IR 05000445/1989030
| ML20246L533 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 07/07/1989 |
| From: | Livermore H Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20246L530 | List: |
| References | |
| 50-445-89-30, 50-446-89-30, NUDOCS 8907180460 | |
| Download: ML20246L533 (62) | |
Text
.. ___- _
-_ - __ _. _ _
- _ _
-*.
,,
..
l l
U. S. NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REACTOR REGULATION NRC AIT Inspection Report:
50-445/89-30 Permits: CPPR-126 50-446/89-30 CPPR-127 Dockets: 50-445 Category: A2 50-446 Construction Permit Expiration Dates:
Unit 1: August 1, 1991 Unit 2: August 1, 1992 Applicant:
TU Electric Skyway Tower 400 North Olive Street Lock Box 81 Dallas, Texas 75201 Facility Name:
Comanche Peak Steam Electric Station (CPSES),
Units 1 & 2 Inspection At:
Comanche Peak Site, Glen Rose, Texas Inspection Conducted: May 15 through June 16, 1989 Team Leader:
h3 L f' b 3 H. H. Livermore, Lead Senior Inspector '
Date Team Members: S. D. Bitter, Resident Inspector, Operations E. N. Fields, Electrical Engineer, NRR R. M. Latta, Resident Inspector (Electrical), NRR M. Malloy, Project Manager, NRR J. N. Rajan, Mechanical Engineer, NRR W. Richins, NRC Consultant (Parameter)
M. F. Runyan, Resident Inspector (Civil / Structural),
NPR P. Stanish, NRC Consultant (Parameter)
i
,
DC Q
_ - - - _ - _ _ _ _ _ - _ - _ _ _ _ _ _ _ - _ _ _ - _ - _ _ - -
__
-
_
_
_ - _ _. _
_ _ _ _ -
,,
.
Executive Summary On April 23, 1989, a misalignment of the turbine driven auxiliary feedwater pump discharge valves during hot functional testing (HFT)
in combination with multiple failures of Borg-Warner check valves induced a backflow.of high temperature water from the steam generators through auxiliary feedwater (AFW) piping to the condensate storage tank.
The backflow event occurred with the reactor coolagt system at normal operating temperature and pressure (NOT/NOp, 557 F and 2235 psig) and lasted approximately 20 minutes.
The resultant excessive heat caused paint on the AFW piping to discolor, blister, and flake although no visible piping damage was evident.
Available AFW temperature indicators were off-scale during this event.
On May 5, 1989, while still at NOP/NOT, valves in the AFW system were again misaligned allowing an even more pronounced intrusion of j
high temperature water into the AFW system.
During this event, i
backflow occurred intermittently for approximately two hours.
Additional paint was discolored and blistered on the AFW piping and one pipe support was damaged by thermal expansion.
Leak testing and radiographic examination performed subsequent to these events identified that at least 10 Borg-Warner pressure seal swing check valves (3 and 4 inch) in the AFW supply lines and miniflow lines were stuck open.
After approximately six weeks of investigations, the applicant determined the root cause to be improper adjustments of the vertical elevation of the bonnet-disc assembly combined with possible excessive axial play in the disc-arm assembly.
The improper adjustments were primarily the result of inadequate installation instructions in the Borg-Warner O&M manual.
The applicant's corrective action included a valve-specific bonnet elevation adjustment (for pressure seal bonnet check valves) and a verification that the axial play component is within a specified envelope (for both pressure seal and bolted bonnet check valves).
All Borg-Warner check valves located in Unit 1 and Common areas will be physically examined / adjusted and retested for reverse flow prevention capability.
The applicant evaluated the piping and containment penetrations for possible damage.
Several areas in the piping were apparently stressed beyond ASME Code allowables.
No unacceptable conditions were identified for the penetrations.
There were three precursor events.
A similar Borg-Warner check valve failure was identified in 1985 at Comanche Peak but not thoroughly addressed by the applicant.
Subsequently, three Borg-Warner check valves in the turbine driven AFW supply lines to the steam generators were found to be leaking on April 5, 1989, prior to HFT.
proper evaluation and resolution of the leakage found on April 5, 1989, might have prevented the high temperature water intrusions on April 23 and May 5, 1989.
In addition, a Borg-Warner
. _ _ _ _.
_ _ _ _ _ -
. _ - _ _ _ _ _ _ - _ _ - - - _ _ _ _ _ _ _ _ _ _ -
_ ______
'
.,
.
i check valve in an AFW miniflow line was found to be leaking on April 19, 1989, and was repaired prior to the April 23, 1989, event.
'
'
The applicant initially concluded that the failure of this valve was an isolated event.
There exists extensive and well documented industry experience with faulty Borg-Warner check valves.
The AIT determined that a lack of aggressiveness by operations management to thoroughly follow-up on the valve failures identified on-April 5 and April 19, 1989, inadequate communications between operations personnel, and lack of adequate manpower for operating valves during the HFT contributed significantly to the AFW events.
While the problem resolution effort by the applicant was protracted (approximately 6 weeks), the results were thorough and represent a basic commitment to corrective action.
1.0 General Backaround Information Comanche Peak Steam Electric Station (CPSES) Units 1 and 2 are Westinghouse pressurized water reactors with steel-lined, reinforced concrete containments.
The units are under construction approximately 40 miles southwest of Fort Worth, Texas.
An extensive corrective action effort to correct numerous design and quality of construction deficiencies has been underway at CPSES over the past several years.
This program has resulted in a significant number of modifications to bring the plants into conformance with NRC requirements.
For various reasons, in March 1988, the applicant temporarily suspended work on Unit 2 to concentrate resources on Unit 1 completion.
The applicant currently plans to begin loading fuel in Unit 1 on October 2, 1989.
Hot functional testing (HFT) on Unit 1 has recently been completed * and integrated leak rate testing is scheduled July 1, 1989.
The NRC has established a policy to provide for the timely, thorough, and systematic inspection of significant events at nuclear power plants.
This includes the use of an Augmented Inspection Team (AIT) to determine the causes, conditions, and circumstances relevant to an event and to communicate its findings, safety concerns, and recommendations to NRC management.
An AIT was formed on May 15, 1989, to review events which occurred during Unit 1 HFT on April 23 and May 5, 1989.
Although AITs generally evaluate events which have occurred at operating nuclear power plants, NRC management determined that these events warr'nted a team l
inspection conducted in accordance with AIT procedures.
- Unit 1 previously underwent HFT in 1985.
_ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ - _ _ _ _
_ _ _ _ _
_ _ _ _ -
r
's H.,
..
1.1 Description of the Events Toward the end of Unit 1 HFT on April 23, 1989, levels suddenly decreased in Steam Generators (SGs) 1, 2, and 4 while all four SGs were being fed by Motor-Driven Auxiliary Feedwater Pump (MDAFWP) 02.
The Turbine-Driven Auxiliary Feedwater Pump (TDAFWP) _ supply linos to SGs 1, 2, and 4 overheated, as evidenced by paint blistering and cracking on the pipes.
The event was caused, in part, by concurrent opening of the TDAFWP test line isolation valve (1AF-042)
and TDAFWP discharge valve (1AF-041).
When both of these valves were opened simultaneously, a flow path to the Condensate Storage Tank (CST) was created from the SGs via
,
TDAFWP' piping (See Figure 1).
On May 5, 1989, a similar event resulted in the blowdown of steam generators Nos. 1 and 3 to the CST.
On this occasion-the MDAFWP test line isolation valve (1AF-055) and the MDAFWP discharge valve (1AF-054) were operated concurrently, creating a flow path through MDAFW and TDAFW piping to the CST.
The second event was compounded after an attempt to close valve 1AF-055 resulted in this valve being left one-quarter turn open, which resulted in an additional blowdown from steam generators Nos. 1 and 3 to the CST through MDAFW piping.
A diagram showing the feedwater system interface with the auxiliary feedwater system, and the backflow path is provided in Figures 2-and 3.
The primary concerns with this event were (1) the equipment failures which could render the auxiliary feedwater system inoperable and (2) the temperature effects of the backflow on the auxiliary feedwater piping.
On May 15, 1989, the NRC Director Comanche Peak Project Division issued a Confirmation of Action Letter (CAL) to Texas Utilities.
The letter confirmed that specified actions were to be taken by the applicant regarding the event of backleakage through the Borg-Warner check valves in the Auxiliary Feedwater System.
The specified actions were subsequently completed by the applicant and the CAL was
fulfilled as was noted in the AIT exit on June 16, 1989.
i On May 19, 1989, TU Electric notified the NRC of a potential 50.55(e) construction deficiency relative to the AFW check valve backleakago events of April 23, 1989, and again on May 5, 1989.
Additionally, the applicant informed Borg-Warner by letter TSC-89159 on June 1, 1989, that a defect, as defined in 10 CFR, Part 21, may exist within certain check valves supplied by them.
1.2 Augmented Inspection Team (AIT) Tasks The AIT investigating the events was composed of a team leader from the NRC site inspection staff, three NRC resident inspectors assigned to Comanche Peak, the Comanche
__.
._-
- _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _
l
- .
.
.
Peak Project Manager from the Office of Nuclear Reactor Regulation (NRR), two technical specialists from NRR, and two NRC consultants assigned to the'NRC Comanche Peak site inspection staff.
AIT tasks were specified in a May 12, 1989, memorandum from the NRR Associate Director for Special Projects to the team leader.
These tasks included:
a.
Develop and validate a detailed sequence of. events associated with the hot water intrusion into the Auxiliary Feedwater (AFW) System at Comanche Peak on April 23, 1989.
I b.
Evaluate the significance of the equipment failures with regard to safety' system performance, safety significance, and plant proximity to safety limits as, defined in the Technical Specifications.
c.
Evaluate the accuracy, timeliness, and effectiveness with which the information on this event was reported to the NRC.
d.
For each equipment malfunction, to the extent practical, determine:
(1) Root cause.
(2) If the equipment was known to be deficient prior te the event.
(3) If equipment history would indicate that the equipment had either been historically unreliable or if maintenance or modifications had been recently performed.
(4) Any equipment vendor involvement prior to or after the event.
(5) Pre-event status of surveillance, testing, and/or preventive maintenance.
(6) The extent to which the equipment was covered by existing corrective action programs and the implication of the'failurescvith respect to program effectiveness.
e.
Evaluate applicant's program for maintaining equipment operable after installation and initial testing /
inspection as it relates to this event.
This should include surveillance testing and maintenance activities, f..
Evaluate the applicant's response to related experience and information, including NRC bulletins and notices and
._.
_____ _ ____ ______ _ __ -__. ___ -.
.
- - _ _ _ - _ _ - _ - _ _ - - _ _ - - _
-
_- _
-
. _ _ _
_ _ _
.-.
-
. _ -.. - _ _ _ - _ _
me q
!L,
'O-
..
'
]
I
,
industry guidance'provided in the INPO SOER on check l
valves and EPRI. Application Guidelines.
i g..' Evaluate the applicant's thermal stress analysis of the piping affected by the-hot water intrusion.
h.
Evaluate the implications of the identified' equipment failuresLduring this. event on other equipment in other safety systems at_ Comanche Peak.
i.
Identify anyfhuman factors / procedural deficiencies related to the' event'.
j. 'Through operator-and technician ~ interviews, determine-if any of the following played a.significant. role in cach'
failure; plant material condition; the quality.of-maintenance;'or.the responsiveness of engineering to identified problems.
~k.
Evaluate operator. action during the event.
1.
Evaluate management involvement during the Unit'l hot.
functional tests and the subsequent recovery from the event.
m.
Evaluate the effectiveness of applicant's program for investigating. events as it relates to the April 23, 1989 AFW intrusion event.
.n.
' Evaluate the coordination of applicant's operations, engineering, maintenance, and other organizations in identifying and resolving the issues raised as a result of this event."
-
The primary focus of the AIT was on fact finding; any potential enforcement matters will be the subject of subsequent correspondence.
2.0 AIT Inspection During the approximate six week period' utilized by the applicant's1 AFW Task Team to address'the resolution of this issue, the AIT-team closely monitored the. applicant's activities.
This process typically involved the witnessing of valve disassembly, review of work controls and procedures, interviews with members of the applicant's staff, and attending selected meetings.
Efforts to reconstruct the precise timing of events during the incidents of April 23 and May 5, 1989, were difficult because the sequence-of-events computer was not in operation.
The applicant was in the process of realigning
.
.. _ _ _ _ _ _... _ _ _ _ _... _ _ _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.. _ _ _ _ _ _ _. _ _ _ _ _ _ _ _ _
. _ _ _ _ _ _ _ -
- ,
'
,,
..
1 the sequence-of-events computer to the emergency response
,
system computer.
The applicant utilized operator logs, strip chart recorders, and operator. interviews to reconstruct the chronology of the individual events.
2.1 April 23, 1989, Event Description (PIR-89-110)
f 2.1.1 Conditions Preceding Event on April 23, 1989, the applicant was nearing completion of an extensive hot functional testing program.
The plant was in operational Mode 3 (hot standby) with the reactor coogant system at normal operating temperature and pressure (557 F
)
and 2235 psig).
The No. 2 motor-driven auxiliary feedwater (MDAFW) pump was running and feeding all four steam generators.
Steam generator levels ranged from 56% to 59% with a feea rate of approximately 30 gpm per steam generator.- The total steam generator blowdown rate was 45 gpm.
The main feedwater isolation and main feedwater isolation bypass valves were closed and the preheater bypass isolation valves were open in each loop.
A blackout start test of the turbine-driven auxiliary feedwater (TDAFW) pump had been completed at 0532 hours0.00616 days <br />0.148 hours <br />8.796296e-4 weeks <br />2.02426e-4 months <br />.
The TDAFW pump was to be realigned to the condensate storage tank and run for three hours in preparation for a hot alignment check.
2.1.2 Event Chronology At approximately 0610 hours0.00706 days <br />0.169 hours <br />0.00101 weeks <br />2.32105e-4 months <br />, realignment of the TDAFW pump for recirculation flow to the condensate storage tank commenced.
Standard Operating Procedure SOP-304A, Section 5.5.3, specifies closing.TDAFW discharge valve 1AF-041 and then opening TDAFW test isolation valve 1AF-042 to perform this alignment.
Contrary to this procedure, the two valves were operated concurrently.
The auxiliary operator first cracked open 1AF-042 and then started to close 1AF-041.
Three additional auxiliary operators were dispatched to provide assistance.
Since valve 1AF-042 takes considerably less effort and time to open than is required to close 1AF-041, valve 1AF-042 was fully open before 1AF-041 was closed.
At approximately 0620 hours0.00718 days <br />0.172 hours <br />0.00103 weeks <br />2.3591e-4 months <br />, the Reactor Operator noticed that levels in steam generators Nos. 1, 2, and 4 were decreasing rapidly.
Temperature indicators 1-TI-2471 and 1-TI-2474 on gFW loops 1 and 4 were high off-scale j
(greater than 200 F) and 1-TI-2177B and 1-TI-2180B on feedwater (FW) loops 1 and 4 indicated approximately 500 F.
The corresponding temperature indications on loops 2 and 3 remained unchanged at 105 F to 130 F.
In an attempt to recover steam generator levels, the No. 2 MDAFW pump
,
discharge flow was increased to 400 gpm.
However, flow to l
i
.
_ _ _ _ -
---_.--_-__--.._-__-___-..>-__------J
e
-
.,
'.
steam generators 1, 2, and 4 indicated 0 gpm and steam
'
generator levels continued to drop rapidly, approaching a
'
level of 45%.
Some flow was noted to steam generator No. 3 which indicated a slowly increasing level.
The applicant stated that steam generator blowdown was secured on all steam generators at approximately 0625 hours0.00723 days <br />0.174 hours <br />0.00103 weeks <br />2.378125e-4 months <br />.
