ML20210E794
| ML20210E794 | |
| Person / Time | |
|---|---|
| Issue date: | 07/16/1999 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML20210E754 | List: |
| References | |
| NUDOCS 9907290031 | |
| Download: ML20210E794 (290) | |
Text
{{#Wiki_filter:m.[ an Q,g e:Qb fx , n$$ y w q e _u
' $m
.m v :m m_&s mn m mn n$ry.mu ha m.:
g w ans fe m&yow Q & _lws.,m$m?&,gg;h.b w y1>o ,nos V_h_,nh; _ Q a.. ,7- % :yf._ h f & n[ [ % > g ? Y,w-i
. PM @i ace .A n %). w dQ?W Xft g %-
Y}. TM A #*M ' 4
~< .-
M:, N'G 4 ~wnn at s , m~ -- a e~ % s<
. .e-e 79W@ Q Q $ h J y W4hN }M $s @h && $y R
&&;!W M g LG $w lJL ?ni' Q 2:?? [kmW 1%
#y$[dw%W :$
n+
'p w~
q m+.'% U) $wQm sa mme @ m M%x a m e uw amm. w< wwn&OFFICIAb?TRANSCRIPEOMPROCEEDIN B9m: e n.~ w
,? o ' s "- o .- ww 4 " ' av ,4 w--:-
vm: w nr~ x, w w, ~ a w +mw
,7 . . v. ,m > / ,
m nn
- s<o. <f& s a '
wv R @WaNW -
%'$W&n&mu&&a ss QQg_5_NU,CELAR-{RE. GUL,ATORE, ,7,CO V
+, ng W
n N Qw m ,&vQh g m W ar mws:s & ;".l @ .n
% ) &-:/ m s - M%VY m
.s,~ : Q. $, . -$ b m
M , ,,,9 @m'"
' & W~lgy Qq
~ 9
~Y.a.
a&v:4;;? h
.M@, m;. r ,wt,s W.Y&<*v
> -:- yn % ^'
n ;b a Nh?h>W qR% ?)H::M :Nr;CMY, ; $ M, E, W, ,v n a V, om%ayf? .30 3~., Q: ' 1 ey g~ g%:: m n w u n i e, $ potx1% wl#a ;w ,m.,T ',) r r 3 , m. . u s eWNW a m m m m a n e x c r o n s e c o. s$ urssroxrwa sc w. & claw Y wkwnwg? gyWWM W ww wm% w;w i 4
% m%a tN wWTW C/39r%y% p m w x . x%WP:h.iaMWNW @ g %g x a w A 4. y-a 6 , as ; M:mwp QWQ ape . 3 n n Wk yu q , y cy, pn+
VM#d% g MyggWe~% n y pg u m gmn w.rg w+i6 h #, %n?7e 4 % PO, u m
- NdhkDMMxdg/nWfW
- n _Ih_%UMQNh. fM( %
pV & M% ISKi.WORKSH, sng; miR,%- yr s L ; > y m;! OE,dNW"
- Qq < '
yc - NM@f@~i
.- v
. e ,
g% ~ vw ;$NbM N _ q M q gh 7 M % %; W f\,;$[D $@+M,[M _i;& , s . 1 W
.- .n 4hN yk w w nn .~n
&h )W $$ ,i . ~a n.s
.tei nnm$fpkhh$~~:a n
kk th
- a. i ,
n~ = ~ . ,n w %gpl&m!Q?$h&m_p$w]$hw)C w A m . ~ 4)j g,A,7,$w}. [hg ^ 'V w% (i '
- ew m g e w% Q m Y
%u Qp@+z&.
$Mw{n Vl f QWW[%nkQl_[% 'h&,0,- QE(lSku. b,h h? ' j ,
~
. m," mm /M,- maa n
4 g <e q.ggm. m :y y: p 'p w m y w;;9: 8
. P y fy .Q.y gQ.gx QQ m m. !g m Ag e;L'u%*iS.X.g$r A. , m g u 4 Q 7 3;p . g@p@M'gi
.* j;hf gf.
't, s iNE@98MW?
wp - 4 Q pn g mmen m yw . w'M& gkps$m;unjNpQM8 PjMf&y$g $m :gg% f vwM&e g MT$4 ?W;3;9po q ' bygg@f
- fw?p, w Eg
, , ym~
n y 48?g w yys;p'% j; pm pp3WQQ y$n e ma%w%m%g%a%k MyqyQ$ w%$b g VMjh
$O w J %%%% L M@ a@$n A s+w&%%p W,3 n?M, m#f , y, h h ay}
w .m
'9% M q:
.gmb QQf mmm4om Q
q s >e n-:,
.ms g.. w:p y:)
px n QV;Q&}S-g yt
- agg W
gy nqu u,w p:pimn&. y x W 21ym-mmb.a m m. ey5 m my
,_up up_x 4;Aq.;
\
Q.%w y w n w~ 4, .: e t W0 g , e3 ~9 Q . -;ng i g;m&:papqA fse p f .@3 7 n n + q7 x . g ,7,.y yy' .
-- w -
ya e x,,
,1
> miwy- ag. ,i i ., i w w%.
2"y NW 3,, D g,g7N 5; +, QQRW ],)N)g'*Wa,7,gp . mm -QE}d;n va g. ,;s.e w ss:qs i w Q. xl; & , m:,)n*,' hfp y>; Fc?i?:: w u~g;.Qj:~ w %.y* i "' g7 g. .am e z f g& pj qW- n. t . 4
-w m n:n
,rt;g% 7p'y y -:pnw%.. ;g. g;vt.i'aN+s e.g w". y.,17 yp p1 we, g p ;a e ..
cN- 1 , v r 4r n . .xy &;nappvg m. 3 g+2 '.ay ,;. .y-_up p.A. qp,. 4a. ? w~ w An#, : e 3p? wa :m
.m&.w.
&,c m W> ><=. ,>
c , Wmx ct W,$c ..:. a%v , , ysf m h hi o w t v o. .
~s
.'g n? :
1 or 4 .
;t av ~ , mo m~ ,d .%, v+aN 3.4. i; a.et 3,y fi-m:, w ,
, 4s.m. w .
- h. m!di.h,g.k m i. x?
! IN 4 ; u , mM ; M, k's m.
1 i x
$ " 's . ~ -alh. u !m. a
- Mj f ~ ..-<nt ",as v# <ra e c>..s o&, . , > ,/ t "A-2Nw
- s' m m 1 ,
w, t h'.:.' ; M' S. .t,y q - 'bI - b:p @p" : aJhp# 4 4T5( t q ue e ,.[. t
,' m . w' s. . '. p a.
ll jf._' h. y .. .Y . y T f n a , _v t+; ..s# a.m if cb U f ,.
'N2 M 'x? >
m
> i e
n,
%} L ,
t * ! 7n $ $m.b MWids%aw! wT 'smy w4sam.i;% c db Q g n.gQw Q mLf QQye iw3%:%n M w2 GLOCyn10W< i:*;:n g.M c m (t m.ef RT,3 ,m T
~
QU c' 'fww , w* sm
&@$,Cathmburg@m' ..tp ~ ~ .
'Dz % $ c g,,i$. m
. ~g :(~ , ..
/ % a, 3 i
M' ~N 3 ^.-a om. p.
~ M' 2 wa :nw , 1 JMue% 1.pn
' y
>@ ' w? ~
y ' r, :.h n. . V
,a -
w p Wi$e&YW&g&p{NQQ u m&p%,)O4ptm t QQlWjW e ?lflQ v Vg, ,&h
. g ', '
s ,1' w( +'.^ 4g fhQ&,ly. . %p/e"" ;,R m i s, l'p.m y - > a ' Yhh0m+ a,m.+hff)-hb,f, h)s w w x w a m &,%na$. J: u , myb "" &u yWn lw ' w(f;>*
~Q.
' Yli,.)
- n ~
we~~w. v,~ vy hj gg ' " " y yxg.wyMp mm. mhTTIdA7[dMIhIdSS$k,> ,mo,,y,y;in,n
+
r :a w m< n Lpr e w:m u . 3gu x:~ w n n -w . PACESQ4M474g, , gse v, q_p r,n e.h. acnwau v,,. ,m~p c ,w yQ m g .e.a im eypW M.. e sy .Mpp;
,,w ,ma g g r ,eo . m. . ~n &m, w,, , 4 ne .< s y m ,. .
- n uyp my qy ,y.W), , . s y m v Wp
- mpL3 a u ,
"m a h:vaw m%y.
. s m,:..0eVN
;n n u r >, g, 1vn-
. vp w,wz z ,
y: ,sx . >
< y qp w.
. ., . t .- , ar.N:. %,
s 4 u , , - ;% m _?dm y w QQm _Q t , pm w' % "_, '4 .s'pp _Q:V'w fytgeytN kln;; v % yl:[N
>% p Tm'q.y;.u:
W_ 3 4 t
- i
..:q .
. v,Q;x d;'
Ap _p _
+s ;_p.y r _r m
7 w N-n.w W a8 p %.Q&Vf?Qi 4 pM4ydw&& J p$! wa % 1 i GNN.ghWS&p%:q r d 4 f1p gf_ug.j:c-yp
%) $y >L . >q . 3s p f;;;'-m$p.
,t c; < r' !:31 n s
< g(,i da ...'i ' 9
@$%m%]lgj QQ&:hpm&m /37 Ug.
4 NQ'& f c fym4 WVQW D r Li4 &x%pq:),W' M r3h > g W, 29': @c+ s t-t n pWo. %p %m%Q a y "h ~ d u
%y :
R:>: .V. nA ;],~D , M. Yp&a?
? x M p b g g s q q p W /[ $e N O :,T W:u% e @@h k h
,,.% % 09" a cs hh hhhhhkb kh;J w $kh h% N hf , q w s N vih 4 m Y:-M"' g'Z #r h @A se-Am g h;st w asm d M' hf . md [ kW: k,
- e C^ D w@#
w w$ d @p %CQ % w w' W1625 W@4 Connedkut Anoue,NW,5u s , . 2MW #4 MVMyyQ fQK QQM g 5% w 4. Mewm."):Qf;4gyd:g:a,.& vpy mm m c
- w. L n:v, w.y p
Q. x . n4: w. .,v en v ym:~a n u w w .,w min &q m,-u} {/. udn2 ,R QWp&j
..Q . Jp +m '
w %Wanklishi40.CS200361 am y
. l%~$
u e ,
,, ' ,' GQ w &w* ,T ,V: ,
o u v c n w.dy n:;- .ev ar <.
, y
~-a ~
' ' vm; Q, f i p%,;%a) m 'w' [
wq <a< u. :
,9. w 7 .h np w x% JW ?
ny +
,v w% ~
4# M #[ 9907290031 990727 l Whdf$$ PDR 10CFR W w. W W
. e , 4 >
zu, . y&%$Qq$y G PT9 w w 7m w e m - - PDR- - M @n Le,i pp :7M :y, , % '
+
siihld
%Q ~ ., ~
l@
OftlG!lsjAl_ i O' OFFICIAL TRANSCRIPT OF PROCEEDINGS UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION
Title:
REACTOR DECOMMISSIONING RISK WORKSIIOP O Work Order No.: ASB-300-857 LOCATION: Gaithersburg, MD DATE: Friday, July 16,1999 PAGES: 241 - 474 ANN RILEY & ASSOCIATES, LTD.
- 1025 Connecticut Avenue,NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 g qgy3 eeo m PT9 '
PDR
F" 241 1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION 3 *** 4
'S REACTOR DECOMMISSIONING RISK WORKSHOP 6
7 8 9 !!oliday Inn Hotel 10 2 Montgomery Village Avenue 11 Gaithersburg, MD 12 13 Friday, July 16, 1999 14
) 15 16 The above-entitled workshop commenced, pursuant to 17 notice, at 8:19 a.m.
18 19 20 21 22 23 24 25 O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
242 1 1 PROCEEDINGS () 2 3 MR. CAMERON: [8:19 a.m.] We're going to get started, but 4 we're going to give Gary Holahan a minute to get in here. I 5 thought what we could do before that, I want to give you a 6 little reprise of yesterday and I would emphasize the 7 little, and then talk about where we might go today. 8 But we do have some new people at the table with 9 us this morning and I just would like them to introduce 10 themselves. Suzy? I 11 MS. BLACK: My name is Suzy Black. I'm the Deputy 12 Director, Division of Licensing, Project Management. 13 MR. CAMERON: Thanks, Suzy. And let's go down the 14 line. () 15 MR. BAGCHI: I'm Goutam Bagchi. I'm an Advisor to 16 the Division of Engineering. 17 MR. ROTHMAN: I'm Bob Rothman. I'm a seismologist 18 in the Division of Engineering. 19 MR. CAMERON: Dr. Kennedy. 20 MR. KENNEDY: Bob Kennedy. I'm a consultant in 21 the area of seismic engineering. 22 MR. CAMERON: And a consultant for NRC. 23 MR. KENNEDY: On this project, for NRC. 24 MR. CAMERON: All right. 25 MR. BAGCHI: In terms of acknowledgement, we need , i ANN RILEY & ASSOCIATES, LTD. \ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 i (202) 842-0034 l 1 l
243 1 to acknowledge that Dr. Kennedy is going to be the () 2 3 independent reviewer. He has not been briefed by the staff and his availability is made through the Office of Nuclear 4 Regulatory Research. Dr. John Staudenmeier is also here 5 from the national lab and is going to work with Dr. Kennedy 6 on this project. Dr. Wah Chang is the Project Manager from 7 Office of Research and he is also here. 8 I just wanted to emphasize the fact that Dr. 9 Kennedy is going to be the independent reviewer and he has 10 not been briefed by the staff. 11 MR. CAMERON: Thanks. While we're on the point of 12 clarification, Dr. Kennedy, are you associated with any 13 particular institution? 14 DR. KENNEDY: Just my own company, RPK Structural O(_,f 15 Mechanics Consulting. 16 MR. CAMERON: Thank you. And let's go over here, 17 and we'll have to share this. You are with us here at the 18 table today. Go ahead. 19 MR. HUBBARD: George Hubbard, Plant Systems 20 Branch, NRC. 21 MR. ATHERTON: My name is Peter James Atherton. I 22 serve as the nuclear safety consultant for the public. 23 MR. CAMERON: Okay. Thank you. And for the 24 benefit of the stenographer, Jon, you see people have 25 switched places. O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
244 1 REPORTER: Yes, thank you. () 2 3 MR. CAMERON: I think that we had a lot of productive discussions yesterday. Yesterday morning, we 4 started out with a discussion of general issues, including 5 the issue of safeguards and security, and we also had a nice 6 concise description of the NRC staff study by Glen Kelly and 7 a useful critique of that by Ed Burns, who is still with us 8 over there. 9 And George Zinke and his colleagues did an 10 impressive job, I think, of presenting some real useful real 11 life information for us, and then we sort of went into a 12 recalibration phase, which I guess is maybe the kindest word 13 to describe it. 14 But it focused on the need, the feasibility of the 15 NRC, taking the information that was brought to us yesterday l 16 by the industry and attempting to use that, factor that into 17 some type of requantification of the study. 18 And just to make sure that we're all in the same 19 place on that, I'm going to ask Gary Holahan to give us what 20 the staff perspective is on that, and I think we'll probably 21 have a little discussion of that and where we should go from 22 there, not only in the future, but in terms of today's 23 agenda. 24 We're sort of behind in terms of the agenda, but 25 that's fine, as long as we're still making progress. But I ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
)
j
245 1 would like to get to the seismic. There are some views that () 2 that may be resolved more quickly than the time we have on 3 the agenda, but we need to have a sort of agenda discussion. 4 But I would just remind everybody that we are 5 short on' time and we'll really have to stay focused on the 6 subjects that we're talking about today. 7 There is a session at the end today where we can 8 have a more wide-ranging discussion of-issues of concern, 9 but we do need to stay focused. 10 Gary? 11 MR. HOLAHAN: Thank you, Chip. As you probably l 12 recall, I wasn't here yesterday afternoon. My understanding 13 is that yesterday's issues had to do with the first 14 presentations by the utilities on what they're actually () 15' doing with respect to their organizational and human 16 activities. 17 I think it's fair to say most people who gave me 18 their views of it were very impressed by what the utilities 19 were doing and, in my mind, what that does is it says the 20 utilities can, in fact, perform at a level beyond what we 21 assumed in the analysis. 22 So there is room to give credit if, in fact, we 23 know how to do that. 24 I think, also, it's worth pointing out that these 1 25 are difficult issues to talk about and sometimes the ANN RILEY & ASSOCIATES, LTD. O' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
1 1 246 1 examples sort of engender emotional reactions. That's () 2 3 because really what we're talking about is the difference between good operators or good organization and one that's 4 nearly perfect. 5 But when we start talking about examples, the 6 examples don't sound like the difference between a good 7 organization and a perfect one. They sound like terrible 8 examples. I think some of the examples, and I won't even 9 repeat them because I think they cause more problems than 10 they solve, are sort of distractions and we're not saying 11 that operators are not paying attention to their job or the 12 organizations are committed to safety. 13 What we're trying to do is distinguish between 14 what would make an organization or operator, beyond what a l () 15 normal human reliability analysis credits for operator 16 actions, which leads to this point of where should we go 17 from here, the requantification of the human reliability 18 analysis. 19 And what I would propose is actually that we start 20 at the other end of the problem. We had some information 21 yesterday provided by Ed Burns. We are, in fact, getting 22 independent expert input. The first thing we ought to 23 reevaluate and requantify is the non-human reliability 24 analysis, so we know what sort of challenges and what sort 25 of expectations we ought to have for the operators. l l
) ANN RILEY & ASSOCIATES, LTD.
[~'/ (- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L _ -
247 1 So we ought to relook at initiating event () 2 3 frequencies and availability of equipment and those sorts of issues, because what I propose as a second step is even 4 though some of what we're talking about here is beyond what 5 I think of as the current state-of-the-art in human 6 reliability analysis, I think we are willing to push the 7 state-of-the-art, but we need to recognize, to do that, we 8 will have to spend some time, some money and invest some 9 expertise that I think hasn't yet been brought to this 10 proj ect . 11 What I would suggest is we look at initiating 12 event frequencies, we look at the thermal hydraulic analysis 13 and how much time is available and all those sorts of 14 issues, so that we have available a more solid set of (Q, 15 information to give the human reliability expert community. 16 On Ed Burns' slide yesterday, he suggested Gareth 17 Parry was one of those experts. I'm inclined to agree to 18 that. But this is a sufficiently difficult sub-; .t, that I 19 don't want any one of our experts or anybody's experts 20 making decisions on their own. 21 So I think some follow-up activity, focused on a 27 clearer question of what is our expectation for the 23 operators, needs to go on, and whether that's a separate 24 study or a workshop or an expert elicitation or whatever it 25 is, I think we need to sort that out. ( ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
j 248 1 I would like to see interested parties offering ( 2 real expertise in the area of human reliability analysis, 3 combined with the kind of information that the utilities 4 brought as to what they're actually doing and what they're 5 capable of doing. 6 And I think the commitment that we could make is 7 we will requantify, to the extent that the state-of-the-art 8 and the experts in this area say that that's the right thing 9 to do. 10 Now, we can probably do some requalification. 11 There may be some areas in which the real experts, of which 12 there are not many in this room, but the rei. experts will 13 say we can't quantify beyond a certain point or there's 14 maybe certain types of activities that can't be 15 requantified, and you can only make qualitative judgments. 16 I think we need to ask the experts how far to go 17 on this issue. Then I think it's our commitment to publish 18 a report that has the best information available. We're not 19 interested in publishing a report that we don't really 20 believe in. So what I imagine is sort of a staged process; 21 let's get the best information available on the other parts 22 of the issue, let's give that information to the expert 23 community, let's get their advice as to how much, how far, 24 what type of requantification can be done, we'll put that in 25 the analysis, and that ought to be -- Alai RILEY & ASSOCIATES, LTD. w' Court Reporters 1025 Connecticut Avenue, NW,' Suite 1014 Washington, D.C. 20036 (202) 842-0034
249 1 To a certain extent, we may be establishing a new () 2 3 state-of-the-art, and that's okay. Now, that takes commitment of resources, time, expertise. We have a few of 4 those experts. We have -- the Office of Research has both 5 expertise and contact to expertise on a contract basis. 6 I think we have resources available. If this was 7 an issue that we were pressed to make a decision next month, 8 I think this-'would not be practical. But we already have a 9 commitment to an ongoing independent technical quality 10 review over the next six months or so. 11 I think we can work this additional activity into 12 that time-frame. But I think that's about where we are. 13 Now, I understand that there is also a sensitivity 14 that we, the utilities, other interested parties will want () 15 to know what kind of activities are most valuable for the 16 operators to do. We may not be able to distinguish between 17 the value of having three operators on shift versus four 18 operators on shift. That level of detail may not be 19 something that even an expert HRA community can tell you. 20 So there's going to have to be some judgments made. 21 But I think the most credible thing we can do is i 22 clarify the issue, bring in the experts, get their best 23 advice and make a commitment to use the best information 24 available. 25 And I'd be very interested in hearing people's f) v ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 1 (202) 842-0034 l l
250 1 views on this subject. I've already discussed it with a () 2 3 number of our PRA and human reliability experts and, frankly, they feel that it's sort of pushing the 4 state-of-the-art, but that it's something that's doable and 5 worthwhile. 6 If there are any other comments, I'd like to hear 7 them. I guess this belongs to Chip. 8 MR. CAMERON: Thank you very much, Gary. And why 9 don't you -- we'll test this microphone out again. But I'd 10 like to go over to Mike Meisner now for comment on that. 11 I'm just taking some notes here to see if whatever 12 we come out of this discussion with, that we'll put that up 13 on the board so everybody knows what it is. 1 14 Mike? () 15 MR. MEISNER: I guess I'll give you a couple of 16 personal comments, off the top of my head, after hearing 17 what Gary said, but I thought that I'd maybe reserve final 18 comments till later, till we can get together and talk as an 19 industry. 20 My view, not being a PRA expert, and I said this 21 when we first got together yesterday morning, was the model 22 the staff put together was pretty good. We were only in 23 disagreement over some of the inputs to the model. 24 I guess what we're hearing now and what I heard 25 last night was basically the staff saying we can trust this ANN RILEY & ASSOCIATES, LTD. [} O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 251 1 model if it gives high numbers as results, but we can't () 2 trust it if it gives low numbers. I'm not sure that's very 3 constructive as a basis for a rulemaking process. 4 One of the things we are all concerned about going 5 into this effort with the Commission charter was the staff 1 6 turning this into a long-term research project. Again, it's 7 just my kneejerk reaction, it looks like that's the way 8 we're going. 9 I, for the life of me, can't understand why you l 10 can't simply take the information that we gave you yesterday 11 and modify the inputs to your model, your PRA numbers. I'm i 12 not sure that this is really such a state-of-the-art or 1 13 pushing the envelope that reasonable PRA practitioners 14 couldn't agree on what this means. () 15 MR. CAMERON: Okay. We're going to go to Gary to 16 address that. It might be useful, if there are other PRA 17 experts out there from industry, academia, the consulting 18 world, whatever, it would be useful to get your input on l 19 this issue, too, so that we can figure out how to move j 20 forward here. 21 Gary? 22 MR. HOLAHAN: I didn't mean to leave you with the 23 impression that we don't trust low numbers, we only trust 24 high numbers. I didn't say those things, so I don't want 25 you to have that impression. ANN RILEY & ASSOCIATES, LTD. O- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 042-0034
252 1 And it seems to me, even if we seem like we're 2 making this more difficult than it needs to be, and if you 3 are right, then it seems to me when we hand this issue to a 4 bunch of-experts, that perhaps they will come back and 5 support your position. 6 But the only thing we've lost is some time and 7 resources in getting there, but I think even if they come 8 out and they fully support your suggestion, I think we have 9 a more credible way of moving forward, so that other people 10 or the rest of the world sees that we've given this -- 11 which, in fact, we've all acknowledged is the most important 12 part of the decommissioning issue. 13 It's really in the absence of automatic systems 14 and all that, we are relying on the organization and the 15 individuals. And I think we need to have people 16 understanding that we've given this our most serious 17 thought. 18 So I think that's -- if it seems a little bit 19 unreasonable to you, I still think it's the best course of 20 action. 21 MR. CAMERON: Before we go back to Mike, I think 22 let's hear from some other people around the table, we'll go 23 to Alan first, and then we want to test out on people the 1 24 proposal that Gary put forward. But let's get some more 25 input. Alan? j h \_/ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 84'-0034
253 1 MR. NELSON: Yes, Alan Nelson, NEI. Yesterday, we () '2 forwarded.a letter to the Commission noting that they delay 3 action on the SECY-99-168, the rulemaking proposal. The 4 reasen being is that we felt that today's workshop was the 5 underpinning of that process ano the rulemaking activity 1 6 that vas moving forward. 1 7 And given that, how would the study or the 8 longer-term activity eventually effect the rulemaking and 9 the Commission vote as we move forward? Because one is -- I 10 see one interlocked with the other. l 11 MR. CAMERON: We're going to give that to Stu 12 Richards. 13 MR. RICHARDS: I think what we would intend to do 14 is to continue to go forward with the idea of establishing a () 15 separate part for rulemaking activities, if you will, to 16 establish the framework. l 17 Of course, it's kind of an empty shell until you 18 start filling it up with the rules that apply. So assuming 19 that the Commission were to give us the green light, we 20 would continue our activities to review Part 50 and the rest 21 of 10 CFR and determine where the various regulations are 22 that apply to decommissioning, to work towards coming out 23 with a rulemaking to establish a new subpart on l l 24 decommissioning. 25 But as you've said, the foundation for some of the O 7JUJ RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l I Washington, D.C. 20036 (202) 842-0034
l 254 1 more significant issues, based on the exemptions of the () 2 3 process, really are going to have to wait until this is put in perspective as far as the risk goes. l l 4 So process-wise, I think we would continue on, as 5 long as the Commission gives us the green light. Obviously, 6 we're not going to go anywhere with EP, security, financial 7 protection, some of those until Gary's effort is complete. 8 MR. CAMERON: Does that answer your question, 9 Alan? 10 MR. RICHARDS: Again, the Commission has the say 11 on this. You've read our paper, we've made a proposal. I 12 don't want to prejudge what they're going to tell us. 13 MR. NELSON: I'm just concerned that the outcome 14 of today's discussions and the ultimate study that, who know () 15 how long it would take, as Mike said, it could take up a 16 protracted length of time, could effect the rulemaking, the 17 rulemaking package, and the focus of the rule itself. 18 MR. CAMERON: Gary, you.seemed to indicate that 19 you were more optimistic about this. 20 MR. HOLAHAN: I am more optistmic, and since 21 yesterday afternoon, we don't have a detailed plan. But in j 22- my mind, we're talking about months, not years, and 23 remember, we've already laid out a schedule that said we 1 24 were going to study this and follow up on this report 25 through the end of the year.
/ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 hashington, D.C. 20036 (202) 842-0034
1 J l' l 255 l. 1 We may be able to do it in that context. It may
) 2 stretch things a few months, but I don't envision this being 3 an impediment to the schedule.
4 I think the other thing I'd like to mention is I 5 don't mean to ignore all the other issues of yesterday, but l 6 I think this is one on which what we're talking about is 7 doing something in addition to what we were going to do 8 before in the other areas. ! 9 People suggested issues like airplane crashes, for L 10 example. I think Mr. Shadis suggested it. That's one of 11 the topics that's already in our report and we will look l 12 into any new information on that subject, or seismic, or -- 13 you know, that's already part of our independent assessment 14 and input. () 15 The reason I raise this human reliability analysis 16 is because I think in addition to just re-reviewing it, we 17 would do something more than we were previously thinking l 18 about. 19 MR. CAMERON: Okay. That's a good point, Gary, 20 and I believe those were some of the issues that Ray Shadis 21 brought forward at the end of yesterday, these other issues. 22 Ray? 23 MR. SRADIS: Please, let's not have it on the 1 24- record that those are issues I brought forward. What I was 25 saying is that there is a great deal of new information l l \ l O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
256 1 regarding the various issues that the staff had listed, and () 2 3. that really that ought to constitute part of the input to determining probabilities on those things. 4 But I.think, also, as long as the door is open 5 here, you really want to take a look at what goes on in a ; 6 decommissioning plant. The staff seemed to limit itself in 7 the outline at least under heavy load drops to cask drops. 8 That was -- at one point, that was the only item listed in 9 one of those cells that you showed yesterday, and there are 10 a number of other situations in which heavy objects can wind 11 up falling into the spent fuel pool. 12 MR. HOLAHAN: Deconstruction type of things. 13 MR. SHADIS: Deconstruction things, you have your 14 overhead crane rail which is pretty heavy, you have masonry () 15 walls which are not seismically qualified, and I think we 16 really need to look at what the impact of that might be, 17 Last. night I was running through different 18 possible scenarios for criticality as a result of an 19 accident, and it occurred to me that in some decommissioning 20 plants, the systems are not segregated as they are at Maine 21 Yankee, and you have chemical stripping of the primary
-22 piping. Somebody may take it, in their mind, that they want 23 to chemically strip the reactor vessel.
L 24 I thought you have a situation in which you have a 25 fairly rapid loss of spent fuel pool coolant &nd someone ( ANN RILEY & ASSOCIATES, LTD.
-%/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I 257 i
)
1 gets on the phone to say what do we do about it and some () 2 other genius decides that maybe there is a lot of liquid 3 available in the primary system. 4 .m trying to remember what my chemistry looked i i 5 like back when, and my grades weren't very good, but I 6 thought, gee, if you dump a lot of caustic material in i 7 there, can you precipitate out the boron. I don't know, but 8 what I'm saying to you is that accident conditions, from my 9 perspective, are conditions of chaos and decisions are not l 10 always made in a clear rational way at that time and the 11 most peculiar things can show up under accident conditions. 12 That falling crane rail could also take out the 13 electricity. So things get complicated in a hurry and I 14 really think that if you're going to go forward to assign () 15 risk probabilities and consequences for this kind of an 16 accident, that you really need to take a -- open the arms up 17 and take a broad sweep and bring in all these potential 18 issues, , 1 19 MR. CAMERON: I think Gary is shaking his head 20 affirmatively. 21 MR. HOLAHAN: I'm shaking my head affirmatively. 22 We didn't put criticality issues as a specific issue on the ! l 23 list. I think it's because it's not fully developed in the 24 draft report. So in a sense, we didn't want to have a 25 meeting on something that the participants didn't have a f \- ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 i (202) 842-0034
258 1 chance to see. Ii 2 Our criticality experts are looking at this issue V 3 and, frankly, the chemical aspect of it is something that I 4 hadn't heard before. I think it's just one more thing you 5 can put on their plate. 6 But they are looking at a whole collection of 7 possible ways of making the spent fuel pool critical. 8 I don't know that we'll have another workshop, but 9 clearly, when we have a draft of that part of the report, 10 we'll put out for comment and we'll want to work through 11 those issues well. 12 MR. CAMERON: Okay. These other issues are useful 13 to discuss, but let's try to keep on the proposal, and it 14 may be too strong a word to call it, Gary, but maybe it () 15- isn't, but to identify it, Gary's proposal. < 16 Ed Burns, what is your thought on that? 17 MR. BURNS: I'd have to say that I agree with much 18 of what Gary said. I think that he summarized pretty much 19 the tact I would take in trying to reexamine what the draft 20 analysis has said. l 21 I'm sort of a pragmatic -- take a pragmatic ! 22 approach to PRA, so his step-wise approach seemed very 23 appropriate and useful. The only few things I would add to 24 it, I would try to relate each of the specific industry 25 practices that would impact it, impact the model O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I 259 1 quantitatively when we get to the HRA part of it, which ones 2- of those are the most important to make sure that the 3 industry understands what the model is saying and that the 4 industry.can properly feed back any information that may 5 influence that. 6 The second point would be that from a quantitative 7 standpoint, I think that by incorporating those inputs into i 8 the model, that you would be able to identify which l 9 sequences might be most important and then which sequences
)
10 you can say are truncated and you can pick a quantitative 11 value that you might truncate those sequences at, so you're 12 not tied basically to some specific number associated with 13 very low frequencies, but pick a way to truncate those 14 sequences and identify how those got trunc.h ted, based on 15 what actual practical things were done or assumed as part of i 16 that analysis. 17 From my own personal point of view, I think that 18 any enhancement to the HRA methodology, as Gary suggested, 19 that from my personal point of view, that's very useful to 20 me, very important to the PRA community and would further 21 the techniques. 22 I'm not sure that it needs to be tied to i 23 decommissioning rulemaking, but that's a judgment for people 24 other than me to make, j 25 MR. CAMERON: Gary, comment on that. 1 ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034
260 1 MR. HOLAHAN: If I was looking around the room, () 2- it's not because my interest was wandering. I was sort of 3 picking up signals from our various experts that say, yeah, 4 your suggestion sounds very reasonable. It's a good way to 5 proceed. 6 MR. CAMERON: The suggestion being two things; 1 7 one, take the specific industry practices, for example, that 8 were talked about yesterday and make sure that that's fed 9 into the PRA and then there was this issue of truncation. 10 MR. HOLAHAN: That's just a technique for dealing i 11 with these. 12 MR. CAMERON: So truncation as related to the l 13 specific industry practices. It's a technique. ' 14 MR. SHERON: The use of truncation I don't think () 15 is important to put up there. It was a -- it's a way to get 16 past the -- get the problem of very low -- showing very low 17 . numbers that are difficult to defend when we don't need to. 18 We don't need to show those. We need to say what's 19 important and why it's important and if you're able to 20' dismiss a sequence or dismiss things based on very low 21 numbers, then I think that's sufficient. 22 MR. CAMERON: All right. 23 MR. HOLAHAN: If I can make a suggestion that we 24 put some new thoughts of various types on the table, on this 25- subject. We are going to have a wrap-up session this O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 I (202) 842-0034
261 l' afternoon. I'd suggest that people spend the rest of the 1 () 2 3 morning and their lunch to think about what's been said and let's pick this up again as part of all the various issues j l t
-4 to do'the wrap-up. j i
5 We're at a relatively thoughtful and calm point 6 and I think this might be a good point to just let it simmer 7 there for a while, pick it up in our wrap-up, and maybe we
, 8 can move on and try to cover seismic and the other things 9 that we said we were going to do.
10 MR. CAMERON: That seems like a good suggestion. 11 Let me ask Mike, for example, is that okay to let it simmer? 12 MR. MEISNER: Sure, let's do that. 13 MR. CAMERON: All right. We're going to let it 14 simmer. We have sort of a crude representation up here
) 15 about this and we'll go back to it. ,
1 16 Do we have any final comments on this issue before l 17 we go to seismic? Glen ~ Kelly. ; 18 MR. KELLY: I wanted to put a little clarification 19 about the independent review that we're going to be having 20 performed of our analysis. That independent review will 21 take into account all of the information that's presented 22 here and any additional information that people want to 23 provide to us. 24 We're also asking that our -- what we've done be 25 looked at from the point of view of its technical adequacy, O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
262 1 as well as whether we have caught the breadth of issues that () 2 should be covered under potential vulnerabilities or 3 problems with spent fuel pools. 4 So as these issues come up, we will make sure that 5 they are.all. covered in our review. Anything that we didn't 6 capture in our program, we'll make sure that it gets covered 7 under there and it'will be reviewed and all addressed in the 8 final report when it comes out. 9 MR. CAMERON: Thank you. Stu, do you have a 10 comment? 11 MR. RICHARDS: Just a clarifier on the criticality 12 issue, and I just wanted to mention that previously, Mr. 13 Atherton has brought this up at other meetings. Like Gary 14 said, it's still under review. We offered Mr. Atherton the () 15 opportunity to meet with us after that issue had been 16 thought out more carefully, and I don't think we've heard 17 back from him yet, but we've offered to hold a separate 18 meeting in.the future and address it at that time. 19 MR. CAMERON: Ray, you have a comment on that? 20 MR. SHADIS: Yes. I feel like the only person I 21 speaking versions around here. I did not bring up the 22 criticality issue, except as an example, speaking primarily 23 about learning what goes on at a decommissioning plant. 24 So that as you develop potential accident 25 scenarios and you're looking then to build on your list of i O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036
.(202) 842-0034 l
263 i l 1 ; initiating events, that's the place to look, because that's () 2 3~ the situation in which this rule or report is going to be applied. 4 That was my point. My point was not necessarily 5 that you need to look at criticality, although it's 6 certainly a wise idea. 7 _MR. CAMERON: Okay. 8 MR. SHADIS: I just wanted to clarify that. 9 MR. CAMERON: Thanks for that clarification. I l 10 think Stu's point is good information for people to know, 11 and we'll stop attributing things to you. I made the 12 mistake first. We'll stop that. 13 Can we get the seismic session going? Is there a f 14 presentation that's going to kick that off? 1 How about that. 4 () 15 Okay.
-16 MR. RICHARDS: I'm the person that's supposed to 17 kick this off, and in the interest of time, I'll keep it 18- 'real short. I think seismic is one of the areas that, even 19 going back to NUREG-1353, was an issue that seemed to be 20 above the others.
21 It's al'so kind-of unique in that nobody is looking 22 to rebuild the spent fuel pool to change the site j i
'23 characteristics. So it seems like the reviews in this area 24 have focused more on what available information there is on 1
-25 seismic data and what it means.
l O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I
264 1 So we're going to have a presentatior by our ( 2 people and then the industry is going to have their 3 presentation. With that, I'll turn it over to our guys. 4 MR. CAMERON: All yours. 1 5 MR. BAGCHI: My name is Goutam Bagchi. I'm with l 6 the staff, as I introduced myself. My co-presenter will 7 also be Gareth Parry. And before I get started with this, I 8 need to tell you that back when, as a young engineer, I 9 started working for a consulting company. Maine Yankee was 10 a plant that was I involved in designing. I remember, I 11 believe, working because of the large penetration in the 12 containment structure, and I had plenty of experience 13 dealing with design of nuclear power plant structures. 14 When I came to this issue, seismic event was 15 perhaps representative of the toughest tree in the forest. 16 And I started looking at it and I said to myself, I really 17 need to pay attention to two sides of the equation. One is i 19 what is the demand of the structure and what is its l 19 capacity. 20 On the demand side, earthquakes are notoriously 1 21 fraught with uncertainty in our ability to predict what kind 22 of a down-motion the structure might be subjected to. Our ! 23 rules and regulations have certain requirements and those
~24 are considered in the design basis
- 25 Having said that, let me just go back and say that ANN RILEY & ASSOCIATES, LTD. 4 Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington,-D.C. 20036 (202) 842-0034
265 1 my first part would be with respect to the reasons for the (~h g
) 2 study and the second would be the perspectives for the study 3 and then what did we do and then some idea about my 4 understanding of where we could go from here.
5 There may be a slide overlap between my 6 presentation and Gareth's. Yesterday, in the discussi0: by 7 the industry, it was recognized that seismic and heavy load 8 drop events could lead to loss of inventory, but I want to 9 remind ourselves that the loss of inventory in some cases 10 could be really catastrophic, such as it could not be 11 recovered, and the consequences then could be significant. 12 So we need to pay attention to the fact that we 13 don't want to get there and the objective of all of this 14 study is how to avoid that. e~s (,) 15 On the demand side, there has been a significant 16 reduction in the seismic hazard perception. Now, this is 17 the mean seismic hazard, and I want to emphasize the word 18 mean seismic hazard, between 1989 and 1993. 19 The median estimates were pretty much about the 20 same. They didn't change very much between the 1989 study 21 by Lawrence Livermore Lab and the eventual 1993 study. And 22 when I started looking at different spent fuel pool 23 configurations, I recognized that the seismic capacity could 24 be different due to spent fuel location, spent fuel pool 25 location, super-structure, overhead crane consideration, and ANN RILEY & ASSOCIATES, LTD. (\ v- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 1 u
266 1 interaction potential with other unrelated structures, and,
) 2 therefore, they are site-dependent, plant-dependent.
3 And another important thing to recognize is that 4 the pool integrity can be determined by the failure mode 5 that has the least seismic capacity and large variations of 6 seismic capacity, as I mentioned, can exist from one plant 7 to another. 8 Next slide, please. 9 So we recognize that plant-specific hazard and 10 seismic capacity combination can raise a risk concern. And 11 this study offered us an opportunity to address this issue 12 in a risk-informed manner. We want to deal with absolute 13 numbers and things like that, but recognize where the 14 failure potentials are and what that might mean with respect () 15 to public health and safety concerns, and then hopefully 16 make a decision on the basis of those risk-informed 17 insights. 18 The first objective I already spoke about and the 19 second, let me concentrate on the second one. I believe 20 that as a result of this kind of review, as a result of even 21 future applicants coming in with a request for an exemption 22 and eventually, of course, an engineering rule, we would be 23 able to identify certain conditions that would screen out 24 plants from seismic demand and capacity related 25 vulnerabilities. O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
267 1 Next slide, please. () 2 What did we assume in the study and our approach? 3 The basic NUREG that went into the previous study, 4 NUREG-1353, there was a NUREG CR report that dealt with the 5 evaluation of seismic capacity. It came up with an estimate l l 6 of seismic fragility of at least about three times the 7 seismic design basis load. 8 And based on my experience and engineering l 9 judgment, I feel that three times the seismic design basis l 10 capacity is readily available provided that plant-specific 1 11 vulnerabilities or weaknesses can be ruled out. 12 Seismic hazard curves from the NRC and the 13 independent industry studies, this is primarily the EPRI 14 study that I'm talking about, they used very similar values, () 15 as it has been demonstrated by the Duke Engineering report. 16 This is the Duke Engineering report that was submitted at 17 the previous public meeting that we had, authored by Tom l 18 O'Hara. 19 Next slide, please. 20 In trying to use that three times the seismic 21 design basis value as a way to screen out certain plants, we l 22 'use information drawn from the NUREG-1407 and its title is 23 Procedural and Submittal Guidance for the Individual Plant 24 Examination of External Events for Severe Accident 25 Vulnerabilities. (N ANN RILEY & ASSOCIATES, LTD.
