IR 05000324/1987012
| ML20235H546 | |
| Person / Time | |
|---|---|
| Site: | Brunswick  |
| Issue date: | 06/30/1987 |
| From: | Shymlock M, Linda Watson NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| To: | |
| Shared Package | |
| ML20235H499 | List: |
| References | |
| 50-324-87-12, 50-325-87-12, IEIN-86-106, NUDOCS 8707150172 | |
| Download: ML20235H546 (31) | |
Text
,_
__
_ _ _ -
.. _ _ _ _ - _ _ - _ - - _ _ _.
'
o UNITED STATES gn etc
,
.jo
. NUCLEAR REGULATORY COMMISSION y.
REGION 11
.
.,
,
g j-101 MARIETTA STREET. N.W.
- t e
ATLANTA, GEORGI A 30323
'+f.
,,
.....
.
!
'
l Report Nos.: 50 325/87-12.and 50-324/87-12
,
l-Licensee:
Carolina Power and Light Company P. O. Box 1551
'
!
l Raleigh, NC 27602
'
Docket Nos.:
50-325 and 50-324 License Nos..
DPR-71 and DPR-62 Facility Name:
Brunswick 1 and 2 Inspection Conducted:
May 4 - 8, 1987 Inspector:
,h
-
." Date Sidned
.To. /f/ 7 L. J, Wa~t s on U
Accompanying Personnel:
B. A. Breslau S. P. Burris P. B. Moore C. ' J. Paulk S. D. Stadler M. B. Shymiocy,'# N
'7h
/)
4 4 86, 887 Approved by:
Chief C Date Signed Operational Programs Section Division of Reactor Safety SUMMARY Scope: This special, announced inspection was conducted in the area of review of operational events, control room activities, maintenance, training, Licensee Event Reports and close out of cpen items.
Results:
No violations or deviations were identified.
.
71%h070707 0500 G
_ _ _ - _ _ _ - _ _
_- __- _- _
_
_ _ _ _
_
_
,
b.
n
!.
,
,
.
REPORT DETAILS 1.
Persons Contacted'
Licensee' Employees
- P. W. Howe, Vice President
~
- C. R. Deitz, General Manager
- E. A. Bishop, Manager, Operations
- B.. S. Strickland, Shift Supervisor, Operations
- J. A.' Smith, Director, Administrative Support
- E. R. Eckstein, Manager, Technical Support
- J. O'Sullivan, Manager, Maintenance
- K. B. Altman, Principal Engineer, Maintenance
- A. G. Cheatham, Manager, E&RG
- L. E. Jones, Director, QA/QC
- K. E. Enzor, Director, Regulatory Compliance,
- R. M. Poulk, Senior Specialist, Regulatory Compliance
- R. E. Helme, Director, Onsite Nuclear Safety
- J. W. Moyer, Manager, Training
- T. H. Wyllie, Manager, Engineering and Construction Other licensee employees contacted included instructors, engineers, technicians, mechanics, and office personnel.
NRC Resident Inspectors
- W. R' land u
- L. Garner
- Attended exit interview 2.
Exit Interview (30703B)
'
The inspection scope and findings were summarized on May 8,1987, with those persons indicated in paragraph I above. The inspector described the areas inspected ~and discussed in detail the following inspection findings:
(0 pen) Unresolved Item 325, 324/87-12-01.
Evaluation of Licensee's Action to Resolve Equipment Failures Associated With Licensee Event Reports (LERs) 1-86-024,1-87-001, 2-87-001, arid 2-87-004, (paragraph 5.e).
(0 pen) Inspector Followup Item 325, 324/87-12-02.
Revision of Administrative Procedures to Include Requirements for Procedure Changes for Reversing Procedure Steps, (paragraph 5.d).
__ _ _ _ _ _ _ _ _ _ _ _ _ - _
_ - -
-
i
-
,
(0 pen) Inspector Followup Item 325, 324/87-12-03.
Review of LER Preparation Process, (paragraph 9).
(0 pen) Inspector Followup Item 325/87-12-04.
Evaluate Results of Licensee's Inspection of Contact Block Assemblies on the AC Operators for Major Flow Path ECCS Valves, (paragraph 5.b).
(Closed) Violation 324/84-39-01.
Failure to Implement 01-01, (paragraph 3.a).
(Closed) Violation 325, 324/85-01-01.
Failure to Remove a Licensed Operator From Licensed Duties Subsequent to Failing the Annual Requalification Exam, (paragraph 3.b).
(Closed) Violation 325, 324/85-24-01.
Failure to Implement AI-59, (paragraph 3.c).
(Closed) Unresolved Item 325, 324/85-01-03.
Interpretation of
" Actively Performing the Functions of an Operator or Senior Operator" with Regard to 10 CFR 55.31(e), (paragraph 3.d).
(Closed) Unresolved Item 325, 324/85-01-05.
Interpretation of
" Annual" for Requalificatio1 Program, (paragraph 3.e).
(Closed) Inspector Followup Item 325, 324/82-26-02.
Response to QA/QC Audit Findings of Training Programs, (paragraph 8.a).
(Closed) Inspector Followup Item 325, 324/84-39-02.
Controller E11-SS-F605A Output Plug Was Placed in Wrong Position, (paragraph 8.b).
(Closed) Inspector Followup Item 325, 324/85-24-02.
Discrepancies not Documented on Attachment 3 as per AI-17 and Failure to Follow AI-58 for Issuance of Clearances, (paragraph 8.c).
No dissenting comments were received from the licensee. The licensee did I
not identify as proprietary any information reviewed by the inspectors.
3.
Licensee Action on Previous Enforcement Matters (92702, 92701)
a.
(Closed) Violation 324/84-39-01.
Failure to Implement 01-01.
The inspector reviewed 01-01, Rev. 19, Operating Principles and Philosophy, and found that the procedure had been revised to provide direction for making temporary changes.
Training on 01-01 was also given to the operators. Based on the above information, the item is i
closed.
- - - - - - - - - - - - -. - - - - - - - - - - - _
. _
_ _ _ _ _ _ _ _ _ -
_.
.
.
.
b.
(Closed) Violation 325, 324/85-01-01.
Failure to Remove a Licensed Operator from Licensed Duties Subsequent to Failing the Annual Requalification Exam.
The inspector reviewed TI-200, Rev. 10, Brunswick Plant Operator Retraining, and found that the procedure had been revised to provide more definitive guidance for identifying and administering requalification requirements.
Based on this information, the item is closed.
c.
(Closed) Violation 325, 324/85-24-01.
Failure to Implement AI-59.
The inspector reviewed AI-59, Rev. 9, Jumpering, Wire Removal, and Designated Jumper, and found the revision requires the Operations Engineer to provide the Plant Nuclear Sefety Committee (PNSC) with updates on a monthly basis.
Based on this information, tne item is closed.
d.
(Closed) Unresolved Item 325, 324/85-01-03.
Interpretation of
" Actively Performing the Functions of an Operator or Senior Operator" with Regard to 10 CFR 55.31(e).
There were no clearly defined guidelines available to the licensee at the time this item was identified. The recently issued 10 CFR 55 provides clear guidance in this area, and also in the area of actions required to resume active performance of licensed duties after an inactive period.
Because of the previous lack of guidance and subsequent issuance of guidance, this item is closed.
l i
e.
(Closed) Unresolved Item 325, 324/85-01-05.
Interpretation of
" Annual" for Requalification Program. The recently issued 10 CFR 55 has modified the requalification program to require a written exam be administered not longer than once every two years, and an operational exam be given annually. Because of the modification in the recently issued 10 CFR 55, this item is closed.
4.
Unresolved Items l
Unresolved items are matters about which bora information is required to determine whether they are acceptable or may involve violations or deviations.
The new unresolved item identified during this inspection is discussed in paragraph 5.e.
5.
Operations (71707)
a.
Review of Licensee Event Report (LER) 1-86-024, Automatic Reactor Scram Resulting From Loss of Main Generator Output Voltage.
l
<
While controlling Unit 1 main generator output voltage wi.h the manual potentiometer of the voltage regulator, an automatic scram from 100 percent power occurred at 1058 hours0.0122 days <br />0.294 hours <br />0.00175 weeks <br />4.02569e-4 months <br /> on September 13, 1986.
,
The generator voltage was under manual control to permit the on-line cleaning of the Automatic Voltage Regulator (AVR). Erratic operation of the Manual Voltage Regulator (MVR), combined with the inability to
_ _ - - _ _ _ _ - _ - _ _ _ _ _ _ _
_ -
._
- _ _ -
-_
_
.
.
i effect a timely retransfer to the AVR resulted in:
de aced volta l
on the emergency buses; Main Steam Line Isolation Vafve (MSIV) ge closure; and a reactor scram on main steam isolation valves (MSIV)
less than 90 percent open.
