IR 05000361/1997017

Insp Repts 50-361/97-17 & 50-362/97-17 on 970706-0816. Violations Noted.Major Areas Inspected:Licensee Operations, Maint,Engineering & Plant Support

ML20216C755
Person / Time
Site:	San Onofre
Issue date:	09/04/1997
From:	NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
To:
Shared Package
ML20216C709	List: IR 05000361/1997017 ML20216C703 ML20216C730
References
50-361-97-17, 50-362-97-17, NUDOCS 9709090134
Download: ML20216C755 (30)

Text

.

.

ENCLOSURE _R U.S. NUCLE AR REGULATORY COMMISSION

REGION IV

Dock et Nos 50 361 50 3G2 License No NPF-10 NPF-15 Report N ,97-17 50 362/97-17 Licensee: Southern California Edison C Facility: San Onofre Nuclear Generating Station. Units 2 and 3 Location: 5000 S. Pacific Coast Hw San Clemente, California Dates: July 6 through August 16,1997 Inspectors: J. A. Sloan, Senior Resident inspector J. G. Kramer, Resident inspector J. J. Russell, Resident inspector S. L. McCrory, Reactor Inspectcr Approved By: D. F. Kirsch, Chief, Branch F Division of Reactor Projects ATTACHMENT: Supplemental Information

!

9709090134 970904 POR ADOCK 05000361 G PDR m

_ --

.

,

LXICUTIVE SUMMARY San Onofre Nuclear Generating Station, Units 2 and 3 NRC Inspection Report 50 361/97-17;50 362/97-17 This routine announced inspection ir cluded aspects of licensee operations, maintenance, engineering, and plant support. The report covers a 6-week penod of resident inspectio QRfLMiE01

The licensee generally conducted midloop operations and related activities in a carefully controlled manner. Emergent issues were conservatively evaluated and resolved. However, some poor routine operator practices and procedural weaknesses were identified while both units were operating in midloop condition These included procedural inconsistencies, a lack of rigor in monitoring reactor coolant system (RCS) levelindications, failing to recognize the need to verify that RCS heatup limits were beinn met, and failing to ensure that procedural i

prerequisites for draining the RCS were documented, as required (Section 01.1 *

A noncited violation was identified by the licensee regarding the failure to use

!

current RCS samples, analyzed for boron concentratior., to support surveillance requirements (SRs) for determining shutdown margin (SDM). Additionally, the

!

I licensee identified that, in one instance, the boron concentration for a refueling

! water storage tank was inadvertently used in place of RCS boron concentration in a shutdown margin calculation (Sections 01.2 and M8.2).

The Operations crews effectively performed the reactor startups of Units 2 and 3 (Section 01.3).

A violation was identified by the inspectors as the result of operators increasing reactor power in Unit 2 at a rate greater than allowed by the procedure when fuel cladding leaks were known to exist. The nonconforming limits, directed by senior managers, were broadly and nonconservatively interpreted by the Operations manager. Communication of the limits to the shift superintendent led the control room operators to incorrectly believe that a review of the data had revealed that fuel cladding leaks did not actually exist. However, the use of the higher power ramp rate limit did not adversely affect the cladding leak (Section 01.4).

The inspectors identified a lapse in attention to detail when Operations f ailed to schedule and perform a weekly update of the rapid downpower reactivity calculations (Section 01.5).

The longstanding presence of an unsecured cart near safety-related equipment in the control room was a weakness in the implementation of the licensee's seismic controls program (Section O2.1).

Inspectors identified two examples reflecting inattention to detailin the use of procedures by operators. In the first example, the procedure used for draining the RCS to midloop conditions had not been reviewed carefully enough before use to identify an obvious conflict in its applicability. The second example demonstrated a

_ _ _ _ .

l l

i l

l lack of rigor in documenting completion of procedural prerequisites before proceeding with the evolution of draining the RCS (Section 03.3).

Operators and supervision f ailed to recognize the requirement to verify that RCS temperature limits were met when heating up the RCS while in Mode 5 (Sections 04.1 and 01.1).

Operations management's expectations for monitoring the reactor vessel sight glass indication of RCS level while operating in midloop conditions were week, in that only a daily reading was required when RCS level was not intentionally being changed. The failure to more frequently use all available indications to confirm the validity of control room indications of RCS level, when small changes in RCS level could have serious safety consequences, was not a conservative approach to ensuring reactor safety. However, operators carefully monitored levelindications while level was being changed (Sections 04.2 and 01.1).

The control room supervisor (CRS) maintained adequate control of two evolutions by substituting his personal direct oversight of the execution of the evolutions for good communications practices. However, the CRS f ailed to reinforce the management expectations for good communications, as specified in licensee procedures, even during special testing (reactor physics testing) involving reactivity manipulations (Section 04.3).

A transient that occurred while operators were attempting to perform an RCS leak rate surveillance after ramping up to 80 percent power at the beginning of the fuel cycle demonstrated that operators' knowledge of slightly negative integrated temperature coefficient (lTC) operation was weak. Additionally, Operations management missed the opportunity to anticipate the problems with performing the surveillance and, accordingly, missed the opportunity to take mitigating action Operations' response to the event was adequate (Section 04.4).

Operator knowledge of the core operating limits supervisory system (COLSS) power indications used to monitor reactor power during a power increase in which the rate was limited was weak (Section 04.5).

A repeat violation was identified by the inspectors when operators failed to follow a procedure requiring the volume control tank (VCT) inlet diversion valve to be placed in manual when the block valve was closed (Section 04.6).

Nuclear Oversight Division (NOD) observers provided valuable input and insights to the operating crews during the concurrent midloop activities (Section 07.1).

Maintenance

• Generally, surveillance and maintenance activities observed were found to be thorough. All work observed was performed with the work package present and in

__

.

.

3-active use. Technicians were knowledgeable and professional. The inspectors frequently observed supervisors and system engineers monitoring job progress, and quality control personnel were present whenever required by procedure. When applicable, appropriate radiation controls were in place (Sections M1.1 and M1.2).

The placement of a radio inside the emergency diesel generator (EDG) local control cabinet compromised the seismic qualification of the EDG. The technician's performance was unprofessional, and the technician did not understand the potential risk to the EDG as the result of his actions. The licensee's assignment of

an inexperienced electrician to the Work-It-Now (WIN) team, although inadvertent,

!

and the f ailure to provide adequate supervision, were isolated weaknesses in the licensee's maintenance program. The licensee's response to the situation was excellent (Section M4.1).

The licensee used properly trained Shop Services and instrumentation Division (SSID) machinists to perform work during the Unit 3 refueling outage or supervised the machinists that were not independent worker qualified (Section M5.1).

A violation was identified as the result of an inadequate work plan in a maintenance order (MO) for recoupling a four fingered control element assembly (CEA), which resulted in the reactor being reassembled with the CEA not coupled. The work plan was inconsistent with written vendor documentation and existing licensee procedures (Section M8.1 *

The licensee determined that an Operations SR for determining SDM was not properly performed (Sections M8.2 and 01.2).

Enaineerinn

A violation was identified by the inspectors in that the acceptance standard listed in weld records (WRs) for liquid dye penetrant examination of pipes with a wall thickness of 5/8 inch or less was not correct, and for rounded indications was not conservative. The nondestructive examination (NDE) performance related to welds to replace a charging system check valve was good (Section E4.1).

