ML20236C232

Responds to NRC Re Questions Raised in Insp Rept 50-362/87-25.Applicability of Tech Spec 3.3.2 to FWIV Operability Carefully Considered Prior to Making Valve Inoperable.Valve Inoperability Did Not Violate Tech Spec

ML20236C232
Person / Time
Site:	San Onofre
Issue date:	09/24/1987
From:	Ray H SOUTHERN CALIFORNIA EDISON CO.
To:	Martin J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V)
References
NUDOCS 8710270066
Download: ML20236C232 (8)

Text

-

l I

j"%prn?- RECEIVED Southem Califomla Edison Company RE$lhy e.o.eoxtee R8) $[p S AN CLEMENTE, C ALIFORNIA 92672 f HAROLO 5. RAY m e ,ar m c n w c m ara September 24, 1987 m e n.o,,c un o~orar m * . ..c o

.it.

E?, o> Cr

- 1 bJ U'

Mr. John B. Martin, Regional Administrator $$

U. S. Nuclear Regulatory Commission, Region V 33 th 1450 Maria Lane, Suite 210 u)

Walnut Creek, CA 94596 SE o

Subject:

Docket No. 50-362 San Onofre Nuclear Generating Station, Unit 3 By letter dated August 10, 1987, Mr. Zimmerman forwarded for our information a copy of questions discussed with San Onofre Plant Staff concerning maintenance activities performed on a feedwater isolation valve and an atmospheric steam dump valve in San Onofre Unit 3. The responses provided orally to those questions are fully reflected in NRC Inspection Report 50-362/87-25 which was ,

issued by Mr. Kirsch's letter dated September 1, 1987.

i SCE considers the proper control of the maintenance activities of interest to be extremely important to safe operation of the plant. We appreciate the opportunity to provide information in response to the specific questions forwarded by Mr. Zimmerman's letter and the accurate summary of our responses which is included in the subsequent inspection report. Attached hereto are written responses to these questions. They are consistent with those provided orally and reflected in the inspection report, and they are provided here for completeness and to elaborate somewhat on several of the points involved.

The fundamental issues concerning proper control of maintenance activities which are raised by the study of these two instances are highly complex and technical. They go to the intent of the regulations and the Technical Specifications, and we agree that they merit discussion and clarification on a generic basis. We will welcome the clarification that results, and will of course strive to comply with the clarified requirements at all times.

&O O Y ll P

' n;%:

n% .c

j. Mr. John B. Martin September 24, 1987 However, I believe it will be helpful in the NRC's further consideration of the issues involved, for SCE to describe more fully than we have by our response to the questions we have been asked, our process of determining the proper controls to be exercised in performing these particular maintenance activities. Accordingly,.this additional information is provided below.

Atmospheric St am Dump Valve (ADV) 3HV-8419 The facts and circumstances concerning the repair of this valve are summarized in the NRC inspection report. Our bases for concluding that this work did not involve an unreviewed safety question are summarized in the response to Question 3 in the attachment hereto.

As important, in our view, are the management considerations which entered into the decision to allow entry into Mode 2 with ADV 3HV-8419 not Operable. In this case, the decision was reviewed by me with Mr. Baskin, Vice President Nuclear Engineering, Safety & Licensing, and approved by both of us prior to implementation. We considered this to be an important decision, and our bases for approval included the following:

o The Standard Technical Specifications, as implemented for example at Palo Verde, permit an ADV to be inoperable during power operation for up to 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.

At the time, San Onofre Unit 3 was returning from refueling, and the core had a low decay heat level. We felt that a very conservative restraint in the absence of Technical Specification requirements on the San Onofre ADVs was to prohibit entry into Mode 1 until the ADV was Operable. I imposed on plant management a restraint against entry into Mode 1 with the ADV not operable, prior to approving entry into Modes 3 and 2.

o In the event of a steam generator tube rupture that would prevent use of the remaining Operable ADV for cooldown, manual operation of the steam generator safety valves, or steam dumping to the condenser at atmospheric pressure (i.e., with steam relief through the rupture discs) were still available. Given the low decay heat level in the core, this steam dumping would be far less than is normally required.

S Y q? namn Mr. John B. Martin September 24, 1987 i.

.The NRC inspection report discusses SCE's development of a self-imposed 7 day limit on ADV inoperability which occurred during plant startup in 1982 and is currently included in plant operating procedures. I believe that the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> ADV inoperability limit did not then exist elsewhere as a prototype for our consideration. Although it is correct that this 7 day limit was referenced in documenta-tion developed during the repair of ADV 3HV-8419 in March l

1987, San Onofre Units 2 and 3 will not be operated at power in excess of 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> with one ADV not Operable, and a unit l will be promptly shut down if both ADVs should become not operable.

l In summary, as a general matter, where safety limits are not established in our Technical Specifications, we recognize an obligation to carefully consider what limits should apply. In this regard, we do review for guidance requirements established in Technical Specifications elsewhere.

