IR 05000397/1989038
| ML17285A972 | |
| Person / Time | |
|---|---|
| Site: | Columbia  |
| Issue date: | 01/05/1990 |
| From: | Johnson P NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| To: | |
| Shared Package | |
| ML17285A970 | List: |
| References | |
| 50-397-89-38, NUDOCS 9001300228 | |
| Download: ML17285A972 (15) | |
Text
U.
S.
NUCLEAR REGULATORY COMMISSION
REGION V
Report No:
Docket No:
Licensee:
Facility Name:
Inspection at:
50-397/89-38 50-397 Mashington Public Power Supply System P. O.,Box 968 Richland, WA 99352 Mashington Nuclear Project No: 2 (MNP-2)
MNP-2 Site near Richland, Mashington Inspection Conducted:
November 27 - December 19, 1989 Inspectors:
C. J.
Bosted, Senior Resident Inspector Approved by:
R.
.
S rens n, Resident Inspector
/
nson, ie Reacts Projects Section
( Q~
ate cygne Summary:
Ins ection on November 27 - December
1989 (50-397/89-38)
Areas Ins ected:
Special inspection by the resident inspectors on the oosening an torquing of safety related motor operator mounting bolts in t residual heat removal system.
During this inspection, Inspection Procedure 62703, 71707, and 93702. were utilized.
Safet issues mana ement S stem (SINS Items:
None.
Res ults:
General Conclusions and S ecific Findin s
Weaknesses Identified During a routine plant tour, the NRC inspector identified a capscrew associ ated with a safety related valve, which had apparently fallen i
of its valve.
In following up on this issue, the inspector identifi three maintenance-related weaknesses, as summarized belo Incorrect Tor ue Selection and Weak Administrative Controls The inspection found that weak maintenance administrative controls resulted in incorr'ect torques being repeatedly used with regard to safety related motor operated valve fasteners.
The torques used on safety related fasteners in the plant have been se')ected by craft personnel from several different procedures, based on application of the fastener.
Varying applications, types of fastener material, and sizes of fasteners have required that a large number of torques be specified.
When maintenance work requests were prepared, they directed the maintenance technici.an to a table in a procedure and expected the technician to select the proper torque for the fastener material and size.
This practice allowed several opportunities for error.
-If the procedure was correct, the craft worker had to determine the type of material and the size of the fastener.
He was also expected to evaluate whether the procedure that he was directed to use was correct.
In several cases during this inspection, it was determined that errors had been made in the selection of procedures (applications)
and size of fastener.
The post-maintenance review process did not appear to have corrected these problems in al').
cases.
II Ineffective Corrective Action for Loosenin of Fasteners.
The inspection found that the licensee's corrective actions to address loosening of fasteners has been limited.
Since startup, the plant staff has been aware of an industry problem involving threaded fasteners which have become loose.
Industry correspondence and Supply System experience have prompted several operations experience reports to address these conditions.
Improved administrative controls and equipment controls have been recommended.
An evaluation of the mechanical aspects of the concern did not recommend the use of any type of anti-rotational devices.
It appears that little has been done to provide a long-term solution to this problem at WNP-2.
Preventive Naintenance Pro ram A solution to the potential loosening of mechanical fasteners was to apply administrative controls in the form of a preventive maintenance program to inspect the bolts for tightness.
However, this program was largely ineffective, since it did not require verification of proper torques.
The inspection interval for these preventive maintenance checks was specified at eighteen months, although no engineering evaluation of this methodology appears to have been made.
The program does not appear to have reflected good engineering judgement.
If the fasteners. were only verified finger-tight, given the amount of vibration and the number of operations'ome of the valves experience,.the fasteners should not have been expected to stay finger-tight for an extended perio Summar of Violations and Deviations:
Th ee apparent violations'ere identified, as discussed in paragraph 7:
{1) application of incorrect torques on four occasion
, { )
take effective corrective actions in resolving bolt loosening problems, and (3) failure to initiate a Nonconformance Report (NCR) for a missing J
1988 Item (3) is considered to be a Level V violation and in view of program improvements already implemented by the
h licensee is not included in the Notice of Violation which accompanies this report.
