ML20129K351

Provides Info Re S/Rv Test Results Obtained Following Plant Unit 2 Reactor Scram on 961029 During 961105 Site Visit

ML20129K351
Person / Time
Site:	Browns Ferry
Issue date:	11/15/1996
From:	Williams J NRC (Affiliation Not Assigned)
To:	NRC
References
NUDOCS 9611210135
Download: ML20129K351 (28)

Text

-

November 15, 1996 i

MEMORANDUM TO:

File FROM:

_ Joseph F.' Williams, Project Manager

/s/

Project Directorate 11-3 Division of Reactor Projects:- I/II Office of Nuclear _ Reactor Regulation

SUBJECT:

INFORMATION REGARDING SAFETY / RELIEF VALVE PERFORMANCE OBTAINED DURING SITE VISIT - BROWNS FERRY NUCLEAR PLANT UNIT 2

, f l,

During a~ site visit,on: November 5,'1996,' Tennessee Valley Authority (TVA) personnel briefed me'regarding' safety / relief valve (S/RV) test results obtained'following the Browns. Ferry Nuclear Plant' Unit 2 reactor scippm on October 29,.1996:

This preliminary data'(enclosed) is being provided to other NRC staff members for theiriinformation.~ Therefore, it is appropriate to place the information in the Public Document Room.

TVA is expected to ' document final test results in Licensee Event Reports discussing the scram and the subsequent S/RV. test program.

. Docket No. 50-260 n

^

Enclosure:

S/RV Test Results cc w/ enclosure:

C.G. Hammer, EMEB M.S. Wegner, AE00 Distribution:

Docket > File -

BF Reading File I

})hh f F. Hebdon j

DOCUMENT NAME:

G:\\BFN\\PDRN0TE.SRV Ta receive a copy of this decament, Indicate in the bon; "C" = copy without attachment / enclosure "E" = copy with attachment / enclosure "s" s No copy 0FFICE PDII-3/PM /p,f l6 PDII-3/LA.

l l

NAME JWilliams # '

BClayton 4/1 C /

DATE 11/ tr /96' 11/;(/96 0FFICIAL RECOR) COPY 190066 g ffLk % h h 9611210135 961115 DR ADOCK0500g0

...- 4

• W4

.t*

1 4$

UNITED STATES i

s j

NUCLEAR REGULATORY COMMISSION "2

WASHINGTON, D.C. 20666 4 001 o

%,J,,

• November 15, 1996 1

MEMORANDUM TO:

File FROM:

Joseph F. Williams, Project Manager

,w Project Directorate II-3 Division of Reactor Projects - I/II Office of Nuclear Reactor Regulati

SUBJECT:

INFORMATION REGARDING SAFETY / RELIEF VALVE PERFORMANCE OBTAINED DURING SITE VISIT - BROWNS FERRY NUCLEAR PLANT UNIT 2 During a site visit on November 5,1996, Tennessee Valley Authority (TVA) personnel briefed me regarding safety / relief valve (5. 'f) test results obtained following the Browns Ferry Nuclear Plant Unit 2 reactor scram on October 29, 1996.

This preliminary data (enclosed) is being provided to other NRC staff members for their information. Therefore, it is appropriate to place the information in the Public Document Room.

TVA is expected to document final test results in Licensee Event Reports discussing the scram and the subsequent S/RV test program.

Docket No. 50-260

Enclosure:

S/RV Test Results 4

cc w/ enclosure:

C.G. Hammer, EMEB M.S. Wegner, AE00 1

s

a f

a i

1 1

1 J

j t

l SAFETY / RELIEF VALVE PRELIMINARY TEST RESULTS 4

J

.I 1

3 i

I i

d k

ENCLOSURE

.--._...._m.

I l

j Olegerens Vess asse6 nam I

,-:a.

-.a I

i l

s e-mm

=

i 1

A&rVA L...$ o I

w

=

% 4.s t

mm Vihe m I

l Ptseen Vess d

Mais Va%

l I

i I

i

,n I///2//////////////,

e

}

~ vh n l

/gwwwwg,,w,,,,,

p r g-:-tw

-~%

i a

m'

(.

I

=

% VA Pisees fllllj l'=

j,..

" v~ ~ +

f4 :: t::::::.,

Dem s m

4

['l d

l y m uniu m m

wssPm e

i r~

s l

3 l

d i

b

f W

l D

il i

1 f

h a

I I

}j l_

inseesse:: s!

i

!,_1_

l s

ee !,,,,,,,...

i_,

p

,, !.2 2, !,,,,,

i i

l_I_

h 2

g_l_

x

, a.u. s s i s.s a.a i. a.s. !_!

ij 111111111ll11 111lllll l l 111 ej I lllllllllllll lilllllllllll d !! !! ! ! ! ! ! !!! !

l

} !! ! ! ! ! ! ! ! !!! !

4 4 : 5 9 9 5 5 5 5 85 5 jt ss s s s s s s s s s a s

_2kkkkkkkkkkkk k

---~~~

~~

~ ~ ~~

i I

I f

I i

Cycle 3 Cycle 4 Cycle 5 Unit 2 Unit 2 Unit 2 Unit 2 Unit 2 Stellite Pt-Stellite (All BFN (All BFN (All BFN Cycle 6 Cycle 6 Cycle 7 Cycle 8 Cycle 9 Units)

Units)

Units)

ONith)

(Without) i Mean 2.41 1.73 2.29 8.86 4.48 4.56 2.52 3.79 3.11 2.76 To')v6 j

Standard Deviaton 2.09 2.5 225 11.57 3.42 3.4 225 3.05 4.81 225 r

t Number 21 35 36 14 12 13 16 13 130 16 l

i l

l I

l l

1 l

l I

._ _. - - - - ~"

~

l U2C3 U2C4 U2C5 U2C6 Setpomt First Pop DetR Percent Setpomt Fast Pop DdR Percent Setpont Frut Pop DriR Percent Selpoet FwstPop DetR Percent 1125 1125 0.00 1115 1114 0.09 1105 1105 0.00 1115 1108 0.63 1125 1130 0.44 1125 1127 0.18 1115 1114 0.09 1125 1136 0.98 1125 1119 0.53 1115 1113 0.18 1125 1127 0.18 1125 1148 2.04 1105 1112 0.63 1115 1112 0.27 1105 1103 0.18 1105 1082 2.08 1125 1135 0.89 1115 1118 0.27 1105 1107 0.18 1105 1130 2.26 1115 1125 0.90 1125 1121 0.36 1115 1110 0.45 1125 1172 4.18 1115 1130 1.35 1125 1129 0.36 1115 1110 0.45 1125 1180 4.89 1105 1120 1.36 1115 1119 0.36 1105 1110 0.45 1125 1185 5.33 1115 1131 1.43 1105 1101 0.36 1125 1119 0.53 1115 1176 5.47 t

1115 1132 1.52 1105 1101 0.36 1125 1132 0.62 1105 1168 5.70 1115 1097 1.61 1125 1120 0.44 1115 1108 0.63 1115 1194 7.09 1105 1123 1.63 1125 1119 0.53 1115 1125 0.90 1105 1250 13.12 1115 1140 2.24 1125 1118 0.62 1125 1137 1.07 1115 1500 >

34.53 I

1105 1130 2.26 1115 1122 0.63 1115 1128 1.17 1105 1500 #

35.75 1115 1142 2.42 1105 1112 0.63 1115 1130 1.35 j9ss.,ip 1115 1144 2.60 1125 1134 0.80 1105 1121 1.45 mun 1125 1177 4.62 1105 1114 0.81 1115 1133 1.61 fox 1125 1181 4.98 1125 1135 0.89 1105 1123 1.63 9144r.

l 1105 1170 5.88 1115 1127 1.08 1115 1135 1.79 vetva>

1105 1036 6.24 1105 1117 1.09 1125 1147 1.96 1105 1184 7.15 1105 1117 1.09 1125 1148 2.04 1105 1118 1.18 1125 1149 2.13 1105 1118 1.18 1105 1133 2.53 1125 1139 1.24 1105 1135 2.71