The LER attributed the scram and transient to the build-up of an oxide film on the MVR potentiometer and the resultant voltage oscillations.
,
The indicated corrective action was to replace the voltage regulators
'
on Unit 1 with another type of regulator.
The Plant Incident Summary (PIS) indicated that the root cause of this event was a procedural deficiency.
The PIS stated " Plant procedures did not adequately cover vendor recommended preventive maintenance of the potentiometer wiper and slide."
This root cause was not addressed in LER 1-86-24.
In. addition to replacing the voltage regulators, the LER " package" indicated a long term corrective action to evaluate cn-line maintenance of the regulators to determine if improvements could be made in preventive and corrective maintenance procedures to reduce the risk of similar scrams.
This corrective action was also not addressed in the LER.
-
In addition to the oxide film on the MVR and AVR resulting from inadequate preventive maintenance, there appears to have been other contributing root causes which were also not addressed in the LER or the LER investigation package.
As described in the LER, the cleaning of the MVR just prior to the transfer from AVR to MVR appeared to be inadequate to remove all of the oxide film.
In addition, the licensee's chronology of events in the LER indicates that the
,
operator began experiencing severe voltage fluctuations just one
'
minute af ter the AVR/MVR transfer.
The operator could not transfer back to the AVR because the technician had already started work on the AVR and there was a perceived potential shock hazard.
Following the transfer to the MVR less than a minute had been allowed for the voltage control to stabilize before the technician started cleaning the AVR.
When the reactor operator began experiencing problems controlling the voltage swings with the MVR, he attempted several times to reach the technician at the AVR on the Public Address (PA)
system.
These attempts were unsuccessful, and since the operator believed he could not return to the AVR without contacting the technician to ensure he was clear of shock hazard, the voltage swings resulted in a degraded voltage on the emergency buses and, eventually, a turbine trip and reactor scram.
Subsequent investigation by the licensee determined that the PA speaker in the area of the AVR had the volume reduced to a less than adequate level for the ambient noise level.
The licensee does not have a portable radio system for communicating, and relies totally on the PA system for normal and emergency communications.
The prohibitive cost and the potential for causing a unit trip through the tripping of relays were reasons cited by the licensee for not obtaining a radio system.
A proposed modification to the PA system
_ _ _ _ _ - _ _ _
- _ _. - _ _
_ _ _
-
~
.
would allow the override of all locally adjusted PA volu' me controls
' to facilitate emergency messages or alarms, but the approval or installation date has not been established.
Had there been an adequate time period allowed for stabilization of the-MVR prior to the technician starting work, or had.the communications been adequate, the operator could probably. have accomplished a transfer back to the AVR in time to prevent the turbine trip and reactor scram.
There was no procedure written to support the transfer because it was considered a " simple" evolution.
A procedure could have ensured a. testing of communications prior to L
the transfer, and an adequate stabilization period'for the MVR before commencing work on the AVR.
Following the MVR failure on Unit 1, the' licensee did not take timely
'
action to ensure that the Unit 2 MVR and AVR were not subject to the-same failure mechanism. The voltage regulators were not replaced as they were on Unit 1, and were not cleaned to remove any oxide film.
Indications are that Unit-2 experienced degraded performance with oscillations worse than those experienced on Unit 1 prior to the September.13,1986 event. There are also indications that there was a reluctance to perform on-line cleaning of the Unit 2 AVR due to the
'
potential for a similar trip. On January 5,1987, the Unit 2 AVR
experienced similar oscillations'resulting in a turbine trip and full power scram.
Additional details.- on this similar occurrence is presented'in paragraph 5.c.
During the recovery period from the September 1986 scram, the licensee experienced a series of additional equipment failures-including High Pressure Coolant Injection (HPCI) flow oscillations, failure of the.Peactor Core Isolation Cooling (RCIC) system suction relief valve, failure of a Reactor Protection System (RPS) motor generator set output breaker to close, reactor feed pump trip, backup nitrogen failure, momentary loss of full core display and CRT, failure of safety relief valve sonic detectors and memory lights, and failures of intermediate and average power range monitors.
Not all
- of these failures were included in the LER, even though they appear to have contributed to the transient, and the corrective actions for several of these failures were also not addressed.
The following is a list of the more significant of these additional equipment failures:
(1) Following the scram, the HPCI and RCIC system tripped on high level. When the operator subsequently restarted HPCI, a dual flow path was established with injection through the F006 injection valve and flow back to the Condensate Storage Tank (CST) through the full flow test line and valve F008. This dual flow path was initially blamed for the flow oscillations experienced on HPCI during this event, but the licensee later determined that the governor was at fault.
Since there is no check valve installed between the two valves involved in this
_ _ _ _ _ _ _ _ _ - _ _ - _ _ _ _ _ - _ - _ - _ _ _ - _ _ - _ -
I
.
.
'6 dual alignment, a backflow can be established from the HPCI
)
injection line to the full flow test line and CST when the pump is secured. The RCIC system has been backfitted with a check valve to prevent this occurrence.
Hot feedwater from the feedwater line which enters the HPCI injection line, apparently backfeeds to the full flow test line.
During this event, it l
appears that. this scenario resulted in severe water hammer in l
the HPCI piping and potential damage.
The licensee's LER i
package indicates that the HPCI, RCIC, and RHR loops "were walked down and inspected for piping and support damage prior to
'
restart." Neither the water hammer nor the piping inspections
.
were addressed in LER 1-86-024. A similar HPCI dual flowpath
.
and the resultant backflow and water hammer occurred following
'
a scram on March 11, 1987. Additional details on this repeat event are contained in paragraph 5.d.
(2) Following the reactor scram on September 13, 1986, the licensee was initially utilizing the RCIC system to maintain vessel level.
The RCIC suction relief valve, E51-F017, failed to reclose following actuation however, and the system could not maintain level.
This open relief valve allowed the discharge of approximately five inches of water in the RCIC corner room.
The licensee declared RCIC inoperable and subsequently repaired the failed relief valve.
The valve disc was determined to be cocked in the disc guide assembly.
The valve internals were cleaned and the relief valve rebuilt.
(3) Safety Relief Valves (SRVs) A and L cycled open, as indicated by their tailpiece temperatures, but were not indicated open by their sonic detectors.
The A valve was subsequently tested with satisfactory results. The L valve experienced a defective accelerometer in the sonic detector.
During replacement of the L accelerometer, a splice in the cable was noted to be defective.
The cause was not definitely determined, however, two possible explanations were presented.
One was that the splice was defective and the other was that the technicians may have caused the damage while removing the heat shrink material.
An Engineering Work Request (EWR 05005) had been submitted on January 27, 1987, for the sonic detectors on both units. With trouble tickets being issued at approximately one per month for two years, an EWR should have been issued much sooner to address the problem or at least initiate consultation with the vendor.
Trouble tickets that were written for the sonic detectors do not mention the speed at whicn the SRVs cycled.
Discussions with both operators and I&C technicians indicated that the valves were cycled rapidly.
Discussions with I&C technicians also indicated that very few of the technicians actually understood how the system worked, but could follow the procedure for
____ ______ ______
o I
.:
calibration.
One reason for this may be the inadequacy of technical materials available. The technical manual indicated a time delay circuit in the sonic detector instrumentation.
The
'
inspector recommended that the licensee contact the vendor and determine if there is an adjustment that can be made for time response.
The licensee agreed to revise the test calibration procedures to include time response acceptance criteria, if the discussions with the vendor indicate that it is appropriate.
(4) Another problem during this transient was Intermediate Range Monitor (IRM) spiking.
Since this incident, the licensee has consulted the vendor and has prepared a plant modification for each unit. The modification involves a short term fix of noise suppression circuits and insulated ground straps. A long term fix of replacing and re-routing instrument cable was also presented by the vendor and is under consideration by the licensee.
b.
Review of LER 1-87-001, Failure of Unit 1 High Pressure Coolant Injection System Turbine Steam Supply Valve to Open.
On January 26, 1987, the high pressure coolant injection (HPCI)
turbine steam supply inboard primary containment isolation valve, E41-F002, would not reopen following a surveillance test. The HPCI system was declared inoperable with Unit I at 81 percent power.
The licensee's investigation indicated that the F002 valve would not open due to failure of the auxiliary contact block assembly of the valve motor close contactor.
The contacts failed in the open position preventing energization of the valve motor contactor. The failed auxiliary contact block assembly was replaced and HPCI restored to an operable status approximately 14 hours1.62037e-4 days <br />0.00389 hours <br />2.314815e-5 weeks <br />5.327e-6 months <br /> following the event.
During the period that the licensee was investigating this event, two additional failures of safety-related ECCS valves occurred due to the same type of contact block assembly. Additionally, 10 safety-related valve failures have been attributed to this same design contact block assembly at Brunswick within a four week period.
l In response to the inspectors concerns regarding the potential for additional contact block assembly failures and the affect on the ECCS, the licensee committed to inspect the assemblies on the AC j
operators for all major flow path ECCS valves on Unit 1 before restart from the refueling outage.
