05000333/LER-1989-015, :on 890918,two Remote Manually Operated Diaphragm Air Operated Isolation Valves Failed Closing Time Acceptance Criteria.Caused by Buildup of Iron Oxide Sludge. Valves Disassembled & Cleaned
| ML19327C135 | |
| Person / Time | |
|---|---|
| Site: | FitzPatrick  |
| Issue date: | 11/07/1989 |
| From: | Fish H POWER AUTHORITY OF THE STATE OF NEW YORK (NEW YORK |
| To: | |
| Shared Package | |
| ML19327C134 | List:
|
| References | |
| LER-89-015, LER-89-15, NUDOCS 8911200238 | |
| Download: ML19327C135 (7) | |
| Event date: | |
|---|---|
| Report date: | |
| Reporting criterion: | 10 CFR 50.73(a)(2)(ii) |
| 3331989015R00 - NRC Website | |
text
- _ _ - - - -
me yt va masas eenwes, enemmen -
mensven me at UC88000E tysstT REPORT (LER) oes u n men ni sesan mean m
]
JAGil A. FITIPAffitt NOCLEAR POWII PL ANT eisieleielain1 i lop l g g7 ]
"a' " Nine Air Operated Cont.ainment 1 solation Valves 1.xhibit Operational Deficiencies Due l to Packinc Problems & 1ron Bull,d,-Up in Reactor Butiding Closed 1.000 Cooline Syst em l
ovem este e ime meene e asem enn m evuen aan w nee esve6ves e j
en, vaan vaan
,,gt mont.
m, vena e.emtv mas seemst mesena r
j i
elate 10l0l l l 1 0l9 1l8 8
9 8l9 0l1l5 0l1 1l1 0l7 8l9 e ssge le g e n i i no ases, e amenee swmanne te ne naammen,e se w een er sm.*== -=== = sw====w nu amam m
n,ewom mnm seses masone
===m n,ewsw tmw l
655
~
i n
[
Doo emannim
[
[ ggg mm.m
==
$m%%@%dkwaw@c
- 2. ",' *c*,m*i
, m,eme,'=
6 eeneae,.e. note.no 1
senes re6eemene m anen Hamilton C. Fish 13 11115 314191-160 1 1
eew6 eve ens tuo een amen eessemeer en.6une semeses a vue messor ne
- @*'6
- 17,"J.'#*
, @ gg e.
cause
everem emme.awr W
g,,ga g cause erstou eewoment 1,cL I I SgV i Hg0;3 3 5 Y
g
, i i
- , i 5
2 1
1 1 l l l l l
l t i I l l swetmentat
,e enen.
mowr=
=,
,p.
s'4em vow.-
smerres mesese= =res O
,,o ano mas,ru== =ese a. _ _ _.
= :
ne.
i EIIS Codes are in []
l Revised Report - Previous Report Sent October 18, 1989 On September 18, 1989 during a scheduled outage and performance of a scheduled ASME,Section XI In-Service Test, two of nine remote manually oaerated diaahragm air operated containment isolation valves [ISV) on t'ae reactor building closed loop cooling water system [CC) failed the acceptance criteria for valve closing time.
One valve would not close except by manual operation.
Another valve closing time exceeded the criteria by 0.4 seconds or 3.6 aercent.
Subsequent outage operations revealed common problems with t1e seven valves which initially passed the closing time test.
Valves appeared to be binding due principally to the buildup of iron oxide sludge in the valve operating internals although there is some indication that the original packing may have contributed to the problem.
Corrective action
Disassembled and cleaned all nine valves and changed packing from seven ring Grafoil to live loade"d five ring type on five valves.
Three other valves had previously been changed to this type of packing in 1988.
Long-term corrective action will flush system to reduce iron oxide buildup and investigate possible changes to internal valve trim which is less susceptible to corrosion product accumulation.
The packing on the remaining valve will be changed during the next refueling outage.
LERs with common elements:
88-005,88-009, and 86-003.