[

1125 1140 1.33 1125 1158 2.93 1115 1136 1.88 1125 1160 3.11 l

1125 1147 1.96 1125 1161 3.20 1115 1143 2.51 1125 1161 3.20 e

1115 1081 3.05 1125 1163 3.38 1105 1140 3.17 1115 1158 3.86 1105 1141 3.26 1125 1170 4.00 i

1115 1154 3.50 1125 1176 4.53 1125 1171 4.09 1105 1156 4.62 1115 1030 7.62 1125 1179 4.80 1105 1250 13.12 1105 1197 8.33 7

1105 1219 10.32 t

U2C7 U2C8 U2C9 Setpomt Fast Pop Drift Percent Setpoet Fust Pop Drift Percent Setpoet First Pop Drift Percent 1115 1115 0.00 1125 1127 0.18 1125 1135 0.89 1125 1134 0.80 1125 1122 0.27 1125 1136 0.98 1125 1147 1.96 1125 1121 0.36 1125 1136 0.98 1125 1161 3.20 1115 1120 0.45 1115 1129 1.26 1125 1163 3.38 1125 1132 0.62 1115 1131 1.43 1125 1163 3.38 1105 1115 0.90 1115 1147 2.87 1115 1158 3.86 1115 1100 1.35 1125 1160 3.11 1105 1148 3.89 1125 1147 1.96 1105 1141 3.26 1115 1159 3.95 1125 1157 2.84 1125 1183 5.16 1105 1167 5.61 1105 1139 3.08 1105 1163 5.25 1115 1202 7.80 1115 1151 3.23 1105 1169 5.79 1105 1225 10.86 1105 1141 3.26 1105 1225 10.86 1125 1167 3.73 1105 1148 3.89 1115 1187 6.46 1115 1202 7.80 l

r

_i 1t!![t

,ii fr

!! liIi!tI' l

j[t1

'll jI!1)i

'1I

'l.

l o" $

a ta o"

E* O g D

t 762 in

)

4.88 6

2 o

41 004 p

C 1

2 22 34 e

t U

S o

o v Ea e An V

/

w R

(

n n

S o

o a

it i

M t

t a

iv i

A a

a R

D v

t e

e e

a n

D D

e v Ee s Aa t

io a

p d

D d

r r

te a

a S

ndt V

nd n t

n an u R

an u V

ea o S

e a o t

t R

MSC M

MSC S

s v Ea e An H

M A

t f

i r

D

)

6 tn C

e v Ee e Aa e

2 c

U re ro P

f tpecx e v Ee

• An e

(

a taD t

n i

o

• v Ea n Aa p

t e

S VR S

n v Ee N Au M

5

~

I n v Ee An l

v Ee o Au 0

5 0

5 0

5 0

5 0

0 5

4 4

3 3

2 2

1 1

3Ro o WEE:sZ

.il i

1 a-.i l

l'

,il

.l-l ili li

i i

i MSRV Setpoint Drift (Last Four Unit 2 Cycles Only) 15 -

y y

j i

U U

i C

C l

J l

j k

k 12 l

i I

i l

1 t

0 n'

t 6

0 r

1 2

3 4

5 6

7 8

9 10 11 12 13 14 15 16 17 l

l EU2C6 mU2C7 OU2C8 OU2C9 l

l l

t l

I

-.. - ~..... -. - -.. - -. - _ -.... -. _ _ _ -. -

MSRV Setpoint Drift - Platinum versus Stellite i

15 ---

I i

12 j

1 i

9 E

e I

i I

a.

6 j

i

^^E,

,c 3

Jlilllll 1 2 3 4 5 6 7 8 9 1011121314151617181920212223242526272829303132333435363738 l

l 5 Platinum Alloy B Stellite (except for U2C6)

O Stellite (U2C6 to present) l

.. -. - -... -... _ ~..

. - -. -.. - - -. _ ~ - -.. -.. - _. -. -.

i

)

I i

i MSRV Prot Valves (AR Data to include U2C6) 15 Ji D

i

=

t b

I I

o9 o

?

}

l 12 iE l

4 i

I I

9 E

i O

f lie R.

6 j

f i

i i

t 3

i i

s

. _....... - - = = = -mam

^^^^

lllllllll 3

l 115 tu lE MSRV Pilot Valves l t

6 f

i t

r MSRV P90t Valves (AR Cycles except U2C6) 15

[

t i

12 t

i i

9 E

O c

0 E

6 i

f 3

i t

1 0

ts t.