Plant Sunnort

General housekeeping in the radiologically controlled area was very good. Materials storage areas were generally neat, and contaminated areas that no longer had ongoing work activities were returned to their preoutage boundaries (Section R1.1 *

The inspectors identified loose surf ace contamination in an area not posted as being contaminated. In this instance, licensee Health Physics personnel were not sufficiently timely in identifying and posting loose surface contamination in excess of limits (Section R1.2).

__

n

.-

Report Detaita Summary of Plant Status Unit 2 began this inspection period in Mode 5 in a forced outage to repair a check valv The dnit entered Mode 4 on July 13,1997, and Mode 3 on July 14. On July 15 the i reactor was brought critical and the generator output breaker was closed the following ;

day. The unit reached 70 percent reactor power on July 17,1997. The subsequent l power increase was interrupted at 70 percent power due to the failure of Valve 2MUO200, I a.sociated with Main Feedwater Pump 2P062, After repairs to the valve were completed, I the power increase was continued, and the unit achieved 100 percent power on

!

July 22,1997, and operated at 100 percent power for the duration of this inspection perio+ j

Unit 3 began this inspection period in Mode 5 in day 86 of the Unit 3 Cycle 9 refueling outage. The unit entered Mode 4 on July 12,1997, and Mode 3 on July 13. On July 14 the reactor was brought critical, and the generator output breaker was closed on July 2 l Power was increased to 100 percent on July 24,1997, and Unit 3 operated at 100 l percer:t power for the duration of this inspection perio I. Operations 01 Conduct ni Operations 01.1 Midloon Goerations - Units 2 and 3 'nspection Scoce (71707)

I 7he aw lors reviewed portions of Procedures SO23-31.8, Revision 11, " Draining the WM SO23 5-1,3, Revision 17, " Plant Startup From Cold Shutdown to Hot l Stanon . SOi3-13-15, Revision 8, " Loss of Shutdown Cooling:" and monitored the licensee's perbrmance during the midloop operations in Units 2 and 3 from July 7-11,1997, b, Observations and Findinas Procedre Documentation j l

On July 9,1997, while operating in midloop, Unit 3 operators prepared to initiate .

draining to further lower RCS level. The CRS provided the inspectors with a copy j of Procedure SO23 3-1.8, Attachment 2, that was to be used to accomplish the s drain down. The inspectors identified that not all of the prerequisites had been l

.

signed as being performed, but that the operators were preparing to initiate the j draining evolution. The inspectors informed the CRS about the observation. The CRS had understood that all the prerequisites were signed off. Upon reviewing the l record copy of tt'e procedure, maintained by the control operator, the CRS l determined that several prerequisites were not signed off. The operators reviewed ,

the prerequisites and signed them off. The inspectors determined that the prerequisites had been performed but the procedure had not been signed. The l

>

-.

!

.

-2-inspectors concluded that the operators displayed a weakness in attention to detail by not ensuring that all documentation was completed prior to draining the RC RCS Heatuo Verification i

I On July 10,1997, at approximately 9:50 a.m., Unit 3 operators initiated an RCS heatup from 95'F to 125'F in preparation for raising RCS level to exit midloop and go to solid plant operations. At approximately 10:05 a.m., the inspectors informed the CRS of Technical Specification (TS) SR 3.4.3.1, which required verification of the RCS heatup rate at 30 minute intervals. The operators then initiated Procedure SO23-5-1.3, Attachment 10, to monitor the verify the heatup rate and completed the verification within the 30-minute requiremen The licensee initiated an action request (AR) to document the failure to identify the requirement to verify the heatup rate. The inspectors observed that the operating crew, including shif t supervision, and Operations management supervision in the control room failed to recognize the need to verify proper RCS heatu Operating procedures trigger performance of the surveillance when performing a heatup to change modes. Since the heatup in this instance was not intended to

[

!

result in a mode change, the procedure that triggers performance of the surveillance was not being used. TS Limiting Condition for Operation 3.4.3.a limits the heatup to 60 F in any 1-hour period, which could be achieved without changing roode The failure of licensee procedures to recognize this is considered a weaknes However, for the July 10 heatup, the temperature was intended to only be increased 30oF, and the limiting condition for operation limit could not be violated by the planned heatu The inspectors concluded that the management oversight in the control room did not sufficiently focus on ensuring that the SR was being properly performed, despite a recent violation for failure to verify cooldown rate of the Unit 2 pressurizer (documented in NRC Inspection Report 50-361:50-362/97 05).

RCS Level Monitorino On July 10,1997, with the RCS level at 17.8 inches in the hot leg, based on refueling water level indication, the inspectors questioned the Unit 2 CRS regarding the sight glass level. The CRS replied that the level had been 19 inches on July 9 when he was last on shift, but did not know what the level was on July 10. He asked if any of the Operations crew knew what the level was. None of the crew members knew the sight glass leve The sight glass level was being monitored once per day, and had not yet been checked by the crew that was on shif _

. . .

.

3-The CRS subsequently dispatched an operator to verify the level and determin that the sight glass level was 18.5 inches. The sight glass level correlated with other levelindications within the 1.5 inch requirement of Procedure SO23 3- The Operations manager stated that the credited levelindications were c rnonitored in the control room, and were alarmed. The sight glass level was j checked daily to correlate the indications. Additionally, the sight glass desig mechanical, and the sight glass is not very sensitive to smalllevel changes. The

_

sight glass levelis not indicated in the control room. The Operations manage stated that a daily correlation of levelindications was appropriate, and that operators had indications of other parameters to use to monitor the shutdown cooling system.

. The inspectors concluded that the performance of Unit 2 operators' monitorin RCS level was weak in that operators did not frequently monitor all the available

[ ' Section indications 04.2. when in a sensitive midloop condition. This issue is also discussed in I

RCS Level Correlation

}-

c_ On July 9,1997, the inspectors observed the operators in both units drain the RCS

.

i to a level of approximately 18 inches above the bottom of the hot leg.- The - I operatorsreference used a heated l j additional point. junction thermocouple (HJTC) located at 21 inches as an

. . The operators performed a slow approach to 21 inches

{ and monitored for the uncovering of the HJTC, which would indicate that the P.CS level was at 21 inches. The operators verified that allindications agreed with each l

'-

.

other and the HJTC within a 1.5 inch band,' The inspectors concluded that the

!.- operators used all available indications to ensure accuracy of the' level

.

. instrumentation when draining the RCS to the 18-inch leve Procedural inconsistencv -

o .

Procedure SO23-3-1.8, was used in both units to drain the RCS to about 18 inches I

above the bottom of the hot legs. Attachment 22, " Draining the RCS Limitations

'

and Specifications Attachment," Step 3,4, stated that " Draining is limited to a

,

nominal level of 26 inches. However, level excursions below 26 inches are E acceptable provided they do not uncover HJTC No. 6 at 21 inches." Other portions

-

of the procedure did provide for draining to a level below 21 inches. The inspectors concluded that the procedure was inconsistent as to the range of levels for which h the procedure could be used.

.l .i

,

r i.

f

i.

,

yr - - -

--r- g- , c.,--%.-r os. ry--.~ w - - , ..- v -2 w, - , . . . -- y- * - - v -- -

w-r- =

.