The AIT could not confirm this assertion as there is no indication of reduced outflow from the steam generators on the strip chart level recorders or any mention of this event in the operator's logs.
At approximately 0630 hours0.00729 days <br />0.175 hours <br />0.00104 weeks <br />2.39715e-4 months <br />, the TDAFW pump room became steamy with a noticeable smell of paint fumes.
The paint on some pipes in this room was observed to be " bubbling and peeling."
Upon hearing this report, the control room ordered the auxiliary operator to shut valve 1AF-042.
At 0635 hours0.00735 days <br />0.176 hours <br />0.00105 weeks <br />2.416175e-4 months <br />, 1AF-042 was shut, and levels in steam generators Nos. 1, 2, and 4 began to recover from a low level in each of approximate 11'44%.
The flow rate was increased to 50 gpm to each steam generator.
Approximately two minutes later, loops 1 and 4 AFW temperature indications returned on scale.
A review of the event indicated that approximately 6000 gallons had drained from steam generators Nos. 1, 2, and 4 to the condensata storage tank (CST) through the TDAFW piping.
Some increase in CST level was noted following the event.
The applicant conjectured that an inadvertently closed motor-operated valve (1-HV-2493B) prevented blowdown of steam generator No. 3.
The backleakage of water from the steam generators to the CST through the TDAFW piping shculd have been prevented by the TDAFW supply line check valves.
Based on the event scenario and subsequent testing, it is evident that these check valves were stuck open during the event.
The other portions of the backflow path, from the steam (:2nerators to the TDAFW piping, could have taken one of four paths, as follows:
a.
Through the two prehester bypass line check valves in the backflow direction, b.
Through the closed split-flow bypass valve and the outboard preheater bypass line check valve.
c.
Through the closed feedwater isolation valve into the preheater bypass line.
d.
Through the closed feedwater isolation bypass valve into the preheater bypass line.
-
_ _ _ - _ _ _ _ _
_ _ _ _ _ _.
m.
, - - -
- - - - _
-- _ _ _
'
4v c'n
-
j qm.
j
.9 j
-
.
Subsequent testing as described later-in this report
L confirmed that steam generator water leaked back through the
]
closed feedwater isolation bypass valve (d above, See
{
Figure 1).
- 2.2 May 5, 1989, Event Description (PIR-89-129)'
)
12'.2.1 Conditions' Preceding Event
on'May 5, 1989, the applicant was performing'the final portions'of hot functional testing and was conducting a series of tests to_ determine which valves were responsible i
for-the~AFW backleakage event of April 23, 1989,
.
'
(PIR-89-110)._ The' plant was in operational Mode 3 1(hot-standby) withLthe reactor coogant system at normal operating temperature and pressure (557 F and 2235 psig).
All AFW pumps were secured and the MDAFW cross-connect valves
_
j 1AF-090-and 1AF-091'were open.
All AFW test discharge
'
valves were closed.
The main feedwater isolation and main feedwater isolation bypass valves were closed and the-preheaterLbypass isolation valves were open in each loop.
2.2.2 Event Chronology At 0055, preparations were initiated to perform a routine operational surveillance test, OPT-206A, " Auxiliary Feedwater System Operability Test."
The purpose of performing this test was to provide precriticality training for operations personnel and to operationally check the surveillance procedure.
The test was scheduled during HFT to take advantage of the then existing (hot) plant conditions.
The No. 2 MDAFW pump discharge valve 1AF-054
.was in-the process of being closed at the'same time that No. 2 MDAFW pump test valve 1AF-055 was being opened.
This is contrary to procedure OPT-206A and SOP-304A (SOP-304A is referenced by OPT-206A) in that these procedures require 1AF-054-to be-closed prior to opening 1AF-055.
This mispositioning of valves was essentially identical to the April 23 event (paragraph 2.1.2).
During the time'both valves were open, a backleakage path similar to the April 23 event had been established from the steam generators through the leaking feedwater isolation bypass valves, through the preheater bypass line to the AFW inlet, into the AFW piping (See Figure 2).
An analysis of steam generator level strip chart recorders revealed that backleakage occurred only from steam generators Nos. 1 and 3.
Because steam generator No.
3 is located on the opposite end of containment from the feedwater penetration area, apparently no water from this steam generator entered the penetrat;.an area during this
event.
The flowpath from steam generator No. 1 was l
determined to be through TDAFW supply line check
,
valve 1AF-078, into the TDAFW supply header, through TDAFW
-
- - - - - - - - - - -
- - -
_-
- - - - _
_
___
+..
,,.
.i
<
I
l l.
<
supply line check valve 1AF-106 (in the normal forward flow
' direction), through MDAFW supply line check valve 1AF-101, and through 1AF-054 and 1AF-055 to the CST (See Figure 3).
'
The backleakage was stopped when valve 1AF 054 was fully I
closed, after which cross-connect valves 1AF-090 and 1AF-091
)
were closed.
At 0132, No. 2 MDAFW pump was started.
After some data had been mollected, this pump was secured at 0145.
Steam
.)
genera or levels had dropped due to steam-off and
backleakage and the operator decided to realign the system I
to increase levels.
At 0208, cross-connect valves 1AF-090 I
and 1AF-091 were opened, valve 1AF-055 was closed (but inadvertently; left one-quarter turn open), and valve 1AF-054 was opened.
This configuration reinitiated the backleakage predominantly through MDAFW supply line check valve 1AF-075,
.
MDAFW cross-connect valves 1AF-090 and 1AF-091, and valves 1AF-054 and 1AF-055 to the CST.
At 0230, No. 2 MDAFW
pump was started, momentarily stopping backleakage from the
,
Although pump total flow indicated l
300 gpm, the total flow to the steam generators was 80 gpm, indicating that 220 gpm from the No. 2 MDAFW pump was being diverted to the CST via valves 1AF-054 and 1AF-055.
The operators did not know where the missing 220 gpm was going.
They secured the No. 2 MDAFW pump at 0249.
With all pumps secured, backleakage from the steam generator was again hydraulically permitted until, at 0251, No. 1 MDAFW pump was started.
The same abnormal flow indications occurred, indicating that not all pump flow was reaching the steam generators.
The No. 1 MDAFW pump was secured at 0305.
This reinitiated the backleakage; however, within the next several minutes cross-connect valves 1AF-090 and 1AF-091 were closed, restricting the backleakage to steam generator No. 3.
At 0323, No. 1 MDAFW pump was started in order to feed steam generators Nos. 1 and 2 and at 0326, No. 2 MDAFW pump was started in order to feed No. 3 and No. 4 steam generators.
Normal flow conditions existed for No. 1 and No. 2 steam generators.
However, a large flow mismatch was observed between No. 2 MDAFW pump flow and the flow to steam generators Nos. 3 and 4.
Based on these indications, the operators at this time suspected that valve 1AF-055 was not fully closed.
At 0340, valve 1AF-055 was found one-quarter turn open and when fully closed, ended the event.
During the approximately two hours of backflow, an estimated 3000 gallons blew down from steam generator No. 1 and a like amount from steam generator No. 3.
Steam generators Nos. 2 and 4 were isolated.
Based on the volume of piping from steam generator No. 3 to the feedwater penetration room, no water from steam generator No. 3 reached the AFW lines.
The AIT notes that steam generhtor No. 1 is located in containment near the feedwater penetration room.
Steam
__ _ _ _ _ _. _ _. _ _ _ _ _. _. _ _ _. _ _ _ _ _ _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
-_
_ - - _ -
__
s..
.
q generator No. 3, on the other hand, is located on the opposite end of containment.
Given the main feedwater piping volumne of Loop 3 (3261 gallons), there was insufficient backleakage from steam generator No. 3 to reach-the main feedwater penetration.
2.3 Precursor Events 2.3.1 Historical Failure of Valves 1MS-142 and 1MS-143 A precursor to the April 23 and May 5, 1989, incidents occurred in 1983 when the auxiliary feedwater turb3ne driven pump steam supply line check valves (1MS-142 and 1MS-143)
failed inspection following tF; first HFT.
Test Deficiency Report (TDR) 1743, initiated in July 1983, described the disks to be eroded, bent, and unable to perform the designed function.
The valves (along with similar Unit 2 valves 2MS-142 and 2MS-143) were returned to Borg-Warner where, on each valve, the stud was shortened and a stop extending below the bonnet was added.
In addition, the face of the stop which contacts the stud was machined to a o
angle to be perpendicular to the stud axis.
This modification was performed per Design Change Authorization (DCA) 18917, and was apparently necessary due to the sudden high pressure differential applied to the valves when steam is released into the line.
The Unit i valves were again inspected on January 17, 1985, after five cold starts of the turbine driven auxiliary feedwater pump (TDAFWP).
Valve 1MS-142 was found to have a damaged seat, cracked disk, and a cracked disk stud bushing.
Problem Report (PR)85-132 stated that the valve had apparently been assembled with the disk not properly aligned with the seat and contacting the bottom of the valve body.
Failure Analysis Report (FA)85-001 was generated by maintenance engineering to address damaged valve 1MS-142.
Revision 0 of FA 85-001 describes the cause of the failure:
"The bonnet and retainer were incorrectly placed too low in the body, thus, preventing the disk from hitting the seat squarely.
Construction procedures were followed.
However, construction and operations procedures and the manufacturer's technical manual omit steps on setting the depth of the bonnet during reassembly."
The action to prevent recurrence stated in FA 85-001, i
Revision 0, wast
"All valves of the same type will be disassembled, inspected for damage, and properly reassembled.
The procedures will be revised to include the correct method for reassembly."
- - _ - _ - _ - _ - _ _ - - _ _ - _ _ _
-_____
__. _
_~_
__
. _ - -.
.
,,.
..
FA 85-001, Revision 1, was later issued to revise the cause-of the-failure of valve 1MS-142 and the required action to prevent recurrence..The revised root cause of the failure was harsh flow conditions during the cold starts of the TDAFWP.
The valve disk and stud were replaced and the valve seat was reconditioned.
The revised actions to prevent recurrence were:
(a) to replace or modify the valve or (b) to modify the system to prevent harsh flow conditions.
Maintenance Engineering contacted Borg-Warner after issuing FA 85-001, Revision 0, and changed the cause of the failure after Borg-Warner confirmed that the failure was not due to incorrect installation and that the earlier modifications (DCA 18917) were apparently unsuccessful.
The two revisions of FA 85-001'were addressed in the engineering review section of PR 85-132.
PR 85-132 states that test engineers involved in the cold starts of the TDAFWP did not observe any indications of water hammer and noted that valve 1MS-142 had indentations which indicated that the disk did not line up with the seat.
PR 85-132 concluded that:
Mince the disk is not available for re-evaluation, the.
possibility that the failure resulted_from incorrect installation cannot be totally dismissed.
Nevertheless, since one or both of the valves have failed after each heatup, a design review of the valves and the system operating conditions is needed."
I' investigation by the AIT revealed that the design review had been requested in TU Electric office memorandum TCF-85227 dated May 20, 1985.
The AIT has requested has additional information from the applicant regarding documentation of the 1985 discussions with Borg-Warner which led to the decision that the valves were correctly reinstalled.
At the conicusion of this inspection, no documentation had been provided.
The AIT also asked the applicant for information regarding the design review requested by memorandum Ter-85227.
Design modification DM-85-273, " Turbine Dri"tr. Auxiliary Feedwater Steam Supply Line Modifications," dated January 29, 1986, i
describes hardware modifications and operational changes to l
the TDAFW steam supply lines to minimize the effects of water hammer.
Apparently, no design review of the adequacy of the check valves was performed even though the design review was specifically requested by memorandum TCF-85227.
After review of the documentation provided to date and discussions with the applicant, the AIT concluded that:
(1) incorrect valve reassembly was initially identified as
.______ _ _ -_ _ _____ _
~...
.
,
.
the cause of check valve failure in 1985, (2) discussions with Borg-Warner convinced the applicant that the valve failure-was due to other factors, and (3) no design review of the adequacy of the check valves was performed.
Thus, in 1985,.the applicant had identified the root cause of.the check valve problem'and had formulated corrective action plans.which would have fully corrected the problem.
The applicant apparently. permitted the vendor to dissuade them from the correct course of action.
2.3.2 Check Valve Failures of April 5, 1989 A'second precursor event occurred prior to heat up for Hot Functional Testing (HFT) activities on'or about April 5, 1989,.18 days before the first AFW backleakage event. This second precursor event ~ identified that three TDAFW supply line check valves were failing to seat properly.
The
. discovery of this condition occurred during the process of.
draining and filling steam generators to, resolve secondary
~ chemistry problems.
During a filling operation, water was-observed flowing into the TDAFW pump In addition, water was discovered on the floor in.the Ti.'FW pump room.
The source of the water was determined to be backleakage.through~
Procedure ODA-408, log No. 1-89-035 was written primarily to forward flush the TDAFW supply
' lines to the steam generators with reactor makeup water.
Additional steps were added to this procedure to determine if the check valves in the remaining three TDAFW supply lines were leaking.
This leak test revealed that two other TDAFW supply line check valves, lAF-078 and 1AF-086, were not seating properly.
Work requests were written to repair the valves and were assigned a normal priority.. The work requests, however, did not quantify the amount of. valve leakage.
Work orders were initiated with a due date of May 26, 1989, after completion of the HFT.
The AIT interviewed the operations manager concerning the-decision made to continue the HFT with three failed AFW check valves.
The operations manager stated that he reviewed in detail only the. original of procedure ODA-408 log No. 1-89-035 and missed the' fact'that the issued procedure included check valve. leak testing.
The three work requests did not.specify the quantity of water leakage, which was substantial, and were not thoroughly reviewed by the operations manager, the systems engineer, or the shift operators for AFW operability.
The operations manager also stated that the main thrust of the HFT at.this time Was to chemically clean the system and that in hindsight, a Plant Identification Report (PIR) should have been issued to give immediate attention to the leaking check valves.
._
.__
-.
.
..
l'.
i
,
'
s Clearly, poor communication among operations personnel and a lack of-operability awareness was evident.
Because the check valve failures were not documented on a higher-profile document, such as a PIR or NCR, and inasmuch as operations supervision failed to follow-up on the fact that the check valves were not seating properly, management-level attention was not focused on this multiple failure of check valves.
This event provided the applicant an opportunity to discover the full extent of the problem and to avoid the backflow events of April 23 and May 5, 1989.
The applicant did not discuss the failed check valves discovered on April 5, 1989, with the AIT until the week of June 1, 1989.
The applicant stated that this event will be used as a learning experience to effect a change in the mindset of plant personnel from a construction to an operations perspective.
The operators in this case considered the check valve failures to be strictly a hardware issue and did not consider the effect of these failures on the operability of the auxiliary feedwater system.
2.3.3 Failure of Valve 1AF-069 A third precursor event occurred on April 19, 1989, when in the course of AFW pump testing and hot functional testing, the suction relief lifted on the "A" MDAFW pump.
Subsequent i
investigation revealed that the miniflow check valve, 1AF-069, was experiencing gross backleakage.
The valve was disassembled and inspected.
The valve disk was found to have rubbed the inside of the valve body on both sides in the open position.
A small flaw was found on the swing arm in the area of the pivot pin (1/8" wide, 1/8" deep).
The damage appeared to be caused by excessive jarring occurring when the valve disk slammed against the stop upon opening and by turbulent flow conditions resulting from the upstream breakdown flow orifice.
NCRs 89-4484 and 89-4632 were issued and the valve was reworked under Work Order C890005265.
The indicated flaw was dispositioned
"use-as-is," whereas the rubbing of the disk was dispositioned " repair."
Additional weld material was added to the end of the valve stop to prevent the valve disk from coming into contact with the back of the valve body (and possibly becoming lodged in the open position).
The gap between the swing arm and the disk was reduced to limit the amount of axial play in the disk as an added measure to ensure the disk would not contact the valve body.