.\\.

.\\...

1.

\\.

\\

1 6 11 16 21 26 31 36 41 43 51 56 61 66 71 76 81 86 91 96 101 106 111 116 121 126 131 lE MSRV Pilot Valves l I

?

I

.I GENuclear Energy l

l SRV SETPOINT DRIFT DATABASE i

W. J. Roit, Sr. Engineer i

Valve Design Engineering

>21 i

O a

BWROG SRV Setpoint Drift Fix Data as of April 18,1 i

$ SRV SETPOINT DRIFT DATABASE Status of Pt-Stellite Alloy Discs

SUMMARY

OF DATABASE (PLANTS WITH 50/50 STELLITE /Pt-STELLITE)

• NUMBER OF DISCS:

33 Pt-STELLITE 24 STELLITE i

• AVERAGE DRIFT:

Pt-STELLITE:

2.7 %

STELLITE:

5.3 %

i

• NUMBER OF DISCS > 3% DRIFT:

Pt-STELLITE:

11/33 (5 < 5%)

STELLITE:

14/24 (4 < 5%)

i

• PLANTS:

%Aor

• 5 Pt-Stellite,6 Stellite I

pteor-B 4 Pt-Stellite,4 Stellite 9,Adr C.

11 Pt-Stellite,11 Stellite i

?otat-D 13 Pt-Stellite,3 Stellite BWROG SRV Setpoint Drift Fix Data as ofApril 18,1996 Page 2 i

$SRV SETPOINT DRIFT DATABASE Status of Pt-Stellite Alloy Discs NOTES ON Pt-STELLITE ALLOY SETPOINT DRIFT RESULTS TO DATE:

f SIX VALVES WITH Pt-STELLITE DISCS EXHIBITED DRIFT >5%.

THREE OF THE SIX HAD DRIFT CONSISTENT WITH HIGH LEAKAGE.

THREE VALVES HAD DRIFT CONSISTENT WITH DISC STICKING.

l VISUAL OBSERVATIONS:

~P(A.n-S DISC: SMOKY / SILVERY APPEARANCE (LOW OXIDE).

Ptaur 6 DISC: SMOKY / GOLDEN APPEARANCE (LOW OXIDE).

? tao r c._

DISCS REPORTEDLY HAD A DARK COATING.

Pt-STELLITE DISCS REPORTEDLY APPEARED LESS OXIDIZED THAN STELLITE.

TLMTD Pt-STELLITE DISCS REPORTEDLY HAD EXCELLENT APPEARANCE.

BWROG SRV Setpoint Drift Fix Data as of April 18,1996 Page 3

i

$ SRV SETPOINT DRIFT DATABASE l

t Histogram Chart of % Drift for Valves with Pt-Stellite and Stellite Discs Histoyam Chart for Group: ALL t

is l

Seedoec AN Diecs SteElle P4fSleEle f

{ Stellite 3,

2, 33 13<

l 8 Pt/ Stellite

^'"*.

3.res s.32s 2m

,2 Stendeed Dev.

4.19%

s.1s%

2 92 %

11 -

to -

3 N

~

~

s.

>.2

>-i.

>o.

> i.

>2

>3

>4

>s.

>e.

>r.

>e.

>e.

>io as

<-2

<.i

<-o

<-i

<-2

<-3

<-4

<-s

<-e

<7

<-e

<-se

,====o een nwon==s,amessospe==

BWROG SRV Setpoint Drift Fix Data as of April 18, IM Page 4 h

l

< ' - - ~

$ SRV SETPOINT DRIFT DATABASE f

Histogram Chart of Second Pop Drift for Valves with Pt-Stellite and Stellite Discs

{

\\

1 Histogram Chart for Group: ALL i

si.ei.eic As os.c.

se.an.

evse.an.

( Stellite 2.

32 8 Pt/ Stellite

^~*

12 St.nd.rd D.v.

1.58%

1.44 %

1.06 %

10 9

1 a.

N j

7 1 0 1 d Ifl d d I d 5

i ifi i t b ti b i

i ti b ti b t i t 3

O f

<= -3%

>-2. <= 1

>0.<=1

>2.<=3

>4.<=5

>6.<=7

>8.<=9

>10

-o--

a,a

(

l BWROG SRV Setpoint Drift Fix Data as of April 18,1996 Page 5

$ SRV SETPOINT DRIFT DATABASE Effect of Leakage on Setpoint Drift from Stellite /Pt-Stellite Database SetpointDriftvs. Leakage i

18.0 % ;c a Pt-Stemte 16.0% -

XN 2 c 14.0 % 3t 5y 12.0 % - -

a 10.0 % -

^

x s.0% u M

a s.0%

b o c x

4.0% -

a g

a

^

2.0% j i aa a

a

,OvX

.A A

0.0%

.a 0

)[

a 10 20 30 40 50 60 70 80 g )

-2.0%

t Lookage SWhd BWROG SRV Setpoint Drift Fix Data as of April 18,1996 Page 6 i

i

$ SRV SETPOINT DRIFT DATABASE Summary of Results / Conclusions

• Pt-STELLITE DISCS HAVE PROVIDED A MAJOR IMPROVEMENT IN SETPOINT DRIFT AS COMPARED TO STANDARD STELLITE DISCS. (AVERAGE DRIFT AND FREQUENCY OF DRIFT CUT IN HALF).

• THREE Pt-STELLITE DISC EXHIBITED HIGH DRIFT WHICH MAY HAVE BEEN CAUSED BY DISC STICKING. THE SPECIFIC CAUSE OF THE DRIFT IS UNKNOWN.

• MORE DATA IS NEEDED TO PROVIDE GREATER SIGNIFICANCE TO THE Pt-STELLITE vs. STELLITE, AND LEAKAGE vs. DRIFT CONCLUSIONS.

l FURTHER OPTIONS MAY BE EXPLORED REGARDING IMPROVEMENT OF THE CATALYST MITIGATION METHOD (CLEANLINESS, ADDITIONAL CATALYST, f

ETC.).

I f

l BWROG SRV Setpoint Drift Fix Data as of April 18,1996 Page 7

ATTACHMENT 6 Will/ John, For your info. Sent to me by Mary Wegner.

John Boseman AEOD TECHNICAL REVIEW REPORT UNIT: Multiple TR REPORT NO.: AEOD/r96-02 DOCKET: Multiple DATE: April 15,1996 LICENSEES: Multiple CONTACT: Mary S. Wegner NSSS/AE: Multiple

SUBJECT:

TARGET ROCK TWO-STAGE SRV PERFORMANCE UPDATE

)

SUMMARY

Beginning in 1982, the NRC issued several information notices (IN) related to the 2-stage Target Rock safety / relief valves (SRVs) failure to open at the expected setpoint: IN 82-41, " Failure of Safety / Relief Valves to Open at a BWR;" IN 83-49, " Failure of Safety / Relief Valves to Open at a BWR - Interim Report;" IN 83-82, " Failure of Safety / Relief Valves to Open at a BWR - Final Report;" IN 86-12, " Target Rock Two-Stage SRV Setpoint Drift;" IN 88-30, " Target Rock Two-i Stage SRV Setpoint Drift - Update;" and IN 88-30, Supplement 1, "T. apt Rock Two-Stage SRV Setpoint Drift - Update."

The cause of the problem was identified as one or both of: (1) binding in the labyrinth seal area caused by tolerance buildup during manufacturing, or (2) disc-to-seat bonding caused by oxides of the disc and seat material forming a continuous film and inhibiting disc movement. Some time after 1990, the Target Rock 2-stage owners' group submitted two potential solutions: (1) the preferred solution was the use of platinum alloy disc to cause radiolytic oxygen and hydrogen in the valve to recombine, reducing oxygen available to cause corrosion and (2) the alternate solution was the installation of a pressure switch and control circuitty to operate the valve electrically when the set pressure of the pressure transmitter was reached.

On September 11,1995, an event occurred at Limerick Unit I which resulted in a stuck-open SRV and an extended reactor blowdown into the suppression pool. Pilot disc leakage was identified as the cause of the stuck open SRV. This event was especially noteworthy in that sludge and fibers in the pool, roiled by the blowdown, settled on the emergency core cooling system strainers nearest the SRV tailpipe, clogging them.