Main Feedwater Isolation Valve (MFIV) 3HV-4048 Again, the facts and circumstances concerning the repair of this valve are summarized in the inspection report. In this case, the Technical Specification requirements are more complex. Our bases for concluding that the work did not involve violation of the instrumentation technical specification are summarized in response to Question 1 in the attachment. The response to Question 2 summarizes our bases for concluding that the work did not involve an unreviewed safety question.

In this case, I reviewed and approved the action that was taken. This review included the considerations described in our response to the questions and the following:

o Unit shutdown imposes an undesirable transient on the plmnt and substantially increases the opportunity for personnel error which can result in a challenge to safety systems. For this reason, the Technical Specifications allow continued power operation with important safety equipment not Operable for various periods.

-ym ,- .

.z- <<

m- +

. _f*M?$?

[,, ,

'Mr.LJohn:B.. Martin - September; 24, 1987'

-o. rI. imposed ajlimittof 72-hours during which'the'FWIV.

could'be not Operable lwithout: plant shutdown, on the' basis that this.was what I' believed would be required

'if the FWIVs were,to be11ncluded=in the. Technical-Specifications. As a-matter of conservatism,.

1 recognizing:the safety function of the component and.

.the need.to minimize the period of valve inoperability,.

I. imposed a further limit on theLwork window of 8

. hours.:

o I carefully considered the view that making the'FWIV

' inoperable would violate the. Technical-Specification. 1

. requirements on' instrumentation response times. .'For-the~ reasons discussed in response to-Question 1 in the attachment, I concludedLthat this was. clearly not the intent'of:the Technical. Specifications, and it was not ,

required'by considerations of plant safety.

-We continue to operate under a_self-imposed 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> limit on FWIV.inoperability at power.-

I-hope that this additional information is helpful to the'NRC in further' review of this matter. -If you have'any-questions, or if you would like additional information,.

please let me know.

Sincerely,

, $b HBR:bam Attachment cc: David J. Fogarty Kenneth P. Baskin Mr. F. R. Huey, USNRC Senior Resident Inspector - SONGS Mr. A. D. Johnson, USNRC Enforcement Officer - Region V i

l

-m.- _ _ - - . _ _ _ _ _ . _ _

< . F e, e p

)

Attachment Responses to NRC Questions 5 l

The following documents responsas to NRC Questions contained in the August 10, 1987, letter from Mr. R. P.

Zimmerman to Mr. K. P. Baskin.

Question 1 Basis for concluding that FWIV repairs conducted on April 10 and April 24, 1987, were not in violation of Technical Specification 3.3.2.

Response

The applicability of Technical Specification 3.3.2 to FWIV operability was carefully considered prior to making the valve inoperable since this is the only place in the Technical Specifications where the valve is referenced. It was concluded that making 3HV-4048 inoperable did not violate Technical Specification 3.3.2 because:

o If Technical Specification 3.3.2 is violated by this action, then it is so to the extent that all four Engineered Safety Features Actuation System (ESFAS) instrumentation channels are made inoperable. This imposes Technical Specification 3.0.3, which requires a prompt shutdown when outside conditions described in the Technical Specifications.

If the NRC had intended to impose a 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> shutdown requirement on FWIV inoperability, then it would not have been addressed via Technical Specification 3.3.2.

Instead, it would have been explicitly addressed by an appropriate LCO and action statement.

o Other components which perform important safety functions, and which are identically actuated by the ESFAS as are the FWIVs, do not have a 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> shutdown requirement. To the contrary, they have allowable inoperability periods which are explicitly set forth in the Technical Specifications.

Also, in applying the Technical Specifications, we generally comply with the most conservative applicable requirement. In this respect, if Technical Specification 3.3.2 imposes a 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> shutdown requirement on the FWIVs because all four ESFAS channels are inoperable, it would do so for all connected components, notwithstanding longer inoperability periods which may be specified elsewhere.

1 pjff?

nr I

o SCE had under in-house review a proposed change to the Technical Specifications which addressed operability requirements for the FWIVs. This change, which was  ;

reviewed again by Site and General Office management at i the time of the decision to proceed with the I maintenance work, provides a 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> limit on l inoperability, as was imposed on the work at the time.