\_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l
(202) 842-0034 l
268 1 It was published in June 1991 and it was part of () 2 3 Generic Letter 88-20, Revision 4. I do have a copy of it with me, but that's not all the reporting. 4 The important point to make there is that plants
'5 are grouped into various levels of seismic hazard. We took 6 that into account.
7 I looked at the annual probability of exceedance 8 of ground acceleration at three times and then at three and i 9 a half times the design level earthquake values at the mean 10 level. This was from NUREG-1488, revised Livermore seismic 11 hazard estimates for 69 nuclear power plant sites east of 12 the Rocky Mountains. It was published in October 1993. 13- Now, this is not in your slide and perhaps I 14 'shouldn't even attempt to show this, but let me dwell on f 15 this point just a little bit. The mean hazard curve is so 16 flat that the difference in a factor of ten to 100 can cause 17 a big difference in the ground motion itself. So this huge 18 uncertainty, keeping in mind that the consequence of this 19 difference in the-ground motion could then mean a difference 20 .between the catastrophic failure of the spent fuel pool i 21 value you could not recover from the loss of water, and 22 another, you could retain the water and deal with the 23 accident or event in a rather safe manner. 24 So that uncertainty is the driving force behind 25 our attempt, at least it was my attempt to see whether or l [~')
\'
ANN RILEY & ASSOCIATES, LTD.. Court Reporters , 1025 Connecticut Avenue, NW, Suite 1014 I Washington, D.C. 20036 , (202) 842-0034 j
269 1 not some ways of screening out these weaknesses that might () 2 3 exist on a plant-specific basis. So in the end, except for a few sites, we found 4 out that three times the design level earthquake has an 5 annual probability of exceedance of the mean value of about 6 two-times-ten-to-the-minus-five per reactor year. 7 Next slide, please. 8 With a seismic capacity of three times the design 9 level earthquake, there is high confidence that the 10 probability of failure is about .05. So the hazard and 11 failure probability, this is the high confidence, there is 12 some conservatism there, is about 13 one-times-ten-to-the-minus-six per reactor year. This is 14 the probability of unrecoverable loss of spent fuel O) \' , 15 integrity and it is about half of the total probability of 16 fuel being uncovered due to a seismic event. 17 This is just putting things in perspective, so 18 that you can understand what two-times-ten-to-the-minus-six 19 came from. 20 Next one, please. 21 As I emphasized earlier, I attempted to discuss 22 various failure modes of the spent fuel pool structure and 23 it is in the report and potential changes in structure 24 response at high level earthquake ground motion. And I 25 needed to emphasize this, that the change in response due to l [~ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
i 270 1 high level ground motion and so that can understand, taking () 2 a design basis response level and scaling up to a factor of 3 three does not necessarily come up with a true picture of 4 how the structure is going to respond at a very high level 5 ground motion. That's what is of concern here. 6 And-then this is my last point, the last bullet. 7 I believe that we could come up with a very simple checklist 8 to_ screen out plants on a plant-specific basis which would 9 have no structural vulnerability or identify very simple 10 compensatory measures that will relieve the burden of 11 dealing with the catastrophic loss of water from the spent 12 fuel pool. 13 Thank you. 14 MR. CAMERON: And this is Bob Rothman? No. 15 Gareth? 16 MR. PARRY: Gareth Parry. 17 MR. CAMERON: All right. 18 MR. PERRY: I'm just going to make a few 19 additional comments to what Goutam said. I'll talk a little 20 bit about the evaluation of the frequency of spent fuel pool 21 failure due to seismic events. 22 What Goutam described was a sort of shortcut 23 method to come up with an estimate of the failure 24 probability. The most -- perhaps the most accurate way of 25 evaluating frequencies of seismic failures is to, in fact, O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
271 1 follow the methodology that was developed more than 20 years h [J 2 3 ago now by Dr. Kennedy and others and has been used in most of the IPEEEs that actually did seismic PRAs. 4 It basically consists of convolving a 5 site-specific hazard curve with a model of the plant that 6 talks about the conditional failure probabilities given a 7 seismic event of a certain magnitude. 8 Actually, I'm being rather sloppy in my use of the 9 words. Usually, the parameter used in the hazard and the 10 conditional probability formulation is some measure of the 11 ground motion acceleration. This is the approach that I 12 think was adopted in the EPRI study that we received a copy 13 of. 14 The, as Goutam mentioned, significant
) 15 uncertainties in the evaluation of both the hazard and also 16 in the evaluation of the capacity of structures and 17 components to ground motion, but the methodology is capable 18 of taking that into account. In fact, it's built into the 19 methodology so that you come up with a fairly -- with a good 20 approach to handling both the uncertainty and the initial 21 problem.
22 As I said, the methodology itself is pretty 23 mature. So in principal, the problem is relatively easy to 24 go about, but then you have to worry about sources of data. 25 We're very fortunate in that at least for all the , ANN RILEY & ASSOCIATES, LTD.
\~- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
272 1 plants in the eastern U.S., that we have hazard curves that [d'3 2 3 have been developed, both by Lawrence Livermore and by EPRI. When it comes to the fragility curves or the curves that 4 represent the conditional probability of failure of the pool 5 as a function of acceleration, what we, in fact, have is 6 representative curves for gross structural failure for a 7 particular BWR,. Vermont Yankee, and a particular PWR, which 8 is Robinson. 9 This was reported in NUREG CR-5176, and these were 10 the curves, I believe, that were used or these are the 11 fragility descriptions that were used in the EPRI report. 12 The NUREG also did address system failures, but I think that 13 it was really done for an operating plant, so that the 14 systems that are in place are perhaps a little different ) ( 15 from the ones that might be in place for a decommissioned 16 plant, and, therefore, that is an area that needs to be 17 looked at, I think, again. 18 But let's get back to the big issue, which is the 19 failure of the pump that leads to a complete draining of the 20 pool, so that it can be refilled, which has been identified 21 as one of the big issues, one of the tall trees in the i 22 forest, if you like. 23- I think our major concern is that the structural 24 fragilities are likely to be a function -- not likely -- 25 they are a function of site-specific design features and the O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
273 1 fragility curves that we used that were developed in the () 2 3 NUREG were, I think, primarily for gross structural failure. It's probably -- there are other failure modes 4 that perhaps ought to be looked at, such as penetration 5 failures, impact of adjacent structures, and dropped loads, 6 and those are going to be, again, very site-specific issues. 7 And just to illustrate the significance of making 8 assumptions -- of developing different fragility curves, we 9 included, on this last slide, a sensitivity study using the 10 simple approach that we used in the report to demonstrate 11 the impact of reducing this HCLPF value, that's the high 12 confidence low probability of a failure value. That's the 13 value at which you have at least 95 percent confidence that 14 your failure probability is greater than -- less than .05. () 15 So if we reduce that HCLPF value by 40 percent, we 16 can lead to increases in frequency of failure by a factor of 17 on the order of five. 18 Now, reduction in HCLPF of 40 percent is fairly 19 substantial. But also a reduction on the order of 20 20 percent can lead to an increase in the frequency of failure 21 by something on che order of two. So it shows you it's 22 . highly non-linear and, therefore, it is a significant area 23 of concern. 24 That, I think, completes what we have to say on 25 this issue. O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
274 1 MR. CAMERON: Can we go to Te- O'Hara now? Tom, () 2 3 if you could just introduce yourself to us. didn't catch you before. I'm sorry I 4 MR. O'HARA: Tom O'Hara. I work for Duke 5 Engineering. Services. The reason you didn't catch me was I 6 wasn't sitting here. 7 MR. CAMERON: All right. Do you want to use this 8 microphone while you do your presentation? 9 MR. O'HARA: I can sit down here and do it. 10 MR. CAMERON: All right. And when Tom is done, 11 we'll have a discussion on the seismic issue. 12 MR. O'HARA: As I say, my name is Tom O'Hara. I'm 13 with Duke Engineering Services. I've got a long line 14 history of experience in areas of seismic hazard and PRAs. () 15 The focus of this presentation, of my 16 presentation, is specifically on spent fuel pool failure 17 frequencies. I want to give some background information 18 just to sort of frame what I think the issues are and put 19' things, I think, in better perspective. Some will be a 20 rehash of what Goutam said, et cetera. 21 NUREG-1353, this document here, which was the 22 regulatory analysis for the resolution of Generic Issue 82, 23 beyond design basis accidents in spent fuel pools, was the 24 template for that EPRI study. As stated in NUREG-1353, 25 seismic was the dominant contributor to the overall risk. ANN RILEY & ASSOCIATES, LTD.
- O. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
275 1 It represented over 90 percent of the risk. 2 One thing that they -- they also evaluated the 3 systems failure due to seismic, loss of makeup ccoling 4 ability, and that was estimated as 5 two-times-ten-to-the-minus-eight in the 1353 study, and it 6 is one-times-ten-to-the-minus-six currently in the draft NRC 7 document. 8 I'm not addressing that, but there is a huge 9 change between what was published and what is cu; rently 10 being used. The NUREG-1353 results are based primarily -- 11 they are based upon the Livermore 1989 seismic hazard 12 results. In April, I gave a presentation over at Rockville 13 and in that presentation, I said that the spent fuel pool 14 failure frequencies due to seismic would be reduced by a ( 15 factor of five to ten if you kept the methodology the same 16 as was used in NUREG/CR-5176, which was a Livermore study, 17 which was input, which provided the inputs to NUREG-1353, 18 So, again, my position at that time was you would 19 reduce the spent fuel pool failure frequencies, holding 20 everything the same, but just changing the inputs and using 21 updated seismic hazard results. 22 I also said that if you use the EPRI results,
. 23 there would be a factor of ten reduction. Using traditional 24 PRA approaches -- that is, as was talked about, where you 25 convolve a family of seismic hazard curves with a family of ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washir.gton, D. C. 20036 (202) 842-0034
276 1 fragility curves. () 2 I used the -- I performed the analysis, but what I 3 did was I used the seismic hazard results for the 1989 4 study, the Livermore 1993 study, and the EPRI study, and the 5 purpose of doing the '89 study is to show the -- you want to 6 do the calculation for '89 so you can show its reduction by 7 using the new results. 8 Based upon this analysis, the mean for the ; l 9 population of plants, we're talking eastern U.S. studies, is 10 less than ten-to-the-minus-six. That's the mean for the 11 population of plants. 12 You will find that some sites do pop above 13 ten-to-the-minus-six. So with that as a background, go to 14 the next slide. You've got it already, stay right there. () 15 In a PRA for spent fuel pools, there's two pieces 16 to the puzzle, basically, the seismic hazard and the 17 fragility. So I want to quickly cover the seismic hazard, l 18 the history of the seismic hazard, Livermore and EPRI. I'm 1 19 going to just briefly gloss over NUREG-4982. That was a 20 ENL, Brookhaven National Lab study, I guess it was the first 21 one.done on the spent fuel pool. Then I'm going to cover 22 5176, which, again, provides the seismic failure 23 probabilities for NUREG-1353, 24 I will briefly summarize the NRC approach, as I 25 see it or saw it, and then I want -- then what I'm going to ANN RILEY & ASSOCIATES, LTD. Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
277 1 do is I will show you results of running all of the sites 2 using these different seismic hazard curves and different 3 methodologies. 4 Next slide. 5 There is the -- the Livermore methodology goes way 6 back, actually to 1978-79. That's when the systematic 7 evaluation program, SEP, was started. One of the things 8 they were looking for was a quick way to provide a 9 consistent seismic design value in terms of probability at 10 these ten older plants. These were ten older plants in 11 terms of they were pre Appendix A and pre SRP. 12 In 1982, I believe it was, the Charleston issue 13 came about and that's where you had -- basically, it is that 14 the USGS said that the Charleston earthquake could not be 15 tied to a structure.and.the way that people decided to 16 handle this was probabilistically. The Charleston i 4 17 earthquake was a magnitude seven earthquake. ) 18 And what the NRC did was they used the NUREG-1582,
)
19 the original Livermore methodology, as the framework for a 20 methodology to handle the Charleston issue and EPRI 21 generated their large study, and these are large studies. 22 They're -- I don't want to say they're continental, but i 23 they're damn close to continental. They go from the Rockies 24 to the east coast. J 25 In 1985, 2421, the results of that study were used
)
i O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
)
278 1 in the BNL study and a point to -- I'll pass on that for a () 2 second. Then Livermore published their final results in 3 1989 in 5250 and those results were used in the Livermore 4 study, NUREG-5176, where they calculated the spent fue? pool 5 failure frequencies at Vermont Yankee and Robinson, which 6 were subsequently used in 1353. 7 EPRI also published their results and in 1993, 8 Livermore revised their seismic hazard results. And I can 9 go into why, what the differences are between Livermore and 10 EPRI, Livermore '89, Livermore '93, that's another time. 11 But one point I will raise is that prior to 1993, 12 the Livermore mean hazard curve was a point of contention. 13 .It was a contentious statistic, shall we say. It exceeded 14 the 90th percentile at times.
) 15 Consequently, when people did PRAs, they did 16 everything they could to avoid using the mean, and I'll just 17 drop it at that.
18 Next slide, please. 19 I had one more point here which I considered to be 20 important. On page 6 of SECY-91-102, this is the IPEEE for 21 severe accident vulnerabilities, there was a response by the 22 NRC to a question by industry and it had to do with why are 23 you making us use both Live?more and EPRI hazard curves in 24 the evaluations. 25 And they said based on the available information D V ANN RILFY & nSSOCIATES, LTD. Court Reporters 2025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
)
279 1 to date, they consider both of them to be valid. Therefore, 2 you use both of them, and that's why I used both of them. ( 3 Next slide, please. 4 The purpose of this slide, this is the Brookhaven 5 National Lab study. Again, what they used was hazard 6 results. These were preliminary Livermore hazard results, 7 published in 1985, and they used the 15th, the 50th, the 8 80th percentile and a best estimate. The best estimate is 9 not the mean, and that's another story. 10 They also used surrogates for the fragility curves 11 and they were based upon PRA studies at Oyster Creek and 12 Zion. > 13 And remember that when you're performing these 14 PRAs, you have the hazards and the fragilities. You need () 15 those two pieces, and that's why I'm pointing them out. 16 Next slide, please. 17 In the Livermore 5176 study, this used the l 18 Livermore 1989 results. They did not use the mean hazard 19 curve. They came up with a method to generate a family of 20 hazard curves, eleven, weighting and et cetera, and this 21 became the hazard input. 22 Next slide, please. 23 The difference between 5176 and the previous BNL 24 study is they didn't use surrogates. They went out and 25 looked at the plants and they came up with these -- they ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
280 1 came up with fragility estimates of the median capacity and () 2 3 the beta R's and the beta U's, and location estimates. the random and uncertainty 4 And they used these along with the Livermore 1989 5 seismic hazard results and they came up with estimates of 6 spent fuel pool failure and that was used in the 1353 study. 7 And as you recall, in that study, they said 8 seismic represents over 90 percent of the risk, it's a big 9 concern. 10 So what you have here, the 5176 methodology is 11 basically you generate a family of hazard curves, you have a 12 family of fragility curves, you convolve the two, you have a 13 family, if you want to generate uncertainty estimates, but 14 bottom line, we're looking at a mean spent fuel failure /~N () 15 frequency. So that's one methodology. 16 The next one -- and there is a variation of this 17 where you can just simply replace the eleven hazard curves 18 with a mean, and I've done that, too. 19 Next slide, please. l 20 This is what Goutam was just talking about, and j i 21 this is my summary of it. It may not be correct totally, J 22 Goutam, but it's probably close. 1 23 MR. BAGCHI: The next one, the next slide, that's i i 24 what you're talking about. 25 MR. O'HARA: Okay. But what I wanted to -- I l l l ANN RILEY & ASSOCIATES, LTD. s- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 j
281 1 thought it was a great write-up, to be honest with you, and () 2 the bottom line is what you're saying is these spent fuel 3 pools are inherently rugged to withstand a lot of seismic 4 shaking. And what you're saying, what you said in the l 5 report was it was good, .4 to .5. Well, I say that's above 6 .45. And so I used that number, 7 Next slide, please. 8 And then, of course, the NRC approach is to -- and 9 it is a pretty clever little approach, very simple, very 10 easy to check. So what one does is you can take your 11 seismic hazard curve, whether it's a Livermore or EPRI, use 12 a mean seismic hazard curve, go in at .45g, come up with a l 13 probability of exceeding .45g, and now you multiply that by 14 a factor of .05, which is their shortcut, and you come up l () 15 with one-times-ten-to-the-minus-six. l 16 So I have also taken this approach and run the 17 numbers using Livermore and EPRI. And, again, I say one of 18 the beauties of this approach is anybody can check any 19 number that I esent here using the Livermore approach. l 20 It's a lot more difficult with the classic PRA approach, as 21 we are convolving mathematically the hazards and the 22 fragilities. 23 Next slide, please. 24 This shows you, on the X axis is site, and I have 25 a site code and I know the name of each site. All the ANN RILEY & ASSOCIATES, LTD. Os Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
282 1 utility people want to know what their site number is, and I 2 . won't tell them that. On the Y axis is annual probability 3 of a spent fuel pool failure. The solid line is the EPRI 4 results and the dashed line are the Livermore results. 5 What I did here was using the 5176 methodology, 6 you have eleven hazard curves, a family of five fragility 7 curves, and you convolve the two together, do the 8 gymnastics, and you come up with a number. The mean for the I 9 Livermore results is nine-times-ten-to-the-minus-seven, for 10 EPRI it's 1.4-times-ten-to-the-minus-seven, and that's 11 across the population of sites. 12 There are a few EPRI sites that weren't included 13 that Livermore did, because you may see a break in the EPRI 14 line at times. 15 Next slide, please. 16 As I said to you before, both approaches are 17 valid, Livermore and EPRI, and there are reasons for the 18 differences between Livermore and EPRI. I mathematically 19 took a pure average of the Livermore and EPRI, that's all it 20 is, and whenever you average four-times-ten-to-the-minus-six 21 with ten-to-the-minus-eight, the answer is l 22 two-times-ten-to-the-minus-six. So the lower probability 23 numbers don't get anywhere -- unless you do it 24 logarithmically. l 25 Anyway, just a pure average, and what you can see O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, hM , Suite 1014 Washington, D.C. 20036 (202) 842-0034
283 1 here is that the average of the two is about () 2 3 five-times-ten-to-the-minus-seven and a few pop above the line. 4 And I want to raise one point here. In the 5 NUREG-1353 analysis, they went all the way to release. In 6 other words, they multiplied the spent fuel pool failure 7 frequency times the conditional probability of release given 8 failure. For a BWR, that was .25. That's not down here. 9 So some of these are BWRs and -- why are you doing that? 10 MR. BARRETT: I wanted to ask a question. 11 MR. O'HARA: Go ahead. 12 MR. BARRETT: I wonder if you could just quickly 13 tell me what did the .25 represent? 14 MR. O'HARA: The conditional probability -- .25 () 15 represented the conditional probability of a zircaloy fire 16 given that you lost your inventory and that was based upon i 17 the spacing of the fuel rods, et cetera. 1 18 You'll have to go back and ask the authors when 19 they came up with that number, but that's what they used. 20 It's a conditional probability given failure, you've lost 21 inventory, of the fire. Okay? 22 MR. BARRETT: I'm sorry to do this. Yesterday, 23 this was the procedure that we agreed upon for asking 24 questions, for trying to get attention. 25 MR. CAMERON: As the facilitator, I think I spoke ANN RILEY & ASSOCIATES, LTD.
,_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
284 1 to you about doing that. () 2-3 to me. MR. O'HARA: I was thinking why is he doing this 4 MR. CAMERON: Are you okay, Tom? 5 MR. O'HARA: Yes, I think I can handle it. 6 MR. CAMERON: All right. 7 MR. O'HARA: So the -- go ahead, Goutam. l MR. BAGCHI: 8 Just one clarifying point. I think 9 the way it has been done, that factor of .25 is not there 10 anymore. 11 MR. ROTHMAN: It's different. 12 MR. BAGCHI: It's different. 13 MR. O'HARA: Okay. I can't debate that one with 14 you, Goutam. We're not using it. It's not in this () 15 analysis. I'm just clarifying it because that was what was 16 done in NUREG-1353. 17 MR. CAMERON: And, Tom, I guess it would be 18 worthwhile saying that what we agreed to yesterday is that 19 we didn't want to get off on discussion during the 20 presentations, but that there might be a need to clarify , 21 certain information just to understand a presentation. 22 So I'll keep an eye out for anybody who might want j 1 23 to ask you a clarifying question. 24 MR. O'HARA: No problem. 25 MR. CAMERON: All right. ; ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
285 1 MR. O'HARA: This is the NRC HCLPF approach and () 2 3 there is one mistake on this slide. On the Y axis, that should not be probability of exceedance. That's just annual 4 probability. I've done too many seismic hazard 5 calculations. 6 Other than that, what you find, again, it's the 7 same process. The solid line is EPRI, the dashed line is 8 Livermore, and the mean for Livermore is 9 nine-times-ten-to-the-minus-seven, the mean for EPRI is 10 four-times-ten-to-the-minus-seven. What you find is that 11 these are closer together-than the previous analysis and the 12 answer -- and the question is, why is that. 13 The answer is, you're not -- there is a difference 14 between Livermore and EPRI in terms of the slopes at the i 15 lower accelerations and we're not going on out -- excuse me 16 -- the higher accelerations, and we're not going -- we're 17 not convolving hazard curves with fragilities. t.'e ' re l 18 picking a value at .459 To it's pretty close there. 19 Next slide, please. 20 Again, the same error on the Y axis, change it to l 21 annual probability. This is the average of Livermore and l 22 EPRI using the -- I'll call it NRC approach. The mean is l 23 about six-times-ten-to-the-minus-seven and, again, you see l l 24 some that pop above the line. 25 Next slide, please. O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1
286 1 This is exactly the 5176 methodology, just using () 2 3 the mean hazard curve. The mean for Livermore is 9.6-times-ten-to-the-minuc-seven, EPRI 1.3. 4 Next slide. 5 The average of the two is about , 6 5.5-times-ten-to-the-minus-seven. And what's of interest is 7 the same plants, of course, pop above the line and you would 8 expect it, so they kind of cross-check with each other. 9 Next slide, please. 10 This is a summary of our conclusions here. Based 11 upon the 5176 methodology, in using -- in other words, using 12 your eleven seismic hazard curves and your five fragility 13 curves, I ran the Livermore 1989 results. I also ran the 14 Livermore '93 and the EPRI. () 15 The purpose of the Livermore '89 was just to check 16 my original statement. Is it about a factor of five to ten? 17 It is. It's seven, seven or eight. l 18 Livermore is about ! 1 19 nine-times-ten-to-the-minus-seven, EPRI is l 20 1.4-times-ten-to-the-minus-seven, the average is about i 21 five-times-ten-to-the-minus-seven. 22 Next slide, please. 23 Then I took the 5176 and just used the means and 24 what you see is the difference between '89 and '93 is over a 25 factor of ten, and I would expect that. fi V ANN RILEY'& ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
287 1 So Livermore '93 is
, 2 9.6-times-ten-to-the-minus-seven, EPRI is 3 1.3-times-ten-to-the-minus-seven, the mean is about 5.5.
4 Next slide, please. 5 If we use the NRC HCLPF approach, you see over a 6 factor of ten reduction going from Livermore '89 to '93 7 results. Using the revised Livermore '93 and the EPRI, 8 they're pretty close, nine and 9 3.5-times-ten-to-the-minus-seven, and the mean is about 10 six-times-ten-to-the-minus-seven. 11 My last slide is the conclusion. I've looked at 12 this a couple of different ways and the fundamental 1 13 conclusion is that the mean probability of a spent fuel pool 14 failure is less than ten-to-the-minus-seven, using the 15 generic assumptions that were published in NUREG-5176 or 16 using the HCLPF approach that the NRC is proposing. 17 If one used EPRI, you would say the numbers are 18 < about ten-to-the-minus-seven. If you use Livermore, it's 19 less than ten-to-the-minus-six; not much, but less than 20 ten-to-the-minus-six. If you use the average, which is 21 probably the right number, it's around 22 five-times-ten-to-the-minus-seven. 23 And that' concludes the presentation. I 24 MR. CAMERON: Thank you, Tom. We're going to open j 25 it for discussion now, and I guess I would like you to think l
; ANN RILEY & ASSOCIATES, LTD. j s Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 {
l Washington, D.C. 20036 l (202) 842-0034 i
288 1 about what you want to get out of this discussion at the () 2 3 end. j The NRC slides identified some areas of concern, ! I 4 do we want to reach some sort of resolution on these. Tom 5 gave us a sort of bottom line here. Just think about that ) i 6 so that we might get a sense of where we're going. 7 But let's go to Glen and open it up with him. 8 MR. KELLY: I would be interested, Tom, in hearing 9 an explanation of why you feel that the mean or the average 10 is a good number to use for discussing or considering the 11 population, because it reminds me of the story about the 12 statistician who was six feet tall, who drowned in a river 13 crossing that only had a mean of three feet deep. 14 So what does the mean value or the average of () 15 population tell us about the population in this particular 16 case? : 17 MR. O'HARA: The mean value -- historically, I l 18 know, from my experience, the mean has been contentious and j 19 oftentimes the seismic hazard analyses, we will try to use 20 the median. 21 But you use the mean because it provides 22 information about the population. Every point is included i 23 in the statistic, the mean statistic, and, therefore, you 24 use means to make a statement of inference about a 25 population. l [d\ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
289 1 MR. CAMERON: Does that answer > .: estion? (O) 2 MR. KELLY: My question really goes to, if you're 3 taking the mean of the means here and then you've got the 4 means of -- you've used the mean value for each one of these 5 sites to tell you something about the potential risk at the 6 site, and then you've taken the mean of all of these 7 together or the average of all of these together and you 8 said, okay, now this tells me something about population. l 1 9 My question is, what does the mean of this mean 10 mean? i 11 MR. CAMERON: Do you want to answer it? 12 MR. O'HARA: I'll give it a shot. Again, the mean 13 is infcrmative about the distribution. It includes all the 14 values. And I think what you're driving at is no, no, no, () 15 you don't want to use the mean, you want to use the highest 1 16 value. 1 17 And I say I don't think so. To describe your 18 population, you take the mean value. 19 MR. CAMERON: Let's have some further discussion 20 on this. Goutam, did you want to address this, also? 21 MR. BAGCHI: I just wanted to point out that I was 22 pleasantly surprised, perhaps not surprised, but I'm pleased 23 that the results came out where they did. 24 But lets not just take a statistician's point of 25 view of the world., Does it mean that having found this, you ( ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l 290 1 don't take into account the potential vulnerability that ' () 2 3 might exist in plants that are coming in for requests to decommission and do certain things. 4 In my mind, I don't think so. I think there can 5 be some vulnerabilities that one could develop some very 6 simple methods of checking to get past that, and let us not 7 worry about the statistics now, or maybe I just want to ask 8 the industry, does it want to pursue that by neglecting the 9 potential for site-specific vulnerabilities. ! 10 MR. CAMERON: I think we're going to get some 11 comment on that question. Dr. Kennedy? 12 MR. KENNEDY: Yes. I've carefully reviewed both 13 the industry, your report, and the NCR's report. I'm 14 extremely comfortable with your fragility levels that you () 15 used for the two plants, the two plants. But the fact is 16 other plants will have different fragility levels. 17 Some will be higher, some will be lower. If you 18 had to take a gross number.or a generic number, the numbers 19 you used are fine. But I think that you will find, if you 20 looked at other plants, that they could easily vary by a 21 factor of one and a half and that factor of one and a half 22 transfers into risk space, at least a factor of five. So at j l 23 least a factor of five higher or at least a factor of five 24 lower than sort of the generic average fragility carve would l i 25 produce. O \_/ ANN RILEY & ASSOCIATES, LTD. Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
291 1 In addition, there is the tremendous variabilities () 2 3 from the hazard curve. Now, from the Livermore hazard curve, the Livermore '93, because that's -- I looked through 4 your results real carefully for that hazard curve, the same 5 kind of variation would exist for EPRI. 6 Twenty-six of your sites were BWRs, for which you 7 used the BWR fragility, which had a HCLPF value, you didn't 8 report it, but there was enough information in your report 9 to back-calculate it, of about .48g, very close to the 10 number that the NRC used. 11 Those 26 BWRs, the seismic risk ranged from 12 12-times-ten-to-the-minus-six to 13 .1-times-ten-to-the-minus-six. Just from the hazard curve 14 variability. () 15 So there is a factor of 100 difference in 16 individual plants just due to the hazard curve variability. 4 17 You put that with an additional factor of 18 potentially a five due to fragility variability, now you 19 can't multiply these together, because it would be very 20 unlikely you'd bo high hazard and low fragility at the same 21 time. But there is a tremendous location to location 22 scatter in these risk results. 23 I think as a generic average, the
]
24 one-times-ten-to-the-minus-six number is a very reasonable 25 number using Livermore '93 hazard curves and if you use EPRI ANN RILEY & ASSOCIATES, LTD. ,. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
292 1 hazard curves, it will be a factor of three to five less () 2 3 than with Livermore hazard curves. But there is this big scatter from site to site 4 that I think you need to somehow consider. 5 MR. CAMERON: Okay. Before we go to Mike, I'd 6 just like to give Tom an opportunity, if he wants to, to 7 respond to this location to location scatter issue that Dr. 8 Kennedy raised. 9 MR. O'HARA: What Bob says is correct and you 10 don't argue with Bob Kennedy, because that's the first -- 11 one of the things you don't do. 12 But to make the -- the bottom line is you try to 13- get a job done. So you use what you've got. And what we 14 had were estimates of seismic fragility and estimates of () 15 seismic hazard and you perform an analysis and you factor in 16 comments like what Bob just raised. 17 But I think overall, if you're looking for a 18 summary kind of statistic or statement about the seismic 19 contribution to spent fuel pool failure, I contend it's got 20 to be something like less than ten-to-the-minus-six or 21 ten-to-the-minus-six, or with EPRI it's going to be on the 22 order of ten-to-the-minus-seven, or it's in that ballpark 23 range between ten-to-the-minus-seven and 24 ten-to-the-minus-six. 25 And maybe you factor in things like what Goutam , ANN RILEY & ASSOCIATES, LTD. \ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
293 1 was talking abouc, that can just get you totally past the () _2 seismic hazard issue. That's what I -- that's all. 3 MR. CAMERON: Okay. Mike? 4 MR. MEISNER: I had a couple questions, first for 5 Goutam and maybe getting at what we're noodling around here. 6 If I look at page nine of your overheads, you talked about a 7 simple checklist can be developed to screen out plants. 8 Could you talk about that a little more, what you 9 had in mind? 10 MR. BAGCHI: In very simple terms, just a 11 walk-down of the site and making sure that the obvious 12 vulnerabilities are not there. I think beyond that, we need 13 to talk to engineers and come up with the list. 14 MR. MEISNER: Okay. Thanks. The other question I () 15 had was maybe a little more general. We've been talking 16 about the catastrophic seismic event, the other portion of 17 the staff's analysis has somewhat equal contribution for a 18 non-catastrophic event that would, I assume, affect 19 ctructures, systems and components and affect cooling. 20 I would assume that we'd probably agree that for 21 that kind of event, that really reduces to the things we 22 were talking about yesterday, where we get into essentially 23 a loss of cooling event and that any kind of adjustments 24 we'd make, say, in the HRA space or otherwise would affect i 25 the probabilities of that kind of event. j j ANN RILEY & ASSOCIATES, LTD. O* Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l 294 1 So I just want to kind of dispose of that, so we () 2 3 don't have to deal with that anymore in seismic terms. MR. HOLAHAN: I think that's a reasonable 4 observation. I think the issues may be a little bit 5 different, finding available equipment following an 6 earthquake, but it's basically a human reliability question. 7 MR. MEISNER: The time periods and all are about 8 the same. 9 MR. HOLAHAN: Yes. 10 MR. CAMERON: Is that significant enough to put up 11 here? 12 MR. BAGCHI: One thing that's really significant 13 is hat there seems to be an understanding that the 14 ca strophic failure portion can be taken off the table if a (h 15 rezJonable checklist can be developed. That's a very 16 significant point and I'd like to have that on the table. 17 MR. MEISNER: Yes. I was just talking about the 18 -- that's a different question. ; 1 19 MR. CAMERON: Okay. 20 MR. HOLAHAN: What I would suggest is, however you 21' capture this morning's discussion, just say including the 22 seismically-induced loss of cooling. 23 MR. CAMERON: Rich, you had a comment on that? l 24 MR. BARRETT: I really have a question. 25 MR. CAMERON: A question. l ANN RILEY & ASSOCIATES, LTD. O- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l 295 1 MR. BARRETT: My question for the people who would 2 have done the analysis here is, I would think that it's 1 f]\ 3 certainly reasonable to make the statement that Mike made. l 4 I would just ask, is there anything about the seismic 5 analysis or the -- let me put it this way -- the level of 6 the seismic event that we're talking about that would change 7 our assessment of the human response, substantively, as 8 opposed to the sequences we were talking about yesterday, in i 9 which the human response is taking place in the absence of l 10 an earthquake event. 11 In other words, are these -- are we talking about l 12 seismic events here that are relatively low accelerations 13 and would not have a significant effect on operator response 14 or are they -- and I -- just for clarification. b) g 15 MR. CAMERON: Anybody who wants to respond. 16 MR. ROTHMAN: When Goutam was talking about two to ; 17 \ three times the SSC potential for failure of the spent fuel i 18 pool, he's talking about fairly significant ground motion. 19 I mean, operators are going to get knocked off their seats, 20 those kinds of things. 21 Following the earthquake, then you have to talk to 22 somebody in psychology that knows how people react after 23 those things. l l 24 MR. BARRETT: I think Mike's question actually l 25 deals with a different scenario, and that's a scenario in l [ ANN RILEY & ASSOCIATES, LTD. l \
' i, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 i (202) 842-0034
296 1 which presumably an earthquake of lower magnitude, I () 2 believe, affects risk by failing the equipment as opposed to 3 failing the pool, and that kind of a sequence then basically 4 becomes just like the kind of sequences we were talking 5 yesterday, in that it requires operator action primarily to 6 recover. 7 My question simply is, are we talking about 8 earthquakes of a low enough severity such that there will be 9 no concern about impacting those operator actions which we 10 would be able to accept under the circumstances for 11 yesterday's sequences? i 12 MR. CAMERON: Mike, does that capture your 13 clarification? 14 MR. MEISNER: Yes. I () 15 MR. CAMERON: Okay. We're going to go to Goutam 16 to address that. 17 MR. BAGCHI: What I had in mind in terms of 18 checklist is pretty much what would encompass Dr. Barrett's
~
19 concern. Would we be able to bring in portable equipment or 20 equipment that's lying around for pumping water into the 21 pool back? Those kinds of things can be developed by some 22 kind of a checklist and let's remember, no matter how hard l 23 the earthquake was, we have a significant amount o.f time l 24 available to respond and make water or water inventory build ! 25 up. O s/ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ' Washington, D.C. 20036 (202) 842-0034
297 1 And people who would react after the earthquake to 2 get involved in those kinds of decisions, it would be ( 3 significantly after the earthquake has passed away and they 4 have come to their senses and they are ready to make common 5 sense decisions. 6 MR. CAMERON: Okay. Now, I know there's a lot of 7 cards up. Ray, I see yours, and Stu had his up. But I'd 8 like to follow this through and I think Dr. Kennedy had 9 something on this. And it's always useful to have someone 10 at the table that no one wants to argue with. 11 MR. KENNEDY: I'll argue with myself. On the 12 issue of these other earthquake events, any earthquake level 13 that_is going to damage equipment in the plant, at least 14 anchored equipment, or knock off off-site power from the ( 15 local switchyard at the plant, is certainly going to be an 16 earthquake that the operators are going to have felt and 17 it's going to be disturbing to them. 18 I think if you had to have rational decisions made 19 reliably within, let's say, 15 minutes after the earthquake, 20 there would be issues associated with that. 21 But I think experience shows that for emergency 22 crews in real earthquakes, if you talk a half-hour after the 23 earthquake, good decisions are made. And I think, from what 24 I understand you're talking about here, you have time. 25 MR. MEISNER: Five days. O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l 298 1 MR. KENNEDY: So that I don't think that there is () 2 going to be any real earthquake influence on whether 3 rational decisions are made or not. 4 MR. CAMERON: Okay. Thank you, Dr. Kennedy. 5 Let's go to Glen, and then we'll go over to Ray Shadis. 6 MR. KELLY: When we calculated the 1 7 ten-to-the-minus-six number for loss of cooling due to a j 8 failure of other equipment on site due to a seismic event -- 9 that is, failure of things other than.the spent fuel pool -- 1 1 10 we assumed that there was no recovery on-site of that 11 equipment; that that equipment was effectively destroyed. 12 And what we looked at was then recovery using 13 off-site equipment, but with a slightly lower probability 14 than we would give it ordinarily for off-site recovery, g_j 15 based on the fact that we had had an earthquake and there 16 might be demands for off-site equipment in other places. 17 But we did not -- we're not worrying about the 18 operators on-site trying to repair the equipment. This is a 19 matter of just being able to get in off-site equipment to 20 the site. So that's where our numbers came from. 21 MR. CAMERON: Okay. Mike, did you have a point on 22 that? 23 MR. MEISNER: That's all I wanted. 24 MR. CAMERON: Okay. We're going to revisit that 25 in the next session on mitigation of events. ( ANN RILEY & ASSOCIATES, LTD. \ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
299 1 MR. MEISNER: No, when the staff does their -- () 2 3 MR. CAMERON: Right. Let me -- on that, this is at least my take, and we can revise it as you guys think 4 necessary, of what we have simmering now for later 5 discussion, and I did put up that this point number three, 6 what happens here would include seismic lost of cooling. 7 Ray? 8 MR. SHADIS: Thank you. Two things. Oct ~.E t a 9 layman's question here. I want to inform my ignorance. On 10 these various studies that were cited, these analyses, do 11 any or all of them include the fragility of the transfer 12 tube or transfer tunnel on PWRs in the fragility numbers? 13 I ask that because Venus D'Milo is pretty rugged 14 precisely because their arms are snapped off. We considered j 15 that transfer tube an integral part of the spent fuel pool, 16 that big penetration. 17 MR. CAMERON: Is there an answer from NRC staff or 18 opinions from Tom? Tom, do you want to? Go ahead. 19 MR. O'HARA: I can't answer that question 20 explicitly, but the documentation for those fragilities is 21 found in 5176, and I guess we can review it and determine 22 if, in fact, your concern is addressed. 23 MR. SHADIS: Yes, I would think that if the 24 concern with an earthquake is that somehow there is a shear 25 of the wall of the spent fuel pool and you have an opening O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
300 1 of.a given size at which it can drain rapidly, that that ( 2 might be in the same ballpark as having the transfer tunnel 3 blown out by stresses put on it. 4 Again, a layman's view, but it seems to me that's 5 a great deal more fragile than the six to ten foot thick 6 walls in'the spent fuel pool. 7 MR. CAMERON: And we need to get an answer for 8 Goutam. State your name for the record. 9 MR. FORD: Brian Ford, Millstone Unit 1. On the 10 BWRs, until you get to the BWR-6's, there is not a transfer 11 tube, as such. There is a transfer canal that's up at a 12 higher level, but there is not really the same thing as a j i 13 tube that you have on the PWRs. l 14 MR. SHADIS: I understand that. I was speaking () 15 specifically to the PWRs. But maybe we can -- and if we can 16 get that worked in some way. 17 MR. CAMERON: Okay. Let's hear Goutam's answer. 18 I put that up there as an item. Goutam, what's your take on 19 this? 20 MR. BAGCHI: My view of that NUREG document 21 indicated that it was based on the capacity of the shear 22 wall itself and it did not take into account specifically 23 those kinds of things. 24 But please remember, throughout '94, the thing 25 that I pointed out are the reasons for various failure modes O ANN RILEY & ASSOCIATES, LTD. Court Reporters
-1025 Connecticut Avenue, NW, Suite 1014 '
Washington, D.C. 20036 (202) 842-0034
301 1 that need to be considered, and this is the reason why I () 2 3 talked about developing a checklist of things that need to be looked at on a site-specific basis and then once you meet 7 4 that, take it off the table. That was my approach. I 5- Thank you. 6 MR. CAMERON: Mike, do you have a comment on the 7 transfer? 8 MR. MEISNER: I wanted to suggest a wrap-up on l 9 this. 10 MR. CAMERON: Okay. Let's hear Ray's second point 11 and then we'll come back to you. 12 MR. SHADIS: I agree with Mike, a wrap-up might ! 13 just be in order. But the conversation went from 14 probabilities of certain kinds of seismic events to the () 15 fragility and the -- but then it kind of jumped to 16 mitigating actions and whether or not it would be reasonable' 17 if a staff that was knocked off its feet would then get up, 18 dust themselves off, and proceed to take care of things. 19 Let me just say that if there is an earthquake 20 sufficient to damage, in any significant way, the equipment 21 at a building, never mind one large enough to crack the 22 spent fuel pool, that the spent fuel pool building, not 23 being seismically qualified, has to be draped over this in a 24 large covering, which would prevent an awful lot in the way 25 of mitigating action. (')
\/
ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
302 1 And you might also have your crew either injured () 2 3 or dead in auxiliary buildings that are not seismically qualified, or even in ones that are, depends on whether or 4 not filing cabinets have fallen around people, who knows. 5 So I think it gets to be a little bit silly to 6 start speculating about how the crew on-site is going to 7 behave. I think you need to speculate on how the crews 8 coming from off-site are going to behave and respond in 9 whatever timeframe is available. 10 MR. CAMERON: Okay. Thank you. Let's go to 11 Peter, and then we're going to go to Mike for a wrap-up, and 12 we're getting towards break time here. 13 MR. ATHERTON: I'm sorry I wasn't here yesterday, 14 but I wasn't able to make it. So I'm trying to catch up as () 15 rapidly as possible. 16 This session went directly to the specifics of the 17 seismic evaluation. The question I have would be more 18 general than that, and I would ask, one, did the seismic 19 evaluation take into account the seismic qualification of 20 the structure at the time the plant was constructed and if 21 so, what would be the effects of the qualification of the 22 equipment and the aging of the facility after a certain 23 number of years that the plant has been or the spent fuel 24 pool has been used and irradiated? : 25 There does not appear to be -- that's question l l
, ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
303 1 , nu daer one . Thers does not appear to be any concern about / 2 the failure of electrical equipment in systems and the
}
3 qualification of those systems for any specific period of 4 time that the spent fuel pool is going to be in use. 5 MR. CAMERON: Is that second one a 6 seismic-related? 7 MR. ATHERTON: It is, and that is I have not heard 8 any talk to the desire or the effort to minimize the 9 probability of the failure of systems that would be required 10 to mitigate the consequences of any design basis or beyond 11 design basis spent fuel pool accident, whatever it is we're 12 considering. 13 MR. CAMERON: Can we get some commentary on 14 Peter's questions? Glen? () 15 MR. KELLY: When we did our analysis, we assumed, 16 first of all, that there -- all off-site power was lost and j I 17 because the -- if there is a seismic event, it's going to -- l l 18 we're going to. lose off-site power, because we also assume i 19 that there aren't any emergency diesel generators that are 20 operating, again, in our simplified design that we analyzed, 21 Therefore, there was no AC power available on-site, nor were 22 there any uninterruptible power sources or batteries to 23 power things. 24 So basically the plant is dead as far as 25 electricity goes. O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
304 1 So really what we're depending on then is being 2 able to bring in things from off-site to recover, and here {) 3 recovery really means either bringing in something to -- l l 4 bringing in equipment, to now bring in a diesel generator 5 and cooling equipment to cool the pool, or, if you're not l 6 successful in that, bringing in fire equiptaent or something 7 to provide water to cool. 1. l 8 It's not a matter of -- we analyzed it as if 9 everything on-site was unable to respond. This is 10 potentially conservative, but again everybody -- when I went 11 around to the sites, some of the sites had their equipment I 12 anchored; other sites, it was sitting on skids. There was 13 spent fuel pool cooling equipment. Some sites had -- nobody 14 had any diesel generators, but some did have -- that's not () 15 t rue . One plant did have a diesel generator that was 16- specifically going to their spent fuel pool cooling system, 17 but they didn't have it for other equipment, and I believe 18 that one was supposed to be seismically qualified. But that 19 was -- nobody else had anything like that. 20 And that's not how -- we analyzed it more 21 generically, as if there were no spent fuel pool cooling 22 systems. 23 About aging and equipment qualification, again, 24 since we assumed that all the equipment that was cn-site 25 failed, it wasn't a matter of whether or not it was aging or I ANN RILEY & ASSOCIATES, LTD.
