ML20245A020
| ML20245A020 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 04/17/1989 |
| From: | Hukill H GENERAL PUBLIC UTILITIES CORP. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| C311-89-2027, GL-87-12, GL-88-17, NUDOCS 8904250014 | |
| Download: ML20245A020 (22) | |
Text
,
4 GPU Nuclear Corporation U Nuclear m ',e r s 48o o
8 Middletown, Pennsylvania 17057 717 944-7621 1
TELEX 84-2386 Writer"s Direct Dial Number:
April 17,1989 C311-89-2027 U.
S. Nuclear Regulatory Commission Attn:
Document Control Desk l
Washington, DC 20555 l
Dear Sir:
1 Three Mile Island Nuclear Station, Unit 1 (TMI-1)
Operating License No. DPR-50 Docket No. 50-289 Inservice Testing (IST) of Pumps and Valves Check Valve Disassembly 1
The purpose of this letter is to supplement our letter of June 7, 1988 in support of our request for additional relief from ASME Code Section XI requirements regarding the disassembly i
frequency currently required for certain check valves in the IST l
program.
This completes our submittal of the information requested l
by the NRC in a conference call on May 16, 1988.
j GPUN's letter of June 7, 1988 provided justification for not testing i
the four EG-V10 valves and requested additional relief to extend the l
frequency of inspection to once each ten years for one valve in each set for the following TMI-l check valves:
DH-V14A/B, BS-V52A/B, BS-V30A/B, and MS-V9A/B As requested in the May 16, 1988 conference call, this letter provides the results from valve disassembly accomplished during the Cycle 7 Refueling Outage (7R) and also the following additional information that was requested:
- 1) A review of the TMI-l history for the same types of valves as those listed above.
- 2) A review of industry experience (as documented in NPRDS) for
$f the same types of valves as those listed above that are used in similar service at other plants, and
- 3) A review of the valve installations addressing the EPRI Application Guides for Check Valves.
M Attachment A and the enclosed photographs provide the results of CD inspections performed during 7R.
These photographs show that each ex valve that was disassembled was found in excellent condition.
88 Inspections of DH-V14A,
'V52B, BS-V30A, and MS-V9A/B were observed
,Q by NRC inspectors.
There were no missing parts, no unusual h
A% v /6de i%:&Mid l h a.
GPU Nuclear Corporation is a subs ary of the General Pubhc Utihties Co poration
C311-8'-2027 9 corrosion, no unusual wear nor other abnormal conditions that would prevent these valves from performing their open safety function (or their closed back-checking function).
One of these valves, MS-V9B was not disassembled specifically for IST inspection, but instead to repair a bonnet gasket leak.
The leaking bonnet gasket resulted in part from disassembly for IST inspection during the 6R Outage.
I Attachment B provides the disassembly history for each of the above I
listed valves.
BS-V30B has been disassembled once.
Each of the others have cach been disassembled at least two times.
Attachment C.1 provides the results from our review of TMI-1 history for other similar type / manufacturer / service valves as those listed above.
Attachment C.2 provides a comparison of the 7R disassembly results for Walworth check valves that are used as pump discharge valves compared to the results for those valves that are not pump discharge valves.
l Attachment D provides the results from our review of industry l
experience (as documented in NPRDS) regarding the same type of j
valves as those listed above in Attachment C,1 that are used in l
similar service at other plants.
l Attachment E provides the results of our review of the EPRI Application Guidelines for Check Valves with respect to the above listed valves.
l Discussion GPUN is requesting additional relief from the disassembly frequency for the-check valves listed above.
The following is a description of the valves for which additional relief is requested along with the applicable service conditions required of these valves:
BS-V52A/B - supply sodium hydroxide to the Reactor Building Spray Pumps (BS-P1A/B).
There is no method to test these valves without contaminating the Reactor Coolant System (RCS) with sodium hydroxide.
The only service these valves see is cycling by hand during disassembly for IST inspections.
BS-V30A/B - are on the discharge of the Reactor Building Spray Pumps (BS-PlA/B) and are located inside the Reactor Building at the bottom of the spray headers.
The only service BS-V30A/B see is a quarterly frequency part-stroke test and the manual cycling by hand during disassembly for IST intmections.
DH-V14A/B - are located between the Borated Water Storage Tank (BWST) and the suction of the Low Pressure Injection (LPI) pumps and the Reactor Building Spray Pumps (BS-P1A/B).
The only service DH-V14A/B see are a refueling flow rate test (2/3 of accident design flow), quarterly IST, and the quarterly Engineered Safeguards Logic Test.
l
C311-8'9-2027 MS-V9A/B - supply main steam from the OTSG's to the Steam Driven Emergency Feedwater Pump (EF-P1).
TMI-l does not use EF-P1 for plant startup.
The only service MS-V9A/B see is a monthly surveillance test.
The NRC has addressed a number of check valve failures in recent correspondence.
Also, EPRI has issued application guidelines for I
certain check valves in nuclear power plants to resolve industry concern over the generic problems with check valves.
The mode of failure indicated in a large percentage of failures is partial disassembly (disc may become a loose part) of the i
check valve internals caused by corrosive / abrasive wear, l
inadequate design of locking mechanisms, flow induced disc chatter resulting in damage to the valve internals, or poor quality control during assembly.
Each of these issues is addressed as follows in regard to the TMI-l check valves listed above:
1
- 1. Corrosio../ Abrasion Concerns:
All of the Decay Heat Removal I
System (DH) and Reactor Building Spray System (BS) valves listed in Attachment B are used in chemistry controlled borated water service, are of stainless steel construction, and therefore are not susceptible to general corrosive attack.
MS-V9A/B are carbon steel but are located in a controlled chemistry environment.
Abrasive wear does not result since main steam does not contain appreciable quantities of i
abrasive.
Also, none of the valves listed above are subjected to extended periods of flow.
- 2. Locking Device Concerns:
Check valve failures have been attributed to inadequate design of locking devices or poor quality control during valve assembly.
Attachment B discusses this type of failure for MS-V9A which has been eliminated as a concern for MS-V9A and B.
The BS and DH check valves are in standby service and seldom if ever see flow, thereby eliminating flow induced wear as a likely failure mode.
Consequently, these valves have shown no problems or potential problems with locking devices during any inspection except for the problem regarding MS-V9A.
As discussed in Attachment B, this problem has been corrected and verified by a followup post service inspection.
- 3. Fatigue / Misapplication Concerns:
Fatigue fracture of the disc stud or other internal parts have also caused failure.
Since the check valves discussed in Attachment A are in standby service, fatigue problems would not be likely to occur.
More frequent disassembly of the valves would not add significantly enhanced knowledge of the valve condition because there would be so little operation and opportunity for exercise / wear subsequent to the last disassembly.
Another cause of failure is valve misapplication.
In this context, misapplication
C311-89-2027 refers to selection of a valve that is too large or too small for the system flow conditions, selection of the wrong type of valve, locating check valves in areas of highly turbulent flow, or incorrect orientation of valves in the system (e.g.,
vertical versus horizontal).
Attachment E discusses GPUN
}
review of the EPRI " Application Guidelines for Check Valves" with respect to the four applications that are involved in this request.
- 4. Flow Induced Wear Concerns:
Flow induced chatter was stated as the cause for some of the failures at the other plants.
All of the valves listed in Attachment A have been disassembled at least twice except BS-V30B.
BS-V30B has only been disassembled once.
These valves have been in service for considerably more than ten years and have not exhibited any damage due to chatter.
- 5. Manufacturing defects:
Defects resulting from poor quality control by the manufacturer during assembly can be determined in the first disassembly inspection.
Attachment B describes defects noted on inspection of DH-V14B and MS-V9A/B where problems were determined to be manufacturing defects.
These problems have been corrected.
Since all of the valves for which relief is being requested have been disassembled at least once, GPUN believes that all manufacturing defects have been corrected.
Decay Heat Removal System (DHR) Operability Concerns During Valve Inspections During refueling outages, the RCS and OTSG's are required to be drained-down (due to geometry, the draindown includes both RCS loops and both OTSG's) for long periods of time for Reactor Coolant Pump seal work and OTSG Eddy Current Testing.
TMI-l Technical Specifications requires maintaining two operable means of decay heat removal (one if the Refueling Canal is flooded) ensuring that a single failure does not reEult in a complete loss of decay heat removal capability.
Limited reduction in redundancy is allowed for preventive or corrective maintenance.
However, as discussed in NRC Generic Letter 88-17, it is prudent to minimize the time that Decay Heat Removal Systems are taken out of service.
Disassembling DH-V14A or B each refueling takes away'one train of Decay Heat Removal for no significant benefit resulting in a net reduction in safety.
This is especially true because DH-V14A and B are proven to pass at least 2/3 of their ac,ident design flow rate each refueling by the performance of Surveillance Procedure 1303-11.54.
There is no credible failure of valve internals that would cause these valves, in this application, to significantly restrict flow.
GPUN's request for relief regarding DH-V14A/B was discussed in our February 8, 1989 response to Generic Letter 88-17, Programmed Enhancement Item No. 5.
J C311-89-2027 ;
Conclusion From the a;'ove it is concluded that DH-V14A/B, BS-V52A/B, BS-V30A/B, and MS-V9A/B are low usage valves.
Furthermore, these valves are not subject to the following concerns: 1) Corrosion,
- 2) Locking device failure, 3) Fatigue or misapplication, 4) Flow induced wear, or 5) Manufacturing defects.
GPUN sees no valid failure mechanism for these particular valves.
j We believe that an inspection frequency of one valve in each set l
cvery ten years will provide adequate assurance that DH-V14A/B, BS-V52A/B, BS-V30A/B, and MC-V9A/B would not cause a restriction of the design flow required during post accident operation of their respective systems.
In addition, disassembling DH-V14A/B I
cach refueling is contrary to the NRC's recommendations communicated by NRC Generic Letters 87-12 and 88-17 on loss of decay heat removal.
l l
GPUN believes this letter provides sufficient justification for l
cxtending the frequency of inspection to once each ten years for one valve in each of the following valve sets: DH-V14A/B, BS-V52A/B, BS-V30A/B, and MS-V9A/B.
Justification for deleting the four EG-V10 valves from the IST Program was provided in our letter of June 7, 1988.
Sincerely, H. D.
1 kill Vice President and Director, TMI-l HDH/MRK cc:
F. Young R. Hernan (2)
W. Russell J.
Stolz H. Gregg (1)
(No. of sets of photographs enclosed) l
C311-89-2027 Attachment A As-Found Photographs of the IST Valves Disassembled During the Cycle 7 Refueling Outage
}
PHOTOGRAPH VALVE NUMBER DATE NUMBER DESCRIPTION 1
7/21/88 DH-V14A Internals, Valve closed, No problem.
2 7/21/88 DH-V14A Bonnet, showing slight peening from disc stud, No problem.
3 7/21/88 DH-V14A Internals with disc fully open, No problem.
4 6/21/88 MS-V9A Internals (As discussed in Attachment B, the disc nut is welded to the disc stud),
No problem.
5 6/21/88 MS-V9A Bonnet, No problem.
6 6/21/38 MS-V9B Internals (As discussed in Attachment B, the disc nut is welded to the disc stud),
No problem.
7 6/21/88 MS-V9B Bonnet, No problem.
8 6/14/88 BS-V52B Internals, No problem.
9 6/14/88 BS-V52B Internals, note pin in cover to prevent incorrect re-assembly, No problem.
10 6/12/88 BS-V30A Internals, No problem.
11 6/12/88 BS-V30A Internals, showing capture weld detail for disc stud nut, No problem.
12 6/12/88 BS-V30A Internals, disassembled, No problem.
13 6/12/88 BS-V30A Body, showing capture cavity j
for hinge pin, No problem.
1
i C
ii 1
wg g
t en n
a n e n n
ni i
eh o h o
o r
r rr ct i t
i t
i dd o
e n
t s a t
eOel d
dtrnie a teose c
l tas e
eauast l gchras A
lhsut h
cet a
l netrw 7
atesr T c ascd,ctaif o e e s tiTua S
a l
a e gi c t r f f c r h v
t sw np Itt p5 fl n d e s ao at f
i l
n o
t u
Ig.u nt e0uli nrnt l e e.
rea R1 naririatnawr c s neos ol at o eoou o 1
e e
.icNod fl e
R ra on ee.
nad h tuo r
e f
e o y gtr
.l u dcS elge.iereTtdcn t
o ex.hs soh cinusp g
Cdl aaut
,f l e aeSy ct edu te a
/nb keshn b
np ktSlte
.r forc P
sam as gos mnoa as6tcrsl rl Nax m e e
gii a eiir e
1 ne iaoag pe e s lnntsw s tg ln3 err icin U
l s
gi o
sdio e
rroSrnri.ed o a titkrr andt tbhaodSeiencsng o s asaara suno artpcn6t t esoun r i eeeeoe iooh eaipne1 af a pi o o i N
P D
Sdsl cw DFcp S c w ai v 3 momodlfr O
I e
y T
t l
A a b U
m m "8
L ife 7
8 3
A xos 8
/
8 VS o s
/
1
/
EI rea 1
2 7
S pt s
/
YA pai 7
4 L
e 4
SP d
3 8
o 5
'C A" M
RS N
r A
e E
r HN u
h TO t
t c
r u
YG a
o BN f
w I
u l
B DL n
a EE a
W t
RU M
n IF e
UE e) m QR zn 4
h E ii 1
c RH S(
a C
t SA d
t EE e
t A V t
h LD c
g AE er i
VL pa l
B xe S
)
S t
w gy y FI r o nl l c OD adl.i rdri e
pef l eneg YE s
k etato.
LB U
yo3hur rLt BMO lr/cfaTa s
)
A F
)
S G*
I nI(
D s
gS n
iPe o
s( r i
e u t
Det 0
i ra 5
d gur 2
n nse
/
r o
isp 5
o C
pem 2
t ire 2
c e
PPT e
c p
i t
s v
n n
r e
i e
m S
n d e
C o
e g
R dr tra N
ii aerk 7
uv rt on y
2 l n oata b
0 FE BWST 2
d r
A e
B 9
ee 4
s 4
8 vb 1
s 1
l m V
V e
1 au n
1 VN H
t 3
H D
i C
D W
1I l
ll)
d.
t ea d
t ne c
l eoon c s eh A
lhcNNu a
t
.l c
7 at a o
Tfl r Tl a T
e s tif v
t swr
.f Ssaa Set S
Ii up I ct It f
i l
n ut ns t s xa n
o t
u I gshoa raue re re c s n
gi w osns o e ol 2
e e
.i gi s f
uo f ne fl r
R rntor h
o iS e
r e i ra dt ol d
. dc e
o y
gt tk r e eiN ee s ex g
Cdl aaaaow l w lb
.h l e.
a
/nb keeec b
n b
np b n
P sam assl r
ms.o mooa mno m e e l o eesi eti r eii e s l n gd a sst s s
t g s
t U
l s
gneus sol o.
sdio sdi b
a tiit st apurr andt and o s asrsur srsra suno sun r i ee ena iueoe iooh ioo M
N P
D SdOTup Dprcw DFcp DFc O
I e
y T
t l
A a b 8
U m m 8
7 L
ife 7
4
/
8 A
xos 8
8 2
/
VS o s /
/
1 2
EI rea 1
6
/
M S
pts 6
YA pai TB ADD EFG l
AN 6 a SI l
4 7i L
e 4
3 c SP d
3 2 e M
o 5
Np
'C A M
S RS N
r A
e E
r HN u
h TO t
t o
c r
c yG a
o y
BN f
w o
I u
l l
B DL n
a l
EE a
W A
t RU M
n I F e
UE e) m QR zn 4
h E
ii 1
8 c RH S(
a C
t SA d
5*
t EE e
t A
V t
h LD c
g e
AE er i n
VL pa l
o B
xe S
N KM EW CE ES HS nr ee CA pt S
t w
gy y tor FI r
o nl l c r a.
OD adl i rdri aduT e
pef l eneg peqS YE s
k et ato k
I LB U
yo3h ur rLt yoh B
l r/c f aTa s
l rcr MO nt2 a euSuSe nt ao ET O s( e rqI qET Osef S
S
)
)F G*
I nI(
D s
gS n
iPe o
s( r i
e u t
Det 0
0 i
ra 5
0 d
gur 2
3 r
n nse /
/
o o
isp 5
0 t
C pem 2
0 c
ire 2
2 e
PPT e
c p
i t
s v
n n
r e
i e
m S
n d e
C R
o e g
N dr t ra ii aerk 7
uv rt on y
2 l n oat a b
0 FE BWST 2
d r
B)
A B
se 9
ee 4d 0
0 s
8 vb 1t 3
3 e
l m Vn V
V n
1 au
- o t
1 VN HC S
S i
3 D(
B B
W C
Hl il
oe e
t tr s
hg e
t r s
c ce o tamc ce o A
awoo k aa awoo e.
7 Tf Nl nael Tf Nl d
TSs
.e e s Sss i e r pd Sss ed S
h v
t Iil lt en Iil ta n
kn I
p f
i l
r a n ". o y( sRa t a n aa a
o t
u b
p r a n ". o r".r e
c s osrwi nu.l osrwi lt owg s
3 e e f
ees do f
ees.
e f eo r R ht nor e
eim) a ht nor tk nt r
dt n ra v
kt oen dt n ra es e
o y
eiI ere o
airvrd eiI ere kad eeh d
o g
Cdl l w kow b
esfl ee l w kow sge lkp a
/nb b
i c -
a l o atl b
i c a
l bi P
sam ms
.l l
pdvnb ms
.l r
gdb mld m e ees"l o s
B e im ees"l o em e e s"h e s sst a
a 2dvf e
sst a
t ce V
l s soldus S erees soldus eas sdc b
a apunst e Vseivs apunst nl s ana o s srsuur m
- osll a srsuur npa sut r i i ueona a Slbeae i ueona oee i ot M
N P
D Dprf O
,up S
Bcorvr Dprf up BRr DFa I
e y
T t
l A
a b U
m m 4
7 1
4 4
8 8
L if e 8
8 8
8 8
/
A xos /
/ /
/
/
4 VS o
1 3
8 1
3 1
EI rea S
pts
/
YA pai 6
TB ADD EFG l
jle AN 6 a m
SI l
7i L
e 3 c SP d
2 e M
o Np
'C A M
S RS N
r A
e E
r HN u
w TO t
o c
c YG a
y BN f
o I
u l
B DL n
l EE a
A t
RU M
n IF e
UE e) m QR zn h
E ii 4
c RH S(
a C
t SA d
t EE e
A V t
LD c
e AE er n
VL pa o
B xe N
TCA se g
a S
yue n
n FI c
S i o.
OD ny t
ie e
eb s tc YE s
gd e
ai LB U
rnyt rv B
eal er MO mt n orpe ET Eso noos S
S
)
A S
) F G*
I nI(
D s
gS n
iPe o
s( r i
e u t
Det 0
i ra 5
d gur 1
n nse /
o i sp 5
C pem 7
i re e
PPT c
i t
v n
r e
e m
S n
d e
o e
g dr t ra ii aerk 7
uv rt on 2
l n oat a 0
FE BWST 2
r A
9 ee 2
1 1
2 8
vb 5
5 l m V V
1 au 1
VN S
S 3
B B
C I
t i
sl a8dk s
t c
w oo
/l c o c
s r_
A
.owd1 ca o,
a t
7 Tsl u1 ur l6 Tfl r e
s St et c
8 Ssaa v
t Il,hsne e/
Ii up f
i l
unT ihreh1 ts o
t u
rso c Toht1 raue c s 2 oei sd t
osns 4
e e
frs.i e r
rn f
uo r R f
ordd.asoi h
o e
r o dyra deeef dt ol gr e
o y
erreeacwi d eiN a
g Cdl 6
l oowh a
f nn l w p
a
/nb bt c t
,l oiou b n
P sam e mc r tpnrio ms o
t m
e g
eal onu etf eesi x
e s
a sf a enndvu sst s e
l s
p ssuse i e l s sol o.
n b
a aist wc wsot apurr o
s e
st urt ssoisr srsra e
r i e ianaeiahhea i ueoe e
N P
D S
D s u pb d w s T r p Dprcw S
O I
e y
T t
l A
a b 8
U m
m 6
8 4
6 L
i f e 8
/
8 5
A xos /
1
/
/
VS o s 1
2 2
1 EI rea 1
/
1 1
S pt s 6
YA pai TB ADD EFG AN E
Sl l
W L
e 0
SP d
5 M
o 3
'C A M
5 RS N
r 6
A e
E r
HN u
h TO t
t c
r YG a
o BN f
w I
u l
B DL n
a EE a
W t
RU M
n IF e
UE e)
I m
QR rn h
E ii 6
c RH S(
a C
t SA d
t EE e
A V
t t
LD c
h AE er g
VL pa i B
xe l
KM EW S
HS e
CA k
S o hr FI r co OD t af e
s e YE s
h LB U
t nt B
r enT MO a poS ET P omI S
S
)
A F
)
S G*
I nI(
D s
gS n iPe o
s( r 0
i e u 0
t Det 5
i ra /
d gur 0
r n
nse 5
o o
isp 0
t C
pem 1
ire c
e PPT e
c p
i t
sn v
n m i
r e a e
m e
C S
n t
o S
R N
dr ii n y
7 uv i b
2 l n a
0 FE M
2 de r
s 9
ee A
8 s
8 vb 9
9 e
l m V V
n 1
au t
1 VN S
S i
3 C
M W
l
!l ll
7 burwinoc e s natdi a
v t
aupe rt f
i l
besat o
t u
roe haea c s oNsdt ww 5
e e
f ol e
r R oed6oe r
d
.l wn8 nS e
o y
ek a9 s
g Cdl l a,e 1d a
/ nb benhd e.hp P
sam ml oTunwga m
e e i ti onr e s sts
.s hig l
s seo dsko b
a akrrce ct o s ssrasddao r i i aoeidnrh N
P, D
Dgcwdaacp O
I e
y T
t l
A a b 8
U m m 8
L if e /
A xos 1
VS o s 2
EI rea /
S pt s 6
YA pai TB ADD E
FG AN E
SI l
W L
e 0
SP d
5 M
o 3
'C A M
5 RS N
r A
e E
r HN u
h TO t
t c
r YG a
o BN f
w I
u l
B DL n
a EE a
W t
RU M
n IF e
UE e) m QR tn h
E ii 6
c RH S(
a C
t SA d
t EE e
A V
t t
LD c
h AE er g
VL pa i B
xe l
KM EW S
HS e
CA k
S ohr FI rco OD t af e
se YE s
h LB U
t nt B
rent MO apoS ET S
PomI S
A
)
S
)F G*
I nI(
D s
gS n i Pe o
s{ r i
e u 0
t Det 0
i ra 6
d gur /
r n
nse 0
o o i sp 5
t C
pem 0
c i re 1
e e
PPT p
c s
i t.
n v
m i
ea r
e r
e C
r S
n t
R o
S N
dr ii n
y 7
uv i b
2 l n a
0 FE M
d 2
e r
)
s 9
ee Bd s
5 vb 9t e
l m Vn n
1 au
- o t
1 VN SC i
3 M(
W C
l l
C311-89-2027 Page 6 of 7
Attachment B MS-V9A/B INSPECTIONS / FINDINGS DURING 6R REFUELING The November 18, 1986 disassembly of MS-V9A revealed the following:
- 1. Disc stud nut, disc stud washer and disc stud cotter key were missing.
- 2. The disc was still in place in the yoke arm.
The fit between the hole in the yoke arm and the s.1oulder of the stud was a sliding friction fit.
There was no damage to the disc stud end, stud threads, disc seat or disc seating surfaces.
The valve was found operable.
The disc stud washer was found in the MS-V6 (downstream control valve for the Turbine Driver (EF-U1) for the Turbine Driven Emergency Feedwater Pump, (EF-P1), during the Eddy Current Test Outage (5M).
This outage was approximately six months prior to 3
the November, 1986 disassembly of MS-V9A.
Viewed from the side, the washer was bent in an "S" shape and was oblong.
On November 20, 1986 the aisc nut was found downstream of MS-V10B.
The nut was in good condition.
However, upon trial reassembly the fit between the nut and the stud was very loose (approximately 1/16" of axial and radial play).
This may have allowed the cotter key (1/8" diameter and 2)" in length) to take the load rather than the threads.
Eventually the cotter key may have experienced fatigue failure and broken into several pieces.
This loose thread fit was determined to be a manufacturing defect, not a problem caused by wear.
A review of the machinery history file indicated that MS-V9A had never been disassembled prior to November 18, 1986.
An unsuccessful search was made for the missing cotter key in drain valves, steam trap and the EF-U1 steam chest / governor valve.
As stated above, the washer was found during the SM Outage and EF-P1 had been operated satisfactorily on Main Steam at least five times with the cotter key missing.
In addition, Test Procedure 636/1, " Full EF-P1 flow to the OTSG's" was performed in October, 1985.
Therefore, it was concluded that either the cotter key was never installed by the valve manufacturer or it is not in a location where it will do damage.
A new cotter key was installed in MS-V9A and the nut was welded to the disc stud.
For MS-V9A and B, GPUN's IST Program Submittal dated December 24, 1986 states, "If degradation is found which would make the valve's full stroke capability questionable, then the other valves (MS-V9A or B) would be disassembled.
MS-V9B was disassembled.
MS-V9B was found
C311-89-2027 Page 7 of 7
Attachment B operable. No unusual corrosion, loose parts or wear was found.
The fit between the nut and stud was somewhat loose but not as loose as MS-V9A.
As a precaution the disc nut was welded to the disc stud.
During the 7R outage (July 25, 1988) EF-U1 was completely disassembled for preventive maintenance.
All internal parts were found in excellent condition.
The cotter key, if it entered the turbine, did not cause any damage.
\\
l
C311-89-2027 Attachment C.1 Review of TMI-1 History for Similar Valves BS-V52A/B and BS-V30A/B:
BS-V52A/B (4" valves) and BS-V30A/B (8" valves) are 300 lb. check valves manufactured by Alloyco (Model N2376 Special).
There are approximately six other Model N2376 Special valves at TMI-1.
These other valves are BS-V23A/B (10" valves on the suction of BS-P1A/B), DH-V16A/B (10" valves on the discharge of DH-P1A/B),
DH-V50 (4" valve on the return from refueling water cleanup) and MU-V112 (4" in the line from the Makeup Tank to the Makeup Pumps).
Like BS-V52A/B and BS-V30A/B, these six other valves are in borated water service.
A review of the Machinery History Files shows that these valves have provided acceptable service.
However, during 1983 the internals of DH-V16A/B were replaced due to wear.
Also during 1983 DH-V16A was found with a missing disc stud nut retaining pin.
DH-V16A/B are the discharge check valves for pumps DH-P1A/B and it is believed that the wear was caused by continuous operation of DH-P1A/B for over 6h years during the extended shutdown mandated by the NRC due to the TMI-2 accident.
TMI-1 was shutdown from March, 1979 until October, 1985 because of the TMI-2 accident.
In addition to the Alloyco 300 lb. Model N2376 Special valves, TMI-1 also has Alloyco 150 lb. Model 376 Special valves.
These 150 lb. valves are very similar to the 300 lb. valves.
These 150 lb.
valves are DH-V49 (3" valve, DH-P1A/B test line back to the Borated Water Storage Tank), MU-V49 (3" valve from the Reactor Coolant Bleed Tanks to the Makeup Tank), MU-V113 (2)" valve at the outlet of the Makeup Demineralized), and SF-V7/8 (8" valve on the discharge of SF-P1A/B), and SF-V50/51 (8" valve from SF-P1A/B into the Spent Fuel Pool).
A review of the Machinery History Files shows that these valves have provided good service.
DH-V49, MU-V49, MU-V115, and SF-V50/51 have no entries in the Machinery History Files.
SF-V7/8 only have entries for bonnet leaks.
DH-V14A/B and MS-V9A/B:
DH-V14A/B, 300 lb. valves and MS-V9A/B, 600 lb. valves (Model 5344 and Model 5350, respectively) were manufactured by Walworth.
As part of TMI-l's response to INPO's Significant Operating Experience Report (SOER) 86-3, TMI-1 disassembled and inspected 12 Walworth check valves including some which are not safety-related during the 7R Outage.
Most of the valves selected to be disassembled were judged to be in the more severe service applications.
Some of the valves were known to be less than full open (disc not against its stop) at normal operating flow rates.
Most of the Walworth check valves disassembled during 7R were C.1-1
C311-89-2027 Attachment C.1 (Contd.)
150 lb. valves and these 150 lb. valves are in essentially the same configuration as the 300 lb DH-V14A/B and the 600 lb. MS-V9A/B.
Attachment C.2 provides the 7R results of disassembly for Walworth check valves that serve as pump discharge check valves compared to that for those which are not pump discharge check valves.
From Attachment C.2, the basic conclusions which can be made for Walworth check valves are:
- 1. Pump discharge check valves do experience wear, if the pumps are operated for extended periods of time.
- 2. Valves that see little service see little wear.
From the above and Attachment C.2 in conjunction with our review of the machinery history files it is concluded that DH-V14A/B and MS-V9A/B will not experience degradation because they are not pump discharge check valves, they see little service, and therefore little wear is expected to occur.
i C.1-2
sa sa sa sa sa f
ue ue ue ue ue o
nw nw nw nw nw u
u u
u u
1 r
r r
r r
oo oo oo oo oo N
N N
N N
e g
s s
s s
s a
.t
.t
.t
.t
.t P
sr sr sr sr sr s
t a ta t a t a t a t
l p lp l p l p l p l
u u
u u
u u
se se se se se s
es es es es es e
ro ro ro ro ro R
o o
o o
o yl yl yl yl yl r
r r
r r
o o,
o,
o o
t n tn t n t n t n co co co co co ai ai ai ai ai f s f s f s f s f s so so so so so ir i r ir i r ir tr tr tr tr t r ao ao ao ao ao Sc Sc Sc Sc Sc SE e
V t
L ay A
Dl V
b em 8
8 5
8 8
K t e 8
8 8
8 8
C as /
/
/
/
/
E ms 1
1 1
3 6
H i a 2
2 2
2 2
C xs
/
/
/
/
/
oi 6
6 7
6 7
H rd T
p R
pf O
Ao 2
WL C
AW tn F
rg e O en b
b b
b b
m ri l
l l
l l
h Y
ut c L t a h0 h0 h0 h0 h0 a B cR t0 t0 t0 t0 t5 t
M a/
r6 r6 r3 r3 r1 t
E fl o/
o/
o/
o/
o/
A S
de w0 w0 w4 w4 w1 S
nd l5 l5 l4 l4 l4 A
ao a3 a3 a3 a3 a3 S
MM W5 W5 W5 W5 W5 I
D R
7 y
r r
a F
s e
e r
O e
v v
p v
iy i y S
S l
rc rc T a dn dn g
n L
V e
e n
i U
eg eg i
a S
k nr nr d
M o
E c
ie i e l
t R
e bm bm i
o s
h rE rE u
t sp C
e u
u B
km c
t n tn mn nu e
i g
v
( e
(
e o
ao aP v
v t
en T
r r
li 1i t(
r a
e Ur Ur k
S ee h
S
- D
- D n
e gt c
F F
a rn aa s
Ee)
Ee)
T ei) rw i
np np lbd od D
oim oim r
ire t e tbu tbu e
out Se p
rP rP t
BTa F
m mu mu a
l e
u aTr aTr Ws yre ty P
e e
e e p rer ac t et t et dm at -
sn t
Sh2 Sha eu i ay ne C
o tw t w tP l wt eg R
n n
d n d a
ide dr N
i re i re rI xef ne e
aoe aoe oP uea om y
a d
t e
a "A
"A s
h "A
"B s
7 t
r 4
A 6
e 2
ee 9
4 1
6 1
n 0
s vb V
V V
V V
t 2
e l m i
v au S
S H
S O
W 9
l VN M
M D
A C
8 a
~
V 1
1 3
C A
l
3 owea t6 d
ow c eF3 d nc.oi ltktdeo8t lt n dw
- t uwslt ecnnrb et ei g n 6 et o
t u
f rb y o a e Rnn rb nas 8 nn.
fl rda o
o c
wrEner o ei el ner aaec nn d
oOova nl rdvaRova
,upl e 2
n ei gu B F S b e e neoeul nEbee ss l r s oe nt rw oehtt aoO rw tueap t
i seisd rrp i s t sviSrp l nst e
l s
l r n.ee e s du t
e e uuosa g
u osoecal pgoc osnl cecrgoc s on a
s raht s a nt u raaf senent u eoli s P
e rw tiAn,o d
rw irupo d
rN a
R o dud 3 oBlde o dsdhf,lde rB crnl 1i er crua t
er y.o/sr aaf eVtdal atwe yCal r
Apa l e h-cn l o l es hll/
,l o ol
,7 oe aWdstCna,at aW etllB at t anVt w u ua I u et u cr Au/et cno s s
wn fAcsr s
sen fAcsr aoiR t
u.ts e
3 nne nsdtu.ihe.d10 nne fi sRtn ns ee
.y1 ait e
ait sto ee u t c r w pl V d t
ut wpnVd t
icrhnv rset ol isu redt ou isu tnrt ne oaihet uCual oahnetoSual auooor N pd t b s f I gwf N pt a b s f N g w f Sf cbbp S
e E
t V
ay L
Dl A
b V
em 8
t e 8
8 8
K as 8
8
/
C ms /
/
1 E
i a 5
6 2
H xs /
/
/
C oi 7
6 7
rd H
p T
pf R
Ao O
2 W
L C
AW t
rg n
F en b
b b
e O ri l
l l
m ut h
Y t a h0 h
0 h0 c
L cR t5 t 5 t5 a
B a/
r1 r 1 r1 t
M fl o/
o /
o/
t E
ue w1 w 1 w1 A
S nd l4 l
4 l 4 S
ao a3 a 3 a3 A
MM W5 W 5 W 5 S
I D
R 7
g.
F B
r B ny
)
O d
a il d
e e
l e S
n t l
A ov T s a a c
eoi L
e l
u s h Ct U
v A e N p t
c S
l r
m ye E
a rp r u rcp R V om y o P ons e
f u t f
f ee k
c P e g
gr c
i e f
e n er e
v vg a
v i.
v e h
r l n s
l l y l mB C
e ai a ol a E S
vl n
v oe v d e
o o Cv gn g
ko N k
i k na r
cC c dt ci
(
a e
e ec eda h
he y h se h l c
ct l c op c is s
a e l s up i
ei v
e Ce eBm D
gdi g
r g
u ret p
amc re, rrP ac ao m
hre hic ht r u
cep cv cce P
st s srd sat ine ien iea e
DI r DSa DRW ra C
t B
/
a
/
B B
h A
/
/
7 t
r 3
A A
2 ee 1
0 7
0 s
vb V 1
V 2
e l m V
v au C
R 9
l VN I
S R
8 a
N V
1 1
3 C
B
3 eoihic el2 rdrt Fawse ul sts NaVee a ot u
eg rdf e rie t
crt eewi r sscn f
o l gadr sRianwbll o.utxi o
f eeogp e.nNv a
a f enreh cshe
,w8i rsl stWe cua o
e dn pod gm 8
e p p'A n r s 3
dnr l
sm eao nn s eaee nrr/t.
t l nvhnaae2 s5 ul0 e
l nNe a
e l
beeTrh h2 a8 epa 2dht be ss g
u mt r s
es/l9l mVlta mt oac a
s en
- t oga2 1bRr uoh en.ows P
e sif.
t nw1 e aNoRo t
siel i
R saoes i r
,r nNho sal ed am vachdnel ee sts ambrr s aoee s
nl ws nowider e
ieoa i wa riro.edve ier hh dvivmde dspshf ncel c dvenTt it r
NtelAnt ont ai ipo Atahag ul a o eieavv Bt oi n
cvg n ngcf ctil r
0 n s
.e eruei.nl 0
arn oiareke 2 eyode 2Vreogsndt ein vrercs Vvl rnw e
- rsrnueust reorepre el rut c
Robhiapt nnuewepxoho Rruooe n
Ncothcosiidst soeccn N pf cfb an S
e e
t E
t V
ay n
L Dl i
A b
a V
em 8
m t e 8
9 K
as /
8 e
C ms 0
/
v E
ia 1
7 i
H xs /
/
t C
oi 2
3 c
rd 1
e H
p r
T pf r
R Ao o
O c
2 W
i f
L C
A o
W t
t rg n
F en b
l e O ri l
u m
ut s
h Y
ta h0 e
c L
cR t5 r
a B
a/
r1 t
M fl o/
ae t
E ue w1 c
A S
nd l4 sn S
ao a3 aa MM W5 n
A
)
) e S
d I
e 8t D
u 8 n n
9i R i 1 a
,m 7
tn F
o re O
C ev
(
bi S
mt T s en L
e ce U
v ev S
l De E
a r
r
(
r R V o
o p
e f
f e
k c
A B
gf c
i e e
ao e
v vp vp t
h r
l m l m ut C
e au au Ol S
vP vP u
e l s g
ke ke ae r
cc cc er a
ei ei S
h hv hv a
c cr cr D
s e
e l s i
eS eS P a.