This commitment will be an inspector followup item 325/87-12-04.
The vendor had recommended that the licensee either inspect these contact block assemblies for signs of binding as discussed above, or that the assemblies be replaced with a new design not susceptible to binding. The licensee should also address corrective actions for other valve operator contactors utilizing these assemblies, including Unit 2 ECCS valves and other safety-related valve applications.
In addition, the
. _ _ _ _ - _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ - _ -
.<
8 licensee is investigating the failure mode of these contact block assemblies.
If this investigation determines that this specific component (GE-CR205X1000) to be defective, the information should be reported under the requirements of 10 CFR Part 21.
c.
LER-2-87-001, Automatic Reactor Scram Due to Main Turbine Control Valve Fast Closure Resulting From Loss of Main Generator Excitation Voltage.
On January 5,1987, an automatic reactor scram on unit 2 occurred from 100's power.
Primary containment isolation also occurred
,
accompanied by autostarting of the high pressure coolant injection (HPCI) and reactor core isolation cooling (RCIC) systems and the emergency diesel generators.
The LER attributed this reactor scram to dirty and/or corroded wiper surfaces on the automatic and manual potentiometer associated with the main generator voltage regulator.
The generator trip, which caused the reactor scram, occurred due to a loss of excitation voltage resulting from erratic generator output and voltage oscillations.
A similar generator trip and reactor scram had l
occurred on September 13. 1986 (LER-1-86-24).
In response to this l
previous event and vendor recommendations, the licensee had planned l
to discontinue external cleaning of the potentiometers, and to substitute manual rotation of the potentiometer to clean the surface.
The procedure change which would implement this revised method of
,
maintaining clean potentiometer surfaces, had not been implemented l
for Unit 2 at the time of the January 1987 event.
There are indications that the Unit 2 voltage regulator was experiencing increasingly erratic operations prior to the January 5 trip, but the licensee was reluctant to utilize external cleaning to remove the oxide film buildup due to the vendor recommendations and the previous trip that occurred during external cleaning.
Prompt corrective actions in response to the erratic Unit 2 voltage regulator operation such as external on-line cleaning, unit shutdown and cleaning, regulator replacement, or earlier implementation of the revised cleaning procedure would have probably prevented this second trip and
,
Following this event, the licensee replaced the unit 2
'
voltage regulator potentiometers with a design that is sealed to reduce the probability of oxide film buildup.
The HPCI ar,d RCIC system automatically initiated during this event due to a icw level (LL) 2 signal, but did not inject, according to the lice.1see, due to the momentary duration of this LL2 signal.
Following roset of the PCIS i solatior, signal, the HPCI and RCIC systems tripped on high reactor water level. The reactor water level subsequently decreased and HPCI was manually initiated to restore the level.
Once the reactor vessel water level was restored, the operators placed HPCI in the full flow test mode and closed the HPCI inboard injection valve (E41-F006). The RCIC system was subsequently l
.
_ _ _
. _ _ _ _ _ - _ _ _ _ _ _ _
_ _ _ _ _ _ _ - _ _ _ - -
.
.
L started ' manually and aligned for reactor injection to maintain reactor water level, but did not provide adequate capacity to main-tain vessel level.
In response to the decreasing reactor vessel i
l water level, the operators' attempted to realign HPCI to the injection mode.
This attempt was unsuccessful because the HPCI inboard 6njection valve (F006) would not reopen.
A local inspection determined that the valve motor operator breaker had tripped on thermal overload and breaker magnetic trips.
These trips were reset but the valve still failed to reopen. Manual opening of this injection valve was not possible due to a recent environmental qualification (EQ) modification which moved it to an inaccessible l
location.
In an effort to stop the decreasing trend in water level, the operators secured the reactor water cleanup (RWCU) rejection, and increased control rod drive (CRD) cooling water flow to the vessel.
The operator also closed a redundant full flow test shutoff valve E41-F011 and gave an additional close signal to E51-F022, the RCIC full flow test isolation valve which already indicated closed in the control room. The reactor water level continued to decrease to LL-1 resulting in a reactor scram signal and PCIS isolation.
Nearly two hours later, an operator found the F022 full flow test valve approximately 50 percent, or 110 turns, open despite the closed control room indication.
The licensee's investigation determined that an anti-rotation device had been improperly installed allowing the device to slip. This in turn resulted in premat.ure picking up of the limit switch and a failure of the valve to fully close. The F022 valve, which was essentially in the "open" position, provided a full flow test path diverting flow from injection and rendering the RCIC incapable of performing its intended function.
Under these plant conditions, with HPCI inoperable and RCIC unable to restore reactor water level, the only water supply immediately available to the vessel was CRD cooling flow of approximately 200 GPM. With the decay heat and the safety relief valve actuations required to maintain reactor paessure, this CR0 flow was inadequate to maintain reactor water level.
With both HPCI and RCIC inoperable there were no high pressure cooling systems available to supply the vessel which was at operating pressure.
Any small break loss of coolant accident (LOCA) under these conditions would have exceeded CRD cooling water flow, and required actuation of the automatic depressurization system (ADS) to lower reactor pressure to allow the low pressure cooling systems to inject into the vessel. Actuation of ADS and depressurization of the reactor results in a severe transient on the plant and can cause momentary uncovering of the core. This potential for a concurrent loss of HPCI and RCIC ano activation of ADS had been addressed in the Onsite Nuclear Staff (ONS) Report on the operability of the HPCI system.
Based on a cost-benefit analysis, however, the only major action recommended by ONS to improve the reliability of HPCI and RCIC was to increase the frequency of surveillance testing.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_ - _ - - _ _ -
- - - _ _.
h
.
.
l d.
LER-2-87-004, Reactor Scram on Low Vessel-Level.Due to a Loss of the
! P Uninterruptiple Power Supply and Subsequent Loss of Feedwater Flow to ~
j.'
the Vessel.
.
L On March 11,1987, a unit 2 reactor scram occurred from 100 percent power due' to low. water level in the : reactor vessel. The low vessel l
>
level was caused by'a loss of the uninterruptible. power supply (UPS)
which in turn caused. the signal to the feed. pump control system to i
ramp back to. minimum speed.
i
. On.. Marc h 10,:1987, at approximately 1900 hours0.022 days <br />0.528 hours <br />0.00314 weeks <br />7.2295e-4 months <br />, operators noted that L
the'inW t power to'the UPS was-cycling between the normal and reserve power supplies. At approximate 1y' 2200 hours0.0255 days <br />0.611 hours <br />0.00364 weeks <br />8.371e-4 months <br /> the power supply again transferred to the reserve power source and the Shift Foreman'placed
{
thel static ~ transfer. toggle switch to manual.
This action locked the' UPS on. the reserve power ' source and prevented damage due to additional cycling.
The operator was then directed to bypass the
~
statit transfer switch due to possible static' switch failure. This L
evolution was accomplished utilizing Section. 8.5 of. Operating Procedure OP-52, 120. Volt UPS, Emergency, and Conventional Power Systems. Completion of this section of the procedure established the following UPS conditions:
UPS supplied from reserve source-
-
-
. Static switch bypassed Static transfer toggle switen (S3)' in Manual l
-
Manual transfer switch (S4) in Reserve
-
At this point the Shi.*t Foreman decided to transfer the UPS from the reserve source to the standby power converter source per Section 8.2 of OP-52. When step 8.2.6.a was reached, the procedure required the
,
operator to verify the manual transfer switch was in the automatic l
'
position. As noted above, however, the performance of Section 8.5 of OP-52 had left this switch in the Reserve position. After reviewing the procedure and the system logic with the Shift Technical ~ Advisor (STA) and the Shift Operating Supervisor (S0S) for several hours, the Shift Foreman decided to have the Auxiliary Operator (AO), who is non-licensed, " backup" through the procedure to a step where a temporary procedure change would allow transfer of the S4 switch from Reserve to Automatic. While backing up through the steps of
'
Section 8.2 of OP-52, the operator failed 'to close breaker CBB, Reserve Bus to Load Switch, before opening breaker CBA, Reserve Bus to Static Switch. Opening CBA with CBB also open resulted in a loss of the UPS, a reactor scram, and a loss of numerous indications and controls supplied by the UPS including:
Turbine electric - hydraulic control (EHC)
-
Feedwater control
-
-
Reactor manual control and full core display
-
-
I
- - - - -... _ - _ _ _ _ _ - _ _ - - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _. - _ _ _ _________
_ - _ _ _ - - _
_-
_
.
L Turb'ine supervisory instrumentation
-
Process computer
-
-
One' reactor water level indication Balance of plant instrumentation
-
This loss of UPS and resultant. plant transient could have been prevented if the transfer of the 54 switch had been properly planned, proceduralized, and supervised.
