ML20083Q039

Ao:On 721229,electromatic Relief Valve Failed to Reset Causing Primary Sys Blowdown Following Reactor Scram.As Result,Three Different Valves Failed to Operate in Intended Manner.Manufacturers & Valve Users Will Be Consulted

ML20083Q039
Person / Time
Site:	Oyster Creek, 05000000
Issue date:	01/17/1973
From:	Ross D JERSEY CENTRAL POWER & LIGHT CO.
To:	Giambusso A US ATOMIC ENERGY COMMISSION (AEC)
Shared Package
ML20083Q029	List: ML20083Q031 ML20083Q035 ML20083Q039
References
NUDOCS 8302220589
Download: ML20083Q039 (10)

Text

-

U b '

b s.

r8 9S a w..

m v,m Jersey Central Po.wer; & Light Company h/

%v M ADISON AVENU E AT PUNCH DOWL RO AD 9 MORRISTOWN, N.J. 07960 e S39-6111 January 17, 1973

,7n Mr. A. Giambusso /' . .i]'/ .

Deputy Director for Reactor Projects 3

Directorate of Licensing > -l+

h United States Atomic Energy Commission ., i Washington, D. C. 20545 '

igg JM 2 21973> 4

Dear Mr. Giambusso:

UW

\eq  :

. ': m

Subject:

Oyster Creek Station N /

Docket No. 50-219 Valve Malfunctions N)M ('rg 1 On December. 29, 1972, an operator opened a cabinet door carrying protective fuses for generator monitoring relays. As a result, a turbine trip was initiated immediately followed by an anticipatory reactor scram. At the time of this event, the reactor was operating at 1830 MWt.

As expected, the associated increase in primary system pressure was terminated by actuation of the electromatic relief valves. One (1) electromatic relief valve f ailed to rescat causing a primary system blowdown into the vapor suppression chamber (torus).

As a result of this event, three (3) different valves failed to operate in their intended manner. Each of these valve failures constitutes an abnormal occurrence and are included as Attachments' I, II, and III.

These failures can be summarized briefly as follows:

1) Failure of the "D" electromatic relief valve to rescat caused by a piece of material from the associated disc retainer be-coming trapped under the pilot seat.

2) Failure of main steam isolation valve NSO4B to close upon receiving an isolation signal due to a sticking pilot operated power valve.

3) Failure of isolation condenser NE01B condensate return valve V-14-35 to open due to excessive valve breakaway torque causing its motor operator to burn out.

Each of these situations were thoroughly investigated and corrected. In addition, due to AEC concern over valve operability problems currently being experienced by the industry, we conducted a complete surveillance test program on all valves related to reactor '

safety. - ~ I' 8302220589 730221 PDR ADOCK 05000219 PDR S

1 f l j

y ,

-- q

~

g g Mr. A. Giambusso January 17, 1973 Valves in thirteen (13) different systems (approximately 115 i valves in all) were judged to fall in this category and all but one (1)

!~ were satisfactorily tested prior to plant restart. One (1) torus to l' drywell' vacuum breaker valve did not meet our normal opening force

criterion and required some maintenance prior to plant start-up. This l deficiency is discussed in detail in Attachment IV to this letter.-

1 .~

Since more reliable valve operation is our desire at Oyster Creek and in response to Directorate of Regulatory Operations request, we intend.to give the matter of valve surveillance frequency and sur-l veillance testing adequacy, additional ~ review. All valves related to j reactor safety will be included in our review. Consideration is being given to a meaningful inercase in surveillance frequency. We will evaluate the method of surveillance testing to assure it is the best-possible means available to determine valve operability. ,

Discussions with appropriate manufacturers and valve users will be pursued to benefit from their experience in these areas. We will continue to keep your office advised of our efforts in this area ,;

in a timely manner. .!

We are enclosing forty copies of this report.

Very truly yours, e W- ~

Donald A. Ross Manager, Nuclear Generating Stations DAR/mp Attachments (I, II, III, and IV)

Enclosures (40) cc: Mr. J. P. O'Reilly, Director -

Directorate of Regulatory Operations, Region I a l

i I

I l

y? .

o .

(o;

.-ATTACHEMNT I Failure of the "D" Electromatic Relief Valve to reseat af ter performing its protective function is considered to be an abnormal occurrence as defined in the Technical Specifications, Paragraph 1.15.F.

On December 29, 1972, with the reactor operating at 1830 MWt thermal, a turbine trip was initiated which caused a reactor, anticipatory scram..

The turbine trip caused an immediate increase in reactor pressure which was properly terminated by automatic actuation of the Electromatic Relief.

