IR 05000275/1990030
| ML16342B748 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 02/28/1991 |
| From: | Morrill P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| To: | |
| Shared Package | |
| ML16341F999 | List: |
| References | |
| 50-275-90-30, 50-323-90-30, NUDOCS 9103200082 | |
| Download: ML16342B748 (60) | |
Text
U.
S.
NUCLEAR REGULATORY COMMISSION
REGION V
Report Nos:
50-275/90-30 and 50-323/90-30 Docket Nos:
50-275 and 50-323 License Nos:
Pacific Gas and Electric Company 77 Beale Street, Room 1451..
San Francisco, California 94106 Facility Name:
Diablo Canyon Units 1 and 2 Inspection at:
Diablo Canyon Site, San Luis Obispo County, California Inspection. Conducted:
December 16, 1990, through February 2, 1991 Inspectors:
P.
P. Narbut, Senior Resident Inspector K.
E. Johnston, Resident Inspector B. J.
Olson, Project Inspector Approved by:
orr>
,
ie
,
e ctor rogects ection ate igne Summary:
Ins ection from December
1990 throu h Februar
1991 and Februar
1 e ort Nos.
an
"
Areas Ins ected:
The inspection included routine inspections of plant opera sons, maintenance and surveillance activities, follow-up of onsite events, open items, and licensee event reports (LERs),
as well as selected independent inspection activities.
Inspection Procedures 30703, 35502, 37701, 40500, 61726, 62703, 64704, 71707, 90712, 92700, 92701, 92702, 92720, and 93702 were used as guidance during this inspection.
Safet Issues Mana ement S stem (SIMS Items:
None Results:
General Conclusions on Stren th and Weaknesses:
The licensee showed indications of strengths or weaknesses in the following areas:
91032000S2 910304 PDR ADOCK 05000275 G
En ineerin
.The engineering area demonstrated three examples of poorly managed engineering work.
.In one case, the applicability to Diablo Canyon of main steam check valve problems at another facility was misanalyzed as not being applicable (paragraph 8e);
and in another case, an inaccurate engineering analysis was performed in response to a notice of violation regarding the fuel handling building ventilation. system..
The analysis
.
was qualitative in nature and did not duplicate the conditions at the time of the violation (paragraph 8d).
A third example was the failure to followup on the engineering recommendation to disa'ssemble an important valve which had been analyzed as overstressed (paragraph 4j).
0 erations Auxiliary operators caused two incidents which were contrary to clear operations policy documents.
In both cases, the auxiliary operators were not using their procedures.
The first event (paragraph 4d) caused the brief loss of both residual heat removal pumps when control power was erroneously removed to both pumps on January 3, 1991.
In the second event (paragraph 4g) an auxiliary saltwater gate was closed on the wrong unit on January 6, 1991.
Additionally, just prior to the reporting period covered by this report and again just after this reporting period, operations shift supervision, personnel experienced reportable events and did not notify their management or the NRC inspectors prior to resuming operations.
In the first case, the event was a feedwater isolation on December 8, 1990, and in the second case, the event was a brief loss of both residual heat removal pumps on February 8, 1991.
Management policy in all the above incidents was clear and operations management's actions in response were clear and unequivocal.
Nonetheless the occurrences were disappointing and worthy of discussion to prevent repetition.
Untimel Actions Additional examples of untimely actions were noted.
Specifically, the licensee was untimely in resolving problems with the steam admission valve to the steam driven auxiliary feedwater pump (paragraph 4j, problems had first occurred in January 1989).
Likewise the licensee was untimely in resolving hardware problems, some since 1985, on the vibration and loose parts monitor (paragraph 4h) which subsequently led to engineering inattention to increasing. alarms for several months in 1990 and 1991.
Additionally, the licensee failed to take timely action to repair an important auxiliary saltwater system valve operator (paragraph 4k).
Although the problem was identified in July 1990, repairs were not completed until January 1991.
In all three examples, NRC inspector attention preceded corrective action Si nificant Safet Matters:
None.
Summar of Violations and Deviations No violations were identified.-
Two deviations were identified.
One deviation (paragraph 8d) dealt with the failure to install warning labels on fuel handling building boundary doors to prevent them from being left open and violating Technical Specification negative pressure requirements.
The door labels had been committed to in response to a previous violation.
The second deviation (paragraph 4h) dealt with the failure since 1985, to maintain the design features described in the FSAR for the vibration and loose parts monitor.
0 en Items. Summar
Five open items were identified.
Eleven open items were close Persons Contacted.
DETAILS
~
"J.
D. Townsend, Vice President, Diablo Canyon Operations 8 Plant Manag'er
- D. B.
Miklush, Assistant Plant Manager, Operations Services M. J.
Angus, Assistant Plant Manager, Technical Services
- B.
W. Giffin, Assistant Plant Manager, Maintenance Services
- W.
G. Crockett, Instrumentation and Controls Maintenance Manager
"W.
D. Barkhuff, equality Control Manager
"T. A. Bennett, Mechanical Maintenance Manager D. A. Taggert, Director equality Support
- T. L. Grebel, Regulatory Compliance Supervisor
"H. J. Phillips, Electrical Maintenance Manager
- J.
S. Bard, Work Planning Manager R.
C. Washington, Instrumentation and Controls Manager
- J.
A. Shoulders, Onsite Project Engineering Group Manager
- C.
R. Groff, Acting System Engineering Manager S.
R. Fridley, Operations Manager R. Gray, Radiation Protection Manager E.
C. Connell, Assistant Project Engineer
"J.
A. Sexton, equality Assurance Manager
"J.
B. Hoch, Nuclear Safety and Reliability Assessment Manager
- J. Griffin, Regulatory Compliance Engineer
- D. Cosgrove, equality Control Specialist
- R.
C. Anderson, Nuclear Engineering and Construction Manager
- W. Fujimoto, Vice President, Nuclear Technical Services
"B.
H. Patton, Reliability Engineering Manager
"D. Oatley, Assistant Plant Manager, Support Services The inspectors interviewed several other licensee employees including shift foremen (SFM), reactor and auxiliary operators, maintenance personnel, plant technicians and engineers, quality assurance personnel and general construction/startup personnel..
"Denotes those attending the exit interview.
0 erational Status of Diablo Can on Units 1 and
The units both started the reporting period at 100K power.
Both units were reduced to 50K power for condenser cleaning several times during the period.
In addition to the reactor trip of December 5, 1990, which occurred just prior to the reporting period, Unit 1 experienced two additional reactor trips during the reporting period.
One occurred on December 24, 1990, and involved a depressurization due to a stuck open pressurizer spray valve, a consequent safety injection and cooldown in excess of technical specification allowable values caused by an unrelated fai lure of a condenser steam dump valve.
The second trip occurred on February 1, 1991, and was caused by scaffolding workers, who closed an air valve to the air operators for feedwater regulating valves.
The feedwater valves went shut causing a reactor trip on low steam generator level.
As a result of the trip, Unit 1 started it's fourth refueling outage two days earlier than planned.
Unit 2 ended the reporting period at full power in a record continuous power run of 273 day.
0 erational Safet Verification 71707 a.
General During the inspection period, the inspectors observed and examined activities to verify the operational safety of the licensee's facility.
The observations and examinations of those activities were conducted on a daily, weekly or monthly basis.
On a daily basis, the inspectors observed control room activities to verify compliance with selected Limiting Conditions for Operations (LCOs)
as prescribed in the facility Technical Specifications (TS).
