ML20155J541

Transcript of 981104 457th ACRS Meeting in Rockville,Md. Pp 1-172.Certificate & Supporting Documentation Encl

ML20155J541
Person / Time
Issue date:	11/04/1998
From:	Advisory Committee on Reactor Safeguards
To:
References
FRN-64FR12117 ACRS-T-3056, AG12-1-042, AG12-1-42, NUDOCS 9811120190
Download: ML20155J541 (238)

Text

- - ;yLy) ' Q_.f6

., . .y . . # . . - / ..,,,;, . s..y  ;, .* . - .; J .; . ge. ; .. . - l,, p , , : 3..? . c s .. .. c s ; . . g py .c ' ' 1' ' aS ..j j

.: _ .; - U.; '

[ y y, } J.f.D:..':..[.?g? %f;h.g..y.qf.p%QY..;.aM.;. ' .

_..(;;;;,;QqQ-&.fl&QQR

': 'j_ :. g :i..' f,r..;

T: .

-ie.- ., ,,.5 . . [ *. ;; ' ' ? b do c ,9: f $ , . .h

• • ' -.i;. s. . . , . _\;. , , O ? ' ' M ' :__ _ '_  : . ,' % ^;. f' } f, ;[f {k,?,,' } j e .; . *

,.., , .. . i . ^ - !'

. .n..,"1: ?,,,

!,~. ',[_\'.

.'::7 w..i. ) ,,, Q' , ;. ' * y <

ywL _, Ql . t., y: .,} g,lo.

^

,',,l'",~,.  :

' ' ' _} ' - -l .

l l .-

2-

, , f . : p , :;[,{,.l'-.,fg .

, ch j ( "', ]( ,p

.-.,. - .. , , . ,'#. ' )- k -[ '

4'- ,n r' A;l'.'. ,_. ... , , i- j' _ , D.:) ..*.; y , . ,' ;

[, ~glNs-' 5. .; f ' 43;

'~!'.'s : ',  : , . . .' . - --

".*-,' ^

• p. 'gg*Q-f- " <e j,', ". s. . ~ '. f .; , ., , '. , ; !.

]~ " ,

y

[. a  ;(( _ ?. J [ b y.

1

a . " " ' [ ., g . ',  ;'

'b.f * ' ,Y [ ! + >g I'. ',.'{'.

j'

_. _*[. - .

_- - .I'; < ::  ; ; l' . . i ,- .

',,h<'.,'

. . . , '..'e .

.,'l'.'..,'- *

.g , ,,j . j

.l . - . . L - .';. * ' , . . . -

. ' . ' . '.u,.._ e.

. **9 4 . 's._j - * *- * - ' ' ' .. *'.

A 3. ., . " - .; i  ; ^?- i. '

l...-

l~  ;

l: .

..s. ..s.. 3. ( , . p. . gg.;[:,e :g c.....q

. .

.... 'y  ;

'o.  ;'f J ;c.:q.

,..,;;_(...K...'.. ' . lf;;

c

.&.. , z

.,W '_ r.,.- .u..~'... R:

"; ;_ .,_ hr y

. r _

.O

.[

^*'

of,  ;

_. f .[ :. % ..

1 I f_ t

.li ' b. + /. *\

'l

, ' t J , f.,

.7,*,-) .[ ; ., .

, .[,[, ' ,f',,j','J '-g; a

S. s

.' ,g j

• ,;- y  :

'gy 'h[, '. y, - if,,' ' '. l,),f.,.Q)'._ f ': }: ;g ,,

.; ' . ; . . . . ~. 79

.

• l .. ,, .-.. ::

'l}'_,',3_l O _;':3.

{' j . , . ^i.: q e ,_ .

W ..,-r

'-;; $. ,: 9

..r. *

,f. ) '- - . .f; I-'-

.N

[.'[.' , ,

^- - t'

_f _ ' _ . , ,. .g . 4 - . .

- :.a - . . .

' n. ' U-'w .. - .; ;q. ; ~'

. .g

~-

j :g'.g -

. ,- , - :: -'C 3..

l ly.ir l. . .

d l& -l l ..

,.; .. ,, .', ; _:.; .j :,- ;. _. e..f.

..s

. ;..y .l'.. .i,.l},$ v a/ g' t"t,Q:

; 'J. .-

s , ;,

'. . . .* . . - ., . . , -' . ;..- ,,...,3 , p. .

4

'[ ., a .j . ,-

• t [ .

.i f^.. .4 7','. * .q'h.,[ - '

. _ ..f _s .'. ...._

• ,A..* .

.' '.-, . . * - .,% ; y: .'s , .: '

.. . . j [: .

)[.- [ . ..

y . ,', c,;,,

, .f .. , y[ ::,, , , : ' .

. 2 ..;;' . . [;._q x 3: mj e . ..

' c.; . .,: __. .. - . y ;i.,. :nn,f. .mQs.:;

.u. >. 5; ;-

. . . . : - , , . , j. ..i..l% f 4' .

. .. ; .g .

_, , ZL ; . , :

p i .. .- ~<: ,

, , ; t/... ~n.;

.. . ,l. ; . . _

!' .* . ' ' ' ? "-l e q.'..'.' ^* l .=. pl;

^-.

_.. . . . . . u

.. ~ . , ,-_ _- .. _... ~. . ' , . ~ . '~. . ::%" N <

-. '  :,~

. St  % ~' f

. .. .'c . -. '  :

n,

. _ . ;.s . ' ;. l . _;' %

' # ?...,;Y3'y.Y a . ' .3 . .y .; _

J; -

. . . . .. . ., . l f. .; . ;:.:. ;;-.7. .lj ^),G j .g

-. 0. .,g l , }

.al .. .

...x.-

. ; ( f,,:Q [ , j:7 j _  :/.yf.QG Q'gfj i

. .. .. I .

g

- 3 .  ; : ..g.y, jj,5 '

; .. , . ;. c.; .i  ; Q.fpf{&y .,p e q&.%jgga t.

. ' L ; f. [.J. . , , ? 'M  ;' : "l .'.R., q;Q:fy% '

1

a. ..  ; _;

J..;.; q,; Q LF f; .{.:[_.Q f Q M yi: {)

y.>

^

-- ; m. .a e: . s. d b .W:QQ'1 ,

_ .; t.; y;p.4 3

~

'..v

'. ;;' a, ,

..y

e.  :.

.:c1 .- .m,  ; ~...e n.: .x . 'k

,, . , Q.y. ., ,, .. _.[: '. .4/4. ;y .e < g.4,jh [a ,f. (QMlj

.) s . . -

-g . ,,

x gw. 49.

• N ' ,. Y L _

Q. ' - [:' . ,}f j. v. .

4 s!, ',.; ...

- . . , .f.

_ yg;.c . .. .:s A. ;. . n g .p 7uny.,..;. '.w... ...v:r.yj,a '

3 m.  ;

.{ .?t}.'..: ,

- ? ' .

....,;;.. j pg :g 999 .

. . . .. . ... ,~. . .  ?.;

. . . , y. . 7,,...c .,g.,. ;3 y, . . ,. , y.:n. .

f

.,,..l,', l :.'*e.G . W*, . , lp. :_ ' , y' 2 . .. : .:. , ': ; .

t

,1 : ...; v'. ...

b' : _x.ev.e;.

. , kp v.. . gs1 _....g.

..x . g.4.x . pu

- . qlQ"I+ g F

p .-

0 -

I - ' - .' ' g[ .* . ,, e' . g# . . . . 14*M ;p e

= - ' '. ' ' ' [~ 1  ; .,

i {.].y, '

. ;. ,; 'l , L -[:. u y; : .k3.).. ., ;

.:. .: 3

t. q.n/y.' g @ t.; p p. ,, ,

. , . . - - .. .. - . . a , . . n .

.' g i '"*."[.t.'l- .*qI

. . a f,

[-. - .

g. ' ' i, p. ,,

. . '. . . . , .p .. .. ..,.f.. ge .1, . . 7 y , . .,.'s .t ..

.. , * ,- . ~. : r. * -

-y'  ; -l .* ..,

.. ..>;_ _ l4 g ', .e ' . . ' . '

. ., n ' (r _ .;

.- y L;.

9:g *, jn[' 1, yf f,y - ,

v. ., a s,.s

' .L,7 *

.1 .

c... ..{, ,, ,

. ;., f m 57,

. . . . .o , _ - -

.a j , - g y , t

. u _;, .g wn 1 , ,7;p

z. . .. ?. . ',f'.

. . -h ' , 'l ', ', ..; . . ;.

'N

~ -

g.

.g.L.,>; g ' 'S j...,.,t.> .

, ,, L'f;( ..

'i ' l l. ' - . h *. .; l .L.{.

,}::. 6 _ .

t, l  :. ; .

. ;, y. .  ; . . .

. e

I ',j .i

_ ~ . . _ , . . . . .. .

..------n.._

y> y V s i * ,

%('

.p'". ^% w ;q- ;s ,,- ....:'

pm ftg< ,g c 2* +

+J, w

, - q. ,, i r 4 r

h.....?ggYh,ah

, 'Y l . Y Y $ N Y. A$0ST~SOSY .,

}-Q%_

-.- j .,,l l b

% IOFFIDIAInTRANSCRIPT OFlPROCEEDINGS

.2

-+r_ ,

S

- ~ . :

. ~ . . . .

&w-%,w_- '

y, s  %; $* iNUCLEA, .RLREG. ULATO, RYMCO.MMISSION? o i - -

-lV R W-l. v a .:a t J !- -

Q4 &_ [l} V p

- > ..'..'~.t . .' ! ! . -

%y:

~

a n? :

Ln

' ' .9.6 4f. .a >. 'k

,g,%g6s .-.

' ,. #ADVISORYJCOMMITTEE ON. REACTOR' SAFEGUARDS N. Is

~- /i -

- - o,f t h 4

%. c

=mg_t;yl, m%. - 4 4 w  !

s , y >

ah i -a c< .. ' - nb ., e s >:. r.:" ,J w

a j

E  ? s

? '

T @;c ~3 '

w&w?w w ~+y M.n?

Title:

i L' JMEETING':7.45.7TH'ADVBORYN ug . . ;n r 4

hq;3[hk,b. " ^  ;

ha p.y~ s gyg we pg ~ t ~..a ;, s i [COMM, ITTEE DNiREICTOR!

^[ ,

[3 $dkgh-3

7; 3

.j

~

-n + X. 'f iSAFEGNARDST m sts

, , YW wy$.:' n$ . 2 s.~

o C. '. e

.u,

_ O y

J .yn: .

' ,_ u 'W gp')K ,; ? ?'R G~ h * 'I, .

g ,,

w

, y s hkha s ' ^ TRO4 (ACRS) 3 y

/

wm '

RE W ORIGINAL 2 0 smurE M)$$3k MMF ,; N M f [t y J i '

M WS T-2E26 U

&,s ~ ~ cx 1

ys ' g37% ~

m-n , 4 415-7130 j f >tgaw.+ % %,

G,b f

~-

u?

4 r 1": i THANKS .'

,hv i

4 3 4 i

@M@h w AM I M W, P. hJ P

t '

,T ~ %, -

g%wm.s we oe wwm

h. c:p ma y 4 .

. 1 g&

t, QNO , \tU:Y,r 5 ., .

• W ' . I, @ 'w~o

~ y  %

' .. g .

1 1

n. M ~ff v 3 N::%yW;. r- Q' 'O. <,

V^ .

e m

HMyWhylischet>N6!:b s

• .' T, x ~ ,%p~d,. g

,

• 4. . ~r e ' T q,,.

I

g.1 t ow' "

y

, m

m. J ,~

1*

m. _.

[- $  : .7 -

,pr .s e #.-T %g'

, i

• j 1

4 s-y , r ,4 j

<k.'

g.

,- sg i '

e t r,.<

, ,, .1 wa g "s' g <,i %,g1 f,y, $ ~wA ,e]',x 3, .

+ >

J y,. /}. J , y e < . , m a > -

4 g .,

/

yNg Ifl T]? *

% d. ~ f'  : j" y'.

%*

• f'

<. : > i ,  ;.

L V6s 'N1 ' -

~

'?

A.c Fh. [MQ :e o H MSd, w b;;,p,n. wlg. @ %, , U:W U /;- G:C /p 107 ,

.o . <

s .

, ~l rn 1 # a W un 2 a x y w,1 t

.4 WMy#p -m  %.. .

, 4 - .

^

'v.

, . . o ,

. , 3d u n m[;e z har160Eder)No.:1 fhSB;300-531 SihdS pg fh? qM N, h sw.;- -

v - + -

,1 t

i-x .

j c,

QDW:A&.Nd

'r e a t, p M w 4 N,hMw 3 - ,

+

r 3

)

91: > . -

i

&yr. , 4$f ,een af l , .~.fr

?' ,, , -

Jl

>.h %-

l y .i x

n - ; \ ;l- y-pgq,u:m,. g o r "a  ;; e- .

. i

. 4 v

, n,. s.

,43 Q ,

yp. .h>.'

%,f'$', t L. ,

i h ,/ ?: P , w & ? lod TIONE

Rockville,MD '

a;w&m%(..Vs:$ [.Q -o.?7%

s i & Q. * .

W W h @.& $ f l"( ~' 9811120190 981104 QMlf ' '

• PDP ACRS M k , e.. W 1-3056 PDR

.4 My . m '

j DATE:

4g

$ry Ma S;

,h&.. . :g' <

,~ u, ,.

(2 Wed..esday, November 4/1998 J PAGES:1 - 172 i c ' ,

Q& us ;;W p .., , f ps , . - . , .

9g OEU  ; ,ANN . , RILEY dc ASSOCIATES, LTD..

r

,j, y x% d C+:

Q Nygp u# % * ,' 7, A . ,192 q.

G \;, w >' ,

- ~

j p.,

x

% ' _c.

m. 5 Connecticut- Ave 6NW, Suite.1014 ' .

V V9..

on, DA.;.. Dh. 'Im , % .

j h. j f

,j l

N _>.,,

q , v " g s- j t, a a me. '

i m%Qgy, -qv %,; ~g - "w su,

, . yMfa. w%; m..,-r~ n w~w4 v

3 ,

, me Me

- 4

~

,. , . . -~ . - . . . - - - - - . - - . . - . - - . . - - - . - . . . . - . . _ . . . . . . . - . - . . - - . . .-

i

(

l

(

l ..

?

I l

r DISCLAIMER UNITED STATES NUCLEAR REGULATORY COMMISSION'S i ADVISORY COMMITTEE ON REACTOR SAFEGUARDS l i NOVEMBER 4, 1998 l

The contents of this transcript of the. proceeding of the United States Nuclear Regulatory Commission' Advisory l I' '

s Committee on Reactor Safeguards, taken on November 4, 1998, i i

as reported herein, is.a record of the discussions recorded 1 at the meeting held on the above date.

(

This transcript had not beex, reviewed, corrected i

and edited and it may contain inaccuracies.

l t-i I l I

l l

l

=

1 1 UNITED STATES NUCLEAR REGULATORY COMMISSION 2

't ADVISORY COMMITTEE ON REACTOR SAFEGUARDS

'3 ***

4 MEETING: 457TH ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 5

6 U.S. Nuclear Regulatory Commission 7 Conference Room 2B3 8' Two White Flint North 9' Rockville, Maryland 10 Wednesday, November 4, 1998 11 '

12- The subcommittee met, pursuant to notice, at 1:02 13 p.m.

14 j 15' MEMBERS PRESENT

)

16 DR. ROBERT L. SEALE, Chairman, ACRS l

17 DR. GRAHAM WALLIS, Member, ACRS 18 DR. JOHN BARTON, Member, ACRS 19' DR. MARIO FONTANA,HMember, ACRS i 20 DR. ROBERT UHRIG, Member, ACRS 21 DR. THOMAS KRESS, Member, ACRS 22 DR. DANA POWERS, Member,'ACRS i 23 DR. WILLIAM SHACK, Member, ACRS 24.

25 I-I'(/[]

l ANN RILEY'& ASSOCIATES, LTD.

Court Reporters L 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l

._ _ . . . _ _ . ._..m . , _ . . . . _ . . _ . _ _ . _ _ . . _ _ _ _ _ _ _ _ . . . . _ . _ . . _ _ .

1 2

1 PROCEEDINGS 2 [1:02 p.m.]

3 CHAIRMAN SEALE: The meeting will now come to 4- order. This is the first day of the 457th meeting of the 5 Advisory Committee on Reactor Safeguards.

6'. During today's meeting the Committee will consider 7 the following: a proposed rule on the use of revised source 8 term at operating plants, plant application of revised 9 source term to the Perry nuclear plant, assessment of 10 pressurized water reactor primary system leaks, and proposed  :

11 ACRS reports.

12 This meeting is being conducted in accordance with j 13 the provisions of the Federal Advisory Committee Act. Dr.

14 John T. Larkins is the designated Federal official for the 15 initial portion of this meeting.

16 We have received no written statements or requests 17 for time to make oral statements from members of the public 18 regarding today's sessions. A transcript of portions of the J

19 meeting is being kept, and it is requested that the speakers 20 use one of the microphones, identify themselves, and speak

21. with sufficient clarify and volume so that they can be 22 readily heard.

23 I'll begin with a few items of interest. You all 24 have I think at your spot -- at your seat a package of about 25 a quarter of an inch thick of items of interest for this Q ANN RILEY & ASSOCIATES, LTD.

L/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034

/ _ . -

. _- .-- . - . - _ _ _ ~ . - . . _ - . - - . . - . . - _ _ _ _ - . - _

. _ ~ _ - . -

3 l

1- . meeting.

i

)

(~'T 2 There are.actually quite a few items there:

r \u,) .

3 several speeches by the Chairman on risk-informed regulation 4- and the redefinition of certain aspects of the Commission's 5 configuration as they move forward, the evolution of the 6 relationship between NRC and DOE, and other things including 7 a statement on international cooperation and regulatory 8 perspectives, which dovetails rather nicely with our recent

9 quadripartite meeting.

10 There have been two Commissioners appointed, 11 bringing the Commission to full strength. One was the l l

12 reappointment of Greta Dicus as a Commissioner, and the ~

13 other was the appointment of -- and I want to be sure I get 14 the name exactly right -- Jeffrey S. Merryfield as the other l[J) %

15 Commissioner. There's a writeup on both of the l 16 Commissioners attached to this handout.

i l- 17 There are several other things, but the one other 18 thing I want to mention that can impact us individually is s

19- the replacement of your travel card, which is currently an 20 American Express card, with a Citibank Visa card, and L 21 there's a drop-dead date in it. Again, that's attached to

'22 your handout, and I'd urge you all to be familiar with that u 23 so you don't get burned.

l 24 There are several other items of interest.

l 25 Unfortunately one of them is the sad obligation to inform J

i 4

['} ANN RILEY & ASSOCIATES, LTD.

' - (_,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 l

(202) 842-0034 t

4 l

l' those of you who don't know of the death of a long-term, l l ("'s 2 long-time senior member of the staff, Long-Sun Tong, who was 1

3 83 -- I didn't realize he was that old -- died on the 31st

-4 of October. He had pancreatic cancer. He had a career that l 5 I think certainly all the Members of the Committee are at l

l 6 least partially familiar with, and he certainly will be 7 fondly remembered by all of us who had the opportunity to l 8 work with him at one time or another. In fact, I understand l

9 the funeral services will be tomorrow afternoon late in 10 Gaithersburg. We'll all miss him.

! 11 Do I have any other things, John, that I ought to 12 mention at this point?

13 DR. LARKINS: We passed out a copy of a proposal 14 for handling the workload for November and December for ll~~s,)

15 individuals to take a look at.

16 CHAIRMAN SEALE: Right. I'm sorry. I forgot to 17 mention that.

18 We have a load both this month and next month that 19 is enormous, and we're going to try to do something a little 20 bit different in the way we do letters to allow us the 21 opportunity to both discipline ourselves in the amount of 22 time we spend and also to not impose as much on the staff 23 and provide Members the opportunity to work on specific 24 inputs they have.

25 The schedule or this piece of paper that you got, i

(g ANN RILEY & ASSOCIATES, LTD.

( ,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034

.. . _._ _ ___. __. _. _ . - - _ _ _ _ _ _ ___.__m . . _ _ . _ . .

l 5

1- each of you, has -- it's entitled " Key Feat. 10 of the

/)i.

2 Proposed Approach," sort of getting it halfway through. It l 3 indicates the allowed discussion and word engineering times, 4 if you will, for the seven, eight topics that are on this l-l i

5 month's agenda. We really need to finish these things 6 because if this one looks tough, December looks ferocious.

7 What we will be doing, if you go to the last page, 8 is we'll schedule our consideration of items according to l 9 the time that's given there and according to the goal that's 10 given there. If we don't use all of the time, each of you 11 will have time available to work on outstanding letters that 12 you may want to modify based on discussions with other l

l. 13 Members of the Committee and feedback.

l 14 We will not take up other letters out of seqJence.

()

15' This will allow the staff to schedule their attendance such

j. 16 that they are here when they need to be here, but we're not 17 going to hold people around here till six o' clock or seven l

18 o' clock, and then as it turns out not need their 19 availability.

20 It's important that each of you examine the drafts 21 as you get them. If you have substantive issues to discuss j 22- with the author, let them know ahead of time. We'll try to 23 handle that as much as we can in that mode. If we get hung 24 up, I'm going to ask the author and whoever the individuals 25 are that seem to have a problem to get together and.see f

-O \ ,/

ANN RILEY & ASSOCIATES, LTD.

s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i

l

r l 6 l

1 if -- outside the full Committee and see if they can-l 7

\-

2, reconcile their problems. I remind you that the alternative i

~3 of additional comments is always available. It's a L 4 time-honored option, and we shouldn't ignore it if we cannot 1

5 resolve any conflicts we have.

6 Dana is, as you all know, the junkyard dog, and l l

7 Dana will, as I commented earlier, you can lead a horse to 8 water, and we're going to find out if the junkyard dog can 9 make him drink. So we're going to be on good behavior and 1

10 hard discipline this time if we're going to get through.

11 Dana, did you want to deny any of the charges that 12 I've made?

13- DR. POWERS: I think it's just important that 14 people recognize this~as an attempt to try a new business ,

,7m 1 15 practice, that our intention is to -- what the Planning and l {% /}

16 Procedures Committee hopes we will evaluate it for the next 17- couple of meetings and then try to assess it at our January 18 meeting to see if we want to adopt it as a continuing 19 business practice. Undoubtedly we'11' learn warts about it 20 as we go through it. It's simply intended to enhance our 21 productivity and efficiency with which we use the staff, 22 especially since productivity is the hallmark of the world 23 nowadays.

l 24 CHAIRMAN SEALE: Yes, sir.

l 25 DR. UHRIG: Is it -- perhaps you or John can i

ANN RILEY & ASSOCIATES, LTD.

)

. i Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 i

(202) 842-0034 l

I 7

1 answer this -- is it official that there will be no report i

2 to Congress?

(~}

\ _- I 3 CHAIRMAN SEALE: We don't know yet. We should 4 know probably sometime this week. And we will discuss that 5 in any event -- i 6 DR. UHRIG: Okay. I 7 CHAIRMAN SEALE: At planning and procedures.

8 Any other comments on that?

9 DR. UHRIG: Well, the question is whether we spend 10 the time --

11 CHAIRMAN SEALE: I appreciate that.

12 DR. UHRIG: On that letter, which may not be sent, 13 or whether we go ahead and work with the -- l 14 CHAIRMAN SEALE: I think if you look at the l (m) 15 schedule that we have,_that's pretty late in the day anyway, 16 so --

17 DR. UHRIG: Okay.

18 CHAIRMAN SEALE: Okay? We're likely to get 19 resolution on that one way or the other.

20 Okay.

21 DR. LARKINS: Bob?

22 CHAIRMAN SEALE: Yes, sir.

23 DR. LARKINS: I'd like to just pool the Members to 24 see if there's anybody who can't be here next month on 25 Wednesday by one o' clock for a meeting with the Commission.

l l

l l

/^ ANN RILEY & ASSOCIATES, LTD.

(_,)g Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

I l

8 l 1 CHAIRMAN SEALE: I think the most likely

~

2 difficulty people will be the two that aren't here today on f')'T 3 Wednesday at one o' clock. So maybe that question would be 4 better posed to them when they show up tomorrow morning.

5 DR. LARKINS: Okay.

1 6 CHAIRMAN SEALE: Okay. Any other comments? Sam, 7 do you have anything? Okay.

1 8 Well, our first item on the agenda this time is a  !

9 proposed rule on the use of revised source term at operating 10 plants and the pilot application of a revised source term at 11 the Perry nuclear plant, l

12 Tom, that's your subcommittee, so I'll ask you to 1

13 introduce our speakers and run the meeting. '

14 DR. KRESS: Thank you, Mr. Chairman. You'll find t

(/~~)

</

15 this background information under Tab 2 of your book. You 16 of course have all been involved in the activities to revise 17 the source term that's in NUREG-1465, and you recognize that 18 the revised source term was originally intended to use with 19 new applications, new plants.

20 Well, the staff had a number of requests or i 21 applications to apply the same revised source term to 22 operating plants, and they have done a great deal of work to 23 look into what the implications of that is, might be, and

! 24 have come to the conclusion that they need to do some 25 rulemaking. So I think we're going to hear today some --

i f ANN RILEY & ASSOCIATES, LTD.

(,N/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036

[

t (202) 842-0034 t

f

9 i

i there's a briefing on rulemaking that would allow such fg 2 application to operating plants and maybe a briefing on the V 3 status of one particular pilot application.

4 So with that as a very brief background, I'll turn j 5 it over to the staff.

6 MR. EMCH: My name is Rich Emch. I'm chief of the l

7 Radiation Protection Section in NRR. With me today at the l 8 table is Jack Roe, the director of the Division of Reactor l

9 Program Management, and Charlie Tinkler, the acting chief of 10 the Accident Evaluation Branch from Research.

[ 'll Just a little bit of a -- a few introductory I 12 remarks here. First, a little bit of history. We've been 13 pursuing several activities related to the implementation or 14 facilitating the implementation of the revised source term

()

%f 15 at operating reactors. Rebaselining was one of the things 16 that we did, and we discussed that with you some time ago.

L 17 There was also that. That was presented and the work was l

l 18 done principally by the Office of Research.

19 The second activity is rulemaking, which is 20 principally what we're here for to discuss today. We have a i

12 1 proposed rule that is on its way to the Commission. This is ,

i '

22 one of the items that's on the' tasking memorandum. It's 23 VI (j ) on the tasking memorandum. And we're here today to 24 make a presentation to you and hopefully get a letter from 25 the ACRS approving or endorsing what we're trying to do.

I

,G . ANN RILEY & ASSOCIATES, LTD.

( ) Court Reporters

., 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036

.(202) 842-0034 L

., . . ._ . . .. . . . - - . - . . . . . - _ ~ . . . . . . .- . - . - ~ - --

I

! l 10 l

1 I want to thank the committee for the flexibility i .

I'i l A~sN 2 that you have shown us on the schedule in terms of how much l

i-3 info we had to get to you when and getting this meeting E 4 . scheduled as early as you did. We appreciate that.

5 I want to point out that one of the things along {

l 6 that line, when we gave you the draft package to begin your 7 deliberations, we also put that package in the PDR. We did 8 'that at the request of the Commission. In the SRM they sent l

9 us, they said they wanted to have that done, so that was I l

10 done.

I 11 The first speaker today will be Steve LaVie. He 12 is going to talk about the rulemaking situation or the 13- rulemaking plan and the proposed rule. That will be most of 14 the presentation. We understand that you do have an

( 15 interest in the third item in our sort of pursuing of the 16 implementation is the review of the pilot plants.

17 There are five pilot plants for implementation of p 18 the revised source term. The so-called, what I would refer 19 to as the lead pilot plant is Perry. It's the one that we 20 are concentrating the most effort on right now, and there l

21 were operational reasons for making them first.

t 22 We are also in active review on the Indian Point 2 t

L 23 application -- that's a PWR -- and the other three are the 24 Grand Gulf, Oyster Creek, and Browns Ferry, and they are at 25 various stages of where we are going to be at with them.

l l-4

/)

\_/

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 (202) 842-0034 l

t

,. - - - .- .. - . ~ _ - - . . - . ._- . . - - ._. -.. - - - _ - .

l 11 1 None of those three are under active review. Their L [~\ 2 submittal isn't here or there is some question about whether

' \~

3 they are going to continue to be a pilot.

4 The active review of Perry and Indian' Point is 5 being undertaken by a task force both of NRR and RES l 6 personnel from all the technical groups in NRR and then with l 7 the help of the people from RES who did the re-baselining l 8 work.

! l 9 Probably if there is any conclusion, we don't plan 10 to give you a lot of conclusions about Perry because we are 11 in'the middle of an active review on that, but one 2

12 conclusion I think we can give you right off the bat was it 13- has turned out to be a more technically challenging issue to l l 14 fit the revised source term to the design basis for a BWR 1 r

( 15 than it was for a PWR, We had a little bit of experience g l

16 doing it for a PWR on the AP600, as you will all remember, l 17 and we have found it a technically more challenging issue to 18 do it for a'BWR as it is more complicated.

19 One of the things that we are trying to be careful 20 of in doing that, you will remember that when we were l

l 21 ' discussing the AP600 you folks had some questions about how i

22 much we were relying on the particular thermal hydraulic i.

23 conditions of a specific sequence for design basis purposes 24 and we have tried to keep that and we are keeping that in I- 25 mind as go through the Perry review.

l' i

k

, /"'g ANN RILEY & ASSOCIATES, LTD.

j jg ,f Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 L'

(202) 842-0034

._ _ __ _. .. _._ . . . _ . _ _ _ _ _ _ _ _ _ _ . . ~ . . . _ . _ . . _ _ _ _ . - _ _ _ _ . .

l 12

l' As I said, there is no conclusions yet. We will L

I 2 have to be a little careful'about what we say today. Some 3 of the information is proprietary and there are, because we 4' are in active review, there's things that we just wouldn't I 5 want to say here that we haven't discussed with the licensee 6 yet, so with that introduction -- Jay Lee will be doing the i 7 presentation on Perry -- so with that introduction, I will  !

I 8 ask Steve to lead the way.

i 9 CHAIRMAN SEALE: While you are doing that, let me l 10 just point out that we appreciate your comments regarding l

11 our ability to fit in the schedule and so on, but we can't 12 - promise that we will do that for every other thing that ,

4 13 comes along. It turns out 1465 is something that a lot of )

14 us had some interest in for a long time and we are more up i _)

1

[ T 15 to speed in that area, and that may have had something to do l

! \ms/ 1 16 with our willingness to go at this level of information.

17 MR. EMCH: Whatever your reasons, we are grateful, f 18- [ Laughter. ]  :

i 19 CHAIRMAN SEALE: Don't count on it again.

a 20 MR. LaVIE: Good afternoon. 'I am Steve LaVie from l-l 21~ Radiation Protection Branch. I am here today to discuss 22 with you the proposed rulemaking package for using the 23 revised source term at operating reactors.

, 24 You have probably already noticed that we made a 1

25' word change in that we have been typically referring to this i

i 1

/"

• ANN RILEY & ASSOCIATES, LTD.

(, ) Court Reporters

_ 1025 Connecticut

Avenue, NW, Suite 1014 Washington, D.C. 20036

( (202) 842-0034

\

i I~

e

. . .- . . - = . . - . . . - . - . - - - ~ . - . . - . . . .

13 l '1 as revised source term. For the purposes of rulemaking we 2 are going to call it " alternative source term" such that it 3 ' will give us the flexibility if we should de'ide to have

4 another revised source term at some later date. We won't be 5 trying to figure out how to word it, so we are going to use 6 " alternative" all the way through the rulemaking.

~7 A little bit of background on this. Of course, we  !

8 have been before the committee before discussing this issue. 1

9 One of the things we saw as the staff is that we  !

10 considered the source term to be a very significant part of 11 the underlying underpinnings of the design basis. The r

'12 source term gets into an awful lot of the play in the design

) 13 ' of the plants and we felt that we needed to be really 14 careful at this point when we start messing with that l

( 15- underlying pinning is that we didn't overlook something. We 16 wanted to make sure.that we pul' led all the strings.

l- .17 As.a result of this, the way we felt about the p 18 source. term, is it kind of drove us into this rulemaking.

19 The objective of our rulemaking was to provide a

. 2 0. regulatory framework to allow current operating reactors to

' 21. - use the insights we learned in the work in 1465 as they L 22 wished to implement cost beneficial licensing actions --

23 DR. KRESS: Before you go on, what you just said 24- is just slowly sinking in.

j 25 MR. LaVIE: Okay.

L /s ANN RILEY & ASSOCIATES, LTD.

L( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L

m .4 e

, ., .. - . . _ ~~~ .. . - _.. . . . . . . . . . . . .

I 14 l l

.1 DR. KRESS: Does that mean you are actually saying 2 that the source terms the licensee ends up using to meet the 3 regulatory-dose requirements now becomes part of his design l

l 4. -basis also and not just the resulting plant design features?

5 I wasn't-sure I heard what --

l l 6 MR. LaVIE: We' consider the source term to be a l

7 very significant input to the plant design. There are an l 8 . awful' lot of the design features in the plant that were

9 contingent on the source term --

l 10 DR. KRESS: On what source term you use, right.

11 MR. LaVIE: Okay? I'm a little reluctant to get

l. 12 into the semantics of whether or not the actual percent of i

13 core damage is in the design basis but is it a significant

.14 assumption that led into that design basis. '

O i i 15 DR. KRESS: It sure is. I agree. Thank you.

\_J l - 16 ' MR. LaVIE: Thank you. As I said, the reason 17 we're doing this is the industry has. expressed an awful lot

[

18 of interest in using the revised source term to regain what l 19~ they believe to be some benefits in plant design changes, 20 operating procedure changes, maintenance procedures.

21 When we started getting into this rulemaking, we L 22 noticed there was a couple places we needed to make some l:

23 conforming-changes. This was also identified when we did l

24 the AP600 work. We found there were some places that didn't L 25 get changed when we previously revised Part 100. We are 1 - '/ ANN RILEY & ASSOCIATES, LTD.

l( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034 i-t I, _ _ _ _ _

15

'l- . including these changes in this rulemaking.

2- DR. WALLIS: The purpose of this is not just to

(}

~3 help industry, is it? It must be to make the rules a closer i 4 reflection of-reality.

5 MR. LaVIE: A closer reflection of reality? I 6 think we could say that.

\

7 DR. WALLIS: And that would seem to me a primary 8 goal rather than helping industry to save money. If you put 9 it first --

10- I MR. LaVIE: Right. 1 11 DR. WALLIS: -- and said now we are going to do it 12 more correctly, it might be more convincing.

13 MR. LaVIE: Well, correctly is --

14 DR. WALLIS: Is it less correct?

r

,~ .

() 15 MR. LaVIE: These analyses were done over the 16 years in design basis space, were always intended to be 17 surrogates to test various plant designs.

18 We have been very conservative as we didn't have 19 the science to do these -- the work that we are currently 20 able to do. In the context of which the source ~ term is 21 used, I hesitate to say.that what we had previously was 22 incorrect.

23 DR. WALLIS: No , no , no -- but is this more

'24 correct, this new one?

J25 MR. LaVIE: We believe --

. /~) ANU RILEY & ASSOCIATES, LTD.

(m /  :

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C.'20036 (202) 842-0034

16

-1 DR. WALLIS: They are all wrong.

2 fMR.'_LaVIE: We believe it is more realistic, more V[) _ _

3 representative of the source term that:might be experienced 4- during an event.

5 DR. WALLIS: Good. That's good.

6' DR. KRESS: I.think that is a better way to say 7 it, rather than more correct, because we're not real sure 8 . what design-basis space -- I

9. MR. LaVIE: Exactly --

10 DR. KRESS: -- things ought to be, because they I 11 are intended to be surrogates --

12 MR. LaVIE: Right.

13 DR. KRESS: -- and whether they are correct or i

14 not, what you want them to do is be sure they result in the

[)

v

' 15 right design features that deals with the things you want 16 them to. 1 17 MR. LaVIE: That's correct.

18 DR. KRESS: And so more realistic I think is a )

19 better way of saying it. I 20 MR. LaVIE: More realistic is a better way of 21 saying it.

22 DR. WALLIS: It's closer to reality. Yes, that is  !

23- what you mean.

24 MR. LaVIE: As we understand it. One has to

25. recognize --

f}- ANN RILEY & ASSOCIATES, LTD.

A ,/

s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014  ;

Washington, D.C. 20036 (202) 842-0034 i

. , .. . , , - ~ - - - ., .- - ~ -.- .. _ - . ~_.. .- - - .

I 17 1 DR. WALLIS: I get bothered when you sort of say

. 2 well, it's design basis space so we can be unrealistic.

3 DR. KRESS: Well, you can be unrealistic if you i 4 want to in design basis.

5 DR. WALLIS: Yes, but that has always bothered me.

, 6. MR. LaVIE: Okay. l i '7 DR. KRESS: The pressure source term in design 8 ' basis space is unrealistic.

l 9 MR. LaVIE: What one has to recognize is that 1465 10 was generated out of the insights that came out of severe 11 accident work and we're applying insights from severe

} 12 accident work to design basis space where there is always a 11 3 ' disconnect.

14~ In developing that source term, the folks that

() 15 worked up 1465 looked ni several accident sequences, so what 16 is in 1465 doesn't represent any particular sequence. It is 17 -intended to be a surrogate for a wide range of events that-18 might occur within what we-consider to be the design basis

19. events. l l

20' The key change in our rulemaking is we identified l 1

21 a new section in Part 50. called 50.67 and there are several l 22 decisions that went into why it became a new section and we 23 will get'into a little bit of those a little bit later.

l 24 The date -- we have had a lot of questions about 25 why this date, why isn't it the effective date of the rule l D ANN RILEY & ASSOCIATES, LTD.

' (,,) . Court Reporters  !

1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 I (202) 842-0034 i

i k

18 1 and so forth. The date was chosen because that was the date 2 Lat which the current Part 100 and the Part 50.34 changes 3 became applicable for future license applications.

4 This class of. plant have to meet the TEDE 5 requirements, so we were dealing with a group of plants 6 before that, so we decided to use the same time cutoff for 7 those-.

8 The new section was also chosen because there was 9 a desire to co-locate development requirements that have to 10 do with the use of the new source term, and we also looked 11 at the philosophy as separating plant design from siting 12 requirements.

13. Under requirements in the section, because of our 14 belief that this is s significant portion of the design

() 15 basis, we'will be requiring licensees that wish to use the 16 new source term to submit an amendment'for their first 17; application. Once they have the new source term in their 18 design basis, then the existing procedures for effecting 19- plant changes would come into play, but for this first 20 change to get into the design basis, we are telling them to 21 apply for a licensing amendment.

22 The reason we are doing that is the staff feels 23 that-the source term, being a significant part of the design 24 basis, there is an awful lot of methodology changes 25 involved. And the staff feels that because of these ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington,. D.C. 20036 (202)' 842-0034 I

s. .. .

i 19 l

l 1 methodology changes, the staff wants to have a prior review. l 2 When we look at the existing practices on the

'./ N4

'w/

3 50.59, we expect that the revised source term in most cases 4 will reduce the dose, and since the criteria in 50.59 is an

)

5 increase in the consequences, 50.59 didn't provide us the 6 assurance that these would get prior staff review, so we 7 specifically ask in this rule that it will be a license 8 amendment. '

9 DR. KRESS: Your little bullet there.

10 MR. LaVIE: Okay.

11 DR. KRESS: What you mean to say there is that by j 12 source term is expected to result in decreased calculated 13 dose.

14 MR. LaVIE: That's correct. In all cases we are

[) 15 talking postulated calculations.

x 16 DR. KRESS: Okay.

17 MR. LaVIE: What actually happens in the event of 18 any real accident is probably going to be very, very 19 different than what we are projecting, as we saw with TMI.

20 DR. WALLIS: But, nonetheless, there is something 21 better about what you are proposing.

22 MR. LaVIE: Right. We are taking some of the 23 conservatism out.

24 DR. WALLIS: Are you going to show us sometime why 25 this is a better approximation to what is likely to really l

L

/% ANN RILEY & ASSOCIATES, LTD.

(,_,/ . Court Reporters

'1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

, . -.- . _~ . . .- .- - . _ _ .

20

-1 happen?

-[T :2' MR. LaVIE: I wasn't planning in this Q,i.

presentation,- but that was --

4 DR. WALLIS: Your presentation is all going to be 5- on regulatory.

-l 6 MR. LaVIE: On the regulatory. A lot of that,  !

7 what you are asking for --

8 DR. KRESS: Yeah. Yeah, we heard that background 4 some time ago. Yeah.

. k

.. DR. WALLIS: So my colleagues have heard all the

11 other stuff that I am asking about?

12 DR. KRESS: Yes. ,

-13 MR. LaVIE: When NUREG-1465 was approved, a very, i 14 very long peer review process, that was gone over in great

/~'\ - 15 detaa.l.

.(j .

16 DR. KRESS: We heard what was wrong'with the 17 ' _ TID-14844 and'why the new one was a good idea.

18 DR. WALLIS: Okay. That has already been done.  !

l

.19- Okay, j 20- MR. LaVIE: Okay. ]

21 DR. WALLIS: Thank you.

22 MR. LaVIE: The'second part of this identifies the 23 licensees, that the application will be required to contain

24. an evaluation of the consequences of the applicable design 25 basis accidents that are currently in their FSAR. We did

. ANN RILEY & ASSOCIATES, LTD.  !

- ["'N s_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. . . . . . .. -. .- . . .- - . - . . .-.. . - - . - .- _ .-.-. . ~.

, l 21 l

l 1 this, the choice of words here, we used evaluation because

!.[~%

M 2 it is not our intent that in every case the licensee will j 3 have to recalculate all of the dose calculations ever done 4 in their plant. There may be cases where an evaluation can i

5 be formed in an engineering space, and say we believe there  !

i1 1

6 is sufficient margin, there may be enough justification for )

7 the staff to accept that. Other cases, calculations may be 1

i 8 necessary, but the rule has been written to require j l 9 evaluation.

i 10 We also specified those in the security analysis 11 report. That will limit the request to what is currently in 12 that plant's design basis. We won't be identifying any new l j

13 events, particularly for the older plants that may not be up 14 to speed with the SRP analyses.  !

l /~%

( ; 15 DR. KRESS: Just a word of clarification on there.

16 MR. LaVIE: Okay.

l 17' DR. KRESS: I don't know if you are planning on 18 getting into any of the details of the activities that led 19 up to your decision to go this route, but my interpretation 20 is that you did enough evaluations to make a judgment that l 21 the risk changes were small enough for likely changes that 22 would result from this rule --

R23 MR. LaVIE: That is correct.

24 DR. KRESS: -- that you don't have to say come 25 back with this, with a 1.174 type analysis.

ANN RILEY & ASSOCIATES, LTD.

, (O

.,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 L (202) 842-0034

_. ~

_ .. . --- .-. ~ ~ . . - - _ . . . . _ . - - . - ~ .. .- .. -. .

l l 22 l 1 MR. LaVIE: That is correct.

2- DR. KRESS: Because you have already done a not l-I L3 fully' generic,.but enough -- '

l 4 MR. LaVIE: Right.

i-

! 5 -- DR. KRESS: -- analysis to give you a good 6 indication that that wouldn't be necessary.

7 MR. LaVIE: We have gotten a good feel as a result 8 of the work that was done by the research folks on the 9' rebaselining to understand that, for the cases we have 10 looked at, that we do not believe that using the new source 11 term is going to create risk problems. We have got a safety l 12 goal, things of that nature. So, however, the Commission

.13 has asked us, in the Regulatory Guide, to address the L 14 concept that there may be applications of the new source i-15 term where'they may cross a risk threshold.

16 DR. KRESS: So you would look at each application 17 individually?

18 MR. LaVIE: Individually. The rebaselining study i 19 looked at.three plants. We tried to pick three generic type 1

20 . plants, but as we found out when they did work on the IPEs 21 and so forth, that there is nothing really generic about

.22 risk studies. So, and it gets even worse in design basis I~

23 space. Plants were licensed against different assumptions.

l 24 So that is why we need to look individually, but we believe L

25' the rebaselining told us that we weren't way out in left i

\

f C ANN RILEY & ASSOCIATES, LTD.

j ( ,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 I

(202) 842-0034 l-

D +~e a 4 G ---m - , , .

1 23 1 field.

i

('] 2 DR. KRESS: Or likely to get --

\s / l 3 MR. LaVIE: Right. But what we have really done 4 in this rule, other than indicate that the first pass is 5 going to be in the license amendment, is that we have 6 brought the TEDE exposure limits that were documented in 7 Part 100 for the advanced plants into this particular )

8 section on the rule. That plants using the revised source 9 term will be required to show compliance with a 25 rem TEDE 1

10 for the worst two-hour period for the exclusionary boundary 11 and 25 rem TEDE for the low population zone for the 12 duration. 1 13 We also included the language of 5 rem TEDE to the I l

14 control room operators. The GDC-19 currently says 5 rem j I )-

k/

15 whole body. In the AP600 licensing proceeding, the ,

\

16 applicant put in a rule exemption request to get around 17 GDC-19 and used 5 rem TEDE. Our change here will force the 18 same thing on the folks that want to use the revised source 19 term.

20 We carried over the footnotes that were in Part 21 100 almost verbatim. We did make a change in footnote 2.

22 The original language in Part 100 had some explanatory 23 information about why 25 rem was chosen. We determined that 24 that really belonged in the statements of consideration, so 25 that was taken out for this particular section. We did,

(N, ANN RILEY & ASSOCIATES, LTD.

T, ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

- . . - . - . _ . - - . . - _ - . - - ~ . - . . . . - . - . . - , . . . . - . . . . ~ . ~ .

24

1 however, carry the caution that the 25 rem TEDE was not

' ' (i ?2 intended to be an acceptable dose for emergency conditions.

V

3- But I do have to point out a change in the slide 4 is -- we did catch this when we were editing the slides --

.5. is that we have taken this phrase out. We are no longer l2 6 referring to these things as accident dose guidelines, they 7 are going to be considered to be criteria. They are 8 guidelines in citing criteria, but they become criteria when 9 we are talking about plant' design. So just for 10 clarification purposes, we get the word guideline out, 1 i

11 Some of the conforming changes we had put in for l 12 Part 50 to make this work is we'have'added a definition into  !

.13 Part 50.2 to define what we mean by the source term. There 14 is currently no definition of Part-50 for the source term.

1 I~ \- -15 We were the'first ones actually referring to source term, so

.V.

16 'we' decided to define it. lWe' wanted to define it to get

- 17, across,the concept, as in 1465, that the source term 18 includes magnitude, composition, chemical, physical form and 19 timing.

H2 O DR. KRESS: I noticed you exclude from that 21 definition non-radioactive. Is that on purpose, is the 22' question.

23 MR. LaVIE: True, I believe it is on purpose 24 ~because what we are dealing with in these analyses is dose.

L 25 DR. POWERS: I wonder if they don't get into that,

'fg ANN RILEY & ASSOCIATES, LTD.

\s,/- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 L

(202) 842-0034 y

i . _ __ - . , _ . _ _ _ - _ -

25 1 Tom, simply because they specify the physical and chemical O

2 form of the effluent.

U~~

3- DR. KRESS: No , I had in terms of the magnitude 4 and composition. It is says, "of. radionuclides," which, to

5. -me,' excludes the --

6 DR. POWERS: But if you were to specify anything 7 about the physical.or chemical form, the non-radioactives 8 would be unavoidable.

9 DR. KRESS: Oh, you think it somehow gets folded

,10 into1that part. I didn't. I just said -- I just thought 11 that meant whether you have vapor forms or aerosol forms of 121 the radionuclides'and not the magnitude associated with the 13 amount of it. I think the quantity of non-radioactive

'14 material does not get involved in this.

( 15 MR..LaVIE: I would tend to agree with that. We 16 are not looking at things that were looked at in severe 17 . accident spaces, concrete thrown up_in the atmosphere, and 18 :so forth, from burn-through. These are beyond design basis

~

19 space t)Te of. considerations. We are really only worrying 20 about the radionuclides.

21L DR. KRESS: I had things in mind like things 22 coming from the zircaloy clad and from the control rods.

23 They come out with these.

~24 MR. LaVIE: Right.

25 DR. KRESS: You know, those could be a substantial

[N ANN RILEY & ASSOCIATES, LTD.

(_ / ~ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202)L842-0034

26 1 quant'ity.of material that one has to deal with. But, you l2 know, I like your thought-that what you are doing here is 3 putting dose criteria on it.

4 MR; LaVIE: Right.

5 DR. KRESS: But I think in the regulation itself, 6 it allows one to talk about mitigation effects.

7 MR. LaVIE: That's correct.

8 DR. KRESS: And if you had these other things

'9 present, then the mitigation processes will be different.

10 But in mind,.it probably would be conservative to leave them

.11 out per the dose.

12 MR. LaVIE: Yes.

.13 DR. KRESS: And that is why I -- I just wanted to 14 be sure you are doing this on purpose.

] 15 MR. LaVIE: Well, we are looking at the chemical 16 and physical form in the extent that we are looking very 17 heavily at aerosol mitigations.

18 DR. KRESS: -Yes.

19 MR. LaVIE: That is factored in, but we really did

20 not look at materials other than the radionuclides with 21' _ regard to this rulemaking.

22 DR. KRESS: Did you think about doing that maybe 23 for equipment' qualification purposes?

24 MR. LaVIE: Equipment qualification has been an 125 issue we-have_been wrestling with for a while. The ANN RILEY & ASSOCIATES, LTD.

')Q)

(_ Court Reporters 1025 Connecticut Avenue,_NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. . - . - . . . - . . - . _ _ - - - - , . . ~ - . ~ . . . . . . . . . . . ~ _ _ - - - . . , _

~

27 l

1. -rebaselining study indicated to us that, for most events we

(~' 2 looked at, and the two -- the three plants we looked at,

( T- /

-3 that the current TID-1484 source term resulted in higher 4 doses to the equipment outside the plant, out until a 5 certain point whe e you had a crossover, obviously, because 6 of the cesium. At that point then the 1465 became more 1

7 limiting. But at the points where we are worrying about in 1 8 design spaces, even that magnitude above was a small .)

l 9  : percentage.

10 So we feel that EQ is something that needs to be I i

11 considered, but we believe we have enough bases not to go

~12 into it, reanalyzing it all the way throughout.

13 DR. FONTANA: Before you leave that.

14 MR. LaVIE: Okay.

() 15 16' DR. FONTANA: A definition of radionuclide release from the reactor core, that means any attenuation in a 17

~

primary system or in a containment or thereon is taken on 18 top of this? And if that is the case, what guidance do they I

19 'have for that? Is that just something else?

, 20 MR. L& VIE: Okay. 1465 did'not -- did not address 21 mitigation in the containment with regard to the fractions I

'22 given in the table. The tables did, however, include

-23 ~ ret in the reactor coolant system.

24 DR. FONTANA: Okay.

25. MR. LaVIE: I.believe. Yeah, they did.

~ 7"% ANN RILEY & ASSOCIATES, LTD.

(,,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

28 1 DR. FONTANA: Yes.

i

, (g 2 MR. LaVIE: This is a standard definition. Your

! \ss/

3 question is interesting.

4 DR. KRESS: I think, didn't you -- aren't you 5 including in the Regulatory Guide some guidance on how to 6 deal with mitigation process?

l 7 MR. LaVIE: That's correct, yeah. But, see, that I i

i 8 -- at that point we are dealing with the source term to the j 1

9 environment.

10 DR. KRESS: Yes. This just is what goes into the i i

11 containment, they will get regulatory guidance -- l i

l 12- MR. LaVIE: Right.

13 DR. FONTANA: Yeah, but the reactor core, from the l I

14 pictures -- right on top of the fuel elements that's coming )

l

() 15 16 off.

primary.

Really, you are talking about out of the reactor

~17 DR. KRESS: This is what goes into containment.

18 DR. FONTANA: Into containment.

19 MR. LaVIE: Right. In the containment.

l 20 DR. FONTANA: And then there is another guidance

21 on attenuation of containment.

22 MR. LaVIE: In the containment.

l 23 DR. FONTANA: Which explains --

24 MR. LaVIE: That is one of the things that is 25 ' coming out -- excuse me. That is one of the things that is 4

f'T ANN RILEY & ASSOCIATES, LTD.

. (, )- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034

(

?

29 1 coming out very heavily out in the pilot application 2 projects is that a lot of the work that was done previously 3 in the NUREG documents that factored in the AP600 or went 4 into the severe accident work is starting to be applied in 5 these cases. And as a result of the pilots, we hope to get 6 -

• a great deal of insights that we can then reflect in the 7 Reg. Guide is what we feel comfortable with accepting in a 8 design basis space.

9 DR. FONTANA: Okay. My question concerned really 10 more the definition of what you are really doing that for.

11 MR. LaVIE: Right.

12 DR. POWERS: It seems to me that you have hit a 13 flaw in this definition.

14 MR. LaVIE: Right.

[N_/ )- 15 DR. POWERS: It has got to released from the core 16 into the containment, it needs that "into the containment."

17 DR. KRESS: It looks like you need to add those 18 words in the definition.

19 MR. LsVIE: Okay- .

20 DR. POWERS: Because, otherwise, other people are 21- going to have that question, too. The distinction is 1

22 usually made between the reactor core and the reactor 23 coolant system.

24 MR. LaVIE: Okay. We will take -- that's a very good point.

f g ANN RILEY & ASSOCIATES, LTD.

["'J.

(_, Court Reporters

, 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036

(202) 842-0034 l

l

- . - - . . . - - - _ - . . _ ~ - . - . - . - - - - . - . - . . - . - ~ . ~ . - . - - . - -

30 1 MR. ROE: Steve, I would add to -- regarding the

1. 'T 2 question concerning non-radioactive aerosol releases. The

, (d-l 3 development of 1465 recognized that there may be substantial.

! 4- non-radioactive aerosol generation and release, and it also 5 -recognized there-was a fair amount of uncertainty associated l

'6 -with that non-radioactive aerosol release. So --

7- DR. KRESS: Coming out of the core. l l

8. MR. ROE: Coming out of the core. And both early )

9 and -- you know, early as a result of control rod material 10 or whatever. And I think it is also true, as was stated, 11 1 - that the additional fission product-removal due to j I

12 additional enhanced agglomeration as a result of additional 13 material.being airborne in the form of aerosols is a 14 conservatism. Now, that is not to say that at some point,

( 15- -someone may not wish to include some minimal treatment of l 16 that. And this wouldn't necessarily -- this might not 17 necessarily preclude that, but it is not recognized in the l 18 development of the regulation, i 19 DR. KRESS: Particularly if you had a filtration '

20 system as an engineered safety feature or a clean-up, you

'21- might want to consider those as part of the loading on the 22' filters.

23- MR. ROE: Correct. And, traditionally, the 24" -non-radioactive aerosols have not been explicitly considered

~25 in these calculations. .

, -ANN RILEY & ASSOCIATES, LTD.

\. Court Reporters 1025' Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i

a .'

l l

l 31 i.

1 DR. POWERS: My recollection of 1465 was that it 2 encouraged anyone using it to justify a description of the

)

3 non -- the magnitude of the non-radioactive materials, both L

4 for the limited consideration of DBAs and for the more p 5 extensive severe accident analyses. So, presumably, the 6 Reg. Guides and standard review plans associated this --

l 7 would have some guidance on when you do your analysis in the 8 containment, look at what they did on the non-radioactive 9 source term. Because, certainly, in the AP600 that was an t

10 area of controversy between the staff and the applicant. ,

11 The difficulty is that we have never had real good 12 data on'the nonradioactive source term coming out of core is 13 because we can't do experiments melting down big cores, and 14 the nonradioactive source terms come from a variety of O

'q f 15 things. Control rods are the easy one. It's all the other 16 stuff within the core, the structures and things like that, 17- that you can't get your handle on from small scale tests, so 18 it's been difficult to get numbers that you could attest had 19 experimental validation.

20 It is easy enough to calculate numbers but you 21 don't know the reality of them and so you need to make some 22 sort of a justification on why these are reasonable, and my 23 assessment of the state-of-the-art of codes for doing things 24 like control rod release and things like that is they are 25 still in a state of flux, i

l L( ANN RILEY & ASSOCIATES, LTD.

L\ Court Reporters l- 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

l-32 1 DR. KRESS: I would go a little further. I would L /~5r 2 have said that if one had to make a justification of why we

!V 3 didn't use those. One would first ask the question of what 4 .we are trying to accomplish with the rula, and does this L 5 _ thing have a' negative effect on what we are trying to

6 accomplish.

l 7 One of the things -- the sort of !.hings you are

!- 8 trying~to accomplish are control room habitability, 9 equipment qualifications, leakage rates from containment to j 10 assure that your site characteristics and population are i 11 . compatible with your design. Then you say what do these 12 other things have to do with any of those? Well, most of 13 -the time they have positive effects in the sense that it

-14 lessens the dose, it-lessens the leakage and lessens or

() 15' allows you to have bigger leakage and things like that, with 16 the possible exception of if these chemically affect the l

! 17 chemical -- if these have a chemical effect on the form of L 18 iodine.

j 19 That is where I think I would say the 20 justification has to be focused.

- 21 DR. POWERS

I agree with you that the chemical p 22 form of iodine could be affected and certainly I am aware of l

l. 23 at least one individual that's argued passionately that it l 24 does. .I think it also strongly affects the ability to i D j employ hydroscopic effects to argue for substantially

? 25

, I ANN RILEY & ASSOCIATES, LTD.

i .g{s /^} Court Reporters

1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

33 1 enhanced gravitational sedimentation of aerosol particles.

l 4 2 DR. KRESS: Oh, I agree with that, yes. So l

\_ .

3 somewhere in the Reg Guides there should be some --

4 DR. POWERS: You have got to bring these things 5 up.

6 DR. KRESS: Yes.

7 DR. POWERS: Cesium iodide is hydroscopic as all 8' get-out but silver iodide is totally non-hydroscopic and so 9 if your non-radioactive material is converting iodine into I

10 silver iodide, you best not be invoking much hygroscopicity l 11 here.

12 DR. KRESS: That's the sort of things.

. 13 DR. POWERS: Similarly, if cesium iodide is 14 reacting with boric acid to form cesium borate and freeing iodine, you better have a better good gas term in here.

}

1

.5 16 MR. LaVIE: Okay. As a result of adding a new 17 .section to.Part 50, the licensees who choose to use the new 18 source term will no longer have to comply with the dose 19- guidelines in Part 100, so as a result we had to make a 20 bunch of conforming changes to various sections where Part 21 100 was referenced, and these such as Part 21.3, 50.2, 22- 50.49, definitions of basic components, safety related, scope of the maintenance rule monitoring program, scope of

~

23;

-24 ' license renewal applicability -- we have coordinated our l

25 work here with people doing the maintenance rule.

i P i(N . ANN RILEY & ASSOCIATES, LTD.

l'(_)'

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 L

l.

(202) 842-0034

- . - . ..- -. . -~ - . . - - - - - . - - - .. . - - . . .

34 1 There is currently a maintenance rule going l \_/

l

() 2 through. We have talked with them, so what we are doing is 3 not inconsistent with what they are doing.

4 The second category of the changes I pointed out l

l 5 earlier. We were making some changes to etiminate the need l 6 for exemptions for future reactors, construction permits, 7 combined operating license or design certifications.

8 In 50.34 (f) , the additional TMI lessons learned l 9 requirement, there were specific references to TID-1484 10 source term in these sections.

! 11 What we have done in this proposed rulemaking is 12 eliminate the reference to TID-1484, make a generic accident 13 source term and we have added the footnote that defines J

. 14 accident source term. This footnote is the same footnote l

()

l 15 that is in Part 100 for consistency.

1 16 This will make it unnecessary for exemptions for 17 the current plants.

18 This rection also has an impact on that small l 19 subset of plants that had a construction permit pending in 20 February 1982. The Staff doesn't see that this is a 21 particular problem. It's unlikely these plants will ever 22 reactivate. If they do reactivate their applications, then t

23 we would expect them to follow the new source term anyways.

! 24. One of the exemptions that the AP600 needed was an i

L 25 exemption from the control room habitability requirements in p

("'s

(,)

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l

r

I l

i i

35 '

1 GDC-19. What we have done here is we have added the 5 rem I

, /~h 2 TEDE dose criterion for that class of licensees.

Q,)

3 The regulatory analysis that went along with our 1 i

4 thought processes in this is that some of our thought 5 processes were directed for us. The SRM told us to do 1 1

6 certain things, but the no action option here obviously I 7 wasn't taken -- if we took that particular approach each 8 licensee who wanted to use the revised source term would 9 need to file an exemption. This is not good rulemaking so 10 that was rejected. I 11 We gave thought for awhile to allowing the current 12 dose guidelines and criterion. In the paper the Staff put 13 together in SECY 96-242 the Staff discussed the pros and 14 cons of this and the Commission directed the Staff to use r~

(s) 15 TEDE and the worst two-hour. With that direction, of 16 course, we needed to include it.

17 We thought about placing the existing Part 100,11 18 and GDC-19. We rejected that because we needed to maintain 19 the current licensing basis for those plants that decide to 20 stay with TID-1484.

21 We thought about the possibility of adding a new 22 section to Part 50. This is what we ultimately went with 23 because it met the needs we had and also satisfied the 24 criteria that Congress put on the Commission a few years ago 25 is to separate design criteria from siting.

(}

! (_/

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

36 1~

Let me mention briefly the Regulatory Guide, i

)

2 although'it is'not on the agenda today. It will be coming 3: to you some time in the spring. The approach we took there 4 was the-no-action one, we rejected that as well. We feel it ,

5 is necessary to have some clear regulatory guidance here, 6 because we are really changing the methods that we were 7 doing things. The number of methods for handling aerosols 8 and so forth the Reg Guides we current?.}-have in place are

'9- no' longer valid in this environment. The guidance in Reg 10 Guide 1.3 is really inconsistent with what we are doing now 11 and in current thinking.

12 So we needed to have a Reg Guide out there so we 13 didn'.t'end up with a whole lot of differences in opinions, 14 -- -

causing delays in reviews and less consistent 15' ' imp1ementation.

16 We gave thought to replacing all the current 17 Regulatory Guides, 1.3, 1,4, 1.25 -- and there's a couple 18- others. We decided against that for the same reason we 3 f 19-- didn't get rid of.Part 100.11 is we need to maintain the 20 ' current licensing basis for those plants that don't choose l 21 toluse the new source term, so what we ended up selecting 1

22' . was to issue a new Regulatory Guide which will address the 12 3 use of.tne new source term in current licensed plants and we 24 will be endorsing in that Regulatory Guide the use of 25- NUREG-1465 as an alternative source term.

i ANN RILEY & ASSOCIATES, LTD.

j. %f Q Court Reporters

1025 Connecticut Avenue, NW, Suite 1014 l~ Washington, D.C. 20036 (202) 842-0034 l

l' _ . _a_ - . . . . , . _ - . _ _ - _ _ ,~ -- .- - _ - _ . - _ a .

37 1 We intend to also include in that Reg Guide I' 2 updated analysis assumptions.

k)s 3 Now is there a value for doing what we are doing, 4 all of us spending our time talking about this? We did not 5 prepare a detailed cost analysis for this. This is a 6 voluntary rulemaking. No one is being asked to do this.

7 That is not a backfit.

8 The source term in itself may not always have a 9 cost benefit. We believe in most cases the cost benefit 10 will come from plant modifications made possible by the 11 alternative source term, either in physical changes to the 12 plant, procedures, policies and what have you.

13 There are-some cases where we have had a licensee 14 suggest to use the new source term to ameliorate an existing

(/~T 15 licensing concern.regarding control room habitability.

'd 16 Using the new source term, he will avoid making 17 modifications, but the physical idea of the source term 18 itself, no cost benefit.

-19 DR. KRESS: Did you address the question of all of 20 the plants wanting to partially use the source term, parts 21 'of the source term and not all of it?

22 MR. LaVIE: In our rulemaking plan, which we

23. presented to you earlier in the year, the Staff had taken a 24 position that we. felt that selective implementation could be 25' used for. things such as the timing only applications, but we

(~'j- ANN.RILEY & ASSOCIATES, LTD.

js_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l

l.

38 1 really felt that if you get into dose space you are going to 2 have to do all of them.

3 The Commission rejected that position. We were 4 ' directed by the Commission to make this rulemaking as

.5 flexible as possible while maintaining a consistent logical 6- clear design basis. We are not exactly sure how we are 7 going to do that yet. It seems to be certainly 8 inconsistent, lacks flexibility and clear design basis but 9 we are going to have wrestle that for the Regulatory Guide.

10 We believe there will be some possibilities for 11' selective implementation.

12 DR. KRESS: My suggestion is you have to evaluate 13 each one of those on its own merits.

14 MR. LaVIE: That's correct. i

() 15 16 criteria.

DR. KRESS: And probably using some sort of a risk 17 MR. LaVIE: Yes.

18 DR. KRESS: Augmenting by other things, of course.

19 MR. LaVIE: You mentioned the risk, which is a 20 good thing. That was the second part of the SRM that kind 21 of surprised us is that the Commission also directed us to 22 consider addressing risk aspects of this in our Regulatory 23 . Guide as well.

24 Back on the cost benefit, we don't believe 25 licensees are going to do this unless they see some benefit i

/"'s - ANN RILEY & ASSOCIATES, LTD.

(,,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034

'39 1 to doing it. The Nuclear Energy Institute did do an lj'~T

J -

2 informal _ survey about a year ago and found that there was 3 about 41 plants that had plans to do modifications based on 4 the alternative source term so there's obviously some 5 interest out there.

6 We can stand back and look at some of the things 7 we have found out already either on baselining or some of 8 the work we have done on the pilots. The Staff fully I

9 believes that there may be concomitant improvements in 10 overall safety, reduction in personnel exposure, and things 11 of this nature as a result of these changes, but very 12 difficult to quantify.

13 The bottom line is since there are so many -- a 14 wide range of possible applications for the revised source

() 15 16 term is that we really can't do a cost benefit analysis --

strictly qualitative.

.j However, we did conclude that there 17 was reason to going ahead with it.

18 DR. POWERS: When I think about 1465, the thing ,

l 19 that pops most into your mind is some nice little tables 20 that give you timing versus fractional releases.

21 I am reminded that in 1465 it says use these 22 tables or anything else you can justify, and I wondered what 23 you're thinking about that, what other people might justify.

24 MR. LaVIE: Okay. This is one of the reasons why 25 our rule is not directly endorsing NUREG-1465. We intend to i

i i

ANN RILEY & ASSOCIATES, LTD.

i y)g Ais Court Reporters 1025 Connecticut Avenue, NW, Suite 1014

. Washington, D.C. 20036 (202) 842-0034

I '. 40 i

l l

i-1 have the Regulatory Guide endorse aspects of it, something 2 like what we might do for an ANSI standard.

-( ) -

3- My current thinking, and-this is strictly current

<4 - thinking, is that we will-reproduce the tables from 1465 in 5 the Reg l Guide without a disclaimer. That is just current l~

6 . thinking though.

7 DR. KRESS: I think that there might be cases for 8 future advanced reactor design -- '

9 MR. LaVIE: That's correct.

l 10- DR. KRESS: --

that certainly could justify

~11 different timing.

12. M<. LaVIE: Right, i

L 13 DR. POWERS: This rule isn't applicable to them.

14 MR. LaVIE: That's true. This is strictly for

() 15 current plants.

l 16 DR. POWERS: You have to have a licensing before

'17 '97.

18 DR. KRESS: Oh, I see. This is restricted 19 strictly to current plants. )

20- ~MR. LaVIE: The future plants are already l

21. addressed 1 22 DR. KRESS: You might very well want to consider 23 not having that statement in there then.

24- DR. POWERS: Well, I think because it's in a Reg l 25I -Guide in fact an applicant is free to do anything he wants. )

. / ANN RILEY & ASSOCIATES, LTD.

l s- Court Reporters I 1025 Connecticut Avenue, NW, Suite 1014 j Washington,-D.C. 20036 (202) 842-0034 L

I- , _ - _ , _ _ . _ _ . . _-

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

l 41

'l MR. LaVIE: -That's correct.

[~' 2' DR. POWERS:

V) I mean it's understood, that he can

.3 do that, It doesn't have to be said explicitly.

4 It does pose some challenges if you are going to

'5 allow selective application that I could get amazingly f-6 creative.

~7- MR. LaVIE: That's correct.

8 DR. POWERS: And --

9 MR. LaVIE: And that is somewhat, kind of

-10 bothering the-Staff a little bit.

(-

l 11 .MR. EMCH: Indeed, the BWR owners group for the

~ 12 BWR sixes.has looked at that aspect of it, Dr. Powers, and 13 they have put in a submittal to propose a slightly-different 14 timing, release timing for the BWR sixes, and that's -- the

) 15 . review of that is about to start. Research is going to be f

l ;16 undertaking that review as part of this effort. 1

..17 DR. POWERS: I think changes in the timing could

- 18 -- would be what.I would call non-imaginative changes. I  ;

.19 mean, I kind of expected that just.because of the way 1465, i- 20 its genesis wasLsuch that you could probably adjust for r.

[ 21- specific plants' timing a little bit. It's the more 22 imaginative' thing like there's a constant release rate i

23 assumed in the*;e. I can dial that. You can argue that they 1

24 could.be-dialed substantially and create very strange I '25; looking. things.

L.

e

'. /~} - ANN RILEY & ASSOCIATES, LTD.

, [(.,/ - Court Reporters  ;

.1025 Connecticut Avenue', NW, Suite 1014 Washington, D.C. 20036 (202).842-0034 L

p

.n . _.. _ , _, . - . . . - __

. _ . _ . . _ . _ .. _. - > ~ _ . . _ . _ . _ . _ . . _ . _ _ _ . - _ . _ _ _ _ _ . _ _ . _ _ _ .

, 42

, 1 MR. EMCH: The potential implications keep some of i

2 - the' members of my staff up late at night, yes.

3 MR. ROE: I think Rich is correct that most of the

4. area identified in 1465 for modification of timing was the d

F 5: onset of GAT release. That was the area that.was most 6  :

highly;fccused for that particular issue, and Rich said, l'

7 'that's already been the subject of some generic

~

! 8 investigation.

9 DR.' POWERS: You could imagine it would be, and 10 .

you can imagine you have to be fairly cautious because, as l 1 11 the speaker. pointed out, accidents are far more diverse than i 12 calculational capabilities ever will be. I 4.

1 L 13 MR. LAVIE: That's correct, very much. That's one  ;

14 of the things we wrestle with very much in this, is that (f 15 while we feel that 1465 is risk-informed, we're applying i ~ 16f ' this in a design basis traditional deterministic approach I

~

17 and conservatisms and what's the proper level of 18 . conservatisms become a concern.

t.

19 DR. POWERS: I think I would be less apologetic, I i

l 20 that, in fact, what you're doing is risk-informing this 21 deterministic process that you've had in the past based on 22 improved science, quite literally. I mean, we know where J23 1465.came from; it came from a time-we didn't understand how

'24 cores degraded, we didn't understand how fission products

-25' - were retained or not retained in fuel. We'd quite frankly ANN RILEY & ASSOCIATES, LTD.

.(m O)- , Court Reporters 1025 Connecticut Avenue, NW, Suite 1014

[ Washington, D.C. 20036 f (202) 842-0034

- - ,m n , ,e ,, , .,-,-w, - - - - - - . - v- -.e.,v , e---

43 1 taken no irradiated fuel and heated it up to melting at the

)

2 time 1465 was written.

3 MR. LAVIE: Right.

4 DR. POWERS: Now we've done a lot of that.

5 MR. LAVIE: One of the things --

6 DR. KRESS: You were saying 1465; you meant TID --

7 DR. POWERS: ~ Yes, I did.

8 DR. FONTANA: They were tiny amounts at Oak Ridge, 9 the size of your little fingernail.

~10 DR. POWERS: Yes, but the irradiation of them was 11 minuscule compared to the --

12 DR. FONTANA: Right.

13 DR. POWERS: -- to the 60 megawatt days per ton 14 that we're talking about.

(

%)

). 15 MR. LAVIE: Well, one of things those of us in the 16 branch I'm in have to continue to consider is that our 17 current surrogates for risk are core damage frequency and 18' large early release fraction, and many;of the events we deal 19 with in licensing space really do not go toward core damage 20 frequency or.large early release fraction.

21 DR. KRESS: I would like to applaud that view.

22 MR. LAVIE: This is one of the reason we think we 23 need some defense in depth in here, which we're trying to 24 . maintain a good balance between risk informed and 25- traditional.

I i

1

-("T

( ,/

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ]

Washington, D.C. 20036 '

(202) 842-0034

-- _ . . . -. _ . _ - .. ~ .- - - .- . .

\

4 44 1 11' DR. POWERS: You're among kindred spirits here, at

(~~') 2 least a couple of them. DR. WALLIS: Could you go back to V.

3 that last -- yes.

4 MR. LAVIE: Okay. I hadn't gone over this one 5 yet.

6 DR. WALLIS: Oh, I see. I thought you were 7 putting it away.

8 MR. LAVIE: I was; I got off track. The 9 discussion was going so well here.

l 10 When we looked at the safety aspects of this 11 proposed rule, were we affecting the public in any way we 12 shouldn't be doing, what we thought of is that the actual 13 accident sequence -- we mentioned this a little bit earlier l l

14 -- and the release has not changed. If the event happens (m) 15 tomorrow, it's no more likely to match TID-1484 than it is j 16 to match NUREG 1465.

17 We think 1465 is more realistic, but it's still a l

18 modellof what will actually occur in a real event. So we 19 really didn't feel we had changed the risk to the public 20 because we haven't changed the sequence of the accident;

.21 just our assumptions about it.

22 As we noted, the use of the alternative source 23 Lterm in itself cannot increase CDF, large early release 24 fraction -- excuse me -- frequency or doses. Just changing 25 the source term doesn't do this.

, ANN RILEY & ASSOCIATES, LTD.

\ss ' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

- . .. . , . - . . - . . . - . . ~ . . . . ~ . . . . . . ~ - - - . - - -.-

45 1 DR. WALLIS: Why doesn't it change doses?

/~'h 2 MR. LAVIE: Okay. You have to look at the third L N~sA i

3 bullet. The alternate source term, however, could be used l

l 1

! 4 to justify plant modifications that could increase CDF large 5 early release fraction or doses.

l 6 To give you-an example, something that has been 7 done here is that several of our pilots -- one you'll hear 8 about shortly -- Jay -- one of the things licensees were l

'9 looking to do was increase the allowable leakage of main L 10 steam isolation valves, and he's using a new source term to l 11 justify this increase in leakage.

12 Now, he may end up with a dose that's equal or 1

13 close to what he currently has, but the source term in 14 itself did not change that; it was the plant modification i (A) 15 that changed it.

16 DR. WALLIS: Why doesn't it change the dose?

17 Maybe it's a question of semantics here.  :

18 MR. LAVIE: I think it is --

19 DR. WALLIS: What you mean by doses and what I 20 think --

21 MR. LAVIE: I think that goes back to my first 22 slide, is the actual dose to the public has not changed by 23 the assumptions we used in postulating it.

24 DR. WALLIS: But the calculated dose --

25 MR. LAVIE: But the calculated dose would change, i'

i ANN RILEY & ASSOCIATES, LTD.

Y h(_/

,- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014

Washington, D.C. 20036 l

(202) 842-0034 e ,, , . , - - - - - . , -

,..,+y

46 1 yes. The actual dose to the public would not.

Just because

/N 2 we. changed-one source from the other --

i I

V 3 DR. WALLIS: Well, CDF isn't a concrete thing; CDF 4 is;your estimate of what might happen.

'S MR. LAVIE: That's true.

6 DR..WALLIS: A dose is also an estimate of what

7. might happen,

-8 DR. POWERS: Let me see if I understand this fine t l

5F ' point lhere. What you're arguing'is with both the TID and '

, 10 the 1465 source term, some amount of radioactivity comes out

~11 of the reactor coolant system -- t 12 MR. LAVIE: Right.

4 13 DR. POWERS: -- and goes somewhere.

14' MR. LAVIE: Right.

! 15 Okay.

) DR. POWERSi You have a variety of 16 regulations that say -- and design criteria that say do

17. something about that and we'll judge how well you did on it 18 . based on the' doses we get to the boundary.

19 MR. LAVIE: Uh-huh.

20- DR.. POWERS: Whether I change the source term or 21 not, you're still evaluating the dose of the boundary and

22. effects of these equipment modifications on mitigating that 23: source. They have to mitigate. It might be different.

24 MR. LAVIE: Right.

25, DR. POWERS: They still have to mitigate, and O

\m,/

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. . .. _~ _ - - _- - - - - - _ . - . _ -. _ . - _

47 1 their failure to do that could change the dose to the

(s 2 public,-but the source term didn't do it.

d 3 MR. LAVIE: Right. The source term itself didn't

4 do it. It will change these postulated doses, what we use  !

5 in licensing space, but the real doses haven't changed.

i 6 Now, this leakage, the increase in leakage to MSIV l I

7 has-potential to actually --

l 8 DR. WALLIS: No regulation changes anything that l 9 really happened.

10 MR. LAVIE: That's true.

11 1 DR. KRESS: Yes. I would recommend you avoid this I E

12 type of semantic argument. l 1

-13 MR. LAVIE: Okay. )

]

14 DR. KRESS: The real argument is about -- is 1 4 \ 15 bullet three.

- (s/ '

16 MR. LAVIE: Right.

17- DR. KRESS: If you use alternate source term to 18 . modify the plant,-you could change and it doesn't 19 necessarily increase --

.20 MR. LAVIE: Right.

21 DR. KRESS: -- CDF, LERF or dose. I mean, that's 22 the bottom line. All those other things are just confusing 23 things that shouldn't be talked about.

24 MR. LAVIE: Okay. Let me tie something to this 25 now because this'is not;the conclusion we wanted to draw, is

(h ' ANN RILEY & ASSOCIATES, LTD. <

b'. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

48 1 that although the alternate source term could be used to r'~'] 2 justify modifications, is those modifications are subject to Q) 3 current regulations, and that the regulatory guide doesn't 4 address the risk consequences of these changes.

5 Also in the safety arena, the baselining study, as 6 we pointed out earlier, did not identify any significant 7 problems with potential plant modifications based on the 8 alternate source term. The re-baselining study looked at 9 several potential modifications that had been identified to 10 us and didn't uncover any real problems with doing those 11 modifications. I 12 The re-baselining study also did some MELCOR runs 13 and showed that there was still margin between the design 14 basis and severe accident space, that we were still x

(j

(

15 adequately conservative.

~16 DR. KRESS: That's an interesting definition of 17 margins. Let me ask you one question about the first 18 bullet, though.

19 MR. LAVIE: Okay.

20 DR. KRESS: One of the things you looked at by 21 reading the material was not just CDF and LERF, but you 22 looked at latent cancers, latent deaths. Was that right?

23 MR. LAVIE: Okay. That language has been changed.

24 The re-baselining study and the work we did on this 25 rulemaking did not look at latent cancers.

(~N ANN RILEY & ASSOCIATES, LTD.

( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I i

J

~ ._ ~ - .-. . - - - . - . . - . . ~ - - . .__..~ - - . - - ~ - . . _ ..

49 l 1 DR.'KRESS: Oh.

L

'~ 2 MR. LAVIE: Okay. The statements of consideration

,w

, 3 for part 100 a year or so ago included in its discussion in 4 the statements of consideration some of those details. We 5' didn't go into them here.

6 IDR. POWERS: When I look at some research that the p 7 staff has. sponsored on consequence analysis, I see them 8 employing expert judgment to look at how plumes behave once 9 they leave the plant.

10 MR. LAVIE: Right.

11 DR. POWERS: And I see experts, which I am 12 definitely not one, claiming that the plume spreads far more 13 than has been traditionally assumed in consequences code.

14- That would appear to me that the reality of the situation is

() 15 16-that latent effects are more probable than has been L calculated up until now, and the prompt effects are less 17 probable than had been accounted up until now, 18 This seems to be a product of a rather admirable 19 .and comprehensive expert elicitation. It seems to be, in 20 other words, an awfully good piece of research. Why don't 21 we take this and do a --

-22 MR. LAVIE: I'm not sure -- it's perhaps something 23 we could address if we were dealing with plants that hadn't

.24 been licensed yet, but we're looking at the licensing basis i

25 for existing plants, and at this point in the basis to ANN RILEY & ASSOCIATES, LTD.

,' A(_,j ' Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034

. -. _ _ _ _ _ _ . . __ _ m . _ ~ . . _ _ _ . . _ _ . . _ . _. . _ . . _ . _ _ _ - _ . _ _ _ . _

50 .

1 1 change the licensing' basis to try to start looking at latent 2 . fatalities would certainly be a backfit which we don't feel 3 may be valid.

4 DR. POWERS: How-could it be a backfit if it's a. j L

l 5 voluntary thing?

i l 6 MR. LAVIE: We would end up with a situation where L

l. 7- we would be regulating one subgroup of licensees to latent l

8 fatalities, another one to some other criteria.

9 DR. POWERS: Regulating one to one source code and l 10 one to another.

11 MR. ROE: Let me go over a couple of points. The 12- risk study that was done in support of re-baselining did i 13 ' consider latent health effects -- it did -- and saw no 14 significant increase in either effect, f s_,) 15 Now, it's true that the improved plume models for i

16- meander and things that were discussed before the committee

-17 some -- ~I guess it was over a year ago where we talked about 18 the' modifications to some of those release models may change

l. -19 some of those factors.

20 We're still using the assumption that for the 21 .off-site dose calculation, someone standing, you know, in 22 the path ~of a plume, there is no probabilistic treatment or 23 distribution that normally accompanies the true risk 24 analysis. These are fencepost received doses.

25 DR. POWERS: My perception of the analysis of I

i l

l

[)

\_/

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

.- . . . -~ ..- .- . . _ . - . - . . - . .

51

{ 1 plume dispersal, the uncertainty analysis on plume dispersal

{

2 was such that it probably does not affect the fencepost 3 dose, that it's the beyond-the-boundary doses that are most 4 strongly affected; and that the clear implication of what '

l l

1 5 the experts were telling you was that they had nothing 6 beyond the issue of distributions and uncertainty 7 distributions. Your mean was just wrong, in their l

l 8 estimation, by a substantial margin. I mean, it was not l

--it_was not just a small difference; it was a big l 9 10 difference, they were saying, in the spread of that plume.

11 DR. KRESS: This is due to the turbulent I 12 dissipation factors as opposed to meandering? You just get 13 a wider --

.14 DR. POWERS: That's right.

() 15 16 DR. KRESS: You just get a wider plume?

DR. POWERS: Straight-line plume, how wide is the i 17 plume at a given distance. The experts were elicited on 18 .that issue and they came back with numbers that were '

l 19 substantially different than the default values that were 20 used by the codes.

21 DR. KRESS: I think the codes used a default value 22 that was intended intentionally to make the plume as narrow 23 as it could be, didn't they?

24 DR. POWERS: No , I -- you've got me. I don't i 25. .know.

f

(

r f("N

~ \m,)

ANN RILEY & ASSOCIATES, LTD.

Court Reporters

1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034

52 1 DR. KRESS: Okay. I thought they did so that they

(T

\s,)

2 could look at really the center line, make it as high as 3 possible.

4 DR. POWERS: As I understand, and I'm subject to a 5 lot of correction because it has been some time since the i

6 committee looked at these studies, but a problem was 7 defined, analyses were done both with the European code set 8 and one of them that's used by the NRC, and I can't remember 9 which one, and then the experts were asked to look at the

+

10 same problem and there was a substantial disparity in the 11- width of the plume, okay, and the experts used whatever 12 tools they used. I mean, there was no constraint. And in

]

13 fact, what the experts I believe were asked was what is the I 14 dose or the amount of material present in the cloud.at p}

(

15 various distances, and from that, they inferred what the 16_ plume dispersion factors were.

17 DR. KRESS: My own opinion is that that particular 18 issue has basically nothing to do with this design basis 19 source term. The reason I say that is what the design basis l l

20 source term ends up doing for you is things that do not have i 21 anything to do with containment failure, very little to do 22- with containment failure, and that you actually -- if you a

d 23 had a severe accident or you fail contair. ment and did a lot 24 of plume transport, really plume transport, I -- I always 25 associate plume transport with containment failure because i

4 0 ("N ANN RILEY & ASSOCIATES, LTD.

' (~ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

l t

1 53 1 leakage out of an intact containment doesn't really behave i ] 2 like a plume.

1 (Q 3 I think we're dealing in severe accident space, i

4 and I don't think whatever the source term does for you 5 doesn't change things one way or the other. No matter what 6 source term you use, it doesn't affect this problem you're l l

7 dealing with. But I do recognize that as a problem in our 8 risk analysis, in our risk criteria, that ought to be dealt l 1

9 with, and I think that's where it might ought to belong. l 10 MR. ROE: We wouldn't see it as being unique or 11 specific to this revised source term. l 12 MR. La VIE: I think part of this may he.1p us in 13 the design basis space, I'll take a stab at this, is that 14 our criterion is based on the maximum individual, not a f}

v 15 population dose. It's my belief -- I can't show you 16 numbers -- it's my belief that the people immediately 17 underneath the plume centerline would be the people -- op j 18 analysts would project a much higher dose to those folks 19 than if we spread it out across the population.

20 DR. KRESS: Yes.

21 MR. La VIE: We're dealing with an individual dose 22 here, not a population dose.

23 DR. KRESS: Yes, and that means your leakage, 24 would that be less?

25 MR. La VIE: Right.

l l

(~'s ANN RILEY & ASSOCIATES, LTD.

(^~'J - Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

54 1 At the present time our only proved alternative

/~h ' 2- will be NUREG-1465; the tables thereof will be extracted.

V 3 We expect that any future proposed alternative source term 4 -will be subjected to the same level of scrutiny as was 5 NUREG-1465. We'll all probably be long retired before 6 -another source term shows up. And it's the expectation that 7_ any future alternative source term that does get approved 8- will have the same magnitude of risk as what's currently 9 implied in 1465.

10 The rulemaking package also contains an 11 environmental assessment. Real quickly, the same argument.

12 We ended up with the conclusion, however, that there is --

13 public health and safety is not decreased by the proposed 14 rulemaking because the dose limits are changed to match.

() 15 The TEDE criteria were previously implemented, and it was 16 concluded at that time that those did not involve a 17 -significant impact, and therefore no environmental impact 18 statement has been prepared for this rulemaking.

19 DR. WALLIS: Could we look at number 3 here? I 20 mean, you're calculating the same limiting doses as before.

21 MR. La VIE: That's true.

22 DR. WALLIS: So you say it's all right.

23 MR. La VIE: Right.

24 DR. WALLIS: But in fact you're going to allow 25 plants to make modifications which in reality will reduce ANN RILEY & ASSOCIATES, LTD.

kT s

l Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

l-55 l 1 some conservatism and therefore in reality will increase the

(~'T 2 risk to public health and safety.

L) 3 MR. La VIE: But they'll still be under --

4 currently we cannot authorize licensees to take the dose all 5 the way up to 300 rem thyroid.

1 6 DR. WALLIS: No, but if they would use your new 7 source term --  !

.8 MR. La VIE: Right.

l 1

9 DR. WALLIS: With the old design, and then used it 10 with their modification, they would have shown an increase 11 in dose by the modification, most likely. 1 l

12 DR. POWERS: But, Graham, I think what you're

13. seeing is a continuum --

• 14 DR. WALLIS: Is what?

e~x i i 15 DR. POWERS: You're seeing a continuous function

%)

16 here.

17 DR. WALLIS: No, I'm seeing something which I 18 don't think is a true statement.

19 DR. POWERS: Well, it's because of the way you're 20 interpreting it. You see --

21 DR. WALLIS: Well, I'm interpreting it the way 22 that any reasonable person --

23 DR. POWERS: Maybe you'd let me finish this 24 statement I want to make here, and then you'll understand 25 more clearly what I'm driving at.

[~') ANN RILEY & ASSOCIATES, LTD.

(._ /

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

56 1 If the dose to the public -- there is some dose to

./ 2 the public in the' event of an accident -- is seen as a 3 continuous function, then any change that allows that to go 4 up causes an increase by definition. But if in fact

'S adequate protection is achieved by not exceeding the 6 threshold and you don't exceed that threshold by a margin 7- that's two percent changed, then there is no increase in the 1

8 public, and consequently no reduction in the protection of  !

9- the public health and safety.

10 DR. WALLIS: Well, there's no change in meeting 11 the requirements of regulation in the way that public health 12 and safety is defined, but in reality there has been --

13 DR. POWERS: The protection --

14 DR. WALLIS: Increase in risk.

' /~%

15 DR. POWEPS:

() The protection is still in place. i 16 DR. WALLIS: But in reality you have to admit 17 there has been an increase in consequences of an accident.

18 DR. KRESS: I think Dana is correct, what he says.

19 There is this threshold. If one looks at another health 20 effect, and that's the linear dose relationship, by making 21 these changes it's not necessarily clear that they do 22 increase the leakage rate of fission products even from 23 unfailed containments. It's likely that some of the changes 24 would do that. Some of the changes might go the other way.

25 I mean, it doesn't preclude making changes that help you.

• \

("' ANN RILEY & ASSOCIATES, LTD.

Court Reporters l

1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

57 1 But let's presume that some changes can be made

/~N 2 like allowing a bigger increase -- an increase in leakage 3 from their containment. That's one of the possible changes 4 that would happen. In the case of any incidents you have 5 over the whole range of spectrum, that means you're 6 releasing more from this containment, and by the linear dose 7 consequence thing, that means you're increasing the number l

8 of cancers. Now that's very clear. So what you're saying 9 is correct that you do run the risk of increasing that, and 10 the question is, is that a correct statement, because are 11 you protecting the health and safety of the public from all i 12 releases.

13 It's clear to me that if you had an accident that 14 failed containment that this has nothing to do with this.

[

%s 15 and that's a clear statement there. It doesn't have 16 anything to do with it. But there are other little things 17 that go on where you might release from an unfailed 18 containment that that's not a true statement. And I think 19 that's where the difference is.

20 MR. La VIE: I think what we're trying to point 21 out here or what I'm trying to point out here, maybe I 22 simplified it too much in trying to get it into a bullet, is 23 that our current regulatory basis prior to this rule 24 establishes a system of protection of the public. Nothing 25 we're doing in this rulemaking is affecting that level of i

' (p

_ ,/

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. . . . ~ . - - - . - - . - . - _ . ~ . . . . . ~ _ - ~ . . . . . - . _ . ~ - - ~ -

- _ _ . ~ . - . _ . - - .

58 1 ' protection of the public. We have not changed the ultimate t

~.i 2' dose criteria.

/)

\- J 3 DR . WALLIS: That's right, you've not changed the-

-4, dose criterion, but in reality if there were an accident and

-5 there were a more leaky containment, as my colleague says, 6 then there would in reality be more dose to the public. So

~

-7 -you're'using words in a way --

l- 8 MR. La. VIE: True,.there would be more --

3 9 DR. WALLIS: Not the way one would normally use. '

10 them.

11 .DR. KRESS: The question I would have is is this

'12- more dose still acceptable in terms of --

'13- MR. La. VIE: That's what.I think -- the I l

l

'14 semantics'--

( ) 15 DR. KRESS: And it's a question of you're still 16 protecting the health.and safety of the public.

l 17 MR. La VIE: Right. I

" i

'18 DR. KRESS: You've just~1ost-some of the margins l 119 probably.

.20 MR. La VIE: Exactly, you've lost some of the l

21 margin.

122 MR. EMCH: This is Rich Emch again. Your 23 statement, Dr. Wallis, about the exact meaning of the 24- phrase, we understand your point. I think what Steve is 25 ' wrestling with is he's trying to reach a conclusionary l

-ANN'RILEY & ASSOCIATES, LTD.

j (,'

^

.^)- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 L (202) 842-0034 l:

l' L

. . . . _ . . , , . . - - , - . _ _ _ _ . _ . - _ . - _ . _ . _ - - _ . . . ~ . _ _ . . _ . _ . _ _ . _ _ . . . . . _ . . , _ . _ _ .

59  ;

1 statement that is where the meaning of this issue of

,Y$ 2. . protection of public health and safety is much more aligned v ~

3 with:the kind of statements that Dr. Powers and Kress have 4_ been making. So we' understand your point, though.

5. DR. WALLIS: Well', I think you ought to be careful i

6 'that you don't lose credibility by making a statement which

, 7 just doesn't seem to be right. Perhaps if you put it a

'8 different way. -

9: MR. EMCH: Right. We'll take a look at that.

10 MR. La VIE: Thank you, 11 DR. FONTANA: The limits are still observed; 25 12- TEDE is still there.

13 MR. La VIE: Yes. The Part 100, the rulemaking of

14. Part 100, the staff demonstrated that the 25-rem TEDE was

( .

15 not'a reduction in the level of protection of the public, 16 that it was -- on a risk basis'it was comparable.

.17 - I don't'have a slide for this, but I handed the

~

-18' . sheet out, is that in the letter where we forwarded this l 19 package _to'the Committee earlier last month is that'we

. 20- pointed out that the office concurrence process was going on i.

l 21 in parallel with the ACRS review. -

22 You know, we told you we would identify any 23' changes that occurred as a result of office concurrence.

L 24 Thel majority of the changes we received were editorial in 25 . nature and will not change the technical content. We point D[ ANN RILEY & ASSOCIATES, LTD.

kJ) '

Court Reporters l 1025' Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i

l'

( . - . - . .

._ _ _. _ - .- _ . -, _ _ _ _.____ . . . _ _ . . . _ _ . ..m . .

L ,

60 l i

l 1- out in the statement's consideration based on discussions we

! /~'t 2 had with OGCs and others that the phrase " accident dose l

!O 3 guideline" has been replaced with " accident dose criterion." -

4 The chief information officer asked for some 5' changes in the Paperwork Reduction Act statement. Under the i 6 proposed rule language, as we pointed out earlier when I 7 showed you the slide, is we have deleted the explanatory 8 information in the beginning of footnote 2 that we felt '

l 9 belonged in the statement's consideration and it is in the 11 0 statement's consideration. We just left the clarification 11 statement that this limit is not an acceptable emergency l

12 dose.

l 13 Research suggests that revising the definition of 14 " source term" could change " mix" to " composition." We'll be  :

. () 15 definitely considering this suggestion that was made by the I 16 Committee about changing that to be "from containment" as 17 well, f

18 OGC asked us to get rid of the -- make changes to l

19 the conclusions section to remove the appearance that there  !

l l 20 was a draft determination under the provisions of NEPA.  :

1 21 They pointed out the draft determination is only supposed to 22 be in the statement's consideration. So we'll be making 23 those changes as the package does on through to the EDO.

l 24 Arn there any additional questions? l

(

l25 [:No response.]

i 1

!- l ANN RILEY & ASSOCIATES, LTD.

()N~

as_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 I

(202) 842-0034 i l

i-

_ _ - .m

- . ~ - - . . - ~ . - - - - . - . - ~ . . ~ . - . . . . ~ . ~ . - ~ . . . . . . . - . - . . _ . - . - . ~ -

61 1 Thank you.

2 DR. KRESS: Thank you.

We're now going to move on l l 3 to-the Perry.

i-4 Perry has made a specific application for a change j 5 .using the new source term, and that's what we're going to -- i 6 we're just going to hear the status of that application now. i 7 It's an. ongoing thing they're still discussing and hearing, 8 and this is mostly just a status report on it.

9 DR. POWERS: Mr. Chairman, I think I have to point i 10' to the Committee I probably have some sort of conflict with i

11 respect to this particular presentation. I note that among 12 the speaker's slides he's going to cite work in which I was

13.  :'he senior author and work that was done by colleagues at  ;

14 .Sandia National Laboratory. I think that's not going to 15 inhibit me from making comments to this, but Members may 16' want to discount those comments based on the fact that I l- 17 obviously have some conflict in this, in connection with i

18 this.

-19 DR '. KRESS: So noted, and I might add it was a 20 very good piece of work.

{-

21 CHAIRMAN SEALE: We'll give him the usual 22 treatment.

L-p 23. DR. KRESS: Well worthy of your name.

V

[ -24 DR. POWERS: I'll try to get that envelope with i-25 the money.in it to you as quick as I can find a good

! ANN.RILEY & ASSOCIATES, LTD. I

,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C 20036  :

(202) 842-0034 L

l 62 envelope, a big enough envelope, j'"\ 2- MR. LEE: Okay. Good afternoon. My name is Jay (s,) 1 3 Lee, and I'm with Emergency Preparedness on the Rattiation l

4 Protection Branch of NRR. Today's presentation to your 5' Committee is the application of a revised source term at

! 6 Perry, how we are using this revised source term for review.

1 7 As Rich briefly stated during introductory 8 remarks, we are not the only branch who were involved in l 9 this review, but there are many other technical branches -- 1 10 Containment Systems Branch, Electrical Engineering Branch, 11 and particularly Office of Research is involved. In 12' addition to that we have three national laboratories 13 assisting us for the Perry review. Pacific Northwest l 14 Laboratory is helping us review meteorological data at the 15 Perry site, and Sandia is helping us to run a RES/ RAD code.

(J) r~

16 This is a new computer code Sandia developed for NRC, and

! 17 .they are helping us to make use of that code. And the u .18 third, Oak Ridge, is helping us review iodine chemistry. So I 19 three national labs are also involved in the review of this, l

20 Rich Emch went over this briefly. About two years i

21 ago or so we provided a Commission with the Commission SECY l

22 paper 96-242 informing the Commission how we intend to use i

I 23 this revised accident source term for operating reactors.

24 Response to that SECY, we received SRM, go ahead

{ 25 and undertake rebaselining and we did perform, rather the l

l p)

(_

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 o

I l

I i

L 63 1 Office of Research did the rebaselining and we completed L

2 .that effort in last early April, this year, Research briefed L(/~Nf 3 you on the result of rebaselining.

4 Steve just went over rulemaking effort after we 5 complete this rebaselining. I 6 Then we started reviewing the pilot plant

( '7 -application. The first one is of course Perry Station, and i 8 in that particular SRM, the Commission directed us to go 9 ahead and use worst two-hour methodology and to use new dose l

! 10 criteria using the 25 rem TEDE and the 5 rem TEDE in the 11 control room. I 12 Perry license amendment request really consists of .

l

!' 13 three major items -- number one, they requested us to remove L 14 main steam isolation -- leakage control system.

( 15 DR. KRESS: Could you educate us a little on what 16, that system, how it works and what it does?-

17 MR. LEE: Yes. This system really collects any I 18 leakage passed --

l 19 DR. KRESS: Passed through the valve.

i l

L 20 MR. LEE: Through the MSIV. l l

.. L21 DR. KRESS: It measures it as a function of time?  !

l 22 MR. LEE: Yes. I don't believe they are measuring L 23 it, but they collect this leakage and then send it through I

i 24 the secondary containment for hold'and then filter prior to 25 releasing to the environment, rather than directly ANN RILEY & ASSOCIATES, LTD. I

, A_, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036

'(202) 842-0034 i l

_ l

64 1 discharging --

[)

%j' 2 DR. KRESS: Are you sure they don't release any 3 radioactivity that might be in that -- 1 4 MR. LEE: Yes --

5 DR. KRESS: -- steam?

6 MR. LEE: In the steam leakage.

7 DR. KRESS: Steam leakage -- it condenses and i

8 then --

9 MR. LEE: That's right, so this is nothing but the 10 leakage collection system. Then they don't process any 11 fission product --

12 DR. KRESS: My question then is what does that 13 have to do with the source term?

14 MR. LEE: Yes. I am going into that.

f~%

L m a) 15 The uniqueness of this particular request is they 16 requested this by taking a credit or using the revised 17 source term. Now this type of request is not new to us.

18 Other BWRs, existing, operating BWRs, requested a similar 19 request and they --

20 DR. KRESS: I see. They take credit for that in 21 reducing the dose to the outside in a design basis accident.

22 MR. LEE: Yes.

23 DR. KRESS: I see.

24 MR. LEE: Yes, and okay, so that is one of the 25 first requests and then the second one is they would like to

/'N ANN RILEY & ASSOCIATES, LTD.

(-) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l

l

65

-1 increase the maximum allowable MSIV leakage rate.

2 DR. KRESS: Does that show up in the tech specs?

3 MR. LEE: Yes,'and currently they have 35 cubic 4- feet per hour leakage per line or 100 cubic feet per hour

-5 forftotal of four main steam lines, but they'd like to go 6 from the 35 to the 100 CFH so it is a possibly threefold 7 increase and they'd like to increase 100 CFH leakage limit 8 to the total of 250 CFH.

9 The third item that they would like to -- they are 10 requesting is use of the sodium pentaborate in a SLC system, 11 the Standby Liquid Control System for containment pool water 12 pH control. Now this is the first time, first BWR, as I 13 know, tried to use this sodium pentaborate for buffering 14' capacity to control the pH after accident.

O u,J .

s. 15 DR. KRESS: Sodium pentaborate is already in the 16 SLC system?

.17' MR. LEE: 'le s .

18 DR. KRESS: And they just wanted to be able to add 19 it into the --

20 MR. LEE: Right, right. They have I believe more 21 .than about 5000 pounds of sodium pentaborate in 22 approximately 4000 gallons of solution in the system.

23 DR. KRESS: Does Perry have a spray?

24 MR. LEE: Yes, they have a containment spray 25- system.

D' ANN RILEY & ASSOCIATES, LTD.

\_s/'

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington,.D.C. 20036 (202) 842-0034

L 1 I

66 )

1 DR. KRESS: That is what this is for?  !

i

(N 2 MR. LEE: No , .the primary purpose of this is for

'b 3 the reactivity control.

4 'DR. UHRIG: It's for emergency, isn't it?

5 MR. LEE: Yes. This is for the backup capacity, f

6 capability for the reactivity control, putting in the boron

-7. to it so they-like to use the same system as a buffer for I 1

8 the pH control purpose.

f 9 DR. FONTANA: Would they use all of it?

10 MR. LEE: Yes. I think -- correct me if I am 11 wrong -- Perry.-- that you just dump it right into it'from 12 the main control room operator's action, and then just dump 13 it right into the containment pool water.

14 DR. WALLIS: What is the real consequence of this I q

,.I

( %J e 15 increased liquid -- it means they don't have to replace some L 16 part of the valve so frequently? Is that what it is?

17 MR. LEE: No. They are-not doing any system j 18 changing, if that is what you mean.

19 MR. BARTON: What they don't want to do is every 20 outage when they do their LLRT after they shut down if they 21 fail the leakage, they don't want to spend a couple hundred 22 thousand dollars rebuilding the valves.

23 DR. WALLIS: Right. That is essentially it.

l 24 So it is real measured leakage rate. It is not 25 some kind of assumption --

l

[' 7004 RILEY & ASSOCIATES, LTD.

( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036

! (202)'842-0034 l-l l

. - . . .. - - - . - - - . _ - -_- -_ -_. ~ - - - .

67 1 MR. BARTON: No, it's a real measure of leakage 2 rate.

3 MR. LEE: And these numbers are in tech spec, as I 4 say, and they do that I believe every refueling --

5 MR. BARTON: Outage, yes.

6- MR. LEE: And they are actually measuring it, i

7 Okay. The way we approach this particular license 8 amendment request, we re-analyzed EAB -- this is Exclusion 9 -Air Boundary -- and the low population zone -- the doses, 10 and the control room operator dose as a result of a 11 postulated LOCA.

12 Perry proposed the same and we do the same way 13 that we are analyzing, re-analyzing the LOCA. I 14 The source term -- of course, we are using, also i

() 15_

16 Perry using the NUREG-1465 for the fission product release magnitude or fraction for release timing and the chemical 17 species.

1 18 Now here I had the no exceptions. What I meant 19 was, and maybe that is not right phrase, but what I meant 1

20 was just they are using the gag-release portion and the 21 .inversel release portion without any exceptions. Of course, 22- they are not using the exvessel or late invessel releases, l 23 but other than that, for those two phases, they are taking the release magnitude and release timing and chemical

'24 25 species the way we specified or the way we provided in the ANN RILEY & ASSOCIATES, LTD.

--(,,(~~T) Court Reporters 1025' Connecticut Avenue, NW, Suite 1014

, Washington, D.C. 20036 i

(202) 842-0034 l-I-

l' l

t i

68

- -l' 'NUREG-1465.

"'g 2 DR. WALLIS: That is the new source term?

(d 3 MR. LEE: Yt s , that is the revised source term 4 that we are discussing here today.

! 5 Those criteria of course, as Steve described i

6 earlier, we are going for 25 rem TEDE at the EAB, any 7 two-hour duration. This is the way the Commission directed l

l- 8 us to do.

9 DR. WALLIS: These are the limits?

10 MR. LEE: Yes.

11 DR. WALLIS: What were the actual calculated doses 12 for the two sources? We don't know yet?

13 MR. LEE: We don't know. We don't have any 14 numbers to share with you as yet.

() 15 16 DR. WALLIS:

will go way down with the new source?

Isn't 17 anticipated that the-dose l

17 MR. LEE: We think it will go down but how much-18 and how significantly we don't know that yet. We are still 19 in the middle of the review and middle of the calculations.

20 DR. WALLIS: Isn't it a big change?

21 MR. LEE: Source _ term, yes, and in terms of a 22 particular chemical species and release timing it is big I

L 23 changes but --

24' DR. KRESS: It's big but it's not an order of

-25 magnitude.

{ f% ANN RILEY & ASSOCIATES, LTD.

'(_)-.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 y

. . - .-~ -._ ~- _ - . - . . . -. - . . . - . .

69 1 DR. POWERS: Well, I mean in truth the amount of f~') -2 iodine that gets into containment goes up.

,  %/^

3 The other radionuclides that go into containment 4- go up a '.ot.

5 DR. KRESS: Yes. It is a bigger source term.

l 6 It's just the timing that affects --

! 7 DR. POWERS: The timing has changed. I mean there 8 is no reason a priori, and I think the previous speaker made 9 this point over and over again to us, there is no reason a 10 priori to think that the dose criteria, the margin that you 11 have between the criteria and what actually gets released in 12 the calculation would change at all.

13 DR. KRESS: Well, there is a reason. They did a 14 baselining study --

.(p). 15 DR. POWERS: Yes, after the fact.

j, 16 DR. KRESS: -- told them to expect it.

17 DR. POWERS: That's a priori. Absent that and the l

18 fact that we know something about aerosol physics and things 19 like that, but the source term itself -- actually it's a 20 bigger source term.

21 I mean it's a little difference on a log plot but 22 it is a big difference on a linear plot.

I 23 DR. WALLIS: Isn't it a more removable form?

24 MR. LEE: Right. Chemical forms --

25- 'DR. WALLIS: Yes, that's the' big change is the ANN RILEY & ASSOCIATES, LTD.

O)

\m- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 E Washington, D.C. 20036 l (202) 842-0034 l.

70 1 chemical form, let's assume.

[~} 2 DR. POWERS: Well, I mean that is arguable. There b/

3 are assumptions built into the way TID has been implemented 4 in the Reg Guides that make it essentially -- make iodine 1

5 essentially an noble gas, but the reality of iodine is that l 6 it is a highly reactive material and all you are doing is 7 recognizing more of that reactivity in two forms in 1465 --

8 it -reacts to make particles and those particles fal] under l 1

9 gravity.

10 DR. KRESS: Not only that, these calculations 11 allow you to use sprays in suppression pools under certain i

12 circumstances which are pretty effective for iodine also.

13 DR. POWERS: Oh, yes, absolutely.

14 MR. LEE: Right. The revised source term, you fn

! ) 15 know, has a chemical form. Most fission products are in the v

16 form of aerosol rather than elemental iodine.

17 DR. WALLIS: So the expectation is that you 18 actually get significantly lower doses calculated.

19 MR. LEE: That we don't know yet. According to 20 the baselining we did, it seems to indicate that that is the 21 case.

22 DR. WALLIS: I am new to this. My impression is 23 that the anticipation is these doses will be significantly 24 lower.

25 DR. KRESS: Well, you can almost be sure in this i

/~ ANN RILEY & ASSOCIATES, LTD.

(,}/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

l 71 l

[ 1 case they will be. Otherwise they wouldn't be here.  ;

i

[ 2 DR. WALLIS: And therefore we should anticipate l l\

3 requests for much more significant changes than the leakage 4 of a main steam isolation valve. j 1

5 MR. LEE: Charlie, you have a comment?

6 MR. ROE: Yes. Although the point has been 7 raised, but it is clear from the rebaselining study that the 8 use of the revised source term with a more realistic timing, i 9 more realistic physical and chemical forms may result in a 10 substantial reduction in the dose depending on the specific L  !

11' calculation. l 12 What is at play here is the accompanying change to 13 the plant in the MSIV leakage. )

14 I would point out that the MSIV leakage  !

'{/s) u.-

15 contribution to the offsite dose for a MARK III can be 16 substantial. It can be a substantial part of the overall 17 dose release, so changes to that system -- it's not a 18' trivial change to the plant. It can be a major portion of 19 the dose calculation may actually come from the MSIV.

20 - DR. KRESS: It's the main bypass. j 21 MR. ROE: Yes, it is.

22 MR. LEE: That's right. It contributes in the 23 case of Perry perhaps as high as 80 percent of doses coming 24 from this particular pathway, MSIV leakage.

25 For-the dose calculation, we will be using what we l'

i (7 ANN RILEY & ASSOCIATES, LTD.

' (_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

- - . -. . . . - - . ~ . - - . . . . . - . . . - . .. ~ - - ~ . . .

72 l' call-the RADTRAD code. This is a relatively new computer (f"} . 2 code developed by Sandia for us and for the rebaselining we Q

3 did use RADTRAD code.

4' The Perry is the first plant that we will be using l- .5 this particular new code for the dose calculation.

-6 DR. WALLIS: This is a simplified model.

7- MR. LEE: Yes, it is --

'8 DR. WALLIS: Usually simplified models are less L '9 accurate.

10 MR.. LEE: Well, title is "The Simplified" but I am 11 not sure that --

} 12 [ Laughter.]

l l

13 DR. KRESS: Has it been benchmarked to CRAC II and l

l 14 MAX?

15 MR. LEE: We benchmarked against the Habit code k.J L 16 for other dose calculational model, but this title I think i-17 'is a Sandia name, and I am not sure we can agree to that 18 simplified anymore. We still have some enhancement work to 19 do on this code.

20' MR. EMCH: It is certainly simplified compared to 21 real life.

22 MR. LEE: Yeah.

23 DR. POWERS: Compared to the Habit code, I would L 24 say that simplified is not quite the term I would have 25 elected, but a certain person started a trend maybe he will 5

j ANN RILEY & ASSOCIATES, LTD.

D)

(_ ' Court Reporters 1025 Connecticut Avenue, Mi, Suite 1014 1: Washington, D.C. 20036 L

(202) 842-0034 L

. . - . . .- . - . ~ , . . . . _

l 73 l 1 learn to--regret in the future.

i L IN 2 [ Laughter.]

l ' b .' MR. LEE:

3 Okay. So that where the area of review, 4 how we are reviewing this Perry license amendment request.

5 First, we are giving, of course, the aerosol deposition 6 credit in the dry well.

7 DR. WALLIS: Everything is simplified, it appears.

l 8 [ Laughter. ]  !

9 MR. LEE: That is, again, all the Sandia l j' 10 = nomenclature.

f 11 DR. POWERS: It is beginning to regret all the j 12 time.

l' 13 DR. KRESS: But in the case of that second l

14 sub-bullet there, that is not necessarily less accurate.

'K js j- 15 [ Laughter. ]

16 MR. LEE: Yes. Right. 'That is the second bullet, t

l 17 that only one has -- I mean doesn't have the word 18 simplified, j 19 DR. WALLIS: I am very pleased to know that a I j H2 O government lab is capable of simplifying things. l 21 MR. LEE: No, I think indeed, though, for this 22 particular case for the NUREGs, it is simplified for use.

23 DR. KRESS: Believe me, that one is well titled, 24 the simplified.

L 25 MR. LEE: Which one?

'q ANN RILEY & ASSOCIATES, LTD.

,)

\, Court' Reporters 1025 Connecticut Avenue, NW, Suite 1014

, Washington, D.C, 20036 (202) 842-0034

l.

(7 74 l- 1 DR. KRESS: The NUREG CR-6189. I L

.t 2 MR. LEE: Yes.

'3 DR. KRESS: That is'a simplified one.

4- 'MR. LEE: Right.

5 DR- WALLIS:

. How simple is it?

! 6 DR. KRESS: I don't even need my computer, I can l

L 7 sit here'and do the calculations.

8. DR. POWERS: The senior author will assure that 9 you get a copy of it, especially if he can find the bulk I

10. mailing envelope. It's a very healthy document.

11 DR. KRESS: .Actually, that document and the one on 12- the bottom, I highly recommend you read it. It's a very 13 good piece of work.

14 MR. LEE: Now, for most of --

if g l( j 15 DR. KRESS: I can say that, I had nothing to do 11 6 with either one of them.

17 DR. POWERS: He really wants to borrow some money 18 from you, I can tell, 19 CHAIRMAN SEALE: Gentlemen, let's get our 20 transactions out'of this.

21 MR. LEE: Most of you, the member, really, just 22 this bullet item perhaps doesn't help very much. All I am 23 saying here is, yes, we are giving some sort of deposition 24- credit. This is a fission product aerosol deposition in the 25 dry well. So I say here, the NUREG number and title, that

!O

, \s /

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut. Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

75 i

' l l 1 doesn't really much of saying anything.

2 But this model is really assuming the well mixed )

~3 model in. dry well and it use a Monte Carlo approach. Each l- 4  !

parameter has a range and'it has a particle size, material l

l L

i 5 . density, and'a shape factors, and it has particular ranges

~6 to assign-to each-important parameter.- Then what'you do is

-7 all' Monte Carlo analysas, I think you have like 3,000 or l'

L '8 10,000 random: calculations and it comes up shows what I 9 percent, how much confidence you have on that particular l

I number you just'have developed, like such as 90 percent R10 11 confidence level or 50 percent confidence level over best 12 estimate, or 10 percent. So this, indeed,.a simplified l

)

13 model for this particular purpose.

l 14 Now, we heard also from the committee that you may )

1

[)

3% /-

15' have some question'on electrostatic;of the particle, how we

{

I' 11 6'- dealt with this'particular phenomena, or we just neglected _

17' this model, it doesn't' incorporate any. electrostatic 18

~

l charges. I guess it is also'my. view that no individual

=

19 particle-will'have a particular charge for so long, it will 1

20 ultimately change to positive to negative, or' negative to  ;

l l - 21' positive, and it will fluctuate, and'the paths will attract i 22 each or. repulse, depending on-the charge of that particular

. 23 particle at the particular time.

12 4 So it is complex issue and we are aware of this 252 phenomena, but --

[.

ANN RILEY & ASSOCIATES, LTD.

fg Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 I. . Washington, D.C.'20036 (202)- 842-0034 g F '. -ry 4 M .si-,, awrr .q.g.,- g q , y -- , -+~p.

_1 76 1 DR. KRESS: Are you doing anything in the way of

(% 2 trying to scope that issue, to see whether it is

.VL 3' conservative, non-conservative or neutral?

41 MR. LEE: Again, my view is we are conservative in 5 neglecting.this particular phenomena and we are still 6 considering, or course.

7 DR. KRESS: Do you think i' ,c.uld cause them to 8 agglomerate faster?

9 MR. LEE: Right.

10 DR. KRESS: Do you have a basis for that belief?

11 MR. LEE: No , I don't. But --

12 DR. POWERS: In truth, I mean -- in truth, in 13 6189, the author makes the argument that it is conservative 14 'to neglect the charging based on the irreversibility of the ex l( 15 agglomeration process, and I believe he says, okay, this is

.s-).

16 an intuitional argument. He can't prove it, but it is an 17 intuitional argument.

l 18 DR. KRESS: Are you qualified to quote this 19 author?

20 IMt. POWERS: Marginally. I suffer CRS as bad as I l

21 you do. But, I mean, there -- in truth, there are some i 22 experiments going on in France that may lead us to at least 23 some insight on this particular issue.

24 DR. KRESS: The PHEBUS?

25 DR. POWERS: Yes, that's right. That is the only

("5p ANN RILEY & ASSOCIATES, LTD.

-( ,j. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

.__________._..___._.____.__..____._.__._m _ . _

77 1 . experimental. program-I know that is capable of j 2 experimentally addressing the issue, because the '

D} - -

'3 radioactivity comes not just from the radioactivity of the 4 particle, but'also the radiation field within, and that is a 5 .. tough experimental environment to create. They do. And if 6 we see either enhanced deposition or enhanced survival of l

[ 7' particles in the atmosphere, that would lead us to think g 8 that maybe there is an electrostatic effect, and I will have 9 to tell you right'now that some of the early results coming j 10 out of the PHEBUS suggest that retention in the piping >

l

11. system may be-less than what the codes calculate. What goes L

12 in the containment is proving challenging for the codes to 13 address. ,

14- So, .I.mean there is work going on, but, in truth, (e ) 15 nobody knows how to deal with it because, as the speaker  ;

l 16 correctly pointed out, it is not a static charging, it is a ,

I l 17- dynamic charging and any given particle has a particular

'18 . charge only.for some period of lifetime and that charge can 19 .go from positive to negative. I mean it is wildly varying. I

{

L 20 MR. LEE: It.is fluctuate constantly.

21 DR. POWERS: Fluctuates tremendously. It may be 22 that fluctuation -- the treatment of fluctuations is the way l l l 23. to go after this problem. But, right now, it is just kind l:

24 of intractable.

l=

l 25 MR. LEE: So we adopted a model in this particular  ;

r .

• i

[ (()

,j ANN RILEY & ASSOCIATES, LTD.

Court Reporters l

j 1025 Connecticut Avenue, NW, Suite 1014

-Washington, D.C. 20036 g (202) 842-0034 l

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

E 78 1 NUREG right into the RADTRAD Code and it is all in the code,

! 1'S g 2 really, it calculate for us. Okay.

'V 3- DR. KRESS: Is that RAD -- is that supposed to 4 have a "T" in it? Is RADRAD supposed to have a "T" in it?

5 DR. POWERS: Brockmann's model.

6- MR. LEE: Yes. Sorry.

l 7' DR. KRESS: The question I have about that bullet i

! 8 is the 1465 source term kind of has implied in it already 9- some effects of the RCS in attenuating the release, and are 10 we not counting twice when we say we are going to allow 11 aerosol deposition in the main steam lines and calculate it 12 again on top of the source term? Is that a double counting 13 there?

14- MR. LEE: No , NUREG-1465 source term is the amount

() 15-16 of a fission. product released from reactor coolant system into the containment.

17 DR. KRESS: Yeah, but it is not --

18 MR. LEE: Into the dry well.

19 .DR. KRESS: Yeah, but it is not the amount 20 released into the~-- from the core. In the case of release 21 through these valves, what -- you.would get it from the 22 core, then it would go through this main steam line and 23 either bypass the suppression pool through this valve or go

.24 to the suppression pool.

.25 Now, doesn't -- don't you normally treat the stuff l

l

'f] ANN RILEY & ASSOCIATES, LTD.

\ ,/: Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

1 79 1 going through the suppression pool as 1465? That is what i 2 goes the suppression pool.

3 MR. LEE: Well, in the Perry case,.the way we l 4 review, we don't assume fission product would go through the 5 suppression pool.

-6 DR. KRESS: I'know, but if you assume -- if you

7 did talk about that part to go through the suppression pool, I

8 you would use.1465. And I am saying, shouldn't you use 1465 9 for the amount that goes through the valve? Just to be

. 10 consistent. And not recalculate some more deposition in the 11 main steam line.

12 MR. LEE: If I understand your question right, Dr.

13 Kress, the amount removed within the reactor coolant system 14 is already counted --

() 15 DR. KRESS: Yes.

16 MR. LEE: When we said so much percent is released 17 to the dry well.

18 DR. KRESS: Yes, that was my point. l 2 .

19 MR. LEE: Right. Yes, we do.

20 DR. KRESS: Okay. Are you not now double counting 21 it with that bullet?

22- MR. LEE: No, we do not.

23 DR. KRESS: Okay. You say it goes from the dry 24 well through this thing.

1 25 MR. . LEE: Yes. Through the dry well. That's '

1 y ) ANN RILEYL& ASSOCIATES, LTD.

\m> - Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

l

80 1 right.

(~' 2 Now, the Brockmann's model is one of the models we

(- 3 are.considering in the main steam line. We are still 4 looking at other models as well, and it happened to have --

5 at this moment we have a Brockmann's model in it. And we 6 are -- particularly the Office of Research is reevaluating 4

7 whether this model is really indeed appropriate to use, 8 considering whether it is a plug flow or a well-mixed. For l'

9 example, the Perry proposal, I think they used a combination 10 as well.

11 Now I understood until now Brockmann's model is 12- plug flow, but I was corrected recently that that's not the l

131 . case, and it could be a sort of a combination of plug flow .

14 as well as well-mixed model, and really they use the same l

()

15 equation in the NUREG, contrary to 6189, this is a basic

]

16. sedimentation equation.

17 DR.,KRESS: Sedimentation --

18 MR. LEE: Right.

19 DR. KRESS: Corroboration calculation.

20 MR. LEE: Right, principal removal mechanism is l

21 indeed gravitational settlement. Now, in Brockmann's model 22 he further refined his basic sedimentation equation to 23 account some additional effect such as well-mixed, so we are

24 not quite sure at this time what part of it is plug flow and L

25 what part of it is well-mixed. '

l.

. f

r- (

ANN RILEY & ASSOCIATES, LTD.

Court Reporters ,

1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. - . . _ . . - . - - - - - - . - - - - - - . .~ -- . - - . -. ..

1 l

l

- 81 i

'l DR. WALLIS: Isn't'there some flow? Isn't there i

L /~"s L 2 some flow in the steam lines? Aren't there some secondary I k _[ s

\

=3 flows and circulations and all sorts of things? l l

4 MR. LEE: No , this is just a steam lake, you 5 know --

6 DR. WALLIS: Yes, but --  !

l 7 MR. LEE: About 100 CFH of --

8 DR. WALLIS: Well, some parts are' hotter than 9- others and there are --

l 10 MR. LEE: You mean temperature-wise?-

I .. 11- DR. WALLIS: Yes, there are reasons why you might  !

12 get convection patterns inside the lines themselves.

13 MR. ~ LEE: Well, this is, you know, large, 24-inch

14. . carbon steel pipe with insulation, and it's a long pipe.  !

l

(-x) .15 I'm not sure whether we have enough temperature gradients, i

16 temperature change along the pipe to have such a well-mixed 17 effect. Again, we are still under consideration --

18 DR. KRESS: You treat the main steam line up to 19 the valve --

l 20L MR. LEE: Yes.

21' DR. KRESS

As one control volume?

22 MR. LEE: No. Actually in the Perry case they 23 have three -- inboard MSIV, outboard MSIV, and the stop u

24 valve. So we have three different control volume per line

.25 except of course the broken one.

i f]

't s,j .

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l

(202) 842-0034

1 82 1 DR. POWERS: When you have a 24-inch pipe --

2 :MR. LEE: Yes'.

3 DR. POWERS: Isn't flow basically turbulent for 4' any flow velocity?~

5 .MR. LEE: I don't think --

6' DR. POWERS: The length dimension is so big 7 that --

8 MR. LEE: But the other hand, the movement of 9- particle cn aerosol, you know, 100 CFH is something like 1.7 10 . cubic feet per minute. So you are really talking about a 11- pretty'small amount of a leak.

12 DR. "Gii"RS: 10h , it's much less than even a 13 centimeter ceccad flow velocities.

14 MR. LEE: I believe so. I didn't calculate that.

( 15l DR. KRESS: If you used a well-mixed assumption, 16 wouldn't that be conservative then?

17 MR. LEF: Well-mixed assumption is conservative;

.1;8 yes.

19. DR. KRESS: In the. sense that well-mixed would 20 give you less agglomeration and fallout than a more -- yes, 21 I think it is conservative if we assume well-mixed rather

.22- .than plug flow.

23 Yes, I think it would be.

24 DR.. POWERS: I would think that it would give 25 you -- the well-mixed would enhance your deposition and O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 0,

o

. __ _ _ . . _ _ . . . _ . . . .- _ _ _ ___ _ _ _ ... _ . m . . . .

i 83 1 minimize your growth. What happens if it's not well-mixed?

("] 2 I mean, is it the center that -- volume is not dropping LJ 3 things -- doesn't have any opportunity to drop things out by 4 any mechanism? Okay, so those particles sit there and grow.

5 Okay, by making it well-mixed, you bring things 6 down close to a boundary where they can fall off. I'll bet 7 you that it would make much difference, because if you're 8 not well mixed, then you grow up to the point that they l 9 could sediment out and if it is well-mixed, that brings them l

10 close to the surface where they do sediment.  !

I 11- DR. KRESS: Probably true. )

12 DR. POWERS: I mean, that's not -- woulcn't be the 13 first time that an aerosol probably was insenbitive to that 14 assumption. And the assumption makes it mathematically I i 15 tractable.

V 16 MR. LEE: You know, also a well-mixed model will 17 reach the end of the pipe quicker, which is more ,

1 18 conservative. l 19 DR. WALLIS: I wasn't sure it was isothermal, one i

20 or two degrees temperature difference may give you natural 21 -convection patterns which are far bigger than the convection 22 that we've been talking about.

23 MR. LEE: I'm not sure whether there is such a 24 temperature difference.

25 DR. POWERS: But you'd still be conservative, I

'\ - ANN RILEY & ASSOCIATES, LTD. 1 s_) Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I

84 1 assuming isothermality, because one or two degrees' 2 temperature difference will drive your thermal phoretic

[~)T L

3 deposition crazy. That's plenty to drive --

4 DR. WALLIS: They're being very conservative.

5 DR. POWERS: Yes. I think they are. Sounds like 6 it to me. Or if they're not conservative, they're in a 7 region where, okay, maybe it makes ten percent difference 8 sort of things, I mean, not huge differences.

9 MR. LEE: Yes, this particular model is different 10 from the other one, that this is a single value and we're 11 using single best value rather than range. This is not the 12 Monte Carlo type analysis. So changing the value for the 13 particle densities or material -- excuse me, material 14 densities or particle size or shape factor will greatly

(,m) 15 influence the deposition velocity. So I think right now 16 Research is doing some sensitivity studies comparing some of 17 the major parameters depending on the range of these 18 important parameters.

19 DR. WALLIS: These are all new calculations since 20 the license basis of the plant?

21 MR. LEE: Yes. This is no model. We didn't do 22 this way in the Perry design basis review time.

23 Next is spray. Yes, they do have safety related 24 containment spray. I can assure you, this is a simplified 25 model that we use. All you need is water flux and water

(~ ANN RILEY & ASSOCIATES, LTD.

l (s,}/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. . - . . . . _ _ _ _ _ . . _ _ . _ _ . _ . . _ _ _ _ . . . _ . . _ . . . _ . _ ___m_

i 85 1 full height. That's all you need, Just plug that number I

2' into the code.

3 DR. WALLIS: Does it measure how small the drops l

4 are?

l l '5 MR. LEE: The particle drop size is already 6 counted. It has range, again, as one of the Monte Carlo

! 7 parameter range, the particle size --

8 DR. WALLIS: The spray.

9 MR. LEE: Sir?

l 10 DR. WALLIS: The spray side.

11 MR. LEE: Yes.

l 12 DR. WALLIS: So all you need is a water flux, so 13 you can --

14 MR. LEE: Water flux, that means the water flow

,yy i 15 1 against a given surface area. DR. WALLIS: So you could

%f 16 have one big jet or you could have a spray, and then it 17 .doesn't make'--

18 DR. POWERS: Fortunately, all of the sprays in 19 these plants use the same spray nozzle.

20 DR. WALLIS: Same nozzle. Okay.

21 DR. POWERS: In fact, the analysis was based on a 22 relatively detailed examination of droplet size, 12 3 distributions, and a relatively detailed examination of 24 range of particle size.

25 DR. WALLIS: So it only gets simplified late in l

ANN RILEY & ASSOCIATES, LTD.

((~'b L(_s Court Reporters L 1025 Connecticut Avenue, NW,-Suite 1014 Washington, D.C. 20036' (202) 842-0034 4

.___ _. . - _ _ _ _ . __ -- ~ _ _ _ . - - _ . _ _ _ _ _ _ _ _ _ . . _ _ _ _ _

86 1 the analysis.

,[y 2 MR. LEE: That's right. Yeah. I mean, that's

\_)

3 what he's saying, is that you do a Monte Carlo analysis on 4 this range, and then you get a distributed output, and then 5 once you have that, then you do the simplifications.

6 Actually, all nuclear power plants have two types 7 of nozzles and the Perry happened to have what we call 8 1713(a) type which is commonly used. So once you have that 9 spray nozzle, we know pretty much the characteristic of a 10 water spray into the containment, and that model is also in 11 the RADTRAD code. We just incorporated all that model right i

12 into the code. 1 13 Okay. Now we are coming to the iodine chemistry 14 part. As I briefly indicated in the previous sides, yes, we l( ) 15 did look into buffering capability with 5200 pounds of 16 sodium pentaborate, and also we evaluated the formation of 17 nitric acid resulting from radiolysis of nitrogen in the 18 containment atmosphere, also in nitrogen in the water.

19- We discussed that formation in, as I recall, AP600 20 question with you. We did look into some sensitivity study 21 where formation of a nitric acid is really whether it's a 22 gas-phased phenomenon or liquid-phase phenomenon. Our Oak l

[ 23 Ridge contractor or Oak Ridge people still think it is

! 24 really a water phase, aqueous phase phenomena, that's where 25 the nitric acid is forming.

/~'g ANN RILEY & ASSOCIATES, LTD.

'\j Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i

i

87 1 But when you look into the G-value, you know, we

(~^) 2 used the G-value for the nitrogen atmosphere, a value of 1.9

(_ '

3 that's suggested by our Sandia report for iodine chemistry 4 for MELCOR that we just reviewed. We took that value. Then 5 water phase, we used the G-value of .007. That's an Oak 6 Ridge number. You can see there's, what, 200 or 300 factor l 7 difference.

8 But on the other hand, the density of water and 9 density of air, there's a lot --

10 DR. POWERS: A thousand.

11 MR. LEE: About a thousand difference there. And I

12 there is also more radiation in the water compared to the 13 atmosphere due to the aerosol deposition and the spray and l 14 so forth. You've got more of a source of radioactivity in l

<~s

( ) 15 the water than atmosphere.

%)

16 But yes, we did look at both phenomena, and we did 17 receive a draft report from Oak Ridge. What they did was 18 they took -- they measured hydrogen ion concentration, which 19 is pH, using the proportional amount of Perry containment 20 water value, then used the proportional amount of sodium 21 pentaborate, and then they mixed that one up and then 22 started titrating it with nitric acid, what you expect it to 23 produce, and then also they used the hydrochloric acid for

'24 what they expected for irradiation of cable insulation 25 material, and they measure the pH actually and they came out

/)

(,/

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 101s Washington, D.C. 20036 (202) 842-0034

. ._ _. . . _ _ _ . _ . . ~ __ ____.___.__ . - _ . . _ , _ _ .

j f

88 1 very close to the value that the Perry proposed.

}

J.

2 Perry's number is -- really calculated a value 3 against our.value. Its experimental value actually measured

'4 at Oak Ridge.came up fairly close; happened to be a pH value

[ 5 of 8.3. So whether this nitric acid is formed in the 6 atmosphere or water, you know, the pH value of 8.3, the 7 difference is really small, like a pH value of perhaps 8.1 L 8 lagainst 8.3, depending on where nitric acid is forming.

9 Now, we also look at hydrochloric acid formation 10 resulting from radiolysis of' electric cable insulation. In L 11 this case, it's hypalon that the Perry is using. Perry has l 12 approximately 29,000 pounds of hypalon inside the 13 containment, and about 3,700 pounds inside the drywell. So 14 you can imagine that the potential is so much that

() 15.

16 hydrochloric acid can be formed.

So that's really part of our partition on iodine 17 from aqueous over gas phase and the effect of hydrogen ion 18 concentration,'which is the pH, on iodine partition.

19 Also,.yes, we did look into dissolution of carbon 20 dioxide into the water. Now, we are talking about here 21 design basis accident, so therefore, there is really no 22 contribution of carbon dioxide from core concrete 23- interaction, but it's coming from the carbon dioxide in the 24 air as well as possibly some organic materials. I think 25 your committee expressed interest whether we considered l

L l-l l /~% ANN RILEY & ASSOCIATES, LTD.

j \m ) ' Court Reporters 1025 Connecticut Avenue,-NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

,,.n. , .,

89 1 carbon dioxide -- yes, we did.

! (~') 2 Now, hydrolysis of a cation, yes, we also

< q.) .

i 3 considered that as a hydroxide. We assumed -- I think Perry 4 assumed also -- 90 percent is in the form of hydroxide, l

5 therefore, yes, we did include it, we did incorporate it and 6 we did consider.

7 DR. KRESS: Now, do the Reg Guides' regulations 1

8 ask for some sort of consideration of partitioning of l 1

9 iodine? '

10 MR. LEE: Yes. That's the future. Steve alluded 11 that next spring, there will be a draft copy of regulatory 12 guide.

13 DR. KRESS: That will be in the Reg Guide.

14 MR. LEE: Yes, as well as -

T ) 15 DR. KRESS: So that's why you did it here, because v

16 you expect it to be in the Reg Guide?

17 MR. LEE: Yes. Certainly we do. You will see 18 that.

19 So as I said, Oak Ridge experimentally measured 20 the pH value and they came out about 8.3 or so.

21 DR. POWERS: How did you handle the formation of 22 organic iodides?

23 MR. LEE: It's coming next.

24 DR. POWERS: Oh, okay.

25 MR. LEE: It's not in the slide.

[N- ANN RILEY & ASSOCIATES, LTD.

(_sl Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

90 1 DR. POWERS: All right.

('j'T

\

2 MR. LEE: Mostly you expressed interest in a few 3 items. The committee expressed an interest on the 4 partitioning of atomic iodine, iodine formation you briefly 1 l

5 alluded, and the formation of a volatile organic iodine, and l i

6 the type and amount of organic materials in the containment I 7 and in the sump.

8 Now, these items are not in review at this moment. l 9 Now, speaking of partitioning of atomic iodine, I guess my 10 feel or my view is that atomic iodine is highly reactive, 11 and I don't think it will stay too long as atomic iodine; it 12 will become I-2 iodine. So the life span of atomic iodine 13 will be perhaps relatively short, and due to the reactive -- i 14 such a great reactive nature of atomic iodine, it will (n%_-)15 quickly react with other molecules in forming other 16 compounds.

17 I know you addressed that issue in iodine 18 chemistry for MELCOR. I just reviewed that and we are 19 sending you the comment -- we are sending the comment to 20 Sandia, and we also have comments of a peer review group on 21 this particular item.

22 So what I'm saying is yes, we are aware of this 23 particular issue. We are reviewing it, we are considering.

24 Even though we are not applying this for the Perry review at 25 this time, we may address next year the Reg Guide or SRP.

/~ ANN RILEY & ASSOCIATES, LTD.

hs,h) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036

(202) 842-0034 l

.J de da.u.A A_d4 i&4&J. -4 4% *4 --h4d. Le.mi+k-._*. rah 4 2a.-*--dm=a-8 aaa s Z.&#.- 4 4an , wa m.. m

+- u 91 1 DR. POWERS: The only people that have advocated t-

'T 2 atomic iodine partitioning in any published form that I know (d 3 of are the: British, and they did it to account for the 4 higher levels of partitioning they were seeing from basic l 5 solutions --

6 MR. LEE: Right.

7 DR. POWERS: -- than what they anticipated.

8 MR. LEE: Yeah,-lower -- I mean the higher pH.

9 DR. POWERS: Yeah, I mean they ran pH 9 and they 10 expected to see very, very low partitioning, and they got 11 lower partitioning than you would at lower pH's, but it 12' wasn't as low as they would have calculated. And so they 13 invoked this atomic iodine partitioning, which, quite 14 frankly, I think -- I think we were all shocked when we

() 15 16 first saw it because nobody had ever thought about atomic --

MR. LEE: We did, too.

17 DR. POWERS: Atomic iodine partitioning. So, I 18 mean it is a very seminal issue here. I mean it is 19 speculation. There is no -- like you say, even if the i

20 atomic iodine does partition, it will turn into something 21 else three nanoseconds after it gets into the atmosphere 22 because atomic iodine has got -- it is a radical species.

23 It is not going to want to stay atomic iodine very long, and

~24 there is no reason to think that it is atomic iodine exactly 25 in solution either. And so there is a lot of speculation on D ANN RILEY & ASSOCIATES, LTD.

,() Court Reporters

. 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 L (202) 842-0034 L

92 1 the part of the British there, but it is one that deserves

~

2 attention just because that was the only way they could 3 explain their experimental results.

L 4 MR. LEE: You know, I tried to get ahold of author 5 and I guess Beahm of Oak Ridge did, too, but we were j 6' unsuccessful to get ahold of him. But all I can say at this J7 time is, yes, we are aware of'this phenomena and we are 8 considering. But for the_ Perry review, we did not look at 9 this issue.

10 The same thing is true for silver iodine. You l 11 know, silver is great for holding onto iodine in the water.

l 12 Once you form-silver iodine, it is very insoluble. So, 13 ~ therefore, we do think'the amount of a partition into the 14 atmosphere. So, not considering silver iodine is i/g -15

! 1 conservative in my view.

1

-Q 16 DR. POWERS: We don't have much silver in this 17 plant anyway. Well, there is some.

l' l 18 MR. LEE: Yeah.

19 DR. POWERS: But there's - 'I mean it is not like R20 a PWR with silver indium cadmium control -- I mean there are 21 other' things that will react with iodine, but there's a lot 22' of speculation there.

23: MR. LEE: The formation.of volatile organic n

24 iodine, we did not consider -- we are considering that in 25 the Perry review either at this time. We just don't have

ANN RILEY & ASSOCIATES, LTD.

"A Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036

! (202) 842-0034

l 93

1 enough reference to look into this at this time. Again, all l- 2 I can say is we are aware of this issue, we are considering, l

d(~N 3 and we may perhaps apply this in the future review.

4 DR. POWERS: Didn't Oak Ridge, some years ago, do i 5 some experiments in which they bubbled methane through 6 irradiated aqueous solutions at various pH's and get iodine 7 conversion factors?

8 DR. KRESS: Yes, they did. And I don't recall the 9 reference but the -- where it is published, but they did 10 that, and, of course, it converts, by -- converts it into 11 methyl-iodine very, very efficiently, even at pretty high 12 pH's.

13 DR. POWERS: And I have just recently seen a, 14 unfortunately, draft report from the Finns in which they

( 15 went back and reproduced the Oak Ridge experiments and then

( )

16 did a much larger test matrix, varying lots of things. It 17 is -- it is not something you can use right now because it 18 is a draft report, but I think somehow you are going to have 19 to factor this in because we have got a certain amount of 20 experimental information.

21 DR. KRESS: Yes. And the key is figuring how much 22 methane you have and where it comes from and how fast.

23 DR. POWERS: Well, I think that -- I think that 24 you have got a pretty good handle on that, or some handle on l

25 that if you can calculate how much chlorine you are getting fT ANN RILEY & ASSOCIATES, LTD.

"(_ji Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

__ . . _ m _m _ _ _ _ _ _ _ _ . _ _ . . . _ . . . _ . . . _ . _ . _ _ . _

94 1 off. Then, assuredly, you can estimate how much methane you

2' are getting off from the same source, and that ought to give

3 you a reasonable number. I mean it may not be 100 percent 4 . accurate, but what you know is you are going to have far i 5 more methane than you are ever going to have iodine in these 6 things.- I mean with 29,000 pounds of insulation to work 7' -with, 8 [ Laughter.]

.9 DR. KRESS: A lot of methane potential there.

-10 DR. POWERS: That's enough. And what, seven

11. kilograms -- seven kilograms of iodine altogether? l l 12 MR. LEE: Something like that. That magnitude.

1 13 DR. POWERS: Versus 29,000 pounds 1of organic.

14 MR. LEE: Right. I mean there is no question it 1

' h"'T 15 l t j is lots of organic material in the containment. How'that )

'16- 'really forming the organic iodine is something we will l 17 consider and one of the-items'that we are looking into. -

[

p.

~18 Again, the same thing is true for-the type and L

L 11 9 amount of organic material in the -- it is the same, really, 20 subject areas as a hyperlock case.

21~ Okay. Next one is thermal hydraulics and the 22 fission product transferred from the drywell to containment.

-231 Now,_ traditionally, in the TID source term, we assume that

!/ 24 fission ~ product release is at the time zero uniformly 25 distributed in the containment, and now NUREG-1465 has B

% ANN RILEY & ASSOCIATES, LTD.

<1j- Court Reporters

, 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l_ (202) 842-0034 l-.

I

i 95 l 1 released timing. It is something like more similar to the

(~)/

2 perhaps small break LOCA type because it is releasing for l \m-3 two hour period.

4 So for the Perry case, we are still reviewing the 5 heat generation rate and also steaming rates. Perry 6 proposed a certain steam production rate and we are still 7 reviewing that. And I think research is looking into coming t

8 up some perhaps range of 3,000 CFM of steaming rate, but, 9 still, --

10 DR. WALLIS: Steaming from where?

11 MR. LEE: Steaming from the heat generated in the 12 core due to the quenching of the vessel at two hours. Now, 13 DBA source term, you know, we are saying that the vessel 14 will not fail and somehow it will be reflooded or quenched 15 with water and so that's when it will generate heat and l

(Ov) 16 steam.

17 MR. EMCH: Jay, you are kind of talking about some j 18 details that we haven't fully discussed with the licensee.

19 I think we had better move on and keep it on a higher plane.

20 MR. LEE: Yeah. Well, that just happened to be 21 next to the last item. And the last one is also we are 22 looking at the medalogical dispersion factors. I mentioned 23 that Pacific Northwest is helping us to review this area, 24 just for the control air intake, and that this is under 25 review and we have a low numbar to discuss.

O ANN RILEY & ASSOCIATES, LTD.

' k ,)

m Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. . . . . .- . ~ . . . ~ ~ - . - . -. . . . - . - . . - . . ~ . ~ . . . . .

I j 96 j i

1 DR. .KRESS: Are you using building wake models? .I

(~} 2 MR. LEE: Yes. I believe we are using the ARCA-96

!%)

3 model_that we just developed. So that's -- concludes my l 4 presentation. If you have any question or anything.

5' DR. WALLIS: Well, you spent a lot of time on l

l6 chemistry and then we weren't allowed to discuss this  !

l i

7 -thermal hydraulic item. 'Isn't there some uncertainty l 8 associated --

o 9 DR. POWERS: It's the relative importance of the r

lc 10 two technical issues.

11 (Laughter.]

12 DR. WALLIS: Are you declining to speak about it l

?

13 because of endurance or --

l 14 MR. LEE: Oh, no, we are still in the preliminary 15 ' stage of review.

[b}

16 MR. EMCH: Most of the discussion you folks had 17 about chemistry was very generic in nature and so I didn't i

l 18 . interrupt, but on the thermal hydraulics,-what Jay was 19 starting to do was get into specifics that have not been ,

l 20 fully developed with the licensee, and I don't think it is i 21 appropriate for us to start getting into it here when we 22- _ haven't finished with the licensee.

23 DR. WALLIS: Well, that doesn't mean it is 24 unimportant.

l 25 MR. LEE: Pardon?

t k

. ANN RILEY & ASSOCIATES, LTD.

j s Court Reporters 1025. Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L

! l

! i l 97 J 1 DR. WALLIS: That does not mean --  !

1 2 MR. LEE: Oh, definitely not. This is, of course, I 3 important. l 4 DR. WALLIS: So when we get the whole story, we  !:

5 will hear about it?

1 6 MR. LEE: Rich, are you going to address that l l

7 question?

8 MR. EMCH: Yeah. Earlier'today Dr. Kress and I  !

l 9 talked about the potential need for us to come back and have i l

10 a more lengthy discussion about the Perry review {

11 specifically when there is more to tell you, yes. j 12 DR. KRESS: I think our letter will not go into 0

13 Perry very deeply because it really is not the purpose. We 14 -will talk about the rulemaking, I think. We will let Perry

> p} ; 15 rest for --

, M' 16 CHAIRMAN SEALE: We have an abiding curiosity for 17 dose calculations, where building wakes and things like that 18 are important, too.

So if we do get into that, we may want 19 to also hear a little bit more about the building wake t

. 20 problem.

21- DR. KRESS: I don't think we have any i i

22 presentations-from the industry, or do we, Paul? You are 23 welcome to make any comments, i 24 MR. BOEENERT

Okay. We was the only one.

25 DR. KRESS: All right, with that then, I will turn ANN RILEY & ASSOCIATES, LTD.

t Court Reporters 1025 Connecticut Avenue, NW, Suite 1014

!. Washington, D.C. 20036 (202) 842-0034 t

t

~

l l- 98 +

1- it back to you.

2 CHAIRMAN SEALE: 'Okay.

3 DR. KRESS: 'You know, not too far behind.

'4 CHAIRMAN SEALE: All right. I think what I am +

5 going'to do is call your attention to the fact that if you 6 want to get coffee, you had better get it in about seven 1

7 minutes. So we will recess for 15 and be back here for the r

8 next.

9: (Recess.]

i 10 CHAIRMAN SEALE: Okay. Our next presentation is '

11 on the assessment of pressurized water reactor primary 12 system leaks. John Barton is the chairman of that 113- subcommittee, so John, I'll ask you to introduce our.

14 speakers. '

15

) DR. BARTON: Thank you, Mr. Chairman. This is 11 6 kind of a wet session, I think.

17 CHAIRMAN SEALE: Our enthusiasm is not dampening.

-18 DR. BARTON: The purpose of this -- too many 19' hurricanes to live through, yes -- the purpose of this 20' session is to discuss with representatives of the NRC staff f 21 and its contractor, INEEL, regarding results of the study of 22 the final draft NUREG-CR-6582, assessment of pressurized

23' water reactor primary system leaks.

b 24 The main objective of this study -- and I'd like 25 to comment, I think the study was well done -- was to review p

E i

'(Y ANN RILEY-& ASSOCIATES, LTD.

,3s,)

^

Court Reporters y 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

99 1 the U.S. experience relating to pressurized water primary

, ()

i \m) 2 system leakage in terms of their number and rates, how aging L 3 affects the frequency of leakage events, the safety 4 significance of such leakage, and industry efforts to reduce 5 the leakage events. The report also discusues current 6 leakage-detection methods.

7 With that I'll turn this over to Jack Rosenthal of 8 NRC staff to introduce the topic.

9 Jack?

10 MR. ROSENTHAL: Thank you. My name is Jack 11 Rosenthal. I'm the branch chief of the Reactor Analysis 12 Branch, AEOD, and sitting next to me is Dr. Ernie Rossi, who 13 is the division director of the Safety Program Division, and 14 George Lanac, a section chief in the Reactor Analysis

(

Q) 15 Branch.

16 We did the -- what you're going to hear about is a 17 rcport, as was said, on primary system leaks. Chuck Hsu of 18 the NRC staff was the tech monitor and intimately involved 19 in the technical work of this report, and then the body of 20 the report will be presented by Vik Shah, who is at INEEL 21 and is the principal author.

22 This report looks at the phenomenology of primary 23 system leaks, and there is a companion piece of work that is 24 being done in the Safety Program Division, and we call that 25 an initiating event study, and you've been briefed on a i

'[h

\_ /

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

100 1 draft version of that report. And so where here we're

' [~} 2 looking at the frequency of primary system leaks and we're i s/

3 looking at the underlying phenomenology, it's that companion

)

4 report which references this one that talks about the --  !

5 that focused on the frequency and the risk of these in more 6 detail. This provides the factual basis for that PRA work, 7 And so that the reports are linked.

8 With that, let me turn it over to Chuck for the 9 introduction.

10 MR. HSU: Okay. I'm Chuck Hsu, mechanical 11 engineer from AEOD. I'm here to provide information abcut j 12 our current study. The title is the assessment of the 13 pressurized water reactor primary system leak.

l 14 In the past we saw a vast number of leaks through j f- i (3) v 15 the primary system monitoring. Some of the leaks challenge 16 the integrity, the pressure boundary. Others disable the 17 system. So we say it depends on leak rate, location, and 18 degradation.

19 So we believe that a comprehensive study is needed 20 to address the concern of the leaks problem.

21 So we begin to study with the technical assistance 22 of the INEEL contractor. The work started from January last 23 year. We finished that job and the issue in December '96 --

24 '97, last year. Let me put a schedule here.

25 So by March we were supposed to comment here. We I

-['T ANN RILEY & ASSOCIATES, LTD.

(_ / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

101

.1 worked _on the dissolution, comment, and we incorporated t .

("N 2 those comments to the report. So last month, the middle of V 3 October, we issued a final draft report.

4 We expected to finish this report, issue a final 5 report at the end of this month. Let me see what -- where I 6 get the comment.

1 7 The comment we get it from is from NRC, the other 8 office, and also from our office, that's NRR Research.

9 Industry comments include from those organizations.

10 I'd like to point out this one appears,'RSA i

11 technology, lots of people may hold this organization  !

i 12 here -- the engineer consulting in California. This SKI is

. 13 a Swedish nuclear regulatory agency, equivalent to NRC in {

14 this country. In this year they undertake a series of

() 15 studies on the variability of nuclear pipe, and using the 16 whole nuclear operation better to determine the variability i

17 of the pipe and all the other passive components. l 18 I. guess they use that PRA methodology, today's PRA 19 methodology. And those -- they don't work by themselves, 20 they contract to two engineer consulting. One of the

- 21 consulting is RSA. That's the reason we stand, you know, 22 for comment.

23 DR. WALLIS: This report is based on collecting

. 24 data and understanding what happened. I just wonder what 25 the role of all these peer reviews is. Is it to make sure i

4 ANN RILEY & ASSOCIATES, LTD.

..xp)_

. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D . C .. 20036 (202) 842-0034

_.____-.___,_._____.__..m_. - _ . . _ - . _ _ . . . _ . _ _

l 102 l

1 that it - that the presentation is understandable. How do i 2 they contribute to the report?

l 3 MR. HSU: Yes, most of the confusion of this l-4 comment is that in the PRA area, you know, the frequency l 5 area, the f requency seems to me we don' t have too many dat a, 6- because we only restrict to U.S. data, you know. So that's 7 what we send to them. And looking, you know, over the 8 worldwide data, so they can give a good comment to us. So 9 we changed some of the LOCA frequency, some of the transient 10 lL10CA frequency, we changed that. It's based on their 11 comment.

12' MR. ROSENTHAL: But across the board as a matter 13 of SPD policy and procedures we are systematically writing i 14 our reports and putting them out in draft form to a wide 15

( ) audience, soliciting comments, because we think that a 16 peer-reviewed document is a higher-quality document, and by 17 going to a broad enough variety of people we should have 18 surfaced issues of factual correctness, completeness, 19 understandability, et cetera. And so we are systematically 20 'doing that with all of our work.

21 MR. HSU: Our next --

22 DR. WALLIS: Well, it seems to me the most 23 . knowledgeable people are the people who actually did the 24 study who were not the peers. So this is just as of a 25 check. But the most important part of -- aspect of the i-I'

.,~

\. .

ANN RILEY & ASSOCIATES, LTD.

[c c l _j]- Court Reporters lq 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

103

! 1 study is the factual part, which is known to the people who

/N 2 did the work.

N) 3 MR. ROSSI: Well, you're absolutely right, and we 4 have somebody here that's going to talk about that just as 5 soon as Chuck finishes with a few introductory remarks.

6 DR. WALLIS: So it took a year for peers to look 7 at the work?

8 MR. HSU: No, because it's not a particular year, 9 because --

10 DR. WALLIS: It's a year from the peer review to 11 now.

12- MR. HSU: Oh, that's -- we had ours about five to 13 six months because of the problem, you know, so you're not 14 taking the whole this time.

mi f, Q 15 Next slide. Briefly we have a study of scope and 16 finance and implementation -- the contract from INEL. You 17 are provided detail or description of this item. We can 18 answer or discuss any question you might have when we do the 19 presentation.

20 (Pause.]

21 MR. HSU: We have five objectives here in our 22 report and the first one is we tried to develop distribution 23 of the frequency and train, both on the reactor year and the 24 calendar year for the various ones in the category.

25 The second one was identify where the location and jT ANN RILEY & ASSOCIATES, LTD.

(_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

- _,p - _i e --: , a _ _

104 1 evaluation of the causes to see if any location not

(~h 2 previously recognized are identified.

\

% jl 3 Number three, the leak -- we assessed it -- as 4 precursor to the core damage and also tried to see which set 5 of the leak have potential -- after the detection.

6 On number four, we have described industry efforts 7 to reduce the number of the leaks.

8 Number five tried to provide a description on the 9 leak detection system, you know, in other words try to see 10 what the system is doing now, can detect a leak in timing or 11 can have operator to locate on the resource before it is too 12 late.

13 Okay. Next is the scope.

14 DR. WALLIS: I think an important part of this is n

w/

} 15 an understanding of the mechanisms for causing leaks and how 16 good an understanding you actually achieve.

17 I hope to get on to that.

18 MR. HSU: We will get on to that. Okay, the 19 second one is the study of scope. We reviewed 240 leak 20 events from the period 1985 through September, 1996. That's 21 about 10 years. That includes interface system leak and 22 steam generator tube leak.

23 DR. FONTANA: Why did you exclude those?

24 MR. HSU: The reason we excluded it is for 25 interface system leak that in '93 we started interface t ANN RILEY & ASSOCIATES, LTD.

l(]/

i Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

-u d ,

1 system LOCA research program. Now at that time they found i-j{

2 out that interface system LOCA frequencies only span about '

3 10 minus 6 on the PWR, 10 minus 8 for the BWR and in 4 addition we have another study -- the events study in the f .5- AEOD,

, 6' WE'found out in the past 10 year -- that's why we 7 exclude that.

8 'In a steam generator -- we also have another study 9 back in '96 we issue a-big study, a generator tube leak fail I

'10 in April '96. That's why we exclude this on this study.

11 DR. UHRIG: Were there many leaks, just steam 12 generator tube leaks, reported?

13 MR. HSU: .Yes. Not in here, not in --'in our 14 other report.

() 15 MR. UHRIG: How many were -- compared to the 240?

16 Was it more or less-in the 10 year period?

~

17 MR. HSU: More. l 18 MR. UHRIG: Twice as many?

19 MR. SHAH: This product we have not studied.

20 DR. UHRIG: You will have to use the mike, please.

21 MR. SHAH: We did not study the steam generator 22 leaks in this project because in other projects they were 23 specially focused on the steam generator tube leaks.

2

.4 Our main focus here was in the leak detection 25 ~ system which was in the containment and mostly the leaks in ANN RILEY & ASSOCIATES, LTD.

()/

sms Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

! 106 1 the steam generator tube would be from primary side to

! (~) 2 secondary side and that would require a different kind of

\ LJ 3 leak detection system which we were not focusing on.

4 There are leaks but we did not -- I do not have a 5 number right now to tell you.

6 MR. UHRIG: I was just trying to get a relative 7 magnitude of t.be two different kinds of leaks, but it is 8 more steam generator?

9 MR. SHAH: Yes. Quite a few, but that's a 10 separate issue and that's being studied in quite a detail --

11 also rupture tube --

12 DR. UHRIG: What do you consider a leak? A few 13 drops?

14 MR. SHAH: Okay. I will talk about that.

'p) g sm/

15 DR. UHRIG: Okay. I'll wait.

16 MR. ROSENTHAL: From a safety standpoint, what we 17 are most interested in is compiling what the operating 18 experience was and then this is a primary leak study.

19 There's been a lot of other work on steam generator tubes 20 and then looking at which of these leaks could have 21 propagated via small break or large break LOCA.

22 That was the ultimate interest.

23 If we let Chuck just finish this slide --

24. DR. WALLIS: I think it would help if we went back 25 a bit and you told us what a leak is, how it is detected, 1

/7 ANN RILEY & ASSOCIATES, LTD.

\s,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

t 107 1 how big it has to be before it is detected.

2 MR. BARTON: I think you are going to get that l \_J

("')).

3 further in the report.

4 DR. WALLIS: --

and what's been found.

l l

5 MR. SHAH: We try to go into detail with that.

l 6 DR. WALLIS: How do you know? I mean if something 7 is slightly damp somewhere, is that a leak? How do you l 8 know --

9 MR. SHAH: It depends upon where it is.

10 CHAIRMAN SEALE: I think they are going to tell us I 11 that.

12 MR. ROSSI: Could I make a suggestion? Why don't 13 we just go directly to Vic Shah's presentation because I 14' think a lot of the questions will get answered when he

() 15 starts, so Chuck, why don't you just let Vic go ahead, and 16 he is going to probably cover-this last slide that you have 17 left anyway, so --

18 MR. HSU: Okay.

19 MR. ROSSI: And then we can get directly to it 20' before -- because time seems to be going by, so Vic, why 21 don't you just go ahead?

22- MR. SHAH: Okay.

23 Okay -- so everybody can hear?

24 DR. WALLIS: Could you start as if you were 25 -explaining to a class of sophomores and tell me what a leak L'(,p/ ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i

l'

108 1 is and what happens when you find one and so on?

(~ s 2- MR. SRAH: I will do that right away, sir.

(.

3 DR. WALLIS: All right.

4 MR. SHAH: A utility, a licensee has to submit a 5' LER if there is a leak. Only certain leaks he has to 6 report. There are three criteria.

7 One, if the requirement specified in technical 8 cracification is violated -- I will tell you what are the 9 requirements; second, if the' limits specified in FSAR for

10. the control room habitability is exceeded; or if there is an 11 unplanned actuation of the safety system, then they have to

-J? report the leak.

13 DR. WALLIS: So there could be many leaks which

-14 are of the few drops variety which never get reported at D 15 all?

.k_)

16 MR. SHAH: Yes. Let me tell you what the 17 technical specifications in the requirement, okay?

18 In the technical specifications there.are three 19 requirements: identify the leak and if it is greater than 20 10 GPM you have to report it.

21 DR. WALLIS: .10 GPM?

22 MR. SHAH: Identified leak -- so you know where it 23 is taking place and where it is going.

24 If it is an unidentified leak, and it is greater

.25 than 1 GPM he has to report it -- and if there is a leak O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i

l 2

l 109 1 through the pressure boundary, any leak, you have to report

'( ) 2 it, maybe a drop, if it is through the pressure boundary.

3 DR. WALLIS: 10 GPM is 10,000 gallons a day or 4 something. Can't miss it.

5 MR. SHAH: Yes, sir, but let me give you an 6 example. In a valve there is a leak offline so that leak 7 goes to a collection then so you know where it is coming 8 from, where it is going. It is not going in the 9 environment. That is identified leak.

10 Unidentified leak, it is release in the 11 containment. That is a one gallon per minute limit, but if 12 it is through the pressure boundary -- it means it is a 13 non-isolatable and it is through the crack in the piping or 14 vessel component, then you have to report any size leak. It

,~

'( ) 15 may be a .1 GPM or smaller. You have to report it.

16 DR. WALLIS: How do you know you have a leak?  ;

17 MR. SHAH: There are three types of leak detection 18 system. One is the leak -- there is some containment sump 19 and the leakage gets collected there. Or there are 1

20 radiation monitor and, third, you can do the inventory (

21 balance of the reactor coolant system. Yes, sir.

22 DR. WALLIS: But you never get a small leak from 23 any of those.

24 MR. SHAH: You can get -- okay. In the sump 25 collection, in the containment, ideally, you can get a one l[~)

x_ /

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

110 1 gallon per minute leak in less than 10 minutes. I will show

[

(-

2 you -- I will give you -- explain to you. In the radiation 3 monitor, you can -- like, if you have particulate -- have 4 one particulate radiation monitor, you -- under certain 1.

5' condition, you can get a 1 GPM leak in less than an hour, or 6 even-smaller. It depends upon the fuel condition. So that 7 will not be one -- the radiation monitor will not give you a 8 good detection under all condition.

9 DR. WALLIS: I guess -- I don't want to go on.

10 MR. SHAH: Okay.

11 DR. WALLIS: But we are talking about what looked 12 to me like big holes.

13 MR. SHAH: No , it can be very -- it can be --

14 DR. WALLIS: No, but this is not a few drops of 15 water.

16 MR. ROSSI: Well, if it is through the pressure 17 boundary, it could be --

18 MR. SHAH: Can be .1 GPM or smaller.

19 MR. ROSSI: -- a fairly small leak. If it is 1 20 through the reactor coolant system pressure boundary, then 21 it is reportable.

22 MR. SHAH: Yeah.

23 MR. ROSSI: And those kind of leaks, from time to 24 time, are found just by visual examination or other 25 examinations of the piping. And, of course, that are the C ANN RILEY & ASSOCIATES, LTD.

,4,,_)) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L

. - . . . . ~- - . .- - .-- . . .. .

111 1 ones that we are interested, they are the ones through the 3

)"~'Y 2 reactor coolant system pressure boundary. But there is no-

'% f 3- question'that not every leak in the plant gets reported. I 4 mean it has got to meet the reporting requirements before it J

5 does get reported.

6 DR. BARTON: But they are identified through 7- leakage collecting systems, leakage plus visual and 8 walkdowns.

9 MR. ROSSI: There is a section of his presentation 10 which is at-the end that talks about leak detection systems 11- and what they can and cannot do. So he will talk about that  !

l 12 towards the end. )

13 DR. UHRIG: What is the limit in these leaks that .

i 14 would require you to shut down? Is there a specific number' ,

i 15 or --

L 16 MR. SHAH: Yes, sir.

17 MR. ROSSI: Tech spec number. It is in the tech 18 spec.

19 MR. SHAH: They say that if you find a leak and if 20 you.cannot control it or stop it in so many hours, the plant i

21 has to be in hot shutdown or cool shutdown. All those 22 requirements are in the technical specification.

l 23 DR. UHRIG: But they are different for the steam l- 24 generators than they are other leaks.

l 25 MR. SHAH: That's right. This is -- we are only L

ANN RILEY & ASSOCIATES, LTD.

s_,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036

! (202) 842-0034 l

., m. .

112

1. going-to talk about;the' leakage in the containment through  !

2 the pressure boundary or walls or pump seals.  ;

3 'Shall we go --

E 4 MR. ROSSI: Again, the limits are typically one 5 gallon per minute of: unidentified leakage and 10 gallons per

• 6' minutes of identified leakage. And'if you go.beyond that, '

7' then, generally,~the technical specifications would require 8 'that the plant be shut'--  ;

9 MR. SHAH: Often the utility submits the LER  ;

10 before the leak rate reaches'that limit because it in the .

11 potential for the leak to grow to that limit and they will 12 submit LER. '

13 DR. WALLIS: This is a leak as water,.not as ,

i L

14 steam?

( ) ' 15 MR. SHAH: It can be both. There is a collector 116' system and from that you can collect the steam --

17 radioactivity monitorLwill detect the steam also.

18 DR. SHACK:

.ou do certainly have leaks that go on 19 for a long time undetected, piled up boric acid.

-20 MR. SHAH: What happens when we -- )

~

i .

21 DR. SHACK: Accumulations of considerable 22 magnitudes.

r-23 MR. SHAH: Agree. Many times -- many times we 12 4 have not -- I mean several small leakage have not been 25 detected by current leak detection system, and I will

[ ANN RILEY & ASSOCIATES, LTD.

[ . Court Reporters

1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

m. ._ . . . _ _ _ _ _ _ _ _ . . . ..=. _ _. _ . _ ,_. _ ._ .. . _ _ _ _ _ _ . . . . _ .

113 1-  : identify that, and only time you find it when you go for a jN .

u 2 containment walkdown and then you will notice it because of

%.)

3 the boric acid crystals or because of dampness, and we will 4 -- we willLtalk about that as my presentation. Okay.

~5 So let me jump -- what we did. Okay. I think we 6 already talk about objectives and other things. Okay.

7. In Oak Ridge NRC had a search system. We used 8 'thatfsearch system to identify the events where the leakage 9- took place. We came across two -- okay. We covered period 10 from 1985 January to 1996, end of third quarter, and we 11 ' identified 215 LER, which referred to the leaks. So we 12 reviewed those LERs. In some LERs there were more than one 13 leakage in two different locations leakage might have taken 14 place. :So we considered that two different leak events.

[ %

15L So we said 215 LER, we had totally 240 leak 16 . events. When we reviewed this LER,'we found that some of 17 the LERs were~ reported because of some other event, and leak

(- 18 was not the main thing. If that other event did not take 19 place, they would not have reported LER. So we took those

l. 20 events out, there were 41 events. So, totally 199 events 21 that were reportable event according to those criteria, 22 leak, he has to report it. And we analyzed those 199 leak j 23 events. Okay.

~2 4 And we developed a database how the leak was 25L detected, what was the rate, what caused the leakage, where  ;

' l

~

ANN RILEY & ASSOCIATES, LTD.

Court Reporters

. 1025 Connecticut Avenue, NW, Suite 1014

[ -Washington, D.C. 20036 L (202) 842-0034 l-

-_ , ~ - - . . . . - - . . . . - . - . - . = _ . - .. . . . . - . . . . . _- ...

114 1- was.it, what were the consequences, and what were -- how it j '2 was' mitigated. According to that, we developed a database,

3. ~and then -- generally,-LER does not provide sufficient i 4 information about the root cause analysis, so we reviewed l

'5 the related technical literature and then we analyzed the 6 five different category of leak events with a detailed I 7 analysis with the root cause, including the root cause, and 8 then we visited Duke Power and Florida Power & Light to get 9 additional information on these leak events. Because some 10 of those events just took place a year ago, or six months 11 ago,-and so there were not sufficient public information.

12 We divided this research into seven categories.

13 The first one is a trend. How the leak event change with 14 time? How do they change with age? What is the more l' I -15 dominating factor? Then we look at all those events, said, V

16 did we learn something new which we did not know before 17 1985, and which was not reported before 1985? Mechanism, 18 failure mode, location.

19 Then we said that each of.these leak events have a 20 potential for electively rapid growth. And then what is the 21 safety significance of thermal fatigue? We identified 8 22 thermal fatigue separately because it appears that thermal 23 fatigue is evolving as a new concern during this time period 24 for'the reactor coolant system piping.

l 25 Now, I think you ask us a question, what did we p.

I i

l

(- [~} ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

115 l

l 1 learn from the. review of the industry? We did all this work 2 and I think we had one comment from NEI that this data can

, Y -

3 be very useful in a risk-informed inspection. We have all 4 the-data here but we nave never put together to address the 5 risk-informed inspection. So we reorganize this information 6 and identified what is relevant to the risk-informed 7 inspection programs.

8 And-then we say,.which leak events may be 9' considered as a core damage precursor? And, finally, how 10 good, how effective are the current leak detection system?

11 And I will give you the important finding for each of these 12 items.

13. Here are the distribution of leak event, 199 leak 14 events which took place from January 1985 to third quarter

-() 15 of 1996. Now, you will see that more than half the leak 16 events --

17 DR. WALLIS: Excuse me. Those are reportable' leak

-18 events?.

19 MR. SHAH: All the -- we are going to talk about 20 only reportable leak events.

21 DR. WALLIS: So I will say a 1 GPM leak --

22 MR. SHAH: No , it can --

23 DR. WALLIS: How long has it been developing 24 before it. reaches 1 GPM?

25 MR. SHAH: Okay. That is a good question. That r

, '[N ANN RILEY & ASSOCIATES, LTD.

' ( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

116 1 is question asked by many reviewers. How long it took

()

Lj 2 before they detect? Because nobody knows when it started.

3 You see, we detected, so we know that there is a leak, but 4 there is no way to know when the leak was started. Okay.

5 So I think it will be -- it will be difficult to answer that 6 question. Okay. Unless we run a test.

1 7 DR. WALLIS: You could have a leak which was not 8 reportable yet, and people are watching it until it gets 9 reportable.

10 MR. SHAH: I think what we found is opposite.

11 What we found is opposite, many leaks were smaller than 12 exiting the technical specification, and they are reported 13 because there was a concern that it may grow to the limit 14 and they report it right away. Okay. Yes.

15 DR. KRESS: I was just wondering how you went from (N-) i i

16 the left side of that curve to the right side.

17 MR. SHAH: I will -- just a second, I will do 18 that. This is -- I am talking about just left side, left 19 part of the curve.

20 DR. KRESS: And you divide it by something to get 21 the --

22 MR. SHAH: I will tell you in a second. So we 23 wanted to know -- it looks like there is a decreasing trend, 24 but we do not know how valid it is, because the number of 25 operating years were different. Because during this four

/ ANN RILEY & ASSOCIATES, LTD.

(._3,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

l l

l 117 1 ~ years, quite a few new PWR got a license. So what we did, 2 we calculated the number of operating years per calendar l- 3 year, and we took this frequency -- this events, divided by 4 those years and got this frequency, which is on the 5 trighthand side. Okay. And so now this data is normalized

! 6 and we are going to analyze this normalized data on the next l

l 7 slide. Okay. Because we want to know whether this is 8 statistically significant or not, or is it just because of l

the randomness of the data we see this.

9 10 Let me go briefly to what kind of statistical 11 measures we use. We assume the Poisson distribution and we 12 assume -- we took the very simple model, assume the Poisson 13 distribution, and then took this log linear model where the 14 rate is equal to exponential of A plus BT, where A and B are (s/b 15 the parameter. i

, 1 16 We used the maximum likelihood method to estimate 17 the parameters, then we say how good is this assumption. We 18 made an assumption; is it good or bad? So we did'the -- we 19 examined the goodness of fit. If this one says this is an 20 acceptable model, then we can go further. And then we said 21 that this is the -- B is the slope, and so how significant 22 is it? Is it a statistically significant number? And so we 23 ' calculated the p-value, probability of the significance, to 24 estimate the significance of the trend, and I will show you

.25 those data, i

l-

' /~T ANN RILEY & ASSOCIATES, LTD.

• \m ,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. . . __ ._.m _ _ . _ _ . . . . ~ _ _ _ _ . . _ _ . _ . . _ , _ . . . . _ . _ _ _ _ _ _ . . - _ _ _ _

.118 1 DR. FONTANA: Does that uncover different effects?

2 Like-for example, I was reading your report. It said that 3 . valve packings --

)

4 MR.1 SHAH: I will show you -- I will tell you the -!

5 trend and then we will go to the reason why that trend is

'6 there.

7 DR. FONTANA: Okay.

8 MR. SHAH: At'that time, I will show you.

9' DR. FONTANA- Okay. '

10 MR. SHAH: lOkay. So here are the same frequency- l 11 data, this red dot. So the first thing we did, we.

12 calculated the confidence interval. The larger the' 13 interval, more uncertainty;: smaller the interval, less )

14 . uncertainty. '

(A) 15 If you have a large number of operating year in a 16- given calendar year, then the uncertainty will be small'er, 17 okay? Then we did the' maximum likelihood of frequency. We 18 estimated this-parameter A and B using the maximum 19 likelihood approach, and we got this particular slope.

20 Then we calculated -- the first thing we did, to 21 determine how good is this fit. Is this model good enough.

22 So we did goodness of fit test and goodness of fit test said 23 that we have used the Poisson distribution. We have 124 ' exponential decay and we see this scatter. What is the 25 probability of seeing that scatter? If that probability is

' ANN RILEY & ASSOCIATES, LTD.

f)N, s,, . Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

1 119 1 greater than five percent,.this is the acceptable model. If ,

l 2 it is greater than ten percent, then that's very good. '

l 3 In all the cases, we found greater than five l

!. 4 percent probability, and in all but three cases, we found f.

5 greater than ten. percent. So the model we have used is 6 acceptable. But there are some questions and I will show l-l 7 you -- some deficiency -- I will show you later on.

m 8 DR. KRESS: Can you go back to your previous slide ,

9 just a second?

I

10 MR. SHAH

Yes, sir. l 11 DR. KRESS: You answered my --

12 MR. SHAH: This one -- )

13 DR. KRESS: Yes, r

I 14- .MR. SHAH: -- or the previous one?

f '15 DR. KRESS: No , the one before that one.

, 16. MR. SHAH: Yes, sir.

t I l

17 DR. KRESS: Or the one -- yes.

l 18 MR. SHAH: Okay.

19 DR. KRESS: If I look at, say, 1985, I get roughly 20' 29 leak events, and to get the -- what looks like about .8 21 or something on the other side, you would divide that 29 by 22 something like 40. Does that say there were only 40 reactor

'23 operating years for that year?

24 MR. SHAH: Well, that's certainly right. Every 25~ operating --

l' 5

l

[~ ) ANN RILEY & ASSOCIATES, LTD.

!(J Court Reporters l 1025 Connecticut Avenue, NW,. Suite 1014 Washington, D.C. 20036-(202) 842-0034

120 1- DR. KRESS: Forty out of the 100 reactors --

2 .MR. SHAH: No. There were totally 75 PWRs.

3 DR. KRESS: There were only 75 in '85, and these i

4 are all PWRs. Okay.

i 5- MR. SHAH: I'm sorry. Now, during these four 6 years, about 16 PWRs got licensed in these four years. So

~7 in 1985, it was a much smaller size. Now, this is operating

8 year, so you take the number of PWR plants mul1

iplied by .75 9 approximately --

10 DR. KRESS: Because a number were shut down?

11 MR. SHAH: Yes. So it was about .75, and that 1

i 12 will come to -- the number really is from 40 to 60, and in j 13 my report, it is the figure 2-2 or 2-3, we have given the

( .14 distribution of those years.

'%J

[\ 15 DR. KRESS: I'll take a look at it.

16 MR. SHAH: Sure.

l 17 DR. POWERS: I'm not exactly sure of your words, 18 When you say a 90 percent confidence band on the fitted l

19 rate, is that -- are those bands indicative of the 20 uncertainties that you have in the values of the parameters

.21 or are the uncertainty that you would have in the estimate i 22 derived from the fit?

23 MR. SHAH: Do you mean this confidence band?

24 DR. POWERS: Those dashed lines there.

l >

25 MR. SHAH: They are the confidence band on the l

' ("N ANN RILEY & ASSOCIATES, LTD.

( -

Court Reporters 1025 Connecticut Avenue, NW, Suite 2014

Washington, D.C. 20036 L. (202) 842-0034 l

L

y I

i 121 1 rate. 1 1

2 DR. POWERS: So they're based on the uncertainty l

3 in the value of B derived from the fitting process?

4 MR. SHAH: That's right.

5 DR. POWERS: Okay. They are not the uncertainty 6 that you would have if you went to estimate what the events 7 per operating year was in, say, for instance, 1990?

8 DR. WALLIS: That's the error bars.

9 DR. POWERS: No, that's on the -- the error bars 10 come from the data, as I understand it. But if I use this l

11- correlation to estimate what that value would be, I would 12 have different dashed lines.

13 MR. SHAH: I should mention one thing, that Dr.

14 Cory Atwood worked with me on an aspect of this, and I think l

[% )l 15 I --

I have to check with him. I do not have an answer for 16 you.

17' DR. POWERS: I think when you make your final 18 version of the report, you need to make that clear because i-19 it's very confusing right now on what exact.ly it is that 20 you're -- these dashed lines represent. And I think you're

[ 21 right, I think they represent the band you get based on the 22 uncertainty in the B value. But you never show the -- I 23 mean, for instance, on the slide, you don't show me a 24 standard deviation or anything about that B value.

25 MR. SHAH: We have that --

l

.Ci ANN RILEY & ASSOCIATES, LTD.

( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i

.- - - -. - . - -.. .. --- .~.. .- - . . . . . - . . - - - . _ - . . . .-.

i 122

1 DR. POWERS

So my initial reaction to them was, l

' 7~'s 2 oh, this is the uncertainty that I would have in the value I

k. .

3 estimated with this correlation. But I think it is not j

i 4 .bect.use it doesn't behave like it in --

5- MR. SHAH: So in our final report, we will put  !

i  :

6 down how we calculated it and what are the standard l

1

-7 deviations.

8 DR. WALLIS: I understand uncertainties in B; I L 9 don't quite understand uncertainties in the actual number of 10 . leaks.

. 11 MR. SHAH: I did not understand the question.

12 DR. WALLIS: Maybe I'm misunderstanding the error 13 bars on the actual data points.

L 14 MR. SHAH: Oh, these error bars, these are the

'I\ '

15 --we call it a confidence interval, a 90 percent confidence Nd 16 interval.

17 DR. WALLIS: That there were that many leaks?.

18 MR. SHAH: Means that this is --

19- DR. WALLIS: That's a data point. That's so many 20 leaks, right?

21 MR. SHAH

Yes. This is the frequency.

22 DR. WALLIS: You're saying plus or minus? Why?

23 MR. SHAH: Oh. Because of uncertainty, because 12 4 - our database is small, the number of reactor --

25 DR. WALLIS: You mean there were real leaks which

/' ANN RILEY & ASSOCIATES, LTD.

!.! Court Reporters 1025-Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

,..~.- .. _.. .- - - - . - _ . - . . . - - . _ . ~ . . - . . ~ . - - . . .

123 1 weren't detected or there were some reported as leaks which 2 weren't leaks or something?

l 3 MR. SHAH: No. No. This is just from -- because l' 4 -- this one is mainly.because of the Poisson distribution.

i

( .5 Now -- okay. In this time period, we have many leaks, okay?

I 6 In this one year, there were several leaks. This is the l

7 number of leaks in calendar year 1987 divided by the number.

t l

8 of operating years, reactor operating years.

l.

l 9 DR. WALLIS: Just the red dot.

i 10 MR. SHAH: That's the red dot, okay? Then this 11 one is -- because -- we assumed the Poisson distribution, so 12 that will give us what is the standard deviation, and'that i

L 13 -- so you take the mean plus or minus the standard deviation 14 divided by the square root of the number'of operating years, 15 and that gives'you this error band.

L 16 That'says that if you have a large -- okay. What 17 this error -- this interval says that if you have a large 18' difference -- many different set of -- data set, and then

[

l 19 you calculate the 90 percent interval, 90 percent of the p

20- time, your true value will lie in these limits, okay?

1 L

21 DR. WALLIS: It doesn't matter. Let's go on.

! 22 MR. SHAH: Okay. And we calculated this parameter 23 '"b," and that is the slope of that model, and the p-value.

24 What the p-value says, that if I have a large set of data 25 and the trend is zero,'then what is the probability of 4

• Q ANN RILEY & ASSOCIATES, LTD.

\)

% Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036

.(202) 842-0034 i-l

. _. ~ .. - . . _ . . . . . . . _ _ , _ . - . ...___ _ ..- _ _ ._ . _ _ _ _ _ . _ _ _ . _ . . . _.

124 1 seeing this trend. If that probability is less than 5

(' 'i 2 . percent, . customarily 5 percent, if there is less than 5

(

3

~

percent, then this.is a statistically significant trend. In 4 our case, we get that probability is equal to 0001. If this 5 probability is greater than 5 percent, that means it is 1

6 possible that the trend is zero, okay?

l J

7 Now so the slope "b" gives you the magnitude of 8 the trend. If the "b" is equal to minus .1, then in 12 9 years this value will reduce byfa factor of 3.3. If it is 10 minus .2, then in 12 years it will reduce by a factor of 11.

11 So this one gives you the magnitude of the trend, while the 1

12 p-value gives you the strength, how statistically strong i

13- this evidence is, i-I 14 DR. FONTANA: I hate to show my ignorance, but I) 15 aren't you adding up apples and oranges, because the valve 16 packings increased with time and the other failures didn't?

17 MR.-SHAH: But this is all averaged together. So I i

18 now the next question is -- this is my next slide. Why do I 19 have a decrease in the leak event? What caused the 20 decrease?

21 DR. FONTANA: Okay.

22 MR. SHAH: Okay? So we want to know why do I have L 23 a decreasing' trend, what caused that.

24 DR. FONTANA: Yes, at this point, make believe you i

L 25 don't know.

I-e ANN RILEY & ASSOCIATES, LTD.

,\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. -- _. . . , .~. - .- . - . . - . -. .-- . . - = ..

125 1 MR. SHAH: I'm sorry?

(~N 2 DR. WALLIS: He's going to tell us. Are you going O.)

3 to tell us?

4 MR. SHAH: Yes, I will tell you everything, one by ,

5 one. But this is the whole picture. So we saw this 6 picture, so we wanted to learn what caused this, and that 7 comes to your next slide.

(

l 8' Here. So we just took the -- there were 29 events 9 that would cause where the leakage was through the valve 10' pack, okay? So we look at those 29 events, and on the l-11' left-hand side this is by the calendar year, and these are 12 the frequency only for those valve-packing leak events.

13 Okay? And we plotted that. Remember that there were no l -14 valve-packing leaks reported after 1991. Now we checked the

[Y \

15 requirements. Reporting requirements were not changed. But l 16 during this time period there was quite a design change.

17 DR. WALLIS: Yes, the valve packings got better.

l 18 MR. SHAH: Much -- no, no. Not only better, there l.

19 was a significant learning --

!~

20 DR. WALLIS: Yes.

l 21 MR. SHAH: And significant design changes. There 22 were a graphite base, there were asbestos-based packing,

[ 23 they.were replaced by the graphite-based packing ring. In l

L 24 Edison they were using the deep stuffing box. Instead of 25 that they went to the shallow stuffing box. In Edison they l

t

' /\ ANN RILEY & ASSOCIATES, LTD.

(,,)

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034 L

i 126 l 1 used the live loading, so you have a thing that's

-F'\

,r i q) 2 continuously packed because of that, L

i 3 DR. WALLIS: So we-can accept that this problem L 4 has been solved. i l.

5 MR. SHAH: Yes. And we wanted to know that --

6 what we tried to do, we tried -- we solved the -- because of l -

7 the -- it is a function of the age, or was it a function of l

8 the calendar time? And so here what we did, we considered 9 how many plants were one year old during our study period, 10 and two years old and three years old, and then say how many 11 events 'ook c pluce in that time. And we calculated a 12 frequency, and we made that plot.

13 What we found was that we had a statistically 14 decreasing trend in both cases. So we do not know'which is

(,.-~I 15 more dominating, okay? What we have to do is to -- we have wi  ;

16 to have a new model where we have two indepenomnt 17 parameters, the calendar time and age, and do the analysis

[ 18 again, and then determine-which one is more significant.

19 But right now we can only say that it appears a decreasing

( 20 trend with both calendar time and age.

21 Here we have a leak event associated with a wide 22 break. Then we did the same thing that we did with the l 23 valve packing. According to calendar time, here we lump j 24 every five-year group.

25 DR. WALLIS: Now could you tell me, the previous ANN RILEY & ASSOCIATES, LTD.

j Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034 l

l-(

l

127

.1 .ones were packings --

t.

lp 2 MR. SHAH: -Yes.

!  %._)

3 DR. WALLIS: Which you expect might leak. These 4 ones here are actual holes in pipes?

5 'MR. SHAH: They rupture. Many times they'll fail.

6 DR '. WALLIS

To me that's the only thing that 7 matters here, that we can forget about the packings --

8 MR. SHAH: Yes.

9 DR. WALLIS: Let's talk about these ones.

10 MR. SHAH: Okay. What is your question, sir?

11 DR. WALLIS: Let's forget about the packings.

12 MR. SHAH: Yes. Okay.

13 DR. WALLIS: We fixed that one.

14 MR. SHAH: Yes,'we fixed that one.

() 15 16 the pipe.

DR. WALLIS: Now let's worry about the holes in 17 MR. SRAH: Yes. Okay. So we are talking about i 18 that. And here on the left-hand side we had a trend with 19 respect to calendar year. On the right-hand side we are a 20 trend with respect to age.

21 What we find that the trend with respect to 22' calendar year, we have a slope "b" equal to minus .08, but 23 it is statistically not significant. This p-value is much 24~ larger than 5 percent.

L 25 The trend with respect to age is statistically i ("'s ANN RILEY'& ASSOCIATES, LTD.

lts,,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036

(202) 842-0031 j'  !

L

L 128 1 -quite significant. So what we are seeing here that the

('N- 2 leakage' caused by vibratory fatigue failure the number of 3 leaks are decreasing and the main reason for this reason is

4 premature aging.

5 If you look at the curve, it is'the front portion 6 of the curve which we are seeing here.

7 DR. WALLIS: It's not because somebody' fixed the 8 vibrations.

l 9 MR. SHAH: Yes. Because when they find the leak, 10 they do two different things. One, if the weld is a 11 fabrication error in the socket, most of_- .all these 12 failures are in the socket weld. They make a better weld, 13 and they inspect:it. Or they make_a butt weld. So that

! 14 that weld is no longer susceptible to cracking.

i

() 15 16 In addition, many of the pipings are not well supported. Some of these failures are in cantilever piping.

17E So they put more support on the piping, they reduce the mass l 18' of the valve, so that the frequency of the pipes -- piping 19 system is increased. And therefore we see the decreasing L 20 _ trend. Okay?

21 So what we try to say, this failure, where, 22 because of premature aging, and it is decreasing.

23 DR. WALLIS: It's decreasing because they l 2 41 learned --

25 MR. SHAH: They learn and they --

1 i

G ANN RILEY & ASSOCIATES, LTD.

f k,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014

' Washington, D.C. 20036 (202) 842-0034 L

129 l

1 MR. ROSENTHAL: Let me point out that this is an 2 important observation with respect to aging, because the

.3 first blush is te just say okay, I could expect over the 4 ~ years of service of my machine that the number of fatigue 5; cycles goes up and as I go out in time I could expect more 6 vibration-induced failures. That would be the first base 7 flush.

8 When you look at the data and you understand that j

i 9 what was -- what Vik calls premature aging, where there was

( 10 a problem that was fixed, and now it looks like the problem 11 is fixed, implies that as you continue to run these plants 1

12 out in time, you don't expect to get vibration-induced I

13 failures. So it has I think an important observation with 14 respect to our concerns about aging of the plants.

p 15 DR. KRESS:

_t ] . Do you think that concept could be 16- extrapolated on out for 60 years? Somewhere it ought to 17- turn back around.

18 DR. SRACK: No, I mean, that trend isn't going 19 down because fatigue gets better with cycles of age --

'20' MR. SEAH: No, they are better prepared.

21 DR. SHACK: The stuff that hasn't failed yet, the 22 cumulative damage is building up. It's getting worse with l 23 age.

24 DR. KRESS: Just hasn't showed up yet.

25 DR. BARTON: Right.

/9 ANN RILEY & ASSOCIATES, LTD.

(/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014

Washington, D.C. 20036 (202) 842-0034 V

i E

l i

130 1 MR. SHAH: Well, what.happens? Generally the 2 vibratory fatigue does not take place all the time, only 3- during certain type of cperations. So that accumulates, l' 4 okay? In some of the operations maybe you are accumulating l 5 slowly. But once you see-that and know how to fix it, then 6 you fix it and it does not happen again. Okay?

7 Here is another aging event where the rate is 8 increasing.

9_ Let me tell you, this is different than what we t

10 saw up till now. Here we are trying to determine the trend 11 of thermal fatigue events.

l 12 Now there were only four thermal fatigue events in 13 this country so it was a very small -- a little bit -- so we 14 went to a worldwide database.  ;

lV) 15 Okay. This is the'first time ~it is a non-U.S.

16 experience included. There's a total of 13 events. Out of

~17 13, four are in this country, three in France,-two in 18 Finland, one in Japan, two in Germany and.one in Belgium.

19 That makes a total of 13.

20 We analyzed them by age. We grouped that in five 21 years'and you can see that there is an increasing trend and 22 we have a strong statistical significance and this trend is 23 valid.

24 DR. WALLIS: It's a bit awkward to expand that to 25 an age which no reactor has at all, like 50 years --

O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036

-(202) 842-0034

- - - - - - .-.- ... - ~ _ - - ~ _ = . . - . . .

131 1 MR. SHAH: This is zero -- because there are only O- 2- 43 reactor years.

k.j '

3 DR. WALLIS: Cut off --

4- MR. SHAH: And also~you can see we have a large 5 uncertainty there so this data has to be reanalyzed in the' 6 future as we get more data.

l

, 7 DR. WALLIS: Well, also this problem isn't being i

c 8 fixed. I 9 MR. SHAH: We will talk about that, okay? -- but 10 we do not understand yet, so we cannot fix it.

l l

t

.11 DR. WALLIS: 'Let's go back to a thermal fatigue l 12 diagnosis.

l 13 MR. SHAH: Yes, sir.

l l

l 14 DR. WALLIS: Is it when you get a leak which

[

i s-)

15 doesn't have any other explanation that someone thinks it I i

11 6 .might be due to thermal fatigue or --

17 MR. SHAH: No.

l p 18 DR. WALLIS: -- or is there actual evidence of L19 - thermal cycling? I l 20 MR. SHAH: All this time there is a destructive l

21_ analysis performed. We know -- because that can be stress 22 corrosion cracking, so it has been checked out and you can 23 see the striation --

L 24 DR. WALLIS: Okay.

25 MR. SHAH: -- and you know what it is --

. /7 ANN RILEY & ASSOCIATES,.LTD.

(_j' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington,'D.C. 20036 (202) 842-0034

. - - _ ~ . . - . - ~ - . . . . _ . ~ .. - . _ . . ~. _ _ - - ... .- ~ -..-.- . _-

l 132 1 DR..WALLIS: All right.

jN[ 2 MR. SHAH: Sometimes it can be a mechanical KJ 3 problem also, okay?

4- Now we want to see what are the new things 5- - reported which were not reported before 1985 -- failure mode locations. Let's talk about that.

I 7 Again we will'go back to thermal fatigue in one of '

.8~ the phenomena which was not that well understood before l- 9 1985.

10~ In 1987 we had a crack, a leakage at Farley. This 11 is the' safety injection line and then isolation valve was 12 leaking and, okay, let-me go back.

13 This is a three-loop Westinghouse plant. The main 14 characteristic of three-loop Westinghouse plant which caused s~~.'.

Ef )

> %_).

15 this problem.is that safety injection' system and the ,

l

\

16 charging system, they both are operated through the same i

17- line and so the safety system is operating at 2250. Okay, 18: so this is the safety injection line and there was a leakage 19 and that leaked through, the cold water leaked through the 20 check valve -- and it cracked here. 4 21 Because of that cracking, six months later there 22 was a similar cracking in Tihange, which is another 23 three-loop plant, so initially there was quite a bit of work 24 done in Japan.to understand what caused that.

25 Then EPRI in this country sponsored a big program

' a("')

s _/

ANN RILEY & ASSOCIATES, LTD.

Court Reporters

,. -1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036

(202) 842-0034

,. . . - -. - . _ . - . _ . . . - - . - . - . ~ . ~ . - ~ - - . - . - - . . - .

133 1- for five years to understand what caused this problem.

' /7% 2 On this sketch they have tried to explain what is D. 3: happening. Here is this check valve and we have got a leak 4-

'and there is a stratified hot water and cold water and here

'S .by cold leg, the main coolant piping and the main coolant 6 flow is going through it and it is turbulent and that

.7 - turbulence penetrates this branch line and this length of 8 the turbulence penetration changes. It is a stochastic 9 parameter. It is not a fixed parameter, so it is cycling 10 axially like this.

11 It interacts with the stratified layer and because I

l' 12 its length changes, this portion of the pipe sometimes 13 ' exists as stratified' layer, sometimes it does not see the j 14. ~ stratified layer. If there is a stratified layer, you have i s-LI 15: got.a thermal stress'and when it is not you don't see that, X ,

16 -so you have cyclic stresses.

17 Those-cyclic stresses causes the cracks so this is l 18- well understood. It can: explain why we have th'e cracks in

-19 the RHR piping in Japan's plant -- also~this phenomena, this 20 understanding can also explain why sometimes we have cracks l 21 in the small diameter pipes connected to the main coolant '

l '22 piping though there is no leakage, in the absence of L 23 leakage. This phenomena can explain it.

H24 So we have now a sufficient understanding to know

.25 under which conditions this may not be present, under which

't ANN RILEY & ASSOCIATES, LTD.

I s,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036

.(202) 842-0034 L

c u .

. -. - . - - - - . ~ . - . . . . - . - . . . . - - . - . - . - - . . . ~ . . - - . . - - . - . . . .

l t

I 134 1 condition it will be present. However, our understanding is not aufficient'enough or precise enough'so we can predict at

~

2 i- 3 which location it may cause cracking, okay?

I 4 So right now we have applied this logic to the l 5' Farley plant and we have cracks at this location but these l:

l' (61 models, they cannot predict this location where the crack 7 may take place, so our models are still not precise enough

-8 and we do not know all the thermal hydraulic phenomena 9 taking place, but our qualitative understanding is quite p '10 good enough so that we know where it will take and where it

.11 may not take.

u i 12 For example, if this' wall is about 25 diameters 13 from this piping, then it will not happen because by that 14 time this cold water will be -- far enough -- and this

( 15 thermal penetration, the turbulent penetration does not 16 reach all the way to the wall, so if there is a certain-17 distance within.the piping and the wall, then this phenomena i

,:U3 : will not happen, so we have a qualitative understanding. We 191 don't have a complete quantitative understanding.

L -20 But one more thing we learned here -- that this l: 21 phenomena may cause cracking in the base metal, like here is i

22 the place where the leakage took place to the elbow. Here

[ 23 is Dampierre -- one -- in 1996. The leakage took place p 24- here, far away from the weld.

25 In this country we do not inspect the base metal.

f i

ANN RILEY & ASSOCIATES, LTD.

(_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 L

t (202) 842-0034 l

_ ._ _ . . . _ __ ._. . . . . . - _ _ _ - = __ __-_. _ __ .

1 135 l l

l 1 It is not required to be inspected. However, we have r

2 leakage through there. In France they have to inspect the l

3 whole length of the pipe, both base metal and the weld, 4 because of this leakage.

l 5 Another interesting thing happened.

l 6 DR. FONTANA: Excuse me. We don't inspect -- but

) 7 wouldn't you see it underneath the insulation?

8 MR. SHAH: I'm sorry. The ASME code requires the 1

-9 inspection of the weld. We have to remove the insulation to 10 inspect that but it does not require the inspection of the 11 base metal. Now this crack starts on the inside surface so 12 we have to do the volumetric inspection to find these

! 13 cracks. 1 l

14 DR. FONTANA:

l r's If it goes all the way through and- ]

l 'l l( _j 15 . leaks, then.you can see it?

l 16 DR. WALLIS: You are talking about cracks rather 17 than leaks. I 18 MR. BARTON: Yes, definitely.

19 DR. WALLIS: You are talking about leaks though.

201 Are you talking about cracks which are not --

21' MR. BARTON: Cracks that end up as leaks.

22 MR. SHAH: In all the thermal fatigue it is a crack -- the coolant leaks out through the cracks.

! 24 DR. WALLIS: You are talking about detected leaks.

25 MR. SHAH: Yes.

i ANN RILEY & ASSOCIATES, LTD.

. (.

/)N, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i

I _,,

. . .-. . . ~ ... .. ~ -. . - - _ - - - - ~ . - - - - . ~ . - -

136 l 1 DR. WALLIS: You are not talking about detected 1

.2' thermal fatigue which has not yet leaked, detected by

.3 inspection?

4' MR. SHAH: Right now, ASME Section 11, our purpose 5' isEnot to find the leaks but to find the cracks before they 6 become through-wall.

'7 In this country ASME code requires that you 8 'inspectithe welds. They don't require that you inspect some

~9 of the base metal.

-10' DR. WALLIS: But do you have statistics on the p 11 cracks which did not -- are not yet leaks?

l '12 MR. SHAH: We have some and we'll talk -- okay. I l I- 13 . don't have statistics -- I have got a few examples, i l

l 14 But in this case let me tell you.what happened. -j f f 15 In this case, EDF replaced this pipe in 1996 and they 16' . analyzed the pipe for all the~known thermal loads, learning l

l 17 what kind ^of phenomena ~took place, transient took place and 11B they installed a new pipe and what they found after eight 19 months, that pipe was cracked for two-thirds'through the 20' wall and they were not expecting any cracking for a long l 21 time, and here you have a very fast initiation and the crack 22 growth so that is one datapoint and I think that I have 23 included in one of the safety concerns for the thermal l 24 , fatigue.

/25 Did I answer you, answer your question?

l t

/~% ANN RILEY & ASSOCIATES, LTD.

k,)

Court Reporters

, 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

l 137 1 DR. WALLIS: I want to know the difference between f (^% 2 cracks and leaks. I mean your graph here says leak events,  ;

t td 3 and'then it says cracks per year. Do you really mean by ,

4 that.-- do you mean leak graph?

5 MR. SHAH: Okay. These are the leak events.

6 DR. WALLIS: Right.

7 MR. SHAH: But what I say that we want to -- we 8 will never detect this. We will never find it because we 9 never look for it. The point is --

l 10- DR. WALLIS: Wherever you have cold water on one-11 side of a valve and hot water on the other side, you had 12 'better start looking for leaks.

13 MR. SHAH: That is what I said. That is what EDF 14 is.doing, they are inspecting the whole length of the pipe.

lj ~-15 But what I am trying to say is that right now, in this

s.

16 country, this kind of' inspection is not yet required. Also, 17 in this country, we do not have a leakage through the base

! 18 . metal either. Okay. Yes, sir.

[ 19 DR. WALLIS: Not yet.

20 MR. SHAH: Yeah, but, no, this is just same as the 21 Westinghouse three loop plant, these are the EDF plants, but

'22 they are the three loop plant, like 900 megawatt, so it is 23 the same as the Farley or same as at Tihange plant.

L 24 DR. UHRIG: So we just wait until it leaks?

i-L25 MR. SHAH: . Well, I think we don't have a choice at rT ANN RILEY & ASSOCIATES, LTD.

< (_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014

l. Washington', D.C. 20036 l

(202) 842-0034 l'

138 1 that time. But --

2 DR. UHRIG: Okay.

3 MR. SHAH: Here is another event that is a i 4 cavitation-induced vibratory fatigue. Generally, we know 5 quite well about the pump-induced vibration which caused the 6 vibratory fatigue failure. Okay. But here in this letdown 7 system is -- what happens, there is a significant change 8 decrease in the pressure and temperature of the coolant so 9 that if the presst.re goes below the vapor pressure, then you 10 will form the vapor bubble and collapse it. And that, it is 11 so -- it happens so rapidly it causes the cavitation and the 12 vibratory fatigue.

13 The main difference between this vibratory fatigue  !

14 cavitation-induced and the pump-induced, that in a

() 15 pump-induced your frequencies are discrete. So if you match 16 one of the discrete frequency with the piping system i

17 frequency, you may get a cracking. In a cavitation-induced I i

18 vibration, you have got a frequency band and so it is very l

19 likely that you.will match one of the frequency with this 20 excitation frequency.

21. In addition, if the temperature of this letdown 22 coolant is higher, if it is not sufficiently cooled down in 23 the region that is heat exchanger and it is higher than its

24 saturation temperature, you may form the steam bubble. And 25 then, again, when you restart your letdown system, the cool ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

139 1 water will cause the collapse of these steam bubble and you

(~') 2 will have a water hammer. And in here, -- this happened in

'v/

3 McGuire in 1988, that the pipe was already cracked because 4 of the cavitation-induced vibration, and then water hammer 5 took place, it caused the rupture.

6 Now, I think that this kind of events did take 7 place before 1985, but there was never a root cause analysis 8 to identify that, and that is what Duke Power did around 9 1988.

10 Now, here is another phenomena and that is a 11 primary water stress corrosion cracking and, generally, we 12 know that in the steam generator tubes this mechanism has 13 caused quite a few damage and quite a bit leakage. But this 14 mechanism has caused like a pressurizer instrument nozzle, (o

J

) 15 pressurizer heater sleeves, and also in the penetration in 16 the primary coolant loop, it has caused cracking during this 17 time period which was not reported before.

18 Oh, now we come to the third. finding. The leaks 19 that have a potential for relatively rapid growth.

20 DR. WALLIS: So what does that mean?

21 MR. SHAH: It means you can have a suddenly 22 enlarged leakage.

23 DR. WALLIS: Does that mean a leak that grows in a 24 day or a year.or --

25 MR. SHAH: No, no, much faster than that, i

f') ANN RILEY & ASSOCIATES, LTD.

, \s / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l

-. - - _ - . - . - _ . - - - - . ~ . - . - ~ _ - . .. - - - . . . _ . . - . _ . . . . - . . . . - -

t 1

140 1 DR. WALLIS: Seconds?-

2 MR. SHAH: Yeah. We are talking about that. Or i 3 maybe instantaneous. Much faster. And there are some

! 4- potential for that, and I would like to talk about that.

5 ~Okay.

l -6 There are leaks, I mean through the thermal 7 fatigue cracks in a branch line could lead to SBLOCA. One 8 example is one of the pipe in a Civaux plant in France in RHR line.

1 9 The thermal fatigue cause cracking-and we had a

{

10 leakage of 132 gallon per minute. Now, this crack took 11 place in isolable portion of the piping. It was not in a 12 non-isolable portion of the piping. But similar leak event 13 can take place in a non-isolable portion of the piping and 14 then we will have a concern for the SBLOCA, because some

() 15 ' plants charging system may not be large enough to handle 16 this leakage and it may help to use the safety injection 17 lines.

18 Also, as I mentioned to you, that in Dampierre we 19 had a very high crack growth rate. Most of the fatigue 20 . cracks we came across -- the crack growth rate was quite i 21 low, quite low. But here, we had a much high crack growth 22- rate because -- it could have become a true wall crack in 23 less than a cycle and our experience with this fast crack 24 growth rate is quite limited, and we do not know, under 25 certain conditions it can become a large leakage.

ANN RILEY & ASSOCIATES, LTD.

\s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

l 141 1 This one indicates the low crack initiation time, 2 high crack growth rate, and a large crack size. Also, 3 another concern we have, that uniform growth -- if the 4 particular. crack is circumferential, all the way around the 5 pipe circumference, and if it grows uniformly, then there is 6 a concern that it may lead.to a long through-wall crack and

! '7 we can have a large leakage.

8 Let me give you an example. This is 1997 event'at L 9 Oconee 2 and this is the high pressure injection in the 10' makeup sketch. Here is the cold leg and the crack was --

! .11 this is the safe end and through-wall crack was right here.

12 Okay.

l 13 Here is -- if you cut this, this is a picture of l 14= the crack. Most of the circumference, the crack grew

('

.(

w -

15 uniformly, and-it appears that there may be some local l 16' bending moment acting, maybe because of the local 17 stratification, which caused the growth of the portion of 18- the crack and it became through-wall.

19 In the absence of that kind of bending moment, L.

.20. acting, in the absence of that local bending moment, this 21 crack would have grown uniformly, and it could have led to 22 the much longer crack and much larger leakage. So that, I 23 think this event has been analyzed in the accident sequence 24- precursor program and by AUD, and we get around 5 times 10 25 to-the minus 2 at a conditional core' damage frequency for L.

ANN RILEY L ASSOCIATES, LTD.

s Court Reporters 1025 Ccnnecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

l 142 1 this kind of event.

l (~T 2 What they did, they used the SKI data, the li_

j 3 Scandinavian data, where there were about 13 events where 4 the fatigue caused the leakage in 7. stainless steel piping 5 with a diameter from one inch to four inch, and based on no 6 rupture. And based on that data, using the -- statistics, 7 they came to that frequency number, or probability number.

8 DR. POWERS: When you say the SKI data, is this l

l 9 the SKI report on leaks?

10 MR. SHAH: Yes, sir, '97 something.

11 DR. POWERS: But they are leaks in U.S. plants, 12 right?

13 MR. SHAH: Okay. I told you had four leaks in l 14 this country. Our database, I think the SKI data, it does n)

( 15 not limit to the reactor coolant system. Our data, my data 16 here is only for PWR reactor coolant system. SKI data looks 17 at all stainless steel piping in BWR, PWR worldwide.

18 DR. POWERS: Okay.

19 MR. SHAH: Okay.

20 DR. POWERS: They have another report that simply 21 addresses U.S. plants. I mean they did what you did, they j 22 looked at the LERs.

23 MR. SHAH: That's right. So they included U.S.

l 24 plant and other data. But SKI does not separate that these 25 are. primary system and secondary system, they have all the 4

/"']

( )

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l

(202) 842-0034

l l

143 l 1 data together.

"~

2 Another event -- vibratory fatigue failure, l 3 rupture caused by vibratory fatigue failure. It can lead 4 to -- this is instantaneous. This is not a slow process and 5- can lead to a large rate leakage rate.

6J What we found in our study period, that the l

7 . maximum leakage rate was about 87 gallons per minute at the L .8- McGuire 1.

I 9 Here is another event, which took place at Oconee l 10- 3. This is a failure of compression fitting. When you have 11 a very small diameter instrument line as so, you will 12 connect the two lines which are changing in diameter, 13- elevation, so you have to make a connection with a 14 compression fitting. The. main part is this ferrule, so what i

! . 15 do you'do?

16 First, you take your tube and put it on the 17 adaptor so this tube contacts the adaptor and then you  ;

18 tighten the nut, so this ferrule will produce -- because '

7 19 this is tapered here, it will produce a plastic deformation

! 20 and it makes you a strong joint. If this is not aligned 21 properly, then there is not sufficient deformation and so 22 this is not a strong joint and it may come apart and that 23 happened at Oconee and we had a leakage of about 130 gallons 24 per minute, so this kind of rupture can cause, can lead to a l-25 .SB LOCA.

l ANN RILEY & ASSOCIATES, LTD.

D)

'(,, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

144 1 One other event, which I do not have here, is what  !

2 we found at some of the drain lines. They have got a a[~N 3 insulation on it and the chemicals present in the insulation 4 can cause stress corrosion cracking, and what we found at j i t

, 5 one plan this year in January, it caused circumferential i l

i-6  : cracking and that -- there is a possibility that that line

7. can rupture because of the circumferential crack and it will l '8- lead to very small break LOCA. l 4

9 Here is one other event. This is a reactor 10 coolant pump -- that will be true for other pumps, 11 Westinghouse' design -- and if all the four seals fail for 12 any reason, then you can have a large leakage in the 13 containment.

14 Before 1980 there were two events where the pump

) 15 seal failed and there was about a 300 gallon per minute 16 leakage up to 1980, because the manufacturers made many 17' changes and licensees have changed their practices, the 18 maximum leakage is about 10 gallons per minute through the 19 leakage, not larger than that.  !

20 But in one event we came across all four seals 21- failed. That happened at ANL-2 and there was a small 22 leakage in the containment but this event, this failure 23 takes place generally slowly and so the operators are trying 24 to control it and shut down the plant and that is the main 25' reason we do not see large leakages from the pump seal.

ANN RILEY & ASSOCIATES, LTD.

C_)i- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

q i

145 I l

l t

1 Next I want to talk about the safety significance -

.2 of leaks caused by thermal fatigue and some of the things we 3 already talked about --

4- DR. WALLIS: These seal leaks are due to wear?

~

5- Are all of them crack?

l 6 MR. SHAH: There are two different reasons.

7 Most of the seal leaks are because of'some 8 transients. You need a transient in this phenomena. For  ;

l L 9 example, in ANL-2 the bleed-off line failed because of 10' vibratory fatigue, so there was a large leakage, so when the 11.. coolant -- okay.

1 ]

12 Here is the integral heat exchanger and then one j 13 gallon per minute goes to cool the sealc but because this '

1 14 bleed-off line failed there was a 40 gallon per minute

[%) l 15 leakage, so the large amount of leakage, reactor coolant, 16' was flowing through this path, so it was not getting 4

i 17 sufficiently cool. It was supposed to go only one gallon

-18 per minute should be flowing this way, but because this line i 19- failed so there were about 40 gallons per minute coolant was 20 flowing this way, and so that coolant was not cold enough.

21 So-if failed the seals and it already failed these 22 vapor seals. The vapor seals are placed below the  ;

23L containment vessel, so generally if it has not failed you I

~

24 cannot have a leak outside leaking into the containment. In 25 this case.the vapor seal also failed and it started to leak f.

l

/ ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 L (202) 842-0034 y

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

l 146 1 in the containment, so this was more of a transient induced 2 than in some other events.

3 If it is possible, the seal itself fails because

^

4 of some-mechanical design problem, okay, and in that case 5- you can also have a leak, but all the leaks we came across,-

6 they were induced by transient, maybe a component -- this 7' coolant was not coming here. This component cooling water 8 stopped or so, and so this coolant, the reactor coolant 9 system' passing through the seal was not cold enough and that ,

10 caused the problem.

R11 May I answer your questions?

12 DR. WALLIS: Okay.

13 MR. SHAH: So now let us look at, briefly, how the 11 4 thermal fatigue is becoming a safety issue.

( 15 One reason is we are seeing the large leak rate 16 like 132 gallons per minute in the French plant. Also there 17 is our experience with the rapidly _ growing fatigue crack is 18 very limited. We have seen it only at a few places like at 19 Dampierre. The main reason we considered it a safety issue 20- is because we do not know yet, we have not well understood

-21 the thermodynamic phenomena that caused the thermal fatigue 22 cracking'in this plant's' life.

23 Also the number of --

l 24 DR. POWERS: When you say the thermodynamic I

25 phenomena, what exactly are you talking about there?

i i

ANN RILEY & ASSOCIATES, LTD.

h(,, /) Court Reporters

-1025 Connecticut Avenue, NW, Suite 1014 I

Washington, D.C. 20036 (202) 842-0034

l 147 I

1 MR. SHAH: What I mentioned earlier, that we

(~~' 2 qualitatively understand why this cracking is taking place, 1 Q) 3 but we still cannot quantitatively predict where it is, 4 okay? -- so we do not have full control.

5 DR. POWERS: It's really the dynamic. It has 6 nothing to do with --

7 MR. ROSENTHAL: Dana, you could do computational 8 fluid dynamics on the entire primary system, which would 9 make a lot of analysts employed once again, but without 10 knowing the boundary conditions, you really don't know where 11 to do the analysis per section of piping, so I mean in 12 principle you can do it, but practically --

13 DR. POWERS: It's just that it is not an issue of 14 entropy and re-energy. It is an issue of dynamics here.

/m

( ) 15 DR. SHACK: It's more a Reynolds problem than

%.J 16 Gibbs problem.

17 DR. POWERS: Right.

18 MR. SHAH: I think right now Duke Power is doing 19 quite a bit of detailed analysis because of the cracking and 20 they are still working on that.

21 DR. WALLIS: There is fluid flow involved. It's 22 not just thermodynamics in the sense of property 23 relationships. You mean thermal hydraulic there.

24 MR. SHAH: Okay. I should have said thermal 25 hydraulic phenomena, okay. I think that is a better

' (' 'g ANN RILEY & ASSOCIATES, LTD.

( ,) Court Reporters i 1025 Connecticut Avonue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

148 1 terminology.

l

((}

LV 2 Also, as I showed you the number of leak events j' 3 caused by thermal fatigue increasing with age, so as the 4 time passes -- and also we do not know exactly why it is 5 taking. We don't know all the details about this thermal 6 hydraulic phenomena. We do not know all the ways to 7 mitigate or control these events.

8 Also, another concern is that like in Farley where

( ..

, 9 you had a leak through the wall, leakage of cold water, and j 10 now suppose the cracks are already present in the piping and 11 a seismic event takes place. That seismic event may l 12 increase the cold water leakage and that will put more extra 13 -load on the cracked piping.

14 Right now we do not know or we do not have any 15 experimental data for the interaction between the amount of 16 cold water leaking through the wall and the earthquake, so

l. 17 this is one of the outstanding concerns the French safety 18- authorities have for the leakage -- for the cracking and 19 leakage in their plants.

20 Also, ASME Section 11 inspection requirements are 21- not required to detect fatigue cracking mostly in the small 22 diameter piping, smaller than four inch. In Class 1 piping, 1:

23 ASME Section 11 requires inspection of the weld for piping L

l' l

24 diameters. smaller than four inches, but they do not require 25 .the volumetric inspection of that piping, and it appears to

/ ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l

~. . , -

149 1 'be more like an error, not intentional omission, so I think 2- it. appears. It will be fixed shortly.

3 Also the detection of thermal fatigue cracking in 4 the small diameter piping is quite difficult. What we 5 learned from what is reported in France, that the cracks l 6 which are one-third to two-thirds through-wall, they were

-7 not-detected during the inservice inspection, so that is one 8 of the' concerns, and also, as I mentioned to you in l- 9 .Dampierre, the possibility the crack which was not present l l

10 at the.beginning of the cycle could have become a 11 ~through-wall in the cycle. In those kind of fast crack 12 growth rate events, the inservice inspection will be 13 ineffective to catch it.

14 So we did all this analysis and the question is

, 15- now what do we do with the data, and here is the 16 application. We can -- what my reviewers from industry 17 say -- that this~ data can be very useful in the 18 risk-informed inspection, so what we are trying to do is to 19' identify which information'can be supplemented to that 20' . program.

21 -Here are some of the findings which can be

-g 22 Lutilized-in a risk-informed inspection. No. 1, all leakage 23 we found given.the smaller diameter piping, less than 10 24 inches, the largest one was RHR'line, which was a 10-inch 25; line. So there was no leakage.from a larger diameter pipe.

l.

L l

l

! ANN RILEY & ASSOCIATES, LTD.

Court Reporters

! 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036

'(202) 842-0034

[

150 1 Second one, the effect of materials in stainless

(~N 2 steel and the alloy 600 and the main mechanical fatigue, 3 vibratory fatigue, thermal fatigue, and stress corrosion 4 cracking, primary stress corrosion cracking, intergranular 5 stress corrosion cracking, and transgranular stress l 6 corrosion cracking, they were the main mechanical that 1

(

l

7. caused leakage.

8 Vibratory fatigue was the dominant mechanism, I

9 because it caused about 29 leak events. The leakage has

.10 generally occurred through the pipe failure generally. But i 11 in case of primary water stress corrosion cracking and

-12 thermal fatigue, it has also occurred with a base metal.

13 Also, generally the through-wall crack initiated  ;

14 on the inside surface, as I mentioned earlier, that our

) .15 current code does not require the volumetric inspection of

~16 these welds in class 1 piping, piping which is smaller than 17 four inches in diameter, and also one of the problems, one 18 of the difficulties with a risk-informed inspection, where 19- do I inspect? How do I identify the locations which are 20 significant from risk point of view? In the event I have 21 fast-growing fatigue cracks, then it may require that I 22 monitor the pipe, monitoring of the pipe for temperature, 23 and the pipe vibration, to identify which location may be 24 susceptible to cracking. And then I can inspect it.

25- One other point we found, that some of the leaks ANN RILEY & ASSOCIATES, LTD.

')\_, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. . _ _ _ _- .. _. . . _ . _ . _ . _ _ . _ . - . . _ _ _ _ - . . _ . . . . ~ . . _ _ ._

b l

151 {

1 which are not reported, but those leaks took place through a  !

! (\ 2 weld, which is a common header to the safety injection line V

~'

3 or so, and when you have that kind of leak, then it will put )

4 the whole safety system, make it unavailable. So those 5 locations can be included in the inspectior .. the program  ;

i a

6 for risk-informed inspection. '

7 DR. POWERS: When you talk about monitoring weld 8 temperatures and the vibrations of pipes, aren't you

, 9 discussing a fairly significant undertaking?

10 MR. SHAH: I think right now the Bulletin.8808 11 requires that all those pipings which are susceptible, which 12 have the stagnant coolant or so, they do the temperature 13 monitoring of those pipings, because right now we still do 14 not fully understand the phenomenon. So that monitoring is

() 15 required and all the PWR plants are doing it.

16 DR. WALLIS: What do you look for? You see a 17 fluctuating pipe weld temperature, what do you do then?

18 MR. SHAH: Oh, then you inspect that, okay?

19 Because there is a possibility now it may crack. Because 20 otherwise you may want -- then you may want to inspect the 21 whole length of piping because it will not be only a weld, 22 it can be a base metal which may crack. And that is a large 23 ' volume of inspection. So this will help you. And also if 24 it is a fast-growing fatigue crack where it can' grow within 25 a cycle, so if you are continuously monitoring, then that (j)

/ ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

l l

152 1 will give you a warning that at some places -- l l

l 2 DR. WALLIS: You monitor temperature because it's

('")}

\m.

3 easier to do than to inspect. That's why you -- I l

4 DR. FONTANA: Just what do you mean there by 5 risk-informed inspections the way you have it there?

6 MR. SHAH: Oh, in Edison, because right now in 7 risk-informed inspection we say these are the locations 8 where to inspect. But how do I know the locations?

9 Sometime -- and also, okay, here my main concern is my 10 cracks are growing fast, so I need something more than just 11 a risk-informed inspection program to identify if it is 12 taking place within a full cycle.

13 DR. FONTANA: But to my thinking a comprehensive 14 risk-informed inspection program would take into r~N t

%J

) 15 consideration some of the things that you're talking about, 16 unless that particular word has a meaning that I don't 17 understand.

18 MR. SHAH: I think --

19 MR. ROSENTHAL: What he's saying is that the 20 requirements are to inspect piping now, which we don't see 21 cracking, the real big-bore stuff, and that we're not 22 looking at the smaller-bore piping where we are observing 23 cracking. And that if you went away from a deterministic 24 this is what "thou shalt do" to more of a risk-informed 25 approach which would consider the operating experience, then l

(}

N ,/

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l

l

153 1 you would.do less inspection of the big-bore piping and some yN

, ' d 2 inspection of.the small-bore stuff that you're not currently 3 doing.

l 4 DR. FONTANA: Okay. What threw me was in addition  !

5 to risk-informed inspection. i 6 I got you. I see what you're driving at. Go j i '

l 7- ahead.

'8 DR. WALLIS: I think your report says something 9 like that. There was a lot of inspection of places where 1

10 there was never any leak, and there was noninspection of l

-11 places where there are.

l 12 MR. SHAH: Let'me ask two questions. Did we see 13 any leak events that can be considered a precursor to core 14 damage? And the second thing, if we do the risk analysis of j) 15' all these leak events and how the reasons compare with our j 16 standard understanding of the significance of the potential 17 for the LOCA events, so we are going to talk about that.

18 And what is the likelihood of the LOCA by looking and how do l 19 our reasons compare with the current. understanding of that. j i

[ 20 What we did, we review the reports published in 21 the accident _ sequence precursor program, and we found that

! 22 nine of the leak events were analyzed in the accident l

L 23 sequence precursor program. If four events were analyzed L

24 out of nine, four events were analyzed because of some other 25 reason than leaks. I mean, they would not have been 5

O ANN RILEY & ASSOCIATES, LTD.

\s,) Court Reporters i- 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

_ _ _ _ _ . __._._m.. - _ _ . - _ . _ . . . _ . _ - . _ _ _ _ . _ . _ . _ _ _ . _ -

154 1 analyzed only because of a leak, but other things took place I~T 2 which was more significant. And so they analyzed.

$, )

3. Going back, in the ASP. program, only those events 4 are' analyzed for which the conditional core damage frequency

5. is greater than 10 to the minus 6. So for these nine events 6- -the conditional core damage frequency was greater than 10 to

l. 7' the minus 6.

i l

p 8 But the first four events were analyzed for some l.

'9 other reasons than leaks. The fifth event caused the f .

l 10- reactor -- leak event caused the reactor trip. The sixth 11 one,.the leak repair action disabled a high-pressure L 12- injection system. And these two, the transient in this leak l

-131 occurred during the seventh event. And there were two l 14' events.where the large leakage took place and can be t

() 15' considered precursor to SB LOCA.

l 16 So using this information we identify the six i- 17- different risk categories. What we are going to do, then we 18 -will take each one of our events and identify where does it

, 19 belong, and we'll characterize that. So we say that there 20 will be events which do not contribute to any of these 21 parameters. And there will be events which will cause --

-22 which will be -- then we consider the reactor trip i I

, -23

. initiating events, and there will be events that contribute i

.24 .to PRA-modeled system unavailability. 'We created a new 25 -category where we combined both of these happen together.

l-c/ ANN RILEY & ASSOCIATES, LTD.

1(_,

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

[. #

- . - ..- . - . . - . - . - . _ . _. .- .-.. ...~.-.-.- - _ .. - _.~_- - . . - - -

155 1 And then we have the events that contribute to transient in 2 this LOCA probability, and events that will contribute 3 directly to SB LOCA initiating event.

4 We look at not only reportable events. We also l

5' included all the events, reportable and nonreportable, '

6 because many times the risk significance of a leak event 7 does not only depend on the magnitude of the leak, often it

i. -8 depends on the location of the leak. And so some of the l

l 9 nonreportable events leaks took place in some system which 10 will make it nonoperable, and therefore it has risk-l i

11 significance. So we included all the events in this L 12 analysis, i

13 Out of 230 we found 199 events which we do j 14 ' contribute to any this parameter. So we had about 31 events l( ) 15 left which we categorize-in these five categories.

! 16 What we found -- we had three events which caused 17 a reactor trip. So.we took those-three events and i

E18 determined the frequency. That is about 4.7 times 10 to the l 19 minus 3 per reactor year. In a typical PRA we use about one l  ;

20 to two events per reactor year. So what this one says, that l

l 21- the leak-induced transients that the reactor trips have no 22- discernible influence on the core damage frequency, because 23 this' number is much smaller than the number used in the 24 . typical PRA.

i 251 We also found that failures caused by reactor f-ANN R.ILEY & ASSOCIATES, LTD.

Court Reporters

1025 Connecticut' Avenue, NW, Suite 1014 l Washington, D.C. 20036 l

(202) 842-0034

, . . . . ~ ~ , _ _ _ _ . . - - . _ _ . - - - _ - _ _.

156 1 coolant system leaks have small influence on the safety

\. 2 system unavailability. We had totally 24 events in this l %-

(Y 3 category. Out of 24 events, 19 events made a single train i

4 unavailable, while five events where we had a leakage 5 through a weld in a common header or piping, that made 6 multiple trains unavailable.

7 So we had in 19 events of this category, five 8 events of this category, and we have taken about ten single

9 trains per PWR on the average, and about 2.5 multiple trains 10 per PWR on the average, and we determined those numbers.

11 Then we compared with the overall safety system failure 12 unavailability from the current PRA analysis. We found that 13 this number corresponds to like this 1 times 10 to the minus l 14 2. So this is a 1 percent like this is about 1 times 10 to

(} 15 the minus 4, so this is about 1 percent of the frequencies 16 we see in the current PRA. And here we find about 10 17 percent of what we have in the current PRA. Therefore we 18 say it is a small influence.

19 Now AEOD has another program called safety system )

l 20 reliability studies, and that will provide a much better 1 21 basis for this comparison.

22 What we found was that the observed frequency of 23 transient-induced relief-valve LOCAs is an order of 24 magnitude lower than that calculated in a typical PRA. We 25 found one event which produced a transient in this ANN RILEY & ASSOCIATES, LTD.

, (/"')s

, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014

! Washington, D.C. 20036 l (202) 842-0034 l

. - . - - - ~ -.- - - - - . - - .. - _ . ._ - - -.. - - - .. ..-. _ ... -..

157 l

1 relief-valve LOCA. This was a very small data base, so we I

{ 2 took the data from the'AEOD initiating event report project.

'3 Here they have considered a time-from 1969 to the 4 current period. So they had a much broader data base. And 5 they~came to this number, 5.1 times 10 to the.minus 3 per 6 reactor year, while they say if you look at this individual I- 7 plant evaluation analysis, that mean value is about 7.8 8 times 10 to the minus 2_per reactor year. This number is 9 about 10 times 1 order smaller than what is currently used 10 in the' typical PRA.

11. This is mainly the relief-valve LOCAs, but we did i 12 not. find in the reactor. coolant pumps a leak event becoming 13 a small LOCA does not exist.in the data base. All those 14 leaks were.much smaller than'10 gallons per minute.

'15 The observed frequency of small-break LOCA

([

16 initiating events is an order of magnitude smaller-than 17 those used in typical PRAs. We found only one event that 18 happened because'of that vibratory fatigue rupture in 19 McGuire. In the individual plant evaluation the mean 2(L frequency is about 1 times 10 to the minus 2 per reactor I 21 year. I think we will have a better base for comparison 22 with this number from the AEOD initiating event project, 23- where they-have.a much larger data base, and right'now they

24 -are working on this. And I think they are planning to 25 -present it later on.

i.

s ANN RILEY & ASSOCIATES, LTD.

,) Court Reporters

. . . 1025 Connecticut Avenue, NW, Suite 1014 l' Washington, D.C. 20036 (202) 842-0034 L

O .. . .. . .. - _ -

158 1- MR. ROSENTHAL: At this point it's just a few

()

Q 2 minutes after 5:00 --

3~ DR. BARTON: That's all right. Go ahead. Keep 4 going.

5L MR. SHAH: Okay. Thank you.

6 Okay. So we come to the last topic. How good are 7 'the current leak-detection systems? Here is a -- I have 8 shown a distribution of leaks by the leakage rate, and here

! 9 we have 45 events. This is the leak only inside the 10 containment, 'There are about totally 153 events or so. I 11 did not include the leak events which take place outside the 12 containment because the leak-detection systems are only 13 designed, are only provided for the detecting leak inside 14 the containment. That's what we are addressing here. So j ) 15 there are 45 events which were less than one gallon per 16 minute, and there are about 70 events that the leakage rate 17 was greater than'1 gpm. And there were 38 events we did not 18 know what was the leak rate, because it was not reported.

19 This may be of some interest, because if there is 20 a leakage, actually unidentified leakage of one gallon per 21 minute or greater, we want to detect it in one hour. So we 22 say that how the events, how the leak events with a leakage 23 rate greater than.1 gpm, what is the trend? And we see that 24 there is a statistically valid decreasing trend. But when 25 swe did the test, this was barely acceptable. And the main i

l-ANN RILEY & ASSOCIATES, LTD.

[

(^/j s_ Court Reporters l

1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

.. . _ _ _m _ _ . . . . . _ - - _ . . . . . .__ . __. __

159

1. reason is that if you look at this data, there is a

()

2 significant drop from here to here, and the exponential

'3 decay did not fit that well. So this is a barely acceptable 4- fit.

5 DR. WALLIS: If you look from 88 to 95 is pretty 6 flat.

7 MR. SHAH: That's right. If you take this data 8 out, we found it several times, like in the total events, if {

l 9 you take the.first full year out, then we will not see any '

10 trend, okay?

11 What we found was that the sump level monitor can 12 detect a 1-gallon-per-minute leak within an hour. Ideally 11 l

-13 it-can detect one in ten minutes at Oconee. But there  !

\

14 are -- and this is very simple. We have got a tank here and  !

() .

15 we take a marking, and it is monitored by computer. So when 16 you reach from one marking to other marking, the computer 17 will give the alarm. It will calculate the rate and give 18 you the alarm if the rate is increased -- is greater than 1 19 gpm in the given time.

20 DR. WALLIS: That's the gpm reaching the sump? Is 21 that considered that water can stick on surfaces?

- 22 MR. SHAH: So ideally if no water is sticking 23 anywhere, and then you can do it in less than ten minutes.

24 But now it depends upon what kind of insulation you have.

- 25 1You have metallic insulation, the leakage can leak through

] f ANN RILEY & ASSOCIATES, LTD.

\_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 f 1

• w

l 160 l 1^ the insulation very fast. If you have asbestos insulation, H l

2 it'may take a long time. And then it will drop on the 3 floor, and how far is the leak location from the sump? l l

4 Also, part of the leakage will become steam, will i 5 be_ evaporated, and then through the condenser cooling it )

l 6 'will'get collect'ed. So ideally you can do it in 10 minutes, 1 7 but generally you can detect that kind of leakage in 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />.

8 That's a much longer time period to collect -- to see the 9 effect of that leakage.

10 DR. WALLIS: That's demonstrated.

11 MR.' SHAH: If you have metallic insulation you 12 can, if you have metallic insulation. If you have asbestos, ,

i 13 then I do not know. l[t'may take a long time. Okay? An 14 airborne particulate can' detect a'1-gpm leak within an hour

() 15 provided.there is an absence of fuel failure, because that will provide you the background radioactivity and a presence

~

16

.17 of corrosion product activity.

i 18 DR. WALLIS: That's a' bit odd. If it's a leak of l 19 water and the water's cold, it doesn't make any -- it i20 doesn't take much in the way of particulate airborne 21 products, does it?

22 MR. SHAH: Because it is a -- generally in a 23- reactor --

DR. WALLIS: It depends on what kind of a leak it 25 is. If it's a steam leak --

i

/'N ANN RILEY & ASSOCIATES, LTD.

k Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. - - - - m .

l I

1 161 '

1 MR. SHAH: Reactor coolant system, some really

('T' 2 high pressure, we have reactor coolant system leakage.

LJ 3 Generally it will be --

1

! 4 DR. WALLIS: It depends on how it comes out. If l 5 it gets stuck in the insulation and drips out, it's very  ;

l 6 different from --  !

l l

7 MR. SHAH: That is true.

8 DR. WALLIS: Squirts out as a steam jet.

9 MR. SHAH: In come of the small leakage, they were 10 not detected by any of --

11 DR. WALLIS: This second bullet is subject to 12 qualification.

13 MR. SHAH: That's right. Provided there is an l 14 absence of fuel failure and presence of corrosion product

( ) 15 activity, plus constraint on the insulation. What you are 16 saying that if you have asbestos insulation and it does not 17 come out, then you will not be able to see that.

18 DR. FONTANA: Is that the nitrogen 13 thing?

19 MR. SHAH: No.

l

'20- DR. FONTANA: NO.

1 I

i 21 MR. SHAH: Okay. In. France they use N-13, in this 22 country there is no place it is used for leak detection. In l l

23 AP600 they are planning to use the N-13, they have proposed.

24 But we do have it in our current operating plants.

l- 25 The airborne gaseous radioactivity monitor cannot I ,

i

(

A ANN RILEY & ASSOCIATES, LTD.

Court Reporters 3 1025 Connecticut Avenue, NW, Suite 1014

(:

l Washington, D.C. 20036 (202) 842-0034 L

.. . . . ~ .-- - .- - - -. . . ~ . - _ - . . - _ - . - . - . - . . - . . _ - - .

162 l l' detect a 1 GPM, mainly because -- in one hour, because of 2 the improved fuel performance. There is no cladding defect, 3 and so there is no gaseous products involved in the leaking 4 coolant, so this --

L 5 DR. SHACK: What is a typical leakage, known 6 leakage rate for a PWR? I mean it sits around and it leaks. '

7 Is it typically leaking a gallon per minute?

l 8 DR. BARTON: No. I would say less. f 9- MR. SHAH: Oh, there is identified leakage and I

'10 unidentified leakage. 1 11 DR. SHACK: Identified leakage. So, you know, you l 12 have a 5 GPM limit. I 13 MR. SHAH: No , 10 GPM.

14 DR. SHACK: 10 GPM.

15 DR. BARTON:

( Administratively, they go to 5 and

( 16 start shutting down. But it is 10 in the tech spec.

-17 MR. SHAH: 10 GPM.

18 DR. SHACK: But I just wonder how much background 19 leakage is there? I mean is it typically on the order of a 20 few tenths of a GPM?

I=

l 21. DR. BARTON: I would say so.

l 22 MR. SHAH: I do not know.

=23 DR. WALLIS: It goes to the sump?

24 MR. SHAH: Yes. The leakage -- okay. Part -- the

'25 leaking coolant with partition, maybe partly in the water

-(g ANN RILEY & ASSOCIATES, LTD.

(_)- Court Reporters 1025. Connecticut Avenue, NW, Suite 1014

. Washington, D.C. 20036 (202) 842-0034 1

163 1 and partly in the steam, and that steam will go through the 2 containment condenser and cool down and get collected into 3 . sump. It takes some time. So all the leakage will part 4 eventually and go to the sump.

5 DR. WALLIS: What is the typical sump flow rate?

6 MR. SHAH: The point is, the important point is 7 how much the sump. flow rate changes and that is what is 8 monitored. Increasing the sump flow rate is monitored and 9 .that will give you the signal about the extent of the 10 leakage.

11 DR. SHACK: I guess my -- are the radioactivity 12 monitors, do they get. saturated by the normal background 13 leakage? Are they always going off? I mean, or do you sit 14 here and you raise the threshold until it stops going off,

() 15 16 and then, you know, when does it go off?

MR. SHAH: The radioactivity monitor has a certain 17 sensitivity, it will detect below that value. I think I

18. have those numbers.

19 MR. LaFAVE: This is Bill LaFave with the Plant 20 Systems Branch. In response to your -- the normal leakage 21 of most of the PWRs, the unidentified leakage is normally 22 down to .1, .2 GPM, as I recall. And if it gets up even to 23' .3, .4, .5, they start to --

s 24 DR. SHACK: They start looking.

25 MR. LaFAVE: Yeah.

./

- ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

,-~,..w 4--y: , . , - - . - __

i 164

'l DR. SHACK: Thank you.

(T 2 MR. SHAH: Thank you. In the containment air V~

3 cooler condenser flow rate, monitor can detect a 1 Gl4 4 leakage from a high energy line, but the results may L'e

! .5 inaccurate in plant operating condition changes and, 1

6. generally, like in AP600, initially, this was specified as I L

l 7 one of the leak detection systems which they drop later on.

8 Also, in France they do not require.to have this particular

! 9 system because it is not that accurate in predicting the l

10 leak rates. I 11 One of the -- two things we are lacking in our i 12 current leak detection ~ system, we cannot locate, it cannot i 13' tell us where.the leaking takes place, so you have to go --

i 14 you have to enter the containment to find the leak location.

t es LI ) 15 And, second, if there is a very small leak, like a .1 gallon

%./

16 per minute or so, then it has failed to locate, failed to 17 detect in several instances.

18 So, the advanced leak detection system can cetect 19 a very small leakage and locate the leak source. It can --

l 20 because, generally, these systems are directly installed on 21 the pressure boundary, and like nitrogen-13 monitor can 22 detect a leak of 0.005 gallon per minute in one hour, and l 23 . acoustic monitors can detect a zero -- it has detected, in

24. France, when there is leakage through the CRDM nozzle 25 penetration, a crack, the leakage rate was 003 gallon per l

![] ANN RILEY & ASSOCIATES, LTD,

' \m) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. . - . . - - . _ - - - . - . . _ . - . ~ _ . . . - . - - . . ~ _ - . - _ . - - - . ~ . _ . . ~ . . . - , .

165 L >

[

1 minute and'they' detected that using the acoustic monitor.

1 2 DR. WALLIS: Leak noise?

3 MR. SHAH: Pardon? Yes.

4. DR. BARTON: . Yeah, leak noise.

i 5' DR. WALLIS: A hissing noise.

~6 MR. SHAH: Yeah. -

7 DR. WALLIS: The way you do it, this is for the i

8 drips or the hissing?

l

9. MR. SRAH: This is'the one -- this leak detection 10 system is different.than all other. In all other leak I 11- detection system, we'look at the accumulated leakage and, .

12 from that, we detect the leak. Here, we detect it because 13 of leakage itself. Okay. This is different than all other l- -14' methods.

115 DR. FONTANA: Excuse me.

16- MR. SHAH: Yes, sir.

17 DR. FONTANA: Can that work while the plant is 18 operating,-above all the background noise, can it 19 discriminate that well?

20 MR. SHAH: Let us go back. Which one?

l 21' DR. FONTANA: Acoustic monitoring.

22 DR. SHACK: Acoustic.

~23' DR. FONTANA: There's a lot.of background. noise

bt ~ .when-the plant is operating. It can discriminate --

2

.5-  ;MR . SHAH: It can discriminate. Actual'--

l lt

"')

~

I ANN RILEY & ASSOCIATES, LTD.

. (\,) Court Reporters

i. 1025 Connecticut Avenue, NW, Suite 1014 L Washington, D.C. 20036 E

(202) 842-0034 l'

, - . ,n . .:, .- . . , . -..,+.a - - .. : . - . -- -.- - _ - -.

l 166 N

1 DR. SHACK: It is typically up at a couple of

/N 2 hundred kilohertz.

N.]

3 MR. SEAH: In one BWR plant, Plant Point Beach, I 4 think they installed this to detect the crack growth, not a 5 leakage. Okay. It can do that. I think ASME, there is a 6 code case called the acoustic monitor application. And in 7

Germany they have developed the local humidity monitoring 8 system which can reliably detect a 2.02 gallon per minute 9 leak, and it has been installed in some of the VVER in the 10 Czech Republic.

11 All this -- generally what you do, you sample, you 12 locally sample the leakage.

13 Here, what they do here, take a long tube and it 14 is perforated at every half a meter. You put it under the I [x /) 15 insulation and every 15 minutes you purge all the air and 16 you record the signal and when you get a high humidity 17 signal, then from that you know where it is, so this will 18 tell you the location and the detection both.

19 And also this event -- here you have have to have 20 a special insulation and you have to have a collecting 21 system right on the pressure boundary, so you know where it 22 is, so you know in your sample of this to monitor the whole 23 pressure boundary. Okay?

24 So I think what I would like to do in my last two 25 slides, if you permit, is to go over the most important

/~'s ANN RILEY & ASSOCIATES, LTD.

\m ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i

_m

167l 1 findings which I presented to you -- the summary of this.

('~i 2 The first one is the reportable leak event showing

\,

3 statistically significar.t event in calendar time. The leak 4 events associated with valve-pncking have mainly contributed 5 to this trend.

6 Thermal fatigue has caused through-wall cracking 7 in the base metal. The thermal hydraulic phenomena that 8 cause thermal fatigue cracking in PWR branch lines are not 9 yet well understood. The leaks through thermal fatigue 10 cracks in branch line could lead to SB LOCA.

11 And here is my last slide. The fast growing 12 fatigue cracks may require monitoring of pipe wall 13 temperatures and pipe vibration. The risk impacts observed 14 in this study and in the recent initiating event frequency

,, ,\

ib 15 study are lower than what are reflected in the typical PRA.

16 I would like to caution here that when we did this  !

I 17 study we did not include the effect of aging in this data, l l

18 but because of aging some of this frequency may be {

19 increasing with time, so we need to, as we get more data, we 1 20 need to come back to this part of the analysis again and 21 check it out.

22 Finally, the sump level monitor and under certain 23 conditions an airborne particulate radioactivity monitor can 24 detect a 1 GPM leak within one hour.

I 25 Thank you. Any questions?

(^ ANN RILEY & ASSOCIATES, LTD.

(,,]) - Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

. - .. - - . . _ . - ~ . . - - -- . - . - . . - . . - . - - - . . . . --

168 l 1 1 DR. FONTANA: Well, what happens next? '

2 MR. SHAH: Okay. Next -- what we are doing right 3- now, using some of this information, to calculate the LOCA 4 frequencies for these initiating event reports, and the next

. i 5 thing we are going to do is to complete this report. I L

6 think that was not your question.

7 What are the next actions items to pursue this?  !

8 What~we would like to do and what we have not done here and i

9 what nmy be important is that the only thing we did here to 10 evaluate the leak detection system, we.went back to Duke 11 Power and tried to find out when they had a leak at Oconee 12 what they did.

13 We would like to go back to their utilities where 14 they had big leak events and see how their leak detection l

1 I] 15 system performed and get more data on that issue.

G l

16. .Also what we would like to do is to do the similar 17 analysis for the BWR plant as for the PWR plant.

18 DR. WALLIS: You are still studying this. What is 19 the Staff position? What is the Staff position on this l 20 issue? I 21- MR. SHAH: It's when we are going to finish this?

22 DR. WALLIS: This.is a study you are doing as a l

23 contractor.

24 MR. SHAH: Yes.

25- DR. WALLIS: Does the Staff have any position on e

i l ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l

Washington, D.C. 20036 (202) 842-0034 l

169 1 this?

j ,[

,a

} 2 MR. ROSSI: Well, we are going to finish up the 3 study and then work with others on the lessons that are 4 learned from it. In particular, I think the thing that i

5 warrants looking at most is how we are doing inspections and 6 whether we are doing them in the right place and as you can 7 see, all of the numbers here are less than what's today 8 assumed in the PRA so we don't have a good basis for 9 requiring anything at this point in time.

10 DR. FONTANA: But it could have an effect on plant 11 license -- plant relicensing and license extensions.

12 MR. ROSSI: It could have effects there, so we 13 will make sure that the results are widely disseminated --

14 DR. SHACK: Wouldn't those initiating frequencies

' r^N (x_-) 15 include other events beyond what we are talking about here?

16 MR. ROSSI: Yes. I think it has them divided up 17 into various things. That study I believe you must have 18 because I think it is out in draft for comment.

1,9 - MR. BARTON: Yes, we do have the draft study, and 20 I guess Mario's question or Graham's question was what are 21 you going to do? Finalize this study and once you put it 22 out, there's a lot of good stuff in there, but there's also 23 a lot of things that you could take out and say, well we're 24 going to make some rulemaking here. We are going to 25 require, you know, advanced leak detection. What are you b

\_/

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l-

l 170 J

1 going to --

-2' MR. ROSENTHAL: We have not made --

ty 3~ MR. ROSSI: Not on the basis of the numbers.

4 MR. BARTON: Because if the PRA numbers are 5 showing, you know, where your leaks are with respect to PRA 6 numbers.

7 MR. ROSENTHAL: If anything, it all looks better 8 than had been previously thought, so they just control the 9: backfit -- there's no regulatory basis. The stratification I

10 issue remains a concern with aging, but the rest of it is a i

11 good news story and I mean it is fact and so we should i 12 accept it.

13 Now it was I think in my mind appropriate to look l 14 at, to try to do a phenomenological understanding of the

/ l

(x )T 15 bases for leaks which led you to an understanding of'the 1G small break LOCA, and not simply a number count.

17- MR. BARTON: Right.

18 MR. ROSENTHAL: We have done that for Ps.

19 For boiling water reactors, small break LOCA is 20 not a dominant, at least in that PRA, is not a dominant 21 sequence the way it is on --

22 MR. BARTON: On PWRs.

23 MR. ROSENTHAL: On BWRs it is not, that way it is 24 on Ps.

25 MR. BARTON: Right.

ANN RILEY & ASSOCIATES, LTD.

<k.C)'

_, Court' Reporters 1025 Connecticut Avenue, NW, Suite 1014

, Washington, D.C. 20036 (202) 842-0034 i

L

. -- . . . - - - - - . ~ . . . .~ . -.-. . - -. . . _ - .-.- . - . -

171 1 MR. ROSENTHAL: So we are not at this point

(}

V 2 motivated to extend the study to --

3 MR. BARTON: -- to the BWRs --

4 MR. ROSENTHAL: -- to the boiling water reactors.. '

5 Now in the broadest picture, what you are hearing 6 coming out of AEOD is a series of factually-oriented -- ,

7 factual reports of the operating experience put into the 8 context of the risk analyst and this is just one more.

9 DR. SHACK: But of course I mean it does support 10 the effort that they are already doing on risk-informed 11 inspection, because what you are learning from this is that 12 Section 11 isn't the greatest inspection.

13 MR. BARTON: Right.

14 MR. ROSENTHAL: That's correct, and we have

!I E 15' discussed that with NRR and RES and I even presented that at lQ 16 a CRGR meeting, and that issue, which Dr. Rossi brought up, l 17 is the one which we'll pursue.

18 Now of course we print up about 2000 copies of 19 these reports and send them out all over the states and 20 internationally, so we will surely be sharing the L 21 information.

22' MR. BARTON: Okay.

23 Any other questions or comments from members of l-24 the committee?

.25 [No response.]

!(1 ANN RILEY & ASSOCIATES, LTD.

(K,,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 H

(202) 842-0034 L

t

. -- . . . . . . - _ _ . . - . . . . . - . . . - . - -~ . -- - _ .

j 172 l

, 1 MR. BARTON: Any other comments from the public?

i i\/

Q 2 (No. response.]

.3 MR '. BARTON: That's.what happens at five o' clock.

4 It precludes a lot of interest.

. 5 Well, if not, I want to thank the Staff for their

! .6 presentation. I think it's a good report and there is a lot 7- of good information in there and also it' raises a lot of l

i 8 questions and you heard some of them today and I guess we 9 will hear some more questions after you get the report out.

10 Thank you very much.

11 MR. ROSSI: Thank you. j

'12 MR. BAGCHI: Since we lost the Chairman, Vice .

. l 13 Chairman, and a couple other members, we are now scheduled 14 for a 15-minute break.

[l[()- 15 ~ They do intend to get back and discuss the other 16 ' items on the agenda, so we will break unt twenty of six.

F 17 [Whereupon, at 5:24 p.m., the recorded portion of

18 the meeting was recessed, to reconvene at 8:30 a.m.,

19 Thursday, November 5, 1998.]

i 20

, 21

]

. 22

, 23 24 l~

25

^- ANN RILEY & ASSOCIATES, LTD.

s,)

m Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034

)

REPORTER'S CERTIFICATE L. .

i' f~'C This is to certify that the attached proceedings before the United States Nuclear Regulatory Commission in

'the matter of:

NAME OF PROCEEDING: MEETING: 457TH ADVISORY COMMITTEE l ON REACTOR SAFEGUARDS

[ l l

I 1

DOCKET NUMBER:

l PLACE OF PROCEEDING: Rockville, MD i

--r N

() were held as.herein appears, and that this is the original

'}

]

transcript thereof for the file of the United States Nuclear Regulatory Commission taken by.me and thereafter reduced to I

{

l i typewriting by me-or under the' direction of the court L reporting company, and that the transcript is a true and accurate record of the foregoing proceedings.

o 1 .

f x9n & _

V Jon Hundley O'

Official Reporter i Ann Riley & Associates, Ltd.

v i

I-l

Changes in the O ^itra tiv = #r= T r- ar a Due to Office Concurrence Comments aaui a =k a-PERB received comments from some of the offices during the office concurrence process that occurred in parallel with the review by the ACRS. The majority of these comments were editorial in nature and will not change the technical content or intent of the documents. The (

following are comments that were deemed to be significant.

Statements of Consideration (SOC)

Based on discussions with OGC and others, the phrase " accident dose guideline"has been  ;

replaced with " accident dose criteria"in the SOC, except in those instances where the l reference is to the guidelines in $100.11. While Part 100 definitions of EAB and LPZ may have been siting guidelines, in design space, the numeric dose values are considered to be criteria.

This change would clarify intent.

OClO requested changes in the language of the Paperwork Reduction Act statement. No )

specific OMB clearance will be requested as stated previously. Proposed wording will indicate that the agency determined that the information collection burden was insignificant.