The NRC issued IN 95 47, " Unexpected Opening of a Safety / Relief Valve and Complications Involving Suppression Pool Cooling Strainer Blockage," IN 95-47, Rev.1, and Bulletin 95-02, " Unexpected Clogging of a Residual Heat Removal (RHR) Pump Strainer While Operating in Suppression Pool Cooling Mode" to address issues resulting from this event.

Attachment

DISCUSSION:

Background:

The original General Electric Company's (GE) BWR-4 design employed the 3-stage Target Rock safety / relief valve (SRV) for over-pressure protection for the reactor coolant system (RCS). In the early 1970s, the valves exhibited a tendency to lift for no apparent reason, and on occasions, to stick open. About half of the plants with the 3-stage Target Rock SRVs experienced this problem.

The solution to the 3-stage problem was the 2-stage Target Rock SRV, which used a 3-stage body, modified to accept the new top works. The plants that experienced problems with the 3-stage valve installed the 2-stage modification. In 1982, a problem with the 2 stage valves became apparent: failure to open at the expected serpoint. On July 2,1982, at Hatch Unit 1, all 11 Target Rock 2-suge SRVs failed to open at their setpoints of 1080 psig,1090 psig, and 1100 l

psig. Pressure in th: RCS rose to 1180 psig before three SRVs on one steam line opened and relieved RCS pressure.

Georgia Power, GE, and Target Rock initiated a study of the cause of the event at Hatch Unit 1.

Other utilities that had installed 2-stage Target Rock SRVs joined Georgia Power in an owners' l

group to look at the nature of the problem and its solution.

l The problem ms ultimately identified as one or both of two situations: (1) binding in the labyrinth seal area caused by tolerance buildup during manufacturing, or (2) disc-to-seat bonding i

i caused by oxides of the disc and seat material forming a continuous film and inhibiting disc l

movement.

i A material, PH13-8 Mo, whose oxide would not form a continuous film with the oxide of the l

seat material, was chosen for new discs. Trial tests of valve performance, after installing the new discs on about 50 percent of valves at each plant, were conducted. After a promising first set of test results, Ha:ch Unit I reported in October of 1989 that one of the valves with the PH13-8 Mo i

disc lifted at 10.54 percent above its setpoint and a second at 9.18 percent above its setpoint. In l

November of 1989, Bnmswick Unit 2 reported a valve with the PH13-8 Mo nu.terial dise lifted at 10.4 percent above its setpoint.

Subsequently, the PH13-8 Mo discs were replaced with the Stellite 6B discs, until an appropriate solution to the corrosion bonding problem could be identified.

The NRC issue d several ins which describe the foregoing events in greater detail. They are IN l

82-41, " Failure of Safety / Relief Valves to Open at a BWR", IN 83-49, " Failure of Safety / Relief Valves to Open at a BWR - Interim Report"; IN 83-82, " Failure of Safety / Relief Valves to Open at a BWR - Final Report"; IN 86-12, " Target Rock Two-Stage SRV Setpoint Drift"; IN 88-30,

" Target Rock Two-Stage SRV Setpoint Drift - Update"; and IN 88-30, Supplement 1, " Target i

Rock Two-Stage SRV Setpoint Drift - Update." IN 88-30, Supplement I was issued on February 2,1990, and discusses the Hatch Unit I and Brunswick Unit 2 problems occurring in 1989.

2

l t

i i

l l

Current Events:

The Owners' Group subsequently submitted two solutions which were to be pursued simultaneously. The preferred solution was the use of platinum (first as an inlay, also as a coating, and finally as a trace element in the disc alloy) to cause radiolytic oxygen and hydrogen in the valve to recombine, reducing oxygen available to cause corrosion.

Initial platinate discs were installed at Cooper, Millstone Unit 1, and Hatch Unit 2 in 1994.

