ML20099F145
ML20099F145 | |
Person / Time | |
---|---|
Site: | Palo Verde ![]() |
Issue date: | 11/12/1984 |
From: | Van Bront E, Van Brunt E ARIZONA PUBLIC SERVICE CO. (FORMERLY ARIZONA NUCLEAR |
To: | Bishop T NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
References | |
ANPP-31094-TDS, DER-84-58, NUDOCS 8411260276 | |
Download: ML20099F145 (11) | |
Text
.
50-527 o MSPGD 3, c,0
- 5rizona Nuclear Power Project P.O. BOX 52034
- PHOENIX. ARIZONA 85072-2034] p'r j g {J - l: } {,
November 12, 1984 ANPP .31094-TDS7TRB W 4
U. E. Nuclear Regulatory Commi9sion Region V 1450 Haria Lane - Suite 210 Walnut Creck, California 94596-5368 Attention: Mr. T. W. Bishop, Director Division of Reactor Safety and Projects
Subject:
?inal Report - DER 84-58 A 50.55(e) Reportable Condition Relating To Auxiliary Feedwater System Solenoid Valve Failed To Open In Response To AFAS Signal.
File: 84-019-026; D.4.33.2
Reference:
A) Telephone Conversation between J. Ball and T. Bradish on August 8, 1984 B) ANPP-30449, dated September 6, 1984 (Interim Report)
C) ANPP-30873, dated October 17,1984 (Time Extension)
D) ANPP-30995, dated October 29, 1984 (Time Extension)
Dear Sir:
Attached is our final written report of the deficiency referenced above, which has been determined to be Not Reportable under the requirements of
-10CFR50.55(e).
Very truly yours, 1
l 2&\%kke E. E. Van Brunt, Jr.
APS Vice President Nuclear Production j ANPP Project Director J l
SEVB/TRB/nj i Attachment 1 cc: See Page Two i
l 8411260276 841112 PDR ADOCK 05000528 8
[FJ7
n,. -; ,
v, "
- i th , ,
- '[$.
, X.
D - ,
- =--
. r
. s
- c. Mr. ~ T. W.1 Bishop ;
' DER 84-58 Page Two. >
=J^
cc : - ' Richard:teYoung; Director.
0ffice of . Inspection 'and Enforcement
~
4 h -
U. S. ;Nuclea' Regulatory .Connaission Washington; D. C. 20555
\
T. G.' Woods, Jr.'
D. ' E . Ka rner, W. E. Ide D. B. Fasnacht A. C. Rogers L. A. Souza
{r D. E. Fowler' '
T. D. Shriver' C. N. Russo B. S. Kaplan
~
J. R. Bynum J. M.-Allen A. C. Gehr W. J.:Stubblefield
' W. G. Bingham -
- R. L.-Patterson R. W. Welcher H. D. Foster D. R. Hawkinson R. P. Zimmerman L. Clyde M. Woods T. J. Bloom D. N. Stover J. D. Houchen J. E. Kirby D. Caaady Records Center Institute of Netlear Power Operations 1100 circle 75 Parkway, suite 1500 Atlanta, CA 30339
_v t ..#- , ,.m - - . - ., .- . -. ,r__.--- ..---r.- r. _ ,,-, ,
~ '
^x -
3 1, > ,f,
-,p." *4 J: 2, n s I. *
. kFINAL REPORT --DER 84-58' a 7- ' DEFICIENCY EVALUATION 50.55(e)-
. ARIZCNA PUBLIC SERVICE COMPANY (APS)
PVNGS UNIT 1.
I, - .,"r , -
- '~I. - Description'of Deficiency '
r: '.
The:following deficiencies.were identified with Target Rock Solenoid. ,
' c, c Valves which are installed in various~s'afety-related systems.
g
- 2A. : NCR~ SM-462'8 and DER 84-58':' 4
')
-The steam supply bypass solenoid valve JSA-A-UV138A is used in
, the slow starting of the auxiliary feedwater turbine. During the
~
pre-operational testing'of the steam supply bypass line to the auxiliary pump (AFA-P01),~ this valve failed to open in response
~
to the AFAS signal as required by design.'-
7 4
.. B . NCR SM-4623 and DER _84-63:
3 Valve JSIB-HV613 is located on a Safety Injection Tenk -(SIT) vent
- -line. This valve is_ required for relieving pressure during cooldown of the reactor. This is'a failed close valve and did.
not close when required during-testing. Ihe' valve was opened when SIT pressure was 600 psig and- would not .close via the handswitch si 200 psig. The power was shut'off to the esive and it still would not'close.
