ascertain the licensee's program ,
requirements for addressing primary water stress corrosion cracking in inconel 600 '
components. The inspector noted that the licensee had ranked the susceptibility of the various nozzle locations to primary water stress corrosion cracking, and mandated an inspection for potential leaks each refueling outage. The inspection included, where applicable, an evaluation by chemistry of deposits to determine whether the leakage was from the reactor coolant system and the approximate time of initiation.
Section 7.0, " Current Plan," of Document 90022, Revision 1, discussed the necessity of integrating the inspection procedure, planned replacement activities, and contingency repair plans. A progressive replacement program, based on the susceptibility ranking, was stated to be the most important of the three activities.
The inspector noted, however, that the text in Section 7.0 of Document 90022, Revision 1, then identified that there were no current planned nozzle replacement activities due to the development of an in house capability to perform exterior nozzle repairs. Document 90022, Revision 1, was dated February 9,1995. The inspector was informed by welding engineering staff that development of in house capability was discussed for a considerable period, but fornd qualification of the welding procedures did not begin until about the third quarter of 1996. The partial Inconel 600 nozzle replacements performed during the Units 2 and 3 Cycle 9 refueling outages were the first totally performed by licensee staff. No information was provided to the inspector that would indicate approval had been received, as of the June 30 through July 3,1997, onsite inspection, with respect to implementation of the progressive replacement program discussed in Document 90022, Revision 1.
The inspector considered that recurring leakage (resulting from development of through wall cracks)in inconel 600 reactor coolant system pressure boundary nozzle penetrations represented a significant condition adverse to quality, and for which Criterion XVI of Appendix B to 10 CFR Part 50 would require corrective action to be taken to preclude recurrence. Of specific concern to the inspector were the four confirmed failures discussed above in nozzle penetrations manufactured from Heat NX7630 that had occurred during 1992 1997. No actions had been taken by the licensee to preclude failure recurrence despitJ the high cracking susceptibility demonstrated by the performance history of the material heat.
A licensee letter dated July 22,1997, pertaining to the licensee position on a violation of Criterion XVI of Appendix B to 10 CFR Part 50 was subsequently received in the Region IV office. This letter, which was received subsequent to a
.
.
- 16, recommendation b) Regional management to review the nozzle cracking circumstances in the context of implementation of the Maintenance Rule, has been included as Attachment 2 to the inspection report. As noted in Section V below, licensee management indicated a willingness to provide a licensee position on its compliance with Criterion XVI of Appendix B to 10 CFR Part 50 in regard to reactor coolant system norrie penetration cracking. The inspector performed a review of this letter and made the following observations:
- The absence of specific quantitative information in the letter relative to the range of actual reactor coolant system nozzle penetration flaw sizes noted in Units 2 and 3 precluded any judgement of potential for unstable crr.ck growth, i
l * The letter indicated that NRC and industry had agreed that there was no
! safety significance to primary water stress corrosion cracking, which would justify replacement of nozzles. The inspector was unaware of any NRC ogreement on this subject.
The inspector noted from review of NRC communications on the subject of primary water stress corrosion cracking, that ste'f views that axial cracks were not considered a significant threat to structural integrity were accompanied by concerns regarding circumferential cracking. For exarnple, Information Notice 9010 documented that limited circumferential cracking was reported in the instrument nozzlus of several foreign reactors, and then noted that circumferential cracking poses a more serious safety concern (than axial) because, if it were to go undetected, it could lead to a structural f ailure. No information was provided in the July 22,1997, letter that provided a basis for excluding circumferential cracking as a possibility.
The letter articulated a position that primary water stress corrosion cracking was not a significant condition adverse to quality based primarily on the following arguments: only axial cracking experienced, slow crack propagation rates, slow increases in leakage rates, and ability to detect cracks before they reach an unstable size. The inspector did not accept this position, in that it appeared to imply a significant structural integrity concern must exist before classifying degradation as a significant condition adverse to quality. Recurring through wall cracking in the reactor coolant system pressure boundary, which permits pressure boundary leakage (a condition prohibited by Technical Specifications while in Mode 1), was considered to fully meet a reasonable interpretation of a significant condition adverse to quality, it was concluded on August 28,1997, following review by an NRC panel, that the f ailure to implement actions to preclude recurrence of f ailures of Heat NX7630 nozzle penetrations was an apparent violation of Criterion XVI of Appendix B to i
10 CFR Part 50 (50 362/9715-01).
_ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ - -
.