Brunswick Unit I and Unit 2 are using a platinum-coated disc, first installed in 1993 on Unit 2 and 1994 on Unit 1. Pilgrim is using Stellite 21 for their disc material which they first used in l

1984. To date, setpoint test results involving SRVs using the platinate discs have been received l

from Millstone, Hatch, and Cooper. Also, results from Brunswick for as-found testing of their SRVs which have the platinum coated discs have been received. Figure I shows the results of the setpoint testing of platinate valve at each of the plants reporting. The shading indicates valve 1, 2, 3, etc.

I

% Deviation from Nominal

(

12 i

i 10 f>

l 4

l

?.

I j

o E

2-l 1

}

-4 witistone 1 MM 1one 1 E timit ti 2 Moac Cr eek L oo acr it'un.m ap 1* tPun.mith 2*

}

\\

Platin um -c o nte:I j

Mgure 1: Setpoint Test Results for Platinum Discs j

The alternate solution proposed was the " Pressure SwitchfTransmitter for Two-Stage Target Rock Safety / Relief Valve," NEDC-32121P. The Target Rock SRV is a dual function valve i

3

4 3

~

l which can be actuated by spring force to open when the system pressure reaches its setpoint (the Code safety mode), or it can be actuated by an electrical signal from a pressure sensor to its l

electro-pneumatic actuator. The actuator removes the spring force so that full steamline pressure lifts the pilot disc. This mode of operation was originally installed for manual operation (the Code relief mode) and automatic depressurization system operation (the nuclear safety-related i

l function). The pressure switch fix would extend the pressure sensor operation to the Code safety -

l mode. GE recommended that the pressure switch sensor setpoints be set at the same values as i.

used for the spring-actuated mode. The NRC found the proposed modification to be acceptable.

The topical report was approved on October 24,1995.

Some licensees made plant-specific submittals prior to the topical submittal. Hatch Unit I and Unit 2 and Brunswick Unit I and Unit 2 have installed the pressure switch fix in addition to changing the disc material and Millstone Unit I has committed to installing the pressure switch fix during the 1996 refueling outage.

4 i

About the same time as the platinate discs were first installed, another problem with the SRVs j

was identified, pilot disc leakage which could lead to a stuck open SRV as occurred in the September 11,1995, event at Limerick Unit 1. The stuck open SRV caused an extended j

blowdown into the suppression pool. Sludge and fibers in the pool, roiled by the blowdown, j

settled on the emergency core cooling system strainers nearest the SRV tailpipe, clogging them.

De Limerick Event i

Limerick Unit I and Unit 2 have 2-stage Target Rock SRVs unlike any other nuclear plant.

2 While the pilot stage is identical to those used by other nucles lants, the main stage is quite

}

different. Figure 2 compares the Limerick orientation to the R :_.tard orientation.

t e

l 6

Figure 2: Umerick SRV Orientation - SRV Orientation 4

The difference in orientation is thought to be the cause of the leakage seen at Limerick - main-stage leakage. Pilot stage leakage can eventually cause the main-stage to lift, but main-stage leakage can not cause the main-stage to lift. After the operability demonstration following the refueling outage in February of 1994, five SRVs were found to be leaking. Because of their history of main stage leakage, Limerick personnel assumed that the five leaking SRVs had main-stage leakage. Four of them did, but not the M-SRV.

l On September 11,1995, the M-SRV lifted I

while Limerick Unit I was operating at 100 percent power and remained open. The reactor was manually scrammed. First the M-SRV and the S-SRV were sent to Wyle Laboratories for examination and testing, then the remaining leaking SRVs were sent.

l l

The physical condition of the M SRV disc and pilot rod is shown in Figure 3. The disc was eroded 360'in the seat area such that the 6

nose separated from the remainder of the disc.

Slots were wom in the body of the disc in l

three places. The disc material was verified to be Stellite 6B by part serial number, material certification, and by Rockwell C hardness testing. The pilot rod was eroded l

and deformed as shown.

The stuck open SRV caused an extended blowdown into the suppression pool.

l The NRC issued IN 95-47, " Unexpected l

Opening of a Safety / Relief Valve and l

Complications Involving Suppression Pool l

Cooling Strainer Blockage," IN 95-47, i

Revision 1, and Bulletin 95-02, " Unexpected l

Clogging of a Residual Heat Removal (RHR) l Pump Strainer While Operating in l

Suppression Pool Cooling Mode" to address issues resulting from this event.