,

-4- Conclusions The licensee generally conducted midloop operations and related activities in a carefully controlled manner. Emergent issues were conservatively evaluated and resolved. However, some poor routine operator practices and procedural weaknesses were identified while both units were operating in midloop condition These included procedural inconsistencies, a lack of rigor in monitoring RCS level indications, f ailing to recognize the need to verify that RCS heatup limits were being met, and failing to ensure that procedural prerequisites for draining the RCS were documented as required. These indicated a need for improvement in the attention to detail and oversight of operational activities during midloop condition Additional comments on oversight of midloop activities are discussed in Section 07.1.

l 01.2 Shutdown Marain (SDM) Verification Unit 3(71707)

On July 4,1997, with one source range nuclear instrument inoperable in Unit 3, operators were required by TS to verify SDM every 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />. On July 7, an operator identified that the SDM calculation was being performed using RCS boron samples that were as much as 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> old, inconsistent with the intent of the TS requirement. The RCS boron concentration semples were being obtained every 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, and operators had been using the most recent sample results to perform the 12-hour SDM verification. Operators then requested a current sample be obtained, and documented the surveillance error in an AR. The SDM was calculated to be within limits. This failure to perform SRs 3.1.1.2 and 3.1.2.1 was a violatio This nonrepetitive, licensee-identified and corrected violation is being treated as a noncited violation, consistent with Section Vll.B.1 of the NRC Enforcement Policy (NCV 361/97017-01).

The licensee submitted a licensee event report (LER) regarding the missed surveillances, in reviewing documentation to prepare the LER, a Compliance engineer identified that, on July 4, operators inadvertently used refueling water storage tank boric acid concentration instead of RCS boric acid concentration in the SDM calculation. The results of both the refueling water storags tank and RCS samples showed nearly identical boric acid concentrations, and an operator l inadvertently picked up the wrong chemistry results sheet and did not verify the sample source, since the results matched his expectation, The LER is discussed in Section M .3 : Reactor Startuo - Units 2 arid 3 Insoection Scone (71707)

The inspectors reviewed Procedure SO23-3-1.1, " Reactor Startup," and observed operations perform reactor startups of Units 2 and l l

. ..

.:

-5 b, Observations and Findinas On -July 1_4 and 15, the inspectors observed the reactor startups of Units 3 and 2,-

respectively. Tl'e inspectors observed the Unit 3 reactor startup briefing and observed that the briefing covered all aspects of the startup as described in Procedure SO23 31_.1.in addition, Reactor Engineering provided ad litional insight

.

to the expected performance ~of the reactor, and the CRS highlighted the entry conditions for emergency boration. The licensee minimized unnecessary work activities to ensure operator distractions were kept to a minimum; however, the inspectors _ observed significant noise from Unit 2 crew turnover during the Unit 3 startu The inspectors observed the Unit 3 operating crew perform a shift turnover at a hold point just prior to the reactor going critical, and observed the Unit 2 operators-perform a shift turnover at the first hold point of the startup. Both crew turnovers were controlled and ensured that the ancoming crew had appropriate knowledge of the plant conditions, ' Conclusions The Operations crews effectively performed the reactor startups of Units 2 and .4 Power Ascension with a Failed Fuel Pin - Unit 2 Inspection Scope (71707)

The inspectors observed portions of the Unit 2 power ascension on July 16,1997, and reviewed Procedure SO23-5-1.7, Revision 10, " Power Operations."

- Observations and Findinas On July 8,1997, Reactor Engineering provided'a Reactor Engineering Data Transmittal to Operations that documented the fuel status of Unit 2. The transmittal stated that, based on postshutdown radiochemistry analysis (following the June _29,1997, shutdown), there were indications of cladding defects. The -

-Nuclear. Fuels engineer informed the inspectors that the fuel was suspected of --

having one moderate sized leak in a once-burned. or twice-burned fuel pin, and that

_

a large iodine spike (about 100 times the normal value) had been observed. Further review by Nuclear Fuels determined that the leaking fuel pin was from Batch K, which was twice-burned fuel. .The transmittal stated that, as a consequence of the cladding defect, a ramp rate restriction of 5 percent per hour was in effect for the power ascension at or above 20 percent power,

, On-July 16,1997, just before the power ascension was to begin, the Operations manager observed that increasing the ramp rate limit to 10 percent per hour would be needed to support achieving 80 percent power in time to conduct a heat treat of (

-

'

,,

.

the saltwater cooling system that evening at high tide. The ne>c. suitable high tide would be about 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br /> later, and the licensee did not want to have to hold power at 80 percent, or increase power and then reduce power again, in order to conduct the heat treatmen At the dail) plant status meeting on July 16, the basis for the 5 percent ramp rate limit was discussed emong the managers of Nuclear Fuels, Operations, Station Technical, the Vice Prcsident - Nuclear Generation, and the Vice President -

Engineering and Technical Services. The Nuclear Fuels manager wanted to leave the ramp rate limit at 5 percent per hour, based on industry experience that indicated that fuelleaks would not be aggravated by implementation of that limi However, he acknowledged that there was no guidance from the fuel vendor on limiting the ramp rate (below the normal 20 or 30 percent per hour) for operation I

!

with f ailed fuel cladding. The Operations manager considered that implementation of a 10 percent per hour limit was sufficiently conservative, and .e licensee had no

information indicating that using the higher ramp rate would lead to increasing the magnitude of the failur The Vice Presidants decided to allow use of the 10 percent per hour ramp rate. The Operations manager then informed the shif t superintendent that he had discussed the fuel f ailure with Nuclear Fuels, and that management had decided to lift the 5 percent per hour limit and implement a target of 10 percent per hour, which was the normal limit when no fuel f ailures existed. During discussions with the inspector, the shift superintendent stated that he had interpreted these instructions to mean that management had determined that the fuel was not failed, and that the normal guidance applie After the unit was synchronized to the grid, the shift superintendent instructed the Unit 2 operators to increase power at a target of 10 to 12 percent per hour, and not to exceed 15 percent per hour. At 10:42 a.m. on July 16,1997, operators initiated the power increase. Power was increased from 20 percent at approximately 11:15 a.m., to 43 percent at approximately 1:30 p.m., for an avarage of 10 percent per hour. However, between 12:30 p.m. and 1:30 p.m., power was increased approximately 14 percent, from 30 percent to 44 percen At approximately 12:55 p.m., direction was given to reduce the ramp rate to a target of 5 percent per hour, due to an emerging problem with a feedwater pump that would preclude performance of the heat treatment that evening. However, power was increased by 9 percent between 1 p.m. and 2 p.m., as operators had difficulty controlling the ramp rate due to the effect of the changing xenon concentratio At approximately 2:30 p.m., the inspectors observed plots of the power increase, and informed the Unit 2 CRS and the Operations management representative in the control room that the ramp rate exceeded the requirements specified in Procedure SO23-5-1.7. Step 6.3.1 of the procedure required following the

_ _ _ _ .

.

.

-7-guidelines of Section 6.2, " Guidelines During Power Ascension." Section 6.2. stated: "Do NOT EXCEED the applicable Maximum Core Power Escalation Rate of Attachment 1." Attachment 1 provided a table of maximum core power escalation rates for various fuel condition categories. The table stated that when fuel cladding leaks are known to exist, the maximum core power escalation rate was 5 percent per hour between 20 percent and 100 percent power. A note to the table directed operators to " refer to the Operations Physics Summary for current status of fuel f ailure per Reactor Engineering Transmittal." The transmittal, as described above, stated that there were indications of fuel cladding defects. Failure to follow Procedure SO23 51.7 was in violation of TS 5.5.1.1.a (VIO 361/97017-02).