C. NCR SE-4721 and DER 84-67:'
Valves JSGBUV-1135A, -1135B, -1136A, -1136B, JSGAUV-1133.and
-1134 are used in steam drain trap applications. These are fail-close valves that did not stroke when required during testing. It was initially reported that the control room indication lights did.not reflect the true state of the valve; however, the actual condition was inoperative valves.
Investigation of the valve conditions mentioned above Indicated 'timt the valves were mechanically stuck and could not be made to operate.
The following evaluation, reportability assessment, and cc rrective action plan are also applicable to DERS S4-63 and 84-67 Evaluation Af ter disassembly of the valves by the vendor's field service representative (Target Rock Corp.), evidence of three types of foreign dirt were found in the valves. The following description of contamination is adapted from Target Rock letter C4990, attached,
~ dated October 2,1984, relating their evaluation.
A
s i < t i =
A
}
Final Report- ~
- Page.Two~
^
4.1 Grey Rust: This is due t'o water chemistry during the running of high temperature wa.ter and steam through' the valve. The gray rust deposits-onto the internal surftess'of the valves and is typical of normal plant operation. The grey rust' flakes'off when-the valve internal surfaces are ' scraped. The flakes are thin and.
brittle and can easily be broken up.
4.2 Ped Rust:- This'is due to rusting in the steam generating system. The red rust jis carried along by the -system fluid during' a a prestart up flu h operation.1 It contains fairly large particles that are spherical in shape. .There are also a. large .
amount of fine particles that have a pawdery . appearance. When the valve internal. parts are dried; the red rust:dep) sits onta j the ' valve luternal surfaces and gives them a red coloration. The dried red rust is soft and becomes a. fine' red powder when the
{ coated surfaces are rubbed.
i ..
4.3 Hard -Dirt: This is due to construction dirt in the lines.'
f It I- may be veld yarticles or slag or metal fragments. Such material is carried along by-the system fluid during a prestart-up: flush 1' operation. If those particles become ' lodged at the valve seat and the valve is pressurized, they can cause indentations in the valve disc and body seat and. subsequent increase in the valve-
', seat. leakage.- ,
The diacoloration and flaky grey rust deposits have been noted l before when valves in similar applications' have been i disassembled. This is considered normal and has not preverted valve operation in the past. Hard dirt usually damages the valve-seat and increase seat leakage, although the possibility of hard i dirt jaaming the disc does exist. The amoutt of dirt in the area
! above the disc piston ring was not sufficient to block the pilot i ports and passages which would have prevented the opening of the i valve under pressure. The valve failed to. open due to the
! presence of a significant amount of red rust.inside and spherical
. particles of red rust on the sain disc below the piston ring. !
l This prevented -the actuation of the valve by wedging the main !
disc. The radial clearance. in this area is 0.005 of an inch. '
The valve could have been easily opened if pressure had been 4
applied to the inlet port before actuating it at the site, because there was not enough dirt to clog the pilot disc and
- pressure provides a strong assist to open the dioc.
i i
i J
f
}
(c; -
D+ '
, ~
w lk
- a- ,
rr
, ' t y
a s-
- . Final Report ' ,
' Page Three a e The roct csuse of the valves' failure to operata is the_ inability
~
- 'to completely clean and/or_ flush moisture and debris 1from these-small branch' lines which accumulated ddring construction s'nd' startup activities. This condition .is unavoidable and not unexpected since'.the' valves are, welded.in the lines and must'be
-installed prior to: hydrostatic ~and startup testing of the 'systens.
'During' testing s'etivitiesis steam and/or . water flow is E
^
insufficien: to' completely remove the bits of debris' remaining in.
the systems. Based on a.ctual experience;with valves of similar '
design;' the' frilure of .these valves to function during tormalf system operation is not expected to occur since removal of remaining particles occurs when the systems are exposed to normal
. operating parameters-(eig.,-flow, pressure). vin addition,- ~
careful control of plant chemistry during . operation minimizes'the formation of corrosion products.which contributed to this:
condition. This condition is only applicable.to Target RocL colenoid valves since these are the only pilot-operated valves -
l_
].
purchased by the project and, as such, the only valves
- - susceptible to this type of failure to operate, t.