17-l l
l c. Conclusions The San Onofre Nuclear Generating Station has experienced, as of the end of the l
'
Cycle 9 refueling outages, cracking in 4 Unit 2 and 14 Unit 3 reactor coolant system inconel 600 nozzle penetrations. The available information suggests that l the mechanism for the cracking, with the exception of one nozzle, was primary water stress corrosion cracking. Two heats of nozzle material, NX7630 and 9294, have each exhibited confirmed cracking in four individual nozzle penetrations, indicating a high susceptibility of these heats to primary water stress corrosion cracking. In addition, three other Unit 3 Heat NX7630 nozzles were replaced during the Cycle 9 outage because evaluation of inspection results indicated that they l were potentially cracked. The licensee engineering program requirements for
- inconel 600 reactor coolant system nozzle penetrations, which were contained in l Document 90022, Revision 1, dated February 9,1995, stated that a progressive replacement program, based on susceptibility ranking, was the most important of the defined activities. Document 90022, Revision 1, also stated, however, that there were no current planned nozzle replacement activities due to the development
, of an in house capability to perform external nozzle repairs. The capability was not l accomplished until late 1996, and, as of the June 30 through July 3,1997, onsite inspection, no information was provided that a progressive replacement plan had been approved for implementation. An apparent violation of Criterion XVI of l
Appendix 0 to 10 CFR Part 50 was identified pertaining to the failure to implement
- actions to preclude recurrence of failures of Heat NX7630 nozzle penetrations.
!
V. Manaatment Meetinas X1 Exit Meeting Summary
,
The inspector presented the results of the initial onsite inspection to members of l licensee management on July 3,1997. Management was informed that a potential l violation of Criterion XVI of Appendix 8 to 10 CFR Part 50 had been identified pertaining to the failure to preclude recurrence of primary water stress corrosion cra:: king f ailures in inconel 600 reactor coolant system nozzle penetrations that had been manufactured from Heat NX7630. The licensee Vice President of Engineering responded indicating that the utility view was that it was in full compliance with l 10 CFR Part 50, Appendix B, and that the utility was willing to furnish a letter supporting that position to the Region IV office.
!
l A meeting, which was made open to attendance by the public, was held with l licensee staff in the Region IV office on August 21,1997, to provide an opportunity for the licensee to present its views on its Maintenance Rule program criteria for the reactor coolant system. Presentations made by licensee staff during this meeting have been included in the inspection report as Attachments 3 and 4.
l l
l
_ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - _ _ _ _ _ _ - _ _ _ _ _ - _ _ _ ______ _ _ _ _ _-
.
.
It was finally concluderi, following review by an NRC panel, that the f ailure of the licensee to implement actions to preclude recurrence of Heat NX7630 nozzle penetration cracking was an apparent violation of Criterion XVI of Appendix B to ,
10 CFR Part 50. It was also concluded that the placin0 of the Unit 3 tractor I coolant system in a 10 CFR 50.65(a)(2) category as of July 10,1996, despite the l identification in 1995 of through wall crackin0 in four reactor coolant system norrle penetrations, was a violation of 10 CFR 50.65 (a)(1) and 10 CFR 50.65(a)(2). The licensee was informed of the inspection findin0s in a telephonic exit meeting conducted on September 2,1997, i
.
.
ATTACHMENT 1 SUPPLEMENT AL INFORMATION PARTIAL LIST OF PERSONS CONTACTED Licenseu l
R. Allen, Supervisor, Site Technical Services D. Axline, Compliance Engineer, Nuclear Regulatory Affairs D. Breig, Manager, Station Technical J. Clark, Manager, Chemistry R. Clark, Manager, Quality Engineering, Nuclear Oversight Division C. Coker, Supervisor, Design Process, Nuclear Engineering and Design Organization G. Gibson, Manager, Compliance, Nuclear Regulatory Affairs K. Knight, Nuclear Construction Supervisor, Nuclear Construction R. Krieger Vice President, Nuclear Generation D. Nunn, Vice President, Engineering & Technical Services G. Plumlee, Supervisor, Compliance, Nuclear Regulatory Aff airs S. Root, Supervisor, independent Plant Review Engineering. Nuclear Engineering and Design Organization M. Short, Manager, Site Technical Services K. Slagle, Manager, Nuclear Oversight Division M. Wharton, Manager, Design Engineering, Nuclear Engineering and Design Organization K. Wright, Engineer, independent Plant Review Engineering, Nuclear Engineering and Design Organization Nf1C J. Kramer, Resident inspector J. Russell, Resident inspector J. Sloan, Senior Resident inspector INSPECTION PROCEDURES USED IP 92903 Followup Engineering ITEMS OPENEQ Onened 50 362/9715 01 APV Faiiure to implement actions to preclude recurrence of failures of Heat NX7630 nozzle penetrations (Section E1.1)
50-362/9715-02 VIO Incorrect 10 CFR 50.65 classification for Unit 3 reactor coolant system (Section M1.1)
.
.
2-l l
LIST OF DOCUMENTS REVIEWED hecadures/ Documents Root Cause Report RCE 92 019, " SONGS 3 Pressurizer LevelInstrument Nozzle Leakage,"
dated June 19,1992 Root Cause Report RCE 97 001, "RTD 3TE 9179 3 Thermowell Cracking," dated May 28, 1997 Safety Assessment Report SEA 97 002, " Assessment of the inconel 600 Nozzle inspection
& Replacement Program and Decision Process Lt ading to the March,1997 Nozzle Failure in Unit 2," dated April 9,1997
' Electric Power Research Institute Report TR luo714, "PWR Primary Water Chemistry Guidelines," Revision 3 Electric Power Research Institute Report TR 104030, "PWSCC Prediction Guidelines," July 1994
Procedure SO123 V Ill 1.123, " Units 2/3 Chemical Control of Primary Plant and Related Systems," Revision 30 through Temporary Change Notice 301 Units 2 and 3 primary water chemistry trend cata Document 90022, " Susceptibility of Reactor Coolant System Alloy 600 Nozzles to Primary Water Stress Corrosion Cracking and Replacement Program Plan," Revision 1 Procedure SO23 V 8.16, ." Reactor Coolant System inconel Nozzle Inspection," Revision 0 Inspection records generated during performance of Procedure SO23 V-8.16, Revision 0, inspections of inconel 600 nozzles Combustion Engineering Owners Group Task 034 "inconel 600 Primary Pressure Boundary Penetrations," dated January 1991 Licensee evaluhtlons of: (a) Information Notice 8418, " Stress Corrosion Cracking in Pressuilzed Water Reactor System"; (b) Information Notice 9010, " Primary Water Stress Corrosion Cracking (PWSCC) of Inconel 600"; (c) Information Notice 9611. " Ingress of Demineralizer Resins increases Potential for Stress Corrosion Cracking of Control Rod Drive
' Mechanism Penetrations"; (d) CE Technical Bulletin 89 06, " Pressurizer Heater Sleeve
-
- Leakage"; and (e) Westinghouse _ Letter 920124 509, " Reactor Vessel Head Adaptor Tube Cracking" Industry Document NUMARC 93 01, " Industry Guideline for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants," Revision 2
.
(
-3
'
Pr9cedure SO123 XIV-5.2.1, " Reliability, Failure and Availability Analysis," Revision O through Temporary Change Notice 0 2 Procedure SO123 XIV-5.3.?, " Determination or Maintenance Rule Performance Criteria,"
Revision O through Temporary Change Notice 0-1 Procedure SO123 XIW5.3.3, " Preparation of Maintenance Rule System Analysis Report,"
Revision O through Temporary Change Notice 01 Procedure SO123 XIV 5.3.6, " Goal Setting for the Maintenance Rule," Revision O through Temporary Change Notice 0-2 Procedure SO123 XIV 5.3.7, " Maintenance Rule Periodic Assessment," Revision O through Temporary Change Notice 01 Procedure SO123 XV 5.3, " Maintenance Rule Program implementation," Revision 0 through-Temporary Change Notice 0 4
.
.
. . . . . . . . . . . . . .. .
.
.
ATTACHMENT 2 SOUTHERN CAllFORNIA EDISON July 22,1997 Letter Mr. A. ill from Mr. G. T. Gibson
. . -
.
.
.. .
. . .
_.
.
$O0T HERN C Allf 0RN14 e EDISON An [DiSO41.W[R \ t t10\ 4L Cewr eny July 22,1997 Mr. A. Ill Director, Division of Reactor Safety U. S. Nuclear Regulatory Commission 611 Ryan Plaza Drive, Suite 400 Arlington, Texas 76011-8064
Dear Mr. Howell:
Subject: Docket Nos. 50-361 and 50-362 Alloy 600 Primary Water Stress Corrosion Cracking San Onofre Nuclear Generating Station, Units 2 and 3 On July 3,1997, the NRC held a primary water stress corrosion cracking (PWSCC)
inspection exit meeting (inspection Report 97-15). The inspector identined a potential violation of 10 CFR 50, Appendix B, Criterion XVI, " Corrective Action." The inspector indiated that the potential violation was for failure to take corrective actions on Alloy 600 heat number NX7630 RCS nozzles, after PWSCC was identined on two NX7630 hot leg nozzles in 1995. Edison does not believe a violation occurred. This letter provides additional information in this regard. Also, Edison requests a meeting, if necessary, to discuss this additional information.
Criterion XVI requires conditions adverse to quality be ". . . promptly identified and corrected." Criterion XVI also requires that for significant conditions adverse to quality
". . . the cause of the condition be determined and corrective actions taken to preclude repetition." In accordance with Criterion XVI, Alloy 600 PWSCC (see Attachment)
was identified and documented on nonconformance reports, and corrected by repair or replacement.