) Court Repurters 1025 Connecticut Avenue, NW,' Suite 1014 l Washington, D.C. 20036 l (202) 842-0034 l
e
305 1 not. We assumed it wasn't there.
\
2 And lastly, about the impact of aging on the (A 3 concrete and the radiation, I don't -- we did not look at c radiation, but generally it's our belief that the concrete 5 generally is considered to get stronger over time and that 6 over the periods that we're talking about, at least I think 7 for the first 100 years, if I've got that right, Goutam, and 8 -- so in this case, from the concrete standpoint, since 9 we're really talking about this giant concrete bathtub, that 10 that actually becomes harder over time or stronger over time 11 than it was initially when it was first constructed. 12 Hopefully, that will answer your comments. 13 MR. CAMERON: Peter, you have a follow-up for Mark 14 or question on what you asked? () 15 MR. ATHERTON: Philosophically, does NRC consider 16 it acceptable to have a complete blackout of the spent fuel 17 storage area as a general proposition, to have a design such 18 that you w ?ld have a complete blackout of the spent fuel 19 pool storage area, without any attempt to prevent the 20 blackout from happening? 21 I'm getting that vibration from you. 22 MR. CAMERON: Let's 9 to Gary to address that. 23 MR. HOLAHAN: The ans-er to that is that's what 24 we're trying to decide. We're looking at probability and l 25 consequences of those various things to decide what levels
, ANN RILEY & ASSOCIATES, LTD. \ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
306 1 of protection are appropriate. 2 We haven't decided before the study. We're going 3 to decide after the study. And if the challenge frequencies I 4 are exceedingly low, then we recognize that, yeah, it may be 1 5 acceptable to have a complete blackout for exceeding the low 6 probability events. For other situations, it might not. 7 But I don't think it's a philosophical or a policy 8 question. It's a technical question about what are the 9 safety implications of that situation. 10 MR. CAMERON: Thank you. We're going to go to 11 Mike for a wrap here. I 12 MR. ATHERTON: One last follow-up question. You 13 did not indicate whether or not the liner, in addition to { 14 the concrete, would strengthen with time. The question I ' 15 would have with regard to the hardening of the concrete 16 would be does that necessarily mean that it becomes 17 stronger, as you. suggested, with time and is able to 18 withstand a greater seismic motion with an increase in 19 hardening and the lack of the potential for flexibility. I 20 see Mr. Kennedy -- 21 MR. CAMERON: Dr. Kennedy, do you want to address 22 this line or issue? 23 MR. KENNEDY: I was not involved in the fragility 24 analyses that Livermore did, but I've been involved in many 25 other fragility analyses and various fragility reports that ANN RILEY & 3SSOCIATES, LTD. m Court Reporters 1025 Connecticut Avenue, IM, Suite 1014 Washington, D.C. 20036 (202) 842-0034
307 1 people have used for doing these analyses. (} 2 For the liner, for the pool, the strength comes 3 from the concrete and the reinforcing steel in the concrete. 4 The way this fragilities ara credicted is as long as the 5 concrete and the steel in the concrete can carry the loads, 6 the liner basically is not part of the load-carrying system. 7 It is a leak-preventing system, because concrete leaks. 8 And so the fact is that the strength is probably 9 not degraded any over reasonable life and as long as the 10 concrete has retained -- the concrete and the reinforcing 11 steel in the concrete have retained their strength, the 12 liner is not really participating in a load-carrying aspect. 13 Now, once the concrete goes non-linear and you 14 start developing cracks in the concrete, that's when you~can () 15 pull the liner apart. So typically these evaluations stop 16 at the point that s;gnificant non-linear behavior of the 17 concrete occurs. 18 MR. CAMERON: peter, we're going to have to move 19 on. 20 MR. ATHERTON: We haven't resolved one issue here. 21 If the concrete leaks, for instance, have you then 22 compromised its strength, as such? 23 MR. BAGCHI: No. 24 MR. KENNEDY: As long as the liner is there, the 25 liner prevents the leakage. The liner is there to prevent ['N ANN RILEY & ASSOCIATES, LTD. \-
% Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ,
Washington, D.C. 20036 (202) 842-0034 l
308 1 the leakage, but the liner doesn't carry the loads. It's () 2 3 the concrete and the steel in the concrete that's carrying the loads. 4 Now, when the concrete goes non-linear, that could 5 crack the liner. If the liner is cracked, then you assume 6 that the pool is going to drain. You don't assume that the 7 concrete is going to prevent drainage. It's the liner that 8 prevents the drainage. It's the concrete that provides the 9 structural support. 10 MR. CAMERON: Dr. Kennedy, could you talk to Peter 11 during the break more about this? I want to go to Mike, but 12 Greg Withrow had a footnote to something that Glen said
-13 about.his particular situation.
14 MR. WITHROW: I think that you'll find at a lot of () 15 the facilities, that they do still have diesel generators. 16 We actually have two on-site, we have a diesel fire pump, i 17 electric fire pump that can be run off either diesel 18 generator. I know Maine Yankee has brought-in a temporary 19 diesel generator. So there are a lot of additional 20 capabilities that exist. 21 Now, that may not be true at every site, though, I 22 -don't know that. ! 23 MR.' CAMERON: Okay. Mike. 24- MR. MEISNER: Thanks. I guess one thing I was 25 struck by in the two different presentations is that there l . ANN RILEY & ASSOCIATES, LTD. (m/ Court Reporters ! 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202)- 842-0034 :
l 309 1 is.a good deal of convergence. I suppose we could argue is () 2 3 it ten-to-the-minus-six or five-times-ten-to-the-minus-seven, but I think what Goutam 4 brought up was maybe the bridge between there, and that's 5 the idea of a checklist. 6 I think that's quite similar to the concept we 7 were talking about yesterday as far as identifying some 8 commitments that the staff could then rely on to go back and 9 change their model. What I heard were fairly 10 straightforward things like a walk-down of the site and 11 ensure that obvious fragilities aren't apparent. 12 If we could kind of flesh that out, I think that 13 might go a long ways to resolving any outstanding i 14 disagreement we have on seismic. () 15 MR. CAMERON: When should we try to flesh that 16 checklist out? And do people agree that the checklist may 17 be the convergence between any discrepancies? Gary? l 18 MR. HOLAHAN: I think that's a reasonable approach 19 and it seems to me the checklist, if you do it right, will 20 also address this issue about whether, on the average, the 21 plants are okay or some of them are substantially different 22 from the average. 23 Ultimately, you have to say that all the plants 24 are safe. That's the way we make the decisions. Some of 25 those plants may be higher than others because of things O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 + (202) 842-0034
310 1 that the checklist could deal with. So I think that's part 2 of developing a useable checklist. b'$ 3 MR. CAMERON: Thank you. Why don't we take a 4 break? 5 MR. NELSON: Then we'll go on?
-6 MR. CAMERON: Yes.
7 MR. NELSON: To do the crane heavy loads. C MR. CAMERON: Do you want to discuss the -- is 9 there a need to discuss the mitigation? I think that was 10 next on the agenda for seismic. I think we're done. 11 MR. MEISNER: We're done, yes. 12 MR. NELSON: That's what I meant. So we would go
-13 on to the heavy loads as a group session and then this 14 afternocn do the thermal.
() 15 MR. CAMERON: Let's talk about it. We'll talk 16 about this during the break. Let's be back at 25 to 11:00. 17 That gives you about 20 minutes. 18 [ Recess.] 19 MR. CAMERON: Dick Dudley is passing out a copy of 20 something. What is that, Dick? i 21 MR. DUDLEY: The industry perspective, j 22 MR. CAMERON: So this is Bill Henry's slides.. 23 Okay. Two things I should mention to everybody. One is -- 24 and I checked this with Alan -- is that on the checklist 25 that we talked about in regard to seismic, NEI is going to l /~T ANN RILEY & ASSOCIATES, LTD. ; ks/ m Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 4
311 1 provide a proposed checklist on this to the NRC and as with () 2 3 the study that you guys are doing, that this will be a public document, I take it. 4 MR. NELSON: But the things that we commit to, 5 we're not going to necessarily wait till they all get 6 together. Everything will be going on. 7 MR. CAMERON: Sure. Okay. Just to give you where 8 we are cn1 the agenda. We did raake some adjustments here. 9 We're going to do the heavy loads discussion now. It was 10 originally scheduled from 1:00 to 2:30. We're going to do 11 it now. That means we will not have to have breakout 12 sessions. So that after lunch, we're going to move into the 13 thermal hydraulics session. 14 After that, we're going to go and see if there's I () 15 any water in this pot that's simmering here, have a 16 discussion of that. There are some other issues that we 17 were going to talk about in terms of process perhaps, but I 18 think this is the most important thing to go to. I 19 But right now, it's heavy loads. After lunch, 20 we'll start with' thermal hydraulics. Our heavy loads person 21 from the NRC is George Hubbard. George, are you ready to do 22 your presentation? 23 MR. HUBBARD: I'm ready. 24 MR. CAMERON: And-is that microphone suitable for 25 you? O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
312 1 MR. HUBBARD: Yes. I'm George Hubbard, with the I 2 , ( -) NRC, and actually, Chip, I'm going to let Ed do the 3 presentation. He's the one who did the study. I'm just 4 here to kind of kick it off. 5 As yesterday, in Rich's comments, this was one of 6 the sequences that was identified as credible. So with that, 7 I'm going to turn it over to Ed and let Ed kick off the 8 study that we did. ' 9 MR. CAMERON: And, Ed, could you just introduce j 10 yourself, and then Brian, and then we'll give Bill Henries a l 11 chance to introduce himself before we get started. I 12 MR. THROM: Yes, My name is Edward Throm. I work ! 13 in the Probabilistic Safety Assessment Branch. I was asked 14 to look at the heavy loads issue when this project first r'T ( ,) 15 started. So that's basically what I did on this project for t 16 this presentation. 17 MR. CAMERON: Okay. 18 MR. THOMAS: I'm Brian Thomas. I work in the 19 Plant Systems Branch. My focus there is sometimes on heavy 20 loads. 21 MR. CAMERON: Okay. Thank you. Bill? 22 MR. HENRIES: I'm Bill Henries. I'm the l 23 Engineering Manager at Maine Yankee. As the other 1 24 gentlemen, I was asked to look at the heavy loads, too. 25 MR. CAMERON: All right. Good. Well, let's go to i [~ ) ANN RILEY & ASSOCIATES, LTD.
'\m / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 j 1
313 1 Mr. Throm. lf%)
%/
2 MR. THROM: What I wanted to do was go through 3 where we looked at information, kind of what that 4 information was telling us, and basically how we got to the 5 bottom line of coming up with a risk estimate, and hopefully 6 I'll be able to clarify why things may be different than 7 they were from the NUREG-1353 numbers. 8 The first -- first of all, we're kind of looking 9 only at cask handling in terms of heavy loads at this 10 particular point. To alleviate Ray right now, I will say, 11 at the end, additional heavy loads is one of the items we'd 12 like to bring up at this meeting. We know that NUREG-0612 13 guidelines would probably stay in effect or would stay in 14 effect, but we don't have a real good handle on the types of () 15 things we may need to anticipate or expect as the plant 16 decommissions. 17 So from the perspective of looking at heavy loads, 18 casks being heavy, and kind of focusing in on this 19 particular study to look at the consequences of heavy loads 20 on pool integrity and loss of inventory. 21 In NUREG-0612, this was the control of heavy loads 22 in nuclear power plants, this was probably the one of the 23 first studies that tried to quantify the risk of heavy load 24 movements and basically some Navy crane data was used to 25 assess what the drop per lift frequency might be. ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034
314 1 There is uncertainty in those numbers because in
/'h 2 this area, one knows when things went wrong, they don't b
3 really have a good handle on the success rate. OSHA tends 4 to be the governing body for reporting accidents and mishaps 5 and they deal with property damage and injury to humans or 6 fatalities to humans. 7 So we know what does go on and we do know why it 8 goes wrong. In terms of picking up a frequency of how often 9 does this happen, that's an area that will always probably 10 remain uncertain. 11 There were two types of handling systems that were 12 looked at in 0612. There was what was called the non-single 13 failure-proof and what was basically touted to be the single 14 failure per system. On the last slide, there is a ! () 15 definition of single failure-proof, so you get an idea of 16 what we mean by single failure-proof. 17 In the NUREG study, part of the non-single 18 failure-proof system included an assessment of common mode 19 failures. 'Again, on the last slide, there is a little more 20 information about what was looked at, and that would be 21 basically an operator or a crew who failed to do some 22 pre-operational checks, didn't realize there was a limit 23 switch failure, and -- not yet. 24 MR. CAMERON: I-was wanting to get to the last 25 slide. N ANN RILEY & ASSOCIATES, LTD. ("'/
\. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I { 315 1 MR. THROM: There are two definitions, just (V ') 2 verbiage, to understand what we mean about a single 3 failure-proof system, and what we mean about common mode. 4 In the NUREG, one other thing they looked at wac a l 5 common mode failure under a single failure -- non-single 6 failure-proof system, which elevated the drop frequency by 7 about a factor of ten. 8 Also, in the NUREG, in figure B-3, there was an 9 evaluation of the single failure-proof crane and it came up 10 with the ten-to-the-minus-four, ten-to-the-minus-seven drop 11 per reactor year estimate, and included in that were events 12 that would lead to the crane failure, the so-called 13 tube-blocking event, which is where the lift and the crane 14 hit, failure of the limit switch, overload of the crane, and rx () 15 failure of backup components. 16 And that was about l 17 eight-times-ten-to-the-minus-five or almost a larger portion 18 of that ten-to-the-minus-four per reactor year frequency. l 19 If we can go to the next slide, please. 20 Another thing we have as background is NUREG-1533, i 1 21 and in that report, alt'..ough it wasn' t necessarily used as 1 22 part of the evaluation, the drop per lift uncertainty range 23 was estimated to be ten-to-the-minus-three times 24 ten-to-the-minus-four. The specific point used in that 25 report was a six-times-ten-to-the-minus-four drop per lift. ANN RILEY & ASSOCIATES, LTD. (~)) (_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l l 316 1 It was convoluted or compared with something on the order of x 2 200 lifts per year and a certain frequency of the casks 3 being over the wall. 4 And that number actually came out to be something 5 on the order of three-times-ten-co-the-minus-four. The A-36 6 contributor was reviewed, the impact of A-36 or NUREG-0612 7 was reviewed by the contractors working on that project, and 8 they determined that about a factor of a chousand reduction 9 was expected from implementation of NUREG-0612 guidelines. 10 That came down to the 11 three-times-ten-to-the-minus-seven number. Part of the 12 Generic Issue 82 study included work done by Livermore, 13 which came up with about a one in ten chance of failure of 14 the wall given the drop on the wall. That was the 15 three-times-ten-to-the-minus-eight number, which is being 16 used as what one would call the best estimate number for 17 cask drops in 1336. 18 What was not considered in 1336 is dropping d. t 19 into the spent fuel pool. I don't know why it got missed, 20 why it wasn't covered, but we went back and we looked at it 21 and we said, well, you only really looked at wall failure 22 . probability in 1353 or as part of Generic Issue 82. l 23 We went back and looked at it and said, well, we l
- 24- really have to contend with the drop into the spent fuel !
25 pool, and we're going to get very plant-specific features a 4 ANN RILEY & ASSOCIATEO, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
317 1 little later on the presentation and say why at some plants () 2 this is and why at some plants this is not necessarily a 3 problem. l 4 The other set of data we had was a study done at 5 Savannah River for non-nuclear facilities. They looked at 6 about 200 drops in 200 crane years of operations, nuclear 1 l 7 plant data, and basico.lly came up with a range of events in j 8 the reports, each one of these is assigned to a category. 9 The ten-to-the-minus-three number would be associated with 10 unusual loads, ten-to-the-minus-four would be associated 11 most likely with the type of load we're concerned with, and 12 ten-to-the-minus-five would be a load that was very, very 13 standard. 14 The question now you have to ask yourself is, is 15 100-ton lift a standard operation or a typical operation as 16 compared to the databases, which tend to be related to like 17 five-ton cranes or the handling of 55-gallon drums in the 18 pilot waste program. ! 19 So that becomes an area where maybe there is 20 better data to look at what type of crane we should be l 21 considering compared to the generic databases. Even the 22 Navy data, when you go over the gamut of the types of cranes 23 and lifting devices, it's open to judgment as to how much of l 24 it is really germane. 25 One more point on this slide is just to r int out
~)
ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
318 1 that when we talk about mechanical electrical crane failure i 2 rates in NUREG-1353, we looked at a [%.)) I 3 three-times-ten-to-the-minus-six per operating hour. If you 4 went back and looked at the Savannah River, for example, and 5 drops per operation and said it took an hour to do an l I 6 operation, you would see that the mechanical electrical 7 contributor to the risk is probably ten percent or less of 8 the overall risk as compared to the potential human 9 component in the drops. l I 10 Human errors right now are the issue that are l 11 probably of concern or at least they're identified to be of 12 concern. In NUREG-0612 or some OSHA data was addressed that 13 said rigging led to 34 percent of the errors, operator 14 errors were 42 percent of the reported incidents. In the () 15 Navy data, it looked like seven percent rigging and operator l 16 errors were about 70 percent. 17 Go to the next slide, please. 18 Doing an internet literature search, which is one 19 of the easy ways of doing business this day, I came upon 20 some information the Department of the Interior, in their 21 Mineral Management Services, and they did a couple of 22 studies, the details are, I think, covered in the report, 23 and their conclusions were employee negligence, 40 percent i 24 of the errors were due to poor maintenance and overloading. l l 25 There was a follow-up study done and a crane accident O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
319 l 1 workshop to that study and they came up with 35 percent of l ( 2 the incidents were due to human error.
~3 They were looking at 34 specific accidents over --
4 incidents over about a three-year period. l 5 The last study that we looked at was the DOE study 6 and for people who are interested, there is the web site, on 7 the bottom, one can go to that web site and that report, at 8 least as of Wednesday, was still available. People have 9 things that disappear off the internet, you can pick it up. 10 And that concluded that the human error was 68 11 percent of the incident rate and work planning led to 12 another large part, 18 percent. An interesting point to 13 make is that training-related deficiencies seemed not to be 14 a' specific problem; in other words, they didn't find () 15 anything directly related to training as being a contributor 16 to the human error, although they did indicate that about 17 nine percent of the problems resulted from inadequate 18 procedures. 19 So what we're basically saying is the human error 20 rate, the incidents have not changed substantially from what 21 we were looking at in NUREG-0612 back in 1980. l 22 Next slide, please. l 23 So we took off as a jumping point basically those I 24 numbers, not having anything'else really to rely on. There l' 25 are two lines there. There is the non-single failure-proof (O ANN RILEY & ASSOCIATES, LTD. , \ms) Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) -842-0034 o
320 1 system. It's on there for completeness. The one that we're () 2 3 really keying in on is the single failure-proof system, and if you go back to NUREG-0612 and you start looking at that 4 ten-to-the-minus-four per reactor year number, what you have 5 to factor in is a couple of additional things. 6 The load path, the entire load is not necessarily 7 being handled in a critical position. While I would not 8 like to see it happen,.if, on the first time on lifting the 9 cask, I happen to drop it, it's not going to lead to a 10 failure of the spent fuel pool. 11 In the NUREG, they were estimating that between 12 five and 25 percent of the total lift would be near or in 13 the spent fuel pool. We took that number and basically said 14 for dropping it on the wall, if that total period of time or k 15 ten percent of that time is over the wall and then said, 16 from the old, from the Lawrence Livermore study, that only 17 ten percent of that time would you have damage to the wall 18 you would come up with a ten-to-the-minus-seven number of 19 damage to the wall. 20 Now, we haven't done any calculations to say 21 whether this is a large leak or a small leak, but that was 22 the number that we came up with for potential damage. 23 And if you want to look at dropping it on the 24 floor, again, all we did was we went back and used the 25 NUREG-0612 number for the failure of the single O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
321 1 failure-proof crane and said that there was a one in ten () 2 3 chance that if it was dropped while it was in the area of the spent fuel pool, that it would cause damage to the pool 4 floor. 5 If we can go to the next slid 2 . 6 Things to be considered with tL number are we are 7 aware that there are plants out there that have segregated 8 cask loading areas which basically isolate the cask, lay 9 down area inside the spent fuel pool, from the remainder of 10 the spent fuel. If you drop it in that area and you did 11 damage that area, depending on whether the gates were in or 12 out, you would not drain the pool below the fuel, that's 13 part of that design consideration and now we're talking 14 about mitigating or making up water to accommodate any 15 boil-off rate. 16 The other thing that there may or may not be in 17 plants are these cask crush pads. They seem to be one 18 answer to the situation, if cask handling becomes a problem, 19 is that these things are designable, they can be 20 implemented, they're energy-absorbing devices that basically 21 mitigate the event. 22 Of course, part of the other consideration that we 23 have in the area is that there have probably been specific 24 cask drop analysis done to demonstrate that the designs are 25 adequate, and I need to make a number of points here. /\ ANN RILEY & ASSOCIATES, LTD. \s- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
322 1 In these numbers that we are putting out, we were ( 2 looking at something on the order of an upper bound of 200 3 cask movements per year, and that's where we're up in the 4 ten-to-the-minus-seven or ten-to-the-minus-six probability 5 of a drop, if you.want to add in what might or might be 6 reasonable cask handling. 7 The last slide, this is where we are hoping to 8 open the dialogue. One of the big questions is, is the data 9 that is generally used to characterize drops per lift, and 10 even the data that comes out of looking at crane and hoist 11 data from the industry, is is it really germane to the 12 issue. Most of these cranes tend to fall outside your 13 concept of the nuclear power plant in that they're mobile 14 cranes. These are the ones you see at construction sites () 15 where heavy load tips them over. 16 There is also a tendency for a lot of this 17 information to be related to hoists and potentially lower i 4 18 load lifting devices, I characteri.2e them kind of in the 19 five ton range. So I don't have a good handle on that. 20 Another issue that needs to be considered is 21 cask-handling plans, not only the frequency that we might 22 expect for these things to happen, but the load path 23 consideration. 24 And, of course, if you really look at this 25 particular issue, when do you do these movements in terms of ) i i O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 j <- 1 I
323 there is a reduction in risk of a release if the fuel has 1 1 l () 2 3 been sitting in the pool for a very, very long time. I would say that when we started looking at the l l 4 200 lifts per year, what was envisioned was a facility that l l 5 had recently shut down, but had 30, 20 year old fuel in the l 6 plant, had just recently discharged and wanted to start l 7 getting that inventory out, and that seemed to be like a 8 maximum number of these types of loads that you might want 9 to consider, even while you may have fuel that would be 10 vulnerable if you did lose the inventory. 11 So the other two questions that we alr1 have for 12 the industry is -- is the human factor consideration. I l 13 mean, in reality, what drives drops per lift is the human 14 involvement. The DOE report, if you read it, it will O) ( , 15 discuss what they identified as management-related problems 16 and personnel-related problems, and what we'd like to know 17 is -- or have a database that says at nuclear facilities, 18 things are quantifiably better. 19 When I look at the DOE information, the reason I 20 bring it up is they were talking about nuclear waste. It 21 may have been low level waste, waste pilot program, but 22 nevertheless it's a risk that the operators would understand 23 in terms of releases as far as a mind set would go. 24 And the other question, E.s Ray brought up this 25 morning, is at this particular point, we don't have any idea
/~ ANN RILEY & ASSOCIATES, LTD.
k_' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D..C. 20036 (202) 842-0034
324 1 of what other heavy load handling near or over the spent 2 fuel pool people might be connidering. We would fully 3 expect that the guidelines in NUREG-0612 would remain in 4 place and that safe load paths would be defined. But the 5 problem is people don't always do it. There are interlocks 6 that can fail, a procedure could be not followed. So these 7 are the alaas that we would like to have the dialogue on to 8 see what we can do about better understanding the heavy load 9 drop risk. 10 That's my presentation. 11 MR. CAMERON: Thank you. That's great, Ed. Thank 12 you. It sets the stage for discussion after we hear from 13 Bill Henries. Bill, you may be going to put some things on 14 the table that relate to what Ed said. 15 MR. HENRIES: Thank you very much. Ed, you know, 16 I think we look at the glass and I think we're very close. 17 Ed's presentation and mine meld in many, many areas. It's a l 18 question of, from our perspective, are we getting an upper l 19 bound or a best estimate again, same issues that were raised 20 . yesterday, 21 We implemented the close-out of A-36 by doing all 22 the work for NUREG-0612, the 1353 report quantified these 23 numbers, as Ed touched on, and yet we've done all this work 24 over ten years and we've lost two orders of magnitude on the 25 risk. Q b/ ANN RILEY & ASSOCIATES, LTD. Court Reporters
-1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036-(202) 842-0034
r 325 1 1 1 So it seems like did we do something wrong or are ! ~x (
%-)
\ 2 we just not getting the credit that we had in these past' )
l l 3 reports. 4 Human error is indeed a higher failure rate in all 5 of these past studies and one of the things I want to bring ) 6 out is that I don't think these past studies are that 7 germane to the way we're going to handle a spent fuel 8 shipping cask at a nuclear power plant. These are one of i 9 the last rate activities we're going to be doing and it will 10 be truly management and oversight intensive. 11 Next slide, please. 12 One thing I think we agree with Ed on completely 13 is there is really only a need to look at single 14 failure-proof cranes. If you're not using a single 0) s, 15 failure-proof crane, following NUREG-0612, you have to do l 16 the consequence analysis, becomes a moot point, you've 17 evaluated the incident, let's just look at it for single 18 failure-proof. 19 Next slide, please. 20 Slight, I guess, disagreement between Ed and I is l 21 that he looked at some of the newer data, primarily the 22 Savannah River, which brought out some more data, but once 23 again, a generic type of information.
- 24 Like Ed pointed out, the five-ton crane out on th0 25 barge out in the back. We're looking here for heavy loads
[))
'Nm ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 , (202) 842-0034 l l
I l 1 l 326 1 over a spent fuel pool as a very specific activity, a very () 2 high profile activity at the power plants, and I think it's 3 going to get much, much more oversight, such that the very ; 4 high human error potential that you saw with wind blowing 5 over yard cranes and things like that. 6 It's really not applicable and I think even though 7 Ed looked at some of the newer data, we ended up going back 8 to the 0612 numbers. Well, with the 0612 numbers, does that ; 9 adequately give us credit.for all these things we . 1 10 implemented in 0612 that made the issue go away? It wasn't 11 clear to us that we were getting any credit really in that t l 12 arena, in that we were once again trying to quantify an 13 upper bound and not a best estimate. 14 Next slide. I () 15 A couple of I think -- I call them somewhat 16 nit-picks, but what I questioned in the report was there did ! 17 seem to be an increase for the single failure-proof crane of 18 going up to 25 percent over the pool instead of ten percent 19 that was in the non-single failure. That seemed to be a 20 factor of two and a half that hadn't been applied before and 21 was now thrown in; once again, is that an upper bound, 22 because I think you argued it was five to 25. So we have 23 the 25 number. i 24 MR. THROM: I'm not here to defend 0612. One of 25 the things in NUREG-0612 is when they looked at the ANN RILEY & ASSOCIATES, LTD. s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l \
327 1 non-single failure-proof system and they looked at the i, (} 2 single failure-proof system, there was some inconsistency in 3 the way the data was presented. 4 In the non-single failure-proof system, they had 5 the ten-to-the-minus-three, ten-to-the-minus-four drop per 6 load and the way they determined whether or not that load ] 7 was over what they called hot fuel at that time was done 8 differently. > 9 If you go to the NUREG, I have the fault tree j l 10 here, if we want to get into that level of detail, and you 11 look at the table or the fault tree for the single 12 failure-proof system and you look at the point at where they 13 came up with what is the probability of that single 14 failure-proof crane failing, it was made up of two parts. (em) 15 It was made up of the ace blocking event, the overload, the 16 failure of the interlock and the backup system failures, and 17 also an assessment of the double-rigging, if the operator 18 failed to rig properly the first time, the second rigging 19 would also fail, and that came up with that 20 ten-to-the-minus-four number, which is on the upper bound. 21 And then when you go further through the system, 22 what you do is what was done in 0612, is they tried to 23 estimate how much of the load path was in a critical area, 24 and that was the five to 25 percent. Again, the number on 25 the top is the 25 percent of the operation and then past
/ ANN RILEY & ASSOCIATES, LTD. \s- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
328 1 that, we kind of used factors of ten to quantify, at that 2 total time, when would you be in a mode to fail the pool. 3 So if it's dropped from the 40-foot elevation, 4 failure; if it's dropped from the 20-foot elevation, not j 5 failure. That's where those other numbers -- so that's the 6 specifics of how the numbers were derived. 7 So I'm not sure where the 2.5 -- I think it l { 8 addresses the 2.5 in that the non-single failure-proof event 9 tree was done differently from the one we were focusing in i 10 on to present what we thought were some relative numbers for 11 the cask drop. 12 MR. HENRIES: That's exactly where I thought you 13 were coming from and that, as you pointed out, five to 25 14 percent was the number and the report used the 25. () 15 And I understand what you're doing and I'm just 16 trying to make the case, to the industry, it appears like l 17 we're focusing more on an upper bound. That's all. 18 Similarly, the 200 lifts per year, I can't see 19 anybody doing more than one a week with the activities that 20 have to go involved. I think the original numbers that were i 21 in WASH-1400 are in 1453, arguing for 100, you know, one a 22 week type of thing. By the time we prep the cask, get it l 23 into the pool, get it loaded, get it out of the pool, get it 24 dried, get it full of helium, get the lid on, get it welded, 25 get it inspected, get it out of the way, we're not going to l ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
329 1 be rushing to do two a week.
/~
( 2 So I think there's another factor of two in there j 3 that one could argue about. It's one of the things that 4 perhaps if it made a pass / fail difference, people could 5 argue whether they wanted that factor of two or not. 6 Next slide, please. 7 I think the heart of our argument is we did 8 implement 0612. We have worked very hard over the past 9 decade to make sure that lifts, and particularly 10 safety-related lifts in the spent fuel pool are very 11 management-intensive. 12 The items -- you know, we're controlled either by 13 tech specs or FSAR. We are going to single failure-proof 14 cranes or else it's a moot point, we'll have addressed the () 15 drops. 16 We'll have analysis and procedural controls, all 17 of them checked generally and all of them checked in 18 accordance with Appendix B. Enhanced management oversight. 19 I envision, just as DOE recommends, everybody is going to 20 have a lift director. You will know who is in charge. 21 These things are going to be coming in under very strict 22 controls. 23 We've always -- well, at least since the institute 1 24 of 0612, we have safe load packs, we have verified rigging, 25 and our operator training has been ongoing along with
\ ANN RILEY & ASSOCIATES, LTD.
[~'/ (_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
330 1 pre-job briefings. 2 At Maine Yankee, this will be considered an (} 3 I unusual event, when we first start out, as a minimum, you 4 know, we have all kinds of procedural controls to have 5 readiness assessments performed. 6 This is going to be, as I mentioned earlier, the 7 last great activity in a nuclear power plant for 8 decommissioning. So the human error aspect, I think, is, by 9 using generic industry or DOE data, has over-sold, to some 10 extent, to a great extent, in my opinion, the likelihood of 11 failure. 12 Next slide, please. 13 These are more opinion than anything else, but 14 it's based on my personal analysis of several PWR pools; () 15 that the one in ten causing significant damage is too 16 conservative. I don't know how much work went into its 17 development, but it's one of the things that was carried on, 18 so I'm not questioning it. 19 But it seemed, once again, where we carry some 20 numbers on, but then we take a worse estimate in a few 21 others, starts to lead the industry towards we're getting a 22 more upper bound evaluation. 23 And as you have mentioned, the timing of the lifts 24 is dependent on the zirc fire. We agree completely on that 25 one.
^
) ANN RILEY & ASSOCIATES, LTD.
' [s / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
331 j 1 Finally, the last slide. 2 ( As I've probably overstated, and I apologize, it 3 really looks to me like we're publishing an upper bound, I 4 particularly when you show 2.5E-to-the-minus-six rc;> Ting to 5 E-to-the-minus-eight and we're at exactly one tail. It 6 would seem that a best estimate has to be somewhere in the 7 middle and the middle is probably in to-the-minus-seven 8 range of a number similar to what was in 0612 or 1353. 9 So I don't think there is a large area of 10 disagreement. It's just a question of are we doing an upper 11 bound versus a beat estimate. . l 12 Thank you. 13 MR. CAMERON: Great. Thanks, Bill. Can we go to 14 the NRC team first to get a reaction to Bill's presentation? 15 Then we'ra going to go back over to Mike. 16 MR. THROM: I think Bill has plotted out the areas 17 where we have concerns. We were not married to the numbers. 18 What we looked at, and, again, this was a very, very quick l 19 review of trying to revisit the issue, was the industry data 20 suggests that the problems exist, still exists, the types of 21 problems that led va to 0612. 22 Maybe Brian Thomas of the staff can address the 23 implementation of 0612 and what it really means, and we know 24 there are a lot of benefits to it. In terms of the damage 25 likely, that is a number that is open to interpretation. Of
; ANN RILEY & ASSOCIATES, LTD.
( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034
332 1 course, it's handled by plants that have done load drop 2' analysis. So you're right in that perspective, that, again, 3 it was just trying to look at not all drops are going to 4 lead to a problem. < 5 To do much better without information in an order 6 of m6gnitude would not tell much, 20, 50, I mean, that's 7 pretty much where we came from. j l 8 So what we'd like to do and I think what we're 9 getting is a better understanding of the care s.nd attention 10 and as you read the report and follow along to what I 11 specifically put into the report was the outcome of the DOE 12 result was that what you really do is have to have a program 13 that instills in both the management and the people actually 14 doing the work that that care and attention needs to be paid 15 to the process. 16 And basically when we go back and say if you were 17 following those guidelines, then you're really addressing 18 the human aspect of the problem and you're addressing the 19 risk from a risk-informed basis. In other words, you've 20 gone out, you've looked at what causes problems and you've 21 instituted programs and plans to assure that your facility 22 or, as an industry, you're not going to get yourself into 23 that situation. 24 So that's the kind of information we're looking 25 at, again,. also, whether or not what we need to do about any O, ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
333 1 other potential heavy loads, maybe we'll get to that, maybe () 2 3 we won't. But it is'a concern and since we don't have a feel for what might be going on, we pointed that out as an 4 area to consider.
'S MR. CAMERON: Brian, are you going to talk a 6 little bit about -- go ahead.
7 MR. THOMAS: Yes. I agree with you, 0612 is a 8 very rigorous document. There is a lot in there that l 9 licensees have implemented and taken advantage of. 10 What I've seen are the licensees that are in 11 decommissioning space is that they have gone -- they have 12 done an exemplary job of utilizing what's in 0612. They 13 have done consequence analyses and where they show the 14 results as being not as consequential as you need to be () 15 concerned about, they have utilized the energy absorbers, 16 the impact limiters. 17 I think Maine Yankee put in a single failure-proof 18 crane. Even after their analysis showed that they didn't 19 necessarily need it. 20 We are concernad that the guidelines, 0612, 21 there's a phase one and a phase two in 0612. There's all ; 22 the administrative and procedural measures that you could l 23 implement and then there is the need to do a consequence j 24 analysis or put a single failure-proof crane in. l 25 We have guidelines out there that say -- and i o ANN RILEY & ASSOCIATES, LTD.
- 1) Court Reporters 1025 Connecticut Avenue, NN, Suite 1014 Washington, D.C. 20036 (202) 842-0034
334 1 that's Generic Letter 85-11, says you don't need to do those
) 2 consequence analyses. We are concerned that the licensees 3 may just do that -- not do that consequence analysis. All 4 the licensees out there are not taking the approach that 5 Maine Yankee and Rancho Seco and Trojan are taking and going 6 to that extent to do extensive consequence analyses, and 7 that's where our concern comes around.
8 The other thing is the predominant contributor to 9 hazards in the plant being the operator error, that is 10 something that we're trying to get our arms around. 11 Data in 0612, we all agree, was full of 12 uncertainties and, again, the current data that Ed has 13 looked at, as he stated, is loaded with uncertainties. 14 We think perhaps there is room here to look at O 15 ( j better data, look at a way of analyzing it, and also look at 16 the need to do with other licensees, as is done in the , i 17 Maine Yankee case, to go beyond what the guideline says. 18 MR. CAMERON: Let's go to Mike, after he's heard i 19 both of you. Mike? ! 20 MR. MEISNER: Thanks. I think this is a good tert 21 case for us on heavy loads. We've been talking foi two days 22 now about crediting licensee commitments and reflecting that 23 in risk reduction. 24 Here is a case where you don't need a licensee i 25 commitment because tne licensee already has their ANN RILEY & ASSOCIATES, LTD. Os' Court Reporters ! 1025 Connecticut Avenue, NW, Suite 1014 Wushington, D.C. 20036 (202) 842-0034
335 1 requirement in their license basis. 0612 is there, that 2 doesn't change when you go from an operating facility to a ) 3 decommissioning facility. 4 And I think this is one where -- that this whole 5 issue should go away by inspection, given that there is 6 credit for 0612 implementation. 7 As Bill pointed out, what the staff did was they 8 took the numbers that were used to develop 0612 and didn't 9 modify them at all for the implementation. 10 I think it was stated that in 1353, when the 11 analysis was done, because of 0612 implementation, there was 12 as factor of about, what, a thousand, is that what you said? 13 MR. THROM: Yes. 14 MR. MEISNER: A thousand reduction in risk. A i ) 15 That's something that the staff should apply right away and 16 we shouldn't be spending a lot more time on this issue. I 17 understand that you need to look at things other than cask 18 drops and, of course, that gets back to the issue of 19 completeness. But here we've got an opportunity to do what 20 we've been saying for two days, and that's credit licensee 21 implementation. l 22 You don't need new commitments from licensees for 23 that. You guys know what we do in 0612 and that's something 24 that we should be able to resolve very quickly. I 25 As far as what was brought up about consequence analysis, ) l /) ANN RILEY & ASSOCIATES, LTD. 1 (~ / Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 ' Washington, D.C. 20036 (202) 842-0034
336 1 that goes -- that's a bit of a concern to me, because a
) 2 licensee over-performs, and almost all of us do, the staff 3 shouldn't be looking to institutionalize new requirements 4 simply because of that. 0612 speaks for itself and it gives 5 you the choice, as it was pointed out, that you can either 6 do your consequence analysis and go with the 7 non-single-failure-proof crane or you can get a single 8 failure-proof crane and dispense with the consequence 9 analysis.
10 I don't think there is any reason to go beyond 11 that. And we shouldn' t be looking for m ' 4.ng a search for 12 new staff requirements coming out of this, but like I said, 13 if this issue can't be resolved real quickly and credit 14 given for licensee implementation, then probably none of the 15 rest can either. 16 I'll throw that over to the staff.