Valves upon reaching the 1070 psig setpoint. The "D" Electromatic Relief Valve failed to reseat, however, after the design blowback of 22 psi _ was achieved.- Upon recognizing that the blowdown had not terminated when the main steam isolation valves closed, checks of the Electromatic Relief Valve temperature sensors were made. These temperature indications verified that the relief valves had operated. The fact that one valve-had not reseated was confirmed by hearing steam collapsing.in the vapor suppression 1 tham'ber. #7 In an attempt to determine which relief valve had not reseated, each valve was individually cycled. When the "D" Relief Valve was selected to open, no change in noise level was noted, thus indicating it was already open.'

Several unsuccessful attempts were made to reseat the "D" Electromatic Relief Valve by operating its control switch.

This event resulted in a primary system cooldown rate of 1580F per hour for the first hour following the scram and rates of 27 F per hour and 37 F per hour for the second and third hours, respectively. Also, as a result of this event, approximately 50,000 gallons of reactor water in the form of steam was blowndown into the vapor suppression chamber.

Upon investigation, the pilot valve for 'the "D" Electromatic Relief -

Valve was found to be open, thus causing the main valve to remain open as long -

as pressure remained on the. main steam line.

Figure 1 attached identifies the valve parts under discussion.

Further investigation revealed that a piece of thread material from

, the disc retainer of the main valve had blown.through the pilot valve tube. '

During the valve actuation it'became lodged between the pilot valve seat and disc, thus the pilot valve was not able to fully 'close which, .in turn,'-

maintained the main valve in an open position.

It was also noted that the tie wire, lock screw and lock arm were missing from the disc-retainer. Since the lock arm and screw were missing from the "D" valve, the remaining 4 relief valves were removed for inspec-l tion. As a result of this inspection, it was found that a small tab from the lock arm of the B" relief : valve had also broken off. . Similar parts e for the other 3 valves were found to be in place and secure. In addition, it was observed that the valve' operator's spring guide for the "E"~ relief

valve was misaligned, thus causing the solenoid plunger to be restricted l lin its upward travel.

t l The safety significance-associated with this event centers around-ll the blowdown event. The attached tabulation develops, the actual primary _

!! system temperature / time relationship.-

I i, .

![ '

s ,

gs ,

L .) ()

l if i

i SOLEMot D EYheoE fo brus 21 opggggy.og 22 013C. 4 20 -

2 g gg i i '

s N ,

N s 23 IB-oPERATIMq LWER 5 q 3'

, \ C w .

~ IU M l 015c. RETNUER-6 { p LCfStug sCRE'd-8 ,

,s .

[.

,.j '

n t_ %~

' ~'

' " 2~ EC"WnJG WK- PL A'rE - 7 '" #ig b- Paor 13- Disc.

' v N _ NET 10 q

+ ,

I

• b  % -

E g Steam lint _ A N e pecssucc. 25 l 28 l $ 21 9

• 28

( MAIM PILOT VALVE VALVE E LECTROR\ ATlc., RELIEF \/ALVE o

FIC3UR E 1 i .

I-I c .. . k

]

.o .

ci  :

\

e4 Primary System Temperature / Time Relationship

{

"0

$a $

M Pressure Sat. Time N Time Time Psig F YN

} o l

Oe O 608 -

1000 j546 I

\

,4,, -4020F/hr 610 2 min *660 - . ,

499

-158 F/hr 618 10 550 479

-132 638 .5 hr 355 437

-98 708 1 200 380

- 26 738 1.5 170 375- --2 7

- 28 808 2 140 361

- 28 838 2.5 115 347 -37

- 46 908 3 80 324 ,

-50 1000 4 30 274

-15 1008 5 20 259

-0.

1208 6 20 259

. I

, . . , ...s. , .. ,

+

..o .

o '

f ATTACHMENT I Page 2 .

This event nearly approximates the' blowdown case analyzed in Report:

y 3

Number CENC-ll43, " Analytical _ Report for Jersey Central Reactor Vessel".'

The actual temperature transient analyzed is as follows:

i 1

(1) 10560F/hr. for .17 hours1.967593e-4 days <br />0.00472 hours <br />2.810847e-5 weeks <br />6.4685e-6 months <br /> (546 F to 366 F) .

(2) 00F/hr. for 1.62 hours7.175926e-4 days <br />0.0172 hours <br />1.025132e-4 weeks <br />2.3591e-5 months <br /> (3660F) 1 1

(3) 1000F/hr. for 2.77 hours8.912037e-4 days <br />0.0214 hours <br />1.273148e-4 weeks <br />2.92985e-5 months <br /> 0 (366 F to 900F) ' +

i 1

Average - 1000F/hr. for 4.56 hours6.481481e-4 days <br />0.0156 hours <br />9.259259e-5 weeks <br />2.1308e-5 months <br /> (5460F to 900F) j According to this report, which is-the basis for Amendment No. 16 of the FDSAR, the reactor head studs are 'the most limiting components for the -

blowdown event, being allowed 308 blowdown cycles. The next most' limiting component (non-replaceable) is the support skirt which is allowed 780 cycles.

j The most limiting reactor components considering a 3000F/hr. cooldown from 546 F to 100 F are again the head studs (280 cycles) and the support skirt

' (780 cycles - based on the more severe stresses caused by the 100 F/hr.

cooldown from 546 F to 1000F). Therefore,no unreviewed safety significance is associated with this event other than counting it as a thermal cycle due ,

} to a blowdown. Concerning the failure of the "D" Electromatic Relief Valve,- ,

the safety significance of this event has been previously analyzed in Amend- '

ment No.15, Section Il of the FDSAR (Primary Containment Design Report).

In order to prevent recurrence of lock arm failures, these - components -

were removed from all relief valve assemblies. A new-method of preventing i

disc retainer rotation was installed. This method consists of pinning the disc retainer to the cage. Dresser Company, manufacturer of the valves, .:

was contacted concerning this repair and approved same, taking'into condider-l ation that the installation of a set screw can in no way 'cause the clearance- -

between the threads to. increase. All other lead threads on each disc retainer ;

l'

' were examined and feathered edges removed as necessary.. The spring guides for the "E" valve were realigned and the solenoid plunger was checked for .,

freedom of movement.

4 The cap screw from the "D" Valve was found lodged in the seat of the.

30" header Main Steam Drain Valve, V-1-65. Attempts made to locate the  !

remaining missing parts were unsuccessful. Discussions-were-held with

[ General Electric, I&SE Representatives, which indicated that th'e strainers

i. provided ahead of the turbine main steam stop valves would trap a piece the size of the missing locking plate. The decision has~.been made to inspect-

i. the main steam stop valve strainers for the missing lock plate' during the i: _ next refueling outage presently scheduled for.this Spring.. ,

l Prior > to putting the plant back on the line, a hot functional test i of each of the five Electromatic Relief Valves was performed. No difficUli ties were i experienced with the operation of these valves during this test.

1

}. -

p ,

I

r

~

a

', -("f . ("%

, . .- u ,

v

'i

' ATTACHMENT I.

Page 3

, In addition to those measures already taken to prevent a repetition of this l event, a review of the operating procedures associated with an inadvertent blowdown is being made. This procedure will be revised as required, if found inadequate.

We intend to investigate with the manufacturer the history of prob-lems associated with this type of Relief Valve and obtain his: recommendations for possible modifications which could improve this valves operation and reliability. (

, a)* .

l l

j i

j.

l i

f

< - J 7, . .- Q -

OL  ;

ATTACHMENT II q

l Failure of NSO4B, Main Steam Isolation Valve to close upon receiving -l an isolation signal is an abnormal occurrence as defined in Section 1.15'.F. l The reactor mode switch key was broken off in trying to switch to the " Start-Up" mode following the reactor scram. This allowed reactor pres-sure to reach 850 psig with the switch in the "KUN", mode and caused closure of the MSIVs. Af ter the isolation, it was discovered that. NSO4B, the south outside isolation valve, was not closed. Its switch was placed in the

" CLOSED" position and the valve still did not close. The switch was then, cycled "0 PEN" and " CLOSED", whereupon the valve closed.

The valve closure problem was determined to be caused by a sticking pilot operated power valve. This power valve is manufactured by N umatics, Inc. (reference Numatics print 91746) and modified by General Electric ,,

Company, FDI 322. <

The AC and DC Solenoid Valves wera checked and operated satisfactorily.

The pilot operated power valve was found to be sticking. A new pilot operated power valve was installed and tested satisfactorily. Similar valves were in-spected on the remaining three (3) main steam isolation valves and found to' be in satisfactory condition. The sticking valve was dismantled and examined.

A small amount of fine red dust was found on the sleeve "O" rings. The piston was removed and cleaned, and although it showed no visible sign of wear, it still exhibited a tendency to stick at one end of its travel.

Each of the two main steam lines are provided with two -(2) isolation, valves in series, one inside and one outside the drywell. As notedLin Amend-ment 65 to the FDSAR, a modification was made to each of the pilot operated power valves during the Fall 1971 outage, increasing the piston and cylinder clearances. No difficulties had been experienced with any of the valves, since that time,until a closure problem occurred with NSO4B on December 3, 1972 prior to a reactor startup.

Ihis difficulty was determined at that time to be a result of a mal-

- functioning control switch which was replaced.- The valve subsequently. cycled.

properly. As discussed with our DRO I inspector, subsequent review-of this earlier event in view of recent findings would indicate the switch did not contribute to the problem;however,it was a sign of difficulty with the pilot-operated power valve.

To prevent reoccurrence of this event we intend to investigate a mod-ification which would permit daily exercise tests to be accomplished by using ;

the main power operated pilot valve. The valve manufac*"rer will be contacted L with regard to instituting a preventative maintenance program on these valves. .

b The main' steam isolation valve full closure test program will continue:

to be followed, i.e. , testing the valves whenever! sufficient' load reductions on plant occur, but with a minimum frequency of 'six weeks between tests rather-than the present Technical Specification frequency of: three months if at power..

We will factor into-the present study on the long-term' suitability of the- ' l 1 main steam isolation valves, the problems associated with -power operated

_ pilot' valves which have.now been experienced.

I 1

U

m s C

~

1 . - . -

ATTACHMENT III

". The failure of -Isolation Condenser NE01B Condensate Return . Valve V-14-35, to open when manually actuated during a plant cooldown is con-sidered to-be an abnormal occurrence as defined in the Technical Specifi-cations, Paragraph l.15.E.

Following a partial depressurization of the primary system,it was:

desired to place the "B" Isolation Condenser in service to assist with the final reactor system cooldown. The control switch for the Condensate Return Valve, V-14-35, was placed in the "OPEN" position ~and left there-for approximately 15-20 seconds. However, when no opening indication was '

evident, the control switch was returned to " AUTOMATIC". At this time, the green " CLOSED" indicating light extinguished indicating that the valve motor supply breaker had' opened.

Knowing. that the control switch- was in the "OPEN" position for 15-20 seconds and then receiving indications that the 70 amp motor supply breaker opened, verifies that an opening signal was applied to the valve motor.

Apparently, the motor could not supply sufficient torque to break the valve

~

away from its seat. Upon investigation the valve motor was found to be burned out. The valve was then operated manually at which time some.

difficulty was experienced in breaking the disc free of its seat.. Once broken off the seat, the valve stroked fully open and fully closed freely ~

by hand. A new motor was installed and operation of the valve checked with no adjustments made to the . torque or limit switch settings. The-motor oper-ator preformed properly. The valve was then ec=pletely ~ dicessembled and!

thoroughly inspected, along with its associated limit torque operator. .The '

only flaw observable during this inspection was a small. burr on the stem which was considered to be of no significance. The valve was then com-pletely reassembled with new motor thermal overloads and motor power supply breakers being installed. The' valve vas then operated satisfactorily.

A complete set of motor currents were obtained for future reference.

The requirement for at least one of the isolation condensers - to act as a means for heat removal is detailed in Amendment #67 to FDSAR. .

. In this case, however, the "B" isolation' condenser would not have been able to' fulfull that function and, therefore, the safety significance of this h event is that the isolation condenser system did not have its intended re-dundancy.

To prevent reoccurrence of this event we plan. to institute a means to obtain more accurate motor amperage data while performing the valve t e operability test and also to' check the repeatability-of the associated torque switch settings. ' This will be done ~on an increased. testing frequency

!- of once per week in order to obtain the additional motor current data.

, Presently the surveillance testing of this component has been required on j a monthly basis.

ll H i-  :

. 1 4 , u j .6 ~

, j --

N ,

t

} . ,j L[ ' i-

t

' ~

- m v

(J ,

ATTACHMENT IV The failure of one drywell-absorption chamber vacuum breaker to satisfy our opening test criteria is considered an abnormal occurrence as defined in Technical Specification 1.15.E. This event was reported to the Directorate of Regulatogr0perations - Region I by telephone on January 8, 1973.

The fourteen (14) drywell-absorption chamber vacuum breakers which are normally tested during each refueling outage, were part of a list of

valves being tested prior to plant start-up. Testing is accomplished by the use of a spring scale to measure the amount of force necessary to open the vacuum breakers and the force necessary to hold them open. Valve V-26-8 failed to meet the maximum allowable one.ning force criteria and also exhibited a tendency to stick partially'open~ Another valve V-26-9 also showed a tendency to stick partially open when allowed to close. The pack-ing was removed from both valves, inspected, and reinstalled. Adjus tments l were made to both valve packing glands allowing both valves to operate satisfactorily.

The safety significance associated with this situation is as follows.

! The Drywell Absorption Chamber Vacuum Breaker System is required to prevent water oscillations in the downcomers due to low steam flow rates in the downcomers and to provide protection against negative pressure conditions in the Drywell portion of the containment vessel. The vacuum relief area to vent area ratio used to size the vacuum relief valves for Oyster Creek was found to limit water oscillations in the downcomers to about two (2) feet from the Bodega Bay tests. Allowing one valve to be inoperative reduces the total vacuum relief area by only 7% based on interpolation of Bodega data and such a reduction would increase the oscillations by only about one-half foot, which is negligible. Only about 25% of the available vacuum relief capacity is required to protect against negative pressure in the containment.

f To prevent reoccurrence of this event we intend to investigate the

. suitability and application of the present packing material for low pres-sure, high humidity environment.

l l

l

{

l l

L _