Because the control room operators misinterpreted the Operations manager's directions, and assumed that there were no fuel cladding leaks despite the transmittal, operators thought that the procedural restriction no longer applie The Vice President - Nuclear Generatior, was unaware that operators had been given guidance that allowed them to exceed the 10 percent per hour ramp rate limit that he had deemed acceptabl Radiochemistry analyses performed after equilibrium conditions were achieved in the RCS indicated that the fuel cladding leak had not been adversely affected by the power ascension, Conclusions A violation was identified by the inspectors as the result of operators increasing reactor power in Unit 2 at a rate greater than allowed by the procedure when fuel cladding leaks were known to exist. Senior managers authorized the use of a higher power ramp rate than that specified in procedures without properly revbing the procedure first. The limits directed by senior managers were broadly and nonconservatively interpreted by the Operations manager. Communication of the limits to the shif t superintendent led the control room operators to incorrectly believe that a review of the data had revealed that fuel cladding leaks did not actually exist. However, the use of the higher power ramp rate limit did not adversely affect the cladding lea .5 Weekiv Rapid Downoower Reactivity Calculation Not Updated - Unit 2 {n_soection Scoce (71707)

On July 28,1997, the inspectors walked down the Unit 2 main control boards, Observations and Findinas Procedure SO23-5-1.7, Revision 10, Attachment 9, " Power Maneuvering Guidelines," had not been updated on a weekly basis. This attachment was posted

_ .

'

I e-8-on the main control boards as an operator aid and used to calculate the amount of boration or CEA position changes needed for various power decrease The calculations implemented a good practice that was not required by NRC regulation Management expectation, as expressed in Procedure S023 5-1.7, was that the attachment would be performed weekly. The attachment had last been performed on July 19,199 The safety consequence was low because fuel depletion was small from week to week. Consequently, the changes in the amounts of boron or rod worth necessary to achieve power changes were also small The inspectors determined that the attachment had not been scheduled to be performed, as reflected in a prirtout of surveillances to be completed, which was kept in a " Red Book" in the control roo Licensee Operations personnel used a computer program to automatically print a

,

l schedule of TS surveillances, including Attachment 9 mentioned above, to be performed on a weekly basis. Licensee Operations personnel explained that, due to a data entry error, Attachment 9 had not been scheduled. The Operations manager explained that there were no more controls over scheduling TS surveillance than over scheduling Attachment 9; however, operator skill of the craft (crew recognition of daily and weekly surveillances to be performed, independent of the " Red Book")

would prevent a similar TS surveillance scheduling error resulting in missed periodicity, Conclusions The inspectors identified a weakness in attention to detail when Operations failed to schedule and perform a weekly update of the rapid downpower reactivity calculation O2 Operational Status of Facilities and Equipment O2.1 Unsecured Cart in Control Room Units 2 and 3 (71707)

On July 13,1997, the inspectors observed that a book cart, with unlocked wheels, was about 4 feet from the control board where the Unit 2 EDG controls were located. The cart had been in the same position for several month Procedure SO123-1-1.20, Revision 4, " Seismic Controls," described this condition as acceptable only because the cart was not unattended. The shift superintendent directed that the cart be removed from the control room, which corrected the deficiency. The inspectors concluded that the longstanding presence of the cart

- _ _ _ _ _

.

.

-9-near safety-related equipment in the control room was a weakness in the implementation of the licensee's seismic controls progra .2 RCS Valve Alionment Verification - Units 2 and 3 (717071 On July 11 and July 14,1997, for Units 2 and 3 respectively, the inspectors

.

observed nuclear plant equipment operators verify that the four vent valves, and four drain valves, downstream of each of the spray valves were each in the required

" closed" position. The inspectors also observed the operator verify as " closed and locked" Valve 2MUO995, the reactor head vent line loss of cooling accident limiter orifice valv Operations Procedures and Documentation (71707)

O3.1 Procedure for Drainina the RCS i

l An inconsistency in the applicability of Procedure SO23-3-1.8 was identified by the inspectors and is discussed in Section 01.1, O3.2 Procedure Documentation i

l Operators did not document completion of procedural prerequisites before draining the RCS while in midloop until the inspectors identified that the prerequisites had not been signed. This was discussed in Section 0 .3 Conclusions Reaardina Operations Procedures and Documentation These examples both reflect inattention to detailin the use of procedures by operators. In the first example, the procedure used for draining the RCS to midloop conditions had not been reviewed carefully enough before use to identify an obvious conflict in its applicability. The second example demonstrated a lack of rigor in documenting completion of procedural prerequisites before proceeding with the evolutio Operator Knowledge and Performance (71707)

04.1 RCS Heatuo Verification Operators and supervision failed to recognize the requirement to verify that RCS temperature limits were met when heating up the RCS while in Mode 5. This issue is discussed in Section 0 .2 RCS Level Monitorina While level was being maintained constant in midloop, operators were monitoring the sight glass level only once per shift, and near the beginning of one shift, the i

. _ _ _ _

.

.

- 10-operators were not cognizant of the level readings since their shift the previous da ~

Operations management did not expect operators to monitor the sight glass level more than once per day, while not intentionally changing level, unless questions arose regarding the validity of the other levelindications. This was a weakness in Operations management's expectations for operators. However, operators carefully monitored level indications while level was being changed. This issue is discussed in Section 0 .3 Communications Practices insoection Scope (71707)

The inspectors observed communications during operational activities in the control roo Observations and Findinaq The inspectors observed two examples in which a particular CRS failed to ensure that operators under his supervision performed in accordance with the good

" operating practices described in Operating Instruction S023 0-44, Revision 2,

" Professional Operator Development and Evaluation Program."

During reactor physics testing in Unit 3 on July 15,1997, the CRS directed an -1

assistant control operator (ACO) to withdraw Group 1 CEAs one step, The ACO did I not repeat the instructions back, as described in Operating Instruction SO23-0-44.

3 The ACO selected the controls to withdraw a single CEA in Group 1, using manual individual mode, instead of selecting manual group mode to select the entire

,

Group 1. The ACO announced that he had selected a single CEA and was j withdrawing it one step. The CRS, who was standing in a position to directly i observe the ACO's actions, then intervened and repeated his direction to withdraw the entire group one step, which the ACO di ]

,

The previous day the same CRS had directed the control operator and the ACO to l

, secure the containment minipurge. The control operator and the ACO did not  ;

repeat the instructions back. The CRS stood close enough to the other operators

'

that he could ascertain that they were taking the intended actions.

.

.

Operations management acknowledged the inspectors' comments and coached the l CRS and the operators, Conclusions

Instead of using good communications practices to ensure that verbal directions l j were heard and understood, a CRS substituted his personal direct oversight of the

execution of the evolutions to maintain adequat. control. The CRS failed to

!