PWSCC, which can result in tiny, slowly developing leakage from the RCS, is not a significant condition adverse to quality, which is supported by the following factors:
1) The combination of plant specific and industry experience and stress analysis shows PWSCC grows axially. Axial cracking is not a significant threat to the nozzle stmetural integrity, and will result in a small, detectable leak.
2) PWSCC axial cracks proceed gradually, and slowly increasing leakage permits plant _ personnel to detect and react to the leakage.
P. O Box 128 San Clemente. CA 92674 0128-i
.
.
,
A. July 22,1997 3) PWSCC axial cracks may grow to a length of two inches axially without exhibiting unstable crack growth. A Combustion Engineering owners group report (CEOG Task 700) fracture mechanics analysis showed that even with a crack two inches long, at normal steady state temperatures and pressures, there is a substantial safety factor (> l 0) against any additional crack growth due to mechanical means.
4) Crack detection occurs well before it can grow to an unstable size. Industry and SONGS experience was consistent with this expectation.
5) Independent industry sources concluded there was no immediate safety concern.
NRC Information Notice 90-10 " Primary Water Stress Corrosion Cracking (PWSCC) of Alloy 600, " dated February 23,1990.
-
NUMARC letter to William T. Russell (NRC) regarding inconel 600, dated June 16, 1993.
-
NRC letter from William T. Russell to William Rasin (NEI), NRC response to NUMARC letter, dated November 19, 1993.
-
NUMARC letter from Alex Marion to the NRC, response to previous letter, dated January 31, 1994,
-
EPRI TR-103696, "PWSCC of Alloy 600 Materials in Primary System Penetrations," dated July 1994.
-
NUREG CR-6245, " Assessment of Pressurized Water Reactor Control l Rod Drive Mechanism Cracking," dated October 1994.
In the absence of SONGS, industry, or NRC criteria establishing a threshold for replacing an entire heat of material once a failure occurs, the replacement decision must consider all potential factors (e.g., safety significance of condition, ALARA, risk factors for the execution of work, etc.). For example, from an ALARA perspective, the dose rate associated with each hot leg nozzle replacement is approximately 1.7 person-rem per RCS nozzle, and 35 to 40 person-rem would have been expended to replace the remaining NX7630 nozzles.
Because the NRC and industry agreed there was no safety significance to PWSCC which would justify replacement of nozzles, Edison concluded that replacement of nozzles from heat number NX7630 was not necessary at that time and proceeded to:
1) qualify a replacement process in house to minimize plant impact: 2) explore other
A. T. Ilowell -3- July 22,1997 f
alternate replacement processes; 3) continue the inspect and repair program through Cycle 9 (current outage); and 4) reevaluate our program plan following the Cycle 9 outages.
The inspect and repair program since 1995 has resulted in the detection of PWSCC at San Onofre well before any significant leakage developed (e.g., the recent 3TWOl38A nozzle PWSCC found while coming out of the Cycle 9 refueling outage). Edison is assessing the Cycle 9 outage PWSCC inspection results. The 1997 Nuclear Organization Business Plan initiative BP-4-4-1, dated January 15,1997, states that a revised Inconel Strategic Plan will be issued by October 31,1997. This letter constitutes a formal Edison commitment to the NRC to complete this plan by October 31, 1997.
If a meeting to further discuss PWSCC with you and your staff would be beneficial, please call me to arrange a meeting. Edison appreciates the opportunity to provide . tis additional information.
Sincerely, G. T. Gibson Manager, Compliance ec: J. A. Sloan, NRC Senior Resident Inspector, San Onofre Units 2 & 3 M. B. Fields, NRC Project Manager, San Onofre Units 2 & 3 NRC Document Control Desk
i NITACllNIENT
<
,
Alloy 600 NX7630 Nozzle Ilistory I -
In 1992 during the Unit 3 Cycle 6 outage, the first indication of PWSCC associated with heat number NX7630 was identified on a pressurizer upper shell nozzle. Edison concluded that all pressurizer upper shell nozzles were the most susceptible to PWSCC, based on the high temperature of the location (653 degrees F). These nozzles were all replaced with a less PWSCC susceptible material (inconel 690).
-
In June 1993, during the Unit 2 Cycle 7 outage, a second nozzle associated with heat number NX7630 evidenced indications of PWSCC. This was the first hot leg nozzle indication of PWSCC. This nczzle was replaced with inconel 690.
In July 1995, a second NX7630 PWSCC hot leg leak was identified during the
-
Unit 3 Cycle 8 outage. The nozzle was replaced with an inconel-690 replacement.
-
In October 1995, another Combustion Engineering nuclear power plant experienced PWSCC in one nozzle involving heat number NX7630 (this was the only other industry recorded PWSCC for heat number NX7630).
Table I below represents the total hot leg nozzle population, and nozzle PWSCC by heat number, through the SONGS Cycle 8 outages. Subsequent inspections performed during the Cycle 9 outages have identified additional indications of nozzle PWSCC in heat numbers NX7630, NX9915, and 9294.