It is believed that valve 1AF-069, prior to being reworked, exhibited a stuck-open configuration (later found in the 4-inch AFW valves) with the top of disk under the lip of the
_. _
_ _ _ _ _ - _ _ _
_ - _ _ _ _ _ _. _ _ -
_
.
,,.
..
seat.
Subsequent backflow tests revealed that the rework effort was effective in stopping the backleakage.
The reduction in the axial play of the disk raised the top of the disk enough to allow the disk to seat properly.
At the time of valve rework, the applicant believed the problem to be isolated to one valve which had excessive-axial play.
An investigation into root cause and generic implications may have presented the opportunity to discover the full extent of the check valve problems.
The proximity of the 3-inch miniflow check valves to the upstream orifice may have contributed t.o the failure of valve 1AF-069 by causing an increase in the axial play of the disk.
In addition, the increased flow turbulence and valve tapping damage resulting from this configuration would greatly reduce the life span of this valve.
The AIT recommends an design change, as soon as possible, to separate the 3-inch miniflow check valves from their associated orifices.
2.4 Equipment Performance and Analysis 2.4.1 Check Valves 2.4.1.1 Component Description The following component descriptions are applicable to the events of April 23, 1989, and May 5, 1989, which involved multiple failures of check valves in the AFW system.
All of the valves that failed were Borg-Warner 900 lb., pressure seal swing check valves.
There are a total of 28 of these valves in each unit.
The failed valves included, for Unit 1, two of the three 3-inch check valves, located on the AFW miniflow recirculation line, which were determined to be partially stuck open and all eight of the 4-inch check valves, located in the AFW discharge lines to the steam generators, which were also identified as being partially stuck open (i.e., the valve disk lodged under the seat ring).
See Figures 4 and 5 for valve details.
In addition to the pressure seal check valves, the applicant
!
utilizes 103 Borg-Warner bolted bonnet swing check valves in selected low pressure applications (i.e. 150 and 300 lb.
systems).
The bolted bonnet valves have, by design, a fixed vertical relationship between the bonnet / disc assembly and the seat ring such that subsequent to assembly at the i
manufacturer's facility the bonnet and disk assembly cannot normally be adjusted.
Therefore, the bolted bonnet valves are not considered to be susceptible to the same failure l
mechanism experienced in the pressure seal valves.
I c
. _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ _ __
-.
__..
-
p..
.
,
...
U=
L
l
Excessive axial play could, however, potentially result in degraded or inoperable check valves.
A design feature which is common to both the pressure seal valve and the bolted bonnet valve is the tolerance stack-up in the disk arm bushing assembly referred to as the " axial tolerance or axial play."
The axial tolerance was not historically regarded as a critical parameter by Borg-Warner.
However, in order to assure that axial play would not affect the operability of the valve, Borg-Warner has committed to establish a maximum / minimum axial play acceptance criteria.
As part of the assessment of the AFW check valve inoperability issue, the following synopsis of check valve applications was provided by the applicant.
A total of 160 Borg-Warner check valves were installed in Unit 1, Unit 2, and areas common to both units.
Out of this total, 114 check valves are located in. safety-related systems, including 16 4-inch.AFW supply line check valves (8 in each unit and all 8 in Unit 1 were determined to leak), 6 3-inch i
AFW pump miniflow recirculation check valves (3 for each
!
'
'
unit, 2 of 3 in Unit 1 were determined to leak), 2 8-inch TDAFWP discharge check valves (1 per unit, tested satisfactory in Unit 1), 4 6-inch MDAP9P discharge check valves (2 for each unit, both tested satisfactory la Unit 1), 2 8-inch TDAFWP suction check valves (1 par unit),
2 6-inch MDAFWP suction check valves (Unit 1 only), and 24 6-inch check valves located in the preheater bypass line
!
to the upper feedwater penetration (12 per unit).
Thus, out of the 114 Borg-Warner check valves located in
'
safety-related systems, 56 are located in the area of interest defined by the backleakege event.
2.4.1.2 Equipment History In order to evaluate the applicant's program for maintaining and ensuring the Borg-Warner check valves operable following installation and initial testing, the AIT reviewed the maintenance records for the pressure seal check valves.
This review included the examination of construction
operation travelers, nonconformance reports, startup work
'
authorization forms, maintenance action requests, work orders, and NIS-2 forms.
This review revealed that the AFW check valves had been i
installed in the 1979-1980 time frame and that all of the
!
check valves were disassembled and inspected in 1983 for the presence of full fillet welds on the disk to the disk stud and on the disk stud to the stud retaining nut.
A change from the original specification of tack welds to full fillet I
--.m
.
- _
_
_ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ -
'
..
.
welds was recommended by the vendor as a result of a valve failure.
In January 1983, while disassembling the containment spray heat exchanger, the disengaged parts of an upstream Borg-Warner check valve were discovered.
Valve failure was determined to be due to a broken tack weld which had previously secured the disk to the stud.
Tack welds were also used to secure the stud to the disk nut.
Other defective tack welds were found in similar valves.
Consultations with the vendor revealed that the problematic tack welds had been replaced with fillet welds as the standard valve design.
The applicant decided to disassemble and inspect all Borg-Warner check valves, even those which had been procured after the vendor's design change.
Any tack welds found were replaced with full fillet welds by site welders.
Approximately 50 percent of the check valves required the installation of full fillet welds.
A vendor representative was present during this modification process and extensive QA and QC oversight was provided.
However, no post-modification retests of the check valves were conducted.
Since all the valves were disassembled and reassembled, the final status was left uncertain in light of the inadequate installation instructions provided in the vendor's O&M Manual.
The vendors O&M manual was inadequate in that it did not provide any instructions for backing off the retainer ring for valve flapper and seat alignment.
For some pressure seal bonnet check valves, this resulted in the full insertion of the retainer ring which had previously been backed off to adjust bonnet elevation. 'This rendered the valve inoperable because the' disk was positioned too low with respect to the seat ring.
The AIT investigation also revealed that the Comanche Peak Review Team (CPRT) in Issue-Specific Action Plan (ISAP) VII.b.2 identified the population of all valves that had been disassembled and reassembled under the construction
QA program.
Included in the population were Borg-Warner i
supplied check valves that were disassembled in 1983.
Borg-Warner valves (1AF-0075, 1AF-0098, and 1FW-0202)
associated with the Unit 1 Auxiliary Feodwater System were included in the CPRT sample.
I CPRT compiled an inspection package for each sampled valve.
J Each package was reviewed for any indications of incorrect
'
valve reassembly including variances in internal component serial numbers.
No such cases were found.
Each accessible valve was then physically inspected to verify that the correct body and bonnet were installed.
No deviations were
,
identified by CPRT for any Borg-Warner valves selected in l
the sample.
No Borg-Warner valves were disassembled by j
CPRT.
I
- _ - _ _ _ _ _ _
_ -.
_
--_____ _
..
..
i l
In 1985, the system underwent initial hot functional and preoperational testing.
These programs did not detect any abnormal check valve backleakage or operational deficiency relative to the valve disk hanging up under the seat ring.
In arriving at this conclusion, it is recognized that the procedures used for preoperational testing did not test these valves in the backflow direction.
It was determined by the AIT (based on interviews with operations personnel) that a thorough flushing of sections of the AFW system could not be accomplished utilizing the existing system drain valves.
Therefore, over the years, the applicant often removed selected check valve internals to allow for increased. flushing flow rates.
The AIT requested clarification on this policy from members of the applicant's AFW Check Valve Task Team.
This practice of removing check valve internals was also used numerous times historically as a means of draining the system in order to effect welding repairs.
The applicant informed the AIT that it was a routine policy at the site to remove check valve internals to enhance system flushing or draining.
The AIT's concern is that the numerous failures of the AFW system's check valves to seat properly may be related to the applicant's " routine" practice of removing check valve internals for the purpose of flushing and draining.
The valves were not designed for routine. disassembly.
The lack of sufficient documentation following the completion of their maintenance activities appears to be historical.
The AIT also determined based on reviews of maintenance histories and discussions with both startup and system engineering personnel that no provisions were made for surveillance testing or maintenance preservation during the period from completion of preoperational testing in 1985 until the recently completed hot functional tests.
2.4.1.3 Check valve Investigative Action AFW Check Valve Testing Subsequent to April 23, 1989 The AIT witnessed the implementation of backleakage tests I
conducted on the AFW check valves subsequent to the April 23, 1989 event.
The purpose of these tests was to detcrmine if the check valves allow backflow past the seats.
The valves tested included:
(1) the eight AFW supply line check valves, (2) the three AFW pump discharge check valves, and (3) the two motor driven AFW pump miniflow check valves.
The turbine driven AFW pump miniflow check valve could not be isolated and tested due to the design of the system.
,
_ _ _ - _ _ _ _..
__-_________._____m_
_ _ _ - - - _ _ _. _ _ _ _,
_ _, _ _ _ _ _ _ _
_
-
_
-
'
'
..
.
The tests required unique valve alignments for each check valve.
The alignments isolated each valve and provided backflow pressures ranging from approximately 22 psig to 95 psig depending on the test procedure.
A drain valve was opened to insure that the presence of flow could be detected should a check valve leak.
Minimum hold times, generally 15' minutes, were specified.
An initial test'of'the AFW supply'line check valves (8) was performed on May 2 and May 4, 1989, using steam generator t
pressure to create a backpressure of approximately 1150 psig.
Additional tests were performed after HFT to provide assurance that similar tests, conducted after the valves were repaired and reassembled, would provide adequate assurance that the check valves were functioning properly.
All of the AFW supply line check valves, and all of the motor driven AFW pump miniflow check valves failed'the tests and showed leakage.
The three AFW pump discharge check valves did not leak.
As a result of these check valve failures, a total of 23 check valves were radiographer (RT'd).
The results of these RTd indicated that ten check valves were partially stuck open.
Of these ten valves, eight were 4-inch valves and two were 3-inch valves.
Additionally, the RTs for valves 1MS-142 and 1MS-143 indicated that the valve discs were contacting the seat ring at the top but that they were laying slightly off the seat ring at the bottom of the valve.
Following the identification of the inoperable check valves in the AFW system, the AIT inspectors witnessed the disassembly and inspection of selected Borg-Warner pressure seal swing check valves.
During this process 14 check valves were disassembled.
Valve disassembles were conducted initially using Mechanical Maintenance Manual MMI-801, Revision 0, titled "Borg-Warner Check Valve Inspection."
This procedure was later superseded by Maintenance Section - Mechanical Manual MSM-CO-8801, Revision 0, titled "Borg-Warner Check Valve Maintenance."
These procedures appeared technically adequate for valve disassembly and the observed work activities were well controlled.
During the disassembly process, various methods were utilized to capture information including the use of video recording equipment as well as boroscopic and radiographic processes.
Physical disassembly of the check valves was typically conducted in a well controlled and disciplined manner by the mechanical maintenance personnel.
The AIT also determined that QC involvement appeared to be adequate and that QC hold
I
b...
,
_ _-
i'
.,
..
,
points were correctly accomplished.
The following is a synopsis of general observations by the witnessing AIT inspectors relative to the 3-inch and 4-inch pressure-seal swing-check valves manufactured by Borg-Warner.
Some of the 4-inch check valve bonnets did not appear to
.
be'$nstalled with the disk assembly parallel to the seat ring..
The bonnet spacers on several of the check valves were
.
deformed inward indicating overtorquing of the bonnet stud fasteners.
Correspondingly, for the 4-inch valves that exhibited
.
concave bonnet spacers, the studs were also deformed (bent) inward which also indicates overtorquing of the fasteners.
Upon disassembly very little internal wear was observed
.
on the disk seating surface and the seat ring was generally in a serviceable condition.
For the 4-inch check valves identified as being stuck
.
open there was some minor indication of disk contact on the seat ring in approximately the 12 o' clock position.
For the one 6-inch check valve which was disassembled
.
(1FW-198) the retainer ring was determined to be backed off approximately 0.150 inches.
The bonnet assemblies were typically installed with an
.
.
I approximate.015 to.030 inch dimensional differential between the top of the bonnet retainer to the top of the bonnet (indicating that the bonnet fasteners were not tightened uniformly and sequentially).
Avarietyofvalvesgatanggeswereencounteredranging
.
from approximately 3 to 12 from the vertical.
Axial play, although not dimensioned on the assembly
.
drawing, was determined to range from 0.124 to 0.315 inches.
Approximately half of the discs exhibited weld bead
.
overlay remnants on the O.D. of the valve disk from the hardfacing process.
Generally the hinge pins showed only minimum play.
.
The disk stud on the 3-inch check valves associated with
.
,
the miniflow lines indicated signs of deformation where it impacted the bonnet stop.
f i
. _ _ _ _ _ _
__
. _ -.
_
. _ _ _
_.
__
_ _ _ _
_ _ _ _ - - _ _
_ _ - _ _ _ -
_
__
-
...
!
l
.
On some of the disk assemblies the disk washer was
.
loose.
Subsequent to the disassembly of the 14 AFW system check valves, the applicant performed detailed dimensional measurements of the valve bonnets and bodies to ensure their i
conformance to the manufacturers drawings.
This review concluded that there were no dimensions outside of the manufacturing drawing tolerances with the exception of the wide variance of axial play dimensions.
Axial play is not a specified dimension on the Borg-Warner assembly drawing.
On May 30, 1989, the applicant sent the internals from 13 check valves (consisting of 3 each 3-inch and 10 each 4-inch valve bonnet / disc assemblies) to Borg-Warner's Nuclear Valve Division in Los Angeles, California.
The same l
valve internals were returned on June 14, 1989, after the l
Vendor had performed dimensional checks and computer aided drawing (CAD) modeled verification of the as-built configuration.
It is noted that while the subject valve internals were at the manufacturer's facility, no disassembly or destructive examination was performed.
The assemblies were returned essentially in the as found condition.
i The AIT determined that the programmatic controls and administrative procedures utilized for the identification, storage, packaging, and shipping of the subject valve internals to Borg-Warner for analysis were adequate.
See Figure 6 for summary of valve findings.
2.4.1.4 Root Cause In order to assess the root cause determination, the AIT reviewed the BW/IP letter to TU Electric dated June 7, 1989, concerning Borg-Warner high pressure, swing check valves.
Specifically, this BW/IP letter identified the cause of the identified failure of the 3 and 4-inch check valves to be inconsistencies between the supplier's valve assembly technique and the procedural guidance contained in Borg-Warner supplied Operation and Maintenance Manual.
The AIT reviewed the applicant's maintenance procedures applicable to the 3 and 4-inch check valves, MMI-801, Revision 0.
The prescribed reassembly technique was to install and bottom out the retainer which ultimately located the disk assembly low enough in the body to allow the disk l
to catch under the seat ring as shown in Figure 5.
Other
'
factors which were identified as contributors included axial play in the valve disc-arm assembly and the residual fillet weld at the juncture of the disk to disk stud.
-
_ - _ _ - _ ---___ -
- _ -
,
'
,,
..
L
!-
l Axial Play Axial play is the total amount of movement within the disk arm socket in the axial direction.
Physically it is a measurement of the distance between the inside of the disk stud washer to the back side of the disk minus the disk arm thickness at the stud bore axis.
The axial play component was a consideration in the applicant's evaluation of the inoperable AFW check valves in that it contributes to the allowed dynamic interaction of the disk to the seat ring for both the pressure seal and bolted bonnet type Borg-Warner check valves.
The relative significance of the axial play component was addressed in Borg-Warner's letter to TU Electric dated June 7, 1989, concerning Borg-Warner high.
pressure swing check valves.
In part, this letter stated that historically the axial play was not considered to be a critical component.
However, in order to assure that the axial play would not adversely affect the operability of the subject valves, Borg-Warner will establish a maximum /
minimum dimensional acceptance criteria for this feature.