II. Analysis of Safety Implications-Based on the above discussion, DER 84-58, 84-63; and 84-67 have been j evaluated as Not Reportable under the requirements of 10CFR50.55(e) .
since the conditions to not represent a Quality Assurance program l
breakdown, a significant deficiency in design or. construction, or a, i deviation from performance specifications as' defined by the.
regulation.
i . .
!- This condition is evalusted as Not Reportable under the requirements i of 10CFR Ptrt 21, since it does noc constitute's defect in a basic--
componaat.
I 9
III. Corrective Action i
! The remedial action will be to disposition NCRs SJ-4628, SM-4623, and 1SE-4721 to eit!.ar clean and reinstall the valves or to replace the valves with clean valves from Unit 3 or from the warehouse.
J
] Copies of this report with the referenced Target Rock letters 'will be transmitted by Engineering to the Transition Manager, Startup
]
Manager, and Maintenance Manager for information and to ensure that j appropriate inscructions are included in work ordere generated in
, accordance with Procedure 30AC-9ZZ01 1
I~
1:
l 4,4,- - - - , c vr +mu, e - :-, m s n 7-+e- ,- vr n w v<w ,,wr--aw p- ne ~-w'r w' -r' - sww-s s<m.m ,-s- - - ----v
rm -
- a tp 5 - - , -
s.
I' Final' Report DER 84-58 Page.Four W
If the condition recurs durlag startup testing on other systems or in Units 2 and 3 the remedial action.shall be to (1) back-flush the line if possible and/or (2) disansemble the valve, clean the internals, and polish'the main disc and pilot dice seats. A light lapping of the valve seat.will be performed if needed. Based on this report, Startup' will be directed to establish this :orrective action plan as a pre-reluisite to startup testing f or similar conditions. -
Proper plant chemistry ecntrol will inhibit the formation of corrosion products which have been identified as a contributing
' factor to the failure of these valves to operate. This type of .
control should prevent these f ailures during futura plant operationa.
The disposition of this DER shall be applicable to all future nonconformances found concerning contaminated Target Rock solenoid j valves. Applicable NCRs shall class-reference this DER for
- reportability disposition.
IV. References
- 1. Field Report /Nonconformance Report (SFR/N23) ,
Startup/SJ-4628; 1AF-3;1 SFR/NCR 1CD-178/SE-4721; SFR/NER ISI-752/SM-4623 '
Startup implemeti stion Documentation for work performed in Unit 1.
I 2. Target Rock letters to W. G. Bingham; C4990, dated -
r Octoter 2,1984, and 41048, dated October 8,1984, i
i
, . _ _ , - - . _ _ _ .l._- _ ., . .-. -
.. .s .
(
.. m *M '
':.. s : ::ve:
Target Rock Corporation, P.O. Box V.1966E Broadhollow Rd., East Farrningdate. N Y.11735 / Phone. (516) 293-3800 ss.m. :p s: <.= s ::c:w s Please refer to. C4990 October 2, 1984 -
. 231%U Bechtel Power Corporation Post Office box 60860 .
Terminal Annex N8 Los Angeles, CA 90060 '
- f, n.gssm Attention: W.G. Bingham "[,, -
Project Engineering Manage.- ~g
~R% a Weitern Power Division [k' hip , -
cras% .
