ML20212E360

Summary of 861120 Meeting W/Bechtel,Southern Co,Svcs & Alabama Power Co in Bethesda,Md Re Proposed Tech Spec Changes to Snubber Surveillance Insp Intervals.Meeting Agenda,List of Attendees & Viewgraphs Encl

ML20212E360
Person / Time
Site:	Farley
Issue date:	12/23/1986
From:	Reeves E Office of Nuclear Reactor Regulation
To:	Office of Nuclear Reactor Regulation
References
NUDOCS 8701050303
Download: ML20212E360 (18)

Text

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _

4 Docket Nos. 50-348 P M Fil d E. Reeves (2) and 50-364 s

NRC PDR D. Miller LICENSEE: Alabama Power Company (APCo)

P.ocal PDR DELD.

PAD #2 Rdg E. Jordan FACILITY: Joseph M. Farley Nuclear Plant, J. Partlow B. Grimes Units 1 and 2 L. Rubenstein ACRS (10)

Gray File NRC Participants

SUMMARY

OF MEETING HELD ON NOVEMBER 20, 1986, BETWEEN NRC AND APCo REPRESENTATIVES TO DISCUSS APCo's PROPOSED I

TECHNICAL SPECIFICATION CHANGES TO SNUBBER SURVEILLANCE i

INSPECTION INTERVALS I

Introduction The NRC Project Manager (G. Requa, backup for E. Reeves) reviewed the purpose of the meeting which resulted from APCo's license amendment request proposed by letter dated September 2, 1986. The meeting was held to assist the NRC staff's understanding of APCo's proposal in order to meet the licensee's request amer <

nt schedule of February 27, 1987. A list of attendees is enclosed (Er. asure 1).

Discussion Southern Company Services (D.P. Hayes), representing Alabama Power Company (APCo),

used viewgraphs to present an overview of the proposed snubber visual inspection methodology. An outline of the discussion is enclosed as pages 1-4 of Enclosure 2.

APCo (S.T. Burns) reviewed the refurbishment program f snubbers as follows:

1.

All snubbers at Farley Unit No. I have been rebuilt.

2.

All snubbers at Farley Unit No. 2 will be rebuilt by the end of the next outage.

3.

The population of snubbers includes:

Unit 1 Unit 2 596 431 INACCESSIBLE 272 223 ACCESSIBLE EE BT4 TOTAL Bechtel (E.W. Thomas) used viewgraphs (Enclosure 2, pages 5-14) to review the details of the statistical methodology. The approach is to maintain a 95 i

percent confidence level that 90 percent of the group remains operable based on visual inspections. Table 1 on page 14 of the Enclosure 2 reflects the proposed changes based on a snubber group size larger than 200. This table has been proposed by APCo in a letter dated September 2, 1986, for replacing Technical Specification 4.7.9 a. visual inspection schedule which is based on a fixed number of inoperable snubbers per inspection period. APCo contends that the existing requirements are overly conservative and result in excessive surveillance requirements.

8701050303 861223 PDR ADOCK 05000348 O

PDR

Q However, the staff indicated that the existing requirements are to maintain an approximately 95 percent confidence level that 90 to 100 percent of the snubbers are operable based on a combination of visual inspection and functional testing.

Conclusion Following discussions between the NRC staff and APCo's staff and this support organizations, APCo agreed to provide additional information at an early date to support the requested NRC staff action date of February 27, 1987.

Edward Reeves, Project Manager Project Directorate #2 Division of PWR-A Licensing

Enclosures:

As stated

• See previcus concurrence

• LA: PAD #2

• PWRA:EB

1. 2 PD:

DMiller TSullivan EReeves:ab LRube stein 12/ /86 12/ /86 12/p/86 12/g/86

4 Mr. R. P. Mcdonald Alabama Power Company Joseph M. Farley Nuclear Plant cc:

Mr. W. O. Whitt Executive Vice President D. Riard MacGuineas, Esquire Volpe, Boskey and Lyons Alabama Power Company 918 16th Street, N.W.