Ei 8911200238 891107 l
PDR ADOCK 05000333 t
a eran
t.
me maa v2 wucuan..ueronyca l
L6CENSEE EVENT REPORT (LER) TEXT CONTINUATION maoveo ow =o me-em o
e, inn n,=
easaurv =ana o.
ooca n =umme* a>
unn,um a i eau o 1Auti A. FIT! PATRICK
" 22.' 6 "An m.
NgtllAR POWit PLANT 0 l1 l 5 2 0 l1 tl2 or t l7 o is o lo o :l: a 8 l9 van n-
- - me = mm.w.
E11S Codes are in []
Revised Report - Previous Report Sent on October 18, 1989
Description
The plant was shutdown for a planned maintenance outage.
During the outage deficiencies were discovered in the operation or internal condicion of each of the nine containment isolation valves [ISV) of the Reactor Building Closed Loop Cooling water system (RBCLCW) [CC).
The valves are air diaphragm operated with miston in-cage valve internals.
The sizes are one 1-1/2 inch, one 6-inch, and seven 4-inch valves.
The stroke closing time of all nine valves was measured on September 18, 1989 during the scheduled performance of ASME Section XI In-Service Testing (IST) Surveillance Test ST-IR, " Reactor Building Closed Loop Cooling Containment Isolation A0V Exercise".
The closing times for seven of the valves were within the IST limits.
The stroke time for valve 15A0V-130B of 11.81 seconds exceeded the maximum permitted closing time of 11.4 seconds by 0.41 second or 3.6 percent.
Also, observation of the valve stem showed an erratic sticking motion.
The operating mechanism for valve 15A0V-131A could be closed only by manual operation of the valve handwheel.
The stroke time for valve 15A0V-131B was within the IST limits.
However, visual observation of the valve stem travel revealed an erratic sticking motion.
The position indicating limit switch for valve 15A0V-130A required adjustment.
Subsequent to the IST, a modification was in progress to replace the existing air supply to the valve operators with a nitrogen supply.
This modification required closing the valves as a part of the protective tagging procedure.
On September 29th and 30th, two valves, i
15A0V-132B and 15A0V-133B, would not close for tag-out although they had previously passed the stroke time test ten days before.
On October 4th, as part of the system valve line-up, in preparation for plant start-up, additional sticking problems were found with several of the valves.
Plant management directed that the remaining four valves in the system be disassembled, examined, and retested.
At this time the pistons for the four remaining valves, 15A0V-130A, 15A0V-132A, 15A0V-133A, and 15A0V-134A, while previously demonstrated to be operable under air pressure, were found by mechanics to be sufficiently fouled by corrosion products that they wore not able to be moved by hand in the valve cage.
n l**'"'
I ans anna v2 =ucuan =.uuoav e UCENSEE EVENT REPORT ILER) TEXT CONTINUAT60N
- aovio oue ao :=-em l
somes ema
. un u.am m one.it== sea e o..uu.,
i.,
JAutt A. Fli2PATRICE
'*a "tm."
- "2 HCLEAR POWER PLANT 0 l1 l 5 0l1 0 l3 or g l7 o
o o o l: a 8l9 van e = so m m.
Cause
There are four contributing conditisns which, because they existed concurrently, jointly may have contributed to excessive resistance to motion of the valve stem or the piston inside of the cage assembly:
L o Valve stem packing o Corrosion product (iron oxide) build-up l
o Inappropriate selection of valve trim for valve application o Infrequent (outage only) cycling of valves closed and open 1.
Packing The valves were originally supplied with seven rings of Grafoil packing.
Inspection of the valve stems found a film of Grafoil packing material adhering to the exposed surface.
Based on this description, tne valve vendor (by telephone) suggested that the t
rings may be shearing on the surface in contact with the valve stem.
These observations indicate that packing problems may be one contributor to the observed slow, erratic, or sticking motion (or lack of motion) of the valve stems.
2.
Corrosion Products All of the valves were disassembled for inspection.
All exhibited an excessive accumulation of a black corrosion product sludge (magnetite iron oxide Fe-3 0-4) on internal surfaces and in critical crevices of the operating parts such as the close clearance space between the piston and cage.
The corrosion product build-up was sufficiently extensive that tha pistons on four of the valves could not be removed from the valve l cage by hand.
Although the three valves which had areviously (1988) had live load packing installed all passed tie initial IST, two of them subsequently stuck during valve line-up.