This dimensional acceptance criteria had not been provided at the conclusion of the AIT inspection and will be evaluated later.
Bolted Bonnet Check Valve Issues Concurrent with the AIT inspection efforts associated with Borg-Warner pressure seal check valve failures in the AFW system, two other similkr but apparently unrelated incidents occurred involving Borg-Warner bolted bonnet check valves.
The first event occurred on May 31, 1989, and involved a 4-inch 150 lb. check valve installed in the Service Water System (1SW-048).
The valve exhibited excessive backleakage and was determined to have the disk separated from the swing arm at a point roughly parallel to the ball disk assembly.
The failure mechanism and root cause for this valve, along with an investigation of known deficiencies associated with the corresponding valve on Unit 2, are currently being conducted by the applicant.
A second suspected check valve failure was reported on PIR 89-168, dated June 9, 1989, and involved the potential leakage of one or both of the 300 lb. check valves located in the discharge piping immediately downstream of the containment spray pump CP-1-01.
The AIT witnessed a special test to determine the nature of the reported check valvo deficiency.
This test was conducted on June 15, 1989, under the auspices of nonstandard alignments and evaluations procedure 1-89-0072.
This test essentially recreated the operational conditions of the containment spray system when the original pressure pulsation (check valve leakage) was identified.
Test observations and procedure review
__
_
- - - _ _ _ _
_ _ _ _ _ _ _ _ _ _ _.
.,
b ' $
..,
conducted by the AIT' indicated that the reported condition was apparently not the result of leaking check valves because the system operated correctly.
2.4.1.5 Corrective Action 2.4.1.5.1 Review of Retainer Ring Calculations To assist in determining the cause of the backleakage, the applicant, based on information obtained by radiographs of several Borg-Warner valves, preliminarily concluded that the cause of the problem appeared to be that the valve disk was stuck in the open position due to interference with the internal valve seat.
To confirm if this was indeed the cause of the backleakage, the applicant developed Computer Aided Design (CAD) models based on dimensions taken from the
"as-installed" valves.
This process was performed on several sizes of Borg-Warner check valves and these models confirmed the suspected cause of the problem (i.e. that the top of the valve disk was binding on the bottom of the upper portion of the valve seat).
This condition was caused by the bonnet being set too low into the valve body.
A secondary, minor contributor to this condition identified by the vendor representative, was the amount of axial play in the valve disk stud.
This additional axial play could cause
'the top edge of the valve disk to sit even lower in the valve body thereby increasing the possibility of interference with the seat.
The applicant intends to restore check valve function principally by backing out the retainer ring attaching the bonne. to the valve body.
This procedure will increase the shear stress acting on the individual threads of the retainer ring due to a reduction in the total shear area available.
The AIT reviewed calculations prepared by the vendor (Borg-Warner Job No. 891-H-2984) concerning the minimum thread engagement required to ensure that the retainer ring can resist the shear stresses anticipated at the design pressure of the AFW system.
From these results, a mazimum retainer ring backout for each size valve was calculated, ranging from 0.25 inches for 4-inch valves to 0.678 inches for 8-inch valves.
The applicant intends to i
set an administrative lLmit for retainer ring backout based
'
on the calculated results.
The applicant reviewed and i
,
I concurred with the vendor calculations.
Likewise, the AIT concluded that the calculations were acceptable and that they were based on conservative design input assumptions.
2.4.1.5.2 Corrective Action Plan To resolve the backleakage concerns for the Borg-Warner check valves associated with the AFW system that were i
i
- _ _ _
-
-
..
i i
d determined to be inoperable, the applicant issued nonconformance report (NCR)-89-6637.
This document defines the measurements that are needed, the methodology to be followed to calculate the required " retainer backout," the additional rework required, and the actual retainer backout for the thirteen AFW check valves known to have been leaking.
The AIT reviewed the methodology for determining the required retainer backout and concluded that the analytical technique was adequate.
This NCR also includes written concurrence from Borg-Warner.
For the remaining Borg-Warner check valves in Unit 1, the applicant issued NCR-89-7476.
This is an explanatory NCR which defines the dimensional data to be obtained in order to calculate the amount of " retainer backout" required to ensure proper function of the remaining Borg-Warner pressure seal check valves..This NCR also provides the direction necessary to determine if the axial play (amount of free movement of the swing arm relative to the bushing) in the Borg-Warner bolted-bonnet check valves is within allowable limits to ensure proper operation.
Borg-Warner is to provide the applicable minimum and maximum value of axial play that will not affect proper operation of these valves.
These two NCRs will ensure that all Borg-Warner check valves in Unit 1 will be inspected prior to fuel load.
The need for rework due to the exploratory NCR will be determined by engineering with all' work committed to be complete as soon as practical prior to fuel load.
Rework for Unit 2 has not been scheduled to date.
2.4.3.5.3 Post Modification Testina After the pressure seal check valves have been disassembled, measured, and reassembled with the proper amount of retainer backout as calculated by the method outlined in NCR-89-6637, the applicant intends to perform post-work testing.
This testing consists of subjecting the valves to a fluid flow in the reverse direction and measuring the relative drop in downstream system pressure after opening the upstream drain valve to confirm that the corrective action was effective.
I Testing for the bolted bonnet valves will be performed to a generic post-work test procedure and will test all valves that can be tested based on current plant conditions (i.e.
l existence of drain connections, etc.).
The applicant is in the process of developing a generic in-service test procedure.
2.4.2 Feedwater Isolation Bypass Valves L
_ _ _ _
_ _ _. - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_ - ___ ___ -.
..
.
2.4.2.1 Valve Description and Design Function
.
The feeowater isolation bypass valves and the feedwater preheater bypass valves are 3-inch globe valves designed for feedwater system isolation and provide a portion of the pressurc boundary of the steam generators.
The valves use air to open and spring pressure to shut.
This design allows
!
for tight shutoff against the maximum postulated inlet pressure.
The valves are used during startup and shutdown of the plant and are closed during plant operations.
The valves receive automatic signals to close within five seconds to isolate feedwater from the steam generators.
They are designed with the capability to isolate against the containment design maximurc pressure of 50 psig with minimal leakage.
Backpressure greater than 50 psig opens the valve against spring pressure.
2.4.2.2 Plant Packleakage Simulation and Valve Leak Tests The AIT witnessed a test entitled "AFW Backleakage Event Simulation Under Controlled Conditions" (ODA-408A, 1-89-049, Section 5.7) conducted May 7-8, 1989.
This test simulated plant conditions existing at the time of the AFW backleakage event of April 23, 1989, and was designed to determine the leak flow path and leak rate associated with that event.
One motor driven AFW pump was lined up to supply 50 gpm to i
each steam generator and valve 1AF-042 (turbine driven AFW pump recirculation to CST) was opened.
Then, separately for each loop, one valve in the preheater bypass line and two valves in the feedwater isolation bypass line were opened to simulate the plant line-up existing during the event (e.g.,
j for loop 4, valves 1FW-0203, 1FW-0207, and 1FW-208 were i
opened).
Backleakage was detected by monitoring temperatures of the upper and lower feedwater penetrations
{
(MI-8 and MV-17 for Loop 4) and by measuring flow rate with a strap-on ultrasonic unit.
The results were nearly identical for each loop, indicating that approximately 120 gpm leaked back through the spring-operated feedwater isolation bn ass valve (valve 1-HV 2188 for Loop 4).
Apparently no leakage occurred through the feedwater isolation valve in any of the loops because after the preheater bypass line valve (1FW-0103 for loop 4) was opened j
(with the feedwater isolation bypass line still isolated),
no signs of any leakage were noted.
Only after the feedwater isolation bypass line was unisolated (1FW-0207 and 1FW-0208 opened for loop 4) was leakage evident.
This test, therefore, demonstrated that the backleakage experienced
.
during the April 23 and May 5 events flowed through the feedwater isolation bypass line and the feedwater preheater j
'
bypass line to the AFW system.
The differential pressure across the feedwater isolation bypass valve apparently overcame spring pressure, unseating this valve in each loop.
J l
.
..
_ - _ -.
_
'
,.
.
'
This valve is designed for 50 psi backpressure for containment isolation purposes.
During the event, approximately 1000 psi backpressure lifted the seat against spring. pressure, allowing a backleakage flow of approximately 120 gpm. A flow path, therefore, was created from the steam generators through the leaking turbine-driven AFW pump supply line check valves to the CST.
The four main faedwater isolation bypass valves were calibratad by the Instrumentation and Control group (I&C) on May 9, 1989.
The valve set points were checked to verify that the valves were actually ful'.y open or fully closed as indicated.
All four valves were found to be satisfactory.
The AIT witnessed the implemr'tation of a backleakage test of the eight main feedwater p;eheater bypass line check valves on May 7, 1989.
The line configurations currently do not allow for the individual isolation of the two check valves in each of the four main feedwater lines.
The two check valves were tested in series and only one nonleaking check valve was needed for satisfactory test results.
The test was conducted in a manner similar co the tests of the AFW check valves described in Section 2.4.1.3 of this report.
The test results were satisfactory leading to the conclusion that at least four of the eight valves held.
All eight main feedwater preheater bypass line check valves have been or are scheduled for disassembly, repair, reassembly, and leak testing.
2.4.2.3 Applicant Intent and Corrective Action The applicant informed the AIT of their intent to administrative 1y isolate the feedwater isolation bypass valves during startup and shutdown conditions except when the valves are actually needed.
This would be done by closing the manual block valves in the feedwater isolation bypass line.
The applicant is also considering eliminating the currently installed interlock between the feedwater isolation bypass valves and the feedwater preheater bypass
valves.
This interlock forces one of these two valves to be
{
open and the other closed at all times other than during a
{
feedwater isolation signal (when both close).
>
l 2.4.3 Analysis of Auxiliary Feedwater Piping, Hangers, and Penetrations i
'
2.4.3.1 Evaluation of Event Effect on Piping I
I Following the April 23 and May 5, 1989 events, significant I
discolorization of the protective coatings of the AFW supply I
lines for steam generators Nos. 1 and 4 was identified.
l This discolorization was most pronounced on the piping for
'
!
_ _ _ _ _ _ _ _ _ _ _ - - - _ - _ _ _ - _ _ - _ _ -
_ _ - _ _
.
__ _ _ __ _ _ - ____
.m 3,
y.
27-(
,
,
Loop 1.
A significant amount of blistering and flaking of.
'the paint _ occurred as'a result of the. higher than anticipated temperatures.
J The piping is designed to ASME Section III Code Classes 2 and 3.
The Class 2~ pipe-from the steam' generator back to the first motor operated valves in the Safeguards.Buildigg on each' loop was. analyzed to a design temperature of 500 F-
and pressure of 1185 psi.
The Class 3 portion from the pump discharce to the class 2 portion,gwhich also saw higher temperatures was designed for 150 F.
Temperatures ~during the event-could have been as high as SG temperature (557 F);
thereforc the thermal portion'of'the piping analysis c
required, at a minimum, a review of stress levels to ensure that there were no excessive stresses' induced by the significantly higher temperature.
The design pressureu for the.feedwater and auxiliary feedwater are essentially the same; therefore, stress levels due to the 1185 psi water pressure are~not a concern.
Due'to the higher temperatures, the pipe supports will-need to be reviewed by the applicant for the effects.of higher than anticipated thermal forces
'
experienced during-the backflow events.
The analysis of the piping associated with the reverse flow event.in the auxiliary feedwater (AFW)-system, is being addressed-in two parts.
The first is associated.with the April 23 event.
During this event, the fluid from the steam generators Nos. 1 and 4 (SGs) flowed toward the condensate.
storage tank (CST) via the discharge lines of the turbine driven auxiliary.feedwater pumps (TDAFWp).
The temperature assumptionforthisportion.oftheaLagysis'wasthatthe piping experienced SG temperature (557 F) from its connection to the feedwater system to the junction of turbine driven and motor driven AFW lines.
From this point bacgtothe'headerpipingthetemperaturewasassumedtobe 325 F.
The reduction in temperature at this point in the piping is based on the fact that the motor driven AFW line was running and circulating water toward the SG at a temperature of 100 F (approximately).
From the hgader back l
to the CST the temperature assumption used is 200 F.
The second backleakage event resulted in a more severe condition from a thermal stress standpoint.
In this event, SGs 2 and 4 were isolated from the Jd5f system and based on volume change in the SGs and capacity of the piping, the-
'
flow path for the reverse flow occurred in loop 1.
The backleakage into the AFW piping was calculated to be
)
approximately 3000 gallons which was sufficient to fill the affected pipes.
Due to intermittent operation of the pumps, the amount of mixing of lower temperature fluid is indeterminate.
Also, the amount of severely discolored pipe suggests higher temperatures.
Accordingly, the temperatures i
l
_ - _ - = _ - _ -.
t '
l..:
L.
<
used for the thermal analysis extend the higher temperature fartger into the system.
Specifically, SG temperature (557 F) past the junction of the turbine and motor driven
.
.
pump discharge lines is considered to have travelled approximately 200 linear feet fartger upstream.
The temperature is then reduced to 400 F from this point back to-the header piping.
fhe temperature is then reduced-incrementally to 200 F back to the CST.
's For each of the two scenarios, there are portions of the pipe which are overstressed, and stresses were most severe for the second event.
After the second event, it was noticed that support AF-1-096-023-S33R had failed.
This
,
support is located in the tunnel at the 810'-0" elevation
'
just before the piping turns south toward the pump rooms.
This support has been replaced in accordance with the disposition of NCR 89-6332, Revision 0.
Also, the piping analysis shows that the location of the maximum thermal stress for both events is adjacent to this failed support.
As mentioned above, there are several areas in the piping which experienced thermal stresses higher than code allowables.
These areas were identified by analyzing the piping using the higher temperatures outlined above.
In determining the effect of these overstressed conditions, SWEC is performing the following steps.
First, the allowable stress level was increased to agree with'the one time allowable provided by the code; further, in determining this allowable stress, actual physical properties from the
!
applicabic certified material test reports (CMTRs) were used.
Based on these values, only two areas of concern remain:
(1) the elbow adjacent to the failed support and (2) some instrument connections.
There exist additional
conservatism for the instrument connections which should
'
climinate these connections as areas of concern:
first, the use of high stress intensification factors (SIFs) for the connections and second, ignoring the existence of gaps which will reduce the actual rigidity in the structural frames restraining the instrument lines.
Evaluating these connections by more precisely modeled field conditions will result in greatly reduced stress values.
For the elbow, even if the assumption is made that the failed support does not exist, a relatively high stress still would have existed.
However, when the worst case analysis is considered in light of actual material behavior, a small amount of yielding would have occurred and then the stresses would be redistributed in the system with a minimal impact on the elbow itself.
To ensure against any potentially adverse conditions, RT and UT were performed on the elbow to determine if any cracks exist.
The results of this nondestructive examination did not disclose any cracks.
_ __
- _ _ _ _ _ _ _ _ _ _ _ _ _
l q
-
..
!
..
!
Therefore, it was concluded-that replacement of the elbow was unnecessary.
The AIT concurs with this assessment.
2.4.3.2 Evaluation of Event Effect on Pipe Supporus/ Restraints On the AFW piping system, there are 563 supports, restraints, and anchors.
To date there are load increases on 59 cases where the deadweight and thern al load due to this backflow event exceeded the design lead used in the original calculations for the particular sapport.
Additionally, there are several levels of review which will be followed to completely evaluate the need for rework.
For example, if the transient load due to the backflow events is higher than the original design load, a review of remaining design margin will be conducted.
At this point, the design margin relates to code allowables based on minimum expected material properties.
The supports that exceed code allowable will be reviewed against a one-time allowable value.
If necessary, the final determination of acceptability will be dependent on a full consideration of actual physical conditions (i.e. gaps to accommodate thermal expression, actual stiffness, actual material properties etc.).
The AIT has reviewed the proposed method for resolution of the actual load increases and concurs with the approach presented.
