ML20084U407

Ao:On 721023,primary Coolant Leakage of 28 Gpm Exceeded Tech Specs Rate of 10 Gpm.Caused by Seal Mechanism Failure of Control Rod Drive Mechanism 19.Seal Assembly Replaced.All Motor Packages Inspected & Replacement Procedures Modified

ML20084U407
Person / Time
Site:	Palisades
Issue date:	11/02/1972
From:	Sewell R CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
To:	Oleary J US ATOMIC ENERGY COMMISSION (AEC)
Shared Package
ML20084U408	List: ML20084U407
References
NUDOCS 8306290016
Download: ML20084U407 (4)

Text

g r

i  :

(a) .. . v

,c- 1 fr '% ConSumBIS h 0 Power

%' iyo.4 nk Company

_$ General Offices: 212 West M chigan Avenue. Jackson, Mic hsgan 49201. Area Code 517 7HR-OSSO Ncvember 2, 1972

'j Mr. John F. O' Leary, Director Re:j9 Docket No 50-255 Directorate of Licensing N License No DPR-20 US Atomic Energy Commission [/

Washington, DC 20545 . N/

_L 7

Dear Mr. O' Leary:

This is written to apprise you of an abnormal occurrence that occurred at the Palisades Plant on October 23, 1972, involving the failure of the seal mechanism of control rod drive mechanism (CRDM) No 19 The

/ failure of this seal mechanism allowed a primary coolant system leakage of 28 gallons per minute, which exceeds the Technical Specifications

• allovable leakt ge rate of 10 gallons per minute, At u.: time the first indications or seal leakage were observed, the plant was generat'ing 405 MRe. The first indications of leakage from the seal of CRDM No 19 followed partial movement testing performed to fulfill Technical Specifications requirements. After receiving the in-dications of the leakage, CFDM No 19 was exercised several times in an attempt to reset the seal. This exercising procedure has been used with some success on previously experienced seal leaks which have all involved very small amounts of seal leakage. During the exercising of CRDM No 19, the seal leak off temperature continued to climb, indicating increasing seal leakage, and went off scale at 4500F. At this time, the primary system leak rate was determined to be 28 gallons per minute. At approxi-mately the same time, both the primary and secondary rod position indi-cations for CRDM No 19 failed. An orderly shutdown was initiated and the plant was cooled to the cold shutdown condition. Reactor coolant water level was lowered to a level below the CRDM seals and the leakage stopped.

The Palisades Plant CRDM seals are composed of three basic parts: collar, rotating seal half and stationary seal half. The collar rests on a shoulder of the CRDM shaft and is keyed to the shaft. In addition, it is held by three 1/h" allen head setscrews and these set-screws also transmit the motion of the collar to the rotating seal half.

The rotating half has a skirt which extends over the collar. The pro-trusion of the collar setscrew is accommodated by three slots in the sides of the skirt. A carbon seal surface is mounted at the top of the rotating seal half. The rotating seal half is sealed to the shaft by

$D' P1 8306290016 721102 d kO54 gDRADDCK05000g

r-O -

. O Mr. John F. O' Leary 2 Docket No 50-255 License No DPR-20 November 2, 1972 means of an 0-ring. The stationary seal half is mounted in the housing above the rotating half and an 0-ring provides a seal between the station-ary half and the seal housing. The seal halves are maintained in contact by springs in the rotating seal half which acts against the collar.

Disassembly of CRDM No 19 seal assembly revealed that one of the three collar setscrews had backed out and lodged between the rotating seal half skirt and the seal housing. The setscrev was destroyed and the rotating seal half skirt was heavily marked. The lodging of the setscrew caused the rotating seal half to cock, allowing the excessive leakage between the seal faces and severe erosion of the carbon seal face. When a seal leak occurs, the leakage path is in an upward direction between rotating seal half skirt and seal assembly, then horizonal between the seal faces and then upward between stationary seal face and CRDM shaft to the seal leak off cavity.

The collar setscrews are locked in place by staking (upsetting the collar and screw thread metal junction with a punch). The setscrew that backed out had not been properly staked.

A vapor seal is installed to prevent leakage from the CRDM mechanical seal leak off cavity to the motor package. . This seal has been bench tested in the cold condition to 1000 psig and has experienced temperatures up to 350 F under operating conditions without failure. The seal leak off cavity normally experiences temperatures of about 1500F to 2000F and atmospheric pressure. The seal was destroyed, allowing primary l coolant to escape along the motor shaft to the motor package as well as '

the normal leak off path, the drain line from the seal leak off cavity.

Most of the steam and water primary coolant mixture that leaked into the CRDM No 19 motor package escaped through the motor package inspec-tion covers. As a result, approximately 50% of the reactor vessel head insulation was vetted. Several CRDM motor packages, junction boxes and electrical connectors showed signs of moisture and/or boron salt de-posits. In addition, both channels of start-up range instrumentation failed during the plant shutdown, most likely due to moisture but as of this writing this conclusion has not been proved or disproved.

The reactor was shut down by driving all CRDMs into their lower electrical limit except No 19 CRDM No 19 was left at an indi-cated position of 128" because it was concluded that either the position indication had failed or the CRDM vould not move. On the day after the leak was detected, attempts were made to drive CRDM No 19 to the full-in position. The primary position indication indicated an inward motion to 120"; the secondary position indication was meaningless. At this time, it was assumed the drive was stuck. However, during disassembly, it was determined that CRDM No 19 was fully inserted and the position indication ,

had been faulty. It is readily apparent that a CRDM will not move as i freely with a collar setscrev lodged between the rotating seal half J skirt and the seal housing as a normal CRDM; however, it was still oper- c' able . This condition is immediately apparent because the setscrew causes the rotation seal housing to cock, allowing gross leakage between the seal faces, which is indicated by an excessively high seal temperature.

n v -

. (v3 Mr. John F. O'Le!.ry 3 Docket No 50-255 License No DPR-20 November 2, 1972 All CRDMs were trip tested from the lower electrical limit on October 26. The clutch halves failed to separate on four CRDMs. Inspec-tion revealed the presence of boron salt and moist tre inside the motor packages of these four CRDMs. This is the first time any CRDM motor packages have shown any signs of boron salt deposits or moisture, with the exception of CRDM No 17 in February 1972 (refer to our letter dated February 14,1972).

The following is a summary of corrective action currently in progress or that has already been completed:

1. The seal assembly for CRDM No 19 has been replaced.

2. All CRDM seal assemblies are being inspected. All collar setscrews found not properly staked are being properly staked.

3 The seal replacement procedure has been modified to pro-vide for inspection of the collar setscrew.

h. All CRDM motor packages have been removed and are being inspected and cleaned as necessary.

5 The clutch spline areas on all CRDM are being vapor blasted and relubricated as described in our letter of September 26, 1972. The CRDMs vill be trip tested as described in the September letter prior to returning the plant to service.

6. CRDM motor package inspection covers, terminal boxes and connectors are being waterproofed to reduce the possibility of in-leakage in the event of a similar occurrence.

7 The detector for one start-up range channel has been re-placed and the channel returned to service. The removed detector has been returned to the vendor for further investigation into the cause of its failure. The detector in the second channel vill be replaced as soon as parts are received from the vendor.

8. All insulation has been removed from the reactor vessel head and the vessel head cleaned. The head will be reinsulated.

9 Plant operating procedures are being revised to incor-porate specific instructions to operators in the event a similar seal failure were to occur. The objective vill be to minimize the total leakage, thus reducing the possibility of demage to other plant equipment.

10. We are investigating the replacement of the vapor seal (the seal between the leak off cavity and the motor package) with a new seal which will withstand higher temperatures.

All the above items, except No 10, vill be completed prior to returning the plant to service.

t

O Mr. John F. O' Leary

.. O 14 Docket No 50-255 License No DPR-20 November 2, 1972 The Palisades Plant Safety Audit and Review Board has reviewed this abnormal occurrence. The Board concluded that restaking the collar setscrews would prevent recurrence of a similar incident. The Bor$rd also concluded that the corrective action outlined above is adequate to return the plant to operation.

It 10 presently anticipated that the plant will be returned to service during the week of November 5 Yours very truly, Ralph B. Sewell (Signed)

RBS/ map Ralph B. Sewell Nuclear Licensing Administrator CC: BHGrier, USAEC L

l l

l