_ __

,

I l

( )

-11-reinforce the management expectations for good communications, as specified in ,

Procedure SO23-0-44, even during special testing involving reactivity manipulation '

04.4 Transient Durino Performance of RCS Leak Rate Surveillance Unit 3 l Insnection Scone (71707)

The inspectors reviewed logs and data trends, interviewed operators, and reviewed RCS leak rate surveillance records associated with a transient that occurred in Unit 3 on July _22,199 Observations and Findinos On July 22,1997, a Unit 3 reactor power ramp to 80 percent was completed at approximately 1:15 Operators were required to perform an RCS leak rate surveillance before 12:10 on July 23 (based on the 120 hours0.00139 days <br />0.0333 hours <br />1.984127e-4 weeks <br />4.566e-5 months <br /> allowed by TS SR 3.4.13.1 when transients prevent meeting the normal 72-hour surveillance interval). Surveillance Operating Instruction SO23-3-3.37, Revision 12 "RCS Water inventory Balance," specified a 1-hour stabilization period, followed by at least 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> of data gathering. The procedure also prohibited changing turbine load, and discouraged RCS dilution or boration during the data-gathering perio Operators first attempted to perform the surveillance at about 2:30 p.m. on July 22, about 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> after reaching 80 percent power. However, RCS cold leg temperature !

lowered more than anticipated, due to increasing xenon and the near zero negative ITC, The surveillance procedure allowed only a l'F temperature change, and the operators aborted the surveillance before completio Operators were not concerned that the RCS leak rate would be too high, since the critical functions monitoring system online leak rate calculation showed that the leak rate was only about 0.07 gpm, well within the limits of TS 3.4.13. However, the critical functions monitoring system calculation does not substitute for the SR procedure specified methodolog Operations personnel changed Surveillance Operating instruction SO23-3-3.37 to provide more tolerance for the surveillance conditions. Only a minor change was made, because there was insufficient time to do a major revision, including all required reviews, before the surveillance interval expired. The one-time change allowed up to a 3 F temperature drop during the data gathering perio At approximately 8'p.m., operators once again attempted to perform the surveillance, but the temperature drop again exceeded the procedural limits. During this attempt, RCS cold leg temperature dropped sharply to about 547 F. Operators performed a quick estimate for a dilution to compensate for the drop and to return e

,

-12-RCS temperature to the top of the operating band. They dilutect the RCS with 1400 gallons of makeup water, which caused the temperature to rise sharply, even though operators used only one charging pump during this interval to reduce the actual dilution rate. Operators had to use both part-length and regulating group CEAs, and limited boration, to control temperature, which had increased from 547 F to 555 F in about 30 minutes. Temperature peaked at approximately 5 5 6.6 * The surveillance was satisfactorily completed at 11:53 p.m. on July 22, without reliance on the wider band for RCS temperatur <

In response to this event, Operations management reinstated the IT-1 Special Test controls that had been invoked for reactor physics testing. This required Reactor Engineering to be more directly involved in overseeing calculations for all reactivity manipulations other than routine burnup compensation, Additionally, Reactor Engineering and Operations developed enhanced guidelines for reactivity control with a near zero ITC. Operators had been trained on operation with a positive ITC, but were not f amiliar with the reactor response and methods of control appropriate for a very slightly negative ITC. Reactor Engineering and Operations management promptly provided special training to each oncoming Operations cre The previous successful performance of the surveillance had been at 12:10 a.m. on July 18,-1997. Plant operation had been steady for approximately 2 days after ,

that, and the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> SR limit had expired just before the main turbine was placed '

on line, and before the power ramp to 80 percent power had begun. The inspectors ;

determined that operators had the opportunity to perform the surveillance before beginning the power ram Conclusions This event demonstrated that operators' knowledge of slightly negative ITC operation was weak. Operations management missed the opportunity to anticipate the problems with performing the surveillance and, accordingly, missed the opportunity to take mitigating actions. Operations response to the event was adequat :

04.5 Knowledae of Reactor Power Indications - Units 2 and 3 a; Insoection Scone (71707)  :

The inspectors questioned operators regarding the meaning of the power indication being used to monitor power increases in both unit I s

f

,.

.e-13-b. . Observations and FintrLq; On July 22,1997, both units were ramping up in reactor power. Unit 2 had a 5 percent per hour ramp limit, and Unit 3 had a 3 percent per hour limi The inspectors questioned the operators regarding which power lev.rlindications they were using to track the ramp limit. In both units, operators indicated that the COLSS power limit, CV9000, was being used. CV9000 is a calculated value with multiple inputs. The inspectors questioned the operators regarding their knowledge of the inputs, or where they could go to find out what they were. This was-important because the various power indications are relevant in different ways, depending on the taason for monitoring the power. However, none of the operators on either uriit could identify what the inputs were or where they could obtain the information. The Unit 3 CRS did know that turbine power was a significant inpu There were no ready references in the control room that described the inputs to the COLSS power indication The Operations superintendent stated that he would have expected the operators to be more familiar with the COLSS power indication . Conclusiong Operator knowledge of the COLSS power indications, used to monitor reactor l

power during a power increase in which the rate was limited, was wea .6 VCT intet Diversion Valve in an incorrect Mode - Unit 2 Inspection Scoce (71707)

On July 29,1997, the inspectors observed the alignment of systems as indicated the control boards in the control room, Observations and Findinas The inspectors identified that the letdown divert valve, Valve 2LV0227A, was in automatic. This is a 3-way valve, allowing letdown to divert to the VCT or to the liquid radioactive waste system. In automatic, the valve positions to divert letdown to the liquid radioactive waste system when VCT levelincreases to 78 percen Tags on the control board indicated that the manual block valve in the flowpath to the liquid radioactive waste system, Valve 2MU924, was closed, and cautioned operators not to divert with the block valve closed. The block valve was closed on the previous shift because Valve 2LV0227A had been determined to be leaking by when in the VCT position.

.

.

i

.

l l-14-Operating Instruction SO2 2.1, Revision 15, " Chemical Volume Control System Charging and Letdown," S ,,6.3.9.1, states that if desired, "due to Valve LV0227A leak by, then close the block valve, place the Valve LV0227A handswitch in Manual, and document the closure by placing caution tags on the block valve handwheel and the Valve LV0227A handswitch." Operators had not reviewed the procedure when they closed Valve 2MUO92 Operators had not identified the condition, which was evident by observing the control boards, during the shift turnover or during routine monitoring of the control board indication The inspectors had previously observed a similar situation in Unit 3, documented in NRC inspection Report 50-361; 362/96 15. As previously documented, the inspectors determined that, during steady-state operations, operators generally controlled RCS inventory, such that additions to the RCS did not result in VCT level changes, by simultaneously diverting letdown to radwaste. Operators did not normally rely on the automatic VCT water level control capability, and Valve 2LV0227A would not normally shift position automatically without a plant transient causing an approximate 8'F elevation in RCS temperature or a control aystem f ailur Because of the licensee's attempts to effectively repair Valve 2LV0227A, operators have had to leave the block valve closed most of the time, opening it as necessary to control VCT level. The previous corrective action, which was to incorporate guidance from an abnormal alignment into the permanent procedure, was generally effective. In this case, the corrective action was not effective because it was not used. The failure to align Valve 2LV0227A as specified in Operating instruction SO23-3-2,1 is a violation of TS 5.5.1.1.a for failure to follow procedures (VIO 361/97017-03).