- 17. MR. CAMERON: Let me ask you to clarify one thing, 18 Mike. When you said be resolved fairly quickly, do you mean 19 in space today or in staff going back and looking at things?
20 MR. MEISNER: It seems.to me you could deal with 21 it in a breakout session, but we're already at 22 ten-to-the-minus-six level. You should be able to, by 23 examination, I would think, give an additional two to three 24 orders of magnitude credit for licensee implementation of 25 0612 today. Go ahead and document that in your report and
\
ANN RILEY & ASFOCIATES, LTD. w- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
i 337 1 go on to other heavy loads issues. ( ); 2 MR. CAMERON: John, let's go to you first on this 3 one. 4 MR. HANNON: Just to follow-up and clarify an 5 assertion that Mike made about the level of management 6 attention that's given to heavy loads in decommissioning 7 reactors. 8 I understand you're saying you've already 9 committed to 0612, but there is a degree involved here for 10 regulatory purposes to enable the staff to give credit for 11 that, and I think we need to have some more discussion on 12 that. 13 For example, I know that some operating licensees 14 have a license condition in their license to adopt 0612 () 15 while they're operating and then when they convert into a 16 decommissioning mode, that license condition is taken away. i 17 So I'm not clear exactly on what the connection is 18 with your commitment. l 19 MR. CAMERON: Okay. Let's get a clarification on 20 that regulatory point. 21 MR. MEISNER: When you go into decommissioning, no 22 license condition changes are automatic. They can only be ; 23 implemented tnrough staff approval. I'm no'_ familiar with 24 licensees having 0612 in license conditions, but regardless, 25 it's up to the staff to approve it if you want to approve / ANN RILEY & ASSOCIATES, LTD. Court ; Reporters 1025 Connecti cut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
338 1 it. () 2 3 I'm talking about the integral part of our license basis has to do with implementation of 0612. That doesn't 4 change going into decommissioning. We can only change that 5 under 50.59 and the only real changes you would see with 6 respect to that is to scope. If we're no longer using kny 7 cranes in containment, or, more accurately, if, in fact, 8 we've moved all the fuel out of containment and it's only 9 sitting in the pool, everything else is non-safety-related, 10 then, of course, we'd reduce the 0612 scope for those 11 applicable cranes. 12 But you surely wouldn't do that with respect to 13 the spent fuel pool system. 14 The staff has all the commitments it needs to go () 15 forward on this and like I said, I think it would be a real 16 good test case of whether we can get anywhere with this 17 commitment and risk reduction approach. We've got 18 everything in all the dockets. You don't need more 19 inspection. You've been inspecting 0612 for years and your 20 inspectors go out and look at that at decommissioning 21 plants, just like they do at operating facilities. 22 The implementation is good. How many cask drops 23 have you heard of in the last five years? You know, I don't 24 know that you need to take a literature search through the 25 nuclear industry to kind of confirm that. / O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1
}
339 1 But anyway, it seems like a straightforward thing, 2 to me. 3 MR. CAMERON: Let's get some responses on Mike's 4 proposition about this being a good test case, should be 5 easily resolved. Stu? 6 MR. RICHARDS: I think we ought to just take the 7 ball and run with it. The industry is going to be committed 8 to 0612, so let's just assume that they are and that anybody 9 who, for some reason, isn't, would fall outside that 10 category and we'll have to deal with them on a case by case 11 basis. 12 So as long as they're willing to make that 13 statement, let's just take it. 14 MR. CAMERON: Do we have some -- when you say -- 15 and what are the implications of taking the ball.and running 16 with it in terms of -- l 17 MR. RICHARDS: I think what Mr. Meisner said is 18 just assume that 0612 applies to all plants, make that a 19 baseline assumption. So the body coming out of the 20 technical working group would be assuming that a plant is 21 applying 0612 and the probability of this kind of event is 22 whatever, and in the rulemaking process, we would take that 23 into account and just say in order to give whatever, you 24 have to have 0612 applicable to your plant. 25 MR. CAMERON: Mike, a comment on that or are there ANN RILEY & ASSOCIATES, LTD. (. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
340 1 other steps or is that good? () 2 3 MR. MEISNER: I'd just say the last step of saying that in order to take advantage of these, you have to use 4 0612 is a redundant step. We already have to use 0612. And 5 I'd also say that I don't want to minimize some of the other 1 6 upper bound assumptions either in the analysis and those 7 need to be addressed, as well. But that in conjunction with 8 0612 implementation clearly drives this down to a low enough 9 level that we needn't spend much more time talking about it. 10 MR. CAMERON: Okay. That's the issue we're 11 focusing on here. Glen? 12 MR. KELLY: In discussing 0612 here, I think the 13 staff has some concerns about what we're seeing out in the 14 industry, that there are some people who are not necessarily q i ) 15 committed to 0612 and we would certainly be interested in 16 not implementing it as fully as might be appropriate. 17 But I think that the idea that you have of using 18 the commitment as the adequate determination may be okay. 19 It may be that the numbers that we've calculated here 20 certainly do have uncertainty, but it's not clear that 21 exactly what is the correct number, because the numbers that 22 were in 0612 were highly uncertain and we continue to have 23 highly uncertain numbers. 24 If we assume that the probability of cask drop is 25 one in 10,000 and we've had a thousand cask movements, which l [~ ) ANN RILEY & ASSOCIATES, LTD. 's / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1 1 f
t 341
-1 I doubt that we've had, it would have been a very low
() 2 3 probability that we've actually had a cask drop event so far, 4 So the fact that we haven't seen one is not 5 unremarkable occurrence, whether or not it was one in 6 10,000, one in 5,000, or one in 100,000 cask movements. 7 But I think that we're certainly willing to 8 entertain that the appropriate implementation of 0612 9 provides adequate assurance for limiting the risk associated 10 with cask drops, to the point that we feel that it would be 11 potentially not cost-effective to go ahead and to do more 12 work. 13- But I think we'd have to sit and caucus some more 14 about that. I think that the potential is there, but we'd
) 15 have to talk a little bit more internally about it.
16 MR. CAMERON: So at a minimum, I guess, what the 17 NRC staff should take away from the table is do they have
- 18. all of the information that they need from the industry to 19 resolve this particular issue. Stu?
20 MR. RICHARDS: Just to make sure I understand, I 21 think what I heard, and I don't know much about it, but I 22 heard Mr. Meisner say that they would have a single 23 failure-proof crane and implement that part of 0612 that i 24 isn't the consequences analysis. Is that right?
-25 MR. CAMERON: I don't --
ANN RILEY & ASSOCIATES, LTD. O' Court Reporters 1025 Connecticut. Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
342 1 MR. MEISNER: What I'm saying is 0612 allows you s
) 2 two paths. One is use of a non-single failure-proof crane, 3 but only if you do a consequence analysis. And given the 4 consequence analysis, then this whole issue goes away. It 5 won't result in breaching the pool. That's a plant-specific 6 licensee initiated evaluation.
7 If you don't do that, now we're into the space 8 you're talking about, and that's single failure-proof crane 9 and implementing all the concomitant things that go along 10 with 0612. 11 MR. RICHARDS: One or the other then, right? 12' That's your commitment. 13 MR. MEISNER: Yes. That's what we're required to 14 do. ( 15 MR. CAMERON: Let me make sure that we all agree 16 on that. You keep referring to it as a commitment, Stu, as 17 i opposed -- 18 MR. RICHARDS: Let me explain why. When you go 19 and look at a 106 licenses, it's not unusual where you won't 20 find that there are outliers out there who somehow, through 21 .the regulatory process, don't fall' into that category. 22 To make it simple, let's just make that an i 23 assumption. Keep it simple. That's an assumption. 24 Somebody falls outside that category, for rulemaking or for
)
1 25 processing purposes, then they're dealt with on an i f ANN RILEY & ASSOCIATES, LTD. Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 I (202) 842-0034 ! l
343 1 individual basis. But that's the envelope. You're inside 2- the envelope, you're dealt with one way; you're outside, V[~D 3 then you're dealt with a different way. 4 MR. CAMERON: Okay. Now, even if you -- then I 5 think even if we accept -- we use that as a starting point, 6 then where are we in terms of resolving this problem, 7 crediting industry with the use of 0612? 8 MR. RICHARDS: I think you asked do we have all 9 the input we need, and I think clarify what the envelope is 10 and then the staff said they need to go back and run it 11 through the mill. But I think the proposal is pretty clear. 12 MR. CAMERON: All right. Then I think that's 13 consistent with the way Rich Barrett feels about this, that 14 we need it consistent; that is, we need to take this () 15 information and go back and do an analysis. Brian, we'll 16 get to you in a minute. 17 MR. BARRETT: I think that what we need to do here 18 is to make sure that we have all the information that we 19 need. The last thing we want to is to come -- to find two 20 weeks from now that we don't have information that covers 21 some case. So that would be our very minimum requirement 22 from the meeting, is that we make sure that whatever l 23 questions we have regarding this analysis, that we can get ! 24 information from the licensees about, that we get that 25 information. I l ANN RILEY & ASSOCIATES, LTD. Ox Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
344 1 Now, with regard to 0612, I've heard it suggested
. 2 that we just posit full implementation of 0612 and have an \
3 analysis that's based on full implementation of 0612, and 4 then any licensee that falls inside of that envelope is i 5 covered by the analysis and any licensee that falls outside i f j 6 of that envelope is not covered the analysis. 7 I don't know if that's the industry's proposal. I y 8 just -- my concern is that right now we're at a -- we're 9 going to make a technical basis that will allow us to go 10 forward and do what we neea to do. If that's the industry's 11 proposal, then we can analyze it that way. 12 But if the industry -- but if we analyze it that 13 way, then that's the analysis I have and we would have to go 14 back and redo it, if a proposal then for a specific plant () 15 was outside of that envelope. 16 So my concern is the same at Mike's. I want to 17 make sure we walk out of here without any loose threads. 18 MR. CAMERON: Sorry, Brian, we'll get back to you. 19 But is that the -- what is the -- the industry proposal, as j 20 Rich stated, comment, Mike? 21 MR MEISNER: I don't know if I'd call it a 22 proposal. I'd call it a statement of fa".ts. We've been 23 talking for two days about what assumptions should we make 24 in the analyses. Here is one that is clear and i 25 straightforward. It's not tied to a commitment that's in ANN RILEY & ASSOCIATES, LTD. Court Reporters 7025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
345 1 the future. It's tied to existing license basis of plants. f 2 If you know of any plant, Glen, I don't know who 3 the outlier is you're talking about that isn't implementing 4 0612, but if you know one, I would suggest that's probably 5 some kind of violation situation. And if you've got some 6 concerns about how well plants are implementing 0612, that's 7 en inspection NOV type situation. 8 You've got everything you need and 1 just thought 9 this would be a very simple case where given that, you can i 10 use this as a test case to see can you, in fact, credit { l 11 licensee actions in your analysis. 12 The 1353 has already established a policy that the 13 staff should follow, and that's give us a three orders of 14 magnitude credit for implementing 0612. We can make it real () 15 complex, but that's it. 16 MR. CAMERON: I think we probably should be 17 careful about talking about outliers that may be in 18 violation space or whatever, but can we -- just assume -- we 19 know that there may -- that the universe may not be 100 20 percent, but can we just assume that for most, that 0612 21 applies, and if 0612 applies, that that could be credited in 22 the NRC analysis? Brian, why don't you go ahead and then 23 we'll go over to Bill Henries. 24 MR. THOMAS: I like what I hear in terms of the 25 commitment, but what I want to remind you of is that we've ANN RILEY & ASSOCIATES, LTD. ['N_)h Court Reporters 1025 Cornecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
346 1 got Generic Letter 85-11 sitting out there, Generic Letter
/T 2 85-11 dismissed the need for licensees to implement phase V
3 two of 0612. So when you say there are a lot of licensees 4 out there that are fully implementing 0612, that's where we 5 have our concern. That's where we have'some uncertainty. 6 There are a few licensees that have come in, as I 7 said, in decommissioning space, the ones that I've seen, I 8 was surprised that they went to the full extent to do -- to 9 put in the cranes, to do the consequence analysis and so 10 forth. 11 MR. CAMERON: Maybe go back to Stu's point, is 12 that we're going to get into an argument about how many 13 plants fully implement 0612. I think there is another 14 question here, which is if they do implement 0612, then what () 15 types of credit can we give them in the analysis. 16 MR. MEISNER: I don't remember exactly the details 17 cf phase twc, but it revolves around the consequence 1 18 analysis. l 19 MR. THOMAS: It's an option to do a consequence 20 analysis and take mitigating measures to manage the 21 consequences or utilize the single f ailure-proof crane, or 22 use the crane and interlocks and so forth to avoid hazards. 23 MR. MEISNER: I think the general industry 24 understanding of this is that it's an either/or, consequence 25 analysis or a single failure-proof crane. And I think you O ANN RILEY & ASSOCIATES, LTD. Court Reporters
.1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 y
347 1 can count on that as far as the industry implementation. 2
) If what you're talking about, though, is you 3 expect the industry to do both, tnen that's a new staff 4 requirement I think you have to justify. But having done 5 the consequence analysis, you all can feel comfortable that, 6 in fact, this whole issue goes away for a licensee, because 7 you can't breach the spent fuel pool.
8 MR. CAMERON: Before you answer that, Brian, let's 9 go out to the audience, the gentleman out here has something 10 to add on this. 11 MR. LAGGART: Mike Laggart, from GPU Nuclear. 12 Brian, I know just what you're talking about, the 85-11, and 13 that it closed down really the phase two portion from 14 further NRC review. () 15 I think there's a couple of other things, 16 especially for decommissioning that you have to consider. i l 17 One is that when we go into decommissioning, the FSAR or the 18 tech specs are revised. So if you come upon a plant that's 19 not in that category of that doesn't have it in their DSAR, 20 I'll call it, you can attack that issue as it comes up, I 21 think, and say, hey, 0612 is what you require. l 22 But a second point is that issue on 85-11 applies l 23 to all operating plants. If it's an open issue with the l 24 staff right now for decommissioning plants who have a load j 25 lift, it should really be an open issue for all operating , i l l i
'~
ANN RILEY & ASSOCIATES, LTD. I Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l L
348 1 plants, as well. () 2 3 If we have plants out there that don't have that consequence analysis or the safety single failure-proof 4 crane, so to apply it to just decommissioning I think is 5 maybe inappropriate. 6 The third item I think is that 0612 was generated 7 more for not only for fuel pools, but for safety-related 8 equipment. And most utilities that don't have single 9 failure-proof cranes have very site-specific technical 10 specifications for modifications to their fuel pool, such as 11 we have a cask drop protection system for casks up to 100 12 tone. 13 So most of the 0612 stuff that you are concerned 14 about or single failure-proof crane or the consequence () 15 analysis was really for all the rest of the safety-related 16 equipment in the plant. That really goes away. We're onlp 17 concerned about the fuel pool and I think they're covered by 18 site-specific tech specs in the mainstream. 19 MR. CAMERON: Thank you for that clarification. 20 Brian, before we go to Bill, do you have anything further 21 you want to say on this? 22 MR. THOMAS: We could spend -- 23 MR. CAMERON: You don't have to say anything. 24 MR. THOMAS: Let me just say that my key point 25 here is that 85-11 exonerates the plants from doing the
': ANN RILEY & ASSOCIATES, LTD.
s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
1
)
I 349 1 consequence analysis or installing single failure-proof
-( ) 2 cranes. What I'm saying is 85-11 says you don't have to do 3 any of that.
1 4 MR. MEISNER: One or the other, 5 MR. CAMERON: Okay. Well, we need to -- { 6 MR. THOMAS: So we need to have some dialogue on i 7 it. 8 MR. CAMERON: The NRC, I think, needs to get clear 9 on what sur regulatory structure is on that. Stu, do you 10 want to offer anything on that? Go ahead. j 11 MR. RICHARDS: I think we ought to just go with 12 what's on the record. Mr. Meisner said they are going to do 13 the single fail-proof crane and the first part of 0612, 14 they're going to do the.second part. Forget about 85-11.
, 15 It didn't say anything about that.
16 MR. CAMERON: 17 MR. CAMERON: Okay. And, Bill, why don't you go 18 ahead. 19 MR. HENRIES: I just wanted to once again harp or 20 reiterate about the use of continual upper bounds instead of 21 best astimate. I'm sure you can handle this as part of what 22 credit you give for implementing 0612, but I think you can't i l 23! continually'take the upper number on each branch of the i 24 sequence and call that a best estimate. 25 MR. CAMERON: All right. Ray Shadis. i
.[ ) .) Uni RILEY & ASSOCIATES, LTD. . \/ Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l.
350 1 MR. SHADIS: Thank you. I just -- Stu Richards, ( 2 I'm not certain what NRC is offering to do in exchange for 3 this commitment by the industry. 4 MR. CAMERON: Good point. It would be nice to 5 find out. 6 MR. RICHARDS: I'm not offering to do anything. I 7 mean, the industry just said let's make an assumption, for 8 simplicity sake in doing this analysis, that the 9 requirements or the guidance contained in 0612 are going to 10 apply to a plant in decommissioning. Make that assumption i 11 of the analysis. 12 What does the analysis then say? We're going to 13 get a report from the technical working group and it's going 14 to say based on these assumptions, here's the risk. When it () 15 comes to me, then we're going to crank that into some kind 16 of a regulatory framework. 17 So rather than argue about a lot of different l 18 variables that can apply, why not just use that simplifying j 19 assumption as a going in point for the analysis and see what 20 comes out the other end. 21 MR. CAMERON: Let's go to Rich Barrett. The step 2:2 is, are -- is the NRC staff going to factor in 0612 in to 23 redo the analysis. 24 MR. SHADIS: What is the drive here to get only 25 one number for each instance or each consideration here? C' ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 J L
351 1 Why are there not two or three numbers? Why not a high () 2 3 number, a low number, a best guesstimate number? What is the push for one number overall to say this is the risk? 4 MR. RICHARDS: I think the response to that is why 5 reinvent the wheel. There's been a lot of work done in the l 6 past and I think it accumulated largely in this NUREG-0612, 7 which is focused on exactly lifting heavy loads. 8 So as long as we're not going to argue about i 9 whether that applies or not for decommissioning, let's use 10 that as a going in assumption, build on the information 11 that's al'reidy available, and move forward. Why start from 12 scratch? l 13 MR. CAMERON: All right. You're thinking about 14 that. () 15 MR. SHADIS: I just did take a minute to think j 1 16 about that. I appreciate that sentiment. For this 17 particular instance that you're going to assume that these 18 things are applied, however, assuming that those conditions 19- are met, in itself, ignores the fact that this wheel was 20 invented before. 21 It ignores the fact that all of these other 22 assumptions were taken into consideration before. 23 MR. RICHARDS: I would say just the opposite. 24 We're not ignoring it. We're recognizing it and using it. 25 MR. CAMERON: All right. Let's come back to that ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 '
3E' 1 dichotomy perhaps. I would like Rich Barrett to tell me ( )- 2 this is true. We assume 0612 applies, NRC relooks at 3 analysis on this basis. Is that what we're going to do? 4 MR. BARRETT: We can certainly look at the 5 analysis on the basis of the assumption that 0612 applies. 6 What I've been told here is that we have Generic Letter 7 85-11, which says, if I understand it correctly, not that we 8 don't expect the licensees to do 0612, but that based on our 9 perception of what the industry has done on its own, we're 10 not requiring -- we're not making that an NRC requirement. 11 Is that what it's all about? I 12 MR. THOMAS: Yes. If I can amplify what you said, 13 it says specifically you do not have to implement the phase i 14 two portions of 0612 and those phase two portions address l () 15 the use of single failure-proof cranes, doing consequence 16 analyses, and using some other compensatory measures. 17 MR. CAMERON: Okay. We're going to go on here to 18 someone who might have some more information on the 85-12 19 issue. Yes, sir. 20- MR. VAN NOORDENNEN: Thank you. Gerry Van 21 Noordennen, from Connecticut Yankee. The one regulation I 22 haven't heard mentioned here, which would trigger an I 23 analysis or safe load path or upgrade of the crane for 24 cperating plants, in addition to decommissioned plants, 25 would be 50.59. l
.h \ms/
ANN RILEY & ASSOCIATES, LTD. Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 ' Washington, D.C. 20036 (202) 842-0034 L
353 1 So that you are moving a heavy load over () 2 safety-related equipment, before somebody does that, say, 3 for example, moving a steam generator over a reactor vessel, 4 you would do a 50.59 to make -- to look at the consequences. 5 So that has to be kept in mind, also. 6 SPEAKER: You're supposed to take things off the 7 table, not put more regulations on, for God's sake. 8 I guess that the -- we keep coming back to this 9 implications of 85-12 here and I don't know that we've 10 gotten a certain answer about what 85-12 does in terms of 11 the consequence analysis, et cetera, et cetera. 12 Ignoring for the moment 85-12 -- 11, excuse me. 13 Let's ignore 12, too. Right. Is the staff going to relook at the analysis based on application of 0612? (f 15 MR. BARRETT: Absolutely, yes. Yes. The staff 16 will take a look or re-take a look at the analysis based on 17 what you see in 0612. Now, if we -- if that's the only 18 analysis we do, if the only analysis that we have is one in 19 which a licensee is fully in compliance with 0612, then 20 that's the -- that's going to be the basis for going forward 21 with any exemptions or rulemakings. 22 And any licensee who comes in who is not in 23 conformance with 0612, phase one, phase two, or however many. 24 phases there might be, then that licensee is going to have 25 to be treated as an exception. O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
354 1- So what we have apparently here is an opportunity () 2 3 to simplify our analysis. That is to say, to eliminate some of the' variability by making the simplifying assumption that 4 all licensees meet 0612. 5 That has advantages in terms of simplifying our analysis, 6 but it has a potential disadvantage for any licensee down 7 the line who doesn't meet that 0612. And if the industry is 8 happy with the situation like that, if the industry is 9 confident that that assumption will encompass all licensees, 10 then that would certainly be a simplifying assumption that 11 ve-could make in our analysis. 12 MR. CAMERON: Let's go to Gary and then we're l 13 going to go to Mike, and then we'll go out to Brian in the 14 audience. () 15 MR. HOLAHAN: I'm not sure this is going to move 16 us forward. It may take us back a step. 17 MR. CAMERON: Let's go to Mike. 18 MR. HOLAHAN: It seems to me there is some value l l 19 in simplifying the analysis and doing it only once and all 20 of that. However, if that analysis doesn't apply to all the 21 plants, you're just leaving yourself with leftover issues l 22 that you haven't decided how to deal with. 23 I kind of like when the analysis was done more 24 than cne way, and this is a sort of factual thing. If you 25 are this kind of plant, you get this answer. If you are O IdRJ RILEY &- ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1
l' 355 1 that kind of plant, you get that answer. () 2 3 Then I think -- and then whether you're conforming with 0612 or not is simply a question of facts. Otherwise, 4 we have the possibility of getting -- mixing the apple and 5 the orange. You're not a plant like this, but you're 6 applying analysis that doesn't apply, and it seems to me 7 it's a little more work to do the analysis both ways, but 8 -ultimately I think it simplifies the understanding where the 9 plant stands. 10 MR. MEISNER: We're making this awful, awful 11 complicated. Rich, as far as 0612, let me be explicit. 12 It's either a single failure-proof crane or it's a 13 consequence analysis that makes your issue go away anyway. 14 And in addition, it's phase one of 0612 that includes the () 15 things like safe load paths, the crane operator training 16 program, and the criteria, there are like five or six items 17 in there. 18 So is that real clear? I'm not talking about 19 pulling in new NRC requirements that you'd like to get in 20 here. It's a single failure-proof crane or the consequence 21 analysis, this phase one, which includes those five to seven 22 items, such as safe load paths. 23 MR. BARRETT: That's my understanding. 24 MR. MEISNER: Okay. As far as having multiple 25 analyses, that's fine, Gary, as long as you give us credit [~ ANN RILEY & ASSOCIATES, LTD. I (- Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034 l I L i
356 1 for 0612. [v 2 MR. HOLAHAN: We should give each plant credit for 3 what they've done. 4 MR. MEISNER: And -- well, you don'? have to do a
)
5 plant-specific analysis. l 1 6 Etery plant, at least the modern ones that are 1 7 going into decommissioning, you don't care about the old 8 ones because they're past your zirc fire concern anyway. 9 MR. HOLAHAN: We said every plant did phase one. 10 MR. MEISNER: Has, within their license basis, 11 what I just described. That's within their license basis. 12 You can argue over the niceties of exactly what the NRC 13 requires or not, but I tried to be explicit. If you take 14 those assumptions and fold it into the analysis, it should .c ( ,3
) 15 make a big difference.
16 MR. HOLAHAN: That was not my understanding of 17 what every plant had. However, li that is what every plant 18 has, that's the way you should do the analysis. 19 MR. CAMERON: Okay. You might afterwards see if 20 this reflects this discussion that's up here. Rich Barrett. 21 MR. BARRETT: What I'm concerned about in this 22 workshop is making sure that we have a complete discovery of 23 all of the factors that are important to driving these 24 analyses and we've had a lot of discussion about NUREG-0612 25 and then I understand that there is a big difference between (', ANN RILEY & ASSOCIATES, LTD. (/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
e 357 1 an 0612 plant and a non-0612 plant or, let's say, a pre-0612 (m) 2 plant in terms of the risk associated here.
-3 And I think with regard to 0612, I think we have 4 enough understanding right now to go back and make a 5 reasonable decision on which way to do the analysis.
6 My bigger concern is do we have enough information 7 from the licensees about all the other factors that are 8 important to this analysis. For instance, Bill made some 9 very interesting statements here regarding the frequency of 10 events being different from what we assumed. Also, that 11 regarding the timing of the handling of large heavy loads, 12 that, in fact, most, if not all of this cask handling would 13 happen after we've passed the window for zirconium fires. 14- So it seems to me that there are other things that have been I 15 raised here and rather than trying to negotiate how we're 16 going to do the analysis, I want to make sure that before we 17 end this session, that people on our staff have heard and 18 understand everything that the licensees can bring to the 19 table that would be the basis for that analysis. 20 MR. CAMERON: We're going to go to someone from 21 the community and then Glen Kelly wants to, I think, just 22 make sure that everybody understands what the analysis was 23 based on, Brian. l 24 MR. FORD: Brian Ford, Millstone. Maybe I just 25 got lost in the discussions. The way I understood the world I I I ( )' (- ' ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 1 (202) 842-0034
I I I J 358 1 was 0612 came out many years ago to address an issue. There ! ( 2 was, I think, a generic letter that said thou must respond 3 within such amount of time, you have to do it on two parts, l l 4 all the utilities went out and did the first part, sent in 5 responses, that kind of thing. l l There is another part that the NRC decided they 6 I 7 did not need to respond on or to do additional level. There 8 was a reason for that decision, basically that the practices 9 and the activities that had already been done adequately 10 mitigated the risks associated with the condition. 11 Sometime -- and that was put out in a generic 12 letter, this 85-11 or 12, whichever number it is. 13 MR. CAMERON: 85-11. 14 MR. FORD: So you have a generic letter. I heard
) 15 someone mention that there was a study or something done 16 after that that identified or in that time that identified 17 that there was about a factor of three reduction in risk 18 associated with doing those activities that had been done 19 for 0612 and what I heard other people asking is why did the )
1 20 credit for those three levels of reduction go away and if 21 they had went away, then are you really calling into 22 question the NCR's decision in 85 to stop the activity. 23 And this came up from someone else. Is this 24 really a decommissioning issue or is this really an issue 25 that the staff is looking bnck at a previous decision for i ANN RILEY & ASSOCIATES, LTD.
\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
359 I 1 operating, shutdown, every plant, and saying we think we () 2 3 have a safety concern potentially that we had not previously recognized. 4 And maybe I just got lost in the discussion. 5 MR. CAMERON: I think Bill Henries did talk about 6 two orders of magnitude in his presentation. But I think 7 that this would be useful for Glen to talk about what is -- 8 what is in the analysis. 9 MR. KELLY: When we performed the analysis, the -- 10 we assumed that the plants that we analyzed were 0612 11 plants. And what I mean by that is that they were basically Jr -- they were following safe load paths. We assumed that 13 they weren't taking the load someplace where they weren't 14 supposed to 90. () 15 What we did not do in our analysin was to give the 16 same level of credit for having a single fai;m e-proof crane 17 that 0612 gave. And we did that, in part, because we felt 18 that the more recent information that we had did not sh;w 19 any reduction in failure rates and, therefore, we didn't see 20 that as much credit as had been given in 0612 was correct. 21 Now, it's probably appropriate for us to go back 22 and look at that in more detail and determine what is the 23 correct amount of credit that one should give for having a 24 single failure-proof crane. This is an issue that is not 25 only for decommissioning. l [~ ANN RILEY & ASSOCIATES, LTD.
\ms Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
360 , I 1 At this time, the Office of Research is looking
} 2 'into whether_or not this is an issue that deserves to be 3 raised again.
4 So the potential is that it may be something of 5 concern for operating reactors. That has not been 1 6 determined yet. But we're still in the process of looking i 7 at it. l 8 I think that, again, as with what we did 9 yesterday, the -- we're not married to the numbers that we 10 have. The question was, what we hoped to do was to point 11 out that our assessment shows that for us, heavy loads are 12 potentially of a concern. And we're interested in 13 understanding what is the practice in the industry or what 14 wi21 industry be doing to help make this concern go away. () 15 And I think we've gotten a good bit of information 16 there and there may be more than industry can tell us that 17 they're doing that we may not be aware of. I l 18 MR. CAMERON: Rich, is this an appropriate point l 19 for you to discuss the issue that you wanted to raise? 20 MR. BARRETT: I don't think it helps. I don't 21 think what I would say would help at this point. 22 MR. CAMERON: Okay. Well, keep thinking about 23 that and let's go to Bill Henries. 24 MR. HENRIES: Thank you. Two things I wanted to 25 touch on, The newer data, so-called, didn't show any ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
)
l 361 1 changes from the 0612. Well, you wouldn't-expect it to. It () 2 wasn't new data based on nuclear power plants. It was new 3 data based on DOE and barges and people out in the back 40 4 with.five-ton come-alongs. The newer data is the same as 5 the older data. 6 You've got to take credit not just for the single
'/ failure-proof crane, I think if you look at the error 8 statistics, it's the other things in 0612, like the operator 9 training, safe load paths, the procedural controls, the 10 management oversight, all of the things that make the human 11 errors go away are the reasons we previously got three 12 orders of magnitude.
13 Now, I wouldn't argue over one to two as long as 14 you didn't use the upper bound on every other number on the () 15 sequence. We still get it back into the 16 ten-to-the-minus-seven realm, which was the number that had 17 been coming out of all the previous studies. l 18 And I think that's all we're looking for. 19 MR. CAMERON: Ray. I 20 MR. SHADIS: I'd like to comment that I was glad l 21 to hear Glen's summation, because I think that what you're 22 being asked to do here is to provide the technical 23 underpinning for a regulatory change, and that's important 24 enough to really assess the' basic data as well as the recent 25 reports. O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l 362 1 A technical evaluation has got to go somewhere f~'i d 2 beyond a book report and so I'm pleased to see that the l 3 staff is still trying to look at what some of these basic l 4 assumptions are. 5 And I hope that they continue on that path. l l 6 That's it. 7 MR. CAMERON: Thank you, Ray. It's about time for 8 lunch here. I don't know what more we can say, other than 9 the staff is going to take the information that they heard 10 from the industry and go back and reevaluate the analysis. 11 Is that at least a true statement, Gary? 12 MR. HOLAHAN: Sure, that's true. Yes. It's the 13 either/or that I'm not sure about under your flipchart. 14 MR. CAMERON: I think that the either/or is we l
, ~ ~
15
!( ) need to capture all categories, do it for all categories of 16 plant, the 0612 plants and non-0612, and whatever the other 17 category is.
I 18 MR. HOLAHAN: I think that's the wrong 19 categorization, to call them non-0612. Everybody did 0612. 20 The question is, what did they do for 0612. Our original 21 study looked at single failure-proof versus not single 22 failure-proof cranes and I think I hear a suggestion that l 23 that's not the real issue, that they're all very good, 24 whether they're single failure-proof or not, and they can be l l 25 put into one category. (~~' ANN RILEY & ASSOCIATES, LTD. (_)s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
i 363 l l l 1 MR. HENRIES: I think the results of 0612 is
~') 2
[V either you're single failure-proof or you've done a 3 consequence analysis, so it's a moot point. 4 MR. HOLAHAN- What's not clear to me is whether 5 that consequence analysis is relevant to this current issue. 6 That consequence analysis had to do with things like load j 7 paths, didn't it? l 8 MR. HENRIES: No. l 9 MR. HOLAHAN: And avoiding, well, I thought -- l 10 MR. CAMERON: Let's hear about the con equence 11 analysis. j l 12 MR. HOLAHAN: The consequence analysis wasn't 13 spent fuel pool. It was operating reactor safety systems 14 and stuff like that, wasn't it? l () 15 MR. MEISNER: Absolutely not. It was specific to 16 the crane involved and the safety significance of what it 17 could do. If we put in a new crane under our 0612 18 commitments over the spent fuel pool, then we have to do 19 either a consequence analysis specific to that crane's use 20 or make it single failure-proof. 21 It's spot-on to exactly what the issue here is, 22 and the key -- there would be two key elements of the 23 consequence analysis. One has to do with impacting fuel, of 24 course. The other would have to do with breaching the pool. 25 And if we showed that the pool wasn't breached, then whether Fj ANN RILEY & ASSOCIATES, LTD. k_/ Court Reporters , 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
( 364 l 1 it's -- you know, there is no probability for this issue. l 2 [u )\ So the only real thing to analyze is a single 3 failure-proof crane. 4 MR. CAMERON: I'm going to ask -- we're going to 5 take one last comment from Ray before we break for lunch, 6 but I'm going to ask the NRC staff to consider thinking l 7 about this over lunch and seeing if there is anything l 8 anymore light that we can put up on the chart in terms of l 9 what exactly is going to be done here. Ray, do you want to 10 take us into lunch? 11 MR. SHADIS: It's a question and then maybe Rich i 12 could answer it and then he could have his shot. 13 MR. CAMERON: And Gary is going to answer that. 14 So okay. l() 15 MR. SHADIS: It's just this. It's just this. You 16 have -- it seems that the backup position here is that the 17 licensee is going to do an analysis specific to their crane 18 and whatever and that that is somehow on one side of an l 19 equation. On the other side of it is further analysis by 20 the staff for the purposes of ultimately changing the 21 regulations. 22 And I wonder if the importance of changing the l 23 regulation and the fact that it affects the, whatever they 24 are, 106 licensees or whatever, if that can really be 25 balanced against this notion that each licensee, if they ANN RILEY & ASSOCIATES, LTD.
\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
365 1 have to, would individually assess the probabilities and 2 consequences at their own plant. 3 " Does that question make sense to you? 4 MR. CAMERON: Rich?
-5 MR. BARRETT: I'm not sure, Ray. Could you just 1 6 run it by again?
7 MR. SHADIS: What I see happening here is that it j 8 is proposed to put away all further analysis on 9 probabilities because they are bounded somehow by conformity l 10 to 0612 and that in conforming to 0612, there are two ways 11- you can go, one is to put in your single failure-proof 12 crane, the other is to do an evaluation or analysis of not 13 using it, j 14 And it seems to me that what happens is that 15 evaluation of not using the single failure-proof crane l 16 somehow gets balanced against the staff going any further tc 17 look at the possibilities. l 18 MR. BARRETT: I think your question goes toward i i 19 the comment I wanted to make, and that is that in a sense, 20 we live in two worlds in this nuclear safety business. We I 21 have a deterministic world in which we develop rules and 22 requirements, NUREGs, generic letters, 0612 being one of 23 them, and when we go through that analysis for a particular i 24 issue, we come to the bottom line that that issue is 25 resolved. l O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l 1
r 366 1 Okay. It's a black and white determination. So ts 2 you're either in conformance with that rule or you're not or (v) 3 that requirement or that regulation and from a regulatory 4 perspective, we walk away from it. Of course, we inspect it 5 periodically, but you're in compliance and you're fine.
- 6. That's the way we do deterministic rule-based
)
l 7 regulation. The business that we're in right now is risk j 8 assessment and risk assessment doesn't make that assumption I 9 that because you're in compliance with a given rule or j 10 because you've met the requirements of a given regulatory 11 guide, that we can walk away from an issue in risk space. 12 I think everyone knows that we have an ECCS rule j 13 and we've had an ECCS rule for a long time, but if you look l 14 at risk assessment, you'll always find large LOCAs, small () 15 LOCAs, medium LOCAs, and all sorts of LOCAs, as important 16 contributors to the risk for many plants. 17 What we're looking at is the residual risk of the 18 plant. What risk is left after you've met all of the 19 standards and requirements? So in this case, we're not 20 talking about saying that for a plant that meets 0612, we're 21 going to walk away and say the risk is zero. What we're 22 talking about here is in the analysis of risk, do we want to 23 do the analysis under the assumption that a plant meets 0612 24 or do we want to do the analysis two ways, one for a plant 25 that meets 0612 and one for a plant that doesn't. 1 l [~' ANN RILEY & ASSOCIATES, LTD.
\m- Court Reporters I 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
367 1 And I don't know if that answers your concern or (} 2 not. But in any event, when you take all of that, heavy 3 loads, and add that along with what we're left with after 4 we've reanalyzed seismic and intermediate and long-term 5 sources, then that becomes the risk basis for what we do in 6 the future on whether it's rulemaking in the future or 7 exemptions in the nearer term. 8 MR. SHADIS: What it seems like, it seems like 9 that what you all are saying to the industry folks is we 10 don't need to analyze further in terms of risk or 11 consequences because if you're going to be applying, 12 implementing 0612, you're either going to be using this 13 crane or you're going to be providing your own consequence 14 analysis. () 15 And therefore, the fact that they're going to be 16 provit-ing their own consequence analysis substitutes for an 17 overall consequence analysis. 18 MR. BARRETT: I don't think that's what we're 19 saying. I think we're still going to come down to a final 20 analysis of what is the risk associated with a zirconium 21 fire that might be initiated by a drop of a heavy cask. But 22 we have to put down numbers as to what's the likelihood of 23 that, and those numbers have to be based on analysis and j 24 assumptions. And when you do that analysis, you're either 1 25 going to assume that they've got a single failure proof I ANN RILEY & ASSOCIATES, LTD. l f~') ss Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 e
368 1 crane or a consequence analysis in conformance with 0612, or [Vl 2 you're going to assume that they don't. 3 And if you assume that they don't, they're going 4 to get somewhat higher numbers, and if you assume that they 5 do, you're going to get somewhat lower numbers. It's just 6 all part of the analysis and it's the way we do risk 7 analysis for any issue. I 8 So it could very well be that when we take into 1 9 account all of the information we got today about the i 10 current status of plants, in addition to the fact that we -- l 11 that everyone is in conformance with 0612, as has been 12 asserted here, it could very well be that this particular 13 contributor does fall out as not being important. 14 But if it does, it will fall out because the l 3 15
/ numerical analysis shows that it falls out. Not because we 16 have made an assumption that because -- that compliance with 17 0612 means that there is zero risk.
18 MR. CAMERON: Gary, do you want to -- let's go to 19 Gary, and we'll finish up with this gentleman in the 20 audience. Then we'll take a break till 1:30. 21 MR. HOLAHAN: Let me see if I can at least clarify 22 my concern with the issue, and it goes exactly to Bill's 23 second viewgraph, where it states that only single failure 24 proof crane load drops need be considered since consequence 25 analysis must be performed prior to using non-single 1 ANN RILEY & ASSOCIATES, LTD. [%dx] Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1
369 1 failure-proof cranes. () 2 3 It's that equation that I'm having a problem with. If you do a consequence -- if you don't have -- if 4 you have a single failure-proof crane, then this is all 5 taken care of very easily. We know how to do the analysis 6 and we know where it should come out'. l 7 If you don't have a single failure-proof crane, do l 8 you mean to tell me that you will not carry over the spent 9 fuel pool a load that is so heavy that the consequence of it 10 would fail the pool? 11 MR. HENRIES: Yes, that's right. 12 MR. HOLAHAN: But does that mean that licensees 13 who are doing decommissioning and are handling heavy casks, 14 unless there is a way to get fuel into casks that I haven't () 15 thought about, who are putting heavy casks in the pool, have 16 necessarily gone to single failure-proof cranes or is there 17 something else like crash pads or something going on that I 18 haven't -- that I don't see here? 19 MR. CAMERON: Bill, do you want to clarify that? 20 MR. HENRIES: Exactly as you've said, you have two 21 choices. You have a single failure-proof crane, in which l 22 case we discussed it, or you've done a consequence analysis. 23 Consequence analysis can be that dropping 100 ton cask -- 24 MR. HOLAHAN: Doing the analysis accomplishes 25 nothing. That's what I'm trying to get past. ( ANN RILEY & ASSOCIATES, LTD.