Proposed Rule Language l

RES suggested revising proposed footnote 2 to $50.67 to remove material that is addressed in the SOC:

• Tra ;, . .f's;..; dc_ eq .._tc.; gt0 ; ef 0.20 Ov (20 recc.; ..:..r.d e:~.; % :;;%.4 tr

.e ;0 re.;;d ee ree Oc.c c , ;t ;%ee e .-%L .~..J: :..; .. Jc-:-:' { ., t a _::: Oe

s. DOL,_ ;.T.pe; ef e l 7;%.;;; a4;,.; pa e;; Ledy er,ene. Tt.e :.....; an ; ilk ;";

sed;e.L,.. de.e of 0.20 Os '20.....; TCO % ,~.._ _ : .; ;m te LOn; .anar rsk n_: _ ::./0 e,.pe .;; ef 0.00 Os (20 i ...; ;G re te% tedi end 000 ;;.-. t 74 0, rem. 7;ek ef ':c.; r

J .j % _:: ee ta ;4k.T.::: re e:na 4 .... ;.; 7.x".h etjes/.;;; Or ;; t.no L.en : _ __-+' d in te 0.cc.cc'- :-n'; O_!_.j C-:# "e;ej. l L ;.;r.The use of 0.25 Sv (25 rem) TEDEinihese ea; den; d . ;d...ree is not intended to imply that this value constitutes an acceptable limit for emergency doses to the public under accident conditions. Rather, this 0.25 Sv (25 rem) TEDE value has been stated in these guides as a reference value, which can be used in the evaluation of .

proposed design basis changes with respect to potential reactor accidents of exceedingly low probability of occurrence, and low risk of public exposure to radiation.

RES suggested revising the definition of source term in 650.2:

Source term refers to the magnitude and mix composition of radionuclides released from the reactor core, their physical and chemical form, and the timing of their release.

Environmental Assessment OGC requested changes to the section entitled " Conclusion" to remove the appearance that this section is a draft determination under the provisions of NEPA. The draft determination is to appear only in the SOC. The SOC already contains the draft determination.

.bo m__ -

O O O j ASSESSMENT OF PRESSURIZED WATER REACTOR PRIMARY SYSTEM LEAKS z

[98REGu0 (M)9 4 4*5 V AEOD Staff Presentation to the ACRS November 4,1998  ;

I b SCHEDULE OF STUDY .

- January 1997: Study began with searching and collecting data

- December 1997: Draft report issued for peer review

- March 1998: Review comments received

- October 1998: Final draft completed with resolution of the comments

- November 1998: Final report a NUREGICR document i

i i

i 1

I O O O

. PROVIDERS OF PEER REVIEW COMMENTS

- NRC NRR: SPLB, PECB, EMCB, SPSB i -

RES: DET, DST .

AEOD: RRAB i

- Industry Nuclear Energy Institute (NEI)/ Westinghouse Owners Group Institute of Nuclear Power Operations (INPO)

Babcock & Wilcox Owners Group t Duke Power Company Susquehanna Steam Electric Station RSA Technologies /Swedish Nuclear Power inspectorate (SKl:1 2

STUDY OBJECTIVES  ;

i Assessment of U.S. experience related to pressurized-water system primary system leaks in terms of:  !

t

- Leak frequencies and rates

- Leak locations and causes  !:

1

- Safety significance of leaks j l

- Industry efforts to reduce the number of leaks  !

- Effectiveness of current leak detection systems f

l i

I 3 i t

STUDY SCOPE

• Reviewed a total of 240 leak events from 1985 thru September 1996 Interfacing systems leaks and steam generator tube leaks not included 41 events considered significant from a qualitative standpoint Nine events analyzed by AEOD accident sequence precursor program-conditional core damage probability range 1.3E-6 to 3.3E-3

• Compiled a database to identify trends, distributions, and causes of leak events

- Visited plants to discuss and collect industry experience

• Reviewed related technical literature, NRC communications, licensees failure reports

• Analyzed selected leak events in detail 4

-_-__ _ _ __ j

~ ~

O O O STUDY FINDINGS The study presents findings in the following areas:

- Leak frequencies and trends

- Degradation mechanisms, failure modes, and locations

- Leaks with potential for relatively rapid growth i

- Safety significance of leaks caused by thermal fatigue

- Leak events that may be regarded as precursors to core damage i - Effectiveness of current leakage detection systems

, i

- Information relevant to risk-informed inspection programs i

. _ - . . ~ . . . . . _ _ _ . . _ _ . _ . . . . _ _ . _ . _ . . _ _ . . . . _ _ . _ _ . . _ _ . _ _ _ - . _ . _ .

./-

e ~i O O -

O 1 ACRS BRIEFING ON

! APPLICATION OF REVISED SOURCE TERM 2

AT PERRY NUCLEAR POWER PLANT AS A PILOT PLANT NOVEMBER 4,1998 JAY Y. LEE t

EMERGENCY PREPAREDNESS AND RADIATION PROTECTION BRANCH DIVISION OF REACTOR PROGRAM MANAGEMENT OFFICE OF NUCLEAR REACTOR REGULATION t

i t

________m_ _ _ _ - _ - - _

O O O

PILOT PLANT REVIEW

• Commission Paper SECY-96-242 (11/96)

!

• Provided the staff's Proposed implementation Plan of

! Revised Accident Source Term for Operating Reactors

. i f

i i

. SRM to SECY-96-242 (2/97)

• Undertake Rebaselining

. Initiate Rulemaking Upon Completion of Rebaselining

. Evaluate Pilot Plant Applications Concurrent with Rulemaking

. Use TEDE Dose Criterion and Worst 2 Hour Methodology in Implementation of Revised Source Terms at Operating reactors b

- - _ - - - - - _ - - - - - - . - - - . - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -----_----.-----J

O O- .O e PERRY LICENSE AMENDMENT REQUEST

• Remove Main Steam isolation Valve (MSIVl Leakage Control System

. Increase Maximum Allowable MSIV Leakage Rate

. Use Sodium Pentaborate in Standby Liquid Control

! system for Containment Pool Water pH control i

i

REVIEW OF PERRY LICENSE AMENDMENT REQUEST e Approach Reanalyzed site boundary (EAB and LPZ) and control room dose as a result of a postulated LOCA e Source Term NUREG-1465 for fission product release magnitude, release timing, and chemical species with no exceptions e Dose Criteria 25 rem TEDE at EAB (any 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> duration?

and LPZ (30 days) 5 rem TEDE in control room (30 days) e Dose Calculation RADTRAD (NUREGICR-6604), "A Simplified i

Model for RADionuclide Transport and Removal And Dose Estimation"

O O O

AREAS OF REVIEW e Aerosol deposition in drywell NUREGICR-6189, "A Simplified Model of Aerosol Removal by Natural Processes in Reactor Containment"

• Aerosol deposition in main steam lines Brockmann's Model in RADRAD e Containment Spray NUREG/CR-5966, "A Simplified Model of Aerosol Removal i by Containment Spray"  ;

i

o o cc  !

+ I e lodine Chemistry l

. Partitioning of iodine from aqueous to gas phase 4

l . Effects of hydrogen ion concentration on iodine l partitioning  !

1

• Thermal-Hydraulic and Fission Product Transport from Drywell to Containment e Meteorological dispersion factors for control room air a intake d

i

-*s.o o o Assessment of Pressurized Water ,

Reactor Primary System Leaks Vik Shah

i

Presentation to Advisory Committee on Reactor Safeguards i

Rockville, MD November 4,1998 l

INF O

FT

o o o

.l .

^

Project Objectives Review the U.S. experience relating to PWR primary system leaks in terms of

- Leak frequencies and rates

- Leak locations and causes

- Effect of aging on leak frequencies

- Safety significance of leaks

- Industry efforts to reduce the number of leaks and their magnitudes

- Effectiveness of current leak detection systems C98 0EM 2

t Project Scope i

Review of leak-related licensee event reports (LERs) submitted from 1985 through the 3rd quarter of 1996

• Included leak events during hot shutdown, hot standby, startup, and power operation

• Did not include intersystem leaks and steam generator tube leaks

o o

.'.o Specific Actions  !

. Searched and reviewed LERs

• Developed database Visited PWR plants

• Reviewed related technical literature

• Analyzed selected leak events .

O O O i PWR Primary System Leaks Findings

= Trends

• Degradation mechanisms, failure modes, and locations identified since 1985

• Leaks having potential for relatively rapid growth

• Safety significance of leaks caused by thermal fatigue

• Relevant information for risk-informed inspection programs

• Leak events as core damage precursors

• Effectiveness of current leak detection systems l con cess a

Distribution of Reportable Leak Events and Their Frequencies 1.2 39 C3 Number of leak events 36 ' e Frequency (leak events / operating year) 33 -1 r

aC 30 1' '

X g 27 -- -

- 0.8 $ ,

$ 24 V8 @

21 \ -

- 0.6 .a 18 - k _

a

.8 15 \ - -

E E - 0.4 5 y 12 - _ _

x 9 \ -

A* -

6 k# s y

/ *w M - 0.2 3 -- y 0 0 85 86 87 88 89 90 91 32 93 94 95 96 Calendar year

o o

.o  ;

Statistical Investigation of Trends .

i

• Number of events in any year is assumed to be a Poisson  ;

random variable

• Loglinear model: A = exp(a + bt), where 1 is the true occurrence rate i

i

• Maximum likelihood method was used to estimate the i parameters a and b

• Examined the goodness of fit

• p-values were calculated to estimate statistical significance of trend

.O O O There is a statistically significant decreasing trend in frequencies of the reportable leak events.

1.25E+00 "

i Point est. and 90% confidence interval Fitted rate 1.00E+00 To c) s

> s

@ 0.75E-00 e s

? N~

a Qm 0.50E-00 '

' &'s,Ns

~

s s

[,

E '"

~

T -

O 's.. s 0.25E-00 '

e

-Q 0.00E+00 85 86 87 88 89 90 91 92 93 94 95 96 Year  ;

slope b = -0.17 c- - '

p-value = 0.0001 t

.o o o

' ~

Leak events associated with valve packing degradation show an apparent decreasing trend with both calendar time and age.

i Point est, and 90% confidence interval I Point est. and 90% confidence intervat u ' Fitted rate Fitted rate 5 - - - 90% confidence band on the fitted rate h

p 4.OE-01

- - 90% confidence band on the fitted rate 4.0E-01 g

e i Y \ E 5 3.0E-01 <

s g 3.0E-01 5 $ E

"\ s E k 2.0E-01 e 2.0E-01 i

$ N

's s

@ \s- -s s

• sm-

> 1.0E-01 N - 1.0E-01 s ,

,~

'sN m W '

~ ,-

-Q'~w_

QQ'- y

< ' ~ ~ -

, ,u __

__~ ~- ~-

~

'~ ~

7--

0.0E+00 0.0E+00 85 86 87 88 89 90 91 92 93 94 95 96 3 5 7 9 11 13 15 17 19 21 23 25 27 29 Year Year Trend with calendar time Trend with age slope b = -0.59 slope b = -0.10 p-value = 0.0001 p-value = 0.002 c- .

o o o

~

Leak events associated with vibratory fatigue show statistically significant decreasing trend with age but not with calendar time. ,

l 0.25 025 1 @',5[,and , 90% confidence interval '

es . M Mm intM

- - 90% confidence band on the fitted rate : - - - 90% confidence band on the fitted rate e e

% 0.15 g 0.15 -

l g 3 o-

a. x '

o o,10 s ., , 3 0.10 -

.\

$ - - E .

~ ~

M- - --- -- --- -- b 0.05 - < N.

i LU --- --~-- -___ _,,

1 -

%si ,  ;

, 0.00 0.00 , , , , , , , , , , , , , , , , , , , , , , , , , , , . . , , t 85 86 87 88 89 90 91 92 93 94 95 96 5 10 15 20 25 30 i Year Year Trend with calendar time Trend with age slope b = -0.08 slope b = -0.07 p-value = 0.16 p-value = 0.025 c._

4 i

Leak events associated with thermal fatigue show statistically significant increasing trend with age.  ;

4 n I

Point est. and 90% confidence interval / -  ;

0.03 Fitted rate / _

---

90% confidence band on the rate / _

l _

~ ~

$ - j _

$m 0.02 -

y/5

,/

,' t

-6 -

e .- j o , _

0.01 m

y' - - - - - -- ------ --- -

_________.____s-o go ._ - -" - - - _2____

1 1 i e 1 i i1 1  ! t I I i 1 1 I e  ! I i e e f f e t t

[

s to 15 20 25 Age (yr) j slope b = 0.0868 p-value = 0.026 -- ,1

o o

~

-O

~

PWR Primary System Leaks Findings f

• Trends

Degradation mechanisms, failure modes, and locations

%

• identified since 1985 '

i . Leaks having potential for relatively rapid growth i

. Safety significance of leaks caused by thermal fatigue Relevant information for risk-informed inspection programs

. Leak events as core damage precursors

• Effectiveness of current leak detection systems i

cas os35 12 I

~

'. O O O Turbulent penetration and thermal cycling have caused thermal fatigue cracking in branch lines.

I t

Turbulent penetration Leaking valve  !

Stratified, coolants U f r i

0 & D $~ b- b >

Branch line

~ '

(Main Loop) Cycli zone (potential site for fatigue crack initiation) " "

I \ \ Stratified flow

~

. - Turbulent decay correlation b e*, Turbulence

~

A Distance frorn header pipe (UD) RCS f cold le9 in leakage of cold Temperature distnbution Loop B

, coolant 2.7 Urnin t i

Distance from header pipe (UD)

Farley 2 Turbulence Interaction Safety inj.ect. ion P. .iping Regions (12/87) c ee ss is

O O O i

Thermal cycling has caused leakage through base metal.

1 t

i l J f I Section A-A RCP 120 VP Weld 9i -

Check valve  ! /N'-

' Crack 1

~ ./ g "g gcrack 1 Section B-B

[

Position of the B.l_. Crack 2 .]B Weldv Section C-C ,

O# ,s

,g BC1 1

Crack 3

" /M Cr

%.-i

~ --

i 1  :

Tihange 1 (6/88) Dampierre 1 (12/96)

{

c . !

~ ~

s.

O O O i

Cavitation-induced -

pressurizer' 1.

vibratory fatigue and J's 4 ~, a L a- -

water hammer have 7' g .

y g 1,,td n heat exchanger caused rupture of a -"

Or' " a  :: lecv letdown system  ?+

,-Drain; 2. l

( * "

I drain line.

.... l i

i Regenerative lV heat exchanger I I ly I Charging

(

545'F flow l I

From 490 7 I

Loop C FCV FCV Legend I

k Flow <ontrolvalve(FCV) To loops and

>4 Normally closed valve cold legs 1.'l Flow measuring orifice l Inside containment  ;

Typical Westinghouse letdown system for a 4-loop plant _ , ,

~ ~

-O O O 1

l Primary water stress corrosion cracking has

caused through-wall cracking in hot-rolled

, Alloy 600 penetrations.

4 Through-wall cracking is mainly in the base meta!

l = Eight leaks took place through pressurizer instrument nozzles and heater penetrations

! . Two leaks through main coolant piping penetrations l

l c e= ..

O O O ,

s PWR Primary System Leaks Findings

• Trends Degradation mechanisms, failure modes, and locations identified since 1985

.

• Leaks having potential for relatively rapid growth

• Safety significance of leaks caused by thermal fatigue

• Relevant information for risk-informed inspection programs

• Leak events as core damage precursors

• Effectiveness of current leak detection systems canas n

> i Leaks through thermal fatigue cracks ,

in branch lines could lead to SBLOCA.  ;

• Leakage rate of 132 gpm through a fatigue crack in a ,

. residual heat removal line at Civaux 1 High crack growth rate reported in a safety injection .

line at Dampierre 1 t

I Low crack initiation time, high crack growth rate, and a large crack size

• Uniform growth of a circumferential fatigue crack may lead to a long through-wall crack c i.

i

I 1997 cracking event at Oconee 2 suggests the possibility of uniform crack growth that ,

may lead to a long through-wall crack. f OO Crack Tip @ 344' (100% TW) l Safe End !r Check valve HP1 o Through-wall Nozzle (x sw ru s., Crack location h OD Crad rp  !

a

1. ' (1 J J (Co d Leg) "Q -

f ,

Isolatio -

7 ';N g g' Welds ,' y ,

(Block) '"

i == ' Valve k

MU/HPI Line N Warming Line Metallography (24m a gg; Typical layout of Angular distribution of depth Oconee 2 MU/HPI line of the circumferential crack i

I

Vibratory fatigue cracking can ,

lead to SBLOCA.

R

.1;p / un pipe

= qg

-:e i

" " / Pipe wall Nek an et Toe g 7 Fillet weld Fillet weld N

.. _j Crack

\ -

p;  !!I _l~ Toe x 7y '"

Small bore w , w.

branch pipe .A x

Half coupling Couplin Socket-welded connection Crack initiated at weld root

• Main mechanism for leakage through piping

• 28 leaks from small diameter (< 2 in.) piping

• Maximum leak rate - 87 gpm

o o

' ~

0 Failure of compression fitting can lead to LOCA.

Adaptor I: 1 .

r Fe<<"ie

-g

\ Adaptorshoulder i

b >

MRIM2!$MP

-* Gap *-

14 leaks from instrument or sample lines having diameter less than 1 in.

• Maximum leak rate 130 gpm ___

(,J A complete failure of all four seal stages '

could lead to SBLOCA.

}

}

r~ '

t-M ,

Vapor sealleakage 4) 1 3 f

. _ _ . . > CortroRedbleedoff po ps.g - to VCT (1.0 gpm)

Upper seal W c5" ,,

p I

f4ddle seal N a,=i --

e V

1500 psa y Lower seat n j _

Auxiliaryimpeller '

ReciruAating

, 2250 pas prirnary coolant  ;

X 0000 c h (40 gpm)

A artegral heat

- Component

- A(f s;@w rpam Ill

.~

_ IE C=. _

1.0 gpm from Mixed Row N. . ./ _ _ .

1 i

I i

o o o W

i PWR Primary System Leaks Findings

• Trends

• Degradation mechanisms, failure modes, and locations identified since 1985 Leaks having potential for relatively rapid growth pas) . Safety significance of leaks caused by thermal fatigue

• Relevant information for risk-informed inspection programs Leak events as core damage precursors

• Effectiveness of current leak detection systems

o o o Thermal fatigue of PWR branch lines is becoming a safety issue.

• Large leak rates have been observed.

Experience with a rapidly growing fatigue crack is limited.

• Thermodynamic phenomena that caused the thermal fatigue cracking are not yet well understood.

Number of leak events caused by thermal fatigue increases with age.

• Stability of big cracks under seismic conditions combined with a cold leak is not well assessed. The cold leak can be enhanced by the earthquake.

-___.__-c..

~

0 0 O i Thermal fatigue of PWR branch lines is

. becoming a safety issue (continued). t f

• ASME Section XI inspection requirements are not 4 adequate to detect fatigue cracking. .

. Detection of thermal fatigue cracks is difficult when the ,

plant is shut down.

t

= Observed fast - ack growth rates can make inservice inspection ineffective.

cescess 25

O O O i

4 PWR Primary System Leaks Findings .

• Trends

• Degradation mechanisms, failure modes, and locations identified since 1985

• Leaks having potential for relatively rapid growth

• Safety significance of leaks caused by thermal fatigue Da3% * . Relevant information for risk-informed inspection programs

• Leak events as core damage precursors Effectiveness of current leak detection systems c=

~ ~

.O O O Our review identified several safety-significant, susceptible sites in piping. Inspection of some of these sites is currently not required.

t Leakage has occurred only from smaller diameter

(< 10 in.) piping.

i

= Affected materials are austenitic stainless steels and .

Alloy 600.

Fatigue and stress corrosion cracking mechanisms have caused leakage from piping.

Vibratory fatigue cracking is the dominant mechanism.

= Leakage has generally occurred through piping welds ~.

~

• But in the case of primary water stress corrosion cracking and thermal fatigue, it has also occurred through base metal. _ , ,

O O O Our review identified several safety-significant, i susceptible sites in piping. Inspection of some of these sites is currently not required (continued).

• Through-wall cracking has been generally initiated on the inside surface. But Code does not require volumetric inspection of Class 1 piping smaller than 4 in.

= Fast growing fatigue cracks may require monitoring of pipe wall temperatures and pipe vibrations in addition to the risk-informed inspections.

c= = .

O o o PWR Primary System Leaks Findings

• Trends

• Degradation mechanisms, failure modes, and locations identified since 1985

• Leaks having potential for relatively rapid growth

• Safety significance of leaks caused by thermal fatigue Relevant information for risk-informed inspection programs leek

• Leak events as core damage precursors

• Effectiveness of current leak detection systems

~

O O O Three reportable leak events could be regarded as core damage precursor.

Nine events analyzed in the Accident Sequence Precursor Program

= Four events were analyzed for reasons other than leaks

~

Fifth event caused a reactor trip During the sixth event, leak repair action disabled a high pressure injection system

. A transient-induced leak occurred during the seventh event

• Large leakage took place during the last two events i

o o o

. Risk Impact Categories

1. Does not contribute to any risk parameter

2. Contributes'to reactor trip initiating event frequency

3. Contributes to PRA-modeled system unavailability

4. Contributes to both trip initiating frequency and system unavailability

5. Contributes to transient-induced LOCA

, probability

6. Contributes directly to SBLOCA initiating event frequency

Leak-induced transients have no discernible influence on the core damage frequency.

Frequency of leak-induced transients

- 4.7E-3/ reactor year

= Overall transient initiating event frequency ,

l - 1 to 2/ reactor year 4

i

. O O O i

Failures caused by RCS leaks have small influence on safety system unavailability.

• PRA-modeled system unavailability because of leaks

- 8.5E-5/ single train / reactor year

- 4.0E-5/ multiple train / reactor year i

. Overall safety system failure unavailability

- 1 E-2/ single train / reactor year

- 5E-4/ multiple train / reactor year

! . Safety system reliability studies being performed by AEOD will provide a better basis for comparison.

O O O '

The observed frequency of transient-induced relief  !

valve LOCAs is an order of magnitude lower than  ;

that calculated in typical PRA.

. Observed one transient-induced relief valve LOCA in

. t the primary. system leak project

. Observed frequency in AEOD Initiating Event Project

- 5.1 E-3/ reactor year

. Mean PRA frequency

- 7.8E-2 / reactor year

. A reactor coolant pump seal leak event becoming a small LOCA does not exist in the database

' ~

~

~O O O l The observed frequency of SBLOCA initiating event is an order of magnitude smaller than those used in typical PRAs.

• Observed one SBLOCA initiating event in the primary system leak project Mean IPE frequency of SBLOCA l - ~1.0E-2/ reactor year l . AEOD Initiating Event Project will provide a better basis for comparison.  ;

t

PWR Primary System Leaks Findings

• Trends ,

• Degradation mechanisms, failure modes, and locations identified since 1985

• Leaks having potential for relatively rapid growth

• Safety significance of leaks caused by thermal fatigue a Relevant information for risk-informed inspection programs

. Leak events as core damage precursors

% = Effectiveness ~of current leak detection systems cosaan m

~ '

O O O

~

Distribution of reportable leak events >

occurred inside conta.inment.

50 45 41 g4 38 so i iso 3

i o

20 5 i 10 9 5_

1 1 1 1

/o/e/p/p/e/*/*/e/p/

/

c /

• 44/ Leakage rate CM 0835 37 i

o o o

~ -

There .is a stat.is tically s.ignif.icant decreasing trend in frequencies of the .

leak events with > 1-9Pm leak rate. _

0.70 _

i Point est. and 90% confidence interval 1 0.60 i Fitted rate {

90% confidence band on the fitted rate :

0.50  : s  :

o \  :

\

0.40  : N  :

N  :

O.30 i \ iN i 0.20

~~,

's N'N i 7

s_Ns x, s' , . .

N[Q r] r-*

0.10.3 , o 5 o

w, d e .-- - - - -

0.00  :

i

-0.10 84 85 86 87 88 89 90 91 92 93 94 95 96 97 slope b = -0.2617 p-value = 0.0001 --

I

.-___-___-__-_________-_-______-___-__--___-___-__-_____-----____c

' ~

O O O

~

Some of the current leak detection systems can detect 1-gpm leak within an hour.

• A sump level monitor can detect a 1-gpm leak within an hour.

• An airborne particulate radioactivity monitor can detect a 1-gpm leak within an hour provided there is an absence of fuel failure and a presence of corrosion product activity.

• An airborne gaseous radioactivity monitor cannot detect a 1-gpm leak within an hour because of the improved fuel performance.

• Containment air cooler condensate flow rate monitor can detect a 1-gpm leakage from a high energy line. The results may be inaccurate if plant operating conditions change.

~

~ 'O O O

~

i Advanced leak detection systems can detect a very small leakage and locate a leak source.

• Directly installed on the reactor pressure boundary

• The nitrogen-13 monitor can detect a leak of 0.005 gpm in one hour.

The acoustic monitoring system can detect a small '

leak (0.003 gpm) and locate it by comparing the leak noise measured at various monitors.

The local humidity monitoring system can reliably detect a 0.0002-gpm leak in 15 minutes.

ces assa ao

.9 9 9 Summary a Reportable leak events show a statistically significant '

decreasing trend in calendar time. The leak events associated with valve packing have mainly contributed to this trend.

• Thermal fatigue has caused through-wall cracking in the base metal.

= Thermodynamic phenomena that caused thermal fatigue cracking in PWR branch lines are not yet well understood.

• Leaks through thermal fatigue cracks in branch lines could lead to SBLOCA.

O O O Summary (continued)

• Fast growing fatigue cracks may require monitoring of pipe wall temperatures and pipe vibrations.

• The risk impacts observed in this study and in the recent initiating event frequency study are lower than

what are reflected in the typical full-scope PRA.

1

• A sump level monitor and, under certain conditions, an airborne particulate radioactivity monitor can detect a j 1-gpm leak within an hour.

4

1 O) t Use of Alternative Source Term at Operating Reactors Proposed Rule to Amend 10 CFR M,, %

g '

Parts 21,50, and 54 '". ]

g ,

b# #+ ~

4

• it Presentation Before Advisory Committee on Reactor Safeguards November 4,1998 Stephen F. LaVie

/"ZU#ihar.U n, m. m m -- ~ ~w

~

O .

objectives

• Provide a Regulatory Framework for the Voluntary implementation of an Alternative SourceTerm (e.g.,

NUREG-1465) as a Change to the Design Basis at Currently Licensed Power Reactors,Thereby Enabling Potential Cost-Beneficial Licensing Actions While Continuing to Maintain Safety Margins and Defense in Depth.

• Maintain Licensing Basis For Operating Reactors That Continue to UseTID-14844.

rh cy .:=. - - - - . . - -

3 1

650.67 Accident SourceTerm ..E' (a) AppIlcability. The requirements of this section apply to all holders of operating licenses issued prior to January 10,1997, who seek to revise the current accident source term usedin their design basis radiologicalanalyses.

• New Section to 10 CFR Part 50.

oDate Selected for Consistency with 550.34 and Part 100.

Applicants for Construction Permits or Combined Operating Licenses After January 10,19g7 are Already Required to Meet TEDE Dose Guidelines.

oNew Section Used to Co-locate Relevant Requirements in Part 50; Separating Plant Design from Siting.

/*

@g)

__ ~ ~ ~~c.- --- w -

0 l

l l 950.67 Accident SourceTerm (b) Requirements. (1) A licensee who seeks to revise its current accident source term in design basis accidents shall apply for a license amendment under 950.90. The application shall contain an evaluation of the consequences of applicable design basis accidents previously analyzed in the safety analysis report.

• Staff has Determined Prior Review is Necessary as Change in Source Term is a Significant Change in Methodology.

i

• Revised Source Term is Expected to Result in Decreased Consequences.

• $50.59 Analysis Could Conclude No USQ, and the Proposed

/*g Change and Any Accompanying Plant Modification May Not l g Receive Staff Review Deemed Necessary.

.:.:a O

~ ~ -n <-- -

2 l

c U

$50.67 Accident SourceTerm (b) Requirements. (1) A Ilcensee who seeks to revise its current accident source term in design basis accidents shall apply for a license amendment l under 950.90. The appilcation shall contain an evaluation of the consequences of appilcable design basis accidents previously analyzed in the safety analysis report.

A Staff intent is Not to Require Complete Re-Calculation of Accident Consequences. OnlyThose Affected by Change Need to be Considered.

A" Evaluation" was Used to Avoid Connotation that Calculations Were Needed in All Ceses.

DBAs Limited to Those in Current Design Basis es Reflected p%g

• by FSAR.

a .. . ~ - , - - ~ -

O 950.67 Accident SourceTerm

& E ASA.FZ Guidesim (2) The Commlesion mayleave the amendment onlyif he p applicant's analyele demonstrates with reneonable assurance that:

E 6fie ed (I) An individuallocated at anypoint on the boundary of in AF400 th* 0xcluelCn aret kr any2 hour period bliowing the Proceedtag. Fusl onset of the postulated Reelon pmduct relenee, would not Derivation Provided receive a radiaNon done in excess of 25 rem (2) total in s.o.c. ettlecove does equivalent (1EM).

A D*** GUI'*H " *** (II) AnIndivir)uallocatedatanypointon theouter j boundary of bse lowpopulation zone, who is exposed to pu n ,g ,g p,,,, the radlonctive cloud resulting kom the postulated flesion smag pmduct relenee (during the entire period ofits peeenge).

would not receive a radiation dose in excess of 25 rem total eNecuwo done equivalent (TEM),

(Ill) Adequate radiation protection le pmvfded to permit access and occupancy of the controlmom under accident f% condluone without personnelreceiving radlaNon f exposureeIn excess of 5 rem totaleNective dose y equivalent (TEDE) kr the duration of the accident D  :.:: . . ~ ~ ~ - , - - ~ -

q 3

O 1

L 950.67 Accident SourceTerm FOOTNOTES 1 The fission product relenee assumed for these 4 Structure of Footnotee hypothesized for purposes of design analysee or 1 k sdentices to nosein Current $100.11 end Postulated from considerations of possINe accidental l gso.u, events, that would result in potential hazarde not l r exceeded by those from any accident considered 4 Footnote 1 le identical. credible. Such accidents have generally been assumed l b Footnote 2 wee revioed to resultin substantialmeltdown of the core with i to remove enplenatort subsequent rolesee of appreclaNe quantitles of fission text reeerding basis of products.

eclectionof 25 rom.

Thie information le in 2 The use of 25 rem TEDE trrtheee eccidenther S.O.C. Remaining test - y " .:: Is notIntended to imply that this value

• constitutes an acceptaNe limit for emergency doses to l

• h* the puNic under accident conditions. Rather, this 25 rem TEDE value has been statedin these guides se a reference value, which can be used in the evaluation of proposed deelgn beels changee with roepect to

'y potentialreactor accidents of exceedinglylow y probabilliy of occurrence, and low risk of puNic exposure to radiation.

\.....J

_, .~~.e,-~

e 950.2 Definitions r

Add:

Source term refers to the magnitude and composition of radionucIldes released from the reactor core, their physical and chemical form, and the timing of their release.

• Added to Establish the Characteristics of the Revised Source Term

'Ihat Need to be Considered.

l l

t 4 N 4

[  !

1 O,

Conforming Changes for Source Term Add Reference to New $50.67 in the Following Phrase: l l

"...comparable to those referred to in f50.34(a)(1),

950.67(bX2), or 9100.11 of this chapter, as l appilcable.... "

AThis Phrase Appears in:

j A 521.3 Definition of Basic Component _ i cl50.2 Definition of Basic Component al50.49 Definition of Safety Related i

cl50.65 Scope of Maintenance Rule Monitoring Program I 4 554.4 Scope of License Renewal Applicability

/*g W

A), ., - , _ .,_ _ .__

l O 1 l

I l

Other Conforming Changes  !

Purpose:

Eliminate Need for Exemptions for Future I CP, COL, DC Applicants.

• $50.34(f), AdditionalTMI-related Requirements.

o Change *... TID-f 4844 source term..'to Read *... accident source term. "in Four Paragraphs.

• Add Footnote to Generically Define " accident source term..."

(Used Language from Part f 00 Footnote.)

• 10 CFR Part 50, Appendix A, GDC-19.

• Add 5 remTEDE Criterion.

l

/*g n /

.;= . ~ - - - - -

r 5

Regulatory Analysis- -

Alternatives - Rule

• No Action.

• Rejected.Would Require Exemptions to implement Desired CBLAs

• Use Current Dose Guidelines and Criteria.

• Rejected. Commission Directed Use of TEDE, Worst 2-hours.

• Replace Existing $100.11 and GDC-19.

• Rejected. Need to Maintain CLB for Plants that Retain TID-14844.

• New Section in Part 50.

oSelected. Meets Needs and Separates Design Criteria from Siting, 6

L(fm~a#;)., . - ~ ~ ~ - - - _ _ _

e Regulatory Analysis -

Alternatives - R.G.

• No Action.

• Rejected. Lack of Clear Regulatory Guidance Would increase l

Costs to Licensees and Staff; would incur delays; Could Cause Less Consistent Regulatory implementation.

• Replace Existing Regulatory Guides (e.g., RG 1.3,1.4, 1.25,..)

• Rejected. Need to Maintain CLB for Plants that RetainTID 14844.

Alssue New Regulatory Guide.

!

• Selected. Applicable Only to Plants that Adopt NUREG-1465.Would l

Endorse NUREG-1465 as an Acceptable Attemative; Provide Updated Analysis Assumptions.

f*\

(M m). .- ~ ~ _.---- - _ g 6

l .. - .

1.

f dc Regulatory Analysis -

I-VaFue / Impact l

• The Staff Did Not Prepare a Detailed Cost Analysis.

• This is a Voluntary Rulemaking. - NOT a Backfit

• Source Term May Not Have,in of itself, Cost Benefit.

• Cost Benefit Comes From Plant Modifications Made Possible By Alternative sourceTerm.

l

• Licensees Won't Pursue Alternative SourceTerm Unless They l Perceive it in Their Benefit to Do so.

• Informal NEl Survey indicated 41 Plants Plan Modifications Based on Alternative SourceTerm.

• May Be Concomitant improvements in overall Safety; Reductions I in Occupational Exposure.

• Wide Range of Possible Applications Make Quantitative

f. Cost-Benefit Analyses infeasible.

l N-). . ~ - ~ - - - ~ -

lO Regulatory Analysis - Safety

• Actual Accident Sequence / Release is Not Changed; Only the Analysis Assumptions Are Changed.

AUse of Alternative SourceTerm, Alone,Can Not l Increase CDF, LERF, or Doses.

• Alternative SourceTerm Could be Used to Justify a Plant ModificationThat Could increase CDF, LERF, or Doses.

• These Changes Are subject to Existing Regulations.

• Regulatory Guide Will Address Risk Considerations.

i (o./ m. _ _ _ _ _ _ _ _

l l

,e Regulatory Analysis - Safety

• Re-baselining Study Did Not identify any Significant Problems with Potential Plant Modifications Based j on Alternative SourceTerm. j l

• MELCOR Runs ShoWed Margin Still Exists. -

i

• Only Approved Alternative is NUREG-1465.

• Any Future Proposed Alternative Source Term Will be Subjected to i s Same Level of Scutiny se Wee NUREG-1465. l oExpectation That Future Proposed Alternettve SourceTerm Will involve the Some Magnitude of Risk se NUREG 1465.

i A

@g -)...

.;;a

.~----e----

O Environmental Assessment

• Use of Alternative SourceTerm, Alone,Can Not increase CDF, LERF, or Doses.

• Plant Modifications Based on Alternative Source Term Are Subject to Existing Regulations.

• Protection of Public Health and Safety is Not Decreased by Proposed Rulemaking.

• TEDE Criteria Previously Determined Not to involve Significant impact (Part 52, Part 100 Rulemaking).

• .. Environmental Impact Statement is Not Required.

_m . _ - . - - _ _ - .

8

. . .. - . _ _ _ _ _ -__=______=__-_____-_____________-_____-_