ML20090L521

Significant Deficiency Rept SDR5 Re High Pressure Svc Water Valve Weld Failure.Initially Reported on 730601.Vendor Weld Stress Analysis Calculations Reviewed & Table of Acceptable Weld Sizes Prepared

ML20090L521
Person / Time
Site:	Monticello, Peach Bottom
Issue date:	10/03/1974
From:	Boyer V NORTHERN STATES POWER CO.
To:	Knuth D US ATOMIC ENERGY COMMISSION (AEC)
Shared Package
ML20090L518	List: ML20090L521
References
NUDOCS 9102110311
Download: ML20090L521 (2)

Text

- i

, (

SICfl1PICAMT I:i FIC1F :CY ilF. PORT - SER NO. 5 ll1Gil PRFJ'CliRP, '" ulTCiO'ATr.R "Alf E UELD FAILUiG

}'EAC11 DOTTOM ATOIC POWFR STATI0ll - U' ITS 2 & 3 AEC C0!!STRUCTICM P?91GT NOS CFPR-37 Ai;D CPPR-3_S l

Dancription of Doriciency i Daring a routino walk-thru of Unit No. 2 plant by the licensees operating personnel, a 12 inch - 300 pound motor operated globe valve in the liich Proscure Servico Water line on the dischargo side of one Residual Iloat Rcmoval heat exchanger was discovered to have erperienced a wcld failure. The failure occurred betwoon the valvo yoko and the taotor operator mounting plato. The reason for the failure has been identified as insufficient fillet veld tluoat dimension caused by the instal 3ation of unauthorized shins batween the yoke legs and the mounting plate, which reduced the effective size of the veld.

t Correct.ive Action l The failed valvo is one of a series of eight valves (four in Unit 2 and four in Unit 3). These eight valves were visually inspected and a second valve sas found to have cracks in the yoke to motor operator mounting plate weld.1 All eight valves were returned to the vendor for rework. The rework involved elimination of the chim:., in the failed valve and the rewelding of the counting plates ,to the yoke legs with full pene-tration welds on all eight valves. i An investigaticn of similar valves (supplied by 'the same vendor) elsewhere in the plant, was undertaken. A total of 108 valves wore iden-tified by the vendor to havo yoko to motor operator mounting plate con-struction similar to that of the failed valve. Fif ty-eight (including the above mentioned eight) of those valves are nuclear valves classified as Group II as defined by Figure A.2.1 of Appendix A of the Peach Bottom Ato:ai.c Power Station FSAR. The remaining valves aro Group III non-nuclear balance of plant valves.

The Vendor's weld stress analysis calcalations were reviewod and a tablo of acceptable veld sizes prepared.

1 This valve was originally reported in the interim report to have chians. The valve was only visually inspected at that T,iro and the cracks were interpreted to indicato the presence of shima.

9102110211 PDR 74cto3 ADOCK 05000263 (R PDR m

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

y ,. mlN.,

q. %y

.; J. . >.

y ATTAcum T a t

. 1 Tys

. % .y

%c.

Q:c

9 w

.i 2 --

l WISCODSin Elecinc rom co?.wm & ggS1e gg' 231 WEST MICHIGAN,f.11LWAUKEE,WISC0fiSIN 53:01 yh.a ,,

, . o %n j

October 29, 1973 i

Mr. John F. O' Leary, Director ,

l Directorate of Licensing l

j U. S. Atomic Energy Commission ,

1 Washington, D. C. 20545

Dear Mr. O' Leary:

1 1

2 DOCKET NOS. 50-266 AND 50-301 l FACILITY OPERATING LICENSE NOS. DPR-24 AND DPR-27 l

POINT BEACH NUCLEAR PLANT BACKSEATING DISC MISLOCATION PROBLEM ON 2" DARLING VALVFS l -

In accordance with Section 15.G.6.A.3.b of tho-Technical Specifications for Point Beach Nuclear Plant (Facility 4

l Operating License Nos. DPR-24 and DPR-27), this report describes a possible generic' problem with a category of 2" gate valves in-l stalled at Point Beach Nuclear Plant. The valves in question are 2", No. S-350 WDD welding end, outside screw and yoke,' double l

disc gate valves 'with lip seals, and are manuf6ctured by the j Darling Valve and Manufacturing Company. The-valves used at Point Beach Nuclear Plant are safety class I, ASA series-1500 ,

lb. valves. . .

! An investigation of excess lei.down line leakage on l

September 15, 1973, lend to an inspection and subsequent repair-of valve 1MOV-1299 on Unit 1 (excess letdown system root valve) on September 26, 1973. Inspection of the valve disclosed that l

its downstream seat protruded from the valve body such that if the. valve disc was fully withdrawn from the guides, as allowed

by its backseating ring, the disc could catch the " lip" of-the

seat ring when reinserting. Four marks on the' lip of the down-

) stream seat ring indicated that the disc had caught there during

previous valve closings. Internal damage to . the va1..ve consisted

! of a fine vertical crack at the 12 o' clock position in the upper portion of the downstream seat ring. Two locating pins between i-the upstream and downstream discs of the split disc valve were j found to be slightly bent also and some facial scratches to the

! down-stream disc were evident. There was no metal loss involved in the damage. -

, l/,O n ') on () $,?