\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1
L
370 1 MR. HENRIES: It does tell you if you pass / fail. [J\ 2 The consequence could be totally acceptable. 3 MR. HOLAHAN: Then you'd never get the fuel out of 4 the pool. 5 MR. HENRIES: No. The consequence -- you're 6 assuming you dropped a cask. The cask drops 40 feet, goes 7 through air, goes through water, impacts the bottom of the 8 pool. If there is no structural damage, no leakage, no 9 off-site dose consequence, no criticality concerns, no zirc 10 fire concerns, the consequence is acceptable. 11 MR. HOLAHAN: I understand that. 12 MR. HENRIES: That's the only way they could bring 13 it up. 14 MR. HOLAHAN: And if we understand that () 15 consequence analysis means either the loads are going to be 16 small and you have to use small casks or crush pads or 17 something that makes this go away, then, in fact, it seems 18 to me that -- 19 MR. HENRIES: We have an analysis at Maine and 20 with a 100-ton cask dropping 40 feet, it goes into the 21 concrete about a foot and a half. We assume it becomes 22 infinitely permeable and then what happens? It starts to 23 try to leak into bedrock. We estimate two to five gpm leak. 24 It's a loss of cooling, loss of inventory. 25 MR. CAMERON: Okay. I hope that that was -- I O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0C34
l
\
371 l 1 think that that was useful, [~) \m ' 2 MR. HOLAHAN: It helped me. I don't know whether t 3 I it helped anyone else. 4 MR. CAMERON: Final comment here, and then we're 5 going to lunch. Go ahead, sir. 6 MR. LAGGART: I heard we're going to go with two 7 types of analyses, Rich, and I would say to make your job 8 easier, and I think Mike has stated, we all believe we're in 9 compliance with both phase one and phase two. Now, the 10 question is, on phase two, whether it involves all licensees 11 or not, because of that Generic Letter 85-11. 12 If you're looking for a commitment for 13 decommissioned plants, I think we can say, and we know 14 everybody meets phase one, we'll meet phase two, as well, () 15 for all the decommissioned plants, because that is our
.6 practice.
17 I think that Mike -- of course, that's subject to l 18 Mike's agreement, Meisner, to make that part a commitment. 19 If there is some uneasiness and the staff says 20 that we're all committed to that, that way it simplifies the 21 analysis, everybody can assume phase one, phase two 22 implementation, which the phase two part is the consequence 23 or -- that would be my proposal. 24 I think I'd ask Mike to comment on that. 25 MR. CAMERON: I think what we'll do is why don't ANN RILEY & ASSOCIATES, LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (.2 02 ) 842-0034
l 372 1 you guys talk about that. NRC, based on Gary's
~2 understanding from this last colloquy with Bill, if there is 3 anything that we can put up on the flipchart when we come 4 back that provides some clarity to the issue that Mike said 5 should be simple to resolve, we can put that on when we come 6 back and then go into thermal hydraulics.
7 If we can't, we can't. But be back at 1:30. l 8 Thank you.. 9 [Whereupon, at 12:20 p.m., the workshop was 10 recessed, to reconvene at 1:30 p.m., this same day.] 11 12 l 13 l 14 () 15 16 17 18 19 20 21 22 23 24 . 25 l O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
373 1 AFT.ERNOON SESSION ( 2 (1:30 p.m.] 3 MR. CAMERON: We're going to get started with the l l 1 4 last push here on these issues and to sort of give you a 5 preview of what we plan to do, I want to quickly close off, 6 finish up the heavy loads segment. We're then going to go l 7 into thermal hydraulics, and then we're going to go into a 8 wrap-up. Part of the wrap-up, a very important part of it, 9 is to see if we can get some agreement on this 10 requantification proposal that is simmering, and probably l 11 more than simmering now because we do have a handout for you 12 that refines what I have up here and that we'll put on the j 13 screen for you as a viewgraph There are some other 14 lingering issues, too, that we want to discuss in that 15 segment. We're going to lead off that segment with some 16 words from Mike Meisner, and then we're going to go into a 17 discussion of this. l 18 In terms of heavy loads, I'm going to read you a 19 statement that we're going to put, we're going to go with as 20 a commitment from that session. It starts off with assume l 21 0612 applied, and NRC should give appropriate credit for 22 this, that is, 0612, in the analysis. That's where we're 23 going to go with that on heavy loads. 24 Okay, since there aren't any questions on that or 25 comments, are we ready for thermal hydraulics, and John, are O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
374 1 you going to set things up? 2 MR. HANNON: Yes. 3 MR. CAMERON: All right, thank you. 4 MR. HANNON: My name's John Hannon. I'm branch 5 chief for NRR. It's the last session of this workshop. We 6 saved the best for last, the thermal hydraulics part. Ed 7 Burns gave me a good lead-in yesterday. I think it was his 8 slide number nine where he talked about the importance of 9 making a good connection between deterministic part of the 10 calculations and the probablistic analysis. That's part of 11 the reason we're having this session today, to make sure we 12 have a good, sound technical basis in the deterministic 13 area. 14 These calculations that we're doing have in the () 15 past provided us a definition of what we're calling critical 16 decay time, and I interpret that to be the time period for 17 which the nuclear safety hazard persists such that 18 regulatory constraints are appropriate to protect the health 19 and safety of the public. 20 Now, in a minute, Diane Jackson and Joe 21 Staudenmeier are going to go into details of what the staff 22 has done and try to identify some of the conservatisms and 23 the parameters and uncertainties of importance and discuss 24 the calculations that we have performed to date and more 25 closely try to approximate the conditions that exist. These O ss ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
j 375 1 calculations rely on computer models that require initial 2 conditions to be set and assumptions to be input. [Vl 3 Another slide that Ed used yesterday, slide number 4 27 in his package, raised a question on the overly 1 5 conservatism aspects of certain of these assumptions that 6 we're using. The one that I recall had to do with the onset 7 of the oxidation runaway. There are some discrepancies 8 there between what the literature has and what we're using I 9 for our analysis. So, maybe we'll have to -- we'll want to 10 go into some discussion on that, too. 11 At any rate, these calculations that we did 12 attempted to be near bounding such that we predict the 13 nuclear time line of concern that would accomplish most of 14 the plants for exemption purposes. The goals of this () 15 session is to have everybody understand what we did, l 16 identify the parameters that are important to obtaining more 17 accurate time estimates on a plant specific basis so that 18 without compromising our safety margins, we can have good 19 technical basis for granting exemptions and in going forward 20 in the rulemaking process. We also want to explore other 21 techniques and methodologies that you all may want to bring 22 to the table. 23 I just would close by saying this is a very 24 technically complicated subject, and we do have a diverse 25 audience, so we need to be careful not to leave anybody y ANN RILEY & ASSOCIATES, LTD. 7 Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I l l 376 1 behind, so I want to be sure we're clear in our ( '1 2 communication, and we can try to reach a line in our N_ 3 thinking. Let me turn it over to Diane for further 4 discussion. 5 MS. JACKSON: Good afternoon. I'm Diane Jackson. 6 I'm a member of the technical working group, and I'm in the 7 plant systems branch with the' staff. For our deterministic 8 calculations, we did two types of calculations, the 9 thermohydraulic analyses and 80 adiabatic heatup 10 calculation. I'm going to talk about the scenario and the 11 existing generic calculations used in the thermal hydraulic 12 analyses, and Joe discussed more of the details of the 13 different types of calculations and analyses that we did in 14 the report. l' Q 15 For the thermal hydraulic analyses, you want to ( ,/ t i 16 analyze the temperature of the spent fuel when it's exposed 17 and only cooled by air. Depending on the code you use, you l 18 can model the whole pool, your specific burn-ups, your decay 19 heat, your configuration that your fuel is arranged in in a 20 pool, and even your pool building, to give you a very good I 21 answer on how high your fuel can heat up. 22 The analyses usually result in calculating a 23 critical decay time, which is the minimum length of time 24 since the reactor is shut down that you have decay heat in 25 your pool or in your most recently discharged fuel that ANN RILEY & ASSOCIATES, LTD. w/
) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l I l 377 1 there isn't sufficient heat to raise the temperature of the ID; q 2 fuel to the point that you would get runaway zirconium v 3 oxidation. 4 In general, the scenario we looked at is do you 5 have a loss of water? At that point, your fuel will start 6 increasing in temperature. To move on to the next slide, 7 early in this sequence, your increase in temperature is due 8 to just the decay heat of your spent fuel. As you start 9 reaching higher temperatures, you have an oxidation reaction i 10 between your air and your zirconium, which is an exothermic l 11 reaction. So. you produce energy as the temperature 12 increases. The heat input from the oxidation reaction I 13 dominates what's driving your temperature increase. Then it 14 can increase in temperature very quickly. ) () 15 Due to the higher temperatures that can be 16 reached, you can heat your fuel to the point that you can 17 have fission products released from the fuel, the solid fuel 18 itself, and at some point you could have a zirconium fire, 19 and this would provide an energy source that could transport 20 fission products offsite, and this is why it's a concern for 21 areas such as emergency preparedness. 22 One of the things that was brought up early on in 23 our technical working group study is why don't we use the 24 existing information. All the information's there. NUREG 25 1353 looked at this, so we went back and looked at the 4 ANN RILEY & ASSOCIATES, LTD. 0s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I i
378 1 applicability of the older thermohydraulic studies to the () 2 3 plants that we've got today or ones we'd see in the future. In the 80's, like I said, generic safety issue 82, 4 Sandia and Brookhaven National Laboratories looked at the l 5 --I wrote probability. It wasn't really probability. It's 6 more like a likelihood. It wasn't a risk analysis like we 7 were talking about earlier today. The phenomena and 8 consequences of zirconium oxidation and a zirconium fire, l 9 and this was based only on operating reactors. 10 We looked at another study that was done by 11 Brookhaven National Lab that was done in 1997 that was done 12 for generic BWR and PWR for decommissioning plants, and it 13 included the potential for a zirconium fire. 14 In looking at these reports, we concluded that the () 15 critical decay times that weie calculated were not accurate l 16 for regulatory decisions, and that the values that they came 17 up with for their critical decay times couldn't bound what 18 we'd see in the future for decommissioned plants or apply to 19 them. It was difficult to match them up, and that's the 20 second point. We'll provide sufficient detail for license 21 amendments. These reports weren't meant to be regulatory 22 guides that you could match your specific plant up to. 23 There's not enough detail. They spoke to an issue, not as a 24 method for regulatory relief or anything. 25 So, why didn't we find the existing studies, why l l ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r: , l 379
.1 wouldn't they apply? Generic safety issue 82, they used two
() 2 3 codes called F fuel and a modification F fuel 1W, and some of the -- two of the things that struck us mostly clearly 4 was the lower burn-ups that they used in the 80's which were 5 representative of plants then didn't apply to the plants 6 today that can burn up to 60,000 megawatts for megaton 7 uranium. They also looked at the storage rackings that was 8- used at the time. They looked at low density BWR racks, ano 9 they looked up to a high density PWR rack, but it's not as 10- high density as we can currently use today. Some plants 11 have a different type of rack than what was modeled then. 12 We also found that these codes didn't model the i 13 actual fuel configuration which some plants wanted to do and l 14 we were looking for at the time. It was a stylized I l () 15 calculation that lumped the hottest fuel in the center and ! 16 then moved them out. It didn't look at any type of 17 configuration that was actual. i 18 For the '97 Brookhaven study, they used their own l 19 developed code called the shark code, and the list below 1 20 that is a mix of things from the study and from the code i 1 21 itself. The first one, the lower burn-up for the BWR that 22 was used today, we find that I think they used 40 gigawatt 23 days per megaton uranium, and BWR's can burn up higher than 24 that. So, that would lower -- for PWR's, it was 25 representative of their burn-ups for today, not as high as O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
380 1 they could possibly go today.
) 2 Beyond the report, though, we've started looking 3 at the code itself, and we found the code, we felt, 4 underestimated, even if it was modeled correctly for a 5 specific pool. They did not have in the code flow losses 6 through the grid spacers through the assembly. They'd only 7 modeled the orifice size. So, you've have a slov'..ng of your 8 air through your assembly that it didn't take -- it didn't 9 account for. So, that would tend to increase your critical 10 decay time that you would need.
l 11 We also found that it was unstable for some input, 12 that it would not converge, and that the code had not been 1 13 verified or validated. This is something we would want 14 before licensees would be able to use that. Additionally, 15 that code could also not model actual fuel configurations. 16 It gave, I guess, a better model than S fuel did. It could l 17 model -- you could input how many spaces you had full, how 18 many spaces you had empty, but it homogenized everything so 19 you couldn't tell where your hot fuel was, where your cool 20 fuel was. So, these are items we found in the existing l 21 reports that we felt couldn't be used for decommissioned 22 plants today. 23 Next slide, please. The longest critical decay 24 time that we found calculated in any of the reports was 700 25 days, and that was in generic safety issue 82 report for a l l l l ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Conneceicut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
l 381 l 1 high density PWR, so we think it's something that would be () '2 3 longer than that. We came up with a generic estimate that Joe will talk about later. We do find, though, that the 4 general conclusions in the reports ~are valid, particularly 5 this'one here, the conditions that could lead to an E- oxidation are dependent on storage configuration and decay t 7 power. That's something they came out with and led to our l 8 -- helped us lead to the conclusion that yes, we need to l 9 look at, if we want a generic solution, looking at 10 thermohydraulic analyses, we need to do calculations that 11 are more appropriate for today's plants or plants that we'll 12 see in the future. 13 That's the end of my presentation. Joe's going to 14 talk more about the thermal hydraulic estimate and the f 15 adiabatic heatup calculations, j 16 MR. STAUDENMEIER: I'm Joe Staudenmeier from l l 17 Reactor Systems Branch. I've been involved in evaluating ) l l 18 the past heat-up ca]culations, we also have some ongoing i l 19 calculations that I'm coordinating and have factored into 20 our analyses. 21 The primary reference is for past calculations, or 22 past credible calculations are NUREG CR-0649, which is a l 23 .Sandia report on spent fuel pool heat-upo, and NUREG 24 CR-4982, which is severe accident studies in spent fuel 25 pools in support of Generic Safety Issue 82. Some i O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
i l l 382 l l 1 additional calculations that we've been looking at, Chris () 2 Boyd of research performing with affluent, we've been 3 assuming a full pool for a large BWR, so it ends up being 4 about 4200 spent fuel assemblies, which is I guess the 5 closest plant to that is probably a Grand Gulf sized plant. 6 We're going to look at high burn-up, or that 7 calculation so far has only looked at 40 gigawatt days per 8 metric ton. We eventually had planned to go up to the upper 9 limits of BWR burn-up. High density racks, we're looking 10 at. I guess one reason we're looking at high density racks 11 is it seems like all the utilities are re-racking their 12 spent fuel pools and putting -- trying to stick as many in 13 the spent fuel pool as they can, putting the racks as close 14 to two inches from the wall, so not allowing much area -- I
) 15 mean, it's fine for looking at water cooling of the spent 16 fuel pool and will give plenty of cooling, but if you go and 17 look at these type of calculations with air cooling, it just 18 severely reduces the amount of cooling you can get.
19 The previously determined generic criteria from 20 the two NUREGS I cited above came up with -- that a critical 21 decay power of six kilowatts per metric ton uranium, and i 22 that very high burn-ups today, like up in the 60 megawatt l 23 days per, or 60 gigawatt days per metric ton. That ends up 24 being about three years for PWR's and I think maybe about 25 somewhere in between two to two-and-a-half years for BWR's. ) ANN RILEY & ASSOCIATES, LTD. I .\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
383 1 The calculations cited in those. reports where they come up l () 2-3 with two years for FWR's, that.was looking at, I think, fuel burn-ups of like 33 gigawatt days per metric ton, so there's 4 -- it's very low compared to the way plants are operating 5 today. I think the BWR calculations were lower than that. 6 These past calculations also had some missing 7 phenomena based on some observations from some more detailed 8 calculations that were being performed. They are very lump 9 calculations. _They didn't include 3D flow and mixing, which 10 can increase the temperature that you're getting down to the 11 fuel racks in the spent fuel pool and has the effect of 12 increasing the maximum temperature. Also, it didn't include 13 grid space or flow resistance, which is also somewhat not 14 --it's a bit of a nonconservatism, although the grid spacer 15 losses aren't dominant, but they can be significant and make 16 significant differences. 17 All of the above information we considered -- the 18 things I just talked about had the effect of lowering that 19 six kilowatts per metric ton to critical decay times, or 20 critical decay powers. That could be as low as half of 21 that, so it could bring that six kilowatts per metric ton 22 down in the range of three kilowatts per metric ton, which 23 extends the time out for critical decay time significantly. 24 Okay, next slide. We still haven't finished our 25 calculations. There are still uncertainties in these
. ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
384 1 numbers, and there is ongoing calculations and there will be 2 more in the future to try to reduce some of that 3 uncertainty. 4 Important parameters that we've seen that these 5 things depend on, rack geometry, obviously, location of the 6 hot bundles compared to cool bundles. Lots of plants have 7 purposely spread out their hot bundles, leaving empty rack ; 8 spaces behind them, but we've seen other plants that have 9 put all their hot bundles in one rack, which is probably not 10 a prudent thing to do if you're worried about zirconium 11 fires. It's the least conservative thing you can do. 12 Building ventilation assumptions have a large 13 effect on the answer. The more ventilation you have going 14 through the building, the cooler the air is in there, the () 15 more cooling you can get. 16 Fuel and rack flow resistance, down in the laminar 17 flow regime which these calculations are in and witn the 18 small distances involved with the tight packing of racks and 19 the small distances involved between rods and grid spacers 20 and things, it's a very high flow resistance configuration. 21 In many cases, especially up in the range of 22 temperature where we're concerned about, you need all modes 23 of heat transfer to really calculate the problem properly. 24
- Convection, radiation and conduction are all important.
25 Zirconium oxidation in air, there's quite a bit of O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NH, Suite 1014 Washington, D.C. 20036 (202) 842-0034
= , l 385 1 uncertainty in the data. It's, in fact, even between 2 different alloys of zirconium, there's factors of four [)
~-
3 difference in zirconium reaction rates. 4 Okay, 3D flow mixing, our calculations so far, 5 especially our CFD calculations have high sensitivity to 6 actually where the fuel is placed in the spent fuel pool. 7 You can get pressure stagnation effects that are small, like 8 on the order of five pascals but then when you figure the 9 whole pressure drop you have available or the whole 10 gravitational potential you have available is less than 40 11 pascals, it can become significant and can actually end up 12 showing that the hottest temperature fuel in the heat-up 13 calculation is not actually the highest power fuel, and it 14 just happens to be because where it's placed in the hole. () 15 As you get to higher temperatures, clad ballooning 16 and severe accident phenomena can become important. Up i 17 above 565C, you start getting clad ballooning. That's a ten 1 la hour clad ballooning rupture criteria, so that's the most 19 relevant experimental data we have for clad ballooning in 20 this type of configuration since we're talking about long . 1 21 heat-up times, Something longer than ten hour clad rupture 22 heat-up time would be a lower temperature than that even, 23 but unfortunately, we don't F ' any data on that. 24 You start gettinc severe accident phenomena 25 at temperatures as low as 64u or 650C where you start I, ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I l 386 l 1 getting melding of aluminum, which the aluminum is in the 2 bore outplates. You start melting that down, I mean, l 3 calculations -- none of the calculations take into account l 4 that so far. There's, I guess, a possibility, depending 5 upon how the plates are in there, that it could cause flow 6 blockages when the aluminum relocates after the melt. 7 That's an additional uncertainty. 8 Oxidation, as I mentioned before, there's some 9 uncertainty in that data, but since we're concerned only -- 10 and there's many more severe accident phenomena. Once the 11 oxidation takes off, I mean, you get into places where 12 zirconium starts dissolving steel, uranium starts dissolving 13 zirconium, get into clad relocation, start releasing fission 14 products up in the range of 14 to 1500C. You start driving () 15 off the more volatile fission products from the fuel, but 16 that's way beyond the criteria that we're considering. 17 I want to say something about the adiabatic heatup 18 calculations that were performed. We realize that not 19 allowing heat removal is unrealistic. I think the adiabatic 20 heatup calculations are misunderstood as to why they're j 21 performed. They were performed to give a timing estimate l 22 only for an unknown fuel configuration after an accident. I 23 mean, the fuel could have been on its side. Something could 24 have been laying on top of the fuel, and then when you start 1 25 getting that type of fuel configuration, suddenly the i ANN RILEY & ASSOCIATES, LTD. O- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D..C. 20036 (202) 842-0034 L .
387 1 adiabatic heatup calculation isn't real conservative at all. n 2 So, that just came as a timing mechanism for time available () 3 for evacuation or ad hoc evacuation if you got rid of 4 emergency planning. Our emergency planning people told us 5 that ten hours was adequate at the time, and we took the 6 simplest way possible to show that ten hours was available 7 for a specific plant, and that's the basis of the adiabatic 8 heatup calculation. It came about between probably 5:00 one 9 day and 11:00 the next day. The concept came out and the 10 calculation was performed for an exemption. 11 Okay, next slide, dominant uncertainties. The 12 dominant uncertainties and critical decay time estimates are i 13 the actual fuel rack geometry itself, the type of racks you 14 have, geometric criteria such as the orifice, size going A () 15 into it. How the fuel is arranged in the racks is another 16 criteria that can make a big difference in plant specific , i 17 calculations. Building ventilation, as I said before, flow 18 resistance and mixing and oxidation rate models. 19 Next slide, I just have Lome examples of models l 20 that we use, orifice loss coefficients, you can see, that as 21 a whole, get smaller, the orifice loss coefficient can 22 become very high. Rod bundle friction is much higher than 23 pipe friction for the same type of flow conditions. Some 24 people have brought in calculations using pipe friction 25 models. Apparently, they didn't look at the code very well
) ANN RILEY & ASSOCIATES, LTD.
l (~'/ s_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
388 1 they were using, and didn't realize the code didn't have an 1
; 2 adequate friction model for what they were doing.
l 3 Grid space losses, a lot of people have come in l 4 originally using grid space loss coefficients from turbulent 5 flow calculations, which are lower than the laminar flow 6 grid space loss coefficients. The flow resistance through 7 the bundle is dominated generally by the rod bundle 8 friction. That's probably an order of magnitude higher than 9 the orifice loss coefficients and grid spacer losses put 10 together. ' 11 Okay, next slide. Okay, zirconium-air oxidation, 12 there's a very high energy release in the zirconium 13 oxidation reaction. What's assumed is a parabolic rate 14 equation reaction rate, that's what the data was fitted to. f 15 It's the same type of reaction equation that we use in LOCA 16 calculations for oxidation reaction, although the constants 17 are different. Oxidation is worse in air than it is in 18 steam. It has a reaction rate quite a bit higher. 19 Once you get into the significant oxidation 20 regime, or the zirconium fire, pretty much the total energy 21 release is dominated by the oxidation energy and the decay 22 heat becomes insignificant. Oxidation energy becomes orders 23 of magnitude higher, like two to three orders of magnitude 24 higher. 25 Next slide. Actually, that's not my slide. There l [~hl ANN RILEY & ASSOCIATES, LTD.
\~s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
389 1 are some back-up slides up there that-I want to get into. () 2 Okay, that's a critical of temperatures versus decay time. 3 This is from the early 1980 study by Sandia from NUREG 4 CR-0649, showing how sensitive temperatures can be to things 5 such as geometry in the pool. You can see back then, the 6 calculation was done at 33 gigawatt days per metric ton. , 1
-7 PWR's now are pushing 60 gigawatt days per metric ton.
8 Since then, just the racking practices alone would push 9 everything to the right-hand side of the plot, so these were 10 configurations done in the early 80's. Now, everything 11 would probably be to the right of the far right curve if you 12 had a plant that didn't shut down prematurely and ended up 13 with a full pool, and you'd be pushing it out to three years 14 and possibly longer the critical decay time. () 15 Also, if you could hold that previously slide. 16 No, don't put it back up, but I'm going to be referring to 17 it again later on. Okay, this is basically a cartoon 18 drawing of the problem that you're locking at, also from 19 NUREG 0649. You can the spant fuel pool. You want to get 20 heat out. This is all ways that you can get heat out, and 21 it's just a cartoon to show this is the problem we're trying 22 to solve. The building and the spent fuel pool itself both 23 have important effects on the problem. Next slide. 24 This is just a sample example of an energy 25 balance. Going through in time, this is a -- I mean, each
\ ANN RILEY & ASSOCIATES, LTD. !
Court Reporters ; 1025 Connecticut Avenue, NW, Suite 1014 1 Washington, D.C. 20036 l (202) 842-0034 1 l
390 1 detailed configuration will have its own energy balance but () 2 you can see heat is being put in through decay heat, and 3 it's different ways that you can remove heat. You can see 4 in this calculation anyway that dominant heat removal is 5 just convective heat transfer. There's some radiation down j l 6 further, so that's how you get heat out of the pool. Next I l 7 slide -- there's also stored energy in the pool from the I 8 heatup. 9 This is a drawing of what these temperature 10 calculations look like. Also, you can see an example of how 11 the diameter of the hole can make a significant difference 12 in the temperatures you rise up to. That's basically a 13 change in the flow resistance. As you increase the flow 14 resistance, the temperatures go up. You can also see the r i 15 difference between having oxidation there and no oxidation. 16 You can see that oxidation probably starts to become 17 important, or where you really need to include it at about 18 600 degrees C, and you can see once you get up above that i 19 800 degree range, I mean, your curve starts to get almost 20 vertical going up. You don't have much time between eight 21 or 900 degrees, and when you're heating the pool up to where 22 you release vision products. 23 Now, there was a lot of talk yesterday, or there 24 was something in the NEI presentation about that how it was 25 an overly conservative temperature criteria and oxidation ANN RILEY & ASSOCIATES, LTD. [~'} (_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
f 391 1 criteria that was used. This is an oxidation compared to ( 2 data that was used in 0649. I mean, you look at this, and 3 compared to data, I don't see that as being conservative at ! 4 all and if anything, I would say it could'be a.little 5 nonconservative. Next slide. 6 So, it's the Sandia curve is also on there, the 7 previous curve. This is from NUREG 4982. It's comparing 8 the Sandia data and another curve where they tried to make 9- one parabolic rate equation fit over the whole temperature 10 range as opposed to the few region Sandia curve to some more 11 recent data. Still, I really don't see the overly 12 conservative oxidation numbers for this Sandia curve. You 13 can see also, there's some steam oxidation curves which are 14 in there to compare to. You can see that air oxidation is
) 15 much higher than the steam oxidation. Next slide.
16 It's back to the one that I had before. Okay, you 17 can see up there, this is another example of why 800 isn't 18 conservative. You can see that once those temperatures get 19 up to 800, they pretty much go straight up. Your increase 20 powerages slightly, and they're going up to infinity or up i 21 to fission product release, so, in fact, looking at this 22 type of information, I would a y if I was going to do 23 anything to the 800, I think you would want to lower it to 24 give yourself a little more margin, especially considering 25 all the configurations that are going to be coming up in the ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
392 L 1 future are on the far right hand of that plot. I mean, you 2 start going vertical below 800 in those cases. 3 That's all I have to say now. I guess if there's l l 4 any questions about it, I can answer them in the discussion 5 period later on. I 6 MR. CAMERON: Okay, thank you, Joe, and thank you, 7 Diane. Are there a couple more slides that you guys need to 8 go through? 9 MS. JACKSON: There's one more slide, and we'll go 10 through it now, but it's really for the discussion period. 11 Currently the staff is looking at doing a few more generic 12 calculations towards this, but we're asking industry, we're 13 asking the public, are there alternates to the values of the 1; parameters we've put in. We assumed a fuel pool because we 15 can see plants in the future going to a fuel pool and 16 shutting down. We didn't think that was conservative for 17 looking at rulemaking. 38 Also, if you look at the items on Joe's slide l ! i 19 that's important parameters and phenomena, if you'd let us ! 20 know what you think industry practice is or will be that's l 21 representative, we'll consider those in our future generic l l 22 calculations. l l 23 If you have an alternative calculation to the )' l 24 adiabatic heatup that would give us just a metric for l 25 sufficient time for the progression of the scenario, we're , ANN RILEY & ASSOCIATES, LTD. I Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ; Washington, D.C. 20036 ! (202) 842-0034 ! I L
393 1 certainly willing to hear and would like to hear what else f'T L 2 can be used. These are just something that we came up with l 3 that would give us an appropriate answer. If there's 4 something else that you think would be more appropriate, we 5 would certainly like to hear them. 6 That's the end of the staff presentation. 7 MR. CAMERON: Okay, and Phil, if you could just 8 hold that last slide ready for when we're done with the next 9 presentation because I think we'll be going back to that. 10 Dr. Niyogi, are you ready for your presentation now? 11 MR. NIYOGI: Yes. My name is Kalyan Niyogi, 12 director of technical division. I think NRC's presentation, 13 Diane and Joe's presentation are pretty good introduction to 14 what I am talking about. My discussion will be based on our
/-
(N) 15 evaluations of a couple of plants we already did. For 16 example, CY and Millstone 1 and also some of the generic 17 information we developed for our work. 18 We will discuss the scenarios we considered to 19 evaluate the consequences of various accidents. We will 20 discuss methodology used for each of these calculations. 21 I'll present some evaluations of specific cases and try to 22 draw some conclusions. 23 I have not tried to develop any generic conclusion 24 out of it because the plans we have considered are typical 25 plans which have gone through years of fuel in their pool. l l l
} ANN RILEY & ASSOCIATES, LTD.
s/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l l t
l l 394 1 1 This is not a plan which has very high burnup. These are
/(_/') 2 not plans which are just shutdown and ready for l 3 dt ommissioning.
4 There are several scenarios we look at. Two of 5 them are important -- spent fuel pool heatup with no spent 6 fuel pool cooling. That's one scenario. The second 7 scenario, the loss of cooling water. In the first scenario, 8 which is spent fuel pool heatup with no spent fuel pool 9 cooling, we considered the consequences when you have the 10 ventilation system available, and also we considered when 11 the ventilation system was not available. By ventilation 12 system, we mean the forced ventilation, and the cooling is 13 done by natural circulation of air. 14 For all these calculations, one tn; g is common, 15 the calculation of decay heat. As it has been pointed out 16 in various discussions that the calculation of decay heat 17 varied from method to method. We used ORIGEN as our basis 18 for all our decayed calculations, and we found it's 19 consistent ~and reliable. . I 20 Another thing which is very important, the pool 21 evaporation model. The information in the industry, ! 22 available information, based on air conditioning, handbooks 23 and stuff, sometimes they don't cover the range of 24 . temperature we are interested in. So, what we did, we did 25 an actual test, and also we developed some basic information
/
} 7dR1 RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
395 1 for evaporation rate and we found these are much better -- () 2 3 these correlations are much better than what you can do fro m the extrapolation of HVAC kind of data available in the 4 industry. Also, we did some actual tests with Millstone 5 Unit 3. They are benchmarked, and they're pretty good 6 there. 7 The spent fuel pool and fuel building 8 temperatures, that these are calculated to find how high the 9 temperature of the building gets because there are equipment 10 that should survive, and as Dan pointed out, the building 11 information on how the circulation is allowed whether you 12 have opening or you have other means of having circulation 13 in the air in case that we have no forced ventilation - 14 very important. () 15 Sometimes we calculate what if kind of situation, like you open the hatch, see how the temperature 16 is brought down to acceptable limits. 17 The loss of cooling water part, we used a CFD 18 model for the pool. 19 We considered heat transfer within the racks and the sounding buildings. Also heat transfer 20 between the fuel assemblies and the racks. Just to show 21 schematically, when you have drain down all water from the 22 pool, the only thing that is remaining to circulate around 23 the racks and cool. Also, it is important how the building 24 is ventilated. 25 Just to refresh your memory, these are typical ANN RILEY & ASSOCIATES, LTD. l Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
396 1 racks, but there are a variety of racks in the industry, and I '\ 2 they are very important. b We could not develop any generic 3 information for that. There are racks with flux straps and 4 side holes and things like that. They play a very important 5 role in thermal hydraulics. 6 The computer program we used for the problem in 7 the industry or find that element calculations, we found the 8 core has sufficient flexibility to incorporate all the 9 factors of elements which are important to our problem. At 10 low temperature range, the dominant factor for heat transfer 11 is the convective heat transfer. At higher temperatures, 12 "he radiat. ' conduction becomes important. 13 MR. CAMERON: Would you rather use that? That 14 looks comfortable. () 15 MR. NIYOGI: Thank you. There's another thing 16 used with the porous media assumption. What it means, the 17 acts and the fuel pool assemblies were considered as porous 18 media, and the resistances are calculated for the flow path 19 of the air very meticulously and converted to the porous 20 media parameters. Since we designed the racks, we have all 1 21 the details of the information. We have a pretty good 22 handle on that and can calculate the losses fairly closely l 23 to the actual value. 24 One other question comes back and forth, is it 2D 25 or 3D assumption, whether we should make it 3D model or 2D
/~)
5 s/ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
397 1 model, what is acceptable. Like any other thing, I believe () 2 3 it's engineering judgment, if you can idolize a 3D pool with 2D assumption, should be good enough. At the same time, I 4 believe you'll be penalizing yourself a little bit because 5 additional conservatism will come into that. It's not that 6 2D-is unacceptable. The other way, it should give a higher 7 number, but if you get lower number, something is wrong in 8 the engineering judgment. 9 We have done both 2D and 3D and compared the 10 results. When there are significant difference, we found Ell the engineering judgment was wrong, but you have to develop 12 some kind of intuition how the flow field looks like, or 13 that the flow will come to the volume first or mix to the 14 top. These are the things which are not sometimes very 15' straightforward. 16 Again, we have done significant amount of work to 17 come to theEconclusion that 2D could be equally acceptable 18 but with some penalty. What do we do with that calculation? ! 19 We calculate the maximum local temperature and then on the 20 basis of the maximum local temperature, calculate the fuel 21 cladding temperature. l 22 Before going into the effects of it, let me show 23 you schematically how we used to do in the past the l 24 calculations. The thermal chimney effect is schematically 25 shown by this. The hot channel go up, the cold channel go l ANN RILEY & ASSOCIATES, LTD. Court Reporters ! 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 ( L
I 398 , down. l 1 Many years ago, early 70's, we used to calculate all j () 2 3 these by hand calculation, and I still remember doing it for operating plants still running. Now, we have the advantage 4 of high speed computer and PC's are very powerful, and we 5 can simulate millions of grade and simulate very precisely 6 the actual geometry and all the details of this. The codes , 7 of the industry are much better now, very reliable, and , I 8 they're based on solid theoretical bases, and they give i 9 results. 10 Let me show you some of the 2D analyses we did. 11 One thing you have to remember, to make a 3D geometry into a 12 2D, you have to distort things. You have to put the hottest 13 flow in a certain way that it is representative. Also, it's 14 not too much penalizing. 15 This is -- the color is a little different there, l 16 but this shows the temperature distribution and I see the 17 hot spots shown. As I mentioned, we do quite a bit of 3D 18 analysis. A pool could look like this with all this small I 19 geometrics there, and this is the flow field. You can see 20 the cold fluid coming from this end going to the bottom and 21 coming back, and because of the higher resistance in the l l 22 racks, the velocities are smaller. There's a lot of l 23 circulation on the top and the sides and the bottom, and 24 because of friction, the velocity drops. 25 This is the temperature profile in one cross
/T ANN RILEY & ASSOCIATES, LTD. \ms/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 1
(202) 842-0034
l 399 L 1 section through the middle of the racks to the hot flows. () 2 This is just to show how I've taken now the highest range of 3 temperature and distributed in the lower range more 4 accurately in this one. 5- This is the velocity vector in the diagonal plane. 6 This is the surface which is interfacing with the building 7 air. Sometimes the question asked, how far do you go? It 8 can make the whole building as a part of this model, but we 9 found we can find a reasonable boundary and still get the 10 results fairly accurate, and that cuts down the number of 11 grids. 12 Again, I have a lot of pictures to show. There's 13 another question whether thermal radiation is important. It 14 is very difficult to accurately model the thermal radiation () 15 because in a traditional-way. So many surfaces are here, 1 16 and we believe there is significant radiation between the 17 fuel rod, the surrounding and so on, and also there is 18 significant conduction at high temperatures. We found a way 19 to get around that. Rather than going into details of each } l 20 and every surface, which is impossible, we developed some l 21 technique to incorporate the radiation. I don't want to get 22 into the details, but of course, fluent has a capability to 23 incorporate some of it, but you have to be careful. 24 Definitely, radiation is important when your temperature I l 25 goes above five, six, 800 centigrade. l l O
\s)
ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 i l
400 1 I mentioned about 2D, 3D, and again, what do we do 2 now in order to the time you spent arguing that 2D is good 3 enough, you can as well do a 3D analysis, and at the same L 4 time, you are not penalized. So, we have started doing 3D 5 analysis irrespective of whether it's needed or not, and let ! 6 the argument go away. 7 Fuel shuffling, for many plans, we found fuel 8 shuffling give significant reduction of temperature rise, 9 especially the tendency has been to put the hot fuels in one 10 place, as somebody mentioned. We found that by shuffling 11 the fuels and spacing out, putting in a checkerboard way, it 12 relieves alot. On'the other hand, the plans which will be 13 having very limited room for shuffling will have difficulty 14 to shuffle and so again, for plans which don't have much 15 room, we have limited success in shuffling. I 16 Rack design. As I mentioned briefly, the rack l 17 designs are important consideration because especially the 18 ones which have the flux trap which is the space which you 19 can take credit for air circulation. There are racks, the 20 channels are not continuously welded, but welded in spaces 21 only. That relieves some of the cross -- that allows some l 22 of the cross current of air and gives much -- leave to l l 23 temperature rise. So, I think your observation, NRC's 24 observation that the variety of racks may give significant 25 relief, it may not be significant when it comes to high ! l l ANN RILEY & ASSOCIATES, LTD. s_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
401 1 density and packed situation because in that case, by [G) 2 allowing cross air flow doesn't help much, but when you have 3 room available, it changes the picture. 4 Conclusions, CFD techniques provide acceptable 5 predictions, and we have verified that by means of various i l 1 6 other means of calculations,_and they're pretty reliable. 7 Plant specific analysis is necessary to factor in burnup, 8 fuels, its history, discharge history, rack design details, 9 etc. 10 I think this is contradictory to what we are 11 .trying to achieve today, to come up with some generic 12 answers, but what I'm saying, the plants which I'm 13 investigating would be penalized if you use some upper bound 14 analysis. The burnup of the half of 60 from 30, 40, even () 15 worse sometimes. The full history based on history I found, 16- that the significant conservatism in the burnup built-in 17 with that too, because the way it is kept record and then 18 you take upper bound of that, basically is conservative 19 estimate by itself. 20 As I mentioned, fuel shuffling may provide 21 significant advantage. Again, it's qualified for plants 22 which have a lot of space you can think of shuffling. Then 23 you get into significant problem of shuffling between hot l 24 fuel and cold fuel, contaminated fuel. So, you found that ! 1 25 there is definitely some advantage if you can do that. O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
402 l 1 Now, as I mentioned, I have not tried to develop () 2 3 any generic information here, just to give what you do and what you have done for several plants and what tools are 4 available,-and the sensitivity of certain parameters be 5 observed. I believe to develop something generic and ' 6 applicable, it is not going to be very easy because the 7 plants are of different categories, but still, it can be 8 developed to some extent and then left to the variety of the 9 plant design and other parameters. Some generic information 10 can be developed, but for the plants I'm involved in, I 11 found they would be somewhat penalizing. 12 That's the end of my presentation. 13 MR. CAMERON: Thanks a lot, Dr. Niyogi. I'd like 14 to go to Diane Jackson and Joe Staudenmeier. Do you have () 15 any. questions or comments on Dr. Niyogi's presentation from 16 the perspective of what you've done on the NRC part of the 17 study, including the need for further information that you 18 mentioned at the end? 19 MR. STAUDENMEIER: I guess we would be interested 20 in discussing some of his CFD calculations in a more 21 detailed manner at some time, because I guess what we've 22 been seeing in our calculations is that these CFD 23 calculations aren't real straightforward all the time, and 24 they can be very sensitive to even things that you would 25 think of as small things like turbulence models that you use ANN RILEY & ASSOCIATES, LTD. O. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 1 i Washington, D.C. 20036 (202) 842-0034
403 1 in the code, and also, you're really pushing, I think, the
T 2 node size that you are able to use in these models and 1 (G 3 especially in 3D models, is probably at the upper limit of 4 what you'd want to use, and you have difficulty performing 5 convergent studies and things like that to make sure you're 6 truly getting converged results, I think, when you're 7 pushing the limits in these 3D calculations.
8 I guess also, we'd be interested in I guess more 9 details about the radiation calculation that's being used 10 since it's not a standard thing, if it's like a diffusion 11 approximation or something like that, that you could maybe 12 implement in that type of model and compare it to some 13 results. PNNL out in Richland is doing some calculations 14 for a decommissioning group with COBRA where they're doing O) ( 15 detailed subchannel modeling and detailed radiation modeling 16 over quite a few bundles in their calculations and seeing 17 how accurately this simplified model is predicting things 18 compared to a more traditional way of doing radiation from 19 surface to surface view factors. 20 MR. CAMERON: So, you're suggesting that it would 21 be useful to sit down and talk further about this with Dr. 22 Niyogi? 23 MR. STAUDENMEIER: Yeah. I don't think we have 24 time in this forum to talk about those type of details. 25 MR. CAMERON: For we laymen out here, it may not l
/T ANN RILEY & ASSOCIATES, LTD.
(_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
404 1 be a question of time, but I think that would be a good 2 idea. Gary? 3 MR. HOLAHAN: I'd like to pursue a little 4 different question, which is how do we expect the thermal J 5 hydraulic calculations to be used in our study or in any 6 licensee activities with respect to spent fuel pools. One 7 example is we heard we had some difficult using the ANS 8 decay heat model. We heard that you were happy in using the 9 origin code, but then yesterday we heard that, in fact, 10 licensees are in a position to do measurements on the pool 11 to get a realistic assessment of what decay heat is. In my 12 mind, this raises the question of how much should we be 1 ) 13 calculating and how much should we just, you know, allow the 14 licensee to rather then sort of keeping track and doing 15 calculations, just make a measurement or a realistic 16 assessmerit of what the heat load is in the pool. 17 The other issue I'd like to put on the table is 18 there are really two rather different thermal hydraulic 19 calculations. Joe talked about heat-up time. If the water 20 is gone from the pool, how long does it have to heat up to 21 self-sustaining oxidation calculation. Are we really going 22 to use that in any of the likely regulatory decisions? Do 23 we need to be able to calculate such a thing?