In response to this violation, the licensee counseled the individuals involved, briefed all Operations crews, and initiated a revision to the procedure in eliminate the requirement to place Valve LV0227A in martual when the block valve was closed, Conclusions A violation was identified by the inspectors when operators failed to follow procedures for alignment of the VCT inlet diversion valve. This was a repeat violation. Additionally, the operators demonstrated a weakness in attention to detail in monitoring the control boards during shift turnover and routine monitoring activitie .7 Conclusions on Operator Knowledae and Performance Operators generally performed well during the many infrequently-performed evolutions that they accomplished during this inspection period. However, several

. _ .

[,

'

.

-15

lapses were observed by the inspectors in attention to detailin the use of procedures, communications practices, the communication and reinforcement of operator expectations, and in the knowledge of plant power indications and plant response with a near-zero IT Quality Assurance in Operations 07.1 NOD Midlooo Activities Insoection Scoce (71707)

The inspectors reviewed NOD Surveillance Report SOS-052-97, "Both Unit High Risk Evolution Surveillance Report," dated July 15,1997, Observations and Findinos i

The surveillance report documented NOD oversight activities during the condition in which both units were in a high risk evolution (midloop).

The NOD observers provided a recommendation to the operating crew to refill the RCS level when Diverse Level Monitoring System inaccuracies were identified in Unit 3. The recommendation was accepted by Operations and the RCS level was increase An NOD observer recommended a contingency plan be developed for the Unit 3 nozzle work should an emergency reflood of the RCS be required. Nuclear l Construction accepted the task and developed a pla l

!

-j An NOD observer recognized a conflict between management guidanco and '

procedure direction regarding the fill rate when raising Unit 2 RCS level from 18 inches and out of midloop. The shift superintendent insured the most conservative procedure guidance was use Conclusions 1

]

NOD observers provided valuable recommendations and insights to the operating j crews during the concurrent midloop activitie !

l

.

.

i l

.

e

<

.

-16-II. Maintenance M1 Conduct of Maintenance M 1.1 General Comments Inspection Scope (62707)

The inspectors observed all or portions of the following work activities:

Liquid dye penetrant testing of pipe ends prior to welding of new auxiliary spray Check Valve 2MUO19 (Unit 2)

Welding of new charging Pump 2MUO21 to Loop 1 A check valve (Unit 2)

• Adjust diverse level monitoring system primary reference constants (Units 2 and 3) l Observations and Findinas The inspectors found the work performed under these activities to be thorough. All work observed was performed with the work package present and in active us Technicians were knowledgeable and professional. The inspectors frequently observed supervisors and system engineers monitoring job progress, and quality control personnel were present whenever required by procedure. When applicable, ;

appropriate radiation controls were in plac '

M1.2 General Comments on Surveillance Activities Insoection Scoce (61726)

The inspectors observed all or portions of the following surveillance activities:

• High Pressure Stop and Governor Valve Testing (Unit 2)

• Water inventory Balance (Unit 3) Observations and Findinas The inspectors found all surveillances performed under these activities to be thorough. All surveillances observed were performed with the work package present and in active use. Technicians were knowledgeable and professional. The inspectors frequently observed supervisors and system engineers monitoring job progress, and quality control personnel were present whenever required by procedure. When applicable, appropriate radiation controls were in plac . ~- - .. - . ._ .. _ - - - - - - . _

.

.

.

,

- 17-In addition, see the specific discussion of a surveillance observed under Secticn 04.4, above.

M4 Maintenance Staff Knowledge and Performance M4.1 Radio 'in EDG Control Cabinet - Unit 2 i insoection Scone (62707 and 71707)

On July 23,1997, the inspectors observed a WIN Maintenance technician replacing the weather stripping on the cabinet door to the EDG 2G002 local control cabinet, ,

,

2L16 I Observations and Findinas A large portable AM/FM radio (approximately 6"x6"x18") was inside the cabinet, between the rows of relays, suspended about 3 feet off the floor by duct tape fastened to each side of the cabinet interior. The radio was plugged into a receptacle inside the cabinet. The inspectors asked if the EDG was operable, and the technician replied that it was and that the weather strip repair would not affect the EDG. When asked if the radio being inside the cabinet had been considered in the approval for the work, the technician replied that it had not. The inspectors stated a concern for the potential impact of the radio on the EDG controls, should the radio fall. The technician then untaped the radio and sat it on the floor inside the cabinet. The inspectors informed Maintenan:o management and Operations supervision of the observatio Maintenance management immediately recognized the significance of the observed condition and began an investigation. The licensee determined that the technician, an electrician, was not experienced in performing work at nuclear facilities, having been on site for about 4 months. Additionally, the technician was not a qualified journeyrnan, as the licensee's program for WIN employees requires. The licensee prohibited the entire WIN team from performing work for 2 days to provide refresher training, and removed the technician from the WIN team. Additionally, the licensee temporarily relieved the two WIN supervisors from their supervisory positions pending the completion of remedial training. The licensee's investigation was documented in AR 97070119 The licensee's seismic controls program, documented in Maintenance Procedure SO123-1-1.20, did not prohibit the observed condition because the technician was in attendance of the radio while it was in the cabine Conclusions The placement of the radio inside the EDG local control cabinet compromised the seismic qualification of the EDG and did not comply with the licensee's

-

.-

.

-18-management expectations with respect to professional work spaces. Additionally, the technician did not understand the potential risk to the EDG as the result of his actions. The licensee's inadvertent assignment of an inexperienced electrician to the WIN team, and the f ailure to provide adequate supervision, were isolated weaknesses in the licensee's maintenance program. The licensee's response to the situation was excellent, MS Maintenance Staff Training and Qualification M S.1 Qualification of SSID Machinists a, insoection Scope (62707)

The inspectors reviewed Procedure SO123-XXI-1.11.13, " Mechanical Maintenance Training Program Description;" American National Standards Institute N18.1,

" Selection and Training of Nuclear Power Plant Personnel;" MOs 97061153, 96051328, 97061681, 97061274, 97020949, 97042057, and 97050396; and the qualifications of SSID machinists that performed work on safety-related equipment during the recent Unit 3 outage, Observations and Findinas The inspectors reviewed the MOs that were performed by six different SSID machinists. The SSID machinists were Southern California Edison employees not normally assigned to the site and were used to augment the normal staff during outages. The machinists performed the work in Unit 3 and in the south yard machine shop. Five of the machinists were independent worker qualified and one was not. The maintenance activity performed by the worker that was not independent worker qualified was properly supervised. in addition, the licensee at times supervised the independent workers, although oversight was not required, Cqnclusions The licensee used properly trained SSID machinists to perform work during the Unit 3 refueling outage or supervised the machinists that were not independent worker qualifie M8 Miscellaneous Maintenance issues (92902 and 90712)

M8.1 (Closed) Unresolved item 362/97012-01: failure to properly couple Unit 3 four-fingered CEA 9 The inspectors reviewed Root Cause Evaluation 97-002, dated July 14,1997; portions of vendor technical manuals " Reactor Vessel Internals Instruction Manual,"

dated July 6,1984, and " Reactor Internals Lift Rig Manual," dated March 6,1981; and interviewed Maintenance supervisory personnel,

i

.

