ML20084U421

From kanterella
Jump to navigation Jump to search
Ao:On 720201 & 02, Safety Injection Sys Containment Sump Isolation Valve Failed & Control Rod Drive Mechanism Discovered Misaligned,Respectively.Caused by Inadequate Review of Design Mods & Procedure of Transient Test
ML20084U421
Person / Time
Site: Palisades Entergy icon.png
Issue date: 02/11/1972
From: Sewell R
CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
To: Morris P
US ATOMIC ENERGY COMMISSION (AEC)
Shared Package
ML20084U424 List:
References
NUDOCS 8306290034
Download: ML20084U421 (4)


Text

.

( ]

y

]

l

t Cppsumers %

4 Powel 4 3 Companf f

gpD g'

f di c..o... on.co.. m w.. u.cn,..n 4.mo.. soa.oo. Michigan 4 92W. Area February 11, 1972 Dr. Peter A. Morris, Director Re:

Docket No 50-255 Division of Reactor Licensing License No DPR-20 United States Atomic Energy Con: mission Washington, DC 205h5

Dear Dr. Morris:

This letter is written to apprise you of two events that occurred recently at the Palisades Plant. These events were the failure of a safety injectior_ system contairment sum valve to open du-ing quiiterly sd1veillance-testliig'y, isolation 7

and the delio-erate misalignment cf a control rod drive mechanir.n (CRDM) during plant cperatien so that an azimuthal zennon transient test could be conducted.

On l'etruary 1, 1972, durinc exercicing of safety injection system valves as required by Article 4.6.4-a of the technical speci-

! fications, CV-3029 (one of two containment sump isolation valves) jfailedtorespondtoanopensignal, This valve is a double acting piston, air-operated, 24-inch gate valve.

At the time the incident occurred, the reactor was oper-ating at a power level of approximately 440 MW. Generator output t

was 105 I&e (gross). The three-month interval safety injection system testing was in progress to meet technical specification requirements.

The results of these tests were all acceptable with the exception of CV-3029 h

Normal air pressure to operate this valve is supplied at gQ 90 - 100 psig. This air pressure supply was increased to 130 psig 4 3 /)

and the valve opened. As it started to open, the gate appeared to

}p

" pop" loose frem the valve seat and then stroke normally to the L

full open position. The valve was stroked several times at both the elevated air pressure and the normal operating pressure. The valve was considered operable following this testing. However, it was de-i

-/\\]

cided to increase the testing frequency to assure that the valve would continue to operate while further investigating the cause of the erratic operation of CV-3029 8306290034 720211 PDR ADOCK 05000255 S

PDR c

e.-

COPY SENT REGIO {

/

i' Dr. Peter A.

ris 2

February 11, 1972 Originally, this valve (and its redundant valve CV-3030) was installed with a double acting air operator, storage tank, air pressure booster and individual pressure regulators to the opening and closing sides of the valve operator. The purpose of the booster and air storage tank was to provide a sufficient capacity of stored air to cycle the valve one time after a complete loss of instrument air. Several other valves in each engineered safeguards room were provided with air boosters and storage tanks to provide the ability to operate these valves one cycle following a complete loss of air pressure. Because of the difficulties experienced in achieving re-liable operation due to problems with the boosters during the initial testing of these v::.1ves, changes were made which eliminated the use of the air boosters and individual storage tanks. A new high-pressure air system was installed in each engineered safeguards room (refer to Section VII of the Semiannual Report of Operations No 1).

These new systems provided accumulators to cycle each valve connected to the system at least once following a loss of the air compresser. The modifications were designed and instaned by the Bechtel Corporation Lnd reviewed by Consumers Power Company personnel.