Upon disassembly all three were found to aave pistons frozen in place.
Because of this, corrosion product build-up is believed to be the principal contributor to the problem.
The water in the RBCLCW system is demineralized.
No chemical treatment is added.
Given the quantity of carbon steel pipe and fittings in the system, the generation of iron oxide is expected.
It is not known at this time whether the quantity of iron oxide is excessive.
A corrosion evaluation of the plant closed loop cooling system is in progress.
This is expected to provide additional information on this subject, g...,.
F e
6 ene sm.
c =vea
.evuton, ca UCENSEE EVENT REPORT RER) TEXT CONTINUATION
- aovio ow no ow-eio.
oei.es om e wiun m
emannues m n,,,,,,,,,,,
lANil A, fli2PAf tlCE t
ultitAt POWit PL ANT 0 l1 0 l4 of g l7 o is 10 lo l0 l 3 l 3 l:
819 0 l1 l5 vari.
mo m. mm 3.
Valve Type and Trim Remote operated containment isolation valves for the RBCLCW system i
were not part of the original design configuration of the plant.
The valves were installed as an upgrade modification in 1983 to provide remote containment isolation operation capability.
To
~
meet a commitment installation deadline, the selection of valves was necessarily limited to those available on a short lead time.
It appears that the internal trim option in these valves was not optimal for this application and may be more appropriate to a flow control application than to a full open to full closed isolation application.
This particular trim option appears to be inherently more susceptible to accumulation of corrosion products and to fouling than a full flow isolation valve would be.
4.
Cycling These valves are cycled only during plant shutdown because there is an inherent risk of recirculation pump trip, containment temperature and pressure transients, and a scram resulting from failure of the valves in a closed position.
This infrequent cycling, together with the original packing system and valve trim, permits build-up of corrosion products and contributes to possible sticking in the packing gland and at the piston interface with the valve cage.
Analysis These nine valves were not installed as part of the original plant design in 1975.
In response to the NRC, they were installed as a modification in 1983 to meet 10 CFR 50 Appendix A General Design Criteria 54 and 57.
The valves are not included in the current plant Technical Specifications.
They are included in a proposed amendment to the specifications which was submitted to the NRC on May 31, 1989 as JPTS-84-005.
They are listed in Table 7.3-1 of the updr.ted Final Safety Analysis Report (FSAR).
As such, they are part of the revised design basis of the plant.
The safety function of the valves is to provide a remote manual method to selectively close and isolate individual sections of the RBCLCW system to isolate identified contamination pathways from primary containment to the environment if they occur during postulated design basis accidents.
These valves fail open.
By procedure, these ralves will be closed only if there is an indication of piping failure in the RBCLCW system or activation of ESW system coincident with increased drywell pressure.
These valves do not receive isolation signals for automatic closure.
Remote manual operation of the valves requires a licensed operator to obtain a key and rotate manual keylock switches on J
I
.,=
c.
oca.....wa UCENSEE EVENT REPORT (LER) TEXT CONTINUATION awaono ous =o owe.
io.n ein.
l
.mm a.= ni o.c.n v ee. m n
,y JABil A.flflPATRICE
" t',"
"'an X
u..
NOCll At P0 Wit PL ANI 0l1 l 5 0l1 gl$ or g l7 01610 Io lo I l 13 8 19 vone-..-
e a back control panel.
Further, these valves are accessible in the reactor buildin; and may be closed manually by use of a handwheel.
Therefore, ncit aer the FSAR nor the 3roposed Technical Specifications specify a maximum closing time for t1ese valves.
While the valves may have been inoperable by remote control, the valves l
were never inoperable by local manual handwheel.
They could have always performed their safety function to be selectively closed.
Nevertheless, because of the number of valves of a common type in one system which exhibited common problems indicative of a potential to fail, this event is reported as a " condition that was outside the design basis of the plant" in accordance with 10 CFR 50.73(a)(2)(ii).
The potential loss of ability to remotely close these valves does not by itself result in adverse safety consequences.