2.4.3.3 Evaluation of the AFW Event Effect on Penetrations The applicant evaluated the structural integrity of auxiliary feedwater cold penetrations (MV-17 to MV-20)
subsequent to the April 23 and May 5, 1989, auxiliary feedwater backflow events.
A preliminary analysis was performed conse5vatively assumin9 the penetrations experienced 550 F as no definitive indication'of the temperature at the penetrations during either event was available.
The actual maximum temperature experienced by thesepenetratgonsisthoughtbytheapplicanttobemuch lower than 550 F.
The check valves inside the containment in the feedwater preheater piping did not leak and these penetrations apparently were not part of the backflow path.
The preliminary analysis was reviewed by the AIT and found to be very conservative in nature and to adequately address expected failure modes.
The analysis included an evaluation of concrete bearing frvm the shear lugs, moment applied to the welding on the lugs, punching shear in the concrete, radial loads in the concrete, and pipe wall stregses.
The analysis concluded that thermal expansion at 550 F of the pipe penetration should have caused spalling and/or crushing of the concrete.
I
_ _ - _ _
_
_
. _ _ _ - _ _
",,
..
i
V!.sual walkdowns of the penetrations by the applicant f.ollowing the AFW events showed no evidence of any concrete distress or pipe movement.
Hairline cracks typical of cracks observed following the structural acceptance test of the Unit 1 containment building were observed radiating outward from some of.the penetrations.
The applicant concludedthatthepenetratgonsdidnotexperience temperatures as high as 550 F.and that the penetrations were l
not adversely affected by the AFW events.
The AIT discussed the occurrence of the hairline cracks with the applicant and inspected the penetrations.
The AIT concluded that the penetrations have not been damaged.
It should also.be noted that ASME B&pV Code,Section III, Division 2, Subsection CC, specifically CC-3430, stipulates that local areas of concrege (containments) are allowed to reach a temperature of 650 F for a short term period, where short term is defined as 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> or less (based on a code interpretation).
2.5 Personnel Actions / Human Factors 2.5.1 Operator Actions The AFW events of April 23, 1989, and May 5, 1989, resulted from combinations of operator errors and equipment failures.
In the first event, the auxiliary operator (AO) operated two valves out of sequence; i.e., he opened valve 1AF-042 prior to fully closing valve 1AF-041.
This operational error, coupled with multiple check valve failures, resulted in an open flowpath backward from the steam generators (SGs) to the condensate storage tank (CST).
A nearly identical out-of-sequence valve operation occurred during the second event (May 5).
This time the operators opened valve 1AF-0.55 prior to fully closing valve 1AF-054.
In both of thesa events, operator actions played a significant role.
T r.
Part 2 of the second back-flow event, however, operator actions figured in less significantly.
Here, the inability of the operators to detect a less-than-fully-shut valve (due to extremely long, articulated valve / handwheel linkage)
resulted in a similar backward flowpath being established from the SGs to the CST.
The first event was initiated while the operators were aligning the turbine-driven auxiliary feedwater (TDAFW) pump to recirculate to the CST.
The applicable procedure, SOP-304A, " Auxiliary Feedwater System," Revision 5, clearly specifies that the TDAFW pump discharge isolation valve, 1AF-041, be closed prior to opening the TDAFW pump test isolation valve, 1AF-042.
The reactor operator (RO)
reviewed the procedure with the AO, ordered the Ao to. shut valve 1AF-041 and open 1AF-042, and dispatched the AO ;o
_ _ - _ ____ -__ - _____ - _ _
.
.
_
_
_
_
-
_ _ _
_ _ _. _
e '
.c
31 accomplish the task.
When the AO arrived in the TDAFW pump room, he noticed that the OPEN/CLOSE direction tag for valve 1AF-041 was missing; therefore, he was confused as to which direction to turn the handwheel to close the valve.
In order to quickly determine the CLOSE direction for-valve 1AF-041, the AO went to valve 1AF-042, similar in design to lAF-041, and spun its handwheel in the OPEN direction while observing its gearbox.
Doing so served two purposes:
(1) by observing the gear motion on 1AF-042 while turning it in a known direction, the Ao could determine which direction to turn lAF-041 handwheel in order to close it and (2) 1AF-042 needed to be opened anyway; with multiple check valve protection, opening it slightly ahead of time should logically not cause any problem.
After cracking valve 1AF-042 off its seat, the AO contacted the control room for assistance; he knew that closing 1AF-041 alone would require one half hour.
Then, he began to close 1AF-041.
When the extra AOs arrived, he directed them to open 1AF-042 and to relieve him in the task of closing 1AF-041.
Just prior to his exiting the TDAFW pump room, the control room contacted him via radio and told him that steam generator levels were dropping rapidly.
At this point, he went to the two motor-driven AFW pump rooms and verified that the test isolation valves for the two motor-driven AFW pumps were shut.
After verifying the valves were shut, he
~
reentered the TDAFW pump room and noticed what appeared to be steam.
He then noticed that the paint on the AFW lines was blistering.
At this point, the AO contacted the control room, apprised the control room personnel of the situation, was ordered to shut lAF-042, and ensure that the TDAFW pump room was evacuated.
In the meantime, the RO stopped SG blowdown.
With 1AF-042 shut and blowdown secured, the event was over.
Within 15 minutes of securing SG blowdown, SG l
levels had recovered to their normal levels.
In evalueting the operator actions for this first event, it is apparent that the AO violated procedure by not operating the valves in the correct sequence.
Furthermore, it is clear that he did not appreciate the potential for check valve backleakage and its consequences.
Finally, it is clear that the AO was under considerable presrure to complete the valve alignment by the end of the shift.
The Ao is not the only operator who performed poorly.
The RO, unit supervisor, and the shift supervisor share some of the responsibility for the poor performance.
It should have been apparent to them that to send one Ao to the TDAFW pump room to manipulate the two valves (1AF-041 and 1AF-042) at the end-of-shift was unreasonable.
Control room personnel should have dispatched more than one AO or left the manipulation for the next shift.
I l
l L
_
-- -
-- - ---i
_
[g
.,
c
,
-
The second event on May 5, 1989, occurred during the performance of the'AFW system operability test.
This test was being performed in_accordance with Procedure OPR-206A,
" Auxiliary Feedwater System Operability Test," Revision 2, as.part of the surveillance test program.
Basically, the test began with placing AFW pump 1-02 in recirculation to the CST per Procedure SOP-304A, " Auxiliary Feedwater System."
Data was taken and the pump was secured.
As in the first event, improper sequence of Jalve manipulation resulted in backflow from the steam generators to the CST.
Because SG levels had decreased during the pump run, valves 1AF-090 and 1AF-091 (cross connect valves) were opened, valve 1AF-055 (test line isolation valve) was closed (but inadvertently left partly open), valve 1AF-054-(motor-driven AFW pump 1-02 discharge isolation valve) was opened, and AFW pump 1-02 was started.
After starting the pump, the operators noticed that total pump flow was 300 gpm (abnormally high), but flow to the steam generators totaled only 80 gpm.
Therefore, AFW pump 1-02 was secured.
The.
operators then started AFW pump 1-01 and again' observed that total pump flow was 300 gpm, again, abnormally high.
Because SG levels were very low, the pump was allowed to run for ten minutes.
After securing it,. valves 1AF-090 and 1AF-091 were closed and AFW pump 1-01 was started to feed SGs 1 and 2.
Several minutes later AFW pump 1-02 was started to feed SGs 3 and 4.
The operators noticed that AFW pump 1-02 total flow was, again, 300 gpm.
Therefore, suspecting backleakage on valve 1AF-055, an AO and the unit supervisor checked it to verify that it was shut.
They discovered that they could turn the valve shut another one-quarter turn.
Upon doing so, total flow for AFW pump 1-02 dropped to 80 gpm.
Soon afterwards, an AO informed the control room that the AFW lines in the TDAFW pump room were hot and reported this information to the control room.
In evaluating Part 2 of the May 5, 1989 event of AFW backleakage, it is apparent that operator actions (errors)
did not play as direct or significant a role as in the April 23, 1989 event or Part 1 of the second event.
Instead, management and supervisory factors figure heavily into this event.
Essentially, management was clearly aware of the events surrounding the first event.
The question comes to mind:
Why perform an auxiliary feedwater system operability test (OPT-206A) knowing that multiple check valve failures would not permit the system to operate as designed?
Performing this test under such circumstances virtually ensured that more piping would be overheated - and it was.
Furthermore, by the time of the second event, it was clear that the TDAFW pump discharge and test line isolation valves (1AF-041 and 1AF-042) were not " operator friendly."
Should
_ _ _ _ _ - -
_
- *
..
33 not the corresponding valves (lAF-054 and 1AF-055) be suspected of having the same, or similar characteristics?
2.5.2-Management Involvement / Oversight Within hours of the April 23, 1989, event (which occurred towards the end of the graveyard shift), a Plant Incident
. Report (PIR-110) was initiated by the shift supervisor; personnel' statements were obtained from him and the auxiliary, operators involved in the event by the next day.
The Operations manager met with the operating crew on. April 24, 1989.
A management meeting, attended by an NRC resident'
inspector, was also conducted on April 24, 1989.
The latter resulted in the development of test plans to leak test check valves between the AFWPs and steam generators with the intent of identifying possible backflow leakage paths.
It was not until May 1, 1989, that the applicant established a task team with responsibility for investigation of all aspects of the AFW system check valve problem.
The team was directed by the Operations manager and comprised of engineering representatives from Unit 1 Projects, Scheduling, Technical Support (Results), Performance and Test, Licensing, and Consolidated Engineering and Construction Organization (CECO).
Within two weeks of the April 23, 1989, event, the Task Team had established an action plan for investigation of the problem, assessment of input on the affected system and equipment, identification of corrective actions, and determination of generic implications for other plant systems / equipment.
The Task Team met daily for team members to report on the status and results of various action plan activities and to identify necessary additional actions.
AIT members attended these meetings.
The AIT found that the Task Team approach, provided a means for coordinating the various organizations in identifying and resolving the issues raised as a result of the events.
Borg-Warner representatives were on site intermittently during activities concerned with assessing the check valve failure mechanism.
Additionally, the applicant used an onsite engineering consultant, Kalsi Engineering, Inc., to provide advice and recommendations during the course of the investigation.
The AIT found the applicant's action plan to be comprehensive.
The AIT observed that the applicant's preparations for execution of valve testing and disassembly were coordinated and conducted in an orderly fashion with an appropriate level of management oversight.
However, it should be noted that more than six weeks passed after the initial event before the applicant arrived at a conclusion on the root cause of the valve failure and determined corrective action to be taken on the AFW system check
_
- - _ -
. _.
._
..
t *
valves.
The results of the thermal stress analysis on the affected piping and the generic implication of the check valve failure for other plant systems and equipment have not yet been completed.
While the slow pace at which action plan activities progressed contributed to the comprehensiveness of the evaluations completed to date, it was indicative of the lesser importance placed by applicant management on the events investigation as compared to other plant activities necessary for Unit 1 licensing.
The AIT considered the events to have significant safety importance and expected that the applicant management would have caused the investigation to proceed in a more expeditious fashion.
In the course of its inspection, the AIT learned that at least some first-level supervision was aware that during a prior flushing operation (April 5, 1989), a number of check valves were found to be leaking.
Despite this information, which appears was not adequately communicated to higher levels of management for evaluation, the applicant proceeded with HFT and subsequently experienced the subject events of i
concern.
The AIT also observed that the applicant's record retrieval capability was slow and this somewhat hampered the progress of several action plan activities.
2.5.3 Procedural /_ Human Factors Deficiencies
'
Personnel directly involved in both the April 23, 1989, and May 5, 1989, events were interviewed.
The purpose of the interviews was to determine the extent to which any of the following played a significant role in each failure: plant material condition; the quality of maintenance; or the responsiveness of engineering to identified problems.
During the April 23 event, the first shift crew was preparing to perform a full-flow hot alignment test on the turbine driven AFW pump (TDAFWP).
This w&s the last of several hot functional tests conducted during that shift.
Based on comments obtained during the AIT's interviews, the crew had been very productive during the shift and there was an apparent press to complete preparation for this last test prior to shift turnover.
The balance of the plant reactor operators (TU) was directed by the unit supervisor to prepare the TDAFWP for full flow recirculation to the condensate storage tank (CST).
The RO subsequently reviewed Procedure SOP-304A with the AO, which describes the steps necessary to operate a motor driven auxiliary feedwater pump or the TDAFWP for recirculation flow to the CST.
The procedure directs the equipment operator to locally close the discharge valve (lAF-041)
on
-__-----____ __ _ ___ _ _ _ J
-7
~ '
..
,,
O
the TDAFWP and to open the recirculation test line isolation valve (lAF-042).
The order in which the valves are to be manipulated is explicit in the procedure.
Responses to questions posed during AIT interviews indicate that the proper order of valve manipulation was not specifically emphasized during the review of the test procedure.
The AO entered the TDAFWP room where he'" cracked 1AF-042 off the seat, approximately 1/4 turn.
then proceeded to
..
start closing 1AF-041."
He requested that the RO provide him with assistance in performing his task.
The RO
!
dispatched three other personnel to the TDAFWP room to help in manipulating the valves.
Two of the Aos responded immediately while the third was delayed for approximately 5 minutes.
Although the three personnel indicated a general familiarity with the alignment being attempted, none had actually reviewed the procedure prior to entering the pump room.
When the two operators entered the pump rocm, one was directed by the original AO to open valve 1AF-042 and the other was requested to manipulate the 1AF-041 valve.
AF a result of these actions, both valves were being manipulated concurrently rather than consecutively.
This resulted in a flow path from the steam generators through the system's faulty check valves to the TDAFWP recirculation test line into the condensate storage tank (CST).
The first AO's rationale for conducting this procedure out
.
of sequence derived from his familiarity with the system, l
his anxiousness to complete his task, and some minor frustration associated with the valve operators.
The first AO's familiarity with the system led him to believe that he could rely on the system's check valves to prevent any back leakage which would possibly result from his operating the valves out of sequence.
He also was aware that his shift was nearly over and he felt a need to expedite the completion of the alignment.
The frustration to which the Ao referred was precipitated by the system's valve design, the valve's physical orientation, and the applicant's practices with respect to maintenance of valve packing.
The 1AF-041 valve requires about 1000 turns of the handwheel to fully stroke while the 1AF-042 valve requires only about 60 turns.
(The discharge valves on the motor driven pumps require about 460 turns to stroke.)
A large electrical junction box is located in proximity to the 1AF-041 valve handwheel.
This box presents an obstacle to ACJ when they attempt to manipulate the valve handwheel often resulting in bumped and bruised knuckles.
Also, the plant policy with respect to the installation and maintenance of valve packing is to tighten the packing to the point where "the valve stem squeaks."
This results in difficulty in manipulating valves "particularly when the l
L---_______
- - - - _ -
_.
_
-
" *^'
l4 i
I
!
valve is backseated."
The remote mechanical linkage by which the TDAFWP valves are manipulated necessitates rotating the hand wheel in the direction opposite that which
,
l would normally be expected if the valve were operated f
directly.
Although there is normally indication affixed to
,
'
the hand whcol to' indicate proper direction of rotation, this indication was not present on the 1AF-042 valve.
This l
required unnecessary distraction on the part of the Ao in determining the proper direction of hand wheel rotation.
This problem alone is not particularly significant; however, in concert with other existing conditions, this problem helped to exacerbate the actions of the AO.
On May 5, 1989,.the second hot water intrusion event occurred at the plant.
This event took place on the second shift.
Part 1 of this event was essentially identical to the first event (April 23).
Part 2 portion was initiated by equipment operability problems.
A full flow test of the No. 2 motor driven AFW pump (MDAFWP) had been conducted in accordance with AFW operability test Procedure OPT-206A.