-19-This item involved the licensee's discovery after reactor reassembly that CEA 91 had not been properly coupled prior to reactor startup. Previously, orientation guides in CEA 91's extension shaft housing had been damaged and consequently removed. The licensee coupled CEA 91 to the extension shaft, with the CEA in the core and the upper guide structure installed. Four-fingered CEAs do not have alignment guides that facilitate proper positioning of the extension shaft relative to thc hub assembly. The licensee had a procedure in place to couple four-fingered CEAs outside the core using a camera to ensure proper coupling. The licensee had not previously coupled a four-fingered CEA in the core. The vendor manuals provided step-by-step guidance for coupling four-fingered CEAs in the core, with the upper guide structure installed, but the licensee did not use this guidance. The vendor manuals called for the use of an extension shaft alignment tool, which was to be inserted into the extension shaft housing past the orientation guides. When the licensee had welded in new orientation guides an insert had been used. The two new guides were welded to the insert, and the insert welded to the interior of the extension shaf t housing. The insert decreased the interior diameter of the extension shaft housing, and the extension shaft alignment tool no longer had sufficient clearance to be inserted in the extension shaft housing, and was not use Prior to repairing the damaged guides, the licensee had contacted the vendor via telephone. The licensee stated that the vendor had said that it was acceptable to couple CEA 91 in the core, in the manner planned. The inspectors found, consequently, that the vendor's statement seemed contradictory because the vendor's written guidance was to use the extension shaft alignment too CFR Part 50, Appendix B, Criterion V, states that activities affecting quality shall be prescribed by instructions appropriate to the circumstances. Contrary to this, MO 9704212000, directing that CEA 91 be coupled in the core, was not appropriate to the circumstances. The MO did not incorporate either established, written method of coupling a four-fingered CEA, either outside the core using a camera to verify coupling, or in the core using an extension shaft alignment tool to facilitate proper alignment of the extension shaft and the spider hub. The method used, which was essentially the method for coupling five-fingered CEAs, was cautioned against being used for four-fingered CEAs in both the licensee procedure and in written vendor guidance (VIO 362/97017-04).

The licensee's root cause report described corrective actions, including establishing

procedural requirements to visually verify proper coupling of iour-fingered CEAs

'

performed in the reactor, a review of maintenance procedures for CEA coupling, and training refueling personnel regarding this occurrence. Since these corrective actions appeared adequate to prevent recurrence of this violation, no response is

'

required. Full compliance was achieved on July 14,1997, the date of the root cause report. As of that date, the licensee had documented an intention to not perform four-fingered CEA coupling in the manner prescribed by MO 9704212000.

.

__

. _ _ . _ .

4

, .

?

,

M8.2 LClosed) LER 361/97011-00 and 361/97011-01: wrong boron sample used for SDM surveillance. This issue was discussed in Section 01.2. The LER was revised to clarify the . method of discovery and the discovery dat Ill. Enaineerina E4 Engineering Staff Knowledge and Performance E incorrect NDE Acceptance Criteria Used for ASME Code Weldina Repairs Units 2 and 3 a. . Irtspection Scooe (37551)

On July 8,1997, the inspectors observed a liquid dye penetrant examination of a weld preparation. 'he preparation area was the two pipe ends involved in the welding in of , new Unit 2 charging to the reactor coolant Loop 1 A injection check Valve 2MUO19. The inspectors reviewed WR 2-97-374 and MO 97070201000, both used for the NDE observed. The inspectors also reviewed Procedure NDEP-PT-007, Revision 0, " Liquid Penetrant Examination;" portions of the ASME

, Code, Section lit, no addenda,1992 edition; ASME Code Case N-4161; AR 970701400 and 970701401 (generated as a result of the inspectors' findings);

WRs 2 97 375,347,348,338,339,328, and 329; and associated records of NDE results. The WRs were for similar welding of the Unit 2 auxiliary spray check valve, Valve 2MUO21, and the two equivalent Unit 3 valves (3MUO19 and 3MUO21).

. Observations and Findinas

!

The pipe ends were of Schedule 160 2-inch pipe with a nominal 0.344-inch wall thickness. A visible, solvent, removable penetrant with a nonaqueous developer was used for the examination.

'

All time requirements stipulated in Procedure NDEP-PT-007 were met, and the surface was properly prepared. The Quality Control technician performing the NDE demonstrated good technique in applying and removing the penetrant and developer. No relevant indications were observe The acceptance standard for the penetrant testing of the weld preparation, as well as for the final weld and pipe examination, was listed in WR 2-97 374 as " ASME 111, NB 5350,92 Ed, No Add." Subsection NB 5140 of the ASME Code requires that the weld, and % inch of the pipe adjacent to the weld, be tested in the final examination. Penetrant testing of the weld preparation was not required, in

'

addition, Subsection NB 5140 states that the acceptance standards for the weld shall be as stated in Article 5100, while the acceptance standards for the base material shall be as stated in Subsection NB 2500. The inspectors determined that Subsection NB 5350 was the liquid penetrant acceptance standard for the welding of Quality Class I components, and listed as " unacceptable" any cracks or linear

.- . _ _. __ _ _ _ .

.

..

l

-21-indications, as well as any rounded indications with dimensions greater than-3/16-inch, and as such, the standard for the weld itself, not the pipe ends. The inspectors questioned the Welds and Codes supervisor as to why Subsection NB 5350 was used instead of Subsection NB 5130 (the acceptance standard for weld edge preparation), and were informed that Subsection NB 5130 allowed linear indications. The Codes and Welding engineers stattJ that linear indications were undesirable in the pipe ends and not permissible in the weld. The engineers were concerned that subsequent radiography might indicate unacceptable weld indications if linear indications appeared in the pipe end The inspectors determined that Subsection NB 5130 was not applicable to base metal thicknesses of less than 2 inches, and that Subsection NB 5140 referred the reader to Subsection NB 2546.3 for acceptance standards of Class I pipe with a wall thickness of less than 2 inches. In addition, for pipe wall thicknesses of less than 5/8 inch, Subsection NB 2546.3 did not allow rounded indications greater than 1/8 inch. The inspectors determined, however, that Subsections NB 5350 and 5130 both allowed up to 3/16 inch rounded indications. The inspectors found, consequently, that the licensee was using a nonconservative acceptance standard for rounded indications in the final (required) examination of the pip The licensee was using Code Case N 416-1, and the 1992 edition of the ASME Code, per authorization dated January 11,1995, from the NRC Office of Nuclear Reactor Regulation. ASME Code Case N-416-1 stipulated that the NDE requirements of the 1992 edition be used. The licensee did not properly identify those requirements Nis is a violation of 10 CFR Part 50, Appendix B, Criterion IX, which states that welding and NDE shall be accomplished with qualified procedures in accordance with applicable codes and specifications (VIO 361; 362/97017-05).

In response to the inspectors' observation, the licensee reviewed all records of postweld NDE performed that had used ASME Code Case N 416-1, which required the 1992 edition of the Code. The licensee subsequently informed the inspectors that other editions of the Code used were evaluated properly for the acceptance standards; it was only the change to the 1992 edition which had been improperly evaluated. The ASME Code Case was used for replacement of Valves MUO19 and MUO21 in both Units 2 and 3, as well as for weld jobs associated with modifications to the reactor level monitoring system. In allinstances, no relevant indications were detected during the NDE. Relevant indications were listed as all indications greater than 1/16-inch Consequently, no additional performance of NDE was require Conclusions The observed NDE performance was good. The acceptance standard listed in WRs for liquid dye penetrant examination of pipes with a wall thickness of 5/8-inch or less was not correct, and, for rounded indications, was not conservative. This was a violation of 10 CFR Part 50, Appendix B, Criterion I .- .