To properly assess the safety consequences it is first necessary to postulate a scenario in which the following events occur simultaneously:
o Loss of coolant accident without the concept of leak before break (and therefore lack of pressure rise above RBCLCW pressure l
in containment) accepted in 52FR41288 of October 27, 1987, l
o High energy pipe break in spite of the augmented 100% in-service inspection of weld locations during each inspection interval, o Pipe whip at the location of, and with sufficient force to, produce a double ended severing of (or a missile which severs) l RBCLCW pipe in spite of the AEC conclusions that ".
. the applicant has provided adequate measures to protect against the occurrence and consequences of missiles and pipe whip."
in U.S.
Atomic Energy Commission Safety Evaluation Report of November 1972, Section 5.2.2.
o And then operators are unable to remotely or locally manually close the single individual isolation valve for the particular section of pipe which was severed.
(The valve which fails to operate remotely must also coincidentally be the isolation valve for the particular section of pipe which was severed.)
To determine the significance, it is necessary to combine the mathematical probabilities of each of these four events occurring within one scenario.
As stated in NYPA letter JAFP-89-0689 of September 21, 1989 to the NRC, "The leakage paths postulated.
are highly improbable.
and.
this is an incredible scenario."
gy...
ene e,. m.a u s wuctaan neout.to.7 co.us.es on UCENSEE EVENT REPORT (LER) TEXT CONTINUATION
- aovio owe =o voo-em i n....
.m o.
L
,wiun i n.
oecisi =vani m ti..u
..e,
.. io JAufl A. Ili! PATRICK
" t!?.L." ' Mu
".a RWCllAR P0W(t PL ANT I
ol6lojol0l3l3l3 8l9 0l1l5 0l1 0 l6 0F 0 l7 ein, u c w an.v on
Corrective Action
P 1mmediate:
All nine valves were disassembled.
The pistons and cage trim were cleaned and polished.
The accumulated black iron oxide sludge, magnetite, was removed from the internal valve body.
The existing seven ring Grafoil packing was replaced with five ring live t
loaded packing on five valves.
The packing had previously been replaced in 1988 on three other valves.
The valves were tested for satisfactory performance prior to returning the system to service.
i Long-term:
1.
Two valves will be disassembled and insmected during the next scheduled outage approximately six months from now (March 1990).
2.
The remaining valve will be repacked during the next scheduled refueling outage to ensure consistency between the valves.
3.
The RBCLCh' system will be flushed during the next outage to attempt to remove some of the accumulated iron oxide.
4.
The plant engineering group will investigate the suitability of alternate styles of internal trim designed for the application assigned to these valves.
5.
Periodic cycling of the valves during operations was considered and rejected due to operational restraints.
The other corrective actions, in conjunction with required cycling during shutdown conditions, should ensure continued operability.
Additional Information
Failed Component Identification:
Valve Manufacturer:
ITT Grinnell Valve-Hammel Dahl Model Number:
V500 Size:
1-1/2",
4",
6" Pressure Rating:
1-1/2" - 600 psig 4" and 6" - 150 psig Stroke:
6" 1/4" 4" 1/2" 1-1/2" 1/8" Number of Valves:
1 - 6" 1 1/2" 7 - 4" pg.o..,
m --
o ireis ana v s wuctun atevuton, e UCENSEE EVENT REPORT (LER) TEXT CONTINUATION maovio ow =o sim-eiw tapsett trate FA514fft h4ABB de pochtt toutSth ge 409 =WRID0R ISI Pett 136 J ANtl A. FITZPATRICK
'i.a "Stit'
."'J ?,i:
NICllAt POW [R PLANT 015 l0 l0 l0 l 3 l1 l3 8l9 0 l1 l5 0l1 0l7 0F 4 l7 e
verv = -.
c w w or, LERs with common elements:
88-005 05/25/88 Emergency Service Water Check Valves Blocked by Sludge Accumulation 88-009 10/21/88 Unit Coolers and Check Valves for ECCS Equipment Fouled by Mud and Silt 86-003 03/12/86 Containment Isolation Valve Inoperable Due to Motor Operator Failure Revision 1 corrects format and editorial discrepancies and revises the
" Analysis" section and the description of the effect of corrosion product build-up on valve operability, f
l gtgeomw w.