Su.bj ect: g --
Arizona Nuclear Powar Pro.iect -
gp Bechtel Job 10407 __ m Defici tncy Evaluation Reports in regard to Target Rock -
1, I Solencid valves, File: JM-603
_ij,,
TRC Pr; ject 76HH ' lig
Reference:
1)Bechtel Letter, (B/TR-E-49052, MOC336995) -% a 70 d; Dated 9/5/84, W.G. Bingham to D. Vater *['[ -
2)Bechtel Letter, (B/ANPP-E-121740, MOC 341716) ~
dated 9/24/84, W.G. Bingha'u to J.D. Hou hen - -
.__ {__
, Gentle: ten:
On 9/17/84 six (6) model 76HH-008 solenoid valves were received at Target Rock "orporation Ocom the Palo Verde haclear Plant. Mr. A. Wissman of Bechtel came to Target Rock to witness the as received inspection and subsequent disasserbly of the valves. Pictures taken of the valves are enclosed. When the electrical enclosure covers were recoved r:om the valves it was noted that all of the switch asserblies were loose or rissing. Available infomation was that before the valves were cut out of the line at Palo Verde, they. were energized to open but failed to actuate. The inlet port was uncovered so that the motion c f the main disc could be viewed. Without cleaning or modifino the as received valves in any way,the valves were energized to open by applying 125 VDC to the solenoid coils. Five out of the six valves opened ir a normal canner. The sounds from the rixth valve indicated that the pilot disc was moving but the main disc did not lift. A brass drift was inserted into Nie outlet port and the underside of the main disc was lightly tapped while the, solenoid coil was energized. The main disc then lifted and thereaf ter operated in a nomal manner. At this poinc all the top works from the valves were removed and the valves were cut open by re-moving the body to bonnet seal welds. Taking care to keep the parts from each asse:bly separate, the valves were disassembled and inspected. The following outlines some of the findings and observations,.
., - it ,
. - .f3 y -(A u '
TAnorr Roca CORPORATION hb5 Page 2 of-4 C4990 October 2,1984 1.-
- All of the sleeves which guide the main discs were examined and no evidence of damage or dirt contamination was found. ~
2.
All of the piston rings that are mounted onto the main disc were examined and all were found to move freelf.
3.
- All of the main disc and pilot disc ports were examined and no evidence of damage or dirt clogging was found.
4.
Identification of Dirt Contamination Evidence of three types of dirt were found in the valves. They can be described as follows:
4.1 Grey Rust:
This is due to water chemistry and running high te=per-ature water and staam through the valve. The grey. rust deposits onte oneration.
plant the internal surfaces of the valves and is typical of normal The grey rust scrfaces are scraped. flakes off when the valve internal be broken up. The flakes are thin and brittle and can easily 4.2 Pf d Rust This is due to rusting in the steam generating system '
and startthe up red flushrust is carried along by the system fluid during a pre-operation.
particles that are spherical in shape.The red rust contains fairly large of fine particles that have apowdery alppearance.There is also a large amount When the valve internal parts are dried, the red rust deposits onto the valve in-ternal surfaces and gives them a red coloration. The dried red tustrubbed.
are is sof t and become a fine red powder when the coated surfaces 4.3 Hard Dirt:
be weld particlesThis is ordue slagto orconstruction dirt in the lines. It may setal fragments.
This material is also carried along by the system fluid during a pre-start up flush oper-ation.
If hard dirt is caught at the valve seat and the valve is pressurized this can cause indentations in the valve disc and body seat and subsequent increase in valve seat leakage.
5.
Dirt contamination of the valves since five of six valves operated normally in the as received condition the amount of dirt found in the valves was not enough to prever t actuation..
(It was reported that all six valves did not acutate at the site ) . Much of the dirt may have been carried out of the valve by the water that was released when the valves were cut out of the line. However, all of the valves had their internal surfaces discolored with a coating of grey rust.
The valves were still wet internally and particles of red rust appeared to be insuspension in the pockets of water.
aritively dirty and one valve was almost without any red rust.Two Theof valve the valves were c
' with the main thediscstuck seat.disc was examined and hard dirt indentations were no ed actuation before the disc was tapped open.This may have caused the di Only two of the valves had significant and the pilot amounts disc. of dirt in the area above the main disc piston rings the piston ring groove.There were some red rust deposits on the disc below
. nm .m
~,
s t . . .
( p ;
.Y 7
.hhhY TAnorr Roca ConPonArrox Page 3 cf 4 C4990 October 2, 1984 Rai}ure Analysis '
The discoloration and flaky grey : rust deposits have been noted before when valves in similar arplfcations have been disassembled. This is considered norral and has not prevented valve operation in the past. The effect of hard dirt is usually to damage the seat and increase seat leakage although the possibility of httd dirt jamming the disc does exist. The amount of dirt in the area above the disc piston ring was not sufficient to block the pilot ports and passa,.s which would have prevented th.e opening of the valve under pressure. It is felt that the failure of the valves to open was due to the presense of a significant amount of red rust in the valve, or by the spherical particles of red rust on the main disc below the piston ring.
This could have prevented the actuation of the valve by wedging the main disc. The radial clearance in this area 0.005 and the relatively sof t red rust could have wedged the disc. However, even if this were the case the valves could have been easily opened if pressure was applied to the inlet port before actuating the valves st the site. This is because there was not enough dirt to clog the pilot disc and pressure provides a strong assist to open the disc. Since the valves failed to actuate several days af ter the ,
start up-flush and that there was still a considerable amount of dirt in the valves indicates that the flushing was incomplete. This gave the dirt in the valve a change to " set" and increased the possibility of interference with valve operation. The fact that the valves opened in the as received I
condition can be explained by a partial drying out and breaking down of the red rust particles which wedged the disc clcsed. This could also have been aided by the shocks and vibrations that the valves experienced during ship-ment. Although it is not always possible, a back flush would probably have been able to clear the valves.
Cenclusions ,
. - _ . , ,u ,. .m ...
Based on the inspection of the valves the debris which was caused to flow through the lines during the initial start up phase at Palo Verde was re-sponsible for the failure of the valves to open. It is felt that an attempt should have bcen made to pressurize the valves and then energizing the valves to open. Indications are that the discs were not tightly wedged into their seats. Also back flushing (if possible) should have been attempted.
In order to return the valves to service the only work that is required is.
a cleaning of the valve internals and a polishing of the main disc and pilot disc seats. A seat leakage check might indicate that a light lapping of the valve seat is needed to reduce seat leakage. -
1 Based on the field experience of over 10,000 Target Rock valves of similar - l design which spans a period of 12 years, the failure of the valves to operate af ter dae ' start up flush.should be classified as a non-typical occurence.
"r.is includes valves that have been used in the same or similar application ha in Palo Verde. It is felt that the present design of the valves can tolerate a significant amount of dirt during normal operation. Also it is important to make sure that a flush operation is maintained for a sufficient time to assure that most of the debris is removed. Durir.g start up operations at other plants j
the only valve operational deficiencies that have been noted is increased seat l
., , t: , .
(i,)
. . . (m)
N3&
TARort Rock ConPoRATION -
Page 4-of 4 C4990 October 2, 1984 g' leakage due to seats that have been dama3ed by hard dirt. This can sometimes be corrected by backflushing but usually requires relapping of valve seats.
Tne screwed and seal weld design of the body to bonnet joint has several advantages. A bolted gasket sealed joint would greatly increase the size-of the valve, especially for high pressure and high temperature valves. Also the seal weld provides a hermetic seal for the body to bonnet joint and this eluminates the maintenance problems associated with leakage of gasketed joints.
Gaskated joints are considered justified where valves mast be subsected to-full line differential pressure while flowing. (As is the case for throttling or control valves) The cutting of the seal weld af ter the valve top works have been remy.ed should take, on the average, about 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. This is with the use of the Target Rock seal weld cutter.
Of the similar val res installed in the field the incidence of cutting the I seal weld to allow repair of.the valve has been minimal. Many seal welded
^ valves have been installed and operating for more than 10 years without ever
' having to be serviced or repaired. Target Rock fee".s that these advantages justify the continued use of the screwed and seal welded bonnet to body joint.
With regard to the use of the valve for st am drain application, Target Rock
- feels that once the plant is beyond the start up stage, the line construction dirt will have been eliminated. And, as demons *. rated in other plants, trouble free operation should be experienced.
Very truly yours, TAROET RM K CORPORATION SAs l&.&t4*
Steven Karidas Project Engineer
/(jij 1 v sroline Bullock '
Contracts Administrator POR: Thomas D. Crowley Contracts Manager /Pover Products 1 CB/kb cc D.M. Pattarini D.K. Vater T.D. Crowley '
i A. Szeglin V. Liantonio
, , - . ,, - - -m- - -. _ _ .
- y. - - - .
it ,
-f.'g-g>, _
,. -v. * -. . .
t
~
Page 2'of.2 TARorr Roca CORPORATION C41018
. October 8,-1984'
- If you have any questions please call the undersigned. .
Very truly yours,
- TARGET ROCK CORPORATION.
hw Au Steven Karidas Project Engir.eer SK/kb cc: V. Liantenio-K. Wenzel T. Crcwley 6
k -
~' '- -
4 I, [g
,o a .
i .
i 6
W ,)
?
i n
4 e
. I p .
t I
d
- 1 )
t 2
_ _ _ _ . _ _ . . _ - . . _ _ . _ .. _ _ _ . _ . ,