Table 1 - Hot Leg Nozzle Heat Numbers Through Cycle 8 Heat Number NX7630 NX9915 9294 7617-4 K259 77604 Unit 2 10 12 5 5 Unit 3 10 7 9 5 1 Total 20 19 14 5 5 i PWSCC 1 U2 Cycle 7 1 U3 Cycle 8 1 U3 Cycle
.
.a
- - . -. - . - - .
.,-
AIJ2f_llilENT 3
.
Licensee Presentation Maintenance Rule and Alloy 600 Nozzles R. G. Allen and M. P Short NRC Region IV August 21,1997
_ -___ _ _ _ . - . _ _ . - _ ., _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ -____ _____ ____ _____ -_ -__ _ - _- _ _ _ _
.
NITACil5 TENT 3
.
Licensee Presentation Maintenance Rule and Alloy 600 Nozzles R. G. Allen and M. P. Short NRC Region IV August 21,1997
e a
,
f n
+
}
.
+
4
4
. + ,~ _ - w
- - - - - _ . _ - _ _ _ _ _ _ _ - - - - - - _ _ _ _
. .
e =
.
SOUlt!ER\ CAttIORNia
. EDISON An i0150% 4% 1I R% Allo %AL Cumpay I
Maintenance Rule and Alloy 600 Nozzles R. G. Allen and M. P. Short NRC Region IV August 21,1997
--- - -
_ - -
, _
. .
g . , m o , - s,.
sf EDISON
- , .- , e . ,c. ,. , m _,.,
ISS G T;ae a narent : failure to place t:ae Reactor Coo . ant System in category (a)(1), goal setting, by Ju .y 10,
~ 996, :for RCS Incone nozz. e earage inc ications o aservec in 1995, in accorc ance wit:a t:ae requirements for imp..ementation 0:Ethe Maintenance Ru.e 10 CFR 50.65.
__ _ _-
. .
a, s<mm -o-i EDISOV
.,,. _ ,,,, ~ -,- . .-
RESPONSE Ecison ae_ieves t:aat we met t:ae requirements o:?
10 CFR 50.65 (a)(1) anc. (a)(2), anc too1 ap;3ro;3riate action with respect to Incone.
nozz.es in accorc ance wit:a t;ae SONGS Maintenance Ru e Program anc. t:ae guic ance .
containec in XUMARC 93-01 anc. Regu atory Guic e 1.160.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
-
. . . . ..
. .
o Maintenance Rule anc Incone 600 Nozz es Presentation Outline
- Industry Guideline Implementation
- Safety Concem Consideration
- Defining Unacceptable Performance
-
1995 Indications
-
1997 Indications
- Maintenance Rule Program Weaknesses
.
NRC Inspection Report 97-18 Issues
_ _ _ _ _ _
_
. ,
. .
- nc ustry Guide ine Im;plementation SCOPING
- Maintenance Rule provides criteria to determine which Systems, Structures, or Components (SSC) must be included in the scope of the rule.
- Reactor Coolant System (RCS) included in the scope of the Rule per the Maintenance Rule Implementation Logic Diagram (NUMARC 93-01, Fig.1) as it is safety related.
'
- Inconel 600 nozzles were included as part of the RCS.
__ - -_- _ -
- - -
. .
. .
Inc ustry Guic e .ine Implementation (coir:inuec )
RISK SIGNIFICANT ANALYSIS
- Established Criteria for defining high safety significance (risk significant) systems based on:
Expert panel assessment.
(NUMARC 93-01,9.3.1)
- RCS determined to be a risk significant system. -
. . .
. . .. . .. . . .
.. .
. - - .
.
- . ..
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _
-
. .
e l
.
. Inc ustry Guic.eline Implementation (continuec)
PERFORMANCE CRITERIA DEVELOPMENT
- Most monitoring done at plant, system or train level. (Reg. Guide 1.160, Rev 2)
-
SSC's can be monitored using:
Existing Monitoring Programs Plant Level Performance Criteria Specific Performance Criteria (SSC Availability / Reliability)
.
- Specific performance criteria are required for high safety significant SSCs. (NUMARC 93-01,9.3.2)
_ ___ __ _ -
_
, ,
. .
Industry Guicle.ine Imp ementation (continuec)
RCS PERFORMANCE CRITERIA ESTABLISHED
- Existing Monitoring Programs Existing programs to monitor other components such as system piping and nozzles. (NUMARC 93-01,7.0)
- Plant Level Performance Criteria Scrams per 7000 hours0.081 days <br />1.944 hours <br />0.0116 weeks <br />0.00266 months <br /> Unplanned Capability Loss Factor Unplanned Shutdosvn Safety Function Events .
Unplanned Safety System Actuations Abnormal Radiological Releases Core Damage Frequency
I
. .