Logs, instrumentation, recorder traces, and other operational records were. examined to obtain information on plant conditions and to evaluate trends.
This operational information was then evaluated to determine if regulatory requirements were satisfied.
Shift turnovers were observed on a sample basis to verify that all pertinent information of plant status was relayed to the oncoming crew.
During each week, the inspectors toured the accessible
'areas of the facility to observe the following:
(a)
General plant and equipment conditions.
(b)
Fire hazards and fire fighting equipment.
(c)
Conduct of selected activities for compliance with the licensee's administrative controls and 'approved procedures.
(d)
Interiors of electrical and control panels.
(e)
Plant housekeeping and cleanliness.
(f)
Engineered safety feature equipment alignment and conditions.
(g)
Storage of pressurized gas bottles.
The inspectors talked with operators in the control room, and other plant personnel.
The discussions centered on pertinent topics of general plant conditions, procedures, security,. training, and other aspects of the work activities.
Fire Door Labels The inspector reviewed the licensee's program for upgrading labeling material on plant fire doors to material which maintains the Underwriters Laboratory (UL) fire qualification for the door.
The licensee was in the process of revising their program for labeling doors.
It was recognized that the material currently used for door labeling had not been tested and approved for use on fire barrier doors.
An effort has been made to identify material qualified for use on fire door b.
Three Lexan products were tested by the UL for the licensee:
Electromark Series 1000, 2000, and 5000.
Both the Series 1000 and
, 2000 materials were found acceptable for use on 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> fire doors.
The Series 5000 material was found to be unacceptable.
Additionally, different adhesives were tested for use.
=
At the time of this inspection, the licensee was proceeding with a program to use the Electromark Series 2000, a self adhering product, for th'eir sign upgrade program.
The inspector discussed the lack of a qualificat>on for sign materials currently installed on fire doors with the Emergency and Safety Services Hanager.
He stated that the interim configuration was acceptable for two reasons:
o The licensee has complied with the Technical Specification action statement for'inoperable fire doors by having a
continuous roving fire watch and operable fire detectors.
o The potential for. combustion was determined to be low based on current fire loadinq and the amount of heat necessary to ignite a sign on the opposing side of the door.
The inspector found this to be acceptable.
Radiolo ical Protection C.
The inspectors periodically observed radiological protection practices to determine whether the licensee's program was being implemented in conformance with facility policies and procedures and in compliance with regulatory requirements.
The inspectors verified that health physics supervisors and professionals conducted frequent plant tours to observe activities in progress and were aware of significant plant activities, particularly those related to radiological conditions and/or challenges.
ALARA considerations were found to be an integral part of each RMP (Radiation Mork Permit),
Ph sical Securit (71707 Security activities were observed for conformance with regulatory requirements, implementation of the site security plan, and administrative procedures including vehicle and personnel access screening, personnel badging, site security force manning, compensatory measures, and protected and vital area integrity.
'Exterior lighting was checked during backshift inspections.
Ho violations or deviations were identifie.
Onsite Event Follow-u 93702)
~
~
~
a.
Or anizational Chan e
On December 19, 1990, the licensee informed the resident inspectors of the assignment of Warren J.
Fujimoto to a newly. created position of Vice President Nuclear Technical Services.
b.
Unit 1 Reactor Tri P
On December 24, 1990; Unit 1 experienced a reactor trip, safety injection, and overcooling event.
The resident inspector responded to the event and observed, recovery actions and evaluated licensee plans for investigation of the event and restart.
The initiating event was the failure of the position feedback mechanism for a pressurizer spray valve due to the mechanism feedback arm becoming disconnected from the pressurizer spray valve.
The condition resulted in the spray valve going full open causing depressurization and eventually safety injection.
Subsequently, an RCS cooldown occurred at greater than 100 F per hour (the technical specification limit).
This was determined to be due to steam leaks and a partially stuck open steam dump valve to the main condenser.
The valve was later found to have a broken stem.
More details regarding the trip are provided in the Licensee Event Report (LER) 1-90-017.
0 er ator Res onse The inspectors evaluated the operators'esponse to the event and considered the response to be properly cautious, actions were deliberate and thought out fully.
The event was complex and not easily understood because the spray valve closed upon isolation of air to the containment (which occurs with safety injection) and then reopened when safety injection was reset.
The added independent event of excessive cooldown was cautiously evaluated and then acted upon by shutting the main steam isolation valves.
Initial Mana ement Res onse Licensee management invoked their procedure for event investigation and initially charted a comprehensive action plan to analyze the event and its causes.
In addition to the event related actions, the licensee planned to perform unrelated previously deferred work during the unplanned outage.
This work included some work on the feedwater components which had caused a reportable event (feedwater isolation)
on December 8, 1990.
Specifically, the work on feedwater components included:
o Disassembly and proper reassembly of main feedwater check valve FW 531.
The valve had been improperly assembled in the last refueling outage in 198 o Disassembly of feedwater regulating bypass valve FCV-1530.
o Partial controls calibration of feedwater regulating valve FCV-530.
r The initial management response was considered go be good by the inspectors.
The.investigative actions led to the discovery of the disconnected feedback linkage on the spray valve.
The broken steam dump valve stem was found during the routine post-trip steam dump visual inspection actions.
The cooldown rate experienced was examined on the simulator and found to be closely duplicated with a steam dump partial failure.
Subse uent Mana ement Res onse and Restart of the Unit Although the residents and the region identified no issues which would technically justify holding plant restart, the licensee actions to examine problems that occurred during the event appeared to be an example of a less than thorough restart review as is explained below:
o Pressurizer S ra Valve The licensee determined that the pressurizer spray valve failed open when the control system feedback linkage became disconnected.
The licensee determined that the feedback linkage became disconnected because a locking device was not installed on the screw holding the linkage to the valve stem.
The inspectors considered the licensee's inspection for other similar missing locking devices to be narrow.
Specifically, the other three spray valve feedback linkage connections were inspected by the licensee, but other important linkage connections (this particular controller is commonly used at Diablo Canyon) were not inspected prior to questioning by the inspectors.
Further, this inspection was apparently performed without precise acceptance criteria since the licensee found and then accepted, during the above inspection, that there were varied configurations of locking devices (double nuts, star nuts, one nut with an elastic stop nut, one elastic stop nut, etc.).
Additionally, the licensee concluded that a locking device on the feedback screw was not specified by instructions and that it had been up to the craft to determine the locking devices necessary.
The inspectors disagreed and concluded that an elastic stop nut was specified in the vendor manual for the positioner.
o Condenser Steam Dum Failure (PCV-1)
PCV-1 failed open when its stem broke inside the valve body.
The licensee did not fully understand the cause of the failure of PCV-l.
In addition, the licensee only inspected three of eleven other steam dump valves to verify their proper functioning.
One of the other steam dumps apparently showed a
number of symptoms exhibited by PCV-l, but was not inspected further.
Management had erroneously assumed that the I8C department would inspect all other valves upon the discovery of PCV-1's broken stem.
o
. Main Feedwater Re ulatin Valve B
ass Valve FCV-1530 FCV-1530 was disassembled and inspected during the outage.
The pre-disassembly investigation was apparently poorly performed.
The valve controller was riot allowing the valve to fully close on a close signal, but this was not tested until after valve disassembly.
As a result, the valve disassembly was probably not necessary.