Si nificant Safet Hatters:
Mhile no system or component inoperability appears o
ave resu e
rom rom the violations identified above, three of eig't motor operator mounting capscrews were found loose on suppression pool cooling lsolatlon valves Inope y
d d
~
~
~
erabilit of either or both of these valves could have affected containment integrity or re uce post-accident containment cooling capability.
0 en Items Summar
were o ened -- the three violations mentioned above and one unresolved item in paragraph
.
ea
>ng wi 5 b d
ith the performance of the scheduled maintenance system in checking the fasteners for tightness,
DETAILS 1.
Persons Contacted
- C. McGilton, Manager, Operational Assurance Programs
"C.
Powers, Plant Manager
- P. Harness, Manager, Mechanical Systems
"J. 8aker, Assistant Plant Manager
~J.
Harmon, Maintenance Manager D.
Kobus, equality Assurance Manager
Safety Assurance D.
Hoon, Acting Manager, Nuc ear a
- C. Edwards, equality Control Manager M. Shaeffer, Assistant Operations Manager
- R. Mebring, Assistant Maintenance Manager The inspectors a so in e
t rviewed,various maintenance, assurance, and management personne
.
- Attended the Exit Neeting on Decemberer
1989.
engineering, qual it~
2.
Hack round
, th
'
ector discoverer
.
on October 31, 1989, e insp t
f id
h t o
(RH eared to be the missing capscrew was found om the motor opera or o 24A Another capscrew appeared to have in an observable gap between the capscr f t th L'o head and the valve yoke ange.
es p
at he would correct the situation by d
'
th t h i 1 t ff
'n to resolve any dditional concern.
mec
'
nsn to reso ve hanical supervisor in the morning to reso ve d t
.m.
on the same day, the inspector discovere ve RHR-II-248.
One capscrew had backed out tdth h df th pos nt where a one-to two-g the valve yoke f'lange.
Another capscrew was oun The inspector informed maintenance managemen an p
shortly thereafter.
" " train test return line isolation valve to th ll h t b t is opened duri g syst f
supp ess~on pool.
It >s norma y s d
'
sis ac 'd t thi
thermal energy from the containment.
fo th o
ool ool'
the same functions or 1 so a contas nment s sol at> on ll o
t d d
'1 (o power operation, these valves are norma y ope for suppression pool coolin. 'e uence of Events The sequence of significant findings associated with this inspection was-as follows:
TTime October
2:00 a.m.
8:00 a.m.
3:00 p.m.
November
Mid Shift Event Inspector discovered capscrew missing from RHR-V-24A and informed shift manager, who directed the shift support supervisor (SSS) to reinstall the missing capscrew.
Management review committee directed that a maintenance work request (MMR) be written to check the torque of capscrews for nine valves that operate similar to RHR-V-24A (significant vibration caused by valve operation under dynamic flow conditions).
Inspector discovered RHR-Y-24B to have one capscrew backed out one to two 'inches, with another capscrew loose.
He informed plant and maintenance management.
Shift manager was informed, and directed the SSS to check the capscrews
~
The SSS tight'ened the capscrews finger-tight.
Plant maintenance personnel checked the torque on eight of nine motor operators installed on similarly configured valves.
RHR-V-24A and B were at this time torqued to 160 FT-LBS.
Dayshift November
The plant staff discovered that the torque applied earlier to RHR-V-24A was wrong.
The MMR was amended to retorque the capscrews.
The inspector reviewed the maintenance work history for RHR-V-24A and B, and determined that a similar incident had occurred. on RHR-V-24B in June 1988.
At that time, plant followup was minimal, with the MMR stating that the capscrew had been torqued to 185 FT-LBS.
A nonconformance report (NCR) was not initiated to document the problem.
The inspector determined that another incident of missing capscrews on RHR-V-24A was described in startup deficiency report SDR M 9011, dated June 1983.
The inspector requested that the events assessment group provide him with any avai labl'e industry information regarding similar experiences.
The inspector also asked the licensee whether valves RHR-V-24A and B were operable, based on the as-found data.
The Nuclear Safety Assurance Group (NSAG) provided the inspector copies of pertinent NRC and industry documents (identified in paragraph 4.a).
These documents described
.
several industry experiences in which bolts or capscrews had backed out of motor operators and in some cases had caused the valve to become inoperable.