]
I 24 The other calculation is, I think, what Diane 25 mentioned earlier, this critical time after which you don't l ANN RILEY & ASSOCIATES, LTD. l Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 405 1 expect that even without water in the pool, that the
/
(m) 2 temperatures will get up to oxidation. It seems to me from 3 all the other things we talked about, what we may be saying, l 4 that the risks for the spent fuel pool appear to be 5 relatively low, and we're sort of trying to settle out how 6 low is low, even for the period of time for which Zirc fire 7 is possible. If I think of this as a time dependent 8 calculation, then a few months, few years after shutdown, 9 presumably you could have a Zirc fire, but it's unlikely. 10 If you go out further past this critical time, it becomes i 11 unlikely to be out of water in the pool for a significant 12 period, but even if it were, there might be a, you know, l 13 worker and shine dose issue, but we don't really have an 14 offsite public health and safety role. So, what role is the l O) ( 15 thermal hydraulic calculation, or any of the thermal 16 hydraulic calculations, going to play in selling out these 17 issues? Which ones do we really need to focus on? What do 18 we need to calculate? How well do we need to calculate it? 19 MR. CAMERON: And Mike, do you have something to 20 say in response to that? 21 MR. MEISNER: Surprisingly, I was going to say 22 almost the same thing as Gary, 23 MR. HOLAHAN: That is surprising. 24 MR. MEISNER: I think -- well, one obvious problem 25 after hearing both of these discussions is it's one thing to ANN RILEY & ASSOC.TATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l 406 1 use -- get out of using bounding values in probablistic i
) 2 t J - ;:e , but boy, it's really rough in deterministic space.
3 Like several people have pointed out, the plants you're 4 talking about now are well below various of these 5 parameters, for instance, on burn-up and the kinds of time 6 periods you're coming up with are very, very conservative 7 for them. 8 I think maybe the red he ag in all this is, and l 9 I think is the direction Gary was L;ing, was that the flow 10 is using these kind of criteria is that they're not risk 11 informed, and when you apply risk, I think they go away, i 12 So, what's the need for those calculations? I'm not sure 1 13 there's any need at all if we focus our time and attention 14 on the risk side because this isn't giving us consequences, j 15 It's not giving us any insight as to probability of these a 16 events, and I think that's the role of the risk evaluation. 17 I'd also like to make a comment on the -- so we 18 don't think it's an appropriate criterion at all, the time 19 beyond which a Zirc fire can happen. Similarly, the same 20 thing as far as this adiabatic heatup, but for a different 21 reason. This is more of a policy and optics issue. That 22 was first applied, I believe, at Maine Yankee, wasn't it, i 23 the adiabatic heatup? I can tell you, in talking to some of 24 the folks there in the community, here's the impression that 25 they got.
/~h ANN RILEY & ASSOCIATES, LTD.
k-} Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
407 1 The NRC said that the licensee doesn't have to j
) 2 have an off site emergency response capability anymore. The 3 NRC says they don't even have to worry about it anymore, and 4 what they've done is, they've essentially thumbed their nose {
I 5 at the public and said, it's your problem. Now you 6 evacuate. You can do it in ten hours. That's the 1 7 impression that was given. I don't think that's good 8 policy, and for that reason, I don't think that's at all an 9 appropriate criterion to be considering for resolving any of 10 these issues. 11 MR. CAMERON: Gary? 12 MR. HOLARAN: I'd like to follow up on a few of 13 the things that Mike said. I would disagree with one point, 14 which is you said that these calculations don't really have I l () 15 16 to do with consequences, and I think they have a lot to do l with consequences. In other words, if you were comfortable 17 that after a certain period of time, you were not going to 18 get a self-sustaining Zirc fire, I think that very greatly 19 affects, you know, the public dose consequences. 20 MR. MEISNER: Yeah, sure. 21 MR. HOLAHAN: Some of the stuff we talked about l 22 earlier and whether it's how the operators are trained or 23 how much instrumentation is available and those sorts of i l 24 things, I think we have to ask the question about whether ' 25 the things that utilities talked about yesterday or that , i > '~ s ANN RILEY & ASSOCIATES, LTD. l Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 4 Washington, D.C. 20036 ' (202) 842-0034
408 1 will be in our follow-up discussions are special precautions () 2 3 that ought to be kept in place early after a plant shutdowns until this critical time when you don't have Zirc fire 4 concerns, or whether these are -- I done know how you want 5 to term it, but are the, you know, normal sort of 6 precautions that a utility would propose to keep in place at 7 all times. 8 So, for example, the level of training and the 9 instrumentation licensees, the licensees are going to keep 10 that level of protection in place before and after such a 11 critical time, then it doesn't become so important to figure 12 out what that time is. If you wanted, for example, say 13 after five years, just arbitrarily, say Zirc fires aren't 14 possible, that at that point you wanted to reduce some of 15 the requirements because the consequences very substantially 16 drop off at that point. Then I think that thermal hydraulic 17 calculation and the consequence calculation play an ; 18' important role. If that's not in the mix, if that's not the 19 way the utility proposes to run its shop, then this becomes 20 -- the details of this become a lot less important. ) 21' MR. MEISNER: That's a very good point, and while i 22 we haven't gotten on to specifics as to what those ; 23 requirements or commitments are, you know, and the rule of i 24 thumb is to what level do they reduce risk. I would say 25 that all the stuff you heard from us yesterday is just good O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ) Washington, D.C. 20036 (202) 842-0034
409 1 common sense, and I don't believe -- I can't think of (Gj 2 anything that we discussed yesterday where we would want to 3 remove it simply because we had gone past the point of a 4 Zirc fire possibility. We have those in place, not for Zirc 5 -fire, but because we don't want to have loss of cooling i 6 events. You know, we don't want to have drain downs, and if 7 we do, we want to be able to respond to them. So, I don't 8 see that that -- that anybody in the industry would be well 9 served by removing those things. 10 The other thing I wanted to say, too, if you're 11 talking about periods now of five years for potential Zirc 12 fire, you know, I think most plants that decommission now 13 and in the future, at that point, they've gone to infancy. 14 You know, they don't even have pools anymore, or they're () 15 just about to get to that point. 16 MR. HOLAHAN: Five years was just a number. You 17 know, I believe the calculations are whatever the 18 calculations are. I guess at least some, most, and maybe 19 all of the utilities who spoke to us yesterday are beyond 20 the point of Zirc fire, or at least they certainly feel they 21 are. 22 MR. CAMERON: At some point we need to get it on 23 the record what the answer is, so just speak into the 24 microphone one at a time. 25 MS. JACKSON: I can answer that. Oyster Creek, / 'N ANN RILEY & ASSOCIATES, LTD. \s- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 ( l 410 1 who was there yesterday, and they haven't shut down. They're just looking for future, and Zion is here today and / 2 v 3 they're undergoing the exemption process. They're on a case 4 by case basis. Oh, and Millstone. 5 MR. HOLAHAN: Did Big Rock yesterday? 6 MS. JACKSON: Yes, and they had an exemption. Big 7 Rock spoke, and they had an exemption, emergency 8 preparedness. 9 MR. HOL' HAN : The point was, is there any 10 difference between the way the utilities are treating their 11 spent fuel pools, whether they think this is early after l 12 shutdown or substantially later, or would they want to treat 13 it differently? i 14 MR. MEISNER: The answer is no. l (ss ) 15 MR. HOLAHAN: Okay. I think that plays a role in 16 our thinking. 17 MR. CAMERON: Is it premature to -- is that 18 something that rises to something we should put up there as 19 a neither action item nor a commitment is perhaps the right 20 word for this, but do we want to make a note of the 21 importance of this issue? 22 MR. HOLAHAN: Yeah, I think it's a significant 23 point and, for example, Mr. Lochbaum's comments yesterday 24 about, you know, beyond the public health and safety issue, 25 there's an issue of worker protection for the people who are f') ANN RILEY & ASSOCIATES, LTD. ( ,/ _ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
411
'l on site. I think if utilities are providing the same level
/ 2 of protection, you know, later as they are earlier, I think 3 it goes at least partly to cover that issue. In other 4 words, you're not reducing the level of protection just 5 because it's been three years, if that's more comfortable to 6 you, because, in fact, all you're really saying is if you 7 have one of these very serious events, you would have a l 8 serious on site source to deal with, and avoiding that, you l 1 9 know, is important to the workers, even if there's no off 10 site consequences. 1 11 MR. MEISNER: And that's wP do it right from 12 the beginning, not because of Zirc fire. l 13 MR. HOLAHAN: Okay. 14 MR. CURRY: Jim Curry, GPU. I think you were () 15 asking about utility input. You know, basically at Oyster 16 Creek, our philosophy is simply we don't change the design I 17 basis of t:.e fuel pool as long as there's fuel in the pool. I I 18 So things like NUREG 0612, heavy load restrictions, 19 administrative controls on chemistry, to maintain the i 20 integrity of the Boraflex level, to prevent a draindown 21 event of any nature, that's really our design philosophy. 22 For us, the calculation of the Zirc fire window is an 23 interesting one, but our philosophies don't drain the pool. 24 So, really, the administrative controls to prevent that, and 25 physical constraints and staffing and training to prevent I ANN RILEY & ASSOCIATES, LTD. O. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
412 1 that event is really our design objective. () 2 3 MR. CAMERON: MR. HUBBARD: George? I've got a clarifying question on 4 that, Jim, just from the standpoint, because I know with 5 most BWRs, you would have an RHR assist mode in the cooling 6 and that, I believe, you know, some plants are -- as the 7 heat load gets down, they would be reducing or taking out 8 that part of the system. Now, could you address, you know, 9 Oyster Creek's approach to -- you know, because in the 10 operating plant, that is a backup system to assist, you 11 know, when you do an off-load. But, how would you factor 12 that in? 13 MR. CURRY: I think -- the first thing, I think 14 you're on to a good question, when you talk about Oyster g 15 Creek and other plants. But, our philosophy, again, is to 16 not change the likelihood of a draindown event, not 17 introduce any new hazards, just because we go into 18 decommissioning. So, when we design our spent fuel pool 19 cooling systems that are sized to the load that we have, we 20 probably won't put in alternative systems the day after we 21 shut down. We'll pick a time that makes sense; probably, a 22 year or so after shut down, when the decay heat loads go 23 down to a certain point. Now, we -- and that curve begins 24 to flatten out a little bit. 25 At that point, the amount of redundancy that we h O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
413 1 havs ;n those systems and the amount of support will be (f 2 designed into the system, such that the probability of a 3 draindown event -- and, again, our thinking is don't drain 4 the pool. You don't drain the pool, you don't have zirc 5 fire, nor do you have any other consequences associated with 6 a draindown event, which just seem fairly nebulous and 7 difficult to define. So, we will -- our intent is to design 8 a system of adequate robustness, to make the probability of 9 a draindown event essentially negligible or incredible. 10 Just a follow-up event, I think when we -- a 11 follow-up question, when we talk about analyzing zirc fire, 12 and we don't want to forget about the embedded assumptions 13 that we have in here. I mean, we talk about this model and 14 that model, but each of these models has key assumptions; () 15 for instance, the gap between, you know, the fuel elements 16 and the wall. So, you have to make sure that those embedded 17 assumptions don't change. So when you get into the 18 consequences business, it's not just, you know, agreeing on 19 zirc fire and how to calculate zirc fire in a given 20 geometry; but all those assumptions that you made when you 21 did that calculation, you begin to spread out and you have 22 to enter those regimes, as well, to make those -- to make 23 sure those assumptions remain valid. So, we think that's a 24 real tough thing to do. 25 MR. CAMERON: Is there a -- is there a question O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
414 1 there that you would like any response form the group on? I~ 2 MR. CURRY: No, I think that was just a comment, k)/ 3 as we talked about the different ways to calculate the 4 duration of the zirc fire window, don't forget that you're 5 making embedded assumptions and you have to keep track of 6 those assumptions or the analysis really isn't any good. 7 MR. CAMERON: All right. Gary? 8 MR. HOLAHAN: I presume if it's difficult for the 9 iicensee, who has the pool and knows what the rack design 10 looks like. If it's difficult for you to do the 11 calculation, it would be really difficult for us to, you 12 know, have a hundred different ones well established. 13 However, it seems to me that we've been talking about, you 14 know, how low is the risk, spent fuel pool, decommissioning. () 15 When you go past this point, difficult to calculate as it 16 may be, there is, in fact, a very substantial drop in the 17 public risk, or you could call it a very substantial 18 increase in the safety margin or whatever. 19 So, I think even if we don't have to have a 20 precise value because it plays a specific role in a 21 decision, you know, we're not going to grant this exemption 22 until three weeks from Thursday, because that's what the 23 calculation says. From a fuller understanding of what the 24 risk and the safety margins are, I think at least that part 25 of the calculation, we ought to continue to pursue, even ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 dashingt:.n, D.C. 20036 (202) 842-0034
i r 415 1 though we would recognize that maybe you -- there's quite a (d \ 2 lot of variability from plant to plant, situation to 3 situation. I think it would be important to understand, in 4 general, what these time period are, even if they're not 5 specifically determining the licensee decisions. 6 MR. CAMERON: Annie, we're going to get to you 7 right in a second. But, it would be useful to get some 8 reactions to what Gary just said. 9 MR. CURRY: There is one comment or question. 10 When you talk about the duration window, that's one thing. 11 Can you comment on what you feel the consequences are, if 12 you did have an ignition of the clouding -- an oxidation -- 13 rapid oxidation of clouding? 14 MR. HOLAHAN: We did some preliminary dose f% Q' 15 calculations, and I think Brookhaven did some in a previous 16 report. We didn't include them in our study, because I 17 think they're not -- they're not quite the same level of 18 readiness for comment. But, if somebody could comment 19 generally on it -- 20 MR. KELLY: This is Glenn Kelly from the staff. 21 Can you help me understand what specifically about the 22 consequences you want to know? 23 MR. CURRY: I think the question is, in terms of 24 how they compare to reactor accident consequences, because 25 if you think of comparable risk profiles, things like that,
/)
(._/ ANN RILEY & ASSOCIATES, LTD. Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 m
416 1 again, I think our view is, look, we want the probability of A i, 2 this accident to be -- or this event to be negligible. U 3 That's really the objective here. But, from what I heard 4 Gary say at an earlier presentation, look, we're interested 5 in the consequences, because if the consequences are so 6 onerous and if I think in terms of the risk profile, that's 1 7 really what I want to try to compare. So, I think we -- we 8 hear that and that's fine. I 9 So, the piece of it is, well, how do the 10 consequences compare to consequences associated with a ; 11 reactor accident or things like that. Because, if we try to 12 compare comparable risk profiles, well, that's the other 13 piece. And as you ask the question, Gary, gee, how much -- 14 how much calculation should I be doing, I think you need 15 that piece. 16 MR. BARRETT: Let me say a couple of words in 17 response, if that's okay with you, Chip. 18 MR. CAMERON: Oh, yes. 19 MR. BARRETT: And then for more detail, I can 20 leave it to Glenn Kelly; but, I think just to provide some 21 context. Qualitatively speaking, how does this compare with 22 a core melt type of accident? The major features that you ) 23 have to consider are, first of all, on the positive side, 24 you generally are at a point in time where the iodides have 25 decayed away. And when the iodides decay away, what that ANN RILEY & ASSOCIATES, LTD. (- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034 l t
417 1 does is significantly reduce the potential for early [') v 2 fatalities, because early fatalities tend to be driven, to a 3 great extent, by the iodides. So, that's a very positive 4 factor. 5 On the negative side of the ledger you have -- you 6 have much larger inventory of fuel than you would have in 7 the core. You have multiple cores present and available for 8 release of radiation, and you don't have the containment -- 9 the level of containment that you would have for a core 10 damage accident. So, in a qualitative sense, what you -- 11 those are the major factors that you uee in the consequence 12 analysis. 13 For -- when you look at the actual consequence -- 14 those consequences, themselves, what you see is, on the one (A) 15 hand, the integrated person rem calculations. And as you 16 probably know, the calculation of latent fatalities is 17 essentially directly proportional to the integrated person 18 rem and those tend to be on an order of magnitude similar to 19 a large release. okay. On the other hand, when you look at 20 the early fatality calculations, because of the absence of 21 the iodine, you tend to see -- and also because you 22 generally give a lot of credit for evacuation in these 23 sequences, although we did calculate cases with and without 24 effective evacuation. But even without effective 25 evacuation, because of the absence of the iodine and because l [' ANN RILEY & ASSOCIATES, LTD.
\s- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 i I
l I
418 1 early fatalities tend to be a threshold effect and tend to () 2 be sensitive to the iodine, the early fatality calculations 3 tend to be much lower than you would see in a spent -- in a 4 core damage accident, resulting in a larger early release. 5 And I'll ask Glenn if that's fair and ask him to 6 expand on that. 7 MR. CAMERON: Thanks Rich. Glenn? 8 MR. KELLY: That's correct what Rich said about 9 the characterization of the risk results. Again, these were 10 based on our assumption that in the pool, two full cores 11 worth of material burned up in the zircaloy fire and that 12 there was no effective containment by the building that the 13 spent fuel pool was in and that we had -- looking at the 14 building -- the heat -- the temperatures from the zircaloy () 15 fire sufficiently high that will melt the building. And we 16 determined that looking at 95 percent effective evacuation, 17 you got around -- these are numbers off the top of my head 18 -- right around 2,000 -- 2,200 latent fatalities. That's 19 not -- that is not a -- that has not been conditioned by the 20 probability and, basically, like three quarters of an early 21 fatality. So, when you look at those times the probability . l 22 of the event, it's a very small number. 23 MR. CAMERON: Okay, thank you, very much, Glenn. 24 Let's go to Ray Shadis. 25 MR. SHADIS: I will remark, Glenn, that that's
/~T IdRJ RILEY & ASSOCIATES, LTD.
kl ms Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r-419 1 just human lives. The property loss stays about the same l' T 2 and, of course, that's always a concern. The question was V 3 raised here as to -- and Gary may have to help on this, 4 because it was his question, but it got lost in the circuit 5 here, is to whether the utilities would maintain the same 6 operating characteristics for their spent fuel pool after 7 what, after the rest of the plant was down or after a 8 certain period of time that expired? I don't recall. 9 MR. HOLAHAN: I think what I heard was so long as 10 there was -- at least from Oyster Creek's point of view, so 11 long as there was any fuel in the spent fuel pool, they 12 would maintain the quality of programs in the vigilant, or 13 however you want to characterize it, as when they had first 14 shut down. O g 15 MR. SHADIS: I see. 16 MR. HOLAHAN: Or, in fact, perhaps as they -- I'll 17 take a picture while I'm pushing on my nose -- 18 [ Laughter.) 19 MR. HOLAHAN: -- as during normal operation. 20 MR. SHADIS: I see. I misunderstood -- 21 SPEAKER: Are we paying for these pictures? Can I 22 veto some of these pictures? 23 [ Laughter.] 24 MR. SHADIS: Okay. But, I misunderstood your 25 question. I thought that you were asking if everything l l (T
\_,/
ANN RILEY & ASSOCIATES, LTD. Court Reporters I 1025 Connecticut Avenue, NW, Suite 1014 2 84 - b34
420 1 remains the same during this period, and the fact is i
/ \ 2 everything does not remain the same.
b l 3 We were discussing yesterday whether or not to 4 give credit in the safety analysis to the safety culture, 5 whatever you would call it, of the companies maintaining 6 these spent fuel pools. And in that same vane, the culture l l 7 at the site changes radi.cally during decommissioning and, in l 1 8 fact, may be out of the hands of the parent company, the 9 owner company that's maintaining the spent fuel pool. 10 Because, you have an influx of contractor workers. i 11 Many of them don't have any specific nuclear training. They 12 may not be, you know, brought up in that same nuclear safety 13 culture tnat the ordinary workers that come in for refueling 14 projects would have. You know, they're not there to meet () 15 specific tolerances and they're not conscious of the 16 regulations or the interactions of components. We had at -- 17 at Maine Yankee, we did have an incident in decommissioning 18 with asbestos workers. It was just hard to find workers. A 19 lot of these guys just didn't have the training. 1 20 And so, what I'm getting to is that that impacts 21 -- if there's any work being done in the vicinity of the 22 spent fuel pool, if you're replacing that crane when yc: re 23 getting ready to do those heavy lifts or doing modifications i 24 on any of that stuff, if you're removing asbestos from any 25 piping that may be in that building or electrical worker, l 5 ANN RILEY & ASSOCIATES, LTD.
~_ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
i I 421 1 any of it, you would have to pay special attention to
)
() 2 3 questions like -- and I don't want to discourse on this, but questions like foreign materials exclusion. Because, if you 4 -- if you drop a sheet of plastic, whatever it may be, some 5 insulation into the spent fuel pool, that's going to adjust 6 the numbers that you have, in terms of the flow through on 7 these fuel assemblies, if you wind up with any amount of 8 that stuff uncar a rack, and especially under a rack that has a freshly discharged fuel assembly. i 9 So, it does play 10 into it. 11 And, you know, if there are any -- when you're 12 looking for inputs tie into what might affect the heat up, 13 then I think you really need again -- and I made this point 14 earlier today, you really need to look at what's going on () 15 during decommissioning, what are the activities -- other 16 than the guys who are sitting there watching the monitors 17 and walking the perimeter of the spent fuel pool, what are i 18 the other activities that are going on around there. And I 19 think that you may find that there are other activities that 20 are going to affect whether or not that fuel is going to 21 heat up. 22 MR. CAMERON: I think we're going to have a 23 comment from Mike on one of the point you've made, at least.
)
24 Mike? l 25 MR. MEISNER: Yeah, Ray is exactly right. But, ' I (~N) ANN RILEY & ASSOCIATES, LTD.
\d Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
422 1 it's a bit of apples and oranges. Things do change when you I \ 2 go into decommissioning. U And recognizing that, I think all 3 decommissioning facilities make a clear distinction between 4 spent fuel management and the D&D work that's going on out 5 in the plant. 6 You don't have contractors coming in, who are 7 responsible for fuel management. In fact, what I think most 8 plants are doing now, and come on out and look, is we're 9 purposely putting in place what we call nuclear islands, so 10 that we clearly isolate the fuel management function, not 11 only procedurally, but physically, mechanically, 12 electrically, and everything else, from what's going on in 13 the decontamination effort. When you have asbestos workers 14 out there removing asbestos in the plant, surely, you've got
/~'T t
y ) 15 -- you don't have nuclear trained workers. But, that's, in 16 fact, why you isolate your pool and don't allow those folks 17 anywhere near it. 18 Another thing I think almost all decommissioning 19 plants do is they carefully refrain from doing any 20 decommissioning around their spent fuel. You know, the 21 spent fuel pool building at Maine, for instance, is about 22 the last thing that gets D&D'd. Once the fuels out of 23 there, then we move in and start doing the decontamination, 24 dismantlement work. So, while Ray's exactly right, that's 25 one of the things that decommissioning management has to be l T ANN RILEY & ASSOCIATES, LTD. s_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 l (202) 842-0034 I
v-423 1 on top of and identify ways to isolate the non-nuclear () 2 3 workero from fuel management. It's a real straightforward thing ~to do and you don't have -- in fact, all of those 4 folks don't -- can't even get access to the -- to our fuel 5 pool building. So, I think, in that sense', it's not really 6 a factor at all, what we're talking about. 7 MR. CAMERON: Thanks, Mike. Ray, do you have 8 anything? 9 MR. SHADIS: Just briefly. We had foreign 10 materials exclusions problems with the reactor cavity. I 11 mean, it's not like these things don't happen, don't get 12 ahead of us, and it may well be that Maine Yankee is an 13 exemplary plant; but, I'm raising this as a generic 14 question. () 15 MR. CAMERON: Thank you, very much, Ray. I think
]
16 we're probably ready to go to our wrap-up session. And we i l l 17 do have some mechanical things to do, right, Vonna, that you 18 need to take care of before we do that. I was going to ! 19 suggest originally that we just forge through. But, we l l 20 might as well take a break, since Vonna has to do some ' 21 things. Okay, so it's now 25 after 3:00. How about a 22 quarter of 4:00, is that too much time for everybody? 23 Okay, we've got a request for 20 to 4:00. That's l l 24 15 minutes. ' 25 [ Recess.] i l l ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I l 424 1 MR. CAMERON: Okay, al ' m all set? All right.
~.
( j 2 All right, we're going to start our wrap-up v 3 session now. And to give you a little preview of that, 4 we're going to start off with a few comments by Mike 5 Meisner. Then, we're going to go to Brian Sheron, one of 6 the top managers in our Office of Nuclear Reactor Regulation 7 for some comments. And then, we're going to have an open 8 discussion of what someone called the path forward, about 9 where we're going to go from here, to try to resolve this i 10 issue and how it might fit into future regulatory context, t 11 whatever. But, we have an immediate task before that. And 12 we did redraft the so-called requantification proposal. I l l l 13 think you have -- all have copies of it. But, I think we 14 need to revisit that, based on what Mike and Brian have to (j-15 say. , 16 So, let's start off with Mike Meisner. Mike? { l 17 MR. MEISNER: Thanks, Chip. I do have more than a j 18 few kind of -- I want to maybe ramble a bit, because it's 19 hard to pull all of this together without some preparation. 20 But, I want to start back at the beginning, after the l 21 Commission meeting and the first time we met with the staff, 22 to get a feel for what the technical working group was 23 coming up with. Remember, we met back in the ACRS room and 24 the staff gave us a presentation about their preliminary 25 results. And in their presentation were a series of
@ 4Ae *
(~ ) ANN RILEY & ASSOCIATES, LTD. l
\/ Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034 i l
425 K 1 sequences with'very high numbers in it. And we probed a lot
/~ -2 i
N)h in that. meeting, trying to understand what was behind it, 3 what drove those high numbers. 4 And I think it was the next day, then, we're in 5 talking to senior NRR management and tried to come up with a 6 success path forward, based on what the staff had done to 7 date. And what sn3 had agreed to, I believe, was this 8 workshop. And the -- what we brought up at the meeting 9 developing that workshop was the willingness on the part of ! 1 10 the industry to come up with commitments that would address 11 those things that seem to drive the high numbers in the 12 staff's preliminary results. And, of course, to do that, we 13 needed to have some information about how that model worked 14 and how it was put'together and what the assumptions were. () -15 So, we came prepared to do that at this workshop. 16 We have talked about a number of commitment areas, and I'm 17 going to go over those again, just to be clear of what we're i 18 talking about. And, you know, in return, we're looking for 19 some explicit reduction in risk that resided in the models 20 and that were tied to these areas of commitment. 21 So, first of all, and we may have lost some of it l ! 22 in passing quickly through it yesterday, when Ed Burns ; 23 talked on the risk insights. We were looking at four areas 24 of commitment -- hardware commitment: one that has to do l 25 with seals, to the extent they're included in the design, 1 O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r-426 1 and committing to the notion of either redundant or soft
/~T 2 loading seals; the second had to do with anti-siphoning l
3 devices associated with the spent fuel pool, both in the -- 4 I'll call it the permanent design, as well as when you use 5 temporary pumps and the like for one reason or another, that 6 you would incorporate the same kind of safeguards; and then, 7 finally, the idea of pre-staging or providing a connection 8 for the diesel fire pump to the spent fuel pool floor, so l 9 that that can be aligned outside of the refueling in the 10 middle of an event. I didn't want those things to get lost 11 in the shuffle. And I want to be able to understand that, 12 to the extent those provide a risk reduction, those are 13 areas the industry is willing to commit to. l 14 Then, we had a longer discussion about -- I'll () 15 call it the software, the procedural aspects of things. And 16 we're looking at things like -- things that really are 17 already required to be in place for, say, internal fires, 18 work controls, combustible controls and the like. I think 19 you can probably verify through our license and design bases 20 that those things are already part of our FSAR. And then to 21 the extent that it's not included in the SSAR, for detection 22 purposes, operator and security rounds, as an example, and 23 various detection devices. And, of course, central to all 24 of this is if you should ever get into a situation where 25 you've lost cooling or you're in a draindown situation, then O
- ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
427 l 1 the idea of clear procedures in place that will identify the 2~ various water sources, means of getting water into the pool, , '3 and power sources that are available to the operation staff, 4 to mitigate that event. 5 Now, we don't think there's any benefit in going ! 6 into detail, like we heard, for instance, when level goes 7 down two inches, you do this; when it's down one foot, you ; 8 do that, whatever. I think what we're talking about is l 9 insuring, as they already are, that operators are aware of l 10 all the potential paths for recovery, that they have a 11 procedure that directs them to not focus on a single path, 12 and that they have some training and ability to respond to 13 these very long-lived, long-term events. 14 There was also some discussion about training. (
) 15 But, I think in that area, clearly, there's no need for 16 additional commitments, at least that I'm aware of, because 17 that's part of Appendix B and our license basis in what 18 we're required to do, given the NRC approved certified fuel 19 handler training program. I think there shouldn't be any 20 concern about lower quality training, for instance, than we l
21 had when we were operating. The quality remains the same 22 and it has to remain the same. And there are a number of 23 regulatory hooks, I'll call them, that you have to come in 24 and if you're n t comfortable with the quality of training, 25 make sure it gets back up to snuff. But, the regulatory i ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l l [
428 1 framework, we believe, is already there. 2 And if ve missed some, somebody can correct me 3 later, but that was kind of a general discussion of those 4 areas that we're willing to provide additional commitment. 5 And the appropriate place, seems to me, to be incorporate 6 that into our license basis through the FSAR. That insures 7 that you've got all the controls associated with 50.59 and 8 50.90, should there be, for instance, an unreviewed safety 9 question and us wanting to change some of that. 10 We kind of throughout the workshop, asked staff 11 for feedback on which ones of these commitments would make a 12 difference. And I guess I thought that was going to be the 13 major portion of the two-day discussion, and I believe we've 14 gotten back very little feedback in that area. So, I don't () 15 know the degree to which any of these things could be 16 credited or the level of possible risk reduction that could 17 be achieved. What we did get from the staff was, you know, 18 that you'll look at it; you'll take it back and look at it. 19 Heavy loads was a bit of a concern. We thought 20 that would be the seminal issue that would be real easy to ! 21 identify; that, in fact, licensees have commitments and 22 requirements in place and have always had them there, and 23 there should have been -- we've had a pretty direct 24 reduction in risk, as a result. I'll mention two that -- 25 although it wasn't emphasized too much by Bill Henries, our O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 >
429 1 position on heavy loads, cask drops, is that the NRC has
!r 2 already resolved that issue in 1353.
( It was -- we feel it 3 was inappropriate to open again. The 1353 resolution for 4 cask drops makes absolutely no distinction between operating 5 and decommissioning plants. And when you're talking about 6 the probabilities of those cask drops, given that the 0612 7 requirements are maintained constant, that, in fact, we 8 think you should have just adopted the 1353 results, or, at 9 a minimum, taken the assumptions in 1353 that credited 10 magnitude risk reduction associated with 0612 11 implementation. 12 As far as the -- what we call the HRA 13 requantification issue, let me talk about. First of all, 14 because I think it was misunderstood, I want to clarify a () 15 statement that Ed Burns made, when he was talking about this 16 proposal, and I've spoken to Ed about this. Is he still 17 here? And Ed can speak for himself, if I misstate it. But, l 18 it's -- it will be real nice for the PRA community to maybe l 19 understand better how to deal with human error events and 20 maybe develop new techniques to do that. That's something 21 really good and probably something we should all pursue, as 22 a result of this. 23 The point is, though, that it's not needed to deal l 24 with the issues that we've been talking about the last two 25 days. We -- in fact, it's absolutely unnecessary. We I O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Aver.de, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034
l 430 i 1 believe the ground rules coming into this was for the staff 7,
/ 2 v
) to take our areas of commitment and adjust their assumptions 3 in the model, which we've told you, time and time again, is 4 a very good model, and to then requantify the model. We 5 don't believe it's at all necessary to embark on a new 6 research project, to advance the state of art of HRA for the 7 PRA community.
8 And related to that is -- buried in the ' 9 requantification issue is the fact that, in fact, we may not 10 requantify a lot of stuff pursuing this approach. The I 11 industry feels strongly that it's absolutely necessary to 12 requantify the staff's model. It doesn't serve anyone to 13 take a sequence and say, well, it's less than ten to the 14 minus fifth or ten to the minus sixth. We need to show some (7-v
) 15 continuity with the draft report that's already out and 16 follow through and requantify it. And the reason is we 17 really do need to risk inform this. We need to know where 18 to apply our resources. And if we're talking about a 19 situation where the probabilities are in the order or ten to 20 the minus seven, ten to the minus six, then it's not good 21 enough in prioritizing our resources to say that a 22 particular sequence is less than ten to the minus six. We, 23 also, believe it was the intent of the Commission to do just 24 this and to really face up to the issues and resolve them 25 once and for all.
[^') ANN RILEY & A$EOCIATES, LTD.* (_ / Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
431 1 I think -- you know, being cautiously optimistic,
'\
I think we made a lot of progress the last two days, but we (b 2 3 don't have anything in our hand to show that. You know, 4 we've got -- we've got, I think, a lot of good intentions to 5 take information back and revisit a lot of issues. But, I l don't think that we got what we expected to get out of the l 6 ' 7 workshop, and that was much clearer commitments on 8 straightforward issues, for instance, like heavy loads, that 9 we felt were achievable in the workshop. So, we're kind of 10 looking for something early on, to give us continuing 11 confidence of the fact we're making progress. So, I have a 12 series of proposals here that I want to lay out, from the 13 industry point of view, 14 First of all, we think it important that the staff l
) 15 provide, in a very short period of time, a detailed 16 near-term plan: what's going to happen in the next one, 17 two, three, and four months; how are we going to resolve 18 these outstanding issues, like in HRA; how are we going to 19 show that there's some credit for heavy load issues; how do 20 we know that we're really bridging the gap on seismic. And 21 we hope that that schedule will include all of these 22 different discussion topics that we've had in the workshop.
23 We think it would be constructive to have periodic 24 management meetings, then, between the industry and the 25 staff. We've done this a lot on larger issue type things, 1 l l o, .- - - a
/N ANN RILEY & ASSOCIATES, LTD.
(,,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
lis I 432 1 where we took, you know, some NEI industry executives that 2 met with NRR executives on a monthly basis. And I think 3 those meetings could then take a look at this schedule, l l 4 track progress, and see if we're tsully getting to where 5 we'd all like to be. l 6 On the HRA issue, I repeat our position that we { 7 feel it's absolutely not necessary and that we came in with 8 the understanding that the model needs to be requantified, 9 based on changed assumptions from the workshop. On the 10 other hand, the biggest concern we have with this is turning 11 it into a research project that's open ended, and that you 12 all are already talking about March of next year to complete l 13 the report, and that's pushing that out even further, and l 14 rulemaking, as a result. The industry is more than willing 15 to participate in this kind of activity, if we can do it on 1 16 a very early start, and actually work through things quickly 17 and get some resolution. We would be happy to provide our 18 experts to participate and look at HRA, maybe a new way to 19 do it, as long as it -- as long as we get some early 20 indication that we're actually being successful. 21 As you know, we've submitted a letter to the 22 Commission yesterday, asking them to delay their vote, which 23 is due today on the staff SECY. We've asked them to delay 24 it for 30 days. And the reason we did that is we wanted to 25 have a good feel about how positive or negative this r t O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202).842-0034
433 1 workshop was going to be, .because once the plan is approved, 2 it goes completely out of the staff's draft report and this 3 workshop. I think the jury is still out, so we're looking 4 for some positive feedback and getting a face-to-face 5 meeting on HRA, say, in the next two to three weeks. And 6 working our way through these issues to some level of l 7 resolution would give us that kind of certainty that we need j 8 to respond to the Commission before that 30-day period is 9 up. 10 On heavy loads, again, we're looking for something l 11 in hand, some actual final decision. We still believe that 12 that's an easy issue to deal with and we would ask the 13 staff, you know, to maybe even as a show of good faith, to 14 get back to us in a couple of weeks and give us the details () 15 of the credit that they intend to give for not only 0612 16 implementation. but those areas of upper bound concerns that 17 Bill Henries talked about, as well. l 18 On seismic, I am optimistic there. The industry l 19 has an action to propose a screening checklist, like Goutam 20 suggested, and Bill Henries has talked to Goutam and will l 21 work with him. We hope to have a first cut of that to you 22 in a couple, three weeks tops. 23 I've heard, maybe more inside discussions than l 12 4 anything else, that perhaps there are some resource 25 restraints, as far as, you know, having to risk inform I h O, ANN RILEY & ASSOCIATES, LTD. Court Re} rters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
434 1 decommissioning, as well as doing other risk inform work for ( 2 the Commission and the staff. I'd like to offer that if the 3 staff will provide us with your model, we will be happy to 4 do the leg work of requantifying that on whatever inputs you 5 give us. And we, as we said at the beginning of the 6 workshop, are going to provide you a report anyway that 7 talks about the substance of what Ed Burns provided. He's 8 already done requantification, to the extent that we can, in 9 understanding the model up to this point. And even if you 10 don't want to take advantage of the offer, I think it will 11 still be nice to give us the model, so we can provide a more 12 accurate response to the -- or more accurate report on what 13 we think is an appropriate requantification. 14 And along those lines, in general, whether it's a O 15 seismic or heavy loads or HRA or whatever, I think we might ( / 16 all benefit from identifying what you might call a tiger 17 team or something to -- composed of the industry and the NRC 18 and any other stakeholders that are interested. Let's 19 attack these problems and run them to ground right away. I 20 don't think it serves anybody to let it drag out. We'd all 21 hate to get to the point six months down the road, where 22 we've not converged, but, in fact, diverged on our support 23 for what's in the staff's final report. l l 24 So, in summary, we feel that we've provided 25 sufficient information, in our minds, to provide a more i ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 j i
435 L 1 accurate assessment of the risk presented in the staff's () 2 3 draft report. And we think that that risk, or those probabilities anyway, are on the order of a couple of orders 4 of magnitude below what the draft report says today. We 5 think that the staff model is good. We think it's real 6 good, as a matter of fact. But, as we indicated up front, 7 that the assumptions, themselves, have some severe flaws i ! 8 associated with it. But, you know, nonetheless, rather than
- 9 waiting for going out to peer groups or independent review 10 or something, to get it corrected, that's why we came into 11 the workshop with these commitments, so that the staff can l 12 feel comfortable in modifying the assumptions that make up i 13 the model. And we think that's fairly straightforward to do 1
14 and achievable. ()- 15 We will reluctantly support an HRA approach, if we 16 can do that quickly; but,.we really believe that all that 17 needs to be done is to go back and understand what we're 18 willing to commit to and then plug that into the existing
'19- model. And we would hope that we could continue to work
( 20 with the staff and very closely, to get some early comfort 21 level. What seems to be the positive strides we've made in 22 this workshop are really there and we keep achieving those 23 as quickly as we can, to provide a real risk informed basis 24 for decommissioning. 25 Thank you. s-ANN RILEY & ASSOCIATES, LTD. O- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 l i 436 l 1 MR. CAMERON: Okay. Thank you, Mike, and thanks () 2 3 for some detailed proposals here. now is hear from Brian Sheron. What we're going to do 4 MR. HOLAHAN: I'd like to hear some more public 5 input, before we go to the staff. > 6 MR. CAMERON: I think we -- I think it might be 7 better to hear from Brian and then let's get input -- and B Brian, you be the judge of this -- and then to go into a 9 discussion on these issues. I mean, we can use the industry 10 proposals to frame the discussion. But, you might want to 11 offer some general thoughts first. But, we can go into a 12 general discussion of this, if you'd like. 13 MR. SHERON: No. Let me -- yeah, let me, I guess, 14 give my perspective. I did not have the benefit of being 15 here yesterday afternoon or early this morning. I do want 16 to thank everyone, the industry, their stakeholders, for I 17 participating. I guess I had a little different perspective 18 cn this than Mr. Meisner did. First off, is that, you know, l l 19 having been in the agency for 23 years, the way we operate ! l 20 is not to make instant decisions at meetings like this. We l 21 never had and I doubt if we ever will. 22 We have just received the industry information at 23 this meeting, I think, or I think shortly before. It's not 24 a matter of somebody coming in and giving us numbers and the 25 staff going, yeah, you're right; we'll fix everything. We ANN RILEY & ASSOCIATES, LTD. ['\--)' Court Reporters , 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034 1 I
437 1 1 1 need to look at it. We need to test it, make sure that it ' l I^) G 2 can withstand technical scrutiny. At the point when we are 3 -- we can accept it, then, yes, we can, obviously, use it. 4 My understanding of going into this meeting, from 5 the previous -- when we had the meeting down in Rockville, I 6 in the ACRS meeting room, was that this was going to be an 7 information gathering meeting. We provided the draft report 8 that we had at the time. The idea was that tnis meeting was 1 9 for the industry to provide us information, where they felt, 10 for example, some of our numbers or the like were wronq or 11 they had other numbers, as well as other stakeholders, .f 12 they had comments. The way we operate is we take all this 13 information, we go back, and we will look it over. And to 14 the extent that, as I said before, that we feel it's valid f~~ l t, 15 and defensible, we certainly will use it. In areas where I 16 think there are still question marks, I will propose that we 17 need to pursue further how to resolve what the differences 18 are. 19 I guess, basically, I don't share the optimism, in 20 the sense that the industry has -- you know, all of your 21 numbers are right and all we have to do is stick them in. 22 We perform an independent assessment. Obviously, we 23 disagreed with the industry in the past in areas and there's j 24 nothing wrong with reasonable people disagreeing. l 25 Nonetheless, I think a lot of progress was made,
) ANN RILEY & ASSOCIATES, LTD.
sm/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034
l 1 438 l l 1 in my mind, at the meeting. We got a lot of information. i l
' l 2 We learned where some of the hard spots are. /'^]N L We learned 3 where there may be areas of where we do agree. I think that 4 basically what we need to do now, as Mike said, is figure j 5 out how we go forward, to try to either resolve these issues 1
6 or agree to disagree on certain issues. The areas I think 7 you've pointed out -- I do want to make one observation. I 8 kind of picked it up when I was sitting in the back there, l 9 and that is that when the NRC resolves a generic issue, such 10 as heavy loads, in the past, you need to understand what i 11 that resolution means, in generic issue space, is that the l 12 staff assessed it and either could or could not support some 13 sort of a backfit. And if there was a backfit that was j 14 required, typically what that does is it reduces the risk to (~~% (v ) 15 an acceptable level, based on a cost benefit analysis, if ' 16 it's not a compliance issue, for example. It doesn't make 17 the risk go to zero. And in my mind, I guess, that the fact 18 that we may have resolved some generic issues in the past 19 doesn't mean the risk is zero from them. 20 And when we're looking at a risk informed 21 approach, we're looking at an integral here. We're not just 22 looking at heavy loads, we're looking at seismic, we're 23 looking at a whole bunch of other things. And you've got to 24 add them all up and see what the integral risk is. And so l 25 the fact that one issue may have been resolved doesn't mean ( ('S ANN RILEY & ASSOCIATES, LTD.