Following reassembly, FCV-1530 would not stroke open.
Upon a second disassembly, the licensee found that a
valve component had deformed such that the valve plug could not fully open.
This happened a second time during reassembly following the machining of the deformed component.
The licensee consequently installed a stem travel stop to limit the stroke of the valve and thereby prevent travel in the deformed area.
questions by the inspector caused the licensee to perform an engineering operability analysis of this configuration prior to restart.
The licensee assumed the torquing of the bonnet nuts caused the valve internals to deform:
Although torquing had been performed in accordance with vendor manual requirements, subsequent vendor updates (which the licensee did not have) required 25K less torque.
The inspectors requested the licensee to perform some scoping calculations to-provide.some assurance that their theory (torque)
was at least capable of the deformation seen and that alternate possible theories (bad material) did not need to be pursued.
The inspectors also asked the licensee to determine why the updated technical manual was not available at the site (see the maintenance section of this report for further details regarding out-of-date vendor technical manuals).
Conclusion The licensee generally handled the initial analysis and recovery from this event well.
As described above however, actions after the initial action planning lacked comprehensiveness and followthrough.
Unit 1 Containment Ventilation Isolation On December 27, 1990, a Unit 1 containment ventilation isolation (CVI) was inadvertently initiated.
The event was considered an Engineered Safety Features actuation and a four hour non-emergency report was made to the NRC.
Only the containment noble gas and air particulate sample return valve, FCV-681, closed as a result of the CVI.
All other valves which receive a close signal were already closed.
While performing a design modification in a radiation monitor cabinet in the control room, an electrician contacted a pair of
pliers with the terminals of a fuse block for a radiation monitor causing the vital power inverter to..experience a voltage transient.
The voltage transient caused RM-ll and RM-12, the containment. noble gas and air particulate monitors,.to spike, resulting in a CVI.
This event is described in LER 50-275/90-18, dated January 28, 1991.
The inspector reviewed the cause and corrective actions descry~bed in the LER and found them to be acceptable.
This item is closed.
Unit 1 Entr Into Technical S ecification 3.0.3 When Both Residual eat emova um s ere e-ener ize On January 3, 1991, at 3:05 am, with Unit 1 at lOOX power, Residual Heat Removal (RHR) pump 1-2 was rendered inoperable when its control power was de-energized during a time period when D/G l-l, the emergency source for RHR pump 1-1, was removed from service.
Subsequently, control power to RHR pump l-l was'de-energized at 3:07 am.
This resulted in a violation of Technical Specification (TS) 3.5.2 and entry into TS 3.0.3 (which requires plant shutdown after one hour).
Control power was restored to both pumps at 3:08 am.
This event was described in LER 50-275/91-01, dated February 4, 1991.
The event resulted from an error by an auxiliary (non licensed)
operator.
Operations was preparing to per form a surveillance test which exercises the RHR pump suction valves (STP V-3M4).
The test requires that control power be isolated to one RHR pump at a time, preventing the pump from starting if called upon while its suction valve is closed.
The senior control operator (SCO), auxiliary control operator (ACO) and an auxiliary operator (AO) conducted a
pre-job briefing in the control room.
The SCO instructed the AO to go to the 4KV vital switchgear rooms and call the control room, at which time he would be instructed to de-energize the DC control power to the RHR pumps.
Following the briefing, the ACO requested permission from the Shift Foreman (SFM) to perform STP V-3M4.
The SFM denied permission because D/G 1-1 was out of service and he recognized the TS implications of de-energizing control power to RHR pump 1-2. It was concluded that the AO would be notified of the decision to not perform the test when he called the control room.
However, shortly following this discussion, the control room received alarms and other indications that the DC control power for both RHR pumps 1-1 and 1-2 had been de-energized.
The AO was immediately contacted to restore control power to both pumps.
The LER described the root cause as "personnel error due to miscommunication."
The LER further concluded that the miscommunication occurred during the briefing.
A contributory cause was attributed to inadequate AO training in that the AO did not demonstrate clear understanding of the "two train" concept (i.e., in any case the power to both pumps should not have been simultaneously cut).
The inspector concurs that the both of the causes described above were valid.
, However, the root cause evaluation should have also recognized that Operations Department policy for the operation of equipment had not been followed.
Operations Department Policy B-l,
"Conduct of Operations", which'escribed'perating practices for the normal execution of operator duties, allowed the AO only two options for performing equipment manipulations.
The first option allowed an operator to perform equipment manipulation with written instructions in hand.
The second option allowed the AO to manipulate equipment while maintaining verbal communication (either directly or over the phone) with an individual who had a written instruction in hand.
Had the AO committed these practices to memory, as was required by the policy, he would have realized that he could not have done what he thought he had been requested to do.
The LER stated that as corrective action-an Operations Incident Summary would be prepared and reviewed with all operations personnel.
The inspector reviewed the Operations Incident Summary and note'd that it had addressed all three of the causes described above.
In summary, while the LER and NCR regarding this event did not recognize compliance with Operations Pol)cy B-1 as a cause of this event and did not require any corrective actions, the Operations Department did.
As a result, the appropriate corrective actions were taken.
Subsequent to the issuance of the LER, the licensee identified an error in the "Analysis of the Event" section.
The LER stated that both trains of RHR were inoperable for three minutes.
However, it was not until the AO de-energized control power to RHR pump l-l that both trains of RHR became inoperable.
This resulted in a total out of service time of one minute.
The weaknesses in the identification of root-cause and the error in the analysis section of the LER indicate a lack of attention to detail.
The quality of LERs has been an issue of previous inspection reports and was discussed at the exit meeting.
The review of LER 50-275/91-01 is closed.
Lack of ualit Control Problem Identification On January 4, 1991, the inspector identified (through review of the distributed shift foreman's log) that certain problems had existed for a long time with little corrective action.
Specifically, the logs made quite a point of the fact that the carbon dioxide fire suppression system for the turbine lube oil reservoir had been out of service for months and the second spent fuel pool cooling pump had been out of service since September 25, 1990.
Although this equipment is not of high safety significance,'quipment out of service for that long raised questions regarding action being taken to fix the equipment.
Limited followup by the inspector indicated the components-did not have significant problems, only that attention had wane Because equality Control (gC) had been'harged with the responsibility to review logs for problems and ensure they were escalated to proper management levels, the inspector queried the gC manager as to whether the problems had been noted or acted upon.
On January 8, 1991, at a weekly resident exit, the gC manager indicated the pi oblems, had been noted in, the log review but not acted upon since the problems did not meet gC's screening criteria for importance.
Subsequent to the.exit, the plant manager indicated the gC screening criteria would be 'changed to encompass such problem areas.
The inspectors will continue to monitor gC performance in problem identification.
Conservative Mana ement Action on Feedwater Re ulatin Valve On January 5, 1991, the licensee made preparations to potentially shutdown Unit 2 to investigate and repair an oscillating feedwater regulating valve, FW 520.
The. valve had oscillated slightly.since the past refueling outage startup of the unit in Hay of 1990.
The licensee had tried several times to determine the cause of the oscillation but had not been successful.
The valve oscillations became worse in December. 1990.
The licensee made the decision to fully investigate FW 520 -and to shutdown if necessary to effect repairs.
The licensee brought in a valve vendor expert and reduced the plant power to 15K, to allow control through the feedwater bypass valve, while the isolated feedwater regulating valve was examined and repaired.
The diagnostic examinations developed by the maintenance team were insightfu) and creative.
The actions included rotating the valve discs 180 to observe any changes.
The problems eventually proved to be a combination of clogged air filters in the positioner and loose fasteners in the air operator.
It was noted by the inspector that on June 14, 1990, I8C planners proposed a theory that the control problem might be related to the air supply to the valve positioner (reference A/R 194201194201.
I8C investigation per work order C0073673 in July found no problem and referred the problem to mechanical maintenance.
The inspector suggested to the licensee at the exit interview that there might be a lesson worthy of their examination in the level of detail provided in ISC trouble shooting procedures, which might have corrected the problem in July 1990.
The inspector commended licensee management at the exit interview for their willingness in this instance to shut the unit down for repair of a valve.
The inspector considered that the licensee's actions very effectively communicated a safety attitude to plant staf Auxi'liar Saltwater Event On January 6, 1991, an auxiliary operator inadvertently closed a
Unit 2 auxiliary saltwater intake gate.
This action, had it gone to completion, would have rendered one train of auxiliary saltwater inoperable.
The auxiliary operator's actions did.not render the
,
auxiliary saltwater train inoperable because the. control room operator noted the gate closing and opened a crosstie valve between the auxiliary saltwater bay and a main circulating water bay.
The shift supervisor'ad the individuals involved make written statements.
The operations manager was called and an action request was initiated to document the event.
No log entry was made noting the event, but the operations manager notified the resident inspector the following day.
The problem apparently arose primarily from working on the wrong unit and working without a procedure in hand.
The auxiliary operator was supposed to open and close the auxiliary saltwater gates on Unit 1.
The procedure requires notifying the unit shift foreman and opening a crosstie gate before closing the auxiliary saltwater gate.
The problem occurred when the auxiliary operator moved from the Unit 1 gates to the Unit 2 gates, which are adjacent.
The operations management responded to the personnel error appropriately.
The licensee is tracking actions on a equality Evaluation.
At the exit interview this'example and the inadvertent disabling of two residual heat removal pumps (discussed in paragraph 4d) were discussed with the licensee.
The inspectors indicated that although the licensee had previously established a healthy attitude toward procedure utilization, these two incidents were worthy of the licensee's attention and action to preclude repetition.
Vibration and Loose Parts Monitor (VELPM Issues In October 1990, operators began to receive an increasing number of control room alarms from the Nuclear Instrument portion of the V8 LPM (9 in October, 22 in November, and 53 in December).
By January 8, having revieved seven alarms the day before, the operations shift supervisor considered the alarms an increasing nuisance and noted it in his log.
Subsequent investigation determined that there was no safety significance to the increased number of alarms, which resulted from an end of core life phenomenon.
However, the inspectors had two concerns:
o Although plant engineering, the "owners" of the VALPM, were aware of the increasing alarms, no significant efforts were made to understand their cause until after January 8, 199 o A considerable portion of the VKLPM data gathering and analysis systems, as described in the FSAR, were not functioning and had not been for an extended period.
Action to address the problems with. the system was slow.
S stem Descri tion The. V8 LPM consists of two separate monitoring systems.
The first is a collection of piezoelectric crystals, located in various parts of the Reactor Coolant System, which sense vibration and transmit an acoustic signal to the control room.
The second is the monitoring of the noise or fluctuations of the four sets of excore power range nuclear instruments (NIs).
The NIs are very sensitive to the amount of water between themselves and the core (with sensitivity on the order of 10 mils) and as such are capable of monitoring vibration of the core barrel or thermal shield.
Both sets of information input to a control room alarm and should input to a tape recorder, a spectrum analyzer, an X-Y plotter and a
chart recorder.
Technical Issue The licensee did not initiate investigative actions prior to January 1991.
In January, the licensee hired industry experts.
The consultants to the licensee concluded that core moderator temperature fluctuations on the order of less than 1 degree Fahrenheit at a frequency of less than 1 Hz were causinq actual power changes (due to the moderator temperature coeffic)ent (MTC)
changes in reactivity).
These power changes, which were picked up by the V8LPM as NI noise, increased in magn'itude as MTC became increasingly negative at the end of core life.
This phenomenon had been documented at other utilities.
These conclusions were discussed with Region V and NRR on January 22, 1991.
The licensee developed an action plan on January 31, 1991 to address questions raised during the discussion.
The action plan committed that a final report from the licensee's contractors was scheduled to be completed by March 15, 1991.
This report will be reviewed during routine followup inspection.
Lack of En ineerin Res onse To Increasin Alarms Plant engineering, who procedurally has responsibility for responding to the indication of the VKLPM and were repeatedly notified by operations of alarms, did not respond to the increase in alarms until January 1991.
The V8LPM had been considered an unreliable instrument.
Little consideration appears to have been given to the idea that the V8 LPM may have been alarming due to an actual vibration condition.
As a result, an unhealthy condition developed where plant personnel essentially ignored control room alarm \\
The annunciator response 'procedure (AR PK ll-ll)associated with the V~LPtl ",equired only notification of the Technical Service Power Production Engineers for guidance.
What evolved was a log,.stored in the V8 LPH cabinet, where new alarms were recorded.
The Technical Services system engineer periodically checked the log for alarms.
AR PK ll-llhad been revised on'November 11, 1988. 'he previous revision included four pages of detailed instructions for the appropriate interpretation and response to alarms.
The procedure revision history form described the reason for the revision as
~ "...Until alarm system works properly, require engineering to determine if an alarm condition is valid.'owever, no equivalent procedure was developed for engineering.
Condition of the V8LPH S stems FSAR chapter 4.4.5 describes the V8 LPM system installed at Diablo Canyon:
"When the output of an individual transducer channel exceeds an adjustable setpoint...
The output of the alarmed channel and of three additional channels (whose selection is based on the location of the alarming transducer)
are recorded on magnetic tape and on strip chart recorders."
Further it states;
"In the event that the output of a loose parts channel exceeds the alarm value, the record of the event will be available to the operator and plant staff for analysis.
The event will be corn~ared with other previously recorded signatures of the RCS.
'ontrary to the above, on January 9, 1991, the inspector found that the magnetic tape and strip chart recorders were out of service and had been for an extended period of time.
A spectrum analyzer and associated X-Y plotter, not described in the FSAR, were also out of service and inoperable.
As a result, there was no installed equipment with the capability to compare event vibration signatures to previously recorded signatures.
An analysis was finally made on January 8, 1991, with separate temporarily installed equipment.
This is an apparent deviation (Enforcement Item 50-275/90-30-01).
On February 11, 1991, the inspector noted that only one gE had been issued to address the V8LPM.
The gE (f0008283), initiated January 10, 1991, was limited to addressing the cause of repeated V8LPH alarms.
The inspector asked the licensee if the lack of an engineering response to repeated control room alarms indicated a
programmatic breakdown and warranted an NCR.
On February 12, 1991, an NCR was initiated.
During an exit meeting on February 15, 1991, the licensee stated that the NCR would address:
o the lack of an engineering procedure for response to V8LPH alarms, o
the issue of out of service times for important equipment not covered in the Technical Speci,fications and not listed as gual ity Class I, o
the adequacy of the installed design of,the system,
o and inaccuracies in the FSAR description.
The licensee'.s resolution of the NCR will be reviewed in conjunction with a review of the licensee's response to the Notice of Deviation.
Removal of the Wron Nuclear Instrument in Unit'2 On January 12, 1991, instrumentation and controls (I8C) technicians inadvertently removed the wrong nuclear instrument from service while performing a planned test.
The unit was operating at lOOX power and power range channel N-43 was removed from service.
One technician inadvertently manipulated test dials which imposed a signal on operating channel N-42.
The act was apparent to control room operators who received quadrant power, tilt and rod stop alarms and immediately ordered the ILC technicians to stop their actions.
A reactor trip was avoided only because the 18C technicians were moving the dial slowly and the control operators assessed the alarms rapidly.
The licensee initiated a nonconformance report on the event.
The licensee's actions included determining the root cause to be lack of self verification and poor labeling.
Further corrective actions were defined, such as discussion of the event with other ILC crews and performing a
human factors review.
Untimel 0 erabilit Review and Lack of A ressive En ineerin nvo vement or t e team riven uxi iar ee water um nit 1 On January 22, 1991, the inspector questioned the action request tag on the control switch for valve FCV-95, the steam admission valve for the steam driven auxiliary feedwater pump.
Detailed followup revealed several licensee weaknesses and concerns about the operability of the valve.
The valve is very important in the event of a loss of AC power.
The valve is motor operated and is the only DC powered motor operated valve in the plant and will open on battery power alone.
The valve admits steam to the steam driven auxiliary feedwater pump, thereby providing the means of removing reactor heat in the event of a loss of all AC power.
The Unit 1 valve first failed to open on January 18, 1989 in a post maintenance test.
Recommendations were made for operational changes to preclude thermal binding at that time but those changes were not made until much later.
The valve failed to open twice more in test situations in December 1989 and May 1990.
The valve failed to open a fourth time (again in May 1990) dur'ing experimentation with valve operation conditions.
Motor operated valve diagnostic test equipment was in place for the analysi The diagnostic equipment revealed at that time (Hay 1990) that the valve had been stuck and to get it unstuck an excessive force had to be used, well in excess of the manufacturer's recommendations.
An action request was written'o engineering, who performed -an evaluation in June 1990 and c'oncluded the vaTve was overstressed and should be disassembled.
The valve was not disassembled at that time and was documented as operable by a gC inspector "per discussion with maintenance."
The inspector's assessment on January 24, 1991, was that the licensee's assessment of the root cause of the valve failing to open and its operability as. a result of being overstressed were not well understood.
The inspectors had meetings with the licensee maintenance management and engineering management on January 24; 25, and 26 in order to focus on technical questions and to ascertain the licensee's basis for operability of the valve.
In summary the technical issues pursued were as follows:
o The diagnostic test report by the equipment vendor dated June 11, 1990, said a similar scenario had occurred at another PMR and the valve had failed in service even though that utility had changed their operating procedure.
The licensee was asked to ascertain the identity of the other utility and determine why Diablo Canyon would not repeat the failure.
o The valve stem in FCV-95 was made by PG8E in 1980 and was made oversize.
The inspectors asked for material identification.
o The inspectors asked about low cycle fatigue.
The valve had been at or near yield four times.
o The inspectors asked about the licensee's root cause (i.e.,
"thermal binding").
The valve vendor said it should not bind.
The "sister" valves do not bind.
These issues were satisfactorily resolved by the licensee on Saturday, January 26, 1991.
The engineering analysis was reviewed by the PSRC.
The inspectors subsequently discussed apparent management issues which evolved out of the attempt to understand the licensee's action Determination of 0 erabilit of Nonconformin E ui ment Several operability deter'minations were made regarding FCV-95-and the following problems are evident:
o There is no single document which is used to support operability.
As a result, operability determinations have been inconsistently documented in a variety of places or in some cases, not at all.
o There is inconsistency in who can make operability determinations and no apparent independent review or verification necessary.
I o
The bases for some determinations were poorly documented and in some cases operability determinations appear to have been based on conversations about undocumented information.
o Operability ha's been determined without a documented understanding of the elemental cause of a problem.
o No management review appears to be necessary for complex operability determinations.
Timeliness of review o
'
root cause review was completed 15 months after the first event.
The root cause, which was determined to be thermal binding, conflicted with vendor information supplied to the licensee.
The first corrective action (revisions to operating procedures)
was taken 19 months following the first event and it was not comprehensive.
o Twenty-four months following the first event, the valve had not been internally mechanically inspected, even though opportunities had been available.
Electronic Trackin Confusion It took several hours to go through the numerous electronic action requests (ARs) to determine with any accuracy the status of the valve:
o There are at least eight ARs which track aspects of this problem, and another two dozen "evaluations 'ssociated with the ARs.
o Several work orders have been initiated, many which were never wor ke o Each document becomes 'essentially a continuing note pad.
Current status of the overall" concern could not be found in any of the documents.
The licensee initiated a nonconformance on the valve and the circumstances surrounding the operability= determinations (reference NCR DC1-91-MM-NOll).'ubsequent to the reactor shutdown on February 1, 1991, for the refueling outage, with licensee personnel the inspector operated the valve FCV-95 manually and felt no abnormal motion.
In addition, the inspector observed disassembly of the valve with the vendor present and noted no deleterious conditions.
At the exit interview, the inspector discussed the weaknesses perceived in the comprehensiveness and timeliness of actions surrounding FCV-95, as well as the administrative weaknesses
.
identified above.
Corroded Auxiliar Saltwater.S stem Crosstie Valve 0 erators On January 16, 1991, the inspector observed that an action request (AR) tag, dated July 2, 1990, was attached to auxiliary saltwater system valve 1-FCV-496 which stated
"manual handwheel badly corroded and will not engage."
1-FCV-496 is a motor operated ASM train crosstie valve.
The ability to manually operate the crosstie valves was assumed in the FSAR.
The inspector questioned:
o The operability of the crosstie valve and the system, The timeliness of corrective actions to repair 1-FCV-496.
The design classification of the crosstie valve manual operators.
0 erabilit of the ASW S stem In response to the inspector's questions, on January 17, 1991, after lubrication with penetrating oil, the handwheel for 1-FCV-496 was dislodged by maintenance and returned to manual operability.
On January 18, 1991, the remaining Unit 1 and Unit 2 crosstie valves (1-FCV-495, 2-FCV-'496, 2-FCV-495, and O-FCV-601, the Unit crosstie)
were examined.
1-FCV-495 was found to be inhibited from manual operation due to excess paint between the handwheel and actuator frame.
The handwheel was engaged manually and was rotated by two men once the paint was broken loose.
Manual operation of 2-FCV-495 was found to have some resistance due to corrosion.
The other two valves were found to be acceptable.
The licensee concluded that all five valves were "manually operable".
In addition, motor operation of the valves had been verified on a quarterly basi The licensee concluded that the. ASM system was operable based on the ability to manually operate 1-FCV-495 and to automatically operate both valves from the control room.
However, a nonconformance report was initiated to address the timeliness of corrective maintenance and the quality classifications of. the crosstie valves (NCR DC1-91-EM-09).
Timeliness of Corrective Actions:
As discussed in Inspection Report 50-275/89-14, on May 23, 1989, 1-FCV-496.failed to close when operated from the control room.
Further inspection revealed that the lever-on the motor operator.
which takes the valve from automatic to local hand control had not returned completely to the automatic position due to rust.
Additionally, the handwheel was also frozen in place due to rust.
A followup item (50-275/89-14-01)
was initiated to follow the licensee's corrective actions.