November
During review of completed MWRs, the inspector determined that RHR-V-24A and 27A were retorqued improperly on
~ November 1 (see paragraph 5.'a).
Maintenance issued a
new MWR to torque RHR-V-24A and 27A to the correct torques.
November
Plant management discussed the engineering evaluation with the inspector.
The inspector pointed out several incorrect assumptions in the evaluation (see paragraph 6).
November
Engineering evaluation of as-found data for valves RHR-V-24A and B concluded that both valves were operable.
4.
Historical Findin s
Following the discovery of this problem, the inspector reviewed the maintenance history, industry events, and operating history for these valves.
He discovered that these valves and others in a similar system configuration had a history of loosening bolts or capscrews.
a.
Notifications of Similar Events at Other Plants The licensee has received several industry notifications during the last six years dealing with loosening of fasteners on motor operated valves (MOVs).
These included an NRC Information Notice, IN 83-70; an INPO event report, SER 1-84; and a General Electric service information letter, SIL 405.
Two of these notifications cited specific examples almost identical to the type of bolt loosening that occurred on RHR-V-24A and 24B.
NRC Information Notice (IN) 83-70, issued in October of 1983, described two different instances wherein yoke-to-bonnet bolts had vibrated loose, one of them in an RHR suppression pool return valve.
INPO SER 1-84, issued in January of 1984, described several bolting failure events on MOVs due to vibration.
Three of them involved loose, sheared, or backed-out motor operator mounting bolts.
In addition, the General Electric SIL, dated February 24, 1984, discussed conditions in which fai lures had occurred when flow-induced vibration caused threaded fasteners to become =loose.
INPO SER 1-84 recommended periodic inspection and the use of some type of mechanical restraint for the capscrews to prevent vibra-tion-induced relaxation of the bolt torque, with resultant loss of bolt captur b.
The licensee's response to these industry operating experiences was to (1) institute administrative controls in the form of periodic t'aintenance checks for bolt tightness on certain d ( )
Bulletins, one in July 1986 and another in arc the awareness of cognizant plant personnel to the potentia or vibration-induced loosening of valve components.
\\
Previous Similar Ex eriences at WP-2 ent occasions, loose or missing capscrews were found on-V-24A and B.
These were documented residual heat removal valves RHR- -
an in startup deficiency report (SDR)
M 9011 and maintenance wor request MMR AT 5356.
SDR M 9011 documented two capscrews found missing from from the RHR-V-24A motor operator-to-yoke 'flange in June 1983.
They were replaced; since no torque value was given in the SDR, the mechanic torque 213 FT-LBS per PPM 10.2.10.
A review of PPM 10.2.10, them to 2 (the revision in effect
"Fastener Torque and Tensioning
, Revision 2,
a e
ime t th t'
showed that 213 FT-LBS was the value to be use or studs on piping flanges.
The correct value, from Table I of the procedure, was 640 FT-LBS for 7/8" socket head capscrews.
Earl in 1985 the licensee identified potential loosening of valve ctor recirculation control valve RRC-V-23B, as documented in NSAG memo
"OERS 80034D an recommended among other actions, that some type of mechanica capture device be used on valves subject t
prevent vibration-induced loosening of the capscr ews and o
s.
These recommendations were rejected by the plant, in a
e p an considered the preventative maintenance program to be adequa e
o address the problem.
In June 1988, loose capscrews were found on RHR-V-24B.
A mainten-ance work request was issued to retorque the motor operator capscrews.
Plant management was not made aware of the p
loose ca screws found i June 1988, and additional corrective actions were not implemen e
that time.
Plant administrative procedure PPM 1.3. 12, "Plant Problems",
required that "a physical characteristic of a componen which does not conform to the requirements of the design document be repor te y a p an d b l t deficiency report/non conformance report (PDR/NCR).
No PDR/NCR was generated at the time to document cument the discovery that t ese capscre h t th apscrews were loose.
Mithout the PDR/NCR n
tl process, a
i iona mana dd t
gement tools were not implemented, and t a
h 7
.
root cause was not corrected (refer to paragrap
).
5.
Additional Findin s After the inspector discovered the capsc
'
g a screw missing from RHR-V-24A, i capscrews loose on RHR-V-24B, the Shift Manager directe e
the loose ones.=
He initiated (PER) 289-0846 documenting the insp
'n ca screw and tighten the oo
.=
'
q d'
d that a maintenance wo t resolution.
s rk M
gement Review Committee,
sn 'MRC) iree e
he oke ana ten and implemen e
o was related test return valves.
Thss w
h ning of October 31 for eight of e
d 24B.
H'
P Co S r y MMR b t ot o k d that evening HPCS-V-23 was included on the MM,
u o direct the craft workers in the iew of the method used to iree ses were inherent in the A review h
MMR evealed that weaknes f the c p rew nd
l f
t
preventive maintenance to period1ca y v a.
Tor ue Selection n
the roper torque or f r these capscrews and ied on the maintenance eng>neer t0
'f'
th o
t l
t'o fo th t
t Th include the procedure number and main e
ize and material and then select th e Th 'o ot d th t th's t
t b
l'dt ft methodology has allowed improper torques related valves on several occasions:
R-V-24B.
MHR a screws were found on RHR--
ue the motor operator capscrews.
AT 5356 was issued to retorque e
"M 'enance and Repair of-LBS s ecified in e
rs" which was for Limitorque moto of PPM 10.2.44, ain r
M't o
ho
"L'to
tor R
ov
incorrect.
ann d th t 405 FT-LBS of tor u
M 10,2.42, hami and Installation," which specific a
applied.
'
ector's findings, the licensee electe cted valves.
On October
198 k
t dt th t
Th
w t
d o T-LBS.
On November 1, the licensee discovei 60 FT-LBS was incorrect The torque va'44 d
to fo ho en.from P
ver bolts for a cer operator to yoke/bonnet bolting.
'"'""'"w"'/.'- --.-.-" "hsa correct torque for the 7 8 )nc cap caation was 405 FT-LBS.
S tl 1 nl 1 a et d an amendment to $8 A
The l>censee subsequen y
on November 1 to check/retorque e ca
nine valves to the torques obtained from Attachment C of PPN 10.2.42.
Five days later, on November 6, during a review of the completed NWR, the inspector observed that the torques recorded in the NWR for redundant valves RHR-V-24A and 24B and for RHR-V-27A and 27B were different.
Each of these pairs of valves is identical, so that the torques applied should have been identical also.
However, while valve Y-24A was torqued to 255 FT-LBS, V-24B was torqued to 405 FT-LBS; and V-27A was torqued to 30 FT-LBS while V-27B was torqued to 145 FT-LBS.
Yalve Y-27A and V-27B, the suppression pool spray valves, have smaller capscr ews with a correct torque of 145 FT-LBS." The inspector brought these observations to the attention of maintenance management, who were unaware of the discrepancies.
The MWR instructions directed the craftsman to determine the torque from Attachment C to PPM 10.2.42.
Maintenance management had been relying on the skill of craft personnel to determine the right capscrew size associated with these valves and to apply the corresponding torques from the procedure.
Different craft personnel determined the capscrews to be different sizes -- based on observation, not on measurement-and obtained different torques from the table.
b.
Scheduled Maintenance S stem A review of the preventive maintenance program was also conducted to determine what administrative controls existed to compensate for the lack of capture devices on the capscrews.
The schedule maintenance system (SMS) records were examined.
~ Two SNS procedures were in existence that checked for loose motor operator (NO) to yoke bolts, These procedures were contained on computer generated cards which direct craft personnel in the per-formance of the task.
Neither SNS computer card specified checking for proper torquing, only that the bolts were "tight".
Statements by mechanics who had actually performed the SNS tasks indicated that they were checking them finger tight.
One SMS card, task ID 03, was apparently developed sometime early in the operational phase of WP-2 at an indeterminate date.
It was performed for both V-24A and V-24B in August 1985, February 1987, and in May 1989.
However, task ID 03 was updated in July 1988, and in the process the guidance for checking bolt tightness was deleted.
The other SMS card, task ID 01, was initiated in early 1985 as action taken in response to industry-identified information which involved failures of threaded fastener devices.
As a precautionary measure, NSAG recommended that an SNS card be initiated, to include inspection for loose yoke-operator bolts every 18 months.
Task ID 01 did include this guidance; however, this SMS task was not initially completed until May 1988 for V-24A, and June l988 for
V-24B. It was again completed on October 1, 198 989 for V-24A and on October 14, 1989 for V-24B.
) de artment had identified the deletion of The quality assurance (g )
epar men the bolt tightness requirement from tas re ort 2-88-208; maintenance had committed to reinstating is uidance into the SNS card, either directly or by reference to a e.
The chose to accomplish this by referencing PPM PPH 10.2 69 did not include the guidance to check
'g
'
1989 when this weakness was for bolt tightness until Novem er ed b
gA.
The gA Manager wrote a memo to the Maintenan
, 1989 to b t k to
'
t PPN 10. 2. 69.
The performance of this SMS task na er on November 6,
1989 reques ing ac io this guidance into in 1989 did not include checking the MO to yoke cap ca screws for tightness.
The latest performances of this SNS w r y
were onl weeks prior to the ntif in the roblem with loosening an ac ing ou St t t
were gathered from the the SNS who stated that they had screws on the valves.
a emen s
i d'dal hoperfor 'd
kP
't od it as written.
A review of a ia ion o
so indicated that the SHS cards were performed by d tion the June 1988 performance these individuals as stated.
In ad i ion, e
of this SNS on Y-24B found-that the capscrew was missing, w ic indicated that the tasks were being performed.
The scheduled maintenance system periodi p
c erformance checks for bolt tightness appear ed not to meet the intended obJec ives o
checks.
The inspector discussed the manner of checking these a t en ineer to determine if this met o
fasteners with a cogniz n
g would discover bolts that were coming oose.
e e
the opinion that checking the fasteners finger tig t was no He stated that if the bolts were finger tig
,
ey c a ver short time.
Based on review of available o nizant en ineering personnel, the documents and discussions with cognizan q
inspector concluded that no engineering review of e
performed.
ews on RHR
'ff't-to determine the last time the capscrew valves RHR-V-24A and 24B were proper y orque
.
ins r c tructions were not detailed enough to provide guidance on ow they should have been checked g
for ti htness.
The inspector iven to the craft personnel performing these at since no standards were speci ie
, insu e ke t of the findings of these checks nd no d
were kept o
- SMS tasks ID 01 and ID 03 were not sufficien o
from becoming 1'oose.
This appeared to be a weakness a
e licensee should address.
o ortunities, via operating experier th'bl d d
'
information and SNS.cards, to address this pro em an e it.
The next scheduled performance of S
as valves RHR-V-24A and wou d 24B ld have been November 1990; however,
the requirement to check these capscrews for tightness had been deleted.
Consequently, the next scheduled SMS that would have checked the capscrews for tightness was task ID Ol, scheduled for April 1991.
The preventive maintenance SMS practice of checking the fasteners finger tight without specifying acceptance values will be inspected further and is considered unresolved (paragraph 8).
(Unresolved Item 397/89-38-01)
0 erabilit Determination On November 1, the inspector asked Engineering for the results of their evaluation associated with the loose and missing bolts and their determination whether the motor operators and valves were operable.
The licensee subsequently performed an operability determination which was initially based on the number of loose/missing capscrews found during the initial maintenance check performed under MWR 2939.
The "as-found data" for valve V-24A, on page 3 of 3 of MWR AS 2939, was listed as having three capscrews loose; one was one full turn and two were one-half turn loose.
Valve V-24B was recorded as having four capscrews loose, each one-quarter turn loose.
This information was different from the inspector's findings.
The inspector had requested that the cognizant Generation Engineering engineer be apprised of the as-found data so that he could make an accurate operability assessment for RHR-V-24A and 24B.
During the November
MWR review, the inspector observed that the as-found data recorded on the MWR were different from his observations.
The inspector had found one capscrew missing and one with a gap between the capscrew head and the valve flange on V-24A, and one capscrew backed out one to two inches and another loose on V-24B.
These were not recorded on the MWR, and the data sent to engineering reflected the less severe case wherein one capscrew was loose one turn and two were 1/2 turn loose.
During a meeting with plant management, the engineering evaluation of valve operability was discussed to insure that the fasteners the inspector found missing or loose were the same as those the licensee's MWR had indicated being loose.
In addition to concern about the difference between his findings and the as-found conditions recorded on the MWR, the inspector was concerned about the quality of the evaluation performed by engineering.
