ML19354C466

Ack Receipt of Documents Re Reactor Pressure Vessel Thermal Shock Event at Facility.Basis Should Be Given for Assuming That flaw,1-inch Deep,Is Conservative

ML19354C466
Person / Time
Site:	Indian Point
Issue date:	03/18/1981
From:	Nichols F ARGONNE NATIONAL LABORATORY
To:	Igne E Advisory Committee on Reactor Safeguards
References
ACRS-CT-1327, NUDOCS 8104090036
Download: ML19354C466 (2)

Text

'

. .emp,%

ARCONNE NATIONAL LABORATORY U- /3 "

9700 Socili CASS h1M, ARCpM, IlrCis 60439 Tdgi<n 312/972- 8292 March 18, 1981 Mr. E. Igne, Staff Engineer U.S. Nuclear Regulatory Commission Advisory Co=mittee on Reactor Safeguards Washington, D. C. 20555

Dear Mr. Igne:

Per the request in your letter dated February 6,1981, I have reviewed the documents concerning the reactor pressure. vessel thermal shock event at Indian Point Unit No. 2. Your letter requested a reply by the end of February, but for reasons unknown to me I received your letter only last week. For that reason, my review has been far from exhaustive, but perhaps the comments will be of some use.

1. Document IP2, page 4.

A basis should be given for assuming that a flaw 1" deep is " conservative".

If the " critical" size is 1.7" this assumption is clearly very critical to the analysis.

A basis for the value of Co in the crack growth equation should be given and shown to be appropriate for the actual conditions of interest, e.g. specific environment, R-value, etc.

2. Document WCAP-9822, page 1-4.

Basis for the 1.7 in. deep critical flaw size should be given.

Also, I consider the statement ".... a flaw of this size is readily detectable during pre-service and in-service examinations and would not be present prior to the event ...." unacceptable. Surely such absolute statements cannot legitimately be made. One can only speak " statistically".

In that vein, as a non-expert in NDE, I find from my discussions with experts that the probability of not detecting a flaw goes pretty close to unity, typically for a flaw size of (20% thickness. Since the critical flav size being discussed is in that vicinity, I think that quantitative basis is required for assuming this non-detection probability to be low enough to neglect.

TIE LMtR5iTy d O*xp ARcpn L.htR5ht5 A55oasim

/4 ET :81040906 N -

.P'

2-

[1 - ,.

i E. Igna. ~ March 18, 1981 i

+

1 1, >

i.

3. -Page 2-27A.'

Bases:are required for assuming 5_ events and 60*F water are in some -

1 sense " worst case".

j. 4. Page 2-C-2. .

Ditto for he = 2000.

~Yours'truly,

, g( .- (-- t & le - 9 Fred A. Nichols, Consultant

- Metal: Components Subcommitte Advisory Committee.on Reactor Safeguards FAN:ph 'T a

k.

f

).

i E

5 e

r I

9.

E D

4 i.

w- -

y w ,- -gy,.9 9yyi-.. e ,We w y ps y- 9 g"*.9 ewe-w9 7-=g w-9...p%yis-rw s-

%9gp.#angmig 7 y yy yy9g. y- +79 9, . 99 p p 9 9 +.q p.y m y .pm La+p - y y p 3- q - g v 9- p e..g enp 9p .y  % w