Post Office Box P641 Washington, DC 20006 Birmingham, Alabama 35291-0400 Charles R. Lowman Mr. Louis B. Long, General Manager Alabama Electric Corporation Southern Company Services, Inc.

Post Office Rox 550 Post Office Box 2625 Andalusia, Alabama 364?0 Birmingham, Alabama 35?07 Chairman Houston County Commission Regional Administrator, Region II Dothan, Alabama 36301 U.S. Nuclear Regulatory Commission 101 Marietta Street, Suite Po00 Atlanta, Georgia 30303 Ernest L. Blake, Jr., Esquire Shaw, Pittman, Potts and Trowbridge Claude Earl Fox, M.D.

2300 N Street, N.W.

State Health Officer Washington, DC 20037 State Department of Public Health State Office Building Montgomery, Alabama 36130 Robert A. Buettner, Esoufre Ralch, Bingham, Baker, Hawthorne, Mr. J. D. Woodard Williams and Ward General Manager - Nuclear Plant Post Office Box 306 Post Office Box 470 Birmingham, Alabama 35201 Ashford, Alabama 36312 Resident. Inspector U.S. Nuclear Regulatory Commission Post Office Box 24 - Route 2 Columbia, Alabama 36319 e

i

ENCLOSURE 1 LIST OF ATTENDEES AT THE NOVEMBER 20, 1986 MEETING ROOM P-422 PHILLIPS PllILDING 9ETHESDA, MARYLAND NAME ORGANI7ATION G. P.eoua PWR-A/ PAD #2 B. J. George Southern Company Services D. P. Hayes Southern Company Services E. W. Thomas Bechtel Eastern Power Corporation S. Sue-Ung Bechtel Eastern Power Corporation K. Gandhi Bechtel Eastern Power Corporation K. Powers Southern Company Services S. T. Burns Alabama Power Company B. D. McKinney Alabama Power Company J. A. Ripple Alabama Power Company J. E. Garlington Alabama Power Company D. Terao NRC/NRR/DPLA/EB G. Johnson NRC/NRR/DPLA/EB S. Lee NRC/NRR/DPLA/EB H. Shaw NRC/NRR/ DBL /EB H. L. Brammer NRC/NRR/PWP-A/EB T. Sullivan NRC/NRR/PWR-A/EB G. Bagchl NRC/NRR/PWR-A/EB

,-w

3 ENCLOSURE 2 PROPOSED SNUB 8ER VISUAL INSPECTION INTERVAL TECHNICAL SPECIFICATION CHANGE PRESENTATION I.

INTRODUCTION A.

BACKGROUND B.

PROPOSED SNUB 8ER TECHilICAL SPECIFICATIONS II.

METHODOLOGY OVERVIEW III.

COMPREHENSIVE METHODOLOGY REVIEW A.

METHODOLOGY 8.

RESULTS Ih.

SUMMARY

/ ACTION ITEMS A.

METHODOLOGY BASED ON SOUND STATISTICAL AND ENGINEERING PRINCIPLES 3.

VISUAL INSPECTION INTERVALS 3ASED ON SAME CRITER AS FUNCTIONAL TESTING (95/90)

C.

FOLLOW UP ACTION ITEMS 0825F t

0

CHANGES TO VISUAL INSPECTION SCHEDULE NO CHANGE TO 100% VISUAL INSPECTION REQUIPEM o

NO CHANGE TO ACTION STATEMENT OR LC0 NO CHANGE TO VISUAL TEST ACCEPTANCE CRITERIA o

NO CHANGE TO FUNCTIONAL TEST REQUI'EMENT o

NO CHANGE TO FUNCTIONAL TEST ACCEPTANCE CRITE No CHANGE TO SERVICE LIFE f10NITORI';G REQUIREMENTS o

r

+

0825F i

I STS METHODOLOGY o

DEVELOPED TO PROVIDE A 95% CONFIDENCE THAT A o

OF THE SNUBBERS ARE-OPERABLE BASED ON PROVIDING CONSTANT PROTE:

o TION BASED ON TRADITIONAL STATISTICAL METHODS o

1 e

0825F

I!

METHODOLOGY 9ASED ON SOUND STATISTICA o

PRINCIPLES VISUAL INSPECTION INTERVALS BASED ON SA I

FUNCTIONAL TESTING (95/90)

OPOSED CHANGE IS SMALL P ART OF C011PREHENS 4

d 0825F b

METHODOLOGY FOR SNUBBER OPERABILITY SURVEILLANCE APPROACH:

• Maintain 95% confidence level that 90% of group remains operable

• Assume unlimited frequency of defined accidents may occur (l. e., probability of occurrence = 1)

• Use exponential distribution to predict time to failure of a given snubber

-Its hazard function constant with time

• Number of failed snubbers before time (t) follows a binomial distribution

• Estimate failure rate 'ising chi-squared function (with a confidence level of 95%)

l

• Assume for all snubbers that fall, they do so at start of inspection period

• Results of previous inspection interval used to establish new inspection interval

• New inspection interval can be increased by only one interval step

• Tolerances are provided on inspection intervals. The extremes on tolerances that creates the most severe result is used in develop-ment of inspection schedule.

G-WRW-11/86

DEVELOPMENT OF METHODOLOGY TIME TO FAILURE OF A GIVEN SNUBBER (X) FOLLOWS AN EXPONENTIAL DISTRIBUTION.

THEN PROBABILITY DENSITY FUNCTION CAN BE EXPRESSED AS :

-O*

f (X) = Oe (I) x WHERE O = FAILURE RATE PROBABILITY THAT AN INDIVIDUAL SNUBBER FAILS BEFORE TIME (t )

Ps = P (x< t } =( fx (5.) d4 = l-e

-et (2)

IF Z IS THE NUMBER OF FAILED SNUBBERS BEFORE OR AT TIME (t ), BY USE OF BINOMI AL DISTRIBUTION P (Z = z ) =,3 ( n _, )3 P

( l - P. ) "~ '

(3)

WHERE n = TOTAL NUMBER IN GROUP

PROBABILITY THAT THERE ARE NO MORE' THAN (s)

SNUBBERS INOPERABLE BEFORE OR AT TIME (t )

P [ z e s ] = $,,("n'_ z ), P', ( l - P, )"- *

(4) z=O USE P[z < s } = CONFIDENCE LEVEL = 95 %

COMBINING EQ'S 284 P{ z <s } = $n!(n-z)!(i-e

) ( e-t}"

=.95 (5) z_o ESTtMATE THE F AILURE RATE $

Total number of follures O

(6) avg.

Total test tirne an failed & unfailed units 2

g _

1 l-c4,2I+ 2 (6a) 2T WHERE

= Chi-SQUARED FUNCTION T

= ( n-1) T I

= TOTAL NUMBER OF SNUBBERS WITH DEFECTS T

= CURRENT INSPECTION PERIOD cK.

= 0.05 CORRESPONDING TO 95% CONFIDENCE LEVEL

i -

L

)

7

(

59

=

2

-n

)

t 2

+1 T 2

)

. 1 K

- n 2i

(

2

_ e

(

F t

2

+ T 1

2

)

.1 d

- n 2I

(

. X 2

_ e i

(

)

z_

6

'" (n 8

! n 5

0

=

sIz S

=

Q E

}

s G

N zi I

N

{

IB P

M O

C

u

!ii I

i 1.

)

I I

i!i i :

s.