During the discussion, several of the assumptions used in the evaluation
,
were questioned by the inspector.
The engineer who performed the evaluation had assumed that the valves were not normally used and had only a few hundred cycles of operation,. and that when used the valves were not opened or closed while the pump was in operation.
The inspector stated that he had other information which indicated that the valves had several thousand cycles and that the valves were normally cycled with the pump in operation, with maximum differential pressure across the valve Later, licensee interviews with the SSS concluded that the inspector's findings had been included in the data that were sent to engineering.
The cognizant engineer subsequently determined that three capscrews with a gap of five mi ls or greater between the head and the yoke flange would be acceptable.
Four or more capscrews with a gap greater than 5 mils, depending on their location, could render the valve inoperable.
It was determined that, in this case, three capscrews were found more than five mils backed out on each valve, which still maintained the valves operable.
Engineering recalculated the evaluation and concluded that with maximum differential pressure across the valve and with several thousand cycles, the valves were still operable with less than four capscrews greater than five mils loose.
No violations or deviations were identified.
7.
Conclusions Evaluation of the licensee's.actions, as documented in the above inspection findings, indicated the following apparent violations:
a.
A lication of Incorrect Tor ues (Para ra h 5.a)
The licensee's failure to apply co} rect torques'o the capscrews on safety-related valves is considered to be a violation of Technical Specification 6.8. 1 and applicable licensee procedures recommended by Regulatory Guide 1.33.
Incorrect torques were applied on four occasions, as follows {Enforcement Item 397/89-38-02):
The capscrews on safety-related valve RHR V-24B were incorrectly torqued to 160-199 FT-LBS in June 1988 after loose capscrews were found.
- The capscrews on safety-related valves RHR-V-24A and V-24B were incorrectly torqued to 160 FT-LBS on October 31, 1989.
This is included in the Notice of Yiolation in this instance, even though the licensee ide'ntified it, because effective corrective actions were not taken following either this or, the previous occurrence.
The capscrews on safety-related valve RHR-Y-24A were incor-rectly retorqued to to 255 FT-LBS on November 1, 1989, after it was determined by the licensee that the torques applied on the previous day were incorrect.
The capscrews on safety-related valve RHR-V-27A were incorrectly torqued to 30 FT-LBS on November 1, 1989.
b.
Ineffective Corrective Actions to Prevent Recurrence of Loose asteners ara ra s
an
.a Through the combination of industry operating experience information and actual previous experience with loosening fasteners at MNP-2,
8.
the licensee had ample opportunity to determine the cause for the loosening, of the motor operator-to-yoke capscrews and to deal effectively with the problem to preclude recurrence.
However, as of October 31, 1989, capscrews were still loosening and backing out of RHR-V-24A and 24B,'which could'ave caused these valves to become damaged or inoperable, in a manner similar to that identified in IN 83-70.
The licensee's failure to determine the cause of this condition adverse to quality, and to assure that corrective action was taken to preclude its recurrence, is considered to be a
violation of 10 CFR 50, Appendix B, Criterion XVI (Enforcement Item 397/89-38-03).
ll c.
Failure to Initiate a Nonconformance Re ort (Para ra h 4.b).
The licensee's failure to initiate a PDR/NCR for the loose capscrews on RHR-V-24B in 1988 is considered to be a Severity Level V
violation of Administrative Procedure PPM 1.3. 12 and
CFR 50, Appendix B, Criteria V and XVI (Enforcement Item 397/89-38-04).
In that the licensee has since instituted significant improvements to the PDR/NCR process, and no further corrective actions are required for this violation, it is not included in the Notice of Violation transmitted with this inspection report.
No response to this violation is required, and this issue is considered closed.
Unresolved Items Unresolved items are matters about which more information is required to determine whether they are acceptable items, violations or deviations.
An unresolved item addressed during this inspection is discussed in paragraph 5.b of this report.
9.
Exit Heetin (30703)-
The inspectors met with licensee management representatives to discuss inspection status at an exit meeting conducted with the indicated per'sonnel (refer to paragraph 1)
on December 7,
1989.
The scope of the inspection and the inspector's findings, as noted in this report, were discussed and acknowledged by the licensee representatives.
The licensee did not identify as proprietary any of the information reviewed by or-discussed. with the inspector during the inspection.