Thus, careful consideration was given, from a safety i viewpoint, to the proper limit at power on valve j f inoperability. 1 In Enrimary, we continue to believe that inoperability )

of a single connected component does not render all channels of ESFAS inoperable, whether or not the cor.ponent is addressed elsewhere in the Technical Specifications.

Question 2 a

Basis for concluding that the FWIV repairs conducted on i April 10 and April 24, 1987, did not involve an j unreviewed safety question, including the following j considerations: i l

a. Accident analysis in the UFSAR  !

b. Plant operation with the FWIV blocked open 1

c. Reduction in the margin of safety of Tecnnical l Specification 3.3.2  !

Response

As indicated above, SCE had under in-house review a l proposed change to the Technical Specifications at the time j the need arose to maintain 3HV-4048. This change provided for a 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> limitation on FWIV inoperability and was to remove any implication that Technical Specification 3.0.3 l applied. Thus, the implications of FWIV inoperability on I the UFSAR accident analysis, and plant operation with the l FWIV not Operable (i.e., open and not capable of closing), l had been recently considered i'1 some detail.

The inoperability of any component which performs a safety function represents a reduction in the margin of ,

safety which is otherwise provided. For this reason, the Technical Specifications limit the time intervals during which such inoperability may exist without taking action to change mode, or otherwise restore the margin normally provided.

_ _ _ _ _ _ _ _ - ________________~

i .

yw[y' Consideration of time limits on inoperability entails the grouping of components into cla:ses, depending on the function performed. Limits are imposed ranging from 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> to indefinite periods, providing compensatory measures are taken. SCE concludes that the FWIV belongs in the class of components which are subject to a 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> limit on inoperability. In the absence of explicit provision for  ;

such a. limit in the Technical Specifications, we imposed the  !

limit administrative 1y. We believe that this limit is consistent with the margin of safety required to be maintained by Technical Specification 3.3.2.  ;

1 With respect to the possible existence of an unreviewed safety question, we concluded that:

o The maintenance did not involve a change, test or experiment. .

i o Imposition of the time limit on inoperability, as l discussed above, resulted in controls which are consistent with those imposed on similar components by the Technical Specifications. Hence, we concluded that the maintenance did not involve an unreviewed safety i question.  !

In response to questions from the NRC, we have subsequently considered the more general questions i associated with maintenance, relative to 10 CFR 50.59 l requirements, and we agree with the. discussion concerning j maintenance in Part 9800 of the I&E Manual in this respect.

That is, we believe that maintenance can be performed  !

without satisfying the requirements of 10 CFR 50.59, i provided that the maintenance does not result in a change  !

and it does not violate the Technical Specifications. i However, we also believe that the maintenance must be l completed in a timely manner, consistent with the safety  !

function performed by the component involved. In this case,  !

we imposed a 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> administrative limit and a further conservative limit on the conduct of the work of 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />.

i If we had removed the valve from service and left it for repair at the next shutdown, this clearly would represent a temporary change and would have required an evaluation to ,

satisfy 10 CFR 50.59. (Of course, such a temporary change  :

would be found to represent an unreviewed safety question and would not be acceptable.) .

j"aW 1

In summary,.the' difference between an impermissible ichange:and permissible maintenance is the time involved.

The time limitation on inoperability must be consistent with that imposed on the class of component involved. In the-case of the FWIV, and in-the absence of explicit provision  ;

in the Technical Specifications, SCE concludes that 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> i is the proper' time limit.

Question 3 4

Basis for concluding'that the ADV repair conducted from i March 1 through March 9, 1987, did not involve an unreviewed safety question, including the following considerations:

a. . Accident analysis in the UFSAR  !

b. Plant operation with ADV isolated

c. Reduction in the margin of safety of Technical '

Specification 3.4.5.2 (operational Leakage)

Response

As with the FWIV, SCE had given consideration to the time limits on'ADV inoperability which would ensure consistency with the. accident analysis in the UFSAR and not constitute an unreviewed safety question. In this case, the review had been. performed in 1982 and had resulted in a 7 day administrative limit which was included in the plant operating procedures.  ;

Review of this limitation prior to mode change following the refueling in 1987 gave consideration to the recent establishment of a 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> inoperability limit for the Palo Verde units. SCE concluded that a conservative limitation to be imposed on the special circumstances involving San onofre Unit 3 (i.e., return to service i following refueling) would be to allow the 7 day limitation  !

to stand, except to not allow entry into Mode 1 until l 3HV-8419 was made operable.  !

l Other considerations, relative to accident analysis,  !

margin of safety and 10 CFR 50.59 are the same for the ADV maintenance as they were for the FWIV, as discussed in the response to Question 2 above.

1 l

I L

1 L