.
I Inc ustly Guic e .ine Ennementation (con :inuec }
- Specific Performance Criteria
.
"Less than 98% availability per functional train ~ system; over the previous four quarters,"
"Two, or more comparable Functional Failures (FF) among similar critical components which caused the loss of the system within 36 months," or
"Three, or more random and unrelated functional failures within 36 months that cause a system loss." .
'
Note: A Functional Failure is a failure of a critical component to perform its specified function.
If a component Functional Failure causes the loss or substantial degradation of the system, it is a Maintenance Rule Functional Failure and is counted against the performance criteria.
. _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ _ _ -
-
. .
. .
Inc ust17 Guic.eline Imp ementation (continuec)
- Loss of System When a substantial degradation or failure of the system occurs NUMARC 93-01 supports system level monitoring. "The object of monitoring at the system level is to evaluate the system against established goals to proceed from the present status of not meeting a performance criteria toward a level of acceptable performance."(NUMARC 93-01,9.4.2.1)
Actual Examples .
- MP402 "RCP 3P002 anti reverse rotation device pump" Failed to start
_ - _ _ _ _ _ _ - _ -
_. . _ _ _
. .
t
- :nc ustry Guide ine Imp ementation (continuec)
_ _ .
Component functional failure
.
No system impact Not a Maintenance Rule Functional Failure
- RCP 3MP004 Primary to atmospheric leak at thermowell for 3TE-9179-3 Component Functional Failure System Impact, loss of system function, went to Mode 5 for repair. ,
Maintenance Rule Functional Failure
.-- -
_ ____- .
.. - . - - - - - - - -. .--
..
.
.. :. .
.. ..
. .
Incustry Guide..ine Implementation (continuec)
-
Hypothetical Examples:
- While operating, RCS unknown leakage is maintained
< 1 gpm, acceptable by Technical Specification During shutdown at next refueling, monitoring per our nozzle program, inspections reveal leakage indications, pressure boundary leakage, that occurred during operation.
Per nozzle program, would be acceptable and within program performance criteria (i.e., shutdown, found it fixed it).
Nozzle program working, not a Maintenance Rule Functional Failure.
II e
- - - -
-
. .
. .
,
.
Inc us:ry Guic eh.ne Im: 1ementation (con ~:inuec)
- While operating, RCS leakage of 0.1 gpm is observed during a containment entry at power.
Exceeds technical specification limit of zero known leakage Power reduced and repair ;
Caused system impact, loss of system ability to perform its functions.
Maintenance Rule Functional Failure
1
_ _ _ _ _ _ _ _ _ _
- -
. .
i
- nc us~:ry Guic e..ine Imp ementa: ion (con:inuec)
INITIAL PERFORMANCE ASSESSMENT
- RCS performance determined to be acceptable as of July 10,1996.
Specific criteria of availability not exceeded.
Specific criteria of functional failures causing a loss of the ,
system not excedded.
Nozzle performance acceptable as monitored within existing Inconel Nozzle program.
Plant level performance criteria not exceeded.
.
- RCS remained in preventive maintenance and condition
-
monitoring program (a)(2).
m
- Appropriate maintenance performed. ,3
- -
. _ .
-_-- _ _
-
. ..
.
Inc ustry Guide ine Implementa~ ion (con:inuec) j
.!
CURRENT PERFORMANCE ASSESSMENT
- Performance determined unacceptable in early 1997.
Unit 2
- 4 random RCS component functional failures causing loss of system (limit 3 unrelated failures)
~
S/G tube plug leak S/G manway gasket leak Shutdown cooling isolation valve packing leakoff plug leak _
Pressurizer instrument nozzle leak
- Plant level performance criteria exceeded.
- (unplanned capability loss factor) n
{ ---
,
- -
-
e l
.
Inc u.stly Guideline Implemerr:ation (con:inuec)
Unit 3
- Repetitive Inconel Nozzle Program functional failure causing system loss in early July caused review of nozzle program standard for acceptable performance.
- Plant level performance criteria exceeded.
. .
.
F
L_ - n
. .
Inc ustry Guide ine Imp. ementation (con ~:inued)
UNIT 2/3 CAUSE ASSESSMENT / CORRECTIVE ACTIONS
- Cause Nozzle failure due to PWSCC Nozzle program failed to identify leak due to:
- Installed shims / clamps that prevented optimum inspection.
- Negative Radionuclides Activity Sample Test Results.
- Corrective Actions
.
~
Repaired nozzle (s)
Revised inspection techniques Procuring seal assemblies _
L_________.___ _
- -
, ,
Inc us~:ry Guideline Imp. ementation (con:inuec )
Developing revised nozzle inspection plan Implementing comprehensive nozzle replacement plant
- Repairing accessible hot leg nozzles during mid-cycle outage.