Although the licensee s administra-tive procedures appear to have permitted " backing up" in a procedure, this is a poor practice which can cause serious problems.
Revising the steps of a procedure can result in undetected detrimental effects on system logics, which can later interfere with the design operation of tnat system.
In addition, reversing the steps without rewriting them requires the operator to remember that all steps are to be performed the opposite of what the procedure requires, i.e., open means closed and closed means open.
In this particular event, Step 8.1.b required the operator to verify breaker'CBB in 0FF when he was in fact required to CLOSE the breaker, placing it in the UN position.
There were a-number of other steps that the operators also had to perform in reverse of, procedure requirements.
It is not difficult to j
understand the confusion that could result, or how a step could be missed.
Reversing the steps of a procedure represents an " intent change" to the procedure, and is not permitted Technical Specifications without prior review and approval b PNSC and the Plant Manager.
If the steps of a procedure must be reversed, the steps should:
(1) be rewritten in the correct order in which they are to be
accomplished.
l (2) be written as they are to be performed (i.e.:
if the operator is to open the breaker the step should state open not closed.)
(3) be reviewed by an adequate level of supervision or management to
'
ensure that interference with system logic or equipment damage will not result from the reversal of steps (4) be provided with check-off or initial blocks to ensure that all steps are performed and in the correct order.
Interviews also indicated that the non-licensed operator who was dispatched to perform the procedure in reverse was not clearly knowledgeable of what was trying to be accomplished and was not directly supervised by a licensed operator or an SRO.
The Shift i
Foreman, who accompanied the operator throughout the earlier UPS evolution, returned to the control room to complete associated paperwork during the evolution which resulted in a loss of UPS.
There was also no communication between the control room and the operator on each step to ensure that the correct action was performed and was in fact accomplished.
__-_-_____--__ -___ - _
_
_ _ _ - _ _ _ _ _ _ _ - _
s
.
.
,
j i
In response to this event the licensee revised OP-52 to allow placing the S4 switch in automatic if in reserve.
Direction was also provided by management not to "back-out" of procedures pending further evaluation.
In response to the additional concerns expressed by i
the inspectors, the licensee committed to revise administrative i
procedures to preclude the practice of backing-out of procedures except in valid emergencies with prior approval of plant management.
This commitment will be inspector followup item 325, 324/87-12-02.
Following the reactor scram on low water level, a ' Group I primary containment isolation occurred, HPCI-and RCIC automatically initiated, the recirculation pumps tripped, and standby gas treatment initiated.
As reactor'essel level increased, the operator reduced the HPCI flow l
v rate utilizing the flow controller.
In preparation to divert HPCI
'
flow from the vessel to the full flow test line and condensate storage tank (CST), the operator opened the common test return line valve (F011).
1This operator action established a dual flow path between HPCI injection to the vessel and HPCI full flow test to the CST. As in the September 13, 1986 event, the dual flowpath, in the. absence of a check valve, can result in hot feedwater (400 F) backfeeding into the full flow test line on a HPCI pump trip. The HPCI. pump did trip on high level, and this backfeed of hot water into the full flow test line resulted in severe water hammer and line heating.
During a subsequent control board review of HPCI alignment, the=
operator noted that the HPCI outboard injection valve, F007, did not have any control power /open-closed light indication. An operator was dispatched to the F007 breaker and it was determined that the F007
)
valve, which was supposed to be open was-closed. Multiple interviews-with licensed operators failed to determine how this valve was closed as only F006, the inboard injection valve, automatically closes on a HPCI trip.
In addition to the valve being closed, the valve operator thermal trips and breaker magnetic trips were in the tripped condition.
The operators attempted twice to open the valve by resetting these trips, but after each attempt the trips reactivated and F007 could not be opened. An attempt to manually open this valve was also unsuccessful due to binding. At this point the'HPCI system was declared inoperable.
The' licensee determined that the F007 valve operator motor had burned up.
This failure was attributed to overloading caused by thermal binding of the valve in its seat.
This thermal binding was attributed to the backflow of feedwater into the full flow test line as addressed above.
Following a cooldown period, _the licensee was able to back the F007 valve out of its seat utilizing the normal torque switches.
The loss of control power indication for the F007 valve and the thermal overload and breaker magnetic trips prior to the attempt to
,
open the valve seem to indicate that the overload and trip may have occurred on closure of the valve into the seat, rather than on
_ _ _.
_ _ _ _ _ _ _ _ _ _ _ _
.
.
..
. _ -- ___
i
.
.
j i
opening, as speculated.
In addition, had the potential effects of the dual flow path and reverse flow previously experienced with j
the HPCI system been adequately evaluated following HPCI/feedwater backfeed and water hammer during the September 1986 event (LER 1-86-024), this failure may have been prevented.
One of the corrective actions taken for this March 11, 1987 event was to revise procedures
!
to require operators to secure one HPCI flowpath, such as full flow test, prior to initiating the remaining flowpath injection. This l
corrective action was discussed following the September 1986 event L
but was never implemented. Additionally, the HPCI inboard injection valve, F006, also failed on thermal overloads and breaker magnetic trips in January 1987 (LER 2-87-001), but the potential contribution of backflow was not adequately evaluated or corrective actions taken.
The licensee did request a vendor to test the valve operator utilized on F006 to determine if it should have been able to operate the valve against the thermal and pressure differentials experienced.
The vendor's conclusion was that the operator should have been capable of opening the valve. The HPCI outboard injection valve, F007, utilizes the next smaller operator than the F006 valve. The licensee did not request the vendor to conduct a similar evaluation of the operator utilized on F007. Another corrective action taken by the licensee was to require a 10 second closure signal to the F008 full flow test valve when securing HPCI full flow test. Although the F008 valve indicated closed during this event, it was suspected that it was actually partially open contributing to hot feedwater backflow to the
.
CST and water hammer. The licensee also walked down the HPCI system
to ensure that the severe water hammer had not resulted iri damage to
{
!
piping, valves, or instrumentation.
The permanent solution to the HPCI/feedwater backflow problem appears to be installation of a check valve, a modification already accomplished on the RCIC system.
The check valve would prevent the backflow of hot feedwater into the full flow test line.
It would also allow a dual flow path which could help prevent high level HPCI trips following initiation by permitting a diversion of flow to the l
CST. Additionally, this check valve could resolve the line break EQ concerns that required moving the F006 inboard HPCI injection valve into the MSIV pit.
This would allow moving the F006 valve to an accessible location for manual opening in an emergency.
As in previous examples, the LER generated in response to this event (2-87-004) appeared to be less than adequate. There were errors in the sequence of events on terminology addressing the UPS trip as well as in the sequence of events surrounding the F007 valve failure.
Other deficiencies that occurred during the transient were only l
mentioned briefly in a table of events, and their root causes and I
corrective actions were not addressed.
Examples included safety relief valves which apparently had incorrect setpoints and did not auto open, an apparent problem with the logic involving high torus level and the torus suction valves, and the failure of 6 control rods to drive full in (the rods stopped at 02 position).
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
--
.
-
-
_
-
_
_ _ _ _ _
.
.
e.
High Pressure ECCS Systems, HPCI and RCIC, Inoperability The four aforementioned LER's, 'l-86-024,1-87-001, 2-87-001, and 2-87-004, all involved on-demand failures of HPCI and/or RCIC.
In a seven month period between Seatember 1986 and March 1987, there were three on-demand failures of H)CI and two on-demand failures of RCIC, all related to valve failures.
In at least one of these events, both the HPCI and RCIC were inoperable, leaving only CRD to supply coolant to the vessel and presenting the potential to utilize ADS to depressurize the vessel for cooling on even a small-break LOCA.
To reiterate the concern over the operability of the Brunswick high pressure ECCS, a brief chronology of the recent failures is included below:
(1) September 13,1986,LER1-86-24(Scram 100percentpower)
HPCI experienced significant flow oscillations which
-
resulted in an initial declaration of inoperability.
These.
oscillations were first attributed to a dual flowpath and later to g,overnor instability.
RCIC suction relief valve failed open, and RCIC was
-
declared inoperable.
Water hammer attributed to a dual HPCI flowpath had the
-
potential to damage piping, valves, and instrumentation.
(2) January 5, 1987, LER-2-87-001 (Scram 100 percent power)
HPCI F006 inboard injection valve failed to reopen
-
following isolation resulting in HPCI being declared inoperable.
A recent plant modification prevented manual opening of the valve.
RCIC full flow test valve, ing declared inoperable.F022, stuck in
-
resulting in RCIC also be The failure of F0e2 was attributed to incorrect installation of the anti-rotation devices.
l 1987,LER-1-87-001(81 percentpower)
(3)
January 26,ing a surveillance test, the HPCI steam isolation Follow J
-
valve, F002, would not reopen resulting in HPCI being declared inoperable.
The failure of the F002 valve was caused by the failure of a contact block assembly.