This test entailed closing the t'o. 2 MDAFWP isolation valve 1AF-055 and running the pump to obtain various operability data.
The test was performed satisfactorily.
Because steam generator levels decreasad during the test, the cross connect valves between the discharge headers ot the two motor driven pumps were opened to allow the No. 2 pump to supply feed flow to all steam generators.
Valve 1AF-054 was opened.
An unsuccessful attempt was made to close valve 1AF-055; however, the fact that the valve remained partially open was not determined until the event was well underway.
The reason the valve was not fully closed is tied to the type of operator used to manipulate the valve.
This valve has been characterized as " difficult to operate."
The remote operator consists of a 15 foot reach rod connected to a 10 foot reach rod through a ninety degree universal joint.
During the event, the valve handwheel had been fully rotated in the closed direction; however, the valve remained 1/4 turn open.
Because of the physical configuration of the reach rod operator, either binding occurred in the universal joint connecting the rod sections or an excessive amount of the force applied by the AO in turning the hand wheel was expended in establishing torsional (twisting) forces in the rods.
One or both of these conditions gave the AO the
" feel" that the valve was seated.
Because of the location of the ve.1ve relative to the handwheel, the Ao was unable to visually determine the degree of closure of the valve.
With the cooning of the No. 2 pump discharge valve, a low pressure flow path was established from the steam generators, through the faulty check valves, and through the open recirculation valve to the condensate storage tank.
- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - _ _
_
. _.
._
~
,p
.
...
L This scenario was similar to the April 23 event and again resulted in the backflow of hot water into the AFW system.
,
After'a period of about.two hours, the open recirculation i
valve was discovered.
The unit supervisor and the AO, together, were able to turn the remote operator an additional 1/4 turn, fully seating the valve and terminating the event.
The AIT asked personnel involved in both the April 23 and
,
May 5 events their impression of the general material i
condition of the plant and whether it p2ayed a role in either event.
The consensus of opinion was that,.not withstanding the severity of the check valve failures and the concerns regarding valve mechanical operators, the I
'
plant's response during the conduct of this hot' functional tect was better than anticipated, given the length of time the plant has been under construction.
However, several of those interviewed indicated adverse personal experiences with remote-valve operators, q
!
There is a perception among those interviewed that the use of remote valve operators at the plant is abundant.
(One person interviewed-told that there was an " overuse" of these operators.)
The design and placement of some of these operators appears to have been executed without proper regard to human factors issues.
For example, the recirculation test line isolation valve on one motor driven AFW pump has a chain operator, while the equivalent valve on the other pump is manipulated with reach rods.
When asked if there had been any attempt on the part of any u
of the personnel interviewed to make their concerns known to appropriate management regarding perceived design or operational deficiencies, responses varied.
Although there are formal procedures in place at the plant for requesting changes to system or equipment design, there was some uncertainty with respect to the appropriate vehicle to be used for a given change request.
This may be attributed to the fact that unit construct $ou is continuing and a concerted effort toward apprialng plant personnel of the availability and proper use of formal plant procedures for requesting changes or reviews is yet to be instituted.
There appeared to be a consensus among those interviewed that management is currently more responcive than in prior years to personnel requests and suggestions.
This altered management attitude is attributed to the nearly complete change in senior management which has occurred at the plant.
The availability of hydraulically operated " man lifts," the construction of " catwalks," and the use of portable air operated wrenches for some "long winded" valves are changes
. _ _ _ _ _ - - _ - _
,
_ _ _ - _....
,...
.. -
p..
,
which have resulted from increased management attention to the needs of operating personnel.
'AIT personnel found that the necessary presence of construction equipment and n,terials;.i.e., scaffolding, test instruments, etc.,:could present significant obstacles to personnel in their attempts to manipulate some equipment.
However, this was not found to be a factor in the operating
!
events currently under review.
!
All~ personnel interviewed were asked their opinions regarding whether engineering at the site was sufficiently responsive when design problems were identified.
Individual views were mixed, but, generally, all felt that the change L
in management at the plant has had a beneficial impact on.
the quality and responsiveness of engineering personnel at the plant.
However, because of the current phase of plant construction, it is often difficult to obtain adequate response to identified problems.
Each of the personnel interviewed was requested to provide an cpinion regarding the quality of maintenance at the plant and his perception of its impact on the events under review.
Again, most responded that the quality of maintenance has improved as a result of the management change that has occurred in recent years.
However, there were two comments which were somewhat critical.
The first comment questioned the policy at the plant of removing check valve internals in the AFW system to facilitate flushing.
The second related to the scarcity of documentation associated with a completed maintenance procedure.
As was previously discussed in paragraph 2.4.1.2 with respect to the policy of removing check valve internals, it was stated to the AIT that a thorough flushing of sections of the AFW system could not be achieved with the existing system drain valves.
Therefore, the applicant removed the appropriate check valve internals to allow for increased flushing flow rates.
This was perceived as a possible system design flaw on the part of the person reporting.
The AIT requested clarification on this policy from members of the applicant's AFW Check Valve Task Team.
The applicant's team informed the AIT that it was a " routine" policy at the site to remove check valve internals to enhance system flushing.
(The task team did not state that this policy existed to allow for back-flushing of the system.)
The AIT's concern is that the numerous failures of AFW check valves to seat properly may be related to the applicant's routine practice of removing check valve internals for the
l purpose of flushing and draining.
The valves were not
designed to be routinely disassembled.
l l
l l
l
_ _ _ _ _ _ _ _ _ _ - _. _
__
__. __- ____-
_--
-
--
.
(..
39 The concern regarding the lack of sufficient documentation following the completion of maintenance activities appears to be historical in nature.
Apparently the maintenance policies in place prior to the major management change discussed above were derived from the practices at fossil-plants.
These practices tended not to be as sensitive to quality assurance requirements as one would expect from a
,
nuclear based system, However, with the change in manage: ment, the person interviewed believes a change in the
attitude regarding the importance of proper and complete documentation in support of maintenance activities is forthcoming.
2.6 Quality Assurance Considerations The events of April 23 and May 5, 1989, represent in part the failure of the applicant's quality assurance program to detect the latent problems underlying these events and to provide corrective measures to prevent them from occurring.
Quality assurance is most effective when events are prevented beforehand rather than as a reaction afterward in an effort to prevent recurrence.
However, the individual elements which combined to create these incidents, for the most part, transcended what is normally construed to be the responsibility of site quality assurance.
The error in the vendor's technical manual regarding check valve installation was clearly the primary root.cause for the backflow events.
Only a highly detailed and somewhat fortunate vendor audit could have detected this problem.
The secondary root cause was the failure of post-maintenance and post-modification testing to perform-backleakage tests of the check valves.
Although these tests would have been prudent and indicative of good engineering judgement, they were not procedurally required, due in part to the fact that the various applicable codes and standards emphasized only the forward-flow capabilities of check valves.
In this perspective, again, the culpability of site
!
'
quality assurance is minimal.
Therefore, it can be stated that quality assurance failures played only a minor role in the two principal root causes for the events under investigation.
The report addresses several precursor events (see paragraph 2.3) which considered collectively should have led a reviewer to suspect that a generic check valve problem existed.
It is here where quality assurance may have failed to notice the adverse trend.
But the timing of these events is critical to the severity of this judgement.
The failure
.
of main steam vrlves 1MS-142 and 1MS-143 occurred in 1983
!
and 1985, but the failures of miniflow check valve 1AF-069 and the three check valves in the AFW system (lAF-106, 078,
_ _ _ _ _ _ _ _ _ _ _ - _ - _ _
._
- -.. __
,,,
-.a-
.
086) all occurred within 2.1/2 weeks of the first backflow event.
Very-little time existed for a quality assurance trend evaluation.
Another quality-related issue that was instrumental in this event was the training of plant operators.
Somewhere in this training, the essentials of in-sequence valve operation were not sufficiently emphasized.
The applicant has committed to conduct' additional training in'this area.
Another training-related issue was the failure of plant operators to document the discovery of.three failed AFW check valves (discussed above) on an NCR or PIR and to recognize the resultant impact on the operability of the AFW system.
The applicant recognized this tailure, pointing out that the mindset of plant l operators is still ingrained'in construction.
The applicant has committt$ to raising the awareness of plant operators to operational issues.
Another area where quality assurance may have gained insight into the check valve problem was the steam binding issue raised by I&E Bulletin 85-01.
This bulletin suggested the possibility of AFW check valves allowing leakage by their seats in sufficient amounts to thermally bind a pump.
The corrective action which resulted from the bulletin was to utilize AFW temperature sensors and to feel the pump discharge piping every shift to detect the presence of leakage.
The NRC considered this commitment to sufficiently address the issues of the bulletin.
Little can be said negatively of the applicant's actions on this_ issue except to suggest the possibility of the more proactive approach of physically testing a few valves to determine whether the problem currently existed.
In summary, the AIT team has concluded that the AFW check valve events do not suggest a major problem in the site quality assurance organization.
These events do, however, point out weaknesses where programmatic enhancements would be prudent.
2.7 Applicant Evaluation 2.7.1 Evaluation of Applicant's Timeliness and Accuracy in Reporting the AFW Incidents to the NRC In the first event, the NRC Senior Resident Inspector was notified promptly.
While all the details were not yet
,
'
apparent at the time of notification, it appears that the applicant reported this first AFW event in a timely and accurate manner.
In contrast, the applicant was not nearly as timely in reporting the second event.
Basically, various NRC inspectors learned at different times that "more pipe had overheated" during the performance of an operability
_
-__ __ _ _ _ _ - _ ______ - _ L
_ _ _ - - _ _.
_.
,..
,,
4" l
surveillance test on the AFW system.
In fact, the second event received so little attention that the Shift Test
'
engineer for the AFW system was not aware of the newly overheated-piping until several days after the event.
Furthermore, a plant event report was not written until more
~
i than a week after the event and then only at the insistence of the AIT.
l l
In summary, the applicant displayed an insensitivity to the seriousness of the second event.
Apparently, the Operations department' felt that running an AFW system operability surveillance test was a routine operational procedure even.
on a system in which multiple mechanical failures were evident.
2.7.2 Evaluation of the Implications on Other Equipment in Other Safety Systems at Comanche Peak In light of the observed failures.of eight 4-inch and three 3-inch Borg-Warner check valves in the Unit 1 Auxiliary Feedwater System, the question exists whether other Borg-Warner check valves located in other safety-related systems may have similar failures and thereby degrade the safety and reliability of the plant.
A total of 58 Borg-Warner check valves are located in safety-related systems other than auxiliary and main feedwater and are distributed as follows:
Component Cooling Water System l
4 3-inch check valves (2 per unit), 150#
10 4-inch check valves (5 per Unit), 150#
2 8-inch check valves (1 per Unit), 150#
2 10-inch check valves (1 per Unit), 150#
Main Steam System (supply to TDAFWP)
4 4-inch check valves (2 per Unit), 900#
Containment Spray System 4 4-inch check valves (Unit 1 only), 300#
8 10-inch check valves (4 per Unit), 300#
12 16-inch check valves (6 per unit), 150-300#
Service Water System 4 4-inch check valves (2 per Unit), 150#
8 10-inch check valves (4 per Unit), 150#
_
__________ __ ____ ___________-_ - _ - ___ m
_
-__
,
'
..
.
.
The applicant has committed to physically examine, make necessary. adjustments, and test the internals of each Borg-Warner check valve in Unit 1 and common prior to fuel load-(Unit 2 is to be addressed at some later time).
This effort should restore confidence that these check valves-will perform.as designed.
2.7.3 Applicant Action on EPRI Guidelines and INPO Significant Operating Experience Report SOER 86-03 As a result of several events involving check valve malfunctions, the NRC contacted the four NSSS Owners Groups in February of 1986, urging them to take a leadership role I
in addressing the design, testing, and maintenance of l
safety-related check valves.
The Institute of Nuclear Power j
Operations (INPO) issued ~a Significant Operating Experience Report SOER 86-03, " Check Valve Failure or Degradation" dated October 1986, on this subject.
In preparing the SOER, check valve failure data on 15,400 check valves included in Nuclear Plant Reliability Data System (NPRDS), Licensee Event Reports (LERs), and previous INPO publications were analyzed.
In addition, check valve manufacturers and architect-engineers were contacted to identify the causes of check valve failures.
Some broad recommendations to prevent check valve failures or degradation were provided in an EPRI report titled, " Evaluation of NUREG-1190 findings on the Adequacy of Check Valve Applications and Maintenance / Surveillance Practices."
This report was developed to provide guidance to utilities in responding to SOER 86-03.
Kalsi Engineering, Inc. is assisting TU Electric in developing and implementing a program based on SOER-86-03 recommendations.
It was initially decided that Kalsi would proceed with its evaluation on a system-by-system basis beginning with the Chemical Volume and Control System.
After the check valve
!
event, Kalsi was asked to shift their effort from the CVCS to the AFW system.
Kalsi has completed their evaluation of the AFW check valves.
Evaluation of the other systems (Main Steam, Service Water, Diesel Generator and Auxiliaries, Chemical and Volume Control, Safety Injection, Residual Heat Removal and Feedwater) is expected to be completed by June 30, 1989.
The main objective of this program is to develop a preventative maintenance schedule and inspection procedure for each check valve located in the above-mentioned systems.
Program priorities are based on many considerations, such as the consequence of valve failure, the location and orientation of the valve, the expected operational environment, and its maintenance history.
Review of the AFW check valves by Kalsi, Inc. is complete and a summary of their recommendations is as follows.
All
_ _ _ - _ _ _ _ _ _ _ - _ _ - _ _ _ - _ - _ _ _ -
_ _ _ _ - _ _
E
-
3-<
v-
the. check valves.in the auxiliary feedwater system were reviewed.
These are 3, 4, 6, and 8-inch Borg-Warner swing check valves.
The 4-inch valves in the. turbine and motor driven supply lines are located anywhere from 18 to 36 inches from 1-inch diameter flow limiting orifices which are treated as high turbulence sources.
-
Based on the design conditions specified in the FSAR, the flow through the 4-inch valves will be 286 gpm and typical usage would be less than 50 hours5.787037e-4 days <br />0.0139 hours <br />8.267196e-5 weeks <br />1.9025e-5 months <br /> per year.
Under these-flow conditions, the disk is predicted to be oscillating at high levels.
Because of the low asage, the calculated wear and fatigue indices are both very low and are considered acceptable.
However, during the plant pre-start-up period the use of these valves is likely to last considerably longer and may be at significantly higher flow rates.
Analyses performed at flow rates of 500 and 570 gpm for operation during Hot Functional Tests predict tapping and oscillating of the disk.
The calculated Wear Index is acceptable because valve usage is very low during a 12-month plant cycle.
Fatigue Index is, however, unacceptable due to high stresses developed when the disk is tapping.
Kalsi therefore recommended that these higher velocities should be avoided.
Kalsi also recommended during inspection of the 4-inch valves, the hinge pin, bushings, disk stud / hinge connection, disk and seat should be checked for wear and damage.
The 3-inch Borg-Warner swing check valves in the AFW pump miniflow lines are even more susceptible to the high flow turbulence.
Based on analysis and review of recent backleakage problems with similar valves and inspection of
.lAF-0069, inspection of each valve prior to plant start up is recommended by Kalsi in order to rectify the disk and seat alignment problems.
During this inspection, the following areas should be checked for wear and damage:
Hinge pin and bushings.
d.
.b.
Disc stud and stud-hinge connection.
c.
Disc and seat.
Inspection of check valve 1AF-0069 revealed signs of considerable damage due to tapping contact with the disk stop such as a bent and peened disk stud and impact depressions on the disk stop.
Kalsi has recommended design revision for the three 3-inch check valves located in the AFW pump miniflow lines.
Kalsi states that if the situation is not corrected, these valves will suffer from exceptionally short lifn due to the high stresses developed during tapping.
In the absence of quantitative assessments on how long these valves could operate without failure, it is recommended by Kalsi that corrective action be initiated I
..__ - _ L
.
__
.-.
,,
.-
'immedi'ately.
In' addition, higher flow rates (>500gpm)
should be avoided due to fatigue-related problems in the AFW motor driven pumps 1 and 2 and turbine driven pump supply lines.
2.7.4 Applicant Action'en other Site Failures and Generic Communications The applicant performed a scarch through INPO and retrieved failures of Borg-Warner check valves at other sites listed in the Nuclear Plants Reliability Data System (NPRDS).
A total of 38 failures of Borg-Warner check valves were retrieved.
Of the total failures, 23 were identified as disk seating failures.
Of these 23 failures, approximately 75 percent were reportedly caused by either foreign material caught between the disk and seat, disk distortion, improper installation of disc-stud-hinge arm assembly, or erosion / corrosion-of valve internals.
The remaining seating failures we.:re attributed to normal wear or indeterminate causes.
Individual contacts were made with four plants identified I
through NPRDS to discuss their specific problems.
No other plant experienced the exact disk binding found at CPSES although all expressed concern with the general quality of their Borg-Warner valves.
One plant, St. Lucie, had to have the clevis of their 12-inch Borg-Warner check valves machined prior to shipment.
They were told by Borg-Warner that this was a "one of a kind" fix and that future maintenance of these valves would have to consider the shorter clevis.
It is unclear at this time whether or not i
this is significant to the CPSES incidents.
Investigation into this item is continuing.
The McGuire Units 1 and 2 experienced full backflow through Borg-Warner pressure bonnet swing check valves under circumstances very similar to CPSES backflow events.
The valves were replaced before a definite cause was determined.
It is suspected; however, that the vertical positioning of the disk assembly caused the failure.
Based on information obtained by the applicant from several plants (St. Lucie and Diablo Canyon) in regards to seal ring and pressure sealed valve applications, it appears that these plants have experienced other problems with Borg-Warner check valves such as bonnet leakage.
The applicant's actions in response to a number of IE Bulletins and Notices on related check valve failures were reviewed by the AIT.
.~,e of those considered significant are summarized below:
. _. _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ - _ _
- - - - - - -
.
..
..
.
IEB 85-01, " Steam Binding of Auxiliary Feedwater Pumps."
.
This IEB was issued because of reported events where hot water leaked into AFW systems and flashed to steam, disabling the AFW pumps.
TU Electric letter TXX-4937 dated August,1, 1986, stated that work instructions for keeping a log for monitoring conditions leading to steam binding had been developed and implemented.
Specifically, the procedure addressed equipment inspections, procedures for handling steam binding, and continued ~use of these methods'until Generic Issue 93 was resolved.
Licensee actions in response to this IEB were reviewed by the NRC and the IEB was closed by NRC Inspection Report 50-445/87-36; 50-446/87-27 dated February 10, 1988.
The licensee had developed and implemented operating procedures and log keeping instructions to address the subject steam binding.
IEB 83-03, " Check Valve Failures in Raw Water Cooling i
.
Systems of Diesel Generators."
This IEB was issued after numerous licensee event reports (LERs) documented check valve failures.
This IEB required operating licensees to review their plant pump and valve in-service test progr&m per Section.XI of ASME Subsection 1WV-3520 and modify, if necessary, to include check valves in cooling water flowpaths.
It also required licensees to develop test procedures and conduct tests to verify valve integrity.
The applicant's action in response to the IEB was reviewed and closed in NRC Inspection Report 50-445/88-12; 50-446/88-10 dated March 17, 1988.
The file included the IEB and several other documents.
The contact / inquiry record forms in the file documented evaluations of this issue.
TU Electric concluded that
,
Crane valve bodies were made from cast iron; however, stainless steel bodies were used at Comanche Peak.
In addition, some valves in the cooling water flow path were identified which were to be tested quarterly per Procedure OPT-207A, Revision 1.
(Service Water System Operability test).
This procedure was developed to ensure compliance with technical specification requirements relative to valve position verification, valve exercising requirements of ASME,Section XI, subsection lWV-3522, and flow, pressure, and vibration measurement during pump start-up.
IEN 80-16, " Shaft Seal Packing Causes Binding in Main
.
Steam Swing Disc Check and Isolation Valves." (Closed 10/07/88).
During disassembly of the main steam isolation valves at Indian Point 2, it was observed that all four reverse flow check valves were stuck at or near fully open.
l l
_ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _
_
. _ _ -.
.
__
- s
-
,
,
+-
During testing in the hot standby mode at the Trojan Nuclear Plant three of the four main steam isolation valves failed to close when manually actuated.
The cause of these events was excessively tight shaft packing.
Although CPSES uses globe valves powered by compressed nitrogen accumulators, the concern of overtightening the shaft packing still affects the main steam isolation valve.
This concern is addressed by MMI-818, Revision 0, "Rockwell MSIV Valve Repair," which has a caution to not exceed 75 foot-lbs. of torque for any reason when tightening the packing gland fasteners.
IEN 80-41, " Failure of Swing Check Valve in the Decay
.
Heat Removal' System at Davis-Besse Unit No.
1," (closed 9/30/87).
During leak rate testing, an RHR pressure isolation check. valve had excessive leakage.
On disassembly, the valve disk and arm were found lodged under the valve cover plate.
The valve is a swing check valve manufactured by Velan Valve Corporation.
IEN 81-30, "Velan Swing Check Valves," (Closed 1/2/87).
.
Upon disassembly of a Velan 6-inch swing check valve at Salem 2, it was found that the valve disk stud had broken and the valve disk was in the bottom of the valve body.
Cracks in the disk and bushings were found, along with a warped hinge pin and elongated hinge pin holes.
Similar check valves at Point Beach 1 were found stuck open due to interference.
CPSES has Velan swing check valves, but not the specific models which failed at the plants described in this report.
Although the specific failure mode is not applicable to CPSES, the general concerns of this report are addressed by SOER 86-03, " Check Valve Failures or Degradation."
The SOER recommended that a check valve maintenance and inspection program be established as discussed in Section 2.7.3.
IEN 81-35, " Check Valve Failures," (Closed 12/31/86).
.
Corrosion of the seat holddown devices caused loose internals in 3-inch 1500# Crane tilting disk check valves at Three Mile Island Unit 1.
Failure of the hinge lugs on a 3-inch Series 900 Mission check valve at Fort Calhoun Unit 1 allowed the valve disk to migrate to the steam generator.
Broken disk pins were found on 4-inch anchor darling swing checks at Arkansas Nuclear Unit 1.
I
- - - -
- _ - - _
_
_
--
_ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _
_ _-
._
',,
'
...
.
i Anchor Darling check valves are not used at CPSES.
All Crane check valves at CPSES are rated at 900# or lower and are of the." swing" disk type, not the 1500#
" tilting" disk type discussed in the report.
Although j
CPSES uses missing duo-check valves, none are the 3-inch size discussed in the report. CPSES has two Velan swing check valves in the Boron Thermal Regeneration System.
While the specific valve failures are not applicable to CPSES, the general concerns are addressed by SOER 86-03.
The applicant should insure that the log keeping instructions and operating procedures developed in response to IEB 86-01, " Steam Binding of Auxiliary Feedwater Pumps" as well as other procedures developed ~as a result of related
!
IE Bulletins discussed above are included in the check valve
. maintenance and inspection program being established at CPSES.
I 2.8 Safety significance of the Identified Check Valve Failures i
A review of the FSAR, TS, design basis documents, and other pertinent material was conducted to determine the safety significance of the identified equipment failures and anomaliss, listed as follows:
All 4 4-inch MDAFW supply line check valves fail to
.
prevent bachleakage.
All 4 4-inch TDAFW supply line check valves fail to
.
prevent backleakage.
Two or three of the three 3-inch AFW pump miniflow check
.
valves fail to prevent backleakage.
The feedwater isolation bypass valve allows backflow at
.
a differential pressure of greater than 50 psid (this is in accordance with valve specifications, but may not be l
adequate for this application).
'
The result of this review, which included consideration of feedwater line breaks, steam generator tube ruptures, AFW piping ruptures, and main steam system breaks, concluded preliminarily that one credible accident occurring in conjunction with the as-found equipment failures could result in the plant exceeding its design basis.
The postulated accident is a rupture of the MDAFW piping upstream of the supply line check valves resulting from an earthquake which also causes a loss of offsite power.
Both main feed pumps are lost as a result of the loss of offsite I
power and the AFW system is automatically started to ensure l
.that the steam generators can reliably remove decay heat
,
j from the reactor coolant system (RCS).
A single active i
l I
I
_ _ _
-. _ _. _ _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
c
.-
9.
..
N
L failure, the loss of one emergency diesel generator, is also postulated.
The MDAFW pump associated with the failed diesel: generator is lost and the other MDAFW pump is assumed l'
to discharge its entire flow to the line break.
This leaves only the TDAFW pump to supply the FSAR required flow of at least 215 gpm to at least two steam generators.
The TDAFW pump is rated at 860 gpm and isLordinarily sufficient by itself to provide adequate flow to the steam generators.
However in.this case, the MDAFW supply line check valves would fail to isolate;the upstream' pipe rupture, allowing the TDAFW pump to feed'the break.
It is doubtful, given the as-found condition of the MDAFW supply line check valves, that a significant amount of ficw from the TDAFW pump would reach the steam generators, and it is very likely that the design basis flow rate would not be achieved.
A line break between the MDAFW supply line check valve and the upstream orifice would exacerbate the accident, since the orifice would not be available to limit the directed flow.
If the design basis steam generator flow rate was not achieved, the decay heat entering the RCS would not be adequately removed in the steam generators.
The RCS could overheat and overpressurize, causing the power-operated relief valve (PORV) and/or safety valves to open and release radioactive steam to the atmosphere.
It is possible that this release
of airborne radiation could exceed the limits of 10 CFR Part 20.
The multiple failure of check valves could have gone undetected as the plant entered the operations phase.
Had this occurred, the plant would have been in a degraded condition and could have exceeded its design basis AFW flow as described above.
It is unlikely, however, that either of the two Reactor Safety Limits would be challenged by this hypothetical accident.
In the event of a loss of all AFW, with the steam generators boiling dry, the reactor coolant system could still be cooled by a procedure known as " feed and bleed."
The power operated relief valve (PORV) is opened to the atmosphere (or vents are opened to containment) and the blowdown is compensated by normal RCS charging.
This procedure should keep the ave 5^98 temperature (T-ave) below the approximate 660 F limit of the Reactor Core Safety Limit.
The pressure-relieving capacity of the PORV should keep RCS pressure below the 2735 psig limit of the RCS Pressure Safety Limit.
Another potential failure mode of the AFW system is steam binding of the AFW pumps caused by backleakage through the inoperable check valves.
Severe steam binding of AFW pumps could result in insufficient flow to the steam generators during emergency conditions.
Prior to the AFW backleakage events, the applicant had committed in response to I&E Bulletin 85-01 to monitor AFW piping for backleakage every
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ - _ - _ _ - _ _
-_
+'
.,
'
.
l
'
cight-hour shift.
An operator will touch the discharge piping of the AFW pumps to detect any increase in temperature and the AFW temperature indicators-will be monitored for any abnormal reading.
Considering the amount of backflow necessary to cause significant steam binding, the applicant's method of detection appears adequate.
It is noted, however, that this process of checking the temperature of the AFW discharge piping would not have detected the existence of numerous inoperable AFW system check valves in that with the absence of system flow to a low pressure point there would have been no thermodynamic migration.
2.9 potential for Re-occurence Discussions with the applicant indicate that in the future if engineering determines that a check valve has a safety-related function, there will be in-service and post-work functional testing which will include backleakage checks.
This procedure will include requirements for QA surveillance.
The periodic and post-work testing as described, should preclude the recurrence of similar incidents during plant operations.
The formal procedure has not been issued.
2.10 Radiological Consequences During the time frame of the event, there were no radiological consequences.
The plant was at normal operating temperature and normal operating pressure, but no fuel was installed in the Reactor Vessel.
Fuel load is currently scheduled by TU Electric for October 2, 1989.
3.0 Findings of Fact Historical Observations A similar Borg-Warner check valve failure was identified
.
in 1985 by Failure Analysis Report 85-001.
Three Borg-Warner check valves in the TDAFW supply lines
.
to the steam generators were found to be leaking on April 5, 1989.
Proper evaluation and resolution of the April 5, 1989,
.
events may have prevented the April 23 and May 5, 1989, events.
Borg-Warner MDAFW pump miniflow check valve 1AF-069
.
leakage was identified April 19, 1989.
_
_
_.
_ _ __-_--_-_
-
. _
__
'
'
..
4.
i Industry experience with faulty Borg-Waner check valves
.
was well documented-.
April 23, 1989 Event Misalignment of valves caused backflow of high
.
temperature water through the AFW piping.
Duration of event was approximately 20 minutes.
..
The event caused the paint on AFW piping to discolor,
.
blister and flake due to excessive heat.
No visible damage to piping during this event.
.
Temperature indicators off scale during this event.
.
Backleakage flow path was through the feedwater
.
isolation bypass valves.
These valves are designed to resist 50 psi backpressure and, when tested, met this design criteria.
May 5, 1989 Event Backflow of high temperature water through the AFW
.
piping due to improper valve alignment.
Duration was approximately two hours.
.
Intermittent pump operation during this event allowed
.
higher piping temperatures to extend further upstream in the AFW system than during the April 23, 1989 event.
One support visibly damaged by thermal expansion.
.
Formal documentation of second event was not timely.
.
Root Causes and Effects of the AFW Events Leak Testing performed subsequent to the April 23, 1989
.
event identified several stuck open Borg-Warner check valves which allowed reverse flow.
l The cause of the stuck open AFW check valves was i
.
determined to be improper adjustments (vertical l
elevation) of the bonnet-disc assembly combined with possible excessive axial play in the disc-arm assembly.
The improper vertical adjustment of the valve bonnet
.
resulted from inadequate installation instructions in the vendor's O&M manual.
_ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_
-
-
,-
-
77,
,
r'
- {
'
'
,
..
l
'
m.
.
['
-51
,
,
.-
A-l contributing.cause of the 3-inch miniflow check valve
inoperability may have been close proximity to an j
upstream. breakdown orifice.
j i
Applicant's' evaluation'of. piping-indicates'that several
.
areas were stressed beyond ASME code allowable..
l Inspection of penetrations revealed no concrete
.
distress.
4.0 Conclusions and Recommendations o4.1 Conclusions 4.1.1 The identification of three inoperable check valves in the TDAFW supply lines ~on April 5 should have been aggressively pursued. 'Instead, it was assigned _a normal work request priority.
This event reflects a lack of understanding of-the system' operability implications of failed components and a lack.of aggressiveness of Operations management to follow-up on the results of.the. system flush they had'
specifically scheduled to determine the scope of the original identified check valve problem.- This event was clearly a missed opportunity to_ discover the full extent of the check valve problem-in time.to prevent the April 23 and May 5 events from occurring.
4.1.2 The overall response by control room personnel to both events (falling steam generator-levels) was weak.(See paragraph 2.1.2).
4.1.3 Continuing.to test the AFW system after_the April-23, 1989 event with known multiple failures of check valves without.
taking_ appropriate precautions shows a potential lack'of respect for degraded plant conditions.
It also shows lack of communications between shifts.
4.1.4 It took an inordinately long period of time for operations to adequately identify the second May 5 ovent and to report it as such, esepecially considering that it had a greater magnitude of severity than the April 23 event.
The applicant's originally stated intent of including this event within the first PIR (110) appeared to be slow.
In fact, PIR-89-129 was only written at the NRC's AIT insistence.