JUSTIFICATION:

• 95% Confidence level of 90% Reliability

- Representative of criteria for safety systems

- Produces very strict acceptance criteria

- Rigorous - less than 0.0076 failure rate to maintain same inspection interval

• Unlimited frequency of accident events is a most conservative

, assumption s

• Use of exponential distribution to predict time-to. failure:

- Commonly used for similar analysis of mechanical systems

-Inplace inspection and maintenance program leads to expectation that if snubber survives to a given time (t),

its chance of survival the next instant is constant with time (Hazard Function)

- Exponential distribution assumed in other approaches to snubber inspection

• Use of binomial probability function is appropriate fori determining number of failures before time (t)

N

- Failure of each snubber independent of others.

l

- Represents a success-failuh event s

- Mathematical exact

(

• Use of Chi-Squared Function L

- For a exponential distribution,' estimate of failure rate follows Chi-Squared distribution (mathematical exact)

- Conservatively uses a failure rate at 95%' probability that its actual value will be lower a

G-WRW-11/86

i JUSTIFICATION CONTINUED

• Establishment of failure rate conservatively assumed all snubbers which fall do so at beginning of inspection interval

• Used latest information in establishing next interval

- Further restricted by allowing only one inspection interval step up from current interval

. Used most severe tolerance in estimate of failure rate and calculation of next inspection interval

- Use lower bound T to establish current failure rate

- Use upper bound T to establish % allowable Inoperative

• Minimum group size used to minimize next inspection Interval

- rn /100 = I

• droup size limit tailored to Farley

. Methodology based on rigorous probabilistic and statistical procedures excerpted from standard texts e9 e

T y

)

i(

\\

\\

MC).

t t

8 so

1. 3 Fig. 313 Typical hazard function.

J hit) 10.0 9.0 8.0-7.0 Uniform distnbubon%

6.0 5.0 40 3 0 '-

g L Normal distnbuten 1.0 Exponenhal

/cstnbuten

.s 0

1.0 20 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 Fig. 3,14' Hazard functions for nonna!, uniform, and exponential dstributions.

e L

n /,(,)

0.3 CH (2) 0.2

("EX(%))

0.1 O

t i

e 0

5 10 15 s

4, /,( s) 0.2 CH (5) 0.1 t

t o

0 5

10 15 s

6 /,1,1 0.2 CH (10) 0.1 0

I I

0 5

10 15 l

x distn*butions. (Adapted from R. 8.

8 ris. BAS Burmston and D.C.May," Handbook of Probability and Statistics," Handbook Publishers, Inc., San-dusky, Ohio,1953.)

i l

(7p

RESULTS:

• Method is valid for group of 200 or more

- As group size increases, conservatism increases

• Requires more frequent inspections with increasing failures

- Number of inoperatives influenced by size of group

+ Applicable to each unit G-WAW-11/86

f l

TABLE 1 visual inspection schedule for any group (category) of snubbers of the Joseph M. Farley Nuclear Power Plant Units 1 and 2.

Percent of Snubbers Found Current Visual Inoperable, r per Current Next Visual Inspection Period

• Visual Inspection Period Inspection Period 24 Months + 25%

r < 0.76%

24 Months + 25%

0.76% T r < 0.99%

18 Months T 25%

0.99% < r T 1.40%

12 Months T 25%

1.40% < r T 2.43%

6 Months T 25%

r > 2.47%

3 Months [25%

18 Months + 25%

r < 0.62%

24 Months + 25%

0.62% T r < 0.76%

18 Months 7 25%

0.76% < r T 1.09%

12 Months 7 25%

1.09% <ri 1.96%

6 Months 25%

r > 1.9T%

3 Months [25%

12 Months + 25%

r < 0.58%

18 Months + 25%

0.58% Tr < 0.76%

12 Months T 25%

0.76% < r T 1.40%

6 Months 7 25%

r > 1.4T%

3 Months [25%

6 Months + 25%

r=0 12 Months + 25%

0<r ( 0.761 6 Months 7 25%

r > 0.7T1 3 Months [25%

3 Months + 25%

r=0 6 Months + 25%

r>0 3 Months + 25%

l Notes:

1.

The above table is generated based on 95% confidence that at least 90% of the snubbers (in a group) being operable all the time.

2.

• Earlier visual inspection periods than required may be utilized.

If this option is chosen, the criteria for determining the next visual inspection period shall be the criteria associated with the earlier visual insepetion period selected.

3.

The above table is not. applicable for the snubber group size (n) smaller than 200