\m ,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1
l t
439
- 1. that the risk is zero. And unless it's something where you 2 might say the risk was so low that it's down in that 3 truncation error with the PRA, we just can't say that 4 because it was solved, we don't have to consider it. We 5 consider it. If the risk is indeed low, then I think the 6 PRA, itself, would probably show that it's a no, never mind.
7 But, it's not something that we can a priori right off and 8 say the staff resolved this in 1985 with some generic issue; 9 therefore, we don't even consider it. We still need to look 10 at it, and that's true for all the issues. 11 With that, I think I don't agree that we're going 12 to turn HRA into a research project. I'm not sure where 13 that perception came from. While there may be more 14 confirmatory work needed, that's NRC's business whether we () 15 16-do that or not. Our plan right now is to try and figure out a way to gather the experts, the HRA experts, in a'short 17 period of time, within 30 days or so, a couple of veeks, as 18 fast as we can get them, you know, in the same room at the
~19 same time, to address this -- that issue. And, hopefully, 20 they will achieve some sort of resolution, what the right 21 numbers are and the like. Our plan right now is to still 22 complete this report by the end of the year, as a draft --
23 as a final draft. 24 I think that where we need to go now is to 25 identify where are the other hard spots and how do we -- how I
- p.
O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 4 I Washington, D.C. 20036 (202) 842-0034 1
440 1 do we deal with them. Do we convene, for example, a group n f i 2 of experts on thermohydraulics, if there are V 3 thermohydraulics issues? We need to define what the 4 specific issues are, what the specific disagreements might 5 be, and then convene the right group of experts, to sit down 6 and try and either come to a resolution and agreement, or to 7 at least identify where the disagreements are. 8 But, you know, in closing, I want to leave on a 9 positive note, that I think this was a very valuable 10 workshop. I think it produced a lot of information, a lot 11 of understanding from both sides, where each side is coming 12 from -- I should say both -- all sides, where everyone is 13 coming from, what the concerns are and the like. And I 14 think we should use it as a stepping stone to identify the () 15 specific issues that need to be resolved and go forward from 16 there. And I think the next steps in the meeting would be 17 to try to identify those and try to identify what are -- 18 what is the right mechanism to go and address those. 19 I guess that's really all I had to say, Chip. 20 MR. CAMERON: Thank you, Brian. I'd like to give 21 Ray and Peter a chance to talk here and, again, along the 22 same vane. I mean, we've been talking about process, okay, 23 and how we move forward. And, of course, Mike put several 24 proposals on the table. I think that Brian has already -- I 25 think, generally, this proposal of requantification, I think (' ' ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
f 441 1 we're gens cally in the same ballpark on how to -- how to 2 handle that, it seems. And we can have -- we can -- people ( 3 can agree of disagree on that. But, that's one thing that I 4 heard Brian say already that seemed to match. 5 Let me ask -- let me ask Ray for any comments that 6 he has, at this time, or Peter, in terms of these process 7 types issues of how we move forward. Ray? 8 MR. SHADIS: Thank you. There's been a good deal l 9 made of what the Commissioner's expectations are, in this 10 process. But, I don't find anything written anywhere, as to ! 11 what their expectations are. If there's a memo or letter or 12 whatever, however you folks communicate, I'd love to see it, 13 to see what the Commissioner's expectations are. I have the 14 feeling that what they've done is they sort of ordered up an l O ( ,7 15 underwater blimp, difficult, difficult project. l 16 And the realities are that maybe the I 17 Commissioners, themselves, need to take another look at 18 what's happening here. Because, this is a construct that is i 19 -- that is out of context. You know, we spoke earlier about j l 20 the narrowed focus of this effort and suggested that there l 21 are a lot of other areas that are as important, if not 22 because of their consequences, because of their l l 23 probabilities, things you're more likely to run into and 24 should be dealt with. But what it really comes down to is 25 trying to get rid of some of the exe:nptions and it really is ANN RILEY & ASSOCIATES, LTD. Os Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
i 442 1 only a few of those. 2 So, I think -- I think what the situation is, is
}
3 that the regulations and regulators did not anticipate the 4 decommissioning mode of life and we've been trying to, you 5 know, hammer regulations intended for operating plants into 6 a fit for decommissioning. And, you know, if I had an 7 opportunity to meet with management on an individual basis, 8 you know, what I would want to say is maybe you really ought 9 to take a look at the project, potentially deciding to can 10 this project altogether and open up the notion that there 11 really needs to be a special categorization in regulation 12 for decommissioned plants, starting with the data -- fuel. 13 That's our point of view out in the field. We're 14 very, very confused by the way that Part 50 is being applied () 15 in places where Part 72 might be more applicable. So, 16 that's our perspective after all. It's been a series of 17 meetings and it's been, you know, highly educational. I'm 18 sure there's been information that's been exchanged. But, 19 maybe, you're going down the wrong track altogether. 20 MR. CAMERON: Okay. I've noted your comment. 21 We're going to get a response from Stu. 22 MR. RICHARDS: There is communication from the 23 Commission. I don't remember, Ray, if you were at the March 24 17th meeting with the Commission or not, but we sat and told 25 the Commission what we intended to do. I think we're (O
\s,)
ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
443 1 following out that mandate. You can argue about the ( 2 details, whether we're doing it fast enough or whether we're l 3 doing it with the right groups of people. But, nonetheless, l 4 we told them that we'd go look at the risk and we'd go look 5 at the process. I can't speak for the Commissioners, but if 6 you read the transcript, I don't remember them having any 7 . objection to us doing that. 8 They did ask us to get back to them with a plan to l 9 speak of and we did that, with a Commission paper, which we 10 made available to the public, that lays it out, and it's 11 under review by the Commission now. They, also, have 12 responded to us on previous Commission papers. The 13 so-called staff requirement memorandum, I believe those are 14 also made available to the public. And to the degree they () 15 have communicated to us, they've encouraged us to look at 16 this integrated approach. i 17 I'm a little bit vexed by your comment about your ! l 18 .being confused on the regulations and you think we need to i l 19 go to a -- you know, a new regulation framework, so all the j 20 decommissioning regulations are in one place. Because, 21 that's exactly what our Commission paper proposes. We've l 22 told people in a couple of meetings before that's what we l l 23- were proposing to do. So, I think we're doing what you say 24 we need to be doing and, yet, you don't seem to acknowledge 25 that. 1 ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
444 1 MR. SHADIS: If I can clarify it for you. It is 2 segmented and, ,au know, we're now engaged in something 3 that's pretty narrowly focused, as opposed to addressing 4 that whole issue, as a package. That's the difference in 5 what-you're saying and what I'm saying, I think. 6 Let me ask if -- just to -- maybe you can put my 7 concern away: is it your perception that the Commissioners 8 would be disappointed at t.he pace and schedule of this 9 process? 10 MR. RICHARDS: I don't think it's proper for us to 11 speculate on the Commission's pleasure or displeasure. 12 MR. SHADIS: Well, I mean, given what -- given 13 what instructions you've been given, you've been told that 14 this is supposed to be a fast track, fast trained () 15 production, and -- or, you know, how are we -- how are we 16 approaching this? I don't understand -- 17 MR. RICHARDS: I think the answer is, is that we 18 -- you know, we told the Commission we'd do this in a 19 relatively fast manner at our March 17th meeting. Perhaps, i 20 we can get you the transcript of that. Again, I don't ! 21 remember anybody objecting. But, at the same time, I'm sure 22 that the Commission expects that the staff produce a quality 1 23 technical product. So, we're not meeting some time frame i 24 just to meet a time frame. We're going to do the job, to 25 the degree that we're satisfied with the technical product, I ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
F 4 445 1 whatever that takes. [d \ 2 MR. SHADIS: I appreciate that. I mean, I'm 3 reflecting what I heard from this side of the table earlier, 4 in terms of what Commission expectations are, and I just i 5 want to have that clear. I' 6 MR. RICHARDS: I want to make sure you understand 7 l that this meeting, the last two days, is to address the risk 8 question. There's other efforts going on. I'll say again, 9 it's spelled out in the Commission paper, that we have a 10 separate effort, looking at the framework, the rulemaking, 11 all that is going on in parallel, and is simply not the 12 topic of this workshop. 13 MR. CAMERON: I think this will be -- this issue 14 will become clearer, in terms of the Commission's response () 15 to the paper that was sent up. Peter James Atherton, do you l 16 have any process points to say, at this juncture, or do you 17 want to wait until we get into discussion? 18 MR. ZWOLINSKI: Excuse me, Chip. 19 MR. CAMERON: Yes. 20 MR. ZWOLINSKI: May I add to Stu's comments -- 21 MR. CAMERON: All right. 22 MR. ZWOLINSKI: -- for context? We proposed to 23 the Commission in the SECY paper some time frame. And it's l 24 our expectation that the Commission will come back to the 1 l 25 staff with what's called a staff requirements memorandum, f~] x_/ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1
l l l l 146 1 and they'll either what we've provided or provide different () 2 direction to us. So, their response back to us is rather 3 important, as to are we fulfilling their expectations of us. 4 So, right now, we don't know, and this is the vote that Mr. 5 Meisner alluded to a little bit. That paper was a proposal 6 from the staff that had the larger framework of a game plan. 7 One piece of that, of course, relates to the technical 8 issues that are being today. But, we were talking in macro, 9 as far as the overall game plan in that Commission paper. 10 MR. SHADIS: Okay. I didn't know. I have a copy 11 of the transcript. I have a' copy of the SECY paper. I 12 didn't know if there was some other, you know, written 13 statement regarding the Commissioners attitude toward this. 14 So, I'm pleased to have that -- to have that cleared up. A) ( 15 MR. CAMERON: Okay. j 16 MR. ZWOLINSKI: Very good. 17 MR. CAMERON: Thank you. Thank you, very much, 18 both of you. Peter, do you have anything to add, at this 19 point? 20 MR. ATHERTON: If I could, please. Over the last i 21 several meetings, I've raised a number of concerns. I've 22 tried to keep them on a generic basis. And I haven't had 23 the opportunity to attend this full two-day meeting. 24 However, I -- it is not obvious to me that these concerns 25 are being looked at. And so, I would like raise them, at (~ ANN RILEY & ASSOCIATES, LTD.
\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
447 1 this point'in time, again, and I'd like to begin with what I () 2 talked about concerning the seismic issue, and that is the 3- aging requirements and the qualification requirements for 4 the spent fuel pools concrete and liner and the racks. This 5 issue was raised by Mr. Kennedy and we discussed it earlier 6 today. 7 ~Per your suggestion, Mr. Cameron, I spoke to him 8 out of -- off the record, in order to try to identify the 9 possibility that there were concerns, which needed to be 10 addressed. And Mr. -- I do not like to paraphrase somebody 11 else's comments off the record, but let me say that there is 12 a concern with the aging requirements upon concrete, 13 especially when it comes to temperature. Concrete does not 14 cure continuously'over a hundred year period. It does cure ,( ) 15 over several years, at the most. And as a result of that,
- 16. the -- what is NRC doing to take into account the 17 degradation of this spent fuel pool with time, in developing 18 it's probabilistic numbers? And I would ask NRC to take this l 19 into account.
20 MR. CAMERON: Okay. We -- I'm going to note that. 21 Do you have some more? 22 MR. ATHERTON: I have; if I could, please. My 23 concern, when I lend my services as, which are in the 24 nuclear engineering and electrical engineering end of the 25 line, to the citizens, who seek that type of help, on ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 448 1 whether or not there is a danger of radiation to the ()
<~
2 environment. And I tried to work towards the radihtion 3 danger from the environmental point of view. I have raised 4 the issue of criticality, which was not a subject of this 5 session today, and I'm in the process of communicating with 6 a member of NRC about that very issue. It doesn't appear to 7 be a major concern with anyone, at this point in time, 8 except that I'm trying to find out why that is the case and 9 I'm looking for written sources of information, which make 10 criticality in the spent fuel pool a non-issue. And I'm 11 working on that separately. 12 The scenario of approaching the accident analysiF 13 and the consequences or the failure modes and affects 14 analysis, as it used to be in my day in the 1970s, when I () l 15 worked for the Commission, took into account first and 16 foremost the worst case criteria. And it appears that one 17 of the worst case scenarios, in this case, would be draining 18 of the spent fuel pool, and despite some people's attempts 19 to keep -- to keep the spent fuel pool covered with water. 20 The consequences to the public of the worst case accident 21 that could occur, as a result of that, needs to be addresst' 22 from a probabilistic perspective, taking into account what 23 the causal factor is and taking into account appropriate 24 probability numbers that are more realistic. 25 It would seem to me, from a purely logical ANN RILEY & ASSOCIATES, LTD. (V) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l 449 1 perspective, that in order to make probabilistic numbers 2 (
) real, it would help to know what, for instance, the 3 probability was or is, depending upon how many different 4 calculations ycu would perform, at the Three Mile Island 5 nuclear power plant partial core meltdown in 1979 would have 6 occurred. I have not spoken to some people and I have not 7 seen any number or series of different numbers, which tell 8 me what that probability is. Now, that accident did, in 9 fact, happen and that probability would be a real life 10 number that we could use as some sort of a basis to 11 determine what probabilistic cutoff point would be for 12 something that would be negligible and didn't have to b<
13 looked at. The number ten to the minus six was put forth 14 for the first time, as I learned it in the Rasmussen study f-( j) 15 and is being bounced around here today. But the Rasmussen 16 study in the mid 1970s was before the Three Mile Island 17 accident. 18 We've also had several other accidents, nowhere 19 near as serious in this country, and we could also determine 20 what the probabilistic numbers would be for that type of an 21 accident. We would then have probabilities of real time 22 accidents, which we could then use as some sort of a basis 23 to set a cutoff point. I haven't -- I haven't seen a basis 24 for the ten to the minus six number. And so, I question why 25 we're accepting that, or we seem to be accepting that, and I ANN RILEY & ASSOCIATES, LTD. O- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
450 1 have been informed that NRS is not set on that number, ten
) 2 to the minus six, if I interpreted Mr. Kelly correctly.
3 MR. CAMERON: Okay. There's some -- because, we , l 4 do -- I'm not trying to minimize your concerns. In fact, 5 we're putting them up there and I think that these are going 6 to be addressed. But, I really do want to make sure that we 7 get back to these process issues, in terms of the path 8 forward here. So, if you have some other issues, could you 9 just give them to us quickly and I'll put them up there. 10 MR. ATHERTON: I am concerned that nobody else 11 seems to be concerned about whether or not the NRC would 12 accept a total blackout condition, as an acceptable 13 scenario, considering any kind of an accident, without 14 having some sort of a process whereby you could get at least 15 some lights made available in the spent fuel pool area. 16 The philosophy behind that, without looking into 17 the numbers, has me -- wa have, for instance, station 18 blackout condition requirements specified in the code. The 19 instrumentation and control branches, when I used to work at 20 the NRC, would not permit, without safety sy' tem s backup, any 21 sort'of power loss requirement in a safety-related area. 22 They didn't have jurisdiction over the spent fuel 23 pool at that time, and it seems to me that you would want to 24 have -- you would want to be able to see what is happening 25 in the spent fuel pool in the event of any postulated i l ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
451 1 accident. 2 MR. CAMERON: Okay, ( I think that what I would 3 like to do, Peter, is have Rich Barrett one constructive 4 word of response to you, and then I want to get on to this 5 process issue again. 6 Rich. 7 MR. BARRETT: Peter, I have been listening very 8 carefully and taking notes here, and I know that, as you 9 said, you are dealing with other members of the staff 10 regarding criticality. But the issues that you have raised, 11 you raised after the criticality issue. The issues related 12 to worst case, draining of the pool, where the 10 to the 13 minus 6 number comes from the relevancy of the TMI, the 14 probability estimates, and the question of whether or not 1 (Qy 15 the NRC would accept a blackout condition, I think we have a 16 lot of discussion in the report on that issue, and we had a 17 fair bit of discussion yesterday that was relevant to that 18 issue. 19 And I think what might be useful is, you know, 20 after this session, if we could spend a little bit of time, 21 I could certainly spend some time with you and at least give 22 the NRC's perspectives on those -- current perspectives on 23 those questions and how those questions relate to what we 24 are proposing to do as next steps, if that would be useful 25 to you. i [~}
\_/
ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034 l l 1 I
f l L l l 452 1 MR. ATHERTON: Whatever you suggest would be I \ 2 helpful to me. V ! 3 MR. CAMERON: Okay. Thank you very much. And l l 4 thanks for putting those concerns on the table, Peter. l 5 I guess I would like to find out now, so that -- 6 because I think that it is fair when there is a proposal on j 7 the table to find out-what the perspectives on that might i 8 ce. We have some proposals that Mike made. Can Gary or l l r 9 Brian, John,~the NRC staff, what do you think of these
'10 proposals?
11 And I have already tried to, and maybe wrongly, 12 say that there was some convergence between Brian and Mike 13 on looking into quantifying -- requantifying the HRA aspects l 14 of the model. There may be a difference in the time () 15 schedule, though. 16 Gary, can I go to you for this? 17 MR. HOLAHAN: Yeah. I took notes and I marked l 18 mine as to how'I thought we ought to proceed. I will turn 19 back and forth to see whether yours are lined up with mine.
-20 With respect to near-term step-by-step plan and 21 schedule, I think that is a reasonable thing to do. We 22 normally do these sorts of things. I think what you are l 23 suggesting is you want to see, you know, that sort of thing.
24 We don't always publish that level of detailed plans. But l 25 since this is a rather interactive exercise with l l O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034 ( l-
1 453 1 stakeholders, I think for planning purposes, I think that is A t i 2 fine, we should -- we will go ahead and do that. G l 3 I captured a separate issue that said you were 4 thinking of periodic meetings with management and I think 5 you ought to take that up with Brian or Sam, because I don't 6 think you mean me. So, it is okay with me if you meet with 7 them. And whatever -- 8 MR. SHERON: I don't think we have any problem 9 with periodic meetings. 1 10 MR. HOLAHAN: Yeah. If those are going to be ' 11 technical meetings, then I presume stakeholders and notices 12 and all of that ought to be -- the normal process will be 13 followed. l 14 MR. CAMERON: HRA aspects. I
) 15 MR. HOLAHAN: No, heavy loads. Are you going to 16 -- okay. Well, okay, let me go to the HRA.
17 What I heard was Mr. Meisner say that the industry 18 was reluctantly willing to support the HRA effort. It is 19 not what he wants, but he is willing to do it. 20 MR. MEISNER: Given that it is done in a short 1 21 period of time. Although we still believe what you need to ! 22 do is just go back and change the assumptions in your 23 current model. 24 MR. HOLAHAN: Yes, that is my understanding. And 25 I think I would propose, therefore, to go ahead and do what I'T ANN RILEY & ASSOCIATES, LTD.
\m Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1
454 l l 1 we proposed on this stage, which includes attempt --
<g (V < 2 MR. CAMERON: Put there up on there, please.
l 3 MR. HOLAHAN: Yeah. 4 MR. CAMERON: All right. Here, give everybody a 5 chance. You had a handout. Take a look at this. The only 6 difference might be here the timeframe in which it is done. 7 MR. HOLAHAN: And there is a timeframe on the 8 bottom, basically saying we would like to take whatever 9 requantification or re-evaluation of the non-PRA issues, any { 10 other input like thermal-hydraulic analysis, information 11 that the -- particularly about the licensees, about their 12 programs and how they run the plants, and provide that to a 13 collective of HRA experts, which we will take the initiative 14 to bring together, and to the extent that it is possible and r (,) 15 pract.. cal, we will take the results from those experts and 16 as they advise us to requantify or to put bounds on given ! 17 values, we will go ahead and do that, and reflect that in ! 18 our final report. l 19 As it says, you know, we will make an attempt to 20 get that started, or as far as we can get within 30 days, 21 and as a goal, we would like it not to derail, you know, our 22 ongoing effort to get this report done. We would like it to 23 be done as an integral part and on the current schedule. If 24 that is possible, you know, we will attempt to do that. 25 MR. CAMERON: Let me ask -- let me go to Mike now. i ANN RILEY & ASSOCIATES, LTD. (/
\_
T Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 l l
455 1 Mike, your response to that, to the staff proposal. 2 MR. MEISNER: I think the key area we disagree is ( 3 in Number 4. We think requantification is feasible under 4 all conditions and that the staff's model in the draft 5 report needs to be requantified. 6 MR. CAMERON: But do you have any objection to 7 having a panel, a collection -- I guess I always wanted to 8 know, like a covey, I guess HRA experts are probably travel 9 and collections, but do you have any objection to -- I mean 10 it is sort of similar to yours, to convening experts to try 11 to talk about it. 12 MR. HOLAHAN: They travel in random patterns. 13 [ Laughter.) 14 MR. CAMERON: Random. () 15 MR. MEISNER: Sort of like Brownian motion. 16 MR. CAMERON: All right. 17 MR. MEISNER: No, and I think, you know, I am sure 18 the staff will make an effort to get truly independent 19 folks, objective people in there. We would like to 20 participate with our experts. 21 But, again, if this -- if the first meeting leads 22 to, geez, we need to make a long-term study of this, then we 23 miss the point. And we think any results need to be 24 factored back into the staff's draft model. 25 MR. CAMERON: Okay. So noted. I put a deviation O \m ,/ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
456 1 up here, HRA staff proposal acceptable. () 3 2 Gary. Any notes? MR. HOLAHAN: Yeah. ' You can say reluctantly, if 4 you like. But I think that is the path we are moving down. 5 MR. CAMERON: All right. 6 MR. HOLAHAN: Okay. Heavy loads, I understand 7 that you would like some final feedback on your proposal to 8 -- or early feedback, a few weeks, on coming to closure on 9 use of the, you know, work that was done under 0612 and, you i 10 know, whether we can close the issue on that basis. And I 11 think that is a fair request. We ought to be able to decide l 12 shortly whether that is the path we are going to go down, or 13 whether that is closed or not and give you feedback. 14 MR. MEISNER: Just to be clear, Gary, I am not () j 15 looking to close any issue. What we are looking for is the 1 1 16 explicit credit in the model, both for 0612 implementation 17 and the upper bound concerns that we proposed. So we would 18 like to see -- we would like to see the model requantified 19 for heavy loads. 20 MR. CAMERON: And is that what you are saying, 21 Gary? 22 MR. HOLAHAN: I am saying we can give him that 23 decision in a short period of time, yeah. I can't tell you 24 today what the numbers are going to be. I hear your input 25 and, you know, we are going to go back and look at 0612. ANN RILEY & ASSOCIATES, LTD. ((^) _/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 t 1 l
-s 457 1 MR. CLMCRON: Okay. I am going to put a caveat on 2 there then. 3 MR. HOLAHAN: The caveat is that Kelly wants to 4 comment. 5 MR. KELLY: The comment that I wanted to make, 6 because I think -- I actually got an opportunity to speak to 7 Mike at lunch time, and I wanted to make clear one of my 8 concerns about these numbers, which I haven't haa the chance 9 to communicate to Gary, and I just wanted to put everybody 10 on the same page here so we have an opportunity to 11 understand. 12 In performing the 1353 analysis, that analysis was 13- performed on the basis of looking at whether there was 14 enough risk such that on a -- there would be enough money () 15 there in order to perform a backfit analysis. Now, it 16 -turned out that the basic case came in at something like 10 l 17 to the minus 6 per year. It ended up being like $33,000 per l 18 man rem, which was way too high, and there was no way that l 19 at that point it was going to be worthwhile doing anything. ! 20 So if you gave a person three orders of magnitude, 21 one order of magnitude, or 22 orders-of magnitude credit for . l 22 doing 0612, it made no difference because if you said the 23 risk went to zero, there still wasn't enough money there to 24 do anything. So the fact that we came up with a number of 25 three orders of magnitude is not necessarily reflective of
/ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
458 1 the real worth of the -- as far as decades of reduction and (} 2 risk for heavy load. It might be four orders of magnitude, 3 but the values that we did there did not have to be done to 4 any great precision. The fact that we gave, you know, one 5 order or 22 orders of magnitude made no difference to the 6 final result. 7 So, therefore, it is going to be hard -- you know, 8 I don't know how quickly we can go back and look at that and 9 determine that number. And that is the only thing I wanted 10 to caution you about from my standpoint. 11 MR. CAMERON: But certainly that would factor into 12 -- the staff is going to take a look at that and come back 13 with a decision or we need the longer period of time, or 14 whatever. Is that correct? ! 1 15 MR. HOLAHAN: What I heard from Mr. Meisner was a [V) 16 desire for early feedback on how we were going to close that 17 issue, and I think he said within a few weeks. And, you 18 know, I still think we ought to be able to do that. I mean 19 I understand there are some questdens to go back and say, 20 why were the numbers picked the way they were? Yeah. But, 21 you know, we ought to be able to identify the closure path. 22 MR. CAMERON: So, it is not a decision in terms of 23 how much -- necessarily a decision on how much credit is 24 going to be given, but a decision as to how this will be 25 clcsed? (' T ANN RILEY & ASSOCIATES, LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
459 1 MR. HOLAHAN: Well, but the number might fall out
/ 2 immediately on that.
V) 3 If we look back and, in fact, we say, ch, now I understand why they did, and we are comfortable 4 with that answer, then the numbers are immediately 5 available. 6 MR. CAMERON: All right. 7 MR. HOLAHAN: Okay. And if they are not, then I 8 think all that is being asked for is to, you know, tell 9 people what our story is. l 10 MR. CAMERON: Okay. j 11 MR. HOLAHAN: Okay. I 12 MR. CAMERON: Proposed checklist for seismic. l 13 MR. HOLAHAN: Yes. I think we said this morning i 14 that that was a good idea. That would be a welcome proposal 15 as part of the addressed the seismic area. (O) 16 MR. CAMERON: Okay. And, Mike, if you have -- we 17 have one more issue, which is the tiger teams. Do you want 18 to wait till you hear -- 19 MR. HOLAHAN: No , I have two more issues. 20 MR. CAMERON: Oh, you have two more. 21 MR. HOLAHAN: I heard a reiteration that there 22 would be the PRA reports sent to us in a few weeks. I 23 think, in fact, we would like to get that. That would be 24 useful, which I think you referred to as the Ed Burns 25 report.
) ANN RILEY & ASSOCIATES, LTD.
s/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
460
\
1 Then I heard a request to have the PRA model 2 available so that you could manipulate it or in fact use as
)
3 a way of looking at the probability numbers. I don't see 4 any reason why it can't be made available. I don't know 5 whether there are any questions about compatibility between, 6 you know, what versions of the code, who is running which 7 sort of codes these days on which sort of machines. 8 I would just suggest that, you know, this is 9 public work that we are doing. Anyone who wants to do this 10 sort of thing, let's get together with Mr. Kelly and look at 11 the practicalities of, is this a matter of data on a disk to 12 hand back and forth? 13 You know, the structure and the numbers reem to me l 14 to already be in the report. Right. The models. () 15 The reality is that we have the fault trees on 16 NUPRA and the event trees are all hand-calculated. 1 MR. CHEOK: No , it is not. 18 MR. HOLAHAN: Well, did I get it wrong, Mike? 19 Okay. Mike did it. 20 MR. CAMERON: Mike, just identify yourself for the 21 stenographer. 22 MR. CHEOK: Mike Cheok from the staff. 23 MR. CAMERON: Thank you. 24 MR. CHEOK: The event trees and fault trees and 25 everything else is in the NURPA model. It is all quantified [] ANN RILEY & ASSOCIATES, LTD. '\_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
~~
E t i 461 1 using the model. What we can provide to you guys is a sets (~ v ) 2 compatible format that you can import to any code you want. 3 Or you can use the NUPRA input, if you like. 4 MR. HOLAHAN: Okay. l 5 MR. CAMERON: Okay. 6 MR. HOLAHAN: Okay. So anyone who is interested 7 in having that level of detail so they can manipulate the 8 model or do sensitivity studies or that sort of thing, just 1 l l 9 let us know and we will work that out. I might try it 10 myself just for fun. 11 Well, the data is not proprietary, okay. You 12 know, it may require someone to have one or another version 13 of a PRA code, okay. We are not going to give them a PRA 14 code, we will just provide them the data. () 15 Tiger teams, I thought was the next issue. 16 MR. CAMERON: Yeah, it is. 17 MR. HOLAHAN: I don't know whether I agree with 18 that or not, because I am not sure what it means. If it 19 means that we are going to work hard on this and have 20 contacts available so that we will meet periodically or 21 share information, that's fine. If it means that we are 22 going to send people off, industry and the staff, privately, 23 to resolve these issues, we don't do things that way. So, 24 in concept, focusing on the issues and getting them resolved 25 early, yes, we are interested in that. I am not sure ANN RILEY & ASSOCIATES, LTD. k',,, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
462 1 exactly how you envision that to work out. () 2 MR. CAMERON: Mike. 3 MR. MEISNER: I think it was closer to the former 4 that you mentioned, which implies some level of l 5 communication. i 6 MR. CAMERON: And it was inclusive of other 7 interests, too, as you stated. 8 MR. HOLAHAN: Yeah. I think, you know, we I think 9 have already identified or we can identify key staff who are 10 working on individual areas. You know, if we want to have, 11 for example, Mr. Atherton is interested in the criticality, 12 and we have set up, you know, meetings on that subject. You 13 know, we can go ahead and do that. Okay. We can have 14 targeted meetings on individual areas where -- you know, to () 15 resolve issues quickly or to identify what it needs to do 16 next. You know, it doesn't need, you know, for all the 17 management to get together and have a big workshop just to 18 go to the next issue. 19 MR. MEISNER: Right. Until we meet. 20 MR. HOLAHAN: Right. That's fine. Okay. 21 MR. CAMERON: So it is -- the understanding is 22 targeted meetings on specific issues. 23 MR. HOLAHAN: Yeah. l 24 MR. CAMERON: All right. 25 MR. HOLAHAN: And what I would expect is that we l l [~}
\ ,/
ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 ' (202) 842-0034 l
463 1 would notify, you know, various stakeholders. In fact, most 2 of us, I would think, wouldn't want to show up at the ( 3 technical meetings at this sort of level. We will just let
'4 the technical people do their job and report back on '5. whatever they found.
6 I think most of the other things I heard I-think 7 are already captured one place or another on their flip 8 chart from earlier. I am not sure what was said about the 9 issue of potential aging degradation in the seismic area. I 10 think that topic can be addressed in the report. You know, 11 whether or not we consider an important area that makes a 12' difference, there is no reason why it shouldn't be 13 addressed. 14 It can be discussed. The issue of concrete and () 15 its strength over time, and degradation mechanism, I think 16 can be addressed. I am inclined to commit to do that, even 17 though the people who would do it are not here. 18 Criticality, I think we have already made a 19 commitment to have discussions with Mr. Atherton, and I 20 think'his other issues, Dr. Barrett is going to speak to him 21 about. And maybe we should go back and look -- 22 MR. CAMERON: Yeah. Let's run through these. ) 23 That is a good idea, because there were a whole bunch of 24 things that I think we agreed to on our way here. i 25 The first points involve Citizens Awareness O ANN RILEY & ASSOCIATES, LTD. Court-Reporters 1025 Connecticut Avenue, NW, Suite 1014 ; j Washington, D.C. 20036 i; (202) 842-0034
t 464 1 Network, New England Coalition on Nuclear Pollution. We do i
) 2 have New England Coalition on Nuclear Pollution, but we will 3 make sure that on this list is Citizens Awareness Network.
4 We talked about security safeguards. And, Gary, 5 we talked about including a safeguards section in the report 6 to explain how we treated that. The same thing with the 7 design basis accident, explained how we treated that, 8 possibly something along the same lines on worker safety. 9 MR. HOLAHAN: And, in fact, that may be something l 10 similar to what we would do on aging degradation of the 11 spent fuel pool, to have a discussion of why we do what we 12 do. It may not in fact change the numbers. l 13 MR. CAMERON: All right. NEI technical report, 14 . well, you just covered that. They are going to send that in 15 within I guess the next couple of weeks or something like 16 that. All right. NRC will evaluate the NEI report. I mean 17 that follows. I 18 Areas beyond the working group study will be 19 addressed in future public meetings, such as organization, 20 QA plan, fitness for duty. These are some of the issues 21 that were raised in a letter from Paul Blanch and I think 22 John put that on there. 23 SPEAKER: That would be part of the rulemaking , 24 process. j 1 1 25 MR. CAMERON: And that would be part of the -- O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 l (202) 842-0034
t .- 5 465 1 let's make it clear, this would be part of the rulemaking 2 process. These broader issues, the broader look that Ray 3 was concerned about, i 4 NRC will notify interested stakeholders of 5 relevant information. In other words, there is going to be 6 things that come out, keep coming out of this process, 7 including meetings, whatever. We will keep people informed. 8 Stu. 9 MR. RICHARDS: We are doing it. 10 MR. CAMERON: All right. 11 MR. RICHARDS: It is all -- there is a tremendous 12 amount of this stuff on the web site. We have made it 13 available electronically. You know, if you are missing 14 something, let us know, but there has been a tremendous () 15 16 amount of this material made available. MR. CAMERON: Okay. 17 MR. RICHARDS: On the web site, send individually 18 to people. I just am -- 19 MR. CAMERON: We are going to keep doing that, and 20 we are going to add Citizens Awareness Network to that. 21 Stu, if we are sending out information to specific 22- people, put Citizens Awareness Network on that. 23 Industry,'this is Bill Henries will submit i 24 information on the pool heatup test for the three New l 25 England plants. And that is where we sort of fell off at l O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 ' (202) 842-0034 I l.
r 466 1 the edge of the earth, I guess, yesterday. f~^ 2 Then we got to today. Okay. Seismic analysis ( ) 3 should include consideration of the impact on the transfer 4 -- I put tunnel -- I think it is tube. Okay. That will be 5 done. 6 We agreed that the checklist -- now, this is the 7 checklist on seismic that we just talked to. That is the 8 convergence between the differences and estimates of seismic 9 probabilities. It also addresses the -- I put the mean 10 average issue that we were talking about. Does that make 11 sense? Okay. 12 And NEI is going to provide a proposed checklist. 13 We just dealt with that again. 14 Heavy leads, assume of 0612 applies and NRC should () 15 give appropriate credit for this in the analysis. We 16 have -- i 17 MR. HOLAHAN: With early feedback. 18 MR. CAMERON: With early feedback. Okay. 19 MR. HOLAHAN: They are also supposed to address 20 the upper bound. 21 MR. BARRETT: I'm sorry. Would you clarify what 22 you meant by that? 23 MR. HOLAHAN: In heavy loads analysis, will we 24 address the issue of realistic value versus upper bounds. 25 MR. BARRETT: Yes. (' ANN RILEY & ASSOCIATES, LTD.
\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
( Washington, D.C. 20036 (202) 842-0034 L
F 467 1 t MR. CAMERON: Lo we will address realistic values 2 versus upper bounds. All right. {} 3 MR. KELLY: This is Glenn Kelly. I am thinking 4 particularly what.we will do there is probably better define 5 why we choose.the values that we do choose, and explain what 6 the basis of those values. It may not be that we take -- we { 7 have upper bounds 200 and a lower bounds of 50, that we add 8 the two and take it in the middle and say that is a 9 realistic number. It may be that 200 is still the best 10 number, or that 50 is the best number, or.that 125 is the 11 best number. But whatever it is, we will explain why we 12 think that is the best number to choose. 13 MR. BARRETT: Or that 200 is the best number for 14 one kind of plant and 50 is the best number for another kind 15 of plant. 16 MR. CAMERON: All right. Then the last thing 17 besides our very important industry proposal, NRC response 18 that we just went through, and this is probably very i' 19 inartfully done -- thermal-hydraulics concerns decrease if I 20 licensees intend /are required to maintain procedures, et 21 cetera, et cetera, after the time of the hazard has passed. i 22 In other words, like do you need to worry about what the 23 time limit is if these procedures are there? And I think i 24 everybody agrees with that, but there is another shoe that 25 is left to fall, which is -- are they going to be applied? ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
468 1 I think that Mike indicated yes. 2 MR. MEISNER: Yes. 3 MR. CAMERON: All right. John. 4 MR. HANNON: Yeah, I think if my memory serves, we 5 did say that we would intend to go ahead and attempt to our 6 best ability to complete those thermal-hydraulics 7 calculations, for other reasons. They may be needed to 8 indemnification or, you know, other purposes. So we do 9 intend to complete the thermal-hydraulic calculations. 10 MR. CAMERON: Okay. And I wasn't trying to send a 11 message otherwise, so that is good that is on the record. 12 Alan Nelson. 13 MR. NELGON: I guess we are kind of wrapping up, 14 but I just want to, you know, express our appreciation. But 15 in the same vein, I guess the first thing that we will be 16 seeing from you is a form of matrix of activities, and then 17 we can work with you after you have identified teams to be 18 certain things and we can support those with expertise. And 19 we will be uorking on those actions that have been -- we 20 have agreed to and provide you with them as they become s 21 available, such as the seismic checklist, a white paper. 22 And if we can go ahead and N e some of the PRA material by 23 Ed Burns and Aaron developed, we will be working toward 24 that. 25 MR. CAMERON: Let me ask a question of the group. O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
I 469 1 Is this, all of this matrix, okay, how we are going te go 2 forward on all this, will this be in the near-term plan? I 3 mean that would be efficient, but I don't know if that is 4 the intent. l 5 Gary? Anybody, any response on that? 6 MR. HOLAHAN: It sounds reasonable. 7 MR. CAMERON: All right. 8 MR. HOLAHAN: It seems to me we have to right 9 these down anyway for some sort of meeting summary and to 10
~
use them as part of the planning process. 11 MR. CAMERON: Okay. Alan, I am sorry I l l 12 interrupted you here. l I 13 MR. NELSON: Well, I am glad you asked that 14 question, because what we have kind of laid out here in the () 15 last half hour or so, or hour, in this process seems to be a l 16 little disconnected from what your opening remarks were, 17 Brian. So I was I was wondering if it does -- 18 MR. SHERON: In what sense? 19 MR. NELSON: Well, you had said that this is not 20 typical of the way you do business. You usually take the 21 information in. You thought this was a fact finding and 22 then we are going to go ahead and do what we normally do, 23 and then we will get back to you with some sort of, you 24 know, kind of results, unless I misinterpreted what you had 25 to say. ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
470 1 MR. SHERON: Yeah. I was not trying to imply that
) 2 the next step might nnt be follow-up meetings. Okay. What 3 I was trying to say was that we don't take information that 4 is presented a: a meeting and make an instant decision, 5 i.e., yes, we agree, we will put this in our model and our 6 data is no good or something like that. Okay. We just !
7 don't -- we don't operate like that. 8 We have got to go back. Okay. We assess, we l 9 analyze, et cetera. What I am kind of looking for is, in 10 terms of the next step here, I guess what I am hearing is 11 that the staff is to put together sort of a plan with a 12 matrix of what issues need to addressed and how, and that is 13 fine. I think what we need to is know who the contacts are 14 that want to be involved then, because, obviously, the next () 15 step will be to set up meetings to discuss these specific 16 issues, you know, whether it is thermal-hydraulics, heavy 17 load, you name it, okay, and we are going to get, obviously, 18 experts in smaller meetings. 19 Do we know how fast we can get this plan out on 20 how to proceed? That is what sort of bothers me. Getting 21 anything out of the agency sometimes is a, you know, I 22 bureaucratic nightmare. And I am wondering whether we 23 should identify what these meetings are right now and 24 identify contacts, and then they can start the process. 25 Even though we are developing a plan, if we can agree that l l ANN RILEY & ASSOCIATES, LTD. ( Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 I Washington, D.C. 20036 (202) 842-0034
471 1 there are certain meetings that ought to go forward fairly 2 quick. 3 MR. NELSON: Let me just ask if you can put out a 4 draft plan to me. I will put the assignments from at least 5 the industry's side on the plan, so that when it is l 6 published -- well,. I mean officially, you know, as a final 7 plan, then you will have all the names, part and parcel part 8 of the plan. 9 MR. SHERON: Yeah. Well, I mean if we put l 10 anything out it will in the public documents and so forth, 11 and we will make sure that all stakeholders get it. But, 12 again, is there something we should be doing right now? 13 That is my question. 14 MR. HOLAHAN: Well, let's see, the people who are () 15 going to have to do the work have something to say. It is 16 always good to listen to them. 17 MR. HUBBARD: George Hubbard. Brian, with regard 18 .to your question, is there something we should be doing 19 right now? I think the answer is no. We have a meeting 20 scheduled Monday with the staff, those that have been in 21 attendance at the workshop to see what are the things we 22 need to do, what are the next steps, and see how that 23 affects the schedule that we had set up previously. 24 MR. NELSON: We can wait till Tuesday. 25 [ Laughter.] ANN RILEY & ASSOCIATES, LTD.