ANALYSIS Figure 3: The Disc and Pilot Rod From the The 2-stage SRV issues remaining are M-SRV

1) "setpoint drift" as the result of disc-to-seat I

corrosion, and 2) leakage past the pilot disc.

i S

The results of setpoint tests of 2 stage SRVs conducted since 1987 (begins near the beginning of the testing of the valves containing PH13-8 Mo discs) until the present are summarized in Figure 4.

l i

l 1

h.

I 1

i I

i 1

1 1

1 t

i i

a f

Pct Deviation f rom Nominal Setpoint_

7 l

d j

l s -

i st i

i

+

j iw t..

i t

a l

l, w

e v.-

i x

3 g.

g - -

j lA T

1 i

i

\\

/

A 1

l S'

'y I

s

t x

i,

.e o;

I i

.i i

1987 1'18d 1989

'%0 1991 1992 1943 199',

'+%

i j

........., l.. j.~

I

,,[,

I

{

i 1

s j

T

- j

.. j

. j

,. {L i

3 1. <

i i

r i. < l' i

I J.,!

l~.j

.~

1 i.

1 7..

+

j 1

s.

7 i

4 t

- t l.

i

)

,s f,.

. i

}

f

}

l i.,...

t

!_.l..

!E :

+, _ _....,,

.. ~

1 v..

u....,

+ ~..

l i

I Figure 4: SRV Setpoint Test Averages s

.. ~ - -. - _ - - - - - -

t l

l By design, setpoints for SRVs were a nominal figure such as 1100 psig

• 1%. The expectation i

was that the deviation from the nominal would be minimal and the average deviation would be 0.

l In fact, the deviation from the nominal has ranged from the minimal to values in excess of 67

{

percent, the majority being positive. The solid line in Figure 4 is the linear regression fit of average setpoints of all SRVs tested during the year. The average setpoint has trended upwards from about 2.5 percent above nominal in 1987 to about 3.5 percem above nominal in 1995. The various points represent the average setpoint for SRVs tested by a particular plant for the year.

l Of these 61 data points,7 or 11.5 percent were within 1%,29 or 45.5 percent were within

• j 3%, and 5 or 8.2% were 5 percent or more above the nominal. There are no negative setpoint test averages.

It was common to stop setpoint testing at or near 1250 psig when an SRV would not lift; however, the raw data shows lift points in excess of 1250 psig for some valves. For representation in Figure 4, all valves which did not lift at or before 1250 psig were called stuck and their setpoints were called 1250 psig whether they lifted or not and whether they were pressurized above that pressure or not. The tolerance was then calculated using that number. It should be noted that the SRVs in Figure 4 do not all have the same disc material. Figure 4 includes SRVs having Stellite 6B, PH13-8 Mo and platinate materials.

A review of the generic implications of significant leakage of Target Rock 2-stage SRVs (particularly pilot valve leakage)is being conducted by the NRC. Factors being considered are:

effects of significant leakage on valve operability, effects of operability testing on valve performance, risk significance of more frequent opening or multiple SRV opraing, adequacy of methods (tailpipe monitoring, torus heatup rate) for detection ofleakage for t..e range of BWR configurations, possible concems about increased suppression pool cooling operation, and what leakage limit monitoring and reporting requirements apply or mit needed. The staff plans to coordinate its review efforts with the Owners Group.

FINDINGS AND CONCLUSIONS The corrosion bonding issue is being addressed by the Owners' Group and being monitored by NRC. The updated information on overall setpoint testing shows a gradual increase in the average liftpoint of the SRVs. 'Ihis report also provides information currently available regarding setpoint testing results on platinate discs, but at least one full set of setpoint testing results are necessary to evaluate the test program. When sufficient information is available on the platinate test results to warrant an update, one will be issued.

The pressure switch fix has NRC's approval for the topical report, requiring a plant-specific submittal for the licensee to adopt it. Should the platinate disc experiment fail, the pressure switch fix is therefore ready for immediate use. Millstone Unit 1 is also installing the pressure

$ witch fix, alleviating concerns for the adequacy of overpressure protection at that site.

8