Failures of major ECCS system valves due to this same type of contact block assembly occurred twice in the two weeks followi'ig this failure.
Additionally,10 safety-related valves failed due to this type of contact block assembly in a four week period.
-
The vendor informed the licensee that the failures of the control block assembly were not design related, and neither the vendor or the licensee have reported the failing part under 10 CFR Part 21.
_ _ - _ _ _ _ _ _ _ _ _ _
________ _
_
_
_
._. -
-
- _ _ _ - _
- - _ - _ _ _ _ - _ - - _ - _ _ _ _
..
.
.
15-(4)- March 11, 1987, LER-2-87-004 (Scram 100 percent power).
Following. initial. automatic initation. and trip on.high
-
level,. the HPCI outboard injection (F007) valve was found out'of position with no plausible explanation.
HPCl outboard injection valve (F007);could not be reopened
-
electrically or manually; therefore, the HPCI was declared -
inoperable. The licensee. attributed the failure to thermal
.
binding caused by a backflow of hot feedwater.
This backflow resulted when a dual injection and full flow test path was established for HPCI, as-previously occurred in September 1986.
These events--indicated that virtually every time the HPCI and/or RCIC high pressure ECCS systems have been called on to' perform their intended function in recent months, they have failed. Each time HPCI was isolated,. three in all, it could not be restored to operability -
by the operator.
The flicen'see's Onsite Nuclear. Safety (ONS) Group. had performed a study of.HPCI. operability in 1985 which did not include the failures-addressed here. The basic. conclusion of the report' was-that ' the
'
decrease in J core ' melt frequency that would be realized by major system modifications did not justify the cost involved.
The only major recommendation was to increase the frequency of HPCI system surveillance testing.
The licensee has more 'recently conducted-an internal Safety. System
-
Functional Inspection (SSFI) of the HPCI system. This SSFI included an evaluation of the impact of the latest HPCI failures, and a simulator evaluation of the cperators' knowledge of the HPCI system and valve logics. The results of the licensee's inspection were not yet assembled, and the report and proposed corrective action were not available during this. NRC irispection.
The licensee agreed to provide the NRC with the results of the SSFI.
The multiple HPCI and RCIC system and valve failures will be an unresolved item 325,324/87-12-01 pending further review of the'
licensee's HPCI, system SSFI, and the licensee's proposed corrective
I actions.
!
f.
Control Room Observation The inspectors observed activities in the control. room' for the day i
and evening shifts on May 5, 1987. These activities included a shift j
turnover.
The inspectors noted that the shift turnovers were conducted in accordance with checklists provided in procedure 01-02,.
!
Control Operator Shift Checklist.
The inspectors subsequently reviewed the status of equipment under clearance and alarms to i
,
.. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _. _ _
_
_ _ _
._.m_
_.. _ _ _
. _ _ _ _ _. _ _ _ _. _. _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _. _ _ _ _ _ _ _ _ _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _. _ _ _.
_ - -. __.
.
_ _ _ _
_
- _ _
_
_ _ __
- -_- _ - _ - - _
,.
i
.
"4
,
determine if the operators were. cognizant of plant conditions. The
- operators = appeared knowledgeable of plant conditions and control board indications.
During the review of valve lineups, the inspectors noted that the licensee was utilizing a valve / breaker exception form to document changes in valve lineup-that were different from. the lineup established by operating procedures-or by clearances.. Use of - the
'
valve / breaker exception form is described in 01-13, Rev.14, Valve and Electrical Lineup Verification, Valve Identification,-and Locked :
Valve Identification and. Locking.
The inspector questioned-independent verification. of restoration of the valve lineups since the valve / breaker exception' form did not indicate independent verification.
The licensee stated that the' operating procedure lineups. were utilized to document the independent verification. When the original-valve / breaker lineup is completed using' the. operating procedure, valves / breakers which.do not meet the required position are left -
blank or a note is made that they are out of position. An entry is made on the valve / breaker exception form for each out of position.
valve / breaker. When the valve / breaker is subsequently returne_d to
- position, the. initials for the. independent verification are made :in the operating procedure lineup by one of three methods:
1) entering initials in the' blanks, 2) marking out notes and entering initials,
,
or 3) marking out initials from the last verification and entering l
new-initials.
Entries for-the third method were the only ones. that were dated.
The inspector was concerned that the documentation for the initial lineup was being revised at a later date when the exception was cleared without indicating that entry was made and the the verification performed at the time the exception was cleared.
The
- exception form, which is the QA record of meeting independent verification requirements was therefore not clear.
The licensee stated that their QA/QC staff would examine the method of documentation and recommend changes' as needed to assure that the record was auditable.
The inspector did not identify any valve / breaker lineups in which it was clear that independent verification was not performed. The inspector interviewed several operators who stated that the methods described above were utilized for independent verification. of valves and breakers restored to service'from the valve / breaker exception list.
The inspector was informed by licensed operators that once the valve / breaker exception form was completed during startup, that the form was not reviewed except when lifting a clearance associated with a valve / breaker on the form or when a valve / breaker was added to the form. The inspector reviewed the valve lineup sheets for the high pressure injection system, the reactor core isolation cooling system, the residual heat removal system and the plant electrical lineup.
Although, these lineups had valve / breaker exceptions listed, the
._ _______ ____________ - _ -
- _ _ - _ _ - _ _ _ - _ -
_ _ _ _. _ _. _ - --.
- - _ __
--
- - _ - - _ _ -
t
-
,
.
'inspector did not find any' required for system operability.
The inspector was concerned that the review of the exception forms was not frequent enough to keep the operators informed of those valves /
breakers which were out of position but not _ captured under the-i clearance program.
The. licensee committed to. review the valve /
'
breaker exception' form on a more frequent interval in order to assure
. operators were aware of the status of equipment and were cleering out-of position valves / breakers on a timely basis.
The licensee indicated the the interval would be at least quarterly.
During the review of the reactor core isolation cooling-system lineup, a note was observed on the valve / breaker exception form that indicated that valve ' 2-E51-V8, the RCIC steam turbine admission valve, was closed from April 31, 1986 to August 6, 1986. This valve was required to be open during the startup which occurred during June
)
1986.
The licensee provided documentation that indicated that valve 2-E51-V8 was opened and tested during the startup. The valve / breaker-exception form, however, was not corrected until August _6, 1986. The'
- i inspector interviewed the Shift Foreman who made the August 6,1986
' '
entry. The Shift Foreman indicated that he could not recall exactly.
what had happened but that he must have had the valve independently verified on August 6,1986 upon discovery of the. incorrect exception
, form.
Initials indicating independent verification were entered for the August 6 date..The shift foreman stated that the valve was tested during the startup and that control room indication was available for the position. The position is verified during control board walkdowns, and alarms indicate if the valve is. closed.
The licensee provided a copy of procedure GP-01, Rev. 102, Startup Checklist, that indicated that the valve / breaker exception form was reviewed during plant startup.
It was determined that the position of E51-V8 should be " closed" on the valve checklist, rather than
"open" since that is the position required during initial startup.
This would allow the operator to verify that the position was correct for startup and therefore not enter an exception on the form.
The valve would be positioned to the open position in accordance with GP-02, Approach to Criticality and Pressurization of the Reactor, and PT-10.1.3, RCIC System Operability Test - Flow Rates.
The licensee stated that the valve lineup would be revised.
The inspectors noted that the licensee was utilizing a combination of full sized information and protective tags, along with improvised plastic caps to denote control board controls and indicators under maintenance or which were defective.
These full sized tags can inadvertently cover controls and indications and are a distraction to operators. This type of tag was determined to have interfered with indications and controls during the TMI event.
The licensee should evaluate alternative methods of providing control board status indications such as the small round tags,- or magnetic numbers and associated status lists, utilized at a number of other facilitie.
.
4 7.
Maintenance (62700)
a.
' Maintenance Program Review The inspectors conducted a review of the licensee's maintenance
! program at the Brunswick facility.
The inspection included reviews of maintenance procedures, routine maintenance surveillance.
activities, selected work requests / job orders (WR/J0) associated with
. selected LERs, in process work activities, maintenance data trending and interviews of maintenance management and staff.
Routine maintenance work was evaluated to ensure that: the activi-ties were completed in accordance with approved procedures, Technical Specifications, and the appropriate industry standards; procedures used were adequate to accomplish the tasks; QC hold points were established 'and correctly signed as required; maintenance activities requiring tagouts were obtained and approved prior t6 work initiation; replacement parts.and materials were prope~rly obtained and certified;
and'that these activities were accomplished by qualified personnel.
The. inspectors' obtained copies of WR/J0s and verified that these.
I work documents had been initiated and issued in accordance with the appropriate administrative controls.
The following WR/J0's were reviewed:
87-AEBD1. Replace Manual Voltage Regulator Potentiometer 87-3LCZ1 Clean Main Generator Automatic Voltage Regulator 87-AAKK2 Investigate and Repair Improper Operation - of Valve.
2-E51-F002
'87-ABBM1 Drain and Flush Sensing Line for 2-SW-PS-1176B 83-AAGR1 4 KVAC Instant Overcurrent Rod Relay Calibration 87-AFFF1 RHR Walkdown on Plant Modification 84-063
87-AFFG1 RHR Walkdown on Plant Modification 84-065 87-AFRIl Walkdown of Interruptible and Non-Interruptible Instrument Air 87-AKYE1 Annunciator System:
Calibration of Pressure Instrument on No. 2 Diesel Generator 86-BPCR1 Annunciator Calibration on 24 VDC Battery Charger 86-BPCQ1 Annunciator Calibration on 24 VDC Battery Charger The inspectors held discussions with maintenance supervision and found that the. licensee has implemented a new Maintenance Trending Analysis Program for the Brunswick site.
The program aids maintenance in identifying potential problem areas by allowing the maintenance. supervision to quickly note an increase in the use of a particular part or material (i.e., oil, water, lubrication, etc.).
Maintenance personnel review the data on a weekly basis to look for adverse trends.
l
-.
.
-
-
_
_--_-- _------_-_- _ _ - -
_-_
_ _ -
.
.__ _
. _ _ - - - _ _ _ _ _ _ _ _
i
.
s
Ins addition to theEabove noted improvement prograns, 'the. ' licensee indicated that they were in the process of actively reviewing problems and/or failures from a generic standpoint.
Examples. of these cases include:
!
Valves with anti-rotatian devices similar to reported failures
-
were reviewed.to ensure that the identified. problem.was
corrected, LER 2-87-00'..
j P
RCIC suction relief ' valve setpoint drift problems wereLreviewed
-
.for all similar valves, LER 1-86-024.
Diode rings (firing rings) for the motor control circuit on the
!
-
Ocean Discharge Pump Motor were replaced with an improved part for all firing circuits when the original diode failed.
Inspection of the maintenance program identified that the licensee's staff was making a consistent effort to improve the program at the-Brunswick facility.
Discussions with both the management and staff revealed that' personnel were professional and knowledgeable in their daily work activities.
No violations-or deviations were identified.
b.
Review of.LERs related to Maintenance Program Due to.a significant number of LERs directly related to maintenance activities which caused unnecessary challenges to safety.and safety-related system / components, the inspectors conducted reviews of LERs related to the maintenance program.
The following is a listing of-LERs reviewed:
1-86-006 Automatic Isolation of Units 1 and 2 Common Control Building Heating, Ventilation, and Air Conditioning System Due to Chlorination System Storage Areas High Chlorine Alarms 1-86-009 Automatic Reactor Scram Resulting from Erroneous High Reactor Level Indication
-
1-86-018 Loss of Reactor Protection System Bus A 1-86-018 Loss of Reactor Protection System Bus A, Supplement 1 1-86-024 Automatic Reactor Scram Resulting From Loss of Main Generator Output Voltage Control 1-86-028 Inoperability of Reactor Building Fire Hose Station 1-86-029 High Pressure Reactor Scram Resulting Fro'm Main Generator Runback i
- _ _ _ _ _ _ _ _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _. _. -. _. _ _... _. _ _._ _ _ _ _ _ _ _ _ _ _
.\\
'
.
i
.
1-86-30 Tripping of Units 1 and 2 Reactor Protection System Division II Motor Generator Set Output Breakers i
1-87-006 Automatic Starting of Emergency Diesel Generator No. 1 Due To Loss of Plant Emergency Bus.E-1 During Reactor
Calibration of Potential Transformer Voltmeter
'
1-87-009 Tripping of Unit 1 Reactor Protection System Division I Motor Generator Set Output Breakers 2-86-004 Primary Containment Group 6 Isolation, Auto Isolation of Reactor Building Ventilation System, and Autostart of Reactor Building Standby Gas Treatment System 2-86-006 Autoclosure of Primary Containment Isolation Valves G16-F004 and F020, Auto Isolation of Reactor Building Ventilation System, and Autostart of Standby Gas Treatment Train 2B 2-86-077 High Radiation Alarm Trip of Reactor Building Ventilation Exhaust Radiation Monitor D12-RM-N010B 2-86-013 Reactor Protection System Actuation During Unit 2 Refueling / Maintenance Outage 2-86-019 Misconfiguration of Transversing Incore Probe (TIP)
System 2-86-020 Reactor Scram While Returning B21-PT-N023B to Service 2-86-024. Core Spray Injection During Short-term Maintenance Outage 2-87-002 Autoclosure of Reactor Water Clean-up (RWCU) System G31-F004 Resulting From Revised Thermocouple leads to RWCU Isolation Instrument G31-TS-N6000 2-87-003 Inoperability of Reactor Core Isolation Cooling System High Reactor Water Level Trip Function Due to Lifted Lead The inspectors' review indicated that the above LERs were in some cases vague concerning the root cause of the events and the information in the LERs did not always present an accurate picture of
'
the events which le.td up to the occurrence. In order to assess the events, the inspectors interviewed technicians, foremen and managers; and reviewed surveillance procedures and other work documents
,
associated with the event.
Examples of the concerns noted are provided in the following:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ - _ _ _ _
_
. _ _ _ _ __ __ - _ __-
_ _ -
.
.(1)
LER 1-86-006 - Automatic Isolation of the Units 1 and 2 Common Control Building Heating, Ventilating, and Air Conditioning Systems Due to Chlorination System Storage Area High Chlorine Alarms The licensee reported in the LER that on February 12, 1986, and February 16, 1986, automatic isolation of the Units.1 and 2 common control building heating, ventilation, and air condition-ing system occurred, per design, due to actuation of the chlorination system storage area chlorine detector, X-AT-2979.
The LER indicated that the cause for the February 12, 1986 event was attributed to a suspected whiff of chlorine gas in the vicinity of the detector resulting from maintenance activities on the chlorinators.
The second ' actuation occurred due to
,
chlorine vapor from a conventional service water system piping
'
(CSWS) flange leak.
The inspector's review of the licensee's investigative package indicated that all the facts were not considered, i.e.,
the
'
project engineer.in charge of the plant modification package for the CSW system proceeded to the service water building to inspect the flanges for leakage when he learned that the 28 CSW pump was started.
The project engineer noted a continuous stream of water (1 GPM) coming from the strainer discharge flange.
He made the decision, without confirming with the control room, te 1ct the leak continue until the next day when the craft personnel would be available to tighten the flange.
This fact would indicate the cause of the second event was improper actions or the project engineer's part in not informing the control room to allow the 2B CWS pump to be stopped and the
,
I area ventilated.
The LER doesn't address whether the engineer had an understanding of the relationship between the leakage and the chlorine alarm detector functions which could have indicated inadequate training of the project engineer. The LER also does
not address whether the pump operation was initiated as a post maintenance test or whether post maintenance testing had been conducted.
Evaluation of these questions would have indicated if the cause of the event was an inadequate maintenance package.
Further interviews by the NRC inspector determined that this test was considered as a post maintenance test and the test pre-briefing was inadequate in that it did not consider actions to be taken if the system leaked.
(2)
LER 1-86-28 - Inoperability of Reactor Building Fire Hose Station From approximately 0300 until 1700 on October 29, 1986, Unit 1 Reactor Building fire hose stations were rendered inoperable due to inadvertent isolation of the stations source of fire suppression.
Unit I was on line at or near 100 percent power. The licensee determined the cause of this event resulted from a procedural weakness involving installation of a mechanical i
_ - _ _ _ _. -_-_
.
22
'
i
. jumper. around hose station supply deluge valve 1-FP-DV-319 to-permit maintenance. The_ procedure which installed the subject mechanical jumper failed to provide controls to ensure its water
- supply remained open. A contributing factor was a personnel error during a subsequent periodic' test (PT) of the Residual
,
Heat Removal Service Water System. An Auxiliary Operator (AO)
failed to recognize the significance of the-fire. suppression deluge supply post indicating valve, PIV-20, not in its normally closed position while performing the PT.
Per the PT, the'A0 closed PIV-20'upon completing the PT, which isolated the Reactor Building Fire Suppression System. The'NRC inspector. determined from a review of the documentation within the licensee's event investigation package that when the A0 went to the control room to get authorization to commence the PT, the shift test engineer, upon his review of the PT failed to note that the PT
'
would-close PIV-20; therefore, he didn't caution the A0 concerning leaving'PIV-20 open.
Further review and interviews indicated that appropriate corrective measure were taken and the revisions to Fire' Protection Procedure, FPP-301, Alternate
. Fire Suppression for Reactor Buildings 1 and 2 Standpipe Deluge System, Revision 1, would provide adequate protection if'
followed, to prevent recurrence of.similar events.
(3) LER 2-86-004 - Primary Containment Group 6 Isolation, Auto IsolationL of Reactor Building Ventilation System, and Autostart of Reactor Building Standby Gas Treatment System Plant modification work was in progress to install environmen-tally qualified temperature switches in the isolation trip logic circuitry of the Reactor Building Ventilation System.
Unit 2 was in a refueling / maintenance outage.. The licensee determined the cause of the inadvertent isolation to be electrical shorting at an electrical solder connection penetrating the back plate of control room back panel PG06. This shorting caused power supply fuse F-5 to blow, deenergizing the trip circuit controlled relay of Reactor Building ventilation radiation monitors (D12-RM-N010A and B) and initiating the incurred isolation and the automatic start of affected systems.
The inspectors' review indicate that the modification package did not specify how the solder connection should be made nor did it provide directions concerning whether insulating sleeving should be used.
Interviews with maintenance technicians revealed that the technicians had a positive confidence level with the periodic test (PT) procedures, maintenance surveillance test procedures (MSTs) and maintenance work packages (MWPs) utilized in carrying out assigned tasks.
The inspectors determined from a review of pts, MSTs, and MWPs that these documents were comprehensive and if followed should, provide quality results.
A review of the above list of LER event investigation packages and procedure reviews coupled with verbal statements from the interviewed technicians, foremen and managers, indicated that
-__-_ ________ _
-
.
y
.
.
the licensee's corrective-actions and root cause determination were adequate. The LERs _ listed above are' closed with the exception of LER 1-86-024 which is. discussed in detail in paragraph 6'of this report and LER - 1-87-009,- which Will remain' open pendirig the licensee's
- submittal of a. supplemental report providing the root cause, corrective action, and consequences of this event.
.The _ inspectors :also conducted numerous interviews with maintenance technicians, foremen and managers to determine if programmatic breakdowns in the maintenance program had contributed to the number of personnel errors reported in LERs in the maintenance area.
The objectives - of the interviews were to determine program adequacy in the following areas:
.,
!
(1) Procedure Adherence The' interviews ~ indicated that_ maintenance personnel had received appropriate training in the requirements and significance of procedure adherence.
Individuals interviewed indicated that L
management _ frequently emphasized the need for procedure adherence, and required _ employees to revise procedures containing errors. Employees involved in events which resulted
,
from personnel errors stated that they were aware - of the
'
requirements. The events reviewed did not appear to be related to problems in this area.
o (2) Training
'
,
The interviews indicated that employees felt they were well trained for the tasks to which they were assigned. In addition, most employees stated that if they did not feel. qualified, they would request additional assistance.
(3)
Independent Verification Requirements l
Individuals interviewed were able to adequately describe the requirements for independent verification of maintenance activities. In addition, technicians stated that they felt that procedures were clearly written to indicate where independent verification was required.
(4) Management Involvement in Maintenance The inspectors determined that management had provided policies to assure that foremen and supervisors were involved in direct i
supervision of maintenance activities in the field. Individuals
!
interviewed indicated that a lead technicien had been designated to assist each foreman in meeting administrative _ requirements, i
The administrative workload appeared to require a significant
'
amount of the foreman's time; however, interviews with techni-cians indicated that supervision in the field was adequate
'
!
I l
. - _ _ _ - -
-_
.
.
-24 l
'and ' technicians requested and received' appropriate nanagement.
E assistance when problems were encountered. The pre-briefing for maintenance activities appeared to be comprehensive.
- (5)- Feedback Mechanisms for Procedure Deficiencies or Work Related-Problems The interviews indicated that technicians understood the feedback mechanisms for correcting procedures and handling work
'!
related problems.
Technicians indicated that it was their responsibility to stop the performance 'of a procedure when an error was. detected and correct : the procedure.
Technicians t
indicated that an' individual from the maintenance section had been assigned to the procedure writing staff. This. individual appeared to provide' an effective interface for the maintenance department, allowing quick resolution and revision of errors.
.
(6)
Labeling / Equipment Identification 1:
L'
The inspectors had noted that several of the LERs indicated wrong unit / wrong train errors and lack of labeling for. certain equipment that contributed to the occurrence. Interviews on the specific events indicated that the particular labeling problems j
were promptly corrected.
Technicians indicated that they had
'
been instructed by management to identify labeling deficiencies and submit work requests to correct the deficiencies.
The interviews indicated - that the licensee ha'd made significant-
.1 progress in identifying and labeling equipment in the last several years and that procedures had been improved to include correct labeling and locations of equipment.
(7) QA/QC Involvement in Maintenance Activities The interviews indicated that QC observation of holdpoints was coordinated with maintenance and was promptly performed.
In Eddition, technicians indicated that QA/QC had conducted audits of maintenance activities in the field. The licensee provided the inspector with a sample of QA/QC audits coriducted in the
'
maintenance area.
The inspector reviewed Surveillance Reports Nos.87-027 (May 5, 1987),87-020 (April 12, 1987),87-003 (January 15,1987), and QASR-87-022 (February 27, 1987). These surveillance audits covered reviews of limitorque motor operator rebuilds, Unit 1 outage activities, inspection of valve stem anti-rotation devices, and ISI valve operability. The audits appeared to be focused on particular problem areas.
The inspector did not have an opportunity during this inspection to determine if audits were being conducted on routine maintenance activities at appropriate time intervcis.
_
- - - _ - _ _ _
_______m_
_ _ - _. _ _ - - _
_ _ _ _ _ _ _ _ - _ _ _
_ _ _
. _ _ _ _ _ _ _ _ _ _ - _
.
.
(8) Appropriate Use of Overtime The interviews indicated that overtime, although heavy during j
outages, was not excessive. Technicians indicated that overtime had decreased significantly over the past several years.
l Technicians'also indicated that work conditions had improved in general, resulting in a significant improvement in morale in the maintenance department.
The reviews in this area indicated that maintenance craft and management believed that the changes in the maintenance area have resulted in improvements in the performance and management of j
maintenance activities.
Craft support of these initiatives was
evident during the interviews.
The inspectors noted that of seven (7) maintenance related LERs which were reviewed and in which the event date and time was known, all but cne occurred on a back shift or on a weekend shift. In addition, the inspectors noted that several of the events involved either problems in the communication between technicians or communications during pre-shift briefings.
During the interviews, the inspectors determined that the maintenance personnel on back shif ts consist of duty personnel assigned from various crews on the day shifts.
The inspectors recommended that the licensee review the composition of these crews and determine if events were related to combining crew members and foremen not accustomed to working with one another. The inspectors did note that adequate pre-shift briefings had been stressed with maintenance personnel.
Within the areas reviewed, no violations or deviations were identified.
c.
Maintenance Training The inspector reviewed the licensee's training program, focusing on training in response to operational events resulting from maintenance activities on the plant.
Procedure MP-49, Volume XII, Rev. 2, Maintenance Subunit Training Program, Section 5.6, Real-Time Training (RTT) offers the following description of RTT: "RTT is conducted to satisfy training requirements that are unique, of a one-time nature, and which require rapid dissemination.
This training fills a need for training in the areas of:
unusual events; operating experience reports; licensee event reports; NRC noted deficiencies; modifications of the power plant; and, special training needs determined by a manager."
This training is conducted quarterly and immediate training sessions may be scheduled at the discretion of the Manager, Maintenance or his designee. All personnel are expected to attend RTT, and those who cannot attend must be formally waivered and should attend make-up training or receive an exemption from the training.
- _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_ _ _ - -
. _ _
_ _.
_
-
. - - - - - _ _ _ ___
y 26.
'LER-2-86-020 ' concerned an automatic reactor scram that occurred on
-
August 23, 1986, as a result of erroneous reactor low level Nos. 2
,
and 3 signals.- The signals occurred while a. reactor ' vessel steam dome pressure instrument was being placed into service. A technician had failed to follow procedure and valved the pressure transmitter intol service prior to equalizing the instrument with reactor pressure.
RTT for this specific event was: conducted on August 27, 1986, and. covered the proper adherence to maintenance instructions and the consequences of reference leg. perturbations. The inspector reviewed attendance' sheets and course notes for the RTT session. All of ~.the I&C ' technicians attended the course _as' well as.the personnel involved' in the event.
The course notes appeared to adequately
. address the causes of the event.
The inspector also reviewed the RTT conducted for th'e first quarter of 1987.
The learning objectives of this training were as follows:
(1) Plant ' modifications planned for turnover before the next RTT E
that may affect equipment installed in the technicians' area of
'
.
. responsibility.
(2) LERs that affect maintenance that were' issued during the past 90 days.
These included thefollowing:
2-86-025, Failure to -Perform Jet Pump Surveillance Testing; 2-86-026, Late Performance of a Fire Watch; 2-87-001, Reactor Scram Due.to Main
Turbine' Control Valve Fast Closure; 2-87-002, Auto Closure of
.RWCU Valve F004; 1-86-029, Reactor Scram-Main Generator Runback; and, 1-86-030, Tripping of Unit 1 and Unit 2 Division II RPS M-G
Output Breakers.