*O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I 472 1 MR. CAMERON: Okay. Thank you, that was a useful 2 clarification 3 Gary, you have something else to say here? 4 MR, HOLAHAN: Yes. There is an issue that came up 5 more than once, and it never got on the board. And it is 6 maybe because it was very general and not a specific issue, 7 but I am not sure I want to walk away from without thinking 8 about it a little bit. And that is, Mr. Shadis commented a l 9 number of times that in' order to do any of this you really I 10 need to understand what goes on when a plant is 11 decommissioning. And I think most of us who are working on 12 this don't.have that complete an understanding of everything 13 that goes on during decommissioning. 14 We sent Mr. Kelly out to look at a collection of 15 plants. I think that was a very valuable activity. He 16 brougut back some information and some ideas about what is 17 going on, and an understanding of the context of all that. 18 I think it was valuable. l 19 So I think we may need to do something to make 20 sure that the staff, and maybe even some of the management 21 who are dealing with these issues really do have a full 22 understanding of what it really means for a plant to be in 23 decommissioning. l 24 MR. CAMERON: Mike, do you have a suggestion on 25 this? l ANN RILEY & ASSOCIATES, LTD. Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034 ; i l' l
r 1 473 1 MR. MEISNER: Yeah, I do. The people in the NRC ( 2 that really know almost day-to-day what is going on out at 3 the plants are regional inspection staff. You know, while 4 we may not have.a permanent resident inspector at some of 5 these sites anymore, we get people coming out all the time, 6 teams of people, and they know intimately what we are doing. 7 So my recommendation would be get those guys involved. If 8 you have specific questions, have them check it out. And I l 9 think you would get a real warm feeling as to what goes on 10 at the plants. 11 MR. HOLAHAN: Thank you. 12 MR. CAMERON: All right. It has been I think a 13 productive two days, and I think you have really worked hard 14 to get some really specific proposals here for a path 15 forward. And in my view, it looks like we are pretty much 16 done. I don't know if anybody has any pressing items. 17 And, Rich, you are going to get together with 18 Peter James Atherton after this? I .19 MR. BARRETT: Yes. 20 MR. CAMERON: And I guess I would just thank you l 21 for being a good group of people to work with. And I was 22 going to say we are adjourned, but George has some pressing 23 need to do something here. So -- 24 MR. HUBBARD: The one action item, and it may have 25 come up when I went out and down the way, was yesterday I ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I 474 1 tnink there was some discussion with regard to heat loads ( 2 and that some of the plants had done actual heat balances. 3 MR. CAMERON: It is on. 4 MR. HUBBARD: Okay. 5 MR. CAMERON: Yeah. Bill Henries is going to get 6 us that. All right. 7 Gary. 8 MR. HOLAHAN: I think before we go, we ought to 9 thank Mr. Cameron for putting up with all of us for two ' 10 days. 11 (Applause.] 12 MR. CAMERON: All right. I guess we are adjourned 13 then. 14 [Whereupon, at 5:13 p.m., the workshop was l () 15 16 concluded.] 17 18 19 20 21 22 23 24 25 l ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
p. l REPORTER'S CERTIFICATE This.is to certify that the attached proceedings l O before the United States Nuclear Regulatory Commission in V the matter of: l' NAME OF PROCEEDING: REACTOR DECOMMISSIONING 1 RISK WORKSHOP l CASE NUMBER: PLACE OF PROCEEDING: Gaithersburg, MD 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 typewriting by me or under the direction of the court reporting company, and that the transcript is a true and accurate record of the foregoing proceedings. I' YW\ 0Mi n Hundley Official Reporter Ann Riley & Associates, Ltd. l l G ,
<yHo %
O Probabilistic Risk Assessment Input to Decommissioning Requirements NRC/ Industry /Public Workshop Risk-Informed Basis for Decommissioning Exemption Guidance July 15-16,1999 Gaithersburg, MD O ; Heavy Load Drops Only single-failure-proof (SFP) crane load I drops need to be considered since consequence analyses must be performed prior to using non-SFP Cranes O 1 L
O Heavy Load Drops NUREG-0612 failure probabilities were assumed without crediting nuclear licensee improvements due to implementing the guidance provided by the NUREG The DOE Savannah River basis for concluding that the NUREG-0612 probability values were adequate used a " generic" data base, therefore, it is not directly applicable to nuclear sites O Heavy Load Drops
- Conservatisms
- SFP crane probability evaluation assumes that it is 2.5 times more likely to drop into the pool than the NSFP crane assumptions
- The assumed 200 lifts / year should probably be closer to 100 lifts / year (factor of 2) i O
2
O o Heavy Load Drops Conservatisms (continued)
- Qualitatively lower failure probability should be assigned to high profile lifts due to increased oversight Technical Specification or FSAR controls
- SFP crane Analysis and procedural controls
. Enhanced management oversight
- Safe load paths and verified rigging
. Crane operator training and pre-job briefing O
Heavy Load Drops
- Conservatisms (continued)
- The assumption that "one-in-ten events"causes I significant pool damage is too conservative and is not justified
- The timing of the assumed drop is ignored. Most (all?) a I
casks will be handled after Zr fire concerns have passed a 3 i
1 O Conclusions "Best estimate" probabilistic analysis results should be used rather than the upper-lower bound values being listed to provide a better perspective on potential dominant risk contributors (Note: A "best estimate" probability of 3.1E-8 is listed in NUREG-1353)
- This heavy load issue was previously closed for operating plants by NUREGs-0612 and 1353. We see no reason for it to be re-opened for plants undergoing decommissioning O
I i 1 1 0 4
O Risk Informed Decommissioning Emergency Planning Evaluation of Spent Fuel Pool Seismic Failure Frequency Tom O'Hara Environmental Sciences Group Duke Engineering & Services BE"l.15"t""""* O Overview Review of LLNL/EPRI Seismic Hazard Studies Summarize NUREG/CR-4982 (BNL) Summarize NUREG/CR-5176 (LLNL) Summarize seismic section ofNRC Draft Report (6/99) Compare Results from alternative methods S_E?.*53L"**""'
i O Chronology of LLNL/EPRI Seismic Hazani Analyses NUREG/CR-1582 (1981) NUREG/CR-3756 (1984) EPRI PROGRAM INITIATED (1984) UCID-20421 (1985) NUREG/CR-5250 (1989) EPRI NP-6395-D (1989) NUREG-1488 (1993) yggggarina O SECY-91-102 IPEEE for Severe Accident Vulnerabilities NRC response to question concerning use of both the LLNL and EPRI seismic hazard curves in the IPEEE.
' Based on the available inforntation to date. the staffis unable to dispute the inerit afeither curve and considers both of thern to be valid. '
O_EY.e*r$"lN**"* )
r I. n U NU REG /CR-4982 (BNL) - Severe Accidents in Spent Fuel Pools ... July 1987 Seismic Hazard Millstone (BWR)- UCID-20421 (1985) Ginna (PWR) - Synthesized Fragility Millstone - Oyster Creek Reactor Building Used as Surrogate Median Fragility = 0.75g Ginna - Zion Auxiliary Building Shear Walls Used as Surrogate Median Fragility = 1.lg SNT5",5L""""' ADdr 6enpCupe O NU REG /CR-5176 (LLNL.) - Seismic { Failure and Cask Drop Analyses of the i Spent Fuel Pools ... January 1989 Seismic Hazard Preliminary Results Came From LLNL Hazard Assumed to be Lognonnally Distributed Truncation of the Hazard Distribution (99%) Family of11 Hazard Curves BN_T5"
- L""""'
O NUREG/CR-5176 (LLNL)- Seismie Failure and Cask Drop Analyses of the Spent Fnel Pools ... .lantlary 1989 (contintied) Based on Explicit Evaluation of Spent Fuel Pool Fragility Vermont Yankee - Median Fragility = 1.4g (BWR) Robinson - Median Fragility = 2.0g (PWR) Note . This methodology used in the draft EPRI SFP e It geering A ber limpGarens O NRC Draft Report (6/99)- Seismic Events Spent fuel structures at operating nuclear power plants are inherently rugged in terms of being able to withstand loads substantially beyond those for which they were designed. SFP are considered robust for seismic events less than three times the SSE. It is assumed that the HCLPF capacity is on the order of 0.45g. I92"s_M"**d"'
l 0 i NRC Draft Report (6/99)- Seismic Events Using mean LLNL (1993) seismic hazard curves, it was determined that the mean annual frequency of exceeding 0.45g is on the order of 2.0 x 10-5 i Using the definition of HCLPF, the working group applied a j mathematical shortcut to get the frequency of a seismic event that will challenge the SFP integrity: j 2.0 x 104 per year x 0.05 = 1 x10-6 per year. l l m Duke Engineering E9 A Services. Alher RampCsepser O SFP Failure Probability Results - NUREG/CR-5176 Methodology
._,.....~......._m g_u,m . <._ _ ,.3
-.m....,
.... . _ _ .._ '* a m- s e .
._. 1 _ _.
? g f 9
, . i .e 4 v..y 4 4 , . 4.__4: .fJ,. ,l .d R I $ g_. "_,a b i*i ? :o-a) e _ . ya 'dj h .j* ** g5 f , ,lg , _ .,l ll
==
o . a:
] ".E:-gl'l i saea - .- __ _
fj '.L b
\
smea I [
---e e ea n r. :
3, e ie a e i : B Duke Engineering [0& Services. A Ddr &asy Campsey
O Average SFP Failure Probability Results - NUREG/CR-5176 Methodology 100E44 Pete + 3c pmpusc ol ehr hie besorm ss abot b onh m asui as imeckeg the assses for e ga en asthod 100E 45 - - v.i... u. .eu.nem e
. ?
~
e ,.". ,'v', - 9
?
' c Io 9 d ^- h a l "o ".
~
l.: . . . . . ..s. a :' : a:
,J s e ." ; ' .
g ;- .!=-.. .a . a a:6 ,a 6 oc: o. , ..,9; :
. .l , oo , - J'
- a. o s oon er a; .y .g; a6 e g
; o b
. .os o.
-: : e e=cza : : es see
- : : : s : a m Duke Engineering Ed .&_S_ervices. ,
l 9 NRC HCLPF Approach - Draft Report w ium .,, . .,
, ug.,
[ x .. . - l 4 . l 1 ....._.t.,- a , _. .p 1_ ._, p. l t Q1 =,. . . .
.f - oI
[,,
- i,' .a ., I 7.._.
/
. . .I .' , 1
- - r : : : u o a . s,n ....,,,2 i e B Duke Engineering rF& Services.
A lhd. Ee.g, Coup.x,
- \
)
1 l l m i U i l NRC HCLPF Approach - Draft Report Average of LLNL & EPRI b v g .. T 1 io .a j *
- _.. . . . ? .... . 'oa. .. .
!=,, ~. . .
9
.- . =
, ,. . =, .,,
44 ,
**. , , = .
=
n.. a . , , . _ 4 . ..
, e I i s.o.
{ 1
-a : e n 1. n a 2,,: i s : i e r e a : '
D Duke Engineering Ed& Services. ain snm O SFP Failure Frequencies Based on Mean llazard Curves & NUREG/CR-5176 Fragilities a . . ..w.. . .., -- - - unacca sits e.ne c l.-. . w.i.i-1"- ! i
?
. - _ . 2.. . r . r . , , _ a 7,, , . , .
i .,; i ,. '. 'i . . ,.
. a ., .iy.,,,1lll : <
y i . ..
.- --._--.-.-.-_.-.--..-_..---,,-._.u_.
184,W
-: : : e n a n . 3, s . . . 4 F E F 3 3 S p Duke Engineering I# A Services.
A lu $my Campany O
1 O Average SFI' Failure Frequencies liase(I on Mean llazar<l Curves & NUREG/CR-5176 Fragilities
'"~
,,, ,. = .g g g .7 _ ...
I me.sp
....m...
?- ... .
r, , , . . ! ,,
/* * : . . , * . s' d
. *l...i.....~'*
.. ~,
. .w .?. ,
g , 1 _ _ . . , .
- - - .-.e n r. ,. ::
g=...,, ..i M Duke Engineering i Ed& Services. A fe 6enerCompey O Conclusions Based on the NUREG/CR-5176 methodology and alternative seismic hazard inputs the mean SFP failure probability estimates for EUS NPPs are: LLNL89 - 6.5E-6 LLNL93 - 9.0E-7 EPRI - 1.4E-7 Average (LLNL93&EPRI) = 5.0E-7 l D Duke Engineering Ed & Services. A Dde hwyCanyser i 1
O Conclusions j i 1 Based on the NUREG/CR-5176 fragilities and { alternative mean seismic hazard inputs the mean i SFP failure probability estimates for EUS NPPs are: LLNL89 -
- 1. l E-5 I LLNL93 -
9.6E-7 EPRI - 1.3E-7 Average (LLNL&EPRI) - 5.5E-7 l y,Duj,e g;neering
)
O ! l
)
Conclusions I Based on the draft NRC HCLPF approach and l alternative mean seismic hazard inputs the mean i SFP failure probability estimates for EUS NPPs are: LLNL89 - 1.2E-5 LLNL93 - 9.0E-7 EPRI - 3.5E-7 Average (LLNL&EPRI) - 6.3E-7 OaserN3U**""* O
Conclusions Based on three alternative methods to estimate SFP failure frequency, it is concluded that the SFP failure frequency for the population of EUS NPPs is < 10 6, , 3it_.'R L""d"' O i f O
O 9 l o n o i o Pst s l en s B B ua e L X Fl P S PB R t ng n S L/S en o f,PA pi n i s eSS So i/S tois s s u c Oz og ) hA CS D 9 9 di tem i s G E 6 *hSr, cDe n ,e i 9 1 O am DR B an r 6 . l eo c }gg ** B s no m 1 y Rce e h (M i
,k l l
u ncd u kDs a oau J it r nJ a Ra d n eS t a n nts y Hah on pe H ip od l hnDe s h i c a o o sc m J J kA r r - o e h W T _ O _ - ~ ~
o s n o 6 i c i s l u i s c a r s D e d e r y c o o r f H- p i s l a e s h a m t g s B r s re l e n i g a c n i h st nrm ye i n T l a m i n n oo n a a h c o o if t ai n a rans M e T e 9 n po l t n ~c uk s i yd d o l ar i et r np t l r arae et n n i s scr o
~
r a o e u s u cf ut c roer l h o e ci ots n dph CSnmS ga S o nmt te= Fi ai o i l h n ni ni t a t sye aeAs cu o omd rf i r o i nghlbn ig s r n u ocatak i i s et e u i st dnlp l a s ueo sn de x mies a u cDF l aUI E mDREM c s o o i s i D00 G0O0 C000 0 D + + 9-
+ -
- 2 Lll ll lllfj
t O d e h t g a e n e s n ey h i g i o o p l e h a e r n b a x e mr t c e al l n n uod y ci e e mfn t a gw a t w c i nwhi tn a i l a e h t S w i ot in u s l e mdhci mur
- t hf t
e u i l s u ul r eh y f t
"e hs sai a vt el m r et o-l a n l l
e n e p i Toul m eu yct f ei u hf . A s un t t e O. c f o aad cer ef o we t n ur g i l ei r e gt r ar c opa u l r a De aei z e s re a u y l anchh be p cisst owo r t al c t h d r n i st m y eo p i t rddt i l n e l ol ot e H y Ceyeb s o ci n l mb " sl pt nirc e i i r ye a ted sa el e el eff u n . t e s m sl eo t ecu a eresri ao t at a e r e zo yc l et wau r c h l umt opad cis t eq c ei n T ad af o nmi l t n n x nd Aa Comit se o Oat o O + + + e ^
^
~
e e O e r r r uo d l u 6 u t mtae ort t a o a r r e sh co ep i r e p e o r s a p m me t t f i n n ml . s ie t c ot h nAr a e ee u y e t c t nf me b b e i cl.o S uh i f et n sstede it it f ut s e n e uf n i s o s ei e s y ap e s cAaa r e ri n eh c r e i bt i l a c h z )y n.cm nmf u f n nt d gi o . aos . i nr edn cr e A. eo f aer si o f ei t O c h t r nuit s s cs 4 a et t et rf i a i adc a r cu f l e,e sr n e u o u ch f ee o cp ar t uo - a ny adss r oyc t d r ea uc num)s od eyo n er pprdgu y qe i t ot at i e o H ed cr apewa mn eo nr s e e(h ai d t i ep l a eh r c t ,s x r s nn sn o s ht nmeue i m t o o r sr hfeiao r nt i e a(ht ge es i e i yu e t ah i h sds - edm a vf i i h dt ee oe T l rd a i x o ri cpo Eis Oeirf x nar hle Tr rh pt O + + + 9
O y
)
e a
- t e i r
d i h f e s l i t f om u e b a e de su s s y c uiei s c n i i l sd o yt s . l a p p sun c r nr e n a i ys e r t o af i ui t A r o t f e )s q (z apm r ogu l c f ales n S (N not o bni i l s ain u e c s c a i Lno a r i d i r e d l ocr u eidi a i d t . r n o n x s. n si s z y sst et a oro oir O H gn Gar a,mt c i t mo
- l na f o nu a N amf l a
i t s p l obe i n e oa t a r ncit m r i xde e n rLmo o ocg een vde t e plani r n a co h wo ei pnezi u oh dr a t hi r p k e T i s sipee o n e g vs na t np o h ei r r et n m i
,Nyi t a o Bg g i
Gm s 2i ,l i t s r ,f oi n b uf o t s o ' i x Wc Te 0 8 )8 a sr 7 9 y d u 9I b Sol e a f
- i 9 dl E ed 1 1 uc h nGr psi n nt si n Tt o i ( I O + + + *
~
t O t one O t ei c y s af i a e rf s u u d s y e c s W t o l a i mcd a e 1 d t L e n yoit d E s A a n v U ul e eer o c F nu i c dr e p S yaf l ahl a u l awo r . d n dt ogt u a c s a t di nca r isst d t ieri n n e L ek y cdo E scle O H u i s m U uad O et id r h scn F n e o agm l a t ge e S t m r an dm htiya d e h t at r o s e t s r s so n n dxt ioe s us pt h eea n s e u sdo T d l ed ne r s ut a l i g uigcnue io dc un o r l u c u n i cdel r r be sg i t ne r o t st e rd e ef i ot arf t s cl 2 u wssdn u ol p oe oo i x Gcpt pi a 8 m cc E I l l Wlaue So Gc O0O Tcsd O + + O
O e s n e d s o s o e i t a e c s r s u s y r e o g yl awo i l t f a u d n p o n tolf c A m s o d de l e c c e t bel i u i P s c f l u R u emt ud l a r A ngseptea l a e n.a tu d H h n sidc y S t a elia O H e Rlyn e l a h t d Weimal Biest vr cr o ve rd oo e r a aodm m r s o rv pf et u f h e y psesefio db l in r d unia e T u no r r cgtst vld g t s u nh noo u n L b r ugunc i t N ee u eee s B wdoddd r o oh ooo i x 7 ct ccc E 9 9 l 1 oO O0O O + e ~
0 e f o 9 s m n e o s~ s m i t i e ig y t ae
'y s t n a dg de ad y yi c l
as t e dix a r od e i a ci d l a o o t e n e x vot s t r A de l a e f on l n c rdn or s re c aa ch i t aa n a _ i i r slet e o s nw i l u t t c t i r r i t a ceg d rdeo ol a e c _ r l t e pu dp n l ui t d an a eo ey cc y t uo l u r cpa c l a a cpr l 0 H c e c mh e 9 ar oi cde d l l e a a y ne a c f rh d og - h ss g nwh l n a m r t t en c n osiga i t u o r e aioi i s nh n l h h tct it s u o o ots t ai e s T sr al u ey xs l ci i t rt i c dn g epc a ndaa r i r a n vt en ce0 l asd t ond u e coag ng i t i l er c e g0 l cl cf i en s briur n7 a r e e n git o s o ara x i eh h f f c e n Li nTt c r e E ar e e op t Sf o gO G0 F o 0 + + + 9
O g i n t - w n l o e p a s l o S i r - e e t F n n t a p i o i r u s i t c - m t t n a r c i a u irr t s e e2 g e e E Heid8 gcI i f nw eo _ l eacS n e urAG o gP - m F o ef c dy - i tn Sr et o t g ea nc O T egv r i n i e y pne o d n s mD r a SrSp p i u k el a c u 2 u o c t ec e s9 e4 a c6 D r98 S i b R dt D n0 e4n- a pyt 'y r i c s C l et i d ui i r rRaRs e n s uU e feCWCl e/ / o mloue r n oT n vM i e R GfGoE o E P uoBD sP e/ G y rRsRl s e 'al s l hh gg rW p K aUoUu GFHH uii d mNLNF e6 i r O W TOOO UO s P0 O + + +
^ ~
l e y O a m c i e i s u o t y a c t d od ece l ad t o G_ s v e rid e edd t deie t rf r _ a pcl a i cen s s mmic i t m s ooi r r f f t r c nr e oar cea y E gs e e s s nnh c a5 n e i si ot n a w3i o - t m st ad nf o ol a t v n s i O i T mr e i s i t e u c G y a ets x e at al a a n b e gR m e n r mc c m oniw foit d y - n ax e o cio s n el et D n ot MlF i eh i ai n i c ed h eht ne adr vget t a r ps r e as aca oud bo e - n ar ec n ibnwp r n e dsi ooS l e ao u G eetaFd r sl i h t d eMec - oyuDr f i cal c3G t l oveu . ana F acOO l l A omd r pi t e r O + + O
r 9 f e . n s a i n n a n s a r e n t m eo t o i a e n mta h h e ol n p n c u o t e i t n h a l c e PC s u d n i d c dp o n c n u i t e o a at p c c e a m n d r r a n o s . e r e u v g i t s a a t eH s s i s , n e s a e n i elo n n r o n d mo r e o i wt a i o n a aP r t t i t o i t a g y a a d l d al e r p l a f a d g n Pu t i t n o k c r i x n i n e g l n o i o t F n e t a x n t a n miv o d g a e d r i o mM i l o e a h t c u l a t r e op k g o n l i d y a n a i e n o v o f dl wB pS l c e l c l e n c u o r D a i a u u e l m R F B D F i C Z 3 C o I
+ + + + + + + + + e i
O y l y . n t e a a r r m e oe i x s t f h o n a et o r o u m r p y i c i t o pr t a ey at e a v e l ym l u bt r b c o l ae o t i m de l a e a o eg l v e t C b a a g al mu e p l i e f i f a h o t - O t u v a t n e st c e r a e a og et e di n e dd H m dee n e i c i t e uwe nl d ooe uh st r t t a a bnn i b m rk s eof a i eoi m2 i d t s e l wNr o .e s i t f eo t t A o a en e l br o d t sdm eia l i at c e s vcr aa i ca vf U Gap Aa O + + + e
)
O t e na _ imsi et _ is t nE i t ae _ r m g ei n O cT i ny i x Ua m s t c n d l e n e n - y r n d aD t r e o a o nl e t t i t a e m _ i a m a c mc o p l i t n t e - ot i e g n a a _ i g n e t r Dr i v s n C k d i c a g s e o - a o n r i t r l i d a l l w d e e l i u o i x u u l F F B F O O + + + + + ~ .
O G t s n i e i c f f e e o c c n s t a s i s l o s e e t a R l p O w t r & l o t n o p F e p e i c ) n u l i f 2 a o s C i f d e t c d n o 2 o i r n u c A f a r B s s (
/
) l e e o 2 d c o
n e a l e A- u R p 5 c 2 b / 0 s 2 i f A ( 'd 0 d 2 o i i r 1 r
= ~
O k R f~ G k O + + + G I
9 e n a h t r e n h i o e i g t t a r a d Z- r h h "e cr i e n i x l o o uf O m i t c mmu r / a s i e n l i i a e e A- k c r n s o a rc 9,. m 2 i o l ei z u 6 t e 2 a r" i n u g s q yn o i e e gi r r c r s e eu i a t a nd Z e r et n a l e c r i l iho e y g o t ay r b e a r da c i n a x e E P Od O. + + + e
~
t a a O n h O e e t md ol i c s n c ns i or s ore t i r e a i t e h "s b a a s u Pn ue s l n i a s e g do i d a l ccn e nt a em e n a a o c u s y i ss l u t t qv e s r nrel c s n ei l a e i o t a i s e st s n c n : n t e t i eC mp l y e r ot e A o m d na a n pco p s a r or at u a eep r u n a c a r hi t e O t o p l Pe o t fs t & a ot H i t s r on n of offs i e e f o pao l o t s oo H g g s l t r o i t an dn e e: eoP o r u usf u pl u o l r s h eo me i u l ct t t f t s a et sniu l al a e a r e u t v ye" F cuc mts m s rl F e oapnt son el ent i u e n e vac vot n q al sp i i a pi ta mS t t e t ap a m e r n n r tu n rdic ei r tuient n G e aiii e a t el f f t e u l W t e l l i Aws l T AhOO Ah O + + + + b .
o l o o n Pst l o B en i L ua s P Fl P s S B t ng e f ,S en S iPe pi n n hS So i o C/ A tois s nS 9 s s . i oS 9 9 di emu %q . tc D 1 o am c e , , t l eo s i
" . Sm ,o 6
1 Rc e D . y kD s "*p ,, drrh aT l J u - s d b it Ra a b d r - o ua nt s Hw on L ed - pe y gE oi d v r o h sc c a e kA e ' G r o H - W - o ~ - ~
r a k e e s l c r a u u c N l i y ) f a l t 7 - n a 7 e s p o s 9 ) l g d d 1 t f mf t n or e r a o i l 4r t s r )s el i i s - D e L 7 e y e e , d i y 9 1 ps pr u m d s v ( sf sl e a n a a p o pi r t y- ys _ o o c H e t ar ooofar r Rs L p f ddpde/ - d 7 g f o u o e8 - e4 o-0 n o n0 r0 y r l o r a1 r xix l u1 m1xd i l c G cy1 0. fa0 n0 t n g n o n a )m e n o v e1 m4 h e u i k C a tol gt o mtof oys t q r N n' o4 e o 2 : "s t n0 4 i s0 c0 resf F r W 1 6 n a o1 x n 1 xth 1xl u o l a g. 0l d5 i o i c n i G Pr e s1 (1 (1 ( o n 5. wi 0. f par hd nl Ee a
+ +
cn Rw B+ - ea Uo - Th NP O -
+ + -
O - n ",3 9- _ g _ e sR i s e _ e sit ai t - e D BliC n hd cR o s t r o n taaS a F i t a e - DrW t _ o y "s r a . f el r a , e p r _ sBo o t f e o . e i i ol r cC o r s",P y l r r u r l u l 2l e E Ne r e p i a a8 e p nr v f r n u s u s aoR e i AeuF p mt c p o n o O yst r sn o r uah a r a r h 3 oie He . d c g t cp d r n4 n 5- l 4 enn9 l aiS ue r 0 t i oa9 0 a c i t a gni n 1 x SNv1 a , 1 x i r r e ee s 0 r r eoS 8 5 t c p RGt n 1 vf e 2 1 e o Rne sy l
- i o E r 3
f o de t o t u ra t / e 5 l p 3 1 ni o c 0 h amh our 0
- a c 8-ic 1 n p gb 1 i
n 0 GuA El t s 0 x n on e al iF 5 x a 1 x oi vet s h 3 Rss 1 av , 1 c e18 e Uea NRB O Se
" DW5 0 M O O + + +
~ ~
n r e % 0 c 7-n o % s r C 2 ;o r r 4 j o s
- %e 7
r r a r o go M %r 4r i n ta r e 3 e ge g h - r ip got RO T e i n a r r ge g p A i RO ) 7 s r 7-o r : 4 7 r t a ( a 2 E 1 a t a 6 d n 0- A d a G H y m S v E a u R O N H U N 0 0
+ .
i L
r o r ) O e r s r c e et e r o n n sn u n a a ua a de t r u n ac e b e m cf i c c i r t n u r i noo t n t i na h j ognr p
)
o l
% ai s - m C om c 5 m se et t
h. r ) g c a r r 3 - i o
)%n i
o o : u r j jo ap
)
I %8 e q t_ s a D 8 (1 i e i o M % ( 6 cd h M : 4 p ( r n ena gi f / s e
)
I 4 u oi w e h D" o r ndi i e nd% ( r r v e O T r e g ae9 r e oc k r nl t
/
h t 5 r inr eeg o a p a (m l i g A t W mkr er e s r e s I ni l g t uo g b l v o r f oen g n e HWn o r
/
v o r d y t d *
;np s
i o g r u nen eei i c dn yt i at r n e o E t S c d. m yd oao A t ueTad c h n l a t rl e Si;i rf c oi e. a c a pl r n n q m i n pme e a Ent Oicg h-s u v r f i Dof ai s t h D "e o C /
/
H c : p e t t T 0 0 o h O + * ~
e h t e l f r o o 4 0 e v r2 o1 o l o 1 x
- 0 aj6t o 0 1 h a 0 e p
)
2 < t om Ga g i n t g ey r
- 4 0
0 u aR sheEmt l 1 1 y . smr x x r a r a t e eUda Ran ea e R d (p 5 1 5 2 m parNf op p l mas u not rh e yd c e n n l o o p eg a)r 4 0 1 4 0 1 ordr nmy - e e o S i t e i t a -
- 7 y u bnuqs su s i l
u dR r l 0 1 x 1 0 x d i r a yr ei R eaw (p sl e 5 5 u t mdf m 1 2 9 t S n oht uuem sth o r a 7-
- l c c el 0 0 a rdl ee no r o 1
x 1 x c onat r aca d op 0 2 0 2 r nma i n et i e rl vuy e )r - - nan cst ahit
- 4 h oft R- 0 0 c at see l pn oe r 1
x 1 x r p e T e mnt ooo D s (p 5 1 5 2 uoe h ch t p t
,oenl f em r
ed e o o uet h eniu t f l gr l is t nr a np f ay e fs eie c aclu n t m e i s er m e nmnae l t n ut s go s l no i Sr e ovpces S y oiy Nf as i r Sp
- O
)
n _ w o k d ) s n e r i R i a u e _ r l i t h a e d a t f c i t r u _ w i _ n o q _ m o e _ e d i e l ( l d _ u c a f a _ u ( _ l l o f e n d r o s _ i e a p e r g s h s t y _ n c i n e l a O i i h y r d v n 7 o a e a W t l o r p p) n ( ong s r e k s d r v di t o i n a c a p k s e c f o d h sd a a e s cl a F s t a u cru r s g r e o e c i f t i l r k c c h f o g s e ru t e a pt O k s s c ss i R O O O O + ~ ~
5 s r 7 ( t f i t o t n s l c e p k e 0 a s n 0 e r s i a 2 r R r c
~
m e v e n o o c s t n o f r r n t e a d d o e a l p e e r s v a u m r e a o e l f i t b s n n s w s m n i e o e e r o d k p t i t e h s r a l e a l d c i r a m u r e n i h e e i f d a n l c s ts t i h W u n s s a (e s i f n a n s e r l py l r o d a gn c )r e r o o c o o t t aft r neea l t s c o y c a i f ) i l uy e t c v a i cn d qr F e r a eo pt neem ar pit f f a n h al eo o r s5 h p e l a a 2 k+ + a r el h i t t 1 s m h e t n a o u of u a t O t e dt c h o P O O 0 O
+
O t a h l t a c gso i t r o n i r idf k eo csr a r n ai b i gssinp a dsoae n el t e re ade h r l i m g)ent l yh e w s ~e tsit dif an n u sw rl i o : m yb s soo O t i e gd e t d. ' n t s i n nr at md . i f y t sy oea e e s n t od ol e sl i D f y a o em l l h ois l o r ( r w et h e p mt e e eh rt ga r e r t h uf n s u st l ot t t i l i yri sof a ye n e f a f ei t s e gml n gb l i r m ot s l i e g n t f t i sipajnadu i l y s a o d S Asacl a O + ll lIf,
d g n e
)
e i 2 l aih n , d 1 cat a 6 i r tr a e 0- t p l cy , G elr dd an m eo oat e l E R d ol s U n pwksys N a e,oc n dl pol l og n l i e r s d w mf el i n e s o aotxt d nn o u l l esf i a i f o ri l oh s t i n ( a t oF a of tyd a ri o i f e n .oi l o e r dtear bat e l u h t D l i aie l paf p era r a f o op pe e dea . d oahi vd k op o l e coro m d f i et d n e sl ah s o b k cu cd da r oqoe i n sreu s elo
. cl n t c a r
o e r t r l i o r C Piof n a pt o
+ e
M l P o L _ o D P s , l t g e n u a n n F l o i n t P i s o n g s i s _ e n pi e s iEA S n S mDS oi o n mi,S 9 t s o ohD 9 di s i cc egy r
, 9 e m s 0 1 s
9, M)++ Dar , t a m u "
- Ba 6 l
e o * + P 1 R c c s p + &m V ath l y u k D s e i D p( I Dur e t J i Ra t c Poa - i ,GG ns m s ot n dr pe i s a od i e h c h s c c S i k R r A t r o a W t u S
n o l e e i e r ueg no s t p f u l i t e ar n c t n f amo _ e r e e el - v p p h yf r E dr s t t t yic s cy a t o b ax i id mu z a e dpe t u e a n iSs h d n c a e e i c t y s i mi cc S h m il ci a r e mty sT s at t si i eic d ea i e pe sr s ad sp ao n3 c r st sc a BF i9 ceh i ya ei c i n9 n s mt t g n s o o1 i rl i t d ge m io c n i ed eh s ss u a s n t t n i ee r De ea d e9 n a n i ssh a t r8 i l n o R 9 s o oh i d s n t n1 et i ot n a n ca nc pa n a o y i c e eo f h oo ot i t f e rl et e i nw f el o sdiat s or n B ie gt f i o o al a Sb Dp Lmvp
+ + +
ll
O y _ d _ t u . S y e t i k s h c i r T a p a r a n n o c re i F cc e in u h c mo sc s s a i ek i s o r ss i i h O p dr t 3-p na s a e s A dsri e r a a d s z r d r t an ae n h a .n e c c t o n v i f n o y a E i ci im t et n d i c pan t u e si r m t b o m r s nm a po pf o i e l Pc Oi n S O + +
~ t y n s a t n d l p ay t u g l pi t S n c i s t u a e s opa h e nc T s e s ed r a r n o f ca F o s d h s on c d t a a h o rym o t ae r e sd s c p p meu ei ds es cm e s A i ni
- m ri s s e s oh n s t
n f t n o i oet t a e i - t i i k d r v st y np o n E i i rl ooi c eb i c rt pa i : nar ypn m e i f ab i v me i t s al n nh t i t m deio i e c xu S j b e Ev I wc O + +
O y f o t n d y e d u t i c c n t t a efl e ps r S apo i . pv c i e e ae e dse h cl p neug r T a s s i i r e sy dla n o t a n n vr i a e F aba c n r e _ dgi s i Ca l n h Ri i c i nie s r Nmg n s e e i a m sE o nd h or y r oct r ek e - O p i i s 5-p t amss uses f s vu ed D A l aen i l e r e e vl
- vsu kn uyt h i
s e eea c sn t n yh ue dei e t t i l s l aw rid ad e i _ v vd . ge zut E s n f amkeet r i t al a at h srt a i c o i c3 qr ue i cys r n mtso i m t p mt uht n i s m s iot r a idsume e e u ebal s Saep Sidn S s - O A + + - ~
l1 st s O ? 7 en e e 0 s cit n 4 1 t n r oipmS t mlai o - Geh e v "s f e 5. cea h i t s n3 D Et Ee i c3 a S l 99 p1 s Rrl t nde a n er a Uoai f ni l r r ee r W Ne rb mct ear daoowb e es f moto t o nxe ces r pc h a r aEn f df l u xmee eiuvrO t l W gu i oV f o3 aLia" v l e , nGnt dc s
- i pl on t yt a ke eu i n
y u taietad i l naisNa d ot ri ni i bouv9 t n $ u gmic aiqe6 t u t bmc 1 bahRr o S t nu aA9 orter",foM r t s l aSxe9 Er1 ple a8 sy c e8 ek pd e. e l a c e 4 n cnn t v n u a l t c e v i s aa e u n va1 o E bralaPr l S J ndeGmR a n n Et l i e u gR s h e uafu)ode l c doiU er s E t-i h dedE h ngeNr d fo m t ciEi s v i s doiEb l akdmzs a e at i e erd sPnIPu t be oaa S U" I ( p Opt h frHE
+ + O ll1 )ll lllll
O
? r e e n
l e p o v e 5-D 0 l _ 1 _ n _ s gfX . a i o2 s yt W t di et u l i o _ a eb b h h a a t bf W so o er e _
- mp u .
y i t l aa l O d 3 uv 7-u ,n n t sn a S t ea e s nm i t s a a n e ws fea f v h o r E a e e a rk c e c o a n y i f a m t pqe u dro s eh e c t i ct r c a e xax e S Eeer O + ~ .
l e h s v gi h tn a l e i h l n e ( y Te .pt o t r a gh a s e o t i s t 5 i l eta00 ief hot t b d h a y sf e et t br os o u e u ot of d c bo h r cl l t sn pa eh ad e t s ee as eel r rbt ue r n md i i ur a ov er e l if t nyt i apeb o v 3 oi f 4 o vac n E l f ci d 0 cius o h ba n1 c ygb aXriet g n di n i t i m i ch o d1 u s asp r rt a uf ne i pi z oo ab e ae e a byt yl e S crel ru h at i i r u f f chi a osig l t i t be f n at n o e i m ,f t s kel a cuc)e b ty vei e a n i d nrol i s ui o l oepobioic aqit rdpi e p a m hd ent hleoi bs f ht t r n iao h o r e Wec Tcif su ps _
+ + _
t O d u snh o y uag nf i o eh ir i et e n r r e a ut a cd vtc si t u e o r ur s o ot s n t b o dy al p ei t o s pi l s so npr e n oba t l er s n ol l o ve e e i s eai en r u sf u tuo rt v dlus E ut cm c eva e s n u - O c bl 9-i i dpt erd s n nram m s u auy s e nor ct r s t i t co i e nh sg s ut et i r a S sf e o ee gk l k s n t s r ps na coe enp e auq h od hch cht m so t eio p wc l r amla i a l t r et n e e or pue t e l mtsl n p eioa l v e i sam Aps l i Rf pl O + + ~
~
~
n e a i o t e r h t . h a t r f o u i) ie l w67 wle t y yc a e i a i n F yr5 v 1 cc na l bi o i bur c C e n bo at Ps u o Ft d ed/ qi om r Sn e nae rG et r o pd i z f e n o v a aR f t e m. h u o y E bhU hd t o cN t n sgr ci c ef i ( su eo e sta nm t ac e i v sr sh g e s ui i mer pu sg ef r o et r pf qe t s sc
- p x o eS r
eey xy en F o t t eiisi l ecn ei o t e e vt r aga t a f v r u r c oeu uf r uq un cuf cof cerd nD c n c y n i o a of i df r e t i ao i t t i i c ed l i st a st ah e ga a u ol uep zt a e a ee r l a mov-s h h c cx f r rl e v ene eie el u c E hoit h h s hi a c Tcs Twi Tf a
+ + +
- O n e
e e r b s u as l i y
. h s a l e dr e F nd v ad Ps Ft i
t t a e r o a Sn e i t n t ut as o v f a u mEs t y E q yE En a cic d e l ePp l O nm e s s s i t vI eg n ui e l aht i
~
qe r e rf t a d or eS r d s r e F o a i i p t e ya o f r gh oeu a l ot uf r o s o ok nD o i t e d o bs i t n h ai r a i a t r eoi c u t r l a e c mfdm v n eeis E h se U Tus O + + ~ ~ lL
l a r e t u e cR t u nreuo t h t rW ts P au n n l pit l h i s ai gf ai s sdd w t e one r ap i t nt n )r i s g a ay r
) o r c/ 4 p .
a l l oe el e ef i t a a f k v pi0 o gE n - u e D r o sn e k f ea v d ri1 osf a f e rye r f eom o l ba8)8 ut e d vr e s c n w eaet t o a e l i a1 4 tyml o t a gd r n c r v- i l r o ul l or at s et e i g e i l u aG aVp v o eE et u t. a hl rR f r ( )2 6 S a U R sbn r o sN e( i eWn7 v t Bs o15 e r n (o ft v rS a nR l uosi d t ai i i t a aie e uU c nf bC f r cr eooI c dr r n e s eR e G E mun geg r r . eiu n at zs puBR t f snqi aa ei l aH U yo ek HE Rf ( N Scf tr a
+ + + e
~
O i c dg e f i c l uni e cn p i nia r s dd t e l ue i l s h ob f o sra n n n e v r o o e i t io:. t cg s e c c a en r n n uri u o u an du l t f vol c c C a e n n e r u r O f e e ai u t 3 1 o r a h , l i s s s t t ol o f aetr n a s n o l ue e id c e r i t dap al ri aa s A i l s elel e u. f jd cn da t i ge s u ar u t af uof ao r u l yht ri f t l a i t t s e n sar tode t t ae l i gep st cp rf u n ads s eap npo t r oe o cg ui r s f e mh t r emr gpid t e hl fl Sd Tao OO0O O + + ~
dra dra e r et o et o - a l c l c n na na _ o af af i t ca ca f a z %o 0 f
%o 0
e r i r 4 e r 2 e r l u t e f u f u i a c oi l ol i F n a r ef a e fa f o r a u f u f y e h c l ao l ao t i l c vy vy i b n y
- c n
F c n b a o t i l F Pe Pe o r C i g Lu Lu P ~ f a r Cq Cq w o f Hrfe Hefr L o s t o e n5 , e . f o a h h n2 s e e i i t f t f r e n eo n eo c n g s r s r e A
, i i n g. n a e n a e e d a .
ord re ord r i f h e i t c o i t c o n o c r cn cn C o a f uie uie - h e d nh dnh g t n eat eat i H ci a l r o n r o n o o - - pn at at F mon P - I O O L - C H
+
- I}}