During the investigation ir.te,the difficulty experienced in opening CV-5029 February 1, 1972, the valve manufacturer was contacted regarding valve timing and operEting pressure requirements under the design conditions of operation. The valve manufacturer informed us 4

th*1t the valve installed in the vertical upright position should be closed with an air pressure that is slightly less than the pressure provided to cpen the valve. In addition, the valve etroke on closing shculd be tudfom and total trevel time shwld net be less than one minute. The ninimum operating requirements change with valve orien-tation because of the consideration of the weights of the moving portions of the valve. CV-3030 is mounted upside down while CV-3029 is mounted about 45 degrees below a horizontal position. The assump-tion of a valve installed in the upright position is the limiting con-dition with regard to differences in opening and closing air pressure required for proper operation of the valve under design conditions.

When the new air systems were installed and the air boosters removed,the pressure provided to the opening and closing sides of CV-3029 and CV-3030 were the same. This was adequate for CV-3030 and marginal for CV-3029 No instructions from the manufacturer existed when the modifications were made; consequently, the personnel review-ing the modifications did not appreciate the relationships cf opening and closing supply pressure for these two large valves.

In light of the above discussion, the cause of this incident has been designated as inadequate review of modifications made to established plant design. Our general office review procedures have

' been strengthened during the past year in accordance with 10 CFR 50, Appendix B.

These procedures are being reevaluated in light of this occurrence for possible areas of improvement.

.a

Dr. Peter A.

rris 3

February 11, 1972 Plans are being made to revise the operating schema of these two valves such that a greater air pressure is available to open the valve than is provided for closing the valve. This will provide ade-quate margin in operating pressure to insure the valves will operate under the design conditions even if they were in the most limiting orientation, ie, upright. Until this modification is completed, the valve will be closed by manually controlling the closing pressures such that they are less than the opening pressure.

Two completely redundant containment sump isolation valves are provided. Therefore, the failure of one valve does not impair the ability of the safety injection system to perform its design func-tion in the recirculation mode of operation. In addition, the tech-nical specifications provide appropriate rules governing plant opera-tien with inoperable components present in the safety injection system. A single component being inoperable does not negate the ability of the safety injection system to perform its function, but it does reduce the redundancy provided by design. The time limits for operr. tion with an inograble camponent prside essurance that the reacter will not be operated for an extended time period with en in-operable component.

The possibility of recurrence of this type of failure will be reduct:d by providing a greater air pressure to open the valve than to close it. Until this modification is completed, the same effect will be provided by manual cont. ol of the clocing pressures.

In the ur.Jikekv erent this valve were again to stic?t in the closed position, it could te opened by manually increasing the air pressure supplied to the opening side of the valve operator.

Ite valvt operator is designed to withstand 400 psi. The air systems installed in the engineered safeguards rooms provide 325 psi air.

On February 2, 1972, control rod drive mechanism (CRDM) No 33 was misaligned from the remainder of regulating Group 3 so that an azimuthal zennon transient test could be conducted.

~

The effects of this delivertte misalignment of CRDM No 33 were analyzed in accordance with Article 3 10.4-c of the technical specifications prior to starting the test. Hot channel factors were shown to be within design limits under the test conditions. In addition, the test was discussed with Division of Compliance personnel prior to starting the test. It was concluded from these discussions that the operation with a deliberately misaligned CRDM should be reported in accordance with Section S-5 of the Special Technical Specifications.

The test procedure required CRDM No 33 to be fully inserted for five hours and then withdrawn to its normal position at the upper electrical limit with the remainder of reactor regulating Group 3.

During the positioning of CRDM No 33, boron concentration was adjusted to maintain essentially constant power level. While the CRIM was

e q

r.

Dr.PeterA.o 9

Morris 4

February 11, 1972 misaligned and after it was withdrawn, data were to be gathered by' monitoring neutron instrumentation. These data were to be analyzed for azimuthal zennon transient effects.

Unfortunately, the reactor tripped due to a feed-water transient before the test could be completed. Therefore, it will be necessary to conduct another test at a later date when the plant has been returned to operation.

Yours very truly, Ralph B.'Sewell (Signed)

RES/dmb Ralph B. Sewell Nuclear Licensing Administrator l

i 1

i t

J

.: