ML20084H159

Confirms 700925 Discussions Re Operational Problems of Turbine Hydraulic & Mechanical Control Sys.Servovalve Filtering Sys Cleaned,Filter Changing Procedure Developed, Two Potentiometers & Electric Pressure Regulator Replaced

ML20084H159
Person / Time
Site:	Oyster Creek
Issue date:	10/08/1970
From:	Finfrock I JERSEY CENTRAL POWER & LIGHT CO.
To:	Morris P US ATOMIC ENERGY COMMISSION (AEC)
Shared Package
ML20084H162	List: ML20084H159
References
03847, 3847, NUDOCS 8305020045
Download: ML20084H159 (5)

Text

9 ,*

~ ;&&. kl{ Q Q i

,. k .

>. p t,:

Jel scy Cen tral PowoM & Ligla t Com pany a

M ADISoN AVENU E AT PUNCH HOWL Ro AD e MoRRISToWN, N. J. 07960 e 539 611 t 4

October 8, 1970

) 1

'T'--

l Dr. Peter A. Morriu, Director Division of Henctor Licensing

-United States litomic Energy Commission <

Washington D. C. 205h5 l

Dear Dr. Morris:

For Div of ' Compliance

Subject:

Docket No. 50-219 l Oyster Creek Unit No. 1 l

l The purpose of this letter is ta confirm recent discussions held with Mr. R. McDermott of the Division of Compliance--Region I, regarding operational problems experienced at Oyster Creek, particularly l the turbine hydraulic and mechanical control system.

The reactor had been operating since May 1970, without any significant operational problems until September 16, 1970, when the reactor was shut down because of an increase in the rate of unidenti-fled leakage in the primary containment. The Technical Specification j limit of 5 GPM had not been reached prior to shutdown. After correcting -

i valve packing leaks, the unit was restarted. The generator was.on the l line at 3:50 p.m. on September 17, 1970.

At 11:01 p.m. on September 17, 1970, the turbine tripped while carrying a load of h00 5Ne as the result of high level in the moisture '

separator drain tank. The reactor scrammed on high flux and all safety systems functioned as required.

The event of the turbine trip was preceded by an oscillation of approximately 2 to 5 MWe in generator load. Steam flow began to fluctuate and reactor pressure decreased slightly. The generator had been operating at 530 MWe, approximately 1600 MWt, and load was reduced-i to h00 MWe by recirculation flow, when the oscillation ceased. At this point, the turbine tripped.

_Upon occasion,'the turbine control valve cams have been known l to contribute toward an oscillation in load at high valve opening posi-

'S I .N tions due yo the control valve loop gain being higher at these positions, l\ \ ~

0 y uem a' rl a gg, Nid .g g x o,qs47

- bb OO4h703oog NT REGION _ OCT 141970 9w p a aoocx o:,ooo m .

a m = _ ena 3 =. -

+ _ ;2

m A (g

'.' ~.

Q Dr. Peter A. Morris Page II October 8, 1970 Up to the valve position associated with approximately 500 MWe, the loop gain is constant. Ilowever, oscillations can occur above 500 MWe if a perturbation, such as a load swing or pressure spike, were to oc cul- at the higher valve open positions. The perturbation which precipitated this event was the result of load swings brought about '

while backwashing the main condensers. The remedy has been to reduce the load to a more stable can position, eliminate the oscillations, and recover to the desired electrical load. In addition, until the cams are replaced, operations which may cause an upset are performed Eh a lower load.

Additional operating problems involving'the turbine hydraulic control system have resulted from fine particles of foreign material in the turbine oil system. This is most noticeable on a restart of the turbine and requires the replacement of the 0.5 micron paper fil-ters in the oil supply to the electric pressure regulator servovalve (moog valve) at least 3 to 4 times after a start up. The interval of change increases as the turbine continues to operate. Any foreign material which gets by the filtering system can cause erratic operation or eticking of the servovalve. The servovalve positions the output piston through the control of hydraulic oil flow. The output piston acts on a torque chaft which is common to the mechanical and electrical pressure regulators and the bypass opening jack.

The remedy for this event is to switch pressure regulation to the mechanical pressure regulator, after reducing the load to approxi-mately 450 MWe because of the previous mentioned cam problem, remove and clean the moog valve, restore it and the electric pressure regulator to service, and return to the desired electrical load.

T On September 20, 1970 at 2:40 a.m., dirt in the moog valve .(3) caused an instantaneous load spike upward of approximately 20 MWe, and upward steam flow spike of an indicated magnitude of 800,000 pounds per hour, a slight pressure decrease, a downward flux spike of approxi-mately 10%, and no apparent change in level. The reactor had been operating at licensed power.

On Sejtember 21, 1970 at 1:35 a.m., with the generator output (f) y at 500 MWe, another spike occurred. Electrical output increased to 550 MWe and decreased to 48u MWe before returning to 500 MWe. The other parameters deviated from normal as follows: Steaa flow up 10 6 lb/hr --

down 600,000 lb/hr, pressure down 8 psig -- up 9 psig, flux down 15% --

up 12%, water level -- minor fluctuation. At 9:22 a.m., load was reduced from 460 MWe to 390 SNe to 359 MWe because of vibrations experienced in

\

L -

. =.

o-- . -

t

- A.

s  % n ,

(h O N.

s.-

Dr. Peter A. Morria q3 Page III -

October 8, 1970

• x

_the turbine lute oil pipfng. Electrical load, pressure, steam flow, <

1 level, and fe6dwater flow were slowly varying. d'he cause was an oscil-lation of the regulator torque bar sficulated by sluggish operation of the servovalve. The oscillati'o.i was stopr.d,dd the entire system returned to steady-state conditions.

x

-In each of the above events, the servovalve was found'to b9 dirty. On the second event, the pria4ry servovalve internal 20 micron filter was plugged. An investigatioll was begun to determine the mecth v nismbywhichforeignmaterialwasbypassingthe.j\.5mic,onpr,<rfilten -

At 9:39 a.m. on September 22, 1970, while, operagGg%;t 500 Mie, g the reactor scrammed as the result of the closure of the Iuain' steam lica.

isolation valves caused by a low pressure in the main stcam line. All [

reactor systems functioned as required. The cause of this operation was a plugging of one or the nozzles in the servovalve torque mqtr, which I' caused the turbine control valves to open, slightly decreadkrig' reactor pressure. The operator reduced reactor power by recirculation' flow in -t an attempt to switch to the mechanical pressure reguhtor. This further .

reduction in power and pressure cau md the main steam line '

isolation -

valve closure. g 'I )

q i t ,-

s '

The following is a summat$oh of the corrective ad., ions taken "

by plant personnel under the direction of a General Electric representa- i tive as a result of this incident: -

The complete servovalvc filtering system was inspected, flushed, and cleaned. A new 0 5 micron filter.was installed in series with the existing filters immediately before the servavalve.

Inspection of the prefiltes ' dis'elosed a bypass flow path, resulting from a missing rubber gaskel en the prefilter.and pinched sealing rubber gasket on a second filt'en inc.he prefilter train. A filter changing procedure was develope:l,A nddill maintenance personnel have been instructed in the proper m g 4 of chaj.e,ing filters.

s ,

The servovalve was c'.eaned, and a new torque motor assembly installed to replace the motor with the plugged nozzle.

The electrical portion of the regulator va4 inspected and all connections checked. Proper response time of th'd ciriovnive was verified, and the electric pressure regulator was checked for foise using an oscil- '

loccope. No spiking or noise was noted in theJaplifier system.

On a routine start-up check of turbine valg'e operation on the norning of September 23,.1970, it was discovered thgtlyu mait: steam ,

bypass valves would not open. A check of the mecha,nidtQ :linkageheua y i

M3 ,,

< M( g2 4J l ' t' i S .

-a. J .%. -u . ,

_Q _  ;

o, e m m ss .s .

t

\

y 1

O v v Dr. Pet,er A. Morric Page IV october 8, 19'(0 to operate thene valveu uncovered a broken aluminum puuh-pull ro 1.

The bar van not, completely nevered; however, it could not operate in the puuh direction (valve opening). The linkage mechanium had operated at the time of the ubutdown on Sept, ember 22,19'(0.

The exact cause of failure of the rod van not determined.

It 13 hypotheuined that pouaibly the rod hal been at,epped on or a

{ chain holut hung from it during conutruction. (The rod 13 mounted horinontally on a upun or approximately 12' 6"). Metallurgleal xamination of the euI will be completed na uoan au the rod la made available to Generul Electric for touting. Thiu rod and the control valve crunu will be replaced during .the week of Get.ober 16, 19'(0. The dunaged rod was repaired by incerting a hollow ut, eel aleeve incide t,he original bar, fautened with boltu.

. General MLectric was requeuted to review thin incident, and reported that the original deuign wau adequate and that the rel;alr t was atronger than the original deuign. The PoitC and GORB reviewed the incident uid t,he repair.

The fact wau eenuidered that the bypana valveu would not have operated if require l on a turbine trip. Thlu uituation had

.')een analyued previously; and therefore, vau not conaldered an an uneuviewed cafety queution. With all the work that could be ac-compliched durin;', unut.down, the final checking had to be made during operat, ion . The unit wau reutarted and the generator was on the line at 8:00 a.m. on September 24, 19'(0.

The aummary of evento discuuued uo far wau reviewed by General Electale and Jeruey Central with Diviulou of Compliance peruonnel on Coptember 25, 1970.

Operation at load with the electric preucure regulator in Jervice indicated uome additional problema not attributed to enmu or dirt, in the servovalve. The firut of thene problemu wau the initiation of a procuure act point change without operator action.

~i Tie corrective action taken van to replace two wire wound poten-tlometern (cet point preuuure control and net point meter control)

(

with compouit, ion type potentiometeru.

Aluo, reuistance and ground checku were made on all inputa to the electrie presuure regulator. A 30,000 ohm ground vaa removed from the preauure uenaar linear variable differential transformer.

The 2 tunplifieru in the electric preuuure regulator were replaced

'\ with new aniplifiers of the unme design.

p

f -^ a - -. ,m

..n.

. O O I>r. Peter A. Morria Pude V october 8, 1970 l After final checking, the EPlt was placed in cervice the j urternoon of September 30, 1970 and has been functioning saticfactorily i uince that time.

1 i

Very truly yours,

,'y j).1/i  ?' *:S ,

l Ivan 11. Fin {fock,fJr.

! Manager,:Iuelear Generating Stations l

IhP/pk I

i ec: I!r.11. W. Kirknan, llegional Director l Divicion of Compliance f

I f

i 1

l i

?

l \

< _J