- Installing clamps on pressurizer nozzles with contingency to clamp emergency leakers.
- Evaluating replacement of remaining nozzles in Cycle 10 outage.
~
GOAL SETTING AND MONITORING
- Goal is no nozzle leaks that result in forced plant shutdown or outage extension.
_
-~ - ~ -
. .
e Safe:y Concern Consic.eration l
-
LEAKAGES DUE TO PWSCC ARE NOT AN IMMEDIATE SAFETY CONCERN.
FAILURE MECHANISM IS RECOGNIZED BY NRC AND INDUSTRY AS LEAK-BEFORE-BREAK' .
- "The staff agrees that there are no unreviewed safety questions associated with CRDM/CEDM penetration cracking. ...The staff believes that catastrophic failure of a penetration is extremely unlikely. Rather, a flaw would leak before it reached the critical '
flaw size and be detected during periodic surveillance walkdowns for boric acid leakage." (November 19,1993 NRC Letter)
I8
- - - .
______ -
-
.
. .
l I
Safety Concern Consideration (continued)
- PWSCC axial cracks proceed gradually, slowly increasing leakage permits detection / reaction.
- PWSCC axial cracks can grow up to 2 inch axial length without unstable crack growth.
- PWSCC grows axially. Axial; cracking not a threat to nozzle structural integrity, results in small leak.
- Experience demonstrates crack detection occurs before it can grow to unstable size.
- Technical Specification specifies no pressure boundary leakage. l I-Iowever, NUMARC- 93-01, 9.3.1. states: " Entry into a Technical Specification Limiting Condition fbr Operation, although
- important, is not necessarily risk significant."
___
-
, u -
-
_ .
.
Defining Unacceptare Performance 1995 INDICATIONS - WOULD NOT HAVE PUT RCS IN (a)(1)
- Results of 1995 inspections were consistent with expectations of existing program for monitoring of boric acid leakage.
- Expectations were that some minor PWSCC and subsequent leakage indications would occur. This was an acceptable outcome.
- Observed minor leakage was within industry expectations.
.
,
- Expectations of next operating cycle were similar.
.
- . . . . . . . . _ . . . .
. .
_ . . . - . . . . _ . .
.
. - . . . . . - . - . - . - - . . - .
. - - .
_ . _
- --
4 4
. .
Defining Unaccepta ale Per:Pormance (continuec )
- Industry experience supported this philosophy over next two years.
-
No indications that failure mechanism was not behaving predictively.
- Consequently, performance supported leaving system in (a)(2).
consistent with monitoring per the existing program as endorsed by the Maintenance Rule.
'
- Plan was that program expectations would change if results .
changed.
- --
_ ,
+
Defining Unacceptable Per:formance (con :inuec )
1997 INDICATIONS - TOOK ACTION TO PLACE RCS IN (a)(1)
- Results of both U2 and U3 outage inspections indicated the need to place RCS systems in (a)(1).
- J2' had 4 different occurrences of pressure boundary leakage having a system effect which were evaluated as Maintenance Rule Functional Failures. One related to a failure of a nozzle monitored by the nozzle program.
- U3 had a similar occurrance of the nozzle program to prevent unacceptable results. .
.
- Aggressive response through a revised Inconel Strategic Plan will be implemented as committed in our July 22,1997 letter.
.-_ -
- -
, .
.
Maintenance Ru e Program Weac1 esses MAINTENANCE RULE PROGRAM WEAKNESSES
- Weakness in documenting use of existing programs.
Self-identified as implementation problem by MRIT (4/97)
Action initiated.to document approach '(Forecast 10/97)
-
CWO Monitoring Oversite Data search method may not capture all potential functional failures.
Initial review indicates isolated occurrence.
Action to assess significance by 10/1.
.
. - _ _ ,
~
. .
. . .
INSP3CZON REPORT 97-18 ISSLES
-
CONFLICTING PRO' GRAM INFORMATION
- System / train level performance criteria are stated somewhat differently in procedure SO123-XIV-5.3.2, " Determination of Maintenance Rule Performance Criteria," versus the Maintenance Rule Program Reactor' Coolant System report, and could be misinterpreted. [MRIT Self-Identified; FUNCTIONAL FAILURE OR NOT
- One of the RCS functions is to " serve as a barrier to the direct .
release of radioactive materials," as specified in the RCS system report.
-
.
..
.. - ..
.
u _ __ _ .
.
_ . .
_
_
_. . _
- --
.
. . . .
INSPECTION RPT 97 ~ 8 ISSES (continuec) .
- However, the PWSCC observed in 1995 did not result in what Edison would consider a " direct release of radioactive materials" to the environment.
- The nozzles are monitored and maintained by the existing nozzle program. Indications found in 1995 were accepted by the program. Thus, would not have been evaluated as a Maintenance Rule Functional Failure.
.
-
__ _
- .
... .
INSPECTION RPT 97-18 ISSLES (continuec)
-
DOCUMENTATION
- The initial MR scoping effort failed to evaluate the previous three years of construction work orders (CWO) under which the PWSCC repairs occurred.
- This was an identified documentation weakness of the initial MR scoping effort, and is being corrected.
PASSIVE COMPONENTS
- Edison's Maintenance Rule system analysis reports did not address passive components.
- -- -
. .
. .
INSPECTION RPT 97-18 ISSMS (continuec)
- Many passive components were included as "non critical" components. Nozzles were evaluated as part of piping and monitored by existing programs.
- Initial NUMARC 93-01 guidance interpreted as allowing full utilization of existing monitoring programs (e.g., ISI, snubber program, etc.) to support the demonstration that SSC performance was being effective.y controlled and monitored through preventive maintenance.
.
,
_ _ _ _
.
SU'AGlilEXi' J d
I.icensee Presengagjpg Alloy 600 Instrument Nozzles NRC Region IV Augus* 21,1997
,
!
l
,
. . . .
. .
.
.
. _ _ , , , , _
- - -
. , , ,, .
.. .
.
U esa~
Alloy 600 Instrument Nozzles NRC Region IV August 21,1997 ;
M. P. Short I
L - - ---- ---_ _ _ -
. _ .
. .
Industry Experience with Alloy 600 PWSCC esensesee.End
- Panisades
h2-s on.or.3
- =.r,-
Ringheds 3
- "" weser Space Noutes CED , ,
]r.essCIheputu.=~~ <
O 1 3 --
occeee t j caevenc m 2
]I
,
Aft 1 I l i
%9y c
-
-U.,o. na,*
( J r
,
/ h
' '
"W ""T" e
- san oncare 2 r=>
Locations Where PWSCC Has Occurred in Ancy 600 Prwessare Boundary Penetra nsons (No Specific Plant Design)(Esclusin of Steam Generator Testnes and Plugs)
., J Leakage Detected
.
. . .. .
. .. . .
_- .
. . .
. . .
_____
. .
San Onofre Experience with Alloy 600 PWSCC Unit 2 Unit 3 Pressurizer
= Heatersleeves 30 30 i
. Instrument nozzles 7 (5) 7 (4) ,
Reactor Vessel Head
. CEDM nozzles 91 91 ;
- Incore instrumentation 10 10 l
Steam Generators 8 8 i i
i Primary Loops
'
.
- Hot leg nozzles 32(1) 32 (10)
e Cold leg nozzles
"
12 12 (1)
!
Total: 190 (6) 190 (15)
'
() Number of nozzles replaced
,
i
.. .. . . _ - .
-
_ _ . _ _ _ _ _ _ _ - _ - - -
.
. .
!
,
a I Alloy 600 Instrument Nozzles i Missed Opportunity l
+ NRC SALP implied i
[
- Pro-active nozzle replacement during '97 j refuelings was warranted by industry and l SONGS experience
- t
!
+ Whereas SONGS ,
l - Continued with its approach ofinspection and
-
replacement as needed i l
i 1
[
!
4
!
'
. . . - - . . - - - -_- ---__ - -- - -
!
. .
Inspect and Replace as Needed
+ EPRI assessed this as the appropriate method (TR-103696)
+ Supported by SCE and EPRI's assessment of safety significance
+ In '95, SCE predicted nozzle leakage trends vcould continue through '97 refuelings and if not, the program approach would be revisited
- Based on engineeringjudgment
- Recently revisited using SG strategic planning techniques with same conclusion
_ _ - - - _ _ _ -
. .. . . .
. .
Hot Leg Nozzles - Units 2/3 Trends 1.00 0.90
[
ON f ,
p 0.70
'[ [ '
M 0.60
[ [ ,
'
3 Om
/ / Uret3 Trerxs in 1995 Ud 2 and kxfusty Trends
.o . l
~ '
'
Urit3,1997 {'-
5 030 ;
On
'
O //
a,0 u~z'~al // ;
Urit3.1995 O.00 O 10 20 30 40 Trne (EFPY) !
-
!
, !
!
'
_ _ .
. .
.
- .
Hot Leg Nozzle Replacement
!
+
Forecast i
!
I
!
i 95 97
! i i, i
.
SO2 FC 1 1
'
,
!, Actual 0 0 i
\
> SO3 FC 1 2
.
,
Actual 2 9 l i
i !
,
r t
i
.i
,
f
~
i
7 I i
i i
r
!
. _ . _ _ - - - . _ - _ -_ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
i
,
N *
Summary
+ Through 1995, Unit trends were consistent with industry experience and managed effectively by the nozzle program
~
+ In 1997 Unit 3's trend exceeds industry
- Aggressive replacement program under development
- Managed under (aX1) of the Maintenance Rule
.
._
-