4.1.5 The out-of-sequence operation of valves in the May 5 event, occurring 12 days after a fundamentally identical out-of-sequence valve operation in the April 23 event, reflects a significant weakness in the applicant's ability to prevent an operational error from recurring.
l a
i
'
_ _ _ _ _ _. _
_ _. _ _ _ _ _ _ _ _ _ _ _. _ _. _ _
_ _ _. _ _ _.
P!.c
'-
.
Lp.
!
4.1.6 Sending only one auxiliary operator near the end of shift to operate valves lAF-041 and 1AF-042 reflects a lack of understanding'in the control room regarding task manpower
. requirements.
'4.1.7 The AIT considers.the difficulty of operation of valves 1AF-041 and 1AF-054 to be a contributing cause to the April 23 and May 5 events, but of minor safety significance.
The AIT supports-the applicant's intent to'make these valves easier to operate.
4.1.8 adua evaluative process, which ultimately determined the root cause for the check valve failures appeared to be unnecessarily protracted in that it required almost six weeks from the inception of the AFW Task Team until the development of a definitive root cause and corrective action program.
This protracted process, although not directly related to any regulatory requirement, is an example of the applicant's lack of management aggressiveness in the resolution of a safety-significant issue.
This issue involved the multiple failures of passive components in a system intended to mitigate the consequences of_an accident.
For an NTOL plant, the applicant's response did not reflect the style of proactive Operations management philosophy normally associated with safe reactor plant operation.
The AIT notesLthat when the applicant's Project Management took charge of the Task Team on May 26, 1989, efforts were significantly more timely and reflected a stronger commitment to corrective action.
The applicant's Task Team went to the vendor Borg-Warner and made things happen.
This aggressive attitude by management brought to light the root cause and brought about a corrective action plan in a timely manner.
4.2 Recommendations 4.2.1 Create a minimum equipment list that would aid Operations personnel to make judgements regarding the effect of failed coraponents on system operability.
4.2.2 Assign system engineers the in-line task of reviewing all work requests related to a given system.
The engineer would evaluate the impact of all component failures in regard to system operability.
4.2.3 Provide training to control room personnel and supervisors regarding manpower requirements for certain types of plant evolutions.
4.2.4 Provide continued emphasis on training plant personnel to comply with procedures.
Steps are to be performed in sequence unless otherwise specifically approved.
. _.
- - - _ - _ - _ _ _
___-_
-
'O
-,
-53 4.2.5 Provide better communications between operations staff, especially during shift changes.
4.2.6 Provice a large and conspicuous plant status board in the control room, sufficient to provide significant " night order" information and to facilitate the transfer of information between shifts.
4.2.7 Initiate an immediate design revision to reparate the 3-inch miniflow check valves from their associated orifices.
The present configuration, if not corrected, lends itself to an exceptionally short lifespan for the check valves due to flow turbulence and valve tapping damage (see paragraph 2.3.3).
4.2.8 The AIT recommends that an Information Notice (IN) be issued in order that all licensees will be aware of necessary corrective action.
The AIT has drafted an IN and submitted Fame to the NRC Generic Communications Branch on June 16, 1989.
5.0 Persons Contacted K. Backus, Engineering Operations, TU Electric M. Bagale, Assistant Project Completion Manager, TU Electric R. Barr, Operations, QA Surveillance, TU Electric C. Bishop, Reg. Adm., TU Electric M. Blevins, NUC Operations Support, TU Electric H. Bruner, Senior Vice President, TU Electric W. Cahill, Executive Vice President, NEO, TU Electric J. Donahue, Operations Manager, TU Electric S. Ellis, Performance and Test Manager, TU Electric B. Garde, CASE W. Galdemond, Licensing, TU Electric B. Hardison, NSSS System Completion Manager, TU Electric T. Heatherly, Licensing Engineer, TU Electric J. Hicks, Chief Engineer, TU Electric C. Hogg, Chief Engineer, TU Electric T. Hope, Licensing, TU Electric J. Kelly, Manager of Plant Operations, TU Electric D. McAfee, QA, TU Electric C. Montgomery, Feedwater System Engineer, TU Electric J. Muffett, Manager of Engineering, CECO E. Ottney, CASE S.
Palmer,.NEA, TU Electric P. Pe11ette, Operations Technical Support, TU Electric C. Rau, Projects Completion Manager, TU Electric M. Samuel, Technical Interface A. Scott, Vice President, Nuclear Operations, TU Electric S. Shuman, Engineering Manager, CECO J. Smith, TU Electric R. Smith, Engineering Management, CECO
, - -
_-
!
- s.
.;*
L, 54'
R. Smith, Operations, TU Electric M,. Street, Projects Scheduling, TU Electric C. Terry, Projects, TU Electric M. Thero, Citizens for Sound Energy (CASE)
O. Thero, CASE
.
G. Trieste, Projects Manager, TU Electric J. Woods, Projects, TU Electric
,
t
_ _ _ - _ _ _ _ _ -. -. - _ _ _ _ _ _. _ _. -
__
-
_ _ _. _. - _. _ _. - _ - _ _ _ _ _
_---.-_-_--__--__-_.----__._______.--__.-_.-.--___..--.-__------_-_-a--
-
"
U
"
D
D
M
,A A,
-
e
- m N,
I_
M I
U.v.
=
v
.s y
-
y rt a
9._ vi
,x o ti
,a
.
i f
l l
g l
D I
O I
g I
v
"
v o
_ = ~
m e_
e e
P
'
e F
'
r
,v v
v
'
r L
I T
8~
A e
$
e-N N'
E
V tv E
p n
e
s
$
.
i
.
,
.
e y
.
y" v Ws-#i
.
ee-5g.g _r.
i 7,
3o e
.
o g
t
-
-
r n
.
v i
M r
a a
i
o'
.
,
.
.
.
.
.
,
d g.-_
&9_
M9
-
L Q
IR
-
P x
A
-
^
EH
-
)
T z>E ys-R
.' j -
OF
)
w *j-H o
.
-
-
r v
-
T e
(
a (
i A
6) *sG; 4 P
-
W
.
- h y
)
Q=.6.
e o
a -
-,.iM, _i i
.
, :,,.
.
.
O
,g 0 R' m-
, 3.,.
p
.
t s
_5
L
5
.
.
s s
,
,
W
,
F
-
,
,,
,,
,.
_
[w i
n.
n
,
,
-
'
e.
-
-
.
.
J.r
..
.
. ' <
..t
_
-
~
_
.
cl P
P V
"v cir
_
V
_ [A
,
_
.
wfD
_-
A
_
_
3A D
.
_
D
_
N H
^yT
_
E
_
C
-
_
T
_
_
A
_
G
>
_
S A
_-
.
N
_
_
R
_
E O C
_
D
_
_
T
_
N S
_
O
- (ig =
-
C
_
_
-
.> !
_
_
_
-
_
-
D D
D D
-
N X
A, uA
~
I a
,
_
I 8+=
x W
.
V S
f
,
T I
D D
_
,v D
I al v
_
-
8" P
_
"
_
m n
- !
-
r r
-
'
,.
o
r r
TR E
A P
a a5 R=
,-
,
T s
e N
r. v
.
,
m
$
S s-
-]g e_MI
__MI%
'a v=
y P
9
.
'
%^t
=.
M
-.
-
.
O
E
-
-
-
a u
g V
ta tt E
w
- e
'
'
-
.,
,-
c I
=
'
.
_
91 A
A9-9m,+
-
m
,
S YA M
E I
>
{>E X:
HT
.? $ -
R
>
O I-
9 F
w H
, }
T
~ss,h A
P
..
'
E_=L
,
g
.
m.
I W,.,J
~,.
W
.
S
.
,0
,M, p_
I I
&
.
IP'
S
.
O F
.
-
-
,
M'M
.
L
=-
F
F
'
y
'M M_
- .e
, 8 e
- Ot
'
. 3 I!
- t
%
-
r e
-
v P
)!.e V
ye F
.
f j
o A
o
] M
-
D E
I m-T E
A G
)
s A N R s
E O C
D g
1 l.
-
T (
N
,
O
- t,5 :s -
C
.{:
mO
"3
~,
D M
,
s
,
Q"
-
~
,A
.
i s)e.
s r
J, e
a-
=
i l
o I
o-I o
I o
,
,_~
e;
,
r
,
"
v
,
'
r TR
_
_
A
_
_
P A
E
,x s*:
,-
A g
&
-
,
,
T n
, m,-m-l m=~= ~_-
- .
.
- .
,
.
)
s
,.
.
N
.
,
t
.
r
.
r.
3s
'
o
.
.
E
-
M
-
_-
g.
V gu E
. :
,.
,
c :
..
89 AA g-9-
_
,- A 4A
_
,
Y
=
_ ~ -
_
A M
,,
E
)
,. i i E HT
- j.
R O
o * j_
-.
F
-
,
H
,
T 6:f; 6 A
P
.
I
.
M~. W.g, !,
.f
,
J.
,,r;FNr'f W,,f f
~
. _,,
a,
,
d W
,
O
,
,
LF
"I
_
-
l
2:f
.
- <
<
~ 3 7__
Fo^v e
g f^e t re v
rio
r
}g o
%o M
s (t E
-
%
TA E GRo S A N
m
.s n
E O c D
i N
i
-
T
<
S O
f l-C 1E
.
l
_ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ - _ _ _ _
..,..
.-
e Typical Borg-Warner Check Valve Assembly
1
Hlh ib
'
-
L Ki
-
F9 F9
Fil_
-
?
T
'
/
9,
.
_
,
-
//~
.
J
\\ad(y
,,-
-
,,_ d.,i
--
-
.
Het nut I4
Is.k washer
Retainer
Spacer
$
Seal g
Stud
8 Hex jam nut
0
Washer
.Q 10 Bt!!
11 Arm 12 Disk 13 Seat 14 Body 15 Neck 6.4 FIGURE 4
. -____ _________ _ _ _ _ -
. _. _ _ -
.
. _ _. _ _ _ _.
- _ _.
_ -.
_ _ _ _ _ _ - _ _ _ _ - _ _ _.
_
,c
- *
.o.
CAD tiodel' of Valve 1 AF-106 ( As Found Condition)
IIlff!1 ff'l ff'l rOffE hi nQ
/
hl f
E
!!
!!
!! R/
/?
8/
/
/
/
/
f
Q(
Nx X
m
,
,,
~s
'
El'9
%Vf~
E ?
/
N N
x x
\\
\\
\\
\\
\\
\\
>
.
4.7* SEAT ANGLE EXISTING DESIGN LAYOUT W/ WELD BEAD 6.5 FIGURE 5
_ _ _ - _ _ -
_ __ _ _____ A
y:
..
i
"6.6-FIGURE 6 Matrix of Unit 1 Borg-Warner Check Valves (As Found Conditions)
s
%
T W
.
J AV-oh9
'3 ISO IAr-nid
lEn IAV-024
l AF-osz B
15 0 1AV-n3B
900 I AF- 045
900 X
OPd N
O T */
H d
.223
'
'
I AV-051
900 IAF-os7
9eo X
oPM Y
o
Y N
.in iAV-nLS
90o
"
'
I AF-o69
9oo x
ctsk Y
o B.3 N
.078 IAV-075
900 V
oPd a
N
&
.l65 IAF-018
.4 908 Y
094 N
n S
Y ll
.160
- AF- 003
9oo r
oP#
Y o
N N
.204 1AF'686
9so Y
09/
V o
n Y
N
.193
__
lAF- 013
9sa x
OP4 Y
o I2 M
N IAF 048 4'
900 X
OPd
6 Y
Y
.141 IAF-fol Y
foo Y
OPd Y
o
N W
. 2Io IAV-104
~4 900 X
OPA N
o
N N
.419 lhF-lL 7
150 IFid-19I t.'
4 oo IVJ-lit.
L
&co
- G-193
Goo IFw-r44-6 doo IFJ-195
&
Goo X
CLsb 1 H -146
Goo X
CLsb IFn/-19 7
6oo i
cLsp lFA-I98-6 600 t
CLsb Y
. /50
$
Y 315 iW-199
6 00 Y
CL5D IFw-zu
foo X
cosb IN-iol
&co X
CLbD l'
600 X
CL6D IM514t.
10o X
014l
'/
o 3.2 Y
V
. li'l l HS-I*LS
90s x
oPd V
o.
3. 6 Y
H
- l2'l
.
_LcnpqB x
'
~
~
ic1"09V X
- _ _ - _ - _ _ - _ - _.
__ _
_
--
__.
4 s
_h c.
d
^QW 7.0 TABLE OF ACRONYMS
'AFW Auxiliary Feedwater AFWP Auxiliary Feedwater Pump AIT Augmented Inspection Team AO
' Auxiliary Operator ASME American Society of Mechanical Engineering-
'
CAD Computer Aided Design CECO Consolidated Engineering and Construction Organization CMTR Certified Material Test Report CPRT Comanche Peak Response Team
'CPSES Comanche Peak Steam Electric Station CST Condensate Storage Tank CVCS Chemical Volume and Control System DCA Design Change Authorization HDM Design Modification EPRI Electric Power Research Institute FA Failure Analysis Report FSAR Final Safety Analysis Report FW Feedwater HFT Hot Functional Test I&C Instrumentation and Control IEE-Information and Enforcement Bulletins l
I&E Inspection and Enforcement INPO Institute of Nuclear Power Operations
'
ISAP Issue-Specific Action Plan LER Licensee Event Reports MDAFW Motor Driven Auxiliary Feedwater MDAFWP Motor Driven Auxiliary Feedwater Pump MMI Mechanical Maintenance Manual MSM Maintenance Section-Mechanical Manual NCR Nonconformance Report NPRDS Nuclear Plan Reliability Data System NRR Nuclear Reactor Regulation NSSS Nuclear Steam Supply System OD Outside Diameter O&M Operation and Maintenance Manual
~ PIR.
Plant Identification Report PORV Power Operated Relief Valve l
PR Problem Report QA Quality Assurance QC Quality Control RCS Feactor Coolant System RHR Residual Heat Removal
'
RO Reactor Operator RT Radiograph Testing
,
'
SIF Stress Intensification Factor i
!
l
_-
_ _ _ _ _ _ _ _ _ _ _ _.
_ - _ _.
-._
_
. _ _ _. -_
. _ _.
_
- _ _ - _ _ _ - _ _ - _ _ - _ _ -
_ _ _ _ _ _ _. - - _ _ _ _ _
3;~ p,
3.__ g gj.,3
'
-
..
~
--
=
.- - }'
..
,
ggf.
J,
.l 1),
-
f
' '.
'
1
>
,
.
.. ;
.
,
34.
pw-
.yy-l
,
, >
.4
-
,,
,:r
<
,
,
}
s-
,
,
g.-
.
.
...
,
'
c.
sp
.
'
,
'3 a f;' Q
.
~
' '
t
'
i i
,.-
,
- <z-
-;
';
-
... i
..
- 4
-
,cj.
,
J SOER
Significant.Op'eriting' Experience. Report;
'
,
~'StoneLand Webster. Engineering.. Corporation.
LSWEC Qj
,ITDAFW. Turbine.. Driven Auxiliary Feedwater'
-
,, " " "
TDAFWP ; Turbine, Driven: Auxiliary.Feedwater' Pump
- TDR
' Test: Deficiency; Report;.
,
'
.TS'
~ Technical-Specifications
'
-
TU -
' Texas.. Utilities' Electric ~ Company.(Formerly.TUGCO)'
'
/
' Ultrasonic-: Testing-
.
.i?
-
if]
'
.;
.
.
'
,.
f '_
L 'l _,
,y
'
'i
',:
g, L
%
- '
M s
?
t f
.L..
_ _ _ _ _ -. - -
__.__,_._._....-,e
- -
~ _ _. - _ _ -
-_-_ -.._ -___ _.--__ _ _-___.
-.~-.___-_.-____--__e.-_
- -