') #

$ 5.

4

. . Mr. John P. O' Leary October 29, 1973 1

1 1 -

, Repair of the valvo involved rounding the lip of the l seat ring to prevent future hangups of the disc. The thin ver-tical crack in the downstream seat could not be fully lapped out

,; during the repair. According1.1, a manual valve was added to the

! system downstream of 1MOV-1299 to bach up the root valvo. Valve 1MOV-1299 thereby remains effectivo and operable as a remotoly 1 controlled root shutoff valvo, but is considorcd not totally I capable of effecting completely tight shutoff without some

) through-leakage.

i At the time, measurements indicated that the location

, of the backscating ring on the valvo stem was too low but this 3 could not be assuredly determined. If such was the caso, this

! would allow the split discs to fully clear the seat rings when the valve was fully open and backsoated. The tendency for inter-4 forence to occur between the downstream disc and seat during I valve closing could be expected to increase if there was flow through the valvo, creating a differential pressure wh! h could j swing the loose hanging disc onto the lip of the seat.

There are six similar 2" Darling valves in each unit i at Point Deach t;uclear Plant. In addition to the above mentioned i 1299 valve, valvos 270A & B (normally open) are installed on the 1 reactor coolant pump seal return lines. These valves are. rarely 4

operated in the life of the plant. Also, valves 598'and 599 on the reactor coolant system drain line are of this type. These valver r.re never operated during normal pressurized and power oporation. The sixth,similar valve on each unit is MOV-427 on

, the normal letdown lino. The function of valve 427 is to close in the event of low pressurizer level and, in closing, cause tho l closure of the containment isolation valvoc 200 A, B and C, via an interlock. None of the Darling valves described in this re-port are containment isolation valves.

1

Valvc 1MOV-427 was investigated during a Unit 1 shut-

, down on October 13, 1973, af ter it was reported that it would not

. fully close remotely. Manual manipulation of the valve on Septem-

)

ber 28, 1973, had shown that at approximately one-half shut and i

again just prior'to closing, the valve operation became sticky.

5 Tests were conducted at that time to verify that IMOV-427 was capabic of performing its primary function of initiating an iso-lation signal for the letdown line. The slightest movement of the valve of f its backseat was found to be sufficient to activate the interlocks and close the AOV-200 lotdown isolation valves.

Measurements indicated that the discs of 1MOV-427 when U

e 4 0

]

PM r -~a v - , < r,- ---w , .n- v-~,- - , ~ . - ,.-v m - --< -nv,,--.- .,-,.n e--

d i . .

Mr. John P. O' Leary -

3- October 29, 1973 l

i i

backscated cleared the seat tings and left the valve open to simi-lar problems as e::perienced in 1MOV-1299. Inspection showed no I

damage to valve IMOV-427 other than a slight marking of the ' o~ ar edge of the seat ring, similar to that found in 1MOV-1299. before

! closing up the valve, the seat ring edges were rounded to aid in i

' guiding the discs down betwoon the coats. The " valve open" limit switch was then set for 2-1/4", 5/16" less than the maximum back-seating position of 2-9/16". Valve cycling tests were then con-ducted satisfactorily.

During the same shutdown, valve IMOV-270B was cycled manually with no evidence of stickiness or d.sc hangup. At the 4

completion of repair of 1MOV-427, on October 13, 1973, it was 4 concluded frca measurement; taken, operating experience and tele-phone discussions with the valve manufacturer, that, indeed , a dimensior. error could exist uith respect to backscat locations on the stem. Uith these confirmations, it was concluded that all twelve valves of this type would require investigation on a sche-j dule com.monsurate with the plant operating schedules.

1 Valves IMOV-1299, 2MOV-1299 and 2MOV-427 will be 4

electrically limited similarly to 1MOV-427. Valvo 1MOV-1299 will be completely changed out during a convenient shutdown fol-lowing the roccipt of a new valve. New valve stems'with back-scats located so that full opening of the valve will not permit the discs to loco the guido effect of the seats have been ordered i

and will be fitted in the remaining valves at convenient shut-j downs. The service of the 598, 599 and 270A & B valves is such

that it is not considered necessary to change the stems of these

• valves until the next refueling shutdo;.n of each unit.

The nuclear stean supply system supplier has boon in-formed about the problems encountered with these valves.

j ,

Very truly yours, i

i XD /

2:5 -

! Sol Burstein Senior Vice President 4

l 1

cc: Mr Jtmes G. Keppler ,

) Reg 3cnal Director

, , D3::cctorato of Regulatory Operations, Region III j . .

l .

e

<- . ,p-- -y . - -

., . - , ,nv -,m.-., -

,'.m

. ,,..ec, -c ., ..,e