Subsequently, the licensee initiated a quality evaluation'gE, a
lower level of nonconformance review).
gE f0006496 attributed the root cause to an "inadequate maintenance procedure".
The
"corrective action to prevent recurrence" was determined to require a revision to maintenance procedures to include the inspection and cleaning of the handwheels and de-clutching levers for motor operated valves.
The procedure changes were not made until October 1990, seventeen months after the problem occurred in May 1989.
In
'he interim, the crosstie operators were not inspected duri,ng the Unit 1 refueling outage in late 1989 nor.were the. Unit 2 operators inspected during the Spring 1990 refueling outage.
Simple periodic checks of the handwheel to freely rotate without the de-clutching mechanism engaged were not made.
Based on the above, and a repeat of the 1989 event, the corrective actions taken in response to gE f0006496 do not appear to have been timely nor comprehensive.
At the end of the inspection period, review of the 1991 NCR had not been completed.
Preliminarily, the NCR focused on two causes; inadequate maintenance and inadequate attention to important equipment not classified as safety related.
Corrective actions were initiated to enhance preventive maintenance performed on the ASM crosstie operator, including periodic manual operation checks, and to initiate a program to identify important non-safety related equipment which requires increased attention and timely maintenance.
The NCR had not considered the adequacy of corrective action resulting from gE f0006496 as part of the root cause.
The inspector discussed this issue with the electrical maintenance manager who agreed the issue should be addressed in the next NCR review-group meeting.
Based on the above, Open Item 50-275/89-14-01 is close Desi n Classification of the ASM Crosstie.Valves:
~His'tor The ASM crosstie valve operators were originally purchased to meet licensee Design Class I requirements, including seismic qualification.
In 1980, the crosstie valve operators were reclassified as Design Class II equipment (equivalent to balance of plant equ>pment).
The design change that accomplished this (DCO-EM-'779) stated that "Crossover capability for auxiliary saltwater heat exchangers is highly desirable, but is not required to perform a safeguards function."
During the Nay 1989 inspection, the inspector. questioned. why the motor operators for the ASM crosstie valve operators were listed as Design Class II, non-safety related (no application of the quality assurance program required and no seismic qualification).
Engineering responded that the crosstie valves were normally open but were required to isolate the ASM trains in the event of a passive failure to one train.
Additionally, the design basis does not assume a passive failure until 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following a design basis accident.
At that time, decay heat is'reduced to the point that time is available to allow manual actuation of the crosstie valves.
Therefore, the valve motor operators were not needed to perform a design function.
2.
Current Ins ection On January 18, 1991; the inspector again questioned the design classification of the ASM crosstie valve operators.
Specifically, the design classification discussed above apparently did not consider the design basis need to manually operate the valves.
The inspector's concern was that separation of two redundant, but.normally crosstied, trains of the ASM required safety related isolation capability.
Nuclear Engineering and Construction Services (NECS) could not identify documentation which summarized the classification decision making for the ASM crosstie valve operators, nor was any documentation found indicating that the licensee had presented this classification determination to the NRC.
The ASM system design criteria memorandum (DCN),
a design basis summary, published in July 1990, provided a vague description of operation requirements for the crosstie valves, stating that they were needed "...to support the long term recirculation phase of post accident operation may require manual action."
NECS initiated a review of the design classification of the crosstie valve operators due to this ambiguous description, the inspector's concerns, and questions provided by the NCR review group.
NECS issued two memos, dated February 12 and February 20, 1991, which summarized the design basis for the ASM
crosstie valves and presented-ah evaluation of their design classification.
In summary,'he February 20, 1991 memo concluded that the ASM crosstie valve operators were not required for post, accident mitigation.
The argument presented was similar to the one presented in May 1989.
However, it went further to say that the design basis passive failure as described in the FSAR is a
cpm leak for 30 minutes occurring at least 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after a
design basis event.
Further, a 50 gpm leak 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after a
design basis event did not impact the ability of an ASM train to supply adequate cooling and therefore train isolation was not necessary.
The inspector discussed the memos with the NECS manager and his staff on February 19 and 22, 1991.
The inspector stated that
,
the licensee's February 12 and 20, '1991, analyses did not fully address the potential for the rupture of an ASM header (without a design basis event such as a
LOCA) while't power as is discussed in FSAR section 9.2.7.2.
This section concludes that an operator can remotely isolate the crosstie header before the temperature of the component cooling water system significantly increases.
However, if a ASM line rupture occurs in one of the pump vaults while the plant is at power, the licensee must close the remaining Design Class II crosstie valve or lose all ASM. If the crosstie valve can not be closed, ASM would not be
"
available for safe shutdown.
This item is unresolved (Unresolved Item 50-'275/90-30-02).
l.
Unit 1 Reactor Tri Februar
1990 On February 1, 1991, at 9:09 a.m.,
a Unit 1 reactor trip occurred due to a steam generator (S/G) low level with a steam flow/feedwater flow mismatch on S/G 1-4
~
Feedwater to S/Gs 1-3 and 1-4 was lost when the associated feedwater regulating valves and the feedwater bypass valves closed.
The feedwater valves went closed as a result of the loss of instrument air.
Instrument air was lost to the four feedwater valves when an instrument air isolation valve was inadvertently shut by a crew building scaffolding to support outage work.
All safety systems functioned as designed and plant conditions were stabilized shortly following the event.
Unit 1, which had been scheduled to start a refueling outage on February 3, 1991, initiated shutdown for refueling.
Although all safety systems operated. as designed, some non-safety related systems did not.
Significantly; o
Circulating water (CM) pump l-l failed to restart following the 12KV bus power transfer.
The operation of one CM pump is necessary to maintain condenser vacuum which allows the
,
operation of condenser steam dumps.
The failure of 'a CW pump
'to restart following a reactor trip has been a recurring problem'.
o Control Rod 'Drive Motor (CRDM) fan E 1-3 failed to restart following the power transfer.
A similar fai lure to restart occurred following the December 24, 1990 reactor trip.
o, The 25KV motor operated di,sconnect (MOD) switch did not fully open when demanded to do so with the control room switch.
Opening the MOD allows operators to backfeed the main transformer with offsite power.
Operators were required to manually open the switch to the full open'position.
The event was the subject of NCR DCl-91-WP-N012 and LER 50-275/91-02.
The licensee committed to. address the equipment problems discussed above in the NCR.
Followup of the licensee's review of the event will be addressed in a future inspection.
No,violations were identified.
One deviation was identified.
5.
Maintenance (62703 37700 The inspectors observed portions of, and reviewed records on, selected maintenance activities to assure compliance with approved procedures, technical specifications, and appropriate industry codes and standards.
Furthermore, the inspectors verified maintenance activities were performed by qualified personnel, in accordance with fire protection and housekeeping controls, and replacement parts were appropriately, certified.
The inspectors examined the maintenance aspects of the events described in paragraph 4 above.
Specifically, the inspectors examined corrective action on positioner feedback linkages, examined steam dump components and reassembly, and examined feedwater regulating and check valve corrective maintenance (paragraph 4b), design change instructions (paragraph Sb),
equipment out-of-service times (paragraph 4e),
feedwater regulating valve troubleshooting and repair (paragraph 4f), extensive maintenance history analysis and observation of disassembly and internals inspection of the steam admission valve for, the steam driven auxH.iary feedwater pump (paragraph 4j), lack of maintenance and repair of the vibration and loose parts monitoring system (paragraph 4h), untimely maintenance of auxiliary saltwater crosstie valves (paragraph 4k), and maintenance controls associated with scaffolding installation (paragraph 41).
a.
Vendor Manual U dates (37700)
In the review of nonconformance reports, the inspectors noted two separate technical review groups had reviewed maintenance problems associated with or caused by a lack of updated vendor manuals.
The review groups both concluded that although the vendor manuals were not up to date, the problem was with the vendor and not with the license The licensee's rationale was that the vendors in question had been approached to participate in the vendor manual update program but
,
had declined the opportunity.
The two examples identified by the inspector were the feedwater regulating valve FCV 1530, whose cage had deformed due to being torqued to old versus updated torque values, and FCV 531, a main feedwater check valve which had not been assembled in accordance with improved assembly instructions.
The inspector met with the Vice President for Technical Services on February 7 and 14, 1991, and discussed the TRG conclusions.
The licensee made the commitment that the vendor manual update program would be reviewed to ensure that vendor manual information accuracy would be verified periodically even if the vendor chose not to do the verification.
The willingness of TRG chairmen to accept a
conclusion that a lack of vendor information was the vendor's problem was discussed with maintenance management, the TRG chairman and with licensee management at the. exit.
The licensee agreed the TRG conclusions were inappropriate and should not be repeated.
Residual Heat Removal Motor Bearin Lube Oil Si ht lass The inspector reviewed non-conformance report DCO-90-EM-N070 dated October 22, 1990,
"Work on RHR Motor 1-2 not in accordance with the instructions provided in the work order."
Electrical Maintenance personnel performing a Design Change Package, (DCP J-45237)
which changed the existing motor bearing lube oil sightglass lens material from lexan to pyrex, did not perform the work in accordance with their work order (WO C0072788).
The craftsmen had implemented the DCP in accordance with the WO through the steps for post-maintenance testing, when a leak was discovered in the new lens.
Under the direction of the Foreman and the General Foreman, the cause of the leak was determined to be a cracked sightglass lens (pyrex).
The sightglass was returned to the pre-DCP configuration with new components, including a new O-ring, new frame, and new lexan sightglass lens.
The removal of the pyrex lens, examination, troubleshooting and return to original configuration were all done without any changes to the WO and without contacting gC.
The cause was determined to be personnel error.
The work crew had not complied with procedures on when and how to implement changes to work orders.
As corrective action, all maintenance personnel were briefed by management on the importance of procedural compliance and adherence to issued work orders.
Additionally, a maintenance bulletin was issued emphasizing the necessity of performing work in accordance with issued work orders and the requirements for making changes to the The inspector found the root cause and corrective actions to be acceptable.
No violations or deviations were identified.
6.
Sur vei 1 1 ance (61726)
By direct observation and record review of selected surveillance testing, the inspectors assured compliance with TS requirements and plant procedures.
The inspectors verified that test equipment was calibrated, and acceptance criteria were met or appropriately dispositioned.
The inspectors examined surveillance aspects of the events described in paragraph 4 above.
Specifically the inspectors examined surveillance testing of steam dump valves (paragraph 4b), surveillance testing of RHR pump suction valves (paragraph 4d), vibration monitoring temporary test equipment (paragraph 4h), surveillance testing of nuclear instruments (paragraph 4i), and surveillance testing of the steam admission valve for the auxiliary feedwater valve (paragraph 4j).
No violations or deviations were identified.
7.
Licensee Event Re ort Follow-u (92700)
The LERs identified below were also closed out after review and follow-up inspections were performed by the inspectors to verify licensee corrective actions:
Unit 1:
90-14, 90-15 Rev.
0, 90-18, 91-01 Unit 2:
90-02 (Revisions 0 and 1)
No violations or deviations were identified.
8.
0 en Item Follow-u (92703 92702)
a.
Untimel Corrective Actions Followu Item 50-275/90-03-03 (Closed This item was related to timeliness and effectiveness of corrective actions, with a concern that no group or individual appeared to feel responsible for managing the overall corrective action program.
This issue had also been raised in other inspection reports,
.
including 50-323/90-13 and 50-323/90-19.
The licensee convened a
special event investigation team (EIT) to address the issue of timely resolution of identified problems.
The results of the EIT were discussed at a management meeting and were provided in Inspection Reports 50-275/90-26 and 50-323/90-26.
'PG8E letter DCL-90-237, dated October 1, 1990, provided the programmatic changes to address the identified root and contributing causes identified by the EIT.
The inspector discussed the implementation of the changes with a lead quality engineer and reviewed changes to Nuclear Plant Administrative Procedures (NPAP)
C-12 and C-23 equality Assurance Procedure (gAP) 15.A.
The procedure
changes included co'nsolidation of responsibilities for problem-categori zati on, initiati on of nonconformance repor ts, and ove'r s ight of corrective action implem'entation.
The inspector noted that a
revision to NPAP C.-18, which provided procedures for event, response plans, was signed but not implemented as of January ll, 1991.
The licensee's letter to the NRC, mentioned above.,
indicated that all changes would be implemented by December '31, 1990.
The issue of the licensee's timeliness and effectiveness of corrective actions remains a concern with the NRC and will continue to be evaluated by the NRC resident and regional staff.
However, this specific open item is closed.
Weak Desi n Chan e Instructions Unresolved Item 50-275/90-15-01 ose This item was related to the design change notice (DCN) for replacement of an 'emergency diesel generator (EDG) droop relay.
Unresolved issues were the function and post maintenance testing of pass through wiring that was disturbed during the design change, and the licensee's actions regarding design change instructions.
The resident determined that the functions of the pass through wiring had been assessed and that the licensee had provided an evaluation to verify that the function of the wiring had been tested after the maintenance.
The inspector reviewed portions of Nuclear Engineering Manual Procedure (NEMP) 3.6DC, Diablo Canyon Power Plant Design Changes.
The procedure had been extensively revised to
'stablish how design changes were initiated, processed, approved',
and documented.
The inspector also reviewed selected design change packages and found no problems similar to those observed for the EDG droop relay DCN.
This item is closed.
Leakin Wel ds Fol 1 owu Item 50-275/90-15-02 (Cl osed This item was related to leaking pipes and welds in the feedwater system for Unit 2.
The licensee had committed to reviewing the history of leaking pipes and welds throughout the facility.
The licensee's review indicated that fatigue appeared to be the cause of most of the piping failures.
The inspector reviewed the licensee's leak history, which was a list of 40 piping and weld cracks dating back to August 1987.
The inspector observed that the licensee's conclusion of failure due to fatigue did not appear to be substantiated in that metallurgical evaluations were performed for only 12 of the 40 failures.
The senior resident had also questioned the evaluation, and the licensee was preparing a followup report.
This item is closed.
Open item 50-323/90-31-01 will be used to follow the licensee's efforts to evaluate the causes of the piping and weld,crack Fuel Handlin Bui ldin Ventilation Enforcement Item 50-323/90-08-01
,ose licensee vent e ort
-
-
-
ose licensee ven e or
-
-
-
ose These items were related to the inability to meet'Technical Specification (TS) 3.9.12 because temporary hoses were blocking open fuel handling building (FHB) personnel doors when irradiated fuel was being'moved.
The effect of the open doors was that the required 1/8" w.g. negative pressure could not be maintained in the FHB; only a 1/16" w. g. negative pressure'was obtained.
The inspector reviewed the licensee's June 25, 1990, response to a Notice of Violation issued in Inspection Report 50-323/90-08.
The corrective actions to prevent recurrence included:
revising Surveillance Test Procedure (STP) M-5, Routine Surveillance of Fuel Handling Building Ventilation System, to add all FHB personnel entrance/exit doors to the list of doors required to be closed for FHB ventilation system operability; installing warning signs on all Unit 1 FHB entrance/exit doors; performing a design criteria review to establish plant doors important to safety; and issuing a
memorandum that described the event to maintenance, operations, and appropriate engineering personnel.
Additionally, the licensee wrote that warning signs were placed on each Unit 2 FHB personnel door.
Revision 8 to STP M-.5 added the personnel doors to the list of doors.
required to be closed to ensure FHB ventilation system operability.
As stated by the engineer who prepared Revision 8, warning signs should have been installed on all of the doors listed in section 6. 1. 1. g of the STP.
The inspector walked through the FHB and checked some of the doors on the list that should have had warning signs.
The inspector noted that warning signs were not installed on Unit 1 door ¹263, Unit 2 Door ¹263-2, and Unit 2 door ¹361-2.
This is an apparent deviation from the licensee's commitment to install warning signs on the FHB personnel entrance/exit doors (50-323/90-30-01).
Licensee Event Report (LER) 2-90-02-01 stated that Nuclear Engineering and Construction Services (NECS)
had performed an evaluation which concluded that during the time period the FHB doors were open, air from the spent fuel pool area continued to be exhausted through the exhaust charcoal filters and the plant stack.
As such, NECS stated that the final safety analysis report accident analysis continued to be met while the doors were open.
The NECS engineer, who had performed the evaluation for the effects on the FHB ventilation system as a result of having doors blocked open, was contacted.
Poorl Conducted En ineerin Evaluation o
The inspector questioned the methods used to perform the evaluation as they were not stated.
The NECS engineer indicated that the evaluation had been performed after walking though the FHB and assessing the apparent effects on ventilation flow-when a door was opened.
Although five doors
25-had been open during the time that the TS requirements were not met, the evaluation was performed with doors open one at a
time.
o The inspector asked the basis of'he TS requirement that was not met, but the NECS engineer could not provide the basis.
o Additionally, the inspector noted that the evaluation did not discuss the configuration of running FHB ventilation fans at the time of the event.
Finally, the evaluation did not appear to quantify the actual effects on ventilation and link those which the calculated effects necessary to increase offsite doses to NRC guideline values, had a fuel handling accident occurred.
Overall, the evaluation did not appear to be thorough, and the assumptions were not stated in the text.-
The NRC resident and project inspectors walked through the Unit 2 FHB (where no fuel movement was in progress)
and opened doors to assess the apparent effects on-the ventilation system.
The NECS evaluation attributed the inability to meet the TS negative pressure requirement to be due, in a large part, to doors 389-2 and 263-2 being open..
However the evaluation went on to state that this allowed air to be exhausted outside the FHB which did not make apparent sense.
The inspectors opened doors 389-2 and 263-2 and found that air appeared to be drawn into the FHB as would be expected.
~0th Fi di The inspectors also noted that Unit 1 door ¹357 was open on its strike and would not always shut, and Unit 1 door ¹285 was shut but not latched.
Comments related to the NECS evaluation, and the inspector's findings related to the doors were provided in an exit meeting on January 11, 1991.
Subsequent to the exit meeting, the inspector learned that the licensee's technical review group would revisit the NECS evaluation of the FHB ventilation system.
On January 18, 1991, the surveillance test to demonstrate the ability to maintain a 1/8" w.g. negative pressure in the Unit 1 FHB was unsuccessful.
The licensee is investigating the cause of the failure to pass the surveillance test.
The issues pertaining to the NECS evaluation, and the inability to pass the surveillance test after additional fuel movement in December 1990 are the basis for a unresolved item (50-275/90-30-03).
The item is unresolved pending the results of the licensee's reevaluation of the ventilation sy'tem performance.
Enforcem'ent Item 50-323/90-08-01 is closed.
LER 50-323/90-02, revisions 0 and 1, are close '6 Informati on Notice 90-79 (C 1 os ed)
The inspector requested the licensee engineering personnel to provide their assessment of the NRC Information Notice 90-?9, Failures of Main Steam Isolation Check Valves'esulting in Disk Separation.
The licensee's program regularly reviews NRC information notices for applicability and possible action.
The inspector reviewed the licensee's assessment dated January 9,
1991, File 4.73 146.10, from the NECS project engineer to the manager of Nuclear Operation Services.
The letter concluded that NECS did not consider the information notice to be applicable to Diablo Canyon.
En ineerin Errors The inspector found the licensee's review to be faulted and based in part on improper assumptions and elemental engineering errors.
For example, the letter states that Sequoyah Power Plant had severe wear in a check valve which the PGKE letter said was indicative of'the disk spinning.
The problem would not occur at Diablo Canyon, stated the letter, since Diablo Canyon valves had anti-rotation pins.
The inspector found, by reading the information notice that the Sequouah disk had separated from its hinge arm due to fatigue, not wear, Further, by calling the Sequoyah NRC resident inspector, the, inspector determined that Sequoyah had anti-rotation pins as well.
The letter also stated upward disc travel of the Diablo Canyon valves was controlled by the "integral tail link stop" which contacted the valve body and would serve to minimize wear due to wobble of the disk.
This conclusion did not make sense to the inspector, since he'knew from past inspections that the disk.-is not firmly connected to the hinge arm and does in fact wobble.
Discussion with the, licensee s
engineer showed that he was using an assembly drawing (which does not show part details} to make conclusions abogt how the parts fit together.
This was a basic engineering error.
Thirdly, the letter stated that the valves were inspected at regular intervals and there has been no abnormal wear noted at the connection.
The inspector considered this information to be extraneous, since the periodic inspection is done by borescope and would only detect part failure and would not be able to see weai ot cracking until failure occurre Revised Action
,
The inspector-contacted the licensee's project engineer who rescinded, the first letter and issued a second letter dated Febr'uary 1, 1991 (File ¹ 146.10).
The licensee had gathered other information (the TVA LER) and discussed the situation with TVA contractors.
The licensee revised their conclusion to include opening and inspecting a mainsteam check valve during their outage, including a liquid penetrant test of the disk post to identify any fatigue cracking.
The inspector observed licensee actions and examined the valve internals during, the week of February ll, 1991.
No wear was apparent and a subsequent liquid penetrant examination was satisfactory.
Mana ement Oiscussion At the exit interview the errors made by engineering were discussed with licensee management.
The opinion that the errors were fundamental and that they should not have occurred in the engineering organization was discussed.
Additionally, the opinion that the engineering management review chain should have identified
'the errors was discussed.
The licensee's engineering management stated that the engineering evaluation could have been improved.
The inspectors will continue to followup engineering performance in the normal course of future inspections.
No violations were identified.
One deviation was identified.
9.
Unresolved Items Unresolved items are matters 'about which more information is required in order to ascertain whether they are acceptable items, violations, or deviations.
Unresolved items disclosed during this inspection are discussed in paragraphs Sd and 4k of this report.
10.
Exit (30703 On February 21, 1991, an exit meeting was conducted with the licensee's representatives identified in paragraph 1.
The inspectors summarized the scope and findings of the inspection as described in this report,
/