(3) Operating Experience Report 87-06, Overpressure of Recirculation-Sample Line.
(4)
IEN 86-106, Feedwater Pipe Thinning - Surry Plant (5) Maintenance Experience Report 86-006, Inadvertent Trip on IB Circulating Water Intake Pump.
l The purpose of the training was to provide factual information, point
out problems, and learn from mistakes.
!
The inspector concluded that RTT accomplishes its stated purpose and is a unique approach to instruction that is usually accomplished through required reading programs.
Interviews with technicians indicated that the training was effective in providing pertinent information that could be applied to their work.
In addition to the RTT program, the licensee provided a copy of a I
'
memo detailing the review of 35 System Operation Evaluation Reports (S0ERs) for incorporation into the Training and Qualification l
l L
L
}
-_
_-
i
.
l
!
[
program.
The memo stated that the SOERs had been reviewed and dispositioned; either incorporated, already addressed, or not j
applicable, as indicated.
The maintenance training program curriculum was examined for evidence of inadequate ' training as a contributing cause of the LERs.
The inspector reviewed the generic training as well as the more specialized training.
The licensee provided the inspector with the training topics that were presented to technicians for each level of experience.
These l
topics were divided into disciplines (I&C, electrical, and maintenance) which in turn contained five levels that followed a logical progression commensurate with an increase in skill and experience.
The inspector had been concerned after review of the LERs that some
!
of the operational events were due to inadequate training at the most basic level. These would have fallen under such topics as procedure adherence, print reading, and labeling. The inspector reviewed the basic course lesson plans for maintenance procedures and drawings.
The lesson plan for maintenance procedures included the following areas:
safety precautions, cleanliness, procedure performance, Q-list statements, Technical Specifications, QA/QC hold points, and removal / return to service.
The lesson plan for drawings was also comprehensive, and both lesson plans indicated that the events were not due to inadequate basic training on procedures and/or drawings.
The inspector reviewed Training Instruction TI-113, Related Technical Training and On-The-Job Training for Selected Maintenance Classifi-cations.
This procedure detailed the methods by which maintenance technicians obtained, documented, and maintained their qualifications.
Qualification Checkout Cards (QCC) are used as a means of documenting the tasks for which the employee had been trained.
The procedure defines the training requirement for qualification as well as the methods used to evaluate the employee's abilities.
The QCCs are evaluated every two years (+ 25%; maximun of 30 months) and
requalification is based upon infrequently performed tasks, complex
'
technical skills, and degraded job performance.
The inspector interviewed several foremen, and each demonstrated knowledge of the QCCs in accordance with TI-113, and also indicated that the QCCs were used for task assignments when they were uncertain of an employee's abilities.
Interviews with employees indicated that the employees had no reservations about requesting assistance or training for unfamiliar or infrequent jobs.
The inspector observed a training class on the topic of Diesel Generator maintenance.
The instructor was prepared and used visual aids.
Notable among the visual aids was a videotape that clearly demonstrated actual Diesel Generator maintenance.
The videotape had been recorded during prior maintenance.
No violations or deviations were identified.
i
- _ _ - - - _, _ - _ _
_ _ _ _ _ _ - _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ _ _ _
_ _ - _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - -
!
.
.
8.
Inspector Followup Items (92700)
!
'
a.
(Closed) Inspector Followup Item 325, 324/82-26-02.
Response to QA/QC Audit Findings of Training Programs. Review and acceptance of the corrective actions were performed by the licensee's Performance Evaluation Unit to complete the actions required by QAA/21-18-19.
This item is closed.
b.
(Closed) Inspector Followup Item 325, 324/84-39-02.
Controller E11-SS-F605A Output Plug Placed in Wrong Position.
The inspector reviewed the licensee's response to this item and was satisfied that the actions taken to discover the cause of the plug being repositioned were acceptable although no cause was found.
The inspector reviewed training records and schedules to verify that MP-14, Corrective Maintenance, and Volume I, Book 1, Section 11.7, of the plant Administration Instructions, was being taught.
The inspector also verified that MP-14 was being taught on a continuing basis along with procedural compliance lessons during the Real Time Training program.
Based on the above information, the item is-closed.
c.
(Closed) Inspector Followup Item 325, 324/85-24-02.
Discrepancies Not Documented on Attachment 3 as per AI-17 and Failure to Follow AI-58 for Issuance of Clearances.
The inspector reviewed AI-17, Rev.13, Plant Housekeeping, and found that the procedure had been i
revised to require that Attachment 3 be filled out only if the discrepancy would require a long time to correct. A sampling of clearance tag sheets from both units indicated compliance with AI-58, Rev. 19, Equipment Clearance Procedure. The procedure was revised to have all personnel who approve clearances first ensure that the clearance sheet is properly filled out.
During the review of active clearances, the inspector found several clearances more than two years old, one which was almost ten years old.
The clearance audit procedure had not been performed since the revision to AI-58 was implemented.
The licensee committed to reducing the number of clearances which are greater than one year old. The licensee also committed to revising the clearance procedure to ensure that the restoration order is defined on the clearances or require that the block be marked N/A if it has been determined that the order of estoration was not important.
Finally, the licensee is evaluating the possibility of making the clearance procedure an operating procedure in order to assure all operations personnel are informed of revisions in a timely manner.
Based on the information above, this item is closed.
9.
Review of LER Preparation Program (92700)
The inspectors observed numerous cases during the review of LERs where the LERs did not provide all of the information pertinent to the review of the event. The inspectors noted, however, that the quality of recent LER input
- _ _ _ -
._
.
.
-
q
'
had improved significantly over reports submitted in 'mid-1986.
The
inspectors were concerned, however, that even in recent events, the j
licensee did not always fully address the root cause of events and
'
describe comprehensive corrective actions to solve the problems.
The reviews of the LER investigation packages available at the. site indicated that the licensee had in most cases performed an adequate review of the events and identified all root causes, but as indicated above, some of
'
this information was not provided in the LERs.
The inspectors reviewed the methods used to investigate and write LERs.
The inspectors determined that the licensee was aware of the deficiencies in the LERs through discussions with the NRC resident inspectors, and
. The had taken action to improve the investigation and review process.
inspector reviewed procedure AI-84, The Scram Incident Investigation Team (SIIT), Rev. 2, and discussed the responsibilities of the team with the team's Chairman.
The SIIT is activated to investigate scrams and, at the' discretion' of the General Manager, other events. Team members are relieved of other duties until the review is completed.
In addition to investigating. the circumstances of the event, analyzing root causes and determining corrective actions as described in 01-22, ' Plant Incident and Posttrip Investigation, Rev. 14; the team is required to evaluate the incident in regards to human factors considerations. : The results of the human factors review is provided to the Institute of Nuclear Power -
l Operations (INPO) for analysis of industry problems.
l The inspector also reviewed RCI-06.1, Licensee Event Report:, (LERs) and Special Reports (SRs) Identification, Investigation, Preparation, and Submittal, Rev. 5.
The inspector determined that RCI-06.1 required the assignment of an Incident Manager for each event.
The Regulatory Compliance section was responsible for determining deportability, processing the paperwork, and reviewing the report for completeness. The
. vari.ous Incident Managers were assigned the responsibility for ensuring that the input was accurate, obtaining input from other organizations, and ensuring reviews were completed.
The Incident Manager was provided guidance in Attachment 1A in RCI-06.1 describing the information required.
l This guidance did not contain all the information required by 10 CFR l:
50.73. The narrative description of Attachment 1A did not specifically require a discussion of the status of structures, components, or systems i
that were inoperable at the start of the event or the Energy Industry Identification System component identifier.
For personnel errors, the guidance did not address a determination of cognitive versus procedural
,
j.-
errors; whether the error was contrary to an approved procedure or not
cavered by procedure; or, whether the work location was a factor.
The L
guidance did not address the incorporation of automatically or manually initiated safety system responses; the manufacturer and model number of each component; or reference to previous similar events in the LERs.
The licensee stated that the Regulatory Compliance section addressed the information discussed above during their preparation of the LER as required by Section 2.11 of RCI-0.61.
This section states that the f
_
_
__
_ _ _ _ _ _ _ _ _ _ _ _
..
,
Director - Regulatory Compliance or his designee shall ensure that the LER is prepared in accordance with the format of Regulatory Guide 1.16 and NUREG-1022.
No further guidance is provided in the procedure.
The licensee committed to review RCI-06.1 and revise the procedure to provide additional guidance which addresses the areas of concern.
Review of the licensee's actions to revise RCI-06.1 to fully describe the LER preparation process and actions to improve the quality of LERs is identified as Inspector Followup Item 325,324/87-12-03.
)
__
____-_______a