..

.

22-

-E8 Miscellaneous Engineering Issues (92903,90712)

E8,1 (Closed) l_nspection Followoo item (IFI) 50 361:362/96002-03: erroneous Updated Final Safety Analysis Report (UFSAR) statements concerning auxiliary feedwater pump controls (Closad) IFl 50-361:362/96003-03_: erroneous UFSAR statements concerning discrete analysis capabilities (Closed) IFl 50-361:362/97002-05: erroneous L73AR statements concerning core reload check These items involved the inspectors' identification of three examples in which the UFSAR was not accurate in describing a plant hardware capability or a human factors process control. As corrective actions for each example, licensee personnel performed a screen, or evaluation, in accordance with 10 CFR Part 50.59, determined that no unreviewed safety question existed, and submitted a "UFSAR Change Request" to the Plant Licensing manager. The inspectors reviewed the three change requests, SAR 2/3-444,446, and 520, and determined that they were sufficient to correct the inaccuracies in the UFSAR. The inspectors also found that these three'. changes did not represent unreviewed safety question E8.2 (Closed) LER 361/97001-01: surveillance not current upon improved TS implementation. This issue was discussed in NRC Inspection Report 50-361: 362/97-1 E8.3 Failure of Containment isolation Valve 3HV6373 - Unit 3 (37551)

On July 17,1997, Unit 3 operators performed a quarterly inservice test of motor operated containment isolation Valve 3HV6373. This valve is located outside containment on the component water cooling outlet for containment emergency cooling Unit 402 and is within the scope of NRC Generic Letter 89-10. The valve f ailed to respond to a "close" signal from the control roo_m, and was declared inoperable. Operators removed power from the actuator, with the valve in its open-

. position, in accordance with Unit 3 TS 3.6.3. The inspectors reviewed the root cause determination in AR 970700860 and interviewed cognizant engineer Valve 3HV6373 has a Limitorque actuator controlled by a Square D Class 8736 reversing line starter. The line starter was replaced and the old starter was evaluated. The "open" contact in the starter was found to have chemical deposits -

that prevented current flow, consequently preventing the valve from receiving an open signal when the "open" contact was closed. The chemical deposits were analyzed and found to be a silicon-based oil, which the licensee surmised came from an agent used while forming the plastic case the contact was housed in. Over time, the oil migrated to the contact, oxidized, and insulated the contact. Licensee engineers found no other similar failures in the maintenance data base. Engineers l

o

.

-23-also examined two other line starters, with the same 1982 date code as the failed starter, but found no evidence of chemical buildup. Consequently, the licensee concluded that the f ailure was not applicable to other line starters and isolated to a manufacturing defect of this particular starter.

.

Valve 3HV6373 was subsequently returned to service and succes; fully tested.

I The inspectors found that the licensee performed a thorough root cause determination, and that the conclusions reached in the root cause were rigorou IV. Plant Suppo R1 Radiological Protection and Chemistry Controls R 1.1 Radioloaical Housekeepina (71750)

During routine plant tours following the completion of the outages in Units 2 and 3, the inspectors found that general housekeeping in the radiologically controlled area was very good. Materials storage areas were generally neat, and contaminated areas that no longer had ongoing work activities were returned to their preoutage boundarie R1.2 Unoosted loose Surface Contamination Area - Unit 3 (717501 On July 21,1997, the inspectors observed that wet boric acid was built up around i the valve stem, due to packing leakage, on Unit 3 Letdown Pressure Control Valve 3PV02018. A licensee Health Physics technician surveyed a small portion of the buildup, and found 5,000 disintegrations per minute (dpm) of activit However, this was not a violation of requirements because there was no specific licensee programmatic requirement, or NRC requirement, to post loose surface contamination as long as radiation exposure was maintained as low as reasonably achievable. Licensee management expectation was that any activity over 1000 dpm/100 cm 2would be posted as a contaminated area. In response, the licensee posted the area as contaminate The inspectors found that the response to their concerns was good, but that, in this instance, licensee Health Physics personnel were not sufficiently timely in identifying and posting loose surface contamination in excess of limits. A similar finding, involving a miscellaneous waste evaporator sample point, was made on February 10,1997, and documented in NRC inspection Report 50-361; 362/97-0 O

,

24-V. Nianaaement Meetinos X1 Exit Meeting Summary The inspectors presented the inspection results to members of licensee management at the exit meeting on August 20,1997. The licensee acknowledged the findings presente The inspectors asked the licensee whether any materials examined during the inspection should be considered proprietary. No proprietary information v<as identifie a

.o

.

\

l A TTACHMENT SUPPLEMENTAL INFORMATION PARTIAL LIST OF PERSONS CONTACTED Licensee D. Brieg, Manager, Station Technical J. Clark, Manager, Chemistry J. Fee, Manager, Maintenance G. Gibson, Manager, Compliance D. Herbst, Manager, Site Quality Assurance J. Madigan, Manager, Health Physics (Acting)

R. Krieger, Vice President, Nuclear Generation D. Nunn, Vice President, Engineering and Technical Services T. Vogt, Plant Superintendent, Units 2 and 3 R. Waldo, Manager, Operations INSPECTION PROCEDURES USED IP 37551: Onsite Engineering IP 61726: Surveillance Observations IP 62707: Maintenance Observations IP 71707: Plant Operations IP 71750: Plant Support Activities IP 90712: Inoffice Review of LER IP 92902: Followup - Maintenance IP 92903: Followup - Engineering ITEMS OPENED AND CLOSED Ooened 50 361/97017-02 VIO five percent per hour ramp rate exceeded with failed fuel pin 50-361/97017-03 VIO VCT inlet diversion valve in incorrect mode 50-361: 362/97017-05 VIO incorrect NDE acceptance criteria used Opened and Closed 50-362/97017-01 NCV inadequate SDM verification 50-362/97017-04 VIO failure to properly couple Unit 3 four-fingered CEA 91

4

.

e 2-Closed 50-361; 362/96002-03 IFl erroneous UFSAR statements concerning auxiliary feedwater pump controls 50-361: 362/96003-03 IFl erroneous UFSAR statements concerning discrete analysis capabilities 50-361; 362/97002-05 IFl erroneous UFSAR stateme its concerning core reload checks 50-362/97012 01 URI failure to properly couple Unit 3 four-fingered CEA 91 50-361/97-011-00 LER wrong boron sample used for SDM surveillance 50-361/97-011-01 50-361/97-001-03 LER surveillance not current upon improved TS implementation LIST OF ACRONYMS USED ACO assistant control operator AR action request CEA control element assembly COLSS core operating limits supervisory system CRS control room supervisor dpm disintegrations per minute EDG emergency diesel generator HJTC heated junction thermocouple ITC integrated temperature coefficient LER licensee event report MO maintenance order NDE nondestructive examination NOD Nuclear Oversight Division 6-PDR Public Document Room RCS reactor coolant system SDM shutdown margin SR surveillance requirement SSID Shop Services and instrumentation Division TS Technical Specification UrSAR Updated Final Safety Analysis Report VCT volume control tank WIN Work-it-Now WR weld record