D g
g rr rr Cy p
aa aa Rr m
he he t
u cl cl mn P
sc sc oe iu i u r
e DN DN f t r
n a
ye rc t
te a
nr h
A B
e 7
t r
0 0
l 2
ee 2
2 l a 0
s vb V
V an 2
e l m no v
au R
R oi 9
l VN N
N it 8
a ti V
3 id 1
dd 1
dA 3
A*
C B
C311-89-2027 Attachment D Review of Industry Experience GPUN searched Nuclear Plant Reliability Data System (NPRDS) data on the following valve types:
DH-V14 14" 300 lb Walworth Model 5344 BS-V30 8" 300 lb Alloyco Model N2376SP BS-V52 4" 300 lb Alloyco Model N2376SP MS-V9 6" 600 lb Alloyco Model 5350 To make the scarch as broad as possible, it was decided to use the following selection criteria:
Manufacturer:
Walworth Co.;
Alloyco Div./Walworth Co.
Valve inlet size:
4 in. to 14 in.
This selection criteria would result in getting information on all Walworth and Alloyco check valves with inlet sizes from 4 in. to 14 in.
The resultant NPRDS data search indicated (116) entries for the selection criteria.
This does not include the (6) TMI-l entries that are discussed in Attachment C.
The (116) entries were reviewed and categorized as follows:
(30)
Non-wear related problems: gasket leakage, dirt, or corrosion; (22) --
Failed Local Leak Rate Testing (LLRT);
(37)
Seat leakage during operation or system operability testing; and
( 2 7 ) --- Wear problems resulting in either replacement of valve internals or replacement of the valve.
The entries in each category were then analyzed for cause, corrective action, and applicability to TMI-1.
The (30) non-wear related problems were predominantly gasket and packing leakage, internal dirt or corrosion, or maintenance related.
These entries are not directly applicable to the issue of extended surveillance intervals because, except for the entries that involve dirt and corrosion, they are r.at related to system operation or valve location.
These entries are not applicable because the TMI-l inspection history demonstrates that the valves for which this request applies are not subject to problems involving dirt and corrosion.
l D-1
C311-89-2027 Attachment D (Contd.)
The (22) failed LLRTs are judged not to be applicable to TMI-1 because none of the valves for which additional relief is being requested are Category A valves requiring seat leakage tests.
Also, disassembly and inspection would not be expected to provide
}
any measure of valve leakage.
None of the examples where a valve failed LLRT gave any indication that excessive wear of valve internals was the cause of the failure.
In fact the TMI-1 LLRT experience on similar valves indicate that high leakage from this type of valve is generally due to dirt / corrosion products on the seating surfaces.
Without such particles of dirt or corrosion products, the seating surfaces experience very little wear.
There were (37) entries that indicated seat leakage discovered during plant operation or system operability testing.
There are many causes for these entries such as wear from normal use, seat wear, water or steam cutting and dirt in the system.
None of these entries indicated any excessive wear or failure of valve internals (except for the seating surfaces).
There are (27) entries describing wear as the cause of valve failure to function properly.
Two of the entries concern cracked seats in valves subject to high pressure reverse flow.
Eighteen of the entries are pump discharge check valves.
Four of the entries are valves downstream of clbows.
The remaining three entries appear to be flow induced wear problems.
The high pressure reverse flow check valve problems do not apply to the TMI-l issue because the valves of concern are not subjected to such service.
The pump discharge check valve problems do not apply because the TMI-1 valves are not pump discharge check valves.
The elbow problems do not exist because the valves are not downstream of clbows.
MS-V9 and DH-V14 are downstream of gate valves. This aspect and related flow concerns are addressed in Attachment E where this has been shown through tests not to be a problem at TMI-1.
None of the (116) entries from the NPRDS data search indicates any failures of check valves due to wear related problems in similar service at other plants.
The failed LLRT and seat leakage entries indicate that excessive wear or failure of the valve internals was not the failure cause.
No B&W plants other than TMI-1 are listed among the (116) entries from the NPRDS data.
This indicates that these plants have not had check valve failures or that the failures have not been entered into NPRDS.
To determine if any failures have occurred in similar valves at B&W plants, GPUN contacted Oconee, ANO, CR-3, and Davis Besse.
l D-2 1
4 W
C311-89-2027 Attachment D (Contd.)
Based upon this information which was provided by telephone, these plants have not experienced any failures for valves similar to those for which GPUN is requesting extension of the disassembly interval.
Based upon the NPRDS data search which indicates there have been no failures in similar valves with service similar to DH-V14A/B, BS-V52A/B, BS-V30A/B, and MS-V9A/B, extension of the disassembly intervals for these valves is justified.
D-3
C311-89-2027 Attachment E Review of the EPRI " Application Guidelines for Check Valves" A design review of the valves' operating conditions and locations was performed per the EPRI Application Guidelines for Check Valves.
The design review indicates that the valves will have sufficient flow velocity to maintain the disc in a full open position during system operation.
The design review also indicated r. hat only BS-V30 is installed sufficiently away from an upstream flow disturbance.
The other valves are within 1.5 diameters of a gate valve, versus the EPRI recommended distance of 10 diameters.
This should not be a problem as the TMI-1 SOER 86-3 test program indicates that upstream flow disturbances do not cause flow induced problems except where l
the disturbance is caused by the valve's proximity to the pump discharge.
Based upon this design review and specific testing which indicates that DH-V14A/B, BS-V52A/B, BS-V30A/B, and MS-V9A/B are not likely to be subjected to flow induced wear problems, extension of the disassembly intervals for these valves is justified.
l l
1 E-1